diff --git a/.gitignore b/.gitignore index 6f38bdf..ea01d78 100644 --- a/.gitignore +++ b/.gitignore @@ -1,3 +1,2 @@ node_modules -test/dataset/out -.cache/* \ No newline at end of file +public/data/* \ No newline at end of file diff --git a/index.js b/index.js index 3c76e6c..450fbb4 100644 --- a/index.js +++ b/index.js @@ -3,119 +3,51 @@ /* jslint indent: 2 */ 'use strict'; -var teeft = require('rd-teeft'), +/* Module Require */ +var path = require('path'), async = require('async'), - path = require('path'), - fs = require('fs'); + argv = require('minimist')(process.argv.slice(2)), + express = require('express'), + utils = require('tdm-utils'); -var jLouvain = require('./lib/jLouvain.js'); +var app = express(); +var root = argv.root || './public', + httpPort = argv.httpPort || 3000; -var myObject = {}; +app.use('/', express.static(path.resolve(__dirname, root))); -myObject.indexAll = function(directory, output, cb) { - var result = {}; // Regroup keywords by document Id - fs.readdir(directory, function(err, filenames) { - if (err) return cb(err); // I/O Errors - async.each(filenames, function(filename, callback) { - var filePath = path.join(directory, filename); - fs.readFile(filePath, 'utf-8', function(err, res) { - if (err) return callback(err); // I/O Errors - var docId = path.basename(filename, ('.txt')); - result[docId] = teeft.index(res).keywords; - callback(); - }); - }, function(err) { - if (err) return cb(err); // I/O Errors - // write data - fs.writeFile(output || './cache/indexAll.json', JSON.stringify(result), 'utf-8', function(err, res) { - if (err) return cb(err); - return cb(null, result); - }); - }); - }); -}; - -myObject.graphs = {} - -myObject.graphs.docToDoc = function(keywords, options, cb) { - if (!options) options = {}; - var terms = {}, // Each key is a term, his value is the list of documents containing it - documents = Object.keys(keywords), // List of document Ids - result = { - 'nodes': [], - 'links': [] - }, - edges = [], // [{'source': '', 'target': '', 'weight': 0}, ...] - nodes = [], // ['id', ...] - matrix = {}, // Matrix of "doc-doc" links (sparse matrix) - output = options.output || './cache/docToDoc.json', - minLinkValue = options.minLinkValue || 0; - // Construction of terms Object - for (var i = 0; i < documents.length; i++) { - var doc = documents[i]; - for (var j = 0; j < keywords[doc].length; j++) { - var term = keywords[doc][j].term; - if (!terms[term]) terms[term] = []; - terms[term].push(i); - } - } - // Construction of matrix Object - for (var key in terms) { - // Fill it with values - for (var i = 0; i < terms[key].length - 1; i++) { - var idDoc1 = terms[key][i]; - for (var j = i + 1; j < terms[key].length; j++) { - var idDoc2 = terms[key][j], - ids = [idDoc1, idDoc2], - id = { - 'min': Math.min(ids[0], ids[1]), - 'max': Math.max(ids[0], ids[1]) - }; - // Only half of it will be fill! - if (!matrix[id.min + ',' + id.max]) { - matrix[id.min + ',' + id.max] = 0; - } - matrix[id.min + ',' + id.max]++; +app.listen(httpPort, function() { + console.log('listening on port : ' + httpPort); + console.log('root : ' + root); + // utils.corpusManager.prepare('./corpora', './public/data', function(err, res) { + // if (err) { + // console.log(err); + // return err; + // } + // console.log('prepare ok'); + // async.each(res, function(item, callback) { + // var indexFile = path.join(item.out, 'indexAll.json'); + // utils.corpusManager.indexAll(item.in, indexFile, function(err, res) { + // if (err) { + // console.log(err); + // return err; + // } + // console.log('indexation ok'); + var res = require('./public/data/out/corpus/indexAll.json'); + utils.graphs.docToDoc(res, { + 'output': 'public/data/out/corpus/graph.json', + 'minLinkValue': 5 + }, function(err, res) { + if (err) { + console.log(err); + return err; } - } - } - // Construction of matrix of links doc-doc - for (var key in matrix) { - var ids = key.split(','); - if (matrix[key] > minLinkValue) { - edges.push({ - 'source': ids[0], - 'target': ids[1], - 'weight': matrix[key] - }); - result.links.push({ - 'source': ids[0], - 'target': ids[1], - 'value': matrix[key] - }); - } - } - // Construction of Nodes object - for (var i = 0; i < documents.length; i++) { - nodes.push(i); - result.nodes.push({ - 'id': i, - 'value': documents[i], - 'group': 0 + console.log('graph ok'); }); - } - // Create the "community" - var community = jLouvain().nodes(nodes).edges(edges), - res = community(); - // Affect community for each node - for (var key in res) { - result.nodes[key].group = res[key]; - } - // write data - fs.writeFile(output, JSON.stringify(result), 'utf-8', function(err, res) { - if (err) return cb(err); - return cb(null, result); - }); -}; - -module.exports = myObject; \ No newline at end of file + // }); + // }, function(err) { + // console.log(err); + // return err; + // }); + // // }); +}); \ No newline at end of file diff --git a/lib/jLouvain.js b/lib/jLouvain.js deleted file mode 100644 index 1181efe..0000000 --- a/lib/jLouvain.js +++ /dev/null @@ -1,398 +0,0 @@ -/* - Author: Corneliu S. (github.com/upphiminn) - This is a javascript implementation of the Louvain - community detection algorithm (http://arxiv.org/abs/0803.0476) - Based on https://bitbucket.org/taynaud/python-louvain/overview - */ -var jLouvain = function () { - //Constants - var __PASS_MAX = -1; - var __MIN = 0.0000001; - - //Local vars - var original_graph_nodes; - var original_graph_edges; - var original_graph = {}; - var partition_init; - - //Helpers - function make_set(array) { - var set = {}; - array.forEach(function (d, i) { - set[d] = true; - }); - - return Object.keys(set); - } - - function obj_values(obj) { - var vals = []; - for (var key in obj) { - if (obj.hasOwnProperty(key)) { - vals.push(obj[key]); - } - } - - return vals; - } - - function get_degree_for_node(graph, node) { - var neighbours = graph._assoc_mat[node] ? Object.keys(graph._assoc_mat[node]) : []; - var weight = 0; - neighbours.forEach(function (neighbour, i) { - var value = graph._assoc_mat[node][neighbour] || 1; - if (node === neighbour) { - value *= 2; - } - weight += value; - }); - - return weight; - } - - function get_neighbours_of_node(graph, node) { - if (typeof graph._assoc_mat[node] === 'undefined') { - return []; - } - - var neighbours = Object.keys(graph._assoc_mat[node]); - - return neighbours; - } - - - function get_edge_weight(graph, node1, node2) { - return graph._assoc_mat[node1] ? graph._assoc_mat[node1][node2] : undefined; - } - - function get_graph_size(graph) { - var size = 0; - graph.edges.forEach(function (edge) { - size += edge.weight; - }); - - return size; - } - - function add_edge_to_graph(graph, edge) { - update_assoc_mat(graph, edge); - - var edge_index = graph.edges.map(function (d) { - return d.source + '_' + d.target; - }).indexOf(edge.source + '_' + edge.target); - - if (edge_index !== -1) { - graph.edges[edge_index].weight = edge.weight; - } else { - graph.edges.push(edge); - } - } - - function make_assoc_mat(edge_list) { - var mat = {}; - edge_list.forEach(function (edge, i) { - mat[edge.source] = mat[edge.source] || {}; - mat[edge.source][edge.target] = edge.weight; - mat[edge.target] = mat[edge.target] || {}; - mat[edge.target][edge.source] = edge.weight; - }); - - return mat; - } - - function update_assoc_mat(graph, edge) { - graph._assoc_mat[edge.source] = graph._assoc_mat[edge.source] || {}; - graph._assoc_mat[edge.source][edge.target] = edge.weight; - graph._assoc_mat[edge.target] = graph._assoc_mat[edge.target] || {}; - graph._assoc_mat[edge.target][edge.source] = edge.weight; - } - - function clone(obj) { - if (obj === null || typeof(obj) !== 'object') - return obj; - - var temp = obj.constructor(); - - for (var key in obj) { - temp[key] = clone(obj[key]); - } - - return temp; - } - - //Core-Algorithm Related - function init_status(graph, status, part) { - status['nodes_to_com'] = {}; - status['total_weight'] = 0; - status['internals'] = {}; - status['degrees'] = {}; - status['gdegrees'] = {}; - status['loops'] = {}; - status['total_weight'] = get_graph_size(graph); - - if (typeof part === 'undefined') { - graph.nodes.forEach(function (node, i) { - status.nodes_to_com[node] = i; - var deg = get_degree_for_node(graph, node); - - if (deg < 0) - throw 'Bad graph type, use positive weights!'; - - status.degrees[i] = deg; - status.gdegrees[node] = deg; - status.loops[node] = get_edge_weight(graph, node, node) || 0; - status.internals[i] = status.loops[node]; - }); - } else { - graph.nodes.forEach(function (node, i) { - var com = part[node]; - status.nodes_to_com[node] = com; - var deg = get_degree_for_node(graph, node); - status.degrees[com] = (status.degrees[com] || 0) + deg; - status.gdegrees[node] = deg; - var inc = 0.0; - - var neighbours = get_neighbours_of_node(graph, node); - neighbours.forEach(function (neighbour, i) { - var weight = graph._assoc_mat[node][neighbour]; - - if (weight <= 0) { - throw "Bad graph type, use positive weights"; - } - - if (part[neighbour] === com) { - if (neighbour === node) { - inc += weight; - } else { - inc += weight / 2.0; - } - } - }); - status.internals[com] = (status.internals[com] || 0) + inc; - }); - } - } - - function __modularity(status) { - var links = status.total_weight; - var result = 0.0; - var communities = make_set(obj_values(status.nodes_to_com)); - - communities.forEach(function (com, i) { - var in_degree = status.internals[com] || 0; - var degree = status.degrees[com] || 0; - if (links > 0) { - result = result + in_degree / links - Math.pow((degree / (2.0 * links)), 2); - } - }); - - return result; - } - - function __neighcom(node, graph, status) { - // compute the communities in the neighb. of the node, with the graph given by - // node_to_com - var weights = {}; - var neighboorhood = get_neighbours_of_node(graph, node);//make iterable; - - neighboorhood.forEach(function (neighbour, i) { - if (neighbour !== node) { - var weight = graph._assoc_mat[node][neighbour] || 1; - var neighbourcom = status.nodes_to_com[neighbour]; - weights[neighbourcom] = (weights[neighbourcom] || 0) + weight; - } - }); - - return weights; - } - - function __insert(node, com, weight, status) { - //insert node into com and modify status - status.nodes_to_com[node] = +com; - status.degrees[com] = (status.degrees[com] || 0) + (status.gdegrees[node] || 0); - status.internals[com] = (status.internals[com] || 0) + weight + (status.loops[node] || 0); - } - - function __remove(node, com, weight, status) { - //remove node from com and modify status - status.degrees[com] = ((status.degrees[com] || 0) - (status.gdegrees[node] || 0)); - status.internals[com] = ((status.internals[com] || 0) - weight - (status.loops[node] || 0)); - status.nodes_to_com[node] = -1; - } - - function __renumber(dict) { - var count = 0; - var ret = clone(dict); //deep copy :) - var new_values = {}; - var dict_keys = Object.keys(dict); - dict_keys.forEach(function (key) { - var value = dict[key]; - var new_value = typeof new_values[value] === 'undefined' ? -1 : new_values[value]; - if (new_value === -1) { - new_values[value] = count; - new_value = count; - count = count + 1; - } - ret[key] = new_value; - }); - - return ret; - } - - function __one_level(graph, status) { - //Compute one level of the Communities Dendogram. - var modif = true; - var nb_pass_done = 0; - var cur_mod = __modularity(status); - var new_mod = cur_mod; - - while (modif && nb_pass_done !== __PASS_MAX) { - cur_mod = new_mod; - modif = false; - nb_pass_done += 1 - - graph.nodes.forEach(function (node, i) { - var com_node = status.nodes_to_com[node]; - var degc_totw = (status.gdegrees[node] || 0) / (status.total_weight * 2.0); - var neigh_communities = __neighcom(node, graph, status); - __remove(node, com_node, (neigh_communities[com_node] || 0.0), status); - var best_com = com_node; - var best_increase = 0; - var neigh_communities_entries = Object.keys(neigh_communities);//make iterable; - - neigh_communities_entries.forEach(function (com, i) { - var incr = neigh_communities[com] - (status.degrees[com] || 0.0) * degc_totw; - if (incr > best_increase) { - best_increase = incr; - best_com = com; - } - }); - - __insert(node, best_com, neigh_communities[best_com] || 0, status); - - if (best_com !== com_node) { - modif = true; - } - }); - new_mod = __modularity(status); - if (new_mod - cur_mod < __MIN) { - break; - } - } - } - - function induced_graph(partition, graph) { - var ret = {nodes: [], edges: [], _assoc_mat: {}}; - var w_prec, weight; - //add nodes from partition values - var partition_values = obj_values(partition); - ret.nodes = ret.nodes.concat(make_set(partition_values)); //make set - graph.edges.forEach(function (edge, i) { - weight = edge.weight || 1; - var com1 = partition[edge.source]; - var com2 = partition[edge.target]; - w_prec = (get_edge_weight(ret, com1, com2) || 0); - var new_weight = (w_prec + weight); - add_edge_to_graph(ret, {'source': com1, 'target': com2, 'weight': new_weight}); - }); - - return ret; - } - - function partition_at_level(dendogram, level) { - var partition = clone(dendogram[0]); - for (var i = 1; i < level + 1; i++) { - Object.keys(partition).forEach(function (key, j) { - var node = key; - var com = partition[key]; - partition[node] = dendogram[i][com]; - }); - } - - return partition; - } - - - function generate_dendogram(graph, part_init) { - if (graph.edges.length === 0) { - var part = {}; - graph.nodes.forEach(function (node, i) { - part[node] = node; - }); - return part; - } - var status = {}; - - init_status(original_graph, status, part_init); - var mod = __modularity(status); - var status_list = []; - __one_level(original_graph, status); - var new_mod = __modularity(status); - var partition = __renumber(status.nodes_to_com); - status_list.push(partition); - mod = new_mod; - var current_graph = induced_graph(partition, original_graph); - init_status(current_graph, status); - - while (true) { - __one_level(current_graph, status); - new_mod = __modularity(status); - if (new_mod - mod < __MIN) { - break; - } - - partition = __renumber(status.nodes_to_com); - status_list.push(partition); - - mod = new_mod; - current_graph = induced_graph(partition, current_graph); - init_status(current_graph, status); - } - - return status_list; - } - - var core = function () { - var status = {}; - var dendogram = generate_dendogram(original_graph, partition_init); - - return partition_at_level(dendogram, dendogram.length - 1); - }; - - core.nodes = function (nds) { - if (arguments.length > 0) { - original_graph_nodes = nds; - } - - return core; - }; - - core.edges = function (edgs) { - if (typeof original_graph_nodes === 'undefined') - throw 'Please provide the graph nodes first!'; - - if (arguments.length > 0) { - original_graph_edges = edgs; - var assoc_mat = make_assoc_mat(edgs); - original_graph = { - 'nodes': original_graph_nodes, - 'edges': original_graph_edges, - '_assoc_mat': assoc_mat - }; - } - - return core; - - }; - - core.partition_init = function (prttn) { - if (arguments.length > 0) { - partition_init = prttn; - } - return core; - }; - - return core; -} - -module.exports = jLouvain; \ No newline at end of file diff --git a/package.json b/package.json index 8273b24..72320d1 100644 --- a/package.json +++ b/package.json @@ -8,14 +8,15 @@ }, "repository": { "type": "git", - "url": "https://git.istex.fr/git/kieffer/tdm-utils.git" + "url": "https://git.istex.fr/git/kieffer/ez-indexation.git" }, "author": "istex", "license": "ISC", "dependencies": { "async": "^2.1.4", + "express": "^4.14.1", + "minimist": "^1.2.0", "mocha": "^3.2.0", - "rd-tu": "git+ssh://vsistex.intra.inist.fr:22222/istex/rd-tu.git", - "rd-teeft": "git+ssh://vsistex.intra.inist.fr:22222/istex/rd-teeft.git" + "tdm-utils": "git+ssh://vsistex.intra.inist.fr:22222/kieffer/tdm-utils.git" } -} +} \ No newline at end of file diff --git a/public/css/force-directed-graph.css b/public/css/force-directed-graph.css new file mode 100644 index 0000000..fa00740 --- /dev/null +++ b/public/css/force-directed-graph.css @@ -0,0 +1,9 @@ +.links line { + stroke: #999; + stroke-opacity: 0.6; +} + +.nodes circle { + stroke: #fff; + stroke-width: 1.5px; +} \ No newline at end of file diff --git a/public/css/style.css b/public/css/style.css new file mode 100644 index 0000000..4cbe397 --- /dev/null +++ b/public/css/style.css @@ -0,0 +1,3 @@ +svg { + border: 1px solid black; +} \ No newline at end of file diff --git a/public/index.html b/public/index.html new file mode 100644 index 0000000..8337656 --- /dev/null +++ b/public/index.html @@ -0,0 +1,14 @@ + + +
+ 10% P < 0.0005), again providing evidence for our model that sees social support as buffering the ill effects of physical disability on distress level. Mastery contributes further to the variance in psychological distress, but at a marginally significant level (2.8% P = 0.05). Taken together, physical disability, social support, and mastery accounted for 42% of the variance in psychological distress.Our third outcome variable, depression, was surprisingly unrelated to physical disability. Nevertheless, it was strongly related to mastery, which accounted for over 28% of the variance in depression. Also, social support and religious commitment added 10% and 5%, respectively, to the variance in depressive symptoms, suggesting that these variables along with a sense of mastery, may help buffer the adverse effects of stressors on mood state. Again, these three variables together accounted for almost 44% of the variance in depression.DISCUSSIONThese data lend empirical support to our proposed model, providing evidence for the importance of internal resources (mastery), external resources (social support), and coping (religious commitment) in buffering the negative effects of physical disability on psychological adjustment in later life. The results are also consistent with a substantial body of research that shows a relationship between physical disability and emotional distress,36–38 and they demonstrate important roles for social support,7–9 mastery,5,39 and coping behavior3,4,9 in mediating the relationship between psychosocial stress and adjustment.In this analysis, the stressor and buffering factors contributed measurably, and differentially, to mental outcomes. Life satisfaction, a broad measure, was only weakly associated with physical disability, but was strongly related to mastery and social support. Religious commitment was also a weak contributor to life satisfaction. Psychological distress, on the other hand, is a more specific mental health outcome than life satisfaction. As measured in this study, it includes anxiety and somatization. It had the strongest relationship with physical disability, our biopsychosocial stressor; both social support and mastery, on the other hand, were inversely related to psychological distress, as predicted by our model.The explanatory model for depression, the most specific mental health construct, was different from either life satisfaction or psychological distress. Like life satisfaction, depression was weakly correlated with physical disability, but strongly and inversely correlated with mastery and, to a lesser degree, social support and religious commitment. Clinically, depression is marked by a sense of helplessness, the opposite of mastery. Social supports are known to be perceived as unavailable or unreliable to depressed persons, although it is unclear whether the absence of social support makes one vulnerable to depression, or whether depression causes one to perceive existing social supports as unavailable, unreliable, or undeserved.7,27,31,33The inverse relationship between religious commitment and depression, independent of social support or mastery, is noteworthy. The early (and even more recent) psychiatric literature has portrayed religious commitment as a symptom of neurosis40 or evidence of psychological disturbance.41 This study adds weight to recent findings of an inverse relationship between religiousness and depression.16,42 Of course, given the cross-sectional nature of our data, as with social support, we cannot say whether greater religious commitment causes less depression or whether depression causes less religiousness. There are both longitudinal data16 and interventional data,43 however, that support the former idea.We recognize that assigning religious commitment as primarily a coping variable is somewhat arbitrary, because it may to some extent serve as an internal and external resource. These constructs, then, may not be entirely independent of each other. First, religious commitment may act as an internal resource by enhancing mastery and making persons less psychologically vulnerable to stressors. Strong religious commitment provides a belief system that may increase mastery by giving the individual a sense that he or she has more control over their situation; that is, if he or she believes in an all-powerful God who cares about him or her and will respond to prayers, then the person will feel that he or she is not powerless, but instead can take control over a situation by influencing God to act on his or her behalf. Second, religious commitment may enhance social support (external resource) by encouraging involvement in the religious community, which may broaden one's social support network (particularly among elderly persons, whose major voluntary social participation outside family is the church); furthermore, by encouraging people to support and care for others, religious beliefs (the “social gospel”) may help form a more supportive community, which helps to reduce the adverse effects of stress.Nevertheless, in our model, we see religious commitment as primarily a coping strategy because it influences how persons perceive and respond to stressors (i.e., cope), independent of internal and external resources. Indeed, our results would suggest that although religious commitment is positively correlated with both mastery (internal resource) and social support (external resource), there is still an independent effect for religion on life satisfaction and depression. Religious belief may facilitate coping directly by altering the perception of the stressor so that consequences are not seen as disastrous. For example, poor physical health may indicate to the older person that he or she is getting closer to death; if the elder believes that there is an afterlife where he or she will be reunited with family and life will improve, then the prospect of dying is not as terror-filled. Finally, religious commitment may facilitate adaptation by providing both cognitive (trusting God, turning problems over to God) and behavioral (praying to God, singing, participating in communal worship) tools for coping with distress.There are many factors accounting for the unexplained variance in our dependent variables (life satisfaction, general distress, depression). One is measurement error. For example, psychosocial variables, like religious commitment, social support, mastery, and the like, cannot be measured with the same degree of accuracy as physiological variables, like blood pressure or heart rate. Consequently, in mental health or social science research, the explained variance in outcome is seldom greater than 50%. Second, we did not measure other important internal resources (hereditary effects or genetic vulnerability) and external resources (health problems not affecting functional status) that could have contributed substantially to our mental health outcomes. Third, there may be factors that we are not yet aware of that affect mental health (for example, until recently religion was not thought to contribute to mental health in any meaningful way).Our study group is unique in several ways. Not only does the sample represent a narrowly defined group demographically and vocationally, but also, as the data suggest, there are other distinctions. Compared with several other studies using general, community-based populations,22–24 our sample of women appears more satisfied with life and less likely to report psychological distress. This observation is all the more striking given the higher prevalence of certain mental disorders, especially depression, in women.44,45 Although only a handful of other studies have looked at mental health outcomes in Catholic religious groups, virtually all have reported similar differences between lay and Catholic religious.46–48Moore46 and Kelly47 both concluded that mental illness was less common among women religious than in the general population. Kvale and colleagues48 studied morale among retired elderly nuns, and found that on nearly all measures, the sisters had higher morale than their lay counterparts. Our study confirms these findings. Despite the lack of spouses and children, and despite being retired from active employment, these women experienced better emotional well-being than women in the general population. However the mechanism explaining the better mental health functioning of this group remains unclear. The healthier outcome could be related to some shared characteristic, such as the overall high degree of religious commitment, the predictable and reliable financial situation, or the more ready access to health care and monitoring.Perhaps women religious are not as distinctive a group as one might assume, at least in terms of the impact of stress on mental well-being. Magee,49 studying 150 retired women religious, reported that the conditions contributing to the life satisfaction of older adults in general also pertained to his study population. Furthermore, many stressors encountered by Catholic religious professionals are similar to the stressors facing many other older Americans. These stressors include the potential for alienation from society and the personal problems they must handle to deal with the changing world. For example, Kennedy and colleagues50 reported that church reforms introduced role confusion, change in occupational focus, and ideological questioning among a group of Catholic priests they studied. Stressors such as these may contribute further to poorer mental health outcomes through the same mechanisms we described previously: loss of mastery and diminished social support.Despite the homogeneity of this sample, it did generate scores on religious commitment that were normally distributed, although still clustered at the high end of the scale (mean score, 27.3 ± 5.0 out of a possible 35; range: 10–35). Perhaps development of a more complex model of religious coping would be a useful approach in future research.17This is, of course, a unique sample of unmarried, celibate, elderly women who have devoted their lives to their religious vocation. Thus, in these and perhaps other unmeasurable ways, this sample is different from the general population of older adults in America. How these differences may have biased results is difficult to predict. Many women in their later years, however, do find themselves unmarried and celibate, living near or with each other, and often congregating in church. Furthermore, the fact that we found similar relationships in this sample to those of other researchers working with nongeriatric and elderly populations helps to dispel some of these concerns. Another potential bias is the lack of information about nonresponders. If nonresponders were selectively impaired in some way that made them refuse the study, such as having poor cognition, mental illness, or physical debility, our study sample may have been a self-selected healthier group. The relatively high response rate, however, adds confidence to the study findings.SUMMARY AND CONCLUSIONSWe have developed and tested a multidimensional stress and coping model in a selected and homogenous population of elderly Catholic women religious. Conceptualizing physical disability as a biopsychosocial stressor and life satisfaction, psychological distress, and depression as mental health outcomes, we found that internal (mastery), external (social support), and coping resources (religious commitment) predicted better mental health in this population. Our internal resource (sense of mastery) was the strongest single predictor of variance in two separate mental health outcomes, life satisfaction and depression, accounting for 27% for life satisfaction and 29% for depression. Higher level of perceived social support (external resource) was also significantly related to all three mental health outcomes. Finally, religious commitment, often unexamined in models of coping and adaptation, emerged as a positive predictor of life satisfaction and lower levels of depression. These findings further support the validity of a multidimensional coping model for adaptation to biopsychosocial stressors in late life.What implications do these findings have for clinicians caring for elderly patients? First, as in other studies, we found a link between disability and emotional distress. This finding should support close attention to the emotional adjustment of older persons experiencing severe disability from health problems. Optimizing physical functioning and the capacity for independent living must be among the first goals of the treating physician. Second, in counseling disabled elders with emotional disorder or those at risk for it, clinicians should pay particular attention to the patient's sense of helplessness or powerlessness about his or her situation. Discussing with patients the different aspects of their life that they can and cannot control may help to resolve some of these feelings. Third, it is important to encourage involvement in social activities that broaden the patient's support network, particularly activities with age-matched peers who may form a support group for them. Finally, clinicians should be encouraged to inquire about and support healthy religious beliefs and activities that may help older persons adapt to chronic illness, disability, and losses over which they have little or no control.51References1DGBlazerKJordanEpidemiology of psychiatric disorders and cognitive problems in the elderlyGLKlermanMMWeissmanPSAppelbaumSocial, Epidemiologic, and Legal Psychiatry1986JB LippincottNew York2612722HGKoenigMSmileyJAPGonzalesReligion, Health, and Aging: A Review and Theoretical Integration1988Westport, CTGreenwood Press3RLazarusSFolkmanStress, Appraisal, and Coping1984SpringerNew York4SFolkmanRSLazarusCoping as a mediator of emotionJ Pers Soc Psychol5419884664755SFolkmanRSLazarusRJGruenAppraisal, coping, health status, and psychological symptomsJ Pers Soc Psychol5019865715796SFolkmanRSLazarusSPimleyAge differences in stress and coping processesPsychol Aging219871711847DWRussellCECutronaSocial support, stress, and depressive symptoms among the elderly: test of a process modelPsychol Aging619911902018PPVitalianoJRussoHMYoungPredictors of burden in spouse caregivers of individuals with Alzheimer's diseasePsychol Aging619913924029WEHaleyEGLevineSLBrownStress, appraisal, coping, and social support as predictors of adaptational outcome among dementia caregiversPsychol Aging2198732333010JSLevinReligious factors in aging, adjustment, and health: a theoretical overviewJournal of Religion and Aging4198813314611DGBlazerEPalmoreReligion and aging in a longitudinal panelGerontologist161976828512KSMarkidesJSLevinLARayReligion, aging and life satisfaction: an 8-year, three-wave longitudinal studyGerontologist27198766066513HGKoenigDOMobergJNKvaleReligious activities and attitudes of older adults in a geriatric assessment clinicJ Am Geriatr Soc36198836237414HGKoenigResearch on religion and mental health in later life: a review and commentaryJ Geriatr Psychiatry231990235315HGKoenigLKGeorgeICSieglerThe use of religion and other emotion-regulating coping strategies among older adultsGerontologist28198030331016HGKoenigHJCohenDGBlazerReligious coping and depression among elderly, hospitalized medically ill menAm J Psychiatry14919921693170017HGKoenigJNKvaleCFerrelReligion and well-being in later lifeGerontologist281988182818KASherrillDBLarsonMGreenwoldIs religion taboo in gerontology? A systematic review of research on religion in three major gerontology journals, 1985–1991American Journal of Geriatric Psychiatry1199310911719MPLawtonMMossMFulcomerA research- and service-oriented multi-level assessment instrumentJ Gerontology371982919920SGranickPsychologic assessment technology for geriatric practiceJ Am Geriatr Soc31198372874221BLNeugartenRJHavighurstSSTobinThe measurement of life satisfactionJ Gerontology15196191592322DGoldbergManual of the General Health Questionnaire1978NFER Publishing CompanyWindsor, UK23BWViewigJLHedlundThe general health questionnaire (GHQ): a comprehensive reviewJournal of Operational Psychiatry141983748124DPGoldbergVFHillierA scaled version of the General Health QuestionnairePsychol Med9197913914525LIPearlinCSchoolerThe structure of copingJ Health Soc Behav19197822126LIPearlinMALiebermanEGMenaghamThe stress processJ Health Soc Behav22198133735627CPetersonMEPSeligmanCausal explanations as a risk factor for depression: theory and evidencePsychol Rev91198434737428LFBerkmanSLSymeSocial networks, host resistance and mortality: a 9-year follow-up study of Alameda County residentsAm J Epidemiol109197918620429SCohenLSSymeSocial Support and Health1985Academic PressOrlando, FL30JWRoweRLKahnHuman aging: usual and successfulScience237198714314931LKGeorgeSocial and economic factorsDGBlazerEWBusseGeriatric Psychiatry1989American Psychiatric PressWashington, DC20323432SCohenTAWillsStress, social support, and the buffering hypothesisPsychol Bull98198531035733CCutronaDRussellJRoseSocial support and adaptation to stress by the elderlyPsychol Aging11986475434SCohenRMermelsteinTKamarckMeasuring the functional components of social supportISarasonBSarasonSocial Support: Theory, Research, and Applications1985Martinus NijhoffThe Hague, Netherlands739435JHKauffmanSocial correlates of spiritual maturity among North American MennonitesDOMobergSpiritual Well-being: Sociological Perspectives1979University Press of AmericaWashington, DC23725436BJGurlandDEWilderCBerkmanDepression and disability in the elderly: reciprocal relations and changes with ageInternational Journal of Geriatric Psychiatry3198816317937GMWilliamsonRSchulzPain, activity restriction, and symptoms of depression among community-residing elderly adultsJ Gerontology471992P367P37238MVon KorffJOrmelWKatonDisability and depression among high utilizers of health careArch Gen Psychiatry4919929110039SFolkmanPersonal control and stress and coping processes: a theoretical analysisJ Pers Soc Psychol46198483985240SFreudThe Future of an Illusion (1927)JStracheyThe Standard Edition of the Complete Psychological Works of Sigmund Freud1962Hogarth PressLondon, UK434441AEllisPsychotherapy and atheistic values: a response to A.E. Bergins's “Psychotherapy and religious values.”J Consult Clin Psychol48198064264542PPressmanJSLyonsDBLarsonReligious belief, depression, and ambulation status in elderly women with broken hipsAm J Psychiatry147199075876043LRPropstROstromPWatkinsComparative efficacy of religious and nonreligious cognitive-behavioral therapy for the treatment of clinical depression in religious individualsJ Consult Clin Psychol6019929410344BRorsmanAGrasbeckOHagnellA prospective study of first-incidence depression: The Lundby Study, 1957–1972Br J Psychiatry156199033634245MMWeissmanPJLeafGLTischierAffective disorders in five United States communitiesPsychol Med18198814115346TVMooreInsanity in priests and religious (letter)American Ecclesiastical Review95198648547MWKellyThe incidence of hospitalized mental illness among religious sisters in the U.S. (letter)Am J Psychiatry11519587248JNKvaleHGKoenigCFerrelLife satisfaction of the aging woman religiousJournal of Religion and Aging51989597049JJMageeA model for the congregational assessment of the life-satisfaction of retired sistersReview for ReligiousNov-Dec 198492392750ECKennedyVJHecklerFJKoblerClinical assessment of a profession: Roman Catholic clergymenJ Clin Psychol33197712012851Koenig HG, Larson DG, Matthews D: Religion and psychotherapy with older adults. Journal of Geriatric Psychiatry 29:155–184
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0ADF61F76363BFE62FEEBA003BFD2581BC290AEA.txt b/test/dataset/in/resources/corpus/Clean_0ADF61F76363BFE62FEEBA003BFD2581BC290AEA.txt
deleted file mode 100644
index d3fcc6b..0000000
--- a/test/dataset/in/resources/corpus/Clean_0ADF61F76363BFE62FEEBA003BFD2581BC290AEA.txt
+++ /dev/null
@@ -1 +0,0 @@
- geronb J Gerontol B Psychol Sci Soc Scigeronb The Journals of Gerontology Series B: Psychological Sciences and Social Sciences J Gerontol B Psychol Sci Soc Sci 1079-5014 1758-5368 Oxford University Press 22010.1093/geronb/59.4.S220 Journal of Gerontology: Social Sciences Is There Equity in the Home Health Care Market? Understanding Racial Patterns in the Use of Formal Home Health Care White-Means Shelley I. Rubin Rose M. Department of Economics, Fogelman College of Business and Economics, University of Memphis, Tennessee. Address correspondence to Dr. Shelley I. White-Means, University of Tennessee Health Science Center, 847 Monroe Avenue, Room 205N, Memphis, TN 38113. E-mail: kingram.utmem.edu 7 2004 59 4 S220 S229 17 2 2004 11 4 2003 The Gerontological Society of America 2004 Objective. This article explores whether the formal home health care (HHC) market is equitable or manifests unexplained racial disparities in use. Methods. The database is the 1994 National Long Term Care Survey. We estimate logit regression models with a race dummy variable, race interaction terms, and stratification by race. We apply the Oaxaca decomposition technique to quantify whether the observed racial gap in formal HHC use is explained by racial differences in predisposing, enabling, need, and environmental characteristics. Results. We find numerous unique racial patterns in HHC use. Blacks with diabetes and low income have higher probabilities of HHC use than their White counterparts. Black older persons have a 25% higher chance of using HHC than Whites. Our Oaxaca analysis indicates that racial differences in predisposing, enabling, need, and environmental characteristics account for the racial gap in use of HHC. Discussion. We find that the HHC market is equitable, enhancing availability, aceptability, and accessibility of care for older Black persons. Thus, the racial differences that we find are not racial disparities. hwp-legacy-fpage S220 hwp-legacy-dochead RESEARCH ARTICLE Distinct racial patterns in older persons' use of both institutional and community-based health care services are well documented (Wallace, Levy-Storms, Kington, & Andersen, 1998). In this article, we explore whether formal home health care (HHC) use among older persons varies by race, and if so, then we seek to explain why this differential use occurs. We assess the use of HHC as an integral part of community health care rather than as one component of the spectrum of long-term care, as is done by Wallace and colleagues (1995, 1998), and we delineate the use of HHC in the context of the broader literature of racial differences in community care to highlight that this market differs from other alternatives. Thus, discussion of racial patterns of use for alternative types of community health care provides important contextual background for our focus on HHC. Owing partly to limited financial access to nursing homes and cultural preferences (Falcone & Broyles, 1994), older Blacks are less likely to use nursing homes and more likely to live in the community with chronic conditions and disabilities that exceed those of older Whites (Wallace et al., 1998; White-Means, 2000). However, lower use of skilled nursing facilities by Blacks is not counterbalanced by higher use of HHC services (Cagney & Agree, 1999). Further, noninstitutionalized older disabled Blacks are less likely to use other community-based services such physician services and prescription medications (White-Means, 2000). Differential use of in-home assistive devices is also found, with Whites more likely to use home modification devices and Blacks more likely to use portable devices (Rubin & White-Means, 2001). What explains these distinct racial patterns in health services use? Racial differences in physician services use and prescription medication are not fully explained by racial differences in average income or socioeconomic or health characteristics (White-Means, 2000). Researchers (Wallace, 1990; Wallace et al., 1998; White-Means, 2000) note that racial differences are explained by three factors: availability, accessibility, and acceptability. Problems of availability arise from the lack of close geographic proximity to services and/or segregation in the use of services. Lack of knowledge about available services and limited referrals to auxiliary services lead to gaps in accessibility. Patient dissatisfaction, culturally insensitive care, and unfavorable provider–patient interactions lead to problems of acceptability. The Institute of Medicine (2002) distinguishes between racial differences and disparities. It reports that some racial differences in health care use can be expected and are acceptable, including differences due to clinical need, patient preferences, and the appropriateness of particular interventions. But if differences are due to characteristics of the health care system, provider discrimination, or incorrect or deficient patient information, then these differences are considered racial disparities in care. Thus, differences in use, reflecting such factors as stereotyping, biases, and uncertainty regarding treatment efficacy by race, limited supplies of racial/ethnic minority health care providers, and poor patient–provider communication and trust, are seen as disparities. We seek to determine whether the racial differences in HHC use are acceptable differences (and thus equitable) or actually reflect disparities, as defined by the Institute of Medicine. If the differences found are disparities, then we would expect to find divergences in availability, accessibility, and acceptability in HHC use. So, disparities are seen as differentials that are not explained by either socioeconomic or health characteristics. We define formal HHC as market-purchased services provided in the patient's home, including both Medicare-financed care associated with subacute care and community-based care that meets needs associated with activities of daily living (ADLs) and instrumental activities of daily living (IADLs). Our comparison focuses on use versus nonuse of community-based HHC and whether the comparative use is equitable by race. The dichotomous dependent variable indicates whether an older Black or White person who lives in the community uses HHC or not. Nonusers of HHC may use informal care support or various alternate formal (purchased or market-supplied) health care services or no care, as distinguished by Wallace et al. (1998), who also explore nursing home users. We hypothesize that HHC is an equitable market. If it achieves the three criteria of availability, accessibility, and acceptability, then this would suggest that HHC represents an equitable market. HHC is more available to the home bound than other community-based care because it does not require transportation (Rubin, White-Means, & Feagin, 2003). Accessibility is broadened by Medicare and by the relatively low price of HHC. Medicare is the largest single HHC payer (about 40%; National Association of Home Care [NAHC], 2000), covering HHC under Part A with no out-of-pocket charges and under Part B with a $100 deductible. Other federal programs and Medicaid also pay for some HHC for the poor or medically indigent (NAHC, 2000; Shi & Singh, 1998). Also, the culturally diverse workforce of HHC (35% are Black and 10% Hispanic) may be more acceptable to diverse groups of older persons than nonhome community-based care and may be perceived to be more culturally sensitive (Stone & Wiener, 2001). As this low-paid workforce is economically disadvantaged, cultural and class barriers tend to be minimized. Additionally, because of the relatively low start-up cost of a home health agency (approximately $150,000), many agencies are owned by ethnic minorities and are more likely to serve communities where a higher proportion of ethnic minorities reside (Brown, 1998). Data from national surveys tend to support the hypothesis of HHC as an equitable market. Indeed, Blacks are more likely to use HHC than Whites (Mui & Burnette, 1994; Wallace et al., 1998). In contrast, regression analysis provides equivocal support for the role of race in increasing home health use. Freedman (1999) reports that non-Whites are significantly more likely to use more HHC services and also to use them for longer terms. This higher intensity of use is associated with the greater likelihood that Blacks postpone use and remain in the community longer with chronic conditions (Cagney & Agree, 1999). Although the coefficients are insignificant, Picone and Wilson (1999) find higher Medicare home health agency use rates for Blacks, but they conclude that socioeconomic factors are not strong determinants compared to health status. Similarly, Miller and colleagues (1996), who do not find racial differences, find that overall health and functional status, as well as Medicaid use, are primary predictors of service use. In contrast, Wallace and colleagues (1998) note that African Americans are less likely than non-Latino Whites to use HHC, with HHC use more likely for those of advanced age, women and single women, and those with more formal education, higher ADLs, stroke or heart conditions, low income, and Medicaid. Research Questions This study makes important contributions to a growing literature on racial disparities in health care. The broad research objective is to discover whether HHC provides an equitable health care market for Blacks. We determine empirically based answers to two specific questions: (a) Are there distinct racial patterns in the use and determinants of use of HHC? (b) What is the role of racial differences in need, predisposing conditions, enabling factors, and environmental factors in explaining observed racial differences in HHC use? Methods Data The database is the 1994 National Long-Term Care Survey (NLTCS), a repeated panel survey designed to estimate chronic disability status, functional problems, and institutionalization rates of older persons and their use of community-based in-home long-term care. The NLTCS provides nationally representative data on the prevalence and patterns of functional limitations, medical conditions, and recent medical problems for Medicare-enrolled Americans aged 65 and older. It also includes data on health services use, out-of-pocket expenditures, financial resources, and sociodemographic characteristics of older persons and their families. The 1994 NLTCS sample is the fourth panel and includes living respondents from previous panels, despite constancy or changes in their disability level (Manton, Corder, & Stallard, 1997). The 1994 dataset is supplemented with additional persons meeting the impaired status screening criteria from those who were previously screened but not interviewed because at that time they did not meet the impaired status criteria or reached age 65 after the 1989 cutoff. The addition of an oversample of persons aged 95 and older and a “healthy” sample not meeting the “impaired” status criteria is unique to the 1994 NLTCS to enhance the national representativeness of the data. Detailed questions are asked if persons meet screening criteria for chronic ADL and IADL disability. However, many carried over were previously chronically disabled but in 1994 had no ADL and/or IADL disabilities so that the sample shows many people with no current ADLs or IADLs (Liu, Manton, & Aragon, 2000). In contrast to some analyses (e.g., Liu et al., 2000), we include going outside as an ADL, because it is grouped as an ADL in the NLTCS screener definitions. The NLTCS screener includes limitations in walking around inside and also going outside as ADLs; this ADL definition may result in relatively high prevalence of ADLs compared with IADLs. The NLTCS includes 5,089 noninstitutionalized older persons in the detailed community interviews, of whom 4,740 are non-Hispanic Whites and non-Hispanic Blacks; our study sample includes 3,448 non-Hispanic White and non-Hispanic Black complete reporters (participants who provided data for all variables). Similar to other survey research data, it has missing data on family income (25%) and education (5%). Typically, persons at upper and lower incomes are least likely to report income or education or both, resulting in patterns of data loss that are nonrandom or systematic. Thus, the factors explaining the likelihood of missing data are unrelated to the current study's exploration of the use of home care. With a sample of complete reporters, we lose some statistical power, but this does not lead to inaccurate parameter estimates. First, statistical tests (Kromrey & Hines, 1994; Switzer, Roth, & Switzer, 1998) indicate that in cases where no more than 30% of the data are missing (as is true for our sample), accurate parameter estimates are generated. Second, standard estimation procedures restricted to the available observations yield valid inferences when the process that generates the missing data is independent of the observed and unobserved data values (Rotnitzky & Wypij, 1994). As the data indicate that it is equally likely that Blacks and Whites have missing income data (27.19% vs 24.3%) and that home health users and nonusers have missing income data (26.81% vs 24.11%), we have further confirmation of lack of bias in the parameter estimates using the available observations. Thus, our analysis uses a sample of 3,448 persons, with 3,136 White non-Hispanic and 312 Black non-Hispanic participants, as described in Table 1. Analytic Design To examine our first research question, we use three regression strategies: one with a race dummy variable, a second with race interaction terms, and a third stratified by race. For the second research question, we apply the Oaxaca decomposition technique (Oaxaca, 1973), quantifying the percentage of the observed racial gap in HHC use explained by racial differences. This approach can contribute to our understanding because it questions whether an identified racial difference in home health use is what we should observe. As the health status of community-residing older Blacks is significantly lower than that of Whites, the Oaxaca approach can help to answer whether we should similarly find differences in home health use, with significantly higher use by Blacks, and if not, then why not? We first estimate logistic regression models of HHC use, including a race dummy variable (Model I) and also race interaction terms (Model II). Next, we estimate logistic regressions of HHC use, stratified by race. The measure of HHC use is a dichtomous variable that equals 1 if paid home services are provided by a helping organization (i.e., a home health agency or other business provider) and 0 otherwise. The NLTCS allows respondents to define a helping organization as any organization that provides in-home assistance, by describing their relationship to each in-home caregiver (or helper). If any caregiver was from a helping organization, the respondent was coded as having HHC. Our choice of explanatory variables and conceptual framework are guided by the Andersen model (1995) and previous research on HHC use. The Anderson model suggests that medical use patterns are influenced by (a) predisposing conditions, (b) enabling factors, (c) environmental factors, and (d) need. Our measures of predisposing conditions are gender, age, marital status, education, and three quantifiable behavioral characteristics (whether the respondent currently smokes, exercises or plays sports regularly, and/or drinks alcoholic beverages one to three times a month or less frequently; Cutler & Sheiner, 1994; Newman, Struyk, Wright, & Rice, 1990). Our measures of enabling factors are income, income squared, health insurance status (Medicare only, Medicare plus supplements, or Medicare and Medicaid dual eligibility), number of adults living in the home, and use of home-delivered meals. We utilize income squared because previous research (White-Means, 1997) documents a nonmonotonic relationship between home health use and income. Home-delivered meals, part of community-based nutrition programs in Title VII of the Older Americans Act and part of community-based in-home care, may enhance nutrition and forestall institutionalization (Shi & Singh, 1998). As home-delivered meals are not part of standard home care and are not covered by Medicare, we view use of home-delivered meals as an enabling factor that facilitates living in the community. Similar to previous research (Picone & Wilson, 1999), we include two proxy measures for geographic environment: urban (respondents living in cities of 50,000 or more residents or their suburbs) and South. We also include measures of (a) whether the state has a large Medicaid home- and community-based care waiver program to reflect financial support for HHC services (Sloan, Picone, & Hoerger, 1997) and (b) the number of primary care physicians per 1,000 residents by state to measure underservice by physicians. State Medicaid waiver program funding of HHC greater than the median of $142.4 million indicates a high level of funding, whereas funding less than or equal to $142.4 million indicates a low level of funding. We examine HHC need with the five most prevalent chronic health conditions in the NLTCS and the ADLs and IADLs used as screening criteria in the NLTCS. The five chronic conditions are (a) joint diseases (rheumatism, paralysis, and arthritis), (b) diabetes, (c) heart conditions, (d) hypertension, and (e) breathing disorders (bronchitis, pneumonia, flu, emphysema, and asthma). Although persons in the sample may have one or more of these conditions, we enter each condition as a separate variable in our regressions. The simultaneous impact of multiple conditions can be explored by summing the impacts of individual conditions. We include the number of ADLs and IADLs as measures of disability (Cutler & Sheiner, 1994; Newman et al., 1990; Picone & Wilson, 1999; Sloan et al., 1997). As data on severity of health conditions are unavailable in the NLTCS database, use of continuous measures of ADLs and IADLs provides an approach to a proxy for severity. The nine ADLs are limitations with eating, getting in and out of bed, getting in and out of chairs, walking around inside, going outside, dressing, bathing, toileting, and controlling bowel movement/urination; and the seven IADLs are limitations with meal preparation, laundry, light housework, grocery shopping, managing money, taking medicine, and making telephone calls. The general form of our logit regression models is as follows:where DG = a vector of demograhic and sociostructural variables (gender, age, marital status, education), PH = a vector of health behaviors (smoking, exercise, and alcohol consumption practices), EN = a vector of enabling factors (income, income squared, health insurance status, number of potential caregivers living in the home, and home-delivered meal use), EF = a vector of environmental factors (urban, South, underservice by physicians, and residence in a state with high- or low-funded Medicaid waiver programs), HS = a vector of chronic health conditions (joint diseases, diabetes, heart conditions, hypertension, and breathing disorders), ADL = number of ADLs, IADL = number of IADLs, and ε = error term. We use three regression strategies to examine our first research question. We begin with regression models that include race dummy variables (Model I) and race interaction terms (Model II), because they are most directly comparable with previous research. If the race dummy variable is statistically significant, then race has a unique and independent influence on HHC use, given the role of other regression variables. The interactive model explores whether the impact of other independent variables on HHC use varies by race. Finally, using models stratified by race, we explore differences in estimated β coefficients and marginals (to measure impact of change in the independent variable on the probability of HHC use) for Blacks and Whites. We seek to determine whether independent variables impact HHC use differently for those of different races. We explore whether factors found significant in the unstratified data are significant in the stratified models and whether they are significant for both Blacks and Whites. For our second research question, we explore the role of racial differences in need, predisposing conditions, enabling factors, and environmental factors in explaining the observed racial difference in HHC use. If the variables included in our model do not fully account for racial differences in HHC use, then other nonmeasured factors are assumed to account for these differences. A priori, we expect Blacks' relatively lower health levels to be associated with greater HHC use compared with Whites. We also predict that the magnitude of the racial gap in HHC use directly corresponds with the size of the racial gap in variables included in the model. Our methodology for the second research question is application of the Oaxaca decomposition technique. The Oaxaca technique is used to quantify the percentage of the observed racial gap in HHC use explained by racial differences in the predisposing, enabling, need, and measured environmental characteristics. This technique is a standard approach in labor economics to determine the extent of wage discrimination. It was originally applied to log-linear models of wage compensation and later to nonlinear logit, probit, and bivariate probit discrimination models (Hinks & Watson, 2001; Mohanty, 2002). With use of this technique, the total racial gap in HHC use is the difference in the probability that Whites and Blacks use HHC services: (PW − PB). PW − PB is equal to F(Xw, ßw) − F(Xb, ßb), where ßw and ßb = regression coefficients for Whites and Blacks, Xw and Xb = characteristics of Whites and Blacks, and F = the cumulative logistic probability function: 1/[1 + e−(α+βXi)]. The total gap emphasized in the Oaxaca technique can also be expressed asThe first term of the above equation measures the gap in the probability of HHC use that is explained by racial differences in characteristics, that is, acceptable differences in use. The gap explained by racial differences uses the regression coefficients for Whites and examines how the probability of HHC use varies when these regression coefficients are paired with the characteristics of Whites versus the characteristics of Blacks. The unexplained gap (i.e., disparity) is the difference in the probability of use of HHC by Blacks assuming their own characteristics (Xb) and using the regression coefficients of Blacks (ßb) and the probability of use of HHC by Blacks assuming their own characteristics and the coefficients (ßw) of Whites. If unexplained differences in service use are found, they may be due to either unmeasured demand side or supply side factors, including racial differences in the knowledge about or preferences for HHC use, barriers to formal medical care, or possibly discrimination in the provision of HHC (i.e., factors in the health delivery system that influence accessibility, availability, and acceptablity of services). If the Oaxaca technique reveals unexplained differences in services use, then an important next step of investigation would be to seek data sources that allow identification of the causal factors among the unmeasured variables. However, if application of the Oaxaca technique reveals no unexplained differences in HHC service use, then an equitable market with acceptable racial differences is indicated. An additional caveat in the use of the Oaxaca decomposition approach is that measurement errors may also generate unexplained racial variation in the use of HHC. If regression variables are systematically measured with error by race, for example, self-reported diseases, then the systematic differences would appear as components of the unexplained gap in service use. Results Sample Characteristics Table 1 presents the study sample characteristics by race. About 16.2% of Whites and 20.2% of Blacks used HHC, or Blacks had a 25% greater chance of utilizing HHC. The sample includes 91% non-Hispanic Whites and 9% non-Hispanic Blacks, with similar mean ages of 78 and gender distribution (63% female). Whereas 40% of Whites live in medium or large cities and 35% are in the South, more than half of Blacks live in medium or large cities, and two thirds live in the South. Fully 70% of Whites, but only 29% of Blacks, have Medicare plus supplemental health insurance coverage, a striking racial differential. Blacks are much more likely than Whites to have only Medicare coverage (30.7% vs 17.3%) and 3.5 times more likely than Whites to be dually eligible (36.7% vs 10.4%). These racial insurance status differences correspond to differences in education and income. The average income is slightly over $21,000 for Whites and only $13,000 for Blacks; about 70% of Whites have at least a ninth grade education compared with 39% for Blacks. Preferences for preventive health and wellness differ for the two racial groups. Whites are more likely than Blacks to report drinking one to two times a week (15% vs 5%) and also getting exercise or playing sports regularly (23% vs 14%), but Blacks are slightly more likely to smoke (13% vs 10%). The ability of older persons with chronic illnesses to remain in the community is affected by the availability of caregivers (measured by the number of other adults in the household) and nonhousehold support services (measured by use of home-delivered meals). The number of other adults living in the home differs by race, with Blacks having an average of slightly over two, while Whites average only one and a half. The data indicate differences in the use of home-delivered meals (7.3% of Blacks vs 4.8% of Whites). Blacks are more likely than Whites to live in states with high physician/population ratios (6.63 vs. 5.00), with high Medicaid waiver programs (25% vs 8%), and also with small Medicaid waiver programs (12% vs 3%). Blacks report the presence of chronic health conditions more frequently than do Whites, with the exception of breathing disorders. The presence of heart conditions is reported by Blacks and Whites with similar frequency (28%). However, Blacks report joint disease, diabetes, and hypertension more frequently. Blacks average more ADLs than Whites (1.78 vs 1.38) and report more IADLs (averaging 1.08 vs 0.81). About 56% of Blacks report four or more ADLs (compared with 47% of Whites), and 29.4% report one or more IADLs (compared with 26.5% of Whites). Note that for both Blacks and Whites, ADL prevalence is higher than IADL prevalence. Based on the NLTCS screener criteria, this is not unexpected, as suggested in the data description section above. The prevalence for difficulty in getting around inside is 33.1%, and the prevalence for difficulty in getting around outside is only 3.9% (3.9% for Whites and 3.8% for Blacks). Thus, the results might be slightly skewed by including difficulty getting around outside as an ADL rather than an IADL. Racial Patterns in Use of Formal HHC Table 2 reports two models analyzing the logistic regression results for HHC use by Blacks and Whites. Model I includes a race dummy variable, and Model II also includes race interaction terms. Chronic medicial conditions, ADLs, IADLs, and enabling, environmental, and sociodemographic factors influence HHC use. In both Model I and Model II, heart condition is the chronic condition most likely to lead to HHC use. Older persons with more ADLs and IADLs have greater HHC use, and both income and income squared are statistically significant. Interestingly, there is a nonmonotonic relationship between income and HHC use. First, HHC use decreases as income increases; then, for those with higher incomes, increases in income are associated with increases in the probability of HHC use. It appears that HHC use is U-shaped with respect to income, as has been found for nursing home care (Headen, 1993; Morrow-Howell & Proctor, 1994). Those participating in meals programs are more likely to use HHC. However, as the number of persons living in the household increases, the probability of HHC use decreases. Dual eligibility increases the likelihood of HHC use, a likelihood that varies with the extent of state funding available for Medicaid community waiver programs. Women and persons who are older and have more formal years of education are more likely to use HHC, but those who are married are less likely to use HHC. The regression indicates that the variable of race is statistically insignificant in influencing HHC use. The role of race is further explored in Model II with race interaction terms. Each variable that was significant in Model I remains significant in the full interaction model, and the variables interacting race and diabetes and race and income are also significant. Thus, if Blacks have diabetes, they are more likely to use HHC than White diabetic individuals and also more likely than Blacks who have other major chronic conditions. Further, when we examine Blacks and Whites who have the same income, Blacks are more likely than Whites to use HHC. Table 3 presents the results of our stratified logistic regressions of the use of formal care by race and the marginals for significant coefficients. The influence of race on some independent variables, as reported in Table 2, allows us to estimate the regressions separately, although a Chow test indicates that race does not affect the impact of all independent variables on the dependent variable. With the exception of the marginals for the income and income-squared terms, in the reported logistic regressions, the marginal is exp(−α − βx)*bi/[1 + exp(−α − βx)]2, where bi = the regression coefficient for the variable for which a marginal is reported. Thus, Table 3 reports the marginal effects of the significant independent variables on the probabilities of HHC use, that is, the percentage point differences in the probability of HHC use when the regression variable changes by 1 unit. For regression variables that are dichotomous, the marginal is the percentage point difference in the probability of HHC use when the regression variable has a value of 1. Computing the marginals for the income and income-squared terms is complicated because we use an interactive term (income squared) in a nonlinear regression model, that is, logistic regression (Ai & Norton, 2003; Norton, Wang, & Ai, 2003). Thus, the marginal for income is (b2 + 2b4x2) * exp(−α − βx)/[1 + exp(−α − βx)]2, where b2 = the estimated coefficient of the income term, b4 = the coefficient of the squared income term, and x2 = the mean of the income variable. Marginals are comparable with reporting odds ratios, but they are more straightforwardly comprehended. Whereas odds ratios provide data on the relative increased probability of use, marginals provide data on the absolute increased probability of use. Odds ratios indicate the effect of a unit increase of the independent variable on the odds of an event [(probability of HHC use)/(1 − probability of HHC use)], but marginals report the effect of a unit increase in the independent variable on the probability of HHC use. The results in Table 3 indicate that despite the difference in sample size for the Black and White samples, there are unique racial patterns in HHC use. As suggested in the interactive model, among Blacks, diabetes increases the probability of using HHC by 10%; among Whites, heart conditions are associated with higher probabilities (4.5%) of HHC use. ADL disability affects HHC use for both Blacks and Whites. The direct effect of each additional ADL is to increase the probability of HHC use by 4.3% for Whites and by 4.6% for Blacks. For both racial groups, participation in delivered meals programs is associated with a greater likelihood of using HHC, with greater probability of 24% for Blacks and 12% for Whites. Several impacts are unique among Whites. More IADLs are associated with greater HHC use, with each additional IADL increasing the probability of HHC use by slightly less than 1%; each year of age increases the probability of HHC use by 0.25%; women have a slightly (3%) higher probability of HHC use; more education increases the likelihood of HHC use by 3%; and married status reduces the likelihood of HHC use by 6%. Three enabling/environmental characteristics also influence HHC use among Whites. Dual eligibility increases the probability of use by 8% in states with high Medicaid community waiver programs and by 7% in states with low Medicaid community waiver programs. There is a nonmonotonic relationship between income and HHC use. At lower income levels, each $1,000 increase in income decreases the probability of HHC use by 1.7%, but higher income (above $49,500) increases the probability of using HHC. For each person living in the house, the probability of HHC use decreases 3.5%. Explained and Unexplained Racial Patterns in the Use of HHC We apply the Oaxaca technique to determine how much of the racial differential in the probability of using HHC is explained by racial differences in medical and socioeconomic characteristics. We calculate the difference between the probability that Whites would use HHC assuming their own actual characteristics (0.1554) and the probability that Whites would use HHC assuming they have the characteristics of Blacks (0.1984), as measured by [F(Xw, ßw) − F(Xb, ßw)]. This calculated differential probability is (0.1554 − 0.1984) or −0.0430, which is approximately the actual racial gap in HHC use (0.162 − 0.202). Thus, the Oaxaca decomposition suggests that differences solely in the medical, socioeconomic, enabling, and environmental characteristics of Blacks and Whites should lead to the observed racial gap in HHC use. Expressed differently, we find that HHC is indeed an equitable health care market. Discussion Consistent with our prediction that HHC provides an equitable community-based health care market, the raw data indicate that in 1994, Blacks were more likely to use HHC than Whites. In this study, we find that predisposing, enabling, environmental, and need factors influence the use of HHC by both Blacks and Whites. Blacks have a 25% higher probability of using HHC than Whites. We find that racial differences in chronic conditions and socioeconomic factors fully predict this greater likelihood of HHC use among Blacks. This finding supports our hypothesis that HHC use is equitable by race. Older Black patients who live in the community with health conditions that are more extensive than those of White patients (possibly due to lower use of nursing home care and earlier-than-average release from hospitals) are absorbed by the HHC market, which presents few barriers to access. What do these results mean when other research (e.g., Wallace and colleagues 1998) finds that racial differences persist? The highlight of our findings is that HHC provides a uniquely equitable market. Similar to other research, our descriptive statistics (see Table 1) note a racial difference in HHC use, with Blacks having greater use. However, using our regression model and an expanded set of variables relative to those of Wallace and colleagues (1998), we find that racial differences in these sample characteristics explain racial differences in use of HHC among community-based older persons. Thus, the racial differences we identify in HHC are not racial disparities (Institute of Medicine, 2002) and do not reflect divergences in availability, accessibility, and acceptability. As described in previous research (Hodgson & Cohen, 1999; Wallace et al., 1998), chronic conditions (e.g., need) significantly influence HHC use, and those with the most impact vary by race. Diabetes is most reported by black HHC users, whereas a heart condition is most reported by Whites. These findings indicate that health policies or policy changes that differentially affect HHC use by chronic condition may also have unintended differential racial impacts. For example, the 1997 Balanced Budget Act, disallowing venipuncture as a sole qualifier for HHC coverage by Medicare, possibly imposed a greater burden on Blacks than Whites, because venipuncture patients are primarily diabetic. Research indicates that loss of HHC among disqualified venipuncture patients reduced their bathing and dressing assistance, requiring them to do more for themselves (Rubin et al., 2003). Thus, an unintended effect of the Balanced Budget Act, decreasing access to a service more heavily used by Blacks, possibly reduced the racial differential in HHC use. This legislation may even have reversed the prior trend of relatively higher HHC use among Blacks, and its impact is worthy of further investigation. ADL limitations increase HHC use for Blacks and Whites, although by different percentages, which is consistent with previous research (Kemper, 1992; Wallace et al., 1998; White-Means, 1997) noting the central role of disability level in affecting HHC use. The unique role that meals programs play in increasing use of HHC needs further exploration. A remaining question is why HHC differs from other health care services for older persons. Examining the differential impacts of race and chronic conditions on older persons' use of assistive devices, Rubin and White-Means (2001) find a 50% unexplained underuse gap for Blacks. White-Means (2000) finds unexplained underutilization of medical services and prescription drugs for Blacks. For HHC, we find higher use among Blacks and no unexplained use differential. It is important to note that whereas the NLTCS involves a complex sampling method, it is nonetheless nationally representative. The NLTCS has only a minimal research design effect of sampling (maybe 10%). Therefore, use of SUDAAN in regression analysis to adjust standard errors for underestimation of the true variation is considered unnecessary. Further, putting the SUDAAN model on top of the regression would possibly portray an inaccurate degree of precision (K. Manton, personal communication, 2004). This is in contrast to the procedures necessary for numerous other complex survey design data sources where the sampling technique leads to substantive correlation among observations, necessitating the use of weights. We predict use of HHC at one point in time and show that this cross-sectional view of HHC use indicates equity. Analyzing HHC use within a dynamic modeling framework remains an interesting area for future research. For example, it is unclear whether Blacks are more likely than Whites to use HHC for extended periods or with more or less continuity than Whites. These are also additional components of equity of care that are beyond the scope of this article. Another area for future research is the importance of including detailed measures of variation in support at the community level, comparable with the enabling variables that Bradley and colleagues (2002) label “availability of support.” We are unable to incorporate these variables analytically, because although we have individual data, the NLTCS lacks detailed community-level variables, giving our research imperfect measures of the community variables of Bradley and colleagues. Inclusion of these community support variables would address significant differences at the small area level. Although our data suggest there are not unexplained racial differences in HHC use, we are unable to determine whether there are racial differences in the types and quality of HHC received, because different types of HHC are not delineated in the data. Thus, further research is still warranted to understand whether there is equity in terms of quality of care provided by different types of helping organizations. In addition, questions of how racial differences in household and family formation affect HHC use and whether the actual intrahousehold relationships are important deserve future investigation. Our results indicate that being married and also the number of other adults in the household are significant for Whites but not for Blacks. The relationship between HHC use and cultural differences is another area for prospective research. Culture, expressed in group values and beliefs, may affect the selection among long-term care options and between institutional and noninstitutional choices (Thornton & White-Means, 2000; Wallace et. al., 1998). Thus, measured racial differences may embody cultural influences that are implicit rather than explicit. Explicit measures of knowledge and cultural preferences for receipt of HHC versus other health care services, personal experiences of discriminaton when interacting with the health care system or with the HHC services market, and charactersistics of patients' community health care systems are further issues needing investigation. These analyses would provide input for effective public policy measures to address racial disparities in use of health care services and enable replication of the equity in use that our data indicate is provided by the HHC market. Decision Editor: Charles F. Longino Jr., PHD Table 1. Characteristics of the Sample. Total Non-Hispanic White Non-Hispanic Black Sample Characteristics % or Mean No. SD % or Mean No. SD % or Mean No. SD N, sample size 100% 3,448 90.4% 3,136 9.6% 312 Home health care usea 16.6% 571 0.37178 16.2% 508 0.3685 20.2% 63 0.40208 Av. age (mean), yr 77.8 7.3624 77.8 7.2659 78 8.28181 Gender Male 36.4% 1,255 36.4% 1,142 36.7% 115 Female 63.6% 2,192 0.48113 63.6% 1,994 0.48109 63.3% 198 0.4823 Marital status Marriedb 46.1% 1,590 0.49851 47.6% 1,493 0.49949 31.0% 97 0.46228 Nonmarried 53.9% 1,858 52.4% 1,643 69.0% 216 Location Medium or large cityb 41.2% 1,420 0.49229 39.9% 1,251 0.48942 54.0% 169 0.49906 Small city, town or rural 58.8% 2,027 60.1% 1,885 46.0% 144 Region Southb 37.2% 1,282 0.4834 34.7% 1,088 0.47597 62.9% 197 0.48406 Non-South 62.8% 2,165 65.3% 2,048 37.1% 116 Insurance status Medicare only 18.5% 638 17.3% 543 30.7% 96 Medicare plus supplementb 66.7% 2,300 0.4712 70.4% 2,208 0.45638 29.4% 92 0.45672 Medicare and Medicaid 12.7% 438 10.4% 326 36.7% 115 Education 9th grade and aboveb 66.9% 2,307 0.47082 69.6% 2,183 0.46001 39.1% 122 0.48876 K through 8th grade 33.1% 1,141 30.4% 953 60.9% 190 Income (mean $)b $20,451 18,422.8 $21,182 18,821.7 $13,096 11,448.1 Other adults (mean)b 1.59 1.12943 1.55 1.06706 2.06 1.55599 Delivered meals Yes 5.0% 172 4.8% 151 7.4% 23 No 95.0% 3,276 0.21893 95.2% 2,985 0.21412 92.6% 289 0.26173 Behavioral characteristics Smokinga 10.3% 355 0.30395 10.0% 314 0.29979 13.5% 42 0.34186 Drinkingb 14.1% 486 0.34772 15.0% 470 0.35669 5.1% 16 0.22093 Exerciseb 21.9% 755 0.41361 22.7% 712 0.41899 13.8% 43 0.34527 Chronic health conditions Joint diseasesb 69.1% 2,383 0.46198 68.2% 2,139 0.46549 77.9% 243 0.41569 Diabetesb 16.4% 565 0.37047 15.5% 486 0.36194 25.6% 80 0.43735 Heart conditions 27.7% 955 0.44757 27.7% 869 0.44764 27.6% 86 0.44755 Hypertensionb 43.6% 1,503 0.49591 42.3% 1,327 0.49409 56.4% 176 0.49667 Breathing disordersb 30.3% 1,045 0.45953 30.7% 963 0.46149 25.6% 80 0.43735 ADL frequency distribution 0 ADLs 51.8% 1,786 0.49973 52.6% 1,649 0.49941 44.2% 138 0.49746 1 or more ADLsb 48.2% 1,662 47.4% 1,487 55.8% 174 ADL (mean)b 1.41 1.92609 1.38 1.91197 1.75 2.03585 IADL frequency distribution 0 IADLs 73.3% 2,527 0.44267 73.5% 2,306 0.44123 70.5% 220 0.45672 1 or more LADLsb 26.7% 921 26.5% 830 29.5% 92 IADL (mean)b 0.84 1.77436 0.81 1.74772 1.08 2.00949 Underserved by physician (mean) 5.15 2.7202 5.00 2.58481 6.63 3.49789 High Medicaid waiverb 9.3% 321 0.29102 7.8% 245 0.26791 25.0% 78 0.43371 Low Medicaid waiverb 3.4% 117 0.18108 2.6% 81 0.15865 11.5% 36 0.32 Notes: ADL = activity of daily living; IADL = instrumental activity of daily living. Source: 1994 National Long-Term Care Study, Center for Demographic Studies, Duke University. aIndicates racial difference significant for variable at.10 or better, based on t test for proportions. bIndicates racial difference significant for variable at.05 or better, based on t test for proportions. Table 2. Results of Logistic Regressions of the Use of Formal Care by Race for Non-Hispanic Black and White Older Persons. Model I Model II With Dummy Race Variable With Interactive Termsa Variables Coefficient t Statistic Coefficient t Statistic Intercept −4.747 −6.70764*** −4.656 −6.0807*** Joint diseases 0.0799 0.6277 0.1508 1.1187 Diabetes 0.0895 0.6303 −0.00669 −0.0430 Heart condition 0.4267 3.7496*** 0.4555 3.7738*** Hypertension −0.0728 −0.6535 −0.1012 −0.8584 Breathing disorders 0.1098 0.9498 0.1571 1.2941 ADL sum 0.4298 15.1873*** 0.439 14.5364*** IADL sum 0.0802 3.0037*** 0.0967 3.3693*** Underserved 0.000251 0.0111 −0.00561 −0.2262 South 0.0709 0.5645 0.0356 0.2681 Gender 0.3057 2.2882** 0.3088 2.1746** Age 0.0262 3.4026*** 0.0256 3.0955*** Married −0.564 −4.0286*** −0.5959 −3.9993*** Education 9th grade+ 0.3204 2.6155*** 0.3331 2.5235** Urban −0.0707 −0.6284 −0.1099 −0.9235 Medicare plus −0.00896 −0.0634 0.0493 0.3237 Dual eligible/high waiver 0.7514 4.0182*** 0.8195 3.8950*** Dual eligible/low waiver 0.5834 2.1599** 0.7364 2.4057** Alcohol 0.0939 0.4798 0.1278 0.6330 Smokes 0.1224 0.6405 0.0901 0.4458 Exercises −0.094 −0.6196 −0.0475 −0.3006 Income ($10,000 units) −0.2 −2.1423** −0.03 −3.0374*** Income squared 0.0000026 3.1257*** 0.000003 3.4788*** Other adults −0.2983 −4.9800*** −0.3545 −5.0861*** Delivered meals 1.2842 7.0599*** 1.1943 6.1689*** Black −0.0487 −0.2542 −1.3254 −0.5915 B Joints −0.6644 −1.4397 B Diabetes 0.7688 1.7946* B Heart condition −0.0963 −0.2402 B Hypertension 0.3958 0.9633 B Breathing disorders −0.6389 −1.4141 B ADL sum −0.0773 −0.7761 B IADL sum −0.1134 −1.3065 B Underserved 0.0694 1.0206 B South 0.4699 1.0021 B Gender −0.2683 −0.5492 B Age 0.00607 0.2409 B Married −0.1268 −0.2498 B Education 0.2349 0.5696 B Urban 0.4882 1.1231 B Medicare plus −0.5015 −0.9734 B Dual eligible/high waiver −0.1561 −0.3228 B Dual eligible/low waiver −0.7839 −1.1313 B Alcohol −0.6826 −0.7285 B Smokes 0.213 0.3208 B Exercise −0.3984 −0.6171 B Income ($10,000 units) 0.97 1.6724* B Income squared −0.000015 −1.299 B Other adult 0.2093 1.3788 B Delivered meals 0.6548 1.0868 Sample size (N) 3,448 3,448 −2 log L 2,368.28 2,340.741 χ2 726.962*** 754.498*** Notes: ADL = activity of daily living; IADL = instrumental activity of daily living. aRace interaction terms begin with “B” preceding a variable name. *Significant at p =.10; **Significant at p =.05; ***Significant at p =.01. Table 3. Results of Logistic Regressions of the Use of Formal Care Stratified by Race. Non-Hispanic White Older Persons Non-Hispanic Black Older Persons Variables Coefficient t Statistic Marginals Coefficient t Statistic Marginals Intercept −4.656 −6.08071*** −5.9814 −2.84017*** Joint diseases 0.1508 1.11869 −0.5136 −1.16383 Diabetes −0.00669 −0.04299 0.7621 1.90955* 0.0975 Heart condition 0.4555 3.77382*** 0.0450231 0.3592 0.93958 Hypertension −0.1012 −0.85835 0.2946 0.74848 Breathing disorders 0.1571 1.29407 −0.4818 −1.10708 Underserved −0.00561 −0.22621 0.0638 1.0079 ADL sum 0.439 14.5364*** 0.0433922 0.3617 3.81138*** 0.04627 IADL sum 0.0967 3.36934*** 0.0095581 −0.0168 −0.20513 South 0.0356 0.26807 0.5056 1.12431 Gender 0.3088 2.17465** 0.0305228 0.0405 0.08665 Age 0.0256 3.09553*** 0.0025304 0.0317 1.33193 Married −0.5959 −3.99933*** −0.058901 −0.7227 −1.48949 Education 9th grade+ 0.3331 2.52348** 0.0329247 0.568 1.4538 Urban −0.1099 −0.92353 0.3783 0.90481 Medicare plus 0.0493 0.3237 −0.4522 −0.91873 Dual eligible/high waiver 0.8195 3.89496*** 0.0810021 0.6634 1.52366 Dual eligible/low waiver 0.7364 2.40575** 0.0727882 −0.0475 −0.07642 Alcohol 0.1278 0.63299 0.5549 0.60645 Smokes 0.0901 0.44582 0.303 0.47913 Exercises −0.0475 −0.30063 −0.4459 −0.7123 Income ($10,000 units) −0.3 −3.03736*** −0.017 0.71 1.2456 Income squared 0.000003 3.47876*** 0.051 −0.00001 −1.0392 Other adults −0.3545 −5.08608*** −0.03504 −0.1452 −1.07715 Delivered meals 1.1943 6.1689*** 0.1180486 1.8492 3.24137*** 0.23658 Sample size (N) 3,136 312 −2 log L 2,109.02 231.722 χ2 669.194*** 82.185*** Notes: ADL = activity of daily living; IADL = instrumental activity of daily living. *Significant at p =.10; **Significant at p =.05; ***Significant at p =.01. References Ai, C., Norton, E. (2003). Interaction terms in logit and probit models. Economics Letters, 80,123-129. Andersen, R. M. (1995). Revisiting the behavioral model and access to medical care: Does it matter? Journal of Health and Social Behavior, 36,1-10. Bradley, E. H., McGraw, S. A., Curry, L., Buckser, A., King, K., Kasl, S., et al. (2002). Expanding the Andersen model: The role of psychosocial factors in long-term care use. Health Services Research, 37,1221-1243. Brown, A. (1998). No place like home care. Black Enterprise, 29,33. Cagney, K., Agree, E. (1999). Racial differences in skilled nursing care and home health use: The mediating effects of family structure and social class. Journal of Gerontology: Social Sciences, 54B,S223-S236. Cutler, D. M., Sheiner, L. M. (1994). Policy options for long term care. In D. A. Wise (Ed.), Studies in the Economics of Aging (pp. 395–434). Chicago: University of Chicago Press. Falcone, D., Broyles, R. (1994). Access to long term care: Race as a barrier. Journal of Health Politics, Policy, and Law, 19,583-595. Freedman, V. A. (1999). Long-term admissions to home health agencies: A life table analysis. The Gerontologist, 39,16-24. Headen, A. (1993). Economic disability and health determinants of the hazard of nursing home entry. Journal of Human Resources, 28,80-110. Hinks, T., Watson, D. (2001). A multinomial logit nondiscrimination approach to estimating racial wage and occupational discrimination. Applied Economics, 33,605-612. Hodgson, T. A., Cohen, A. J. (1999). Medical care expenditures for diabetes, its chronic complications, and its comorbidities. Preventive Medicine, 29,173-186. Institute of Medicine. (2002). Unequal treatment: Confronting racial and ethnic disparities in health care. Washington, DC: National Academy of Sciences. Kemper, P. (1992). The use of formal and informal home care by the disabled elderly. Health Services Research, 27,421-451. Kromrey, J., Hines, C. (1994). Nonrandomly missing data in multiple regression: An empirical comparison of common missing-data treatments. Educational and Psychological Measurement, 54,573-593. Liu, K., Manton, K. G., Aragon, C. (2000). Changes in home care use by older people with disabilities: 1982–1994. AARP Public Policy Institute, 2000–02, January. Manton, K., Corder, L., Stallard, E. (1997). Chronic disability trends in the elderly United States populations: 1982–1994. Proceedings of the National Academy of Science, 94,2593-2598. Miller, B., Campbell, R. T., Davis, L., Furner, S., Giachello, A., Prohaska, T., et al. (1996). Minority use of community long-term care services: A comparative analysis. Journal of Gerontology: Social Sciences, 51B,S70-S81. Mohanty, M. (2002). A bivariate probit approach to the determination of employment: A study of teen employment differentials in Los Angeles County. Applied Economics, 34,143-156. Morrow-Howell, N., Proctor, E. (1994). Discharge destinations of Medicare patients receiving discharge planning: Who goes where? Medical Care, 32,486-497. Mui, A., Burnette, D. (1994). Long term care service use by frail elders: Is ethnicity a factor? The Gerontologist, 34,190-198. National Association of Home Care. (2000). Basic statistics about home care. Retrieved October 2, 2002, from http://www.nahc.org/consumer/hcstats.html. Newman, S. J., Struyk, R., Wright, P., Rice, M. (1990). Overwhelming odds: Caregiving and the risk of institutionalization. Journal of Gerontology: Social Sciences, 45B,S173-S183. Norton, E., Wang, H., Ai, C. (2003). Computing interaction effects and standard errors in logit and probit models. Working paper. September 9, 2003. Oaxaca, R. (1973). Male–female wage differentials in urban labor markets. International Economic Review, 14,693-709. Picone, G., Wilson, R. M. (1999). Medicare Home Health Agency utilization, 1984–1994. Inquiry, 36,291-303. Rotnitzky, A., Wypij, D. (1994). A note on the bias of estimators with missing data. Biometrics, 50,1163-1170. Rubin, R. M., White-Means, S. I. (2001). Race, disability, and assistive devices: Sociodemographics or discrimination? International Journal of Social Economics, 28,927-941. Rubin, R. M., White-Means, S. I., Feagin, T. (2003). Impacts of the 1997 Balanced Budget Act on venipuncture beneficiaries. Journal of Applied Gerontology, 22,252-268. Shi, L., Singh, D. A. (1998). Long term care. In Delivering health care in America: A systems approach (pp. 344–391). Gaithersburg, MD: Aspen. Sloan, F., Picone, G., Hoerger, T. J. (1997). The supply of children's time to disabled elderly parents. Economic Inquiry, 35,295-308. Stone, R., Wiener, J. (2001). Who will care for us? Addressing the long-term care workforce crisis. Washington, DC: Urban Institute and American Association of Homes and Services for the Aging. Switzer, F., III, Roth, P., Switzer, D. (1998). Systematic data loss in HRM settings: A Monte Carlo analysis. Journal of Management, 24,763-774. Thornton, M. C., White-Means, S. I. (2000). Race versus ethnic heritage in models of family economic decisions. Journal of Family and Economic Issues, 21,65-86. Wallace, S. (1990). The no-care zone: Availability, accessibillity, and acceptability in community-based long-term care. The Gerontologist, 30,254-261. Wallace, S. P., Levy-Storms, L., Ferguson, L. (1995). Access to paid in-home help among disabled elderly people: Do Latinos differ from non-Latino whites? American Journal of Public Health, 85,970-975. Wallace, S., Levy-Storms, L., Kington, R., Andersen, R. (1998). The persistence of race and ethnicity in the use of long term care. Journal of Gerontology: Social Sciences, 53B,S104-S112. White-Means, S. I. (1997). The demands of persons with disabilities for home health care and the economic consequences for informal caregivers. Social Science Quarterly, 78,955-972. White-Means, S. I. (2000). Racial patterns in disabled elderly persons' use of medical services. Journal of Gerontology: Social Sciences, 55B,S1-S14.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0AE1EC5958BBE547537B2D57668C9B9FD6BD7723.txt b/test/dataset/in/resources/corpus/Clean_0AE1EC5958BBE547537B2D57668C9B9FD6BD7723.txt
deleted file mode 100644
index b2cf652..0000000
--- a/test/dataset/in/resources/corpus/Clean_0AE1EC5958BBE547537B2D57668C9B9FD6BD7723.txt
+++ /dev/null
@@ -1 +0,0 @@
- gerona J Gerontol A Biol Sci Med Scigerona The Journals of Gerontology Series A: Biological Sciences and Medical Sciences J Gerontol A Biol Sci Med Sci 1079-5006 1758-535X Oxford University Press 10410.1093/gerona/60.1.104 Journal of Gerontology: Medical Sciences Nighttime Oxygen Desaturation and Symptoms of Sleep-Disordered Breathing in Long-Stay Nursing Home Residents Martin Jennifer L. 1 Mory Aaron K. 2 Alessi Cathy A. 1 1University of California, Los Angeles, Multicampus Program in Geriatric Medicine and Gerontology, and VA Greater Los Angeles Healthcare System, Geriatric Research Education and Clinical Center, Los Angeles, California. 2Jefferson Medical College, Philadelphia, Pennsylvania. Address correspondence to Jennifer L. Martin, PhD, VA Medical Center, GRECC (11E), 16111 Plummer Street, North Hills, CA 91343. E-mail: jemartin@ucla.edu 1 2005 60 1 104 108 20 8 2003 17 7 2003 The Gerontological Society of America 2005 Background. Sleep-disordered breathing (SDB) is common in older adults and has been implicated as a cause of decreased quality of life and even death. Sparse data exist on SDB in the nursing home setting. The authors evaluated SDB (using attended nocturnal pulse oximetry) in nursing home residents with daytime sleepiness and nighttime sleep disturbance. Methods. Pulse oximetry was used to estimate the prevalence of nighttime oxygen desaturation in 109 long-stay nursing home residents (mean [standard deviation] age = 86.2 [9.2] years; 74% women). Pulse oximetry findings were compared to a structured observational measurement of symptoms of SDB, the Observational Sleep Assessment Instrument. Seventy-one participants had concurrent wrist actigraphy to estimate total sleep time during oximetry recording. Results. Using the oxygen desaturation index (ODI; average number of oxygen desaturations 4% or more below the baseline level per hour), the authors found that 40% of the residents had abnormal ODI (ODI more than 5, which is suggestive of SDB). Of all observational variables assessed, only loud breathing during sleep was significantly correlated with ODI (r =.284; p =.003). When ODI was adjusted for estimated total sleep time, higher adjusted ODI was associated with higher body mass index (kg/m2). Conclusions. Abnormal ODI is common in nursing home residents. Observed loud breathing at night and high body mass index may suggest that further assessment of SDB is indicated. Future research should determine the importance of SDB and abnormal nocturnal oxygen desaturation on functioning and quality of life in nursing home residents. hwp-legacy-fpage 104 hwp-legacy-dochead RESEARCH ARTICLE AGE-RELATED changes in sleep patterns and sleep structure have been described in older adults. Increased daytime sleepiness, taking longer to fall asleep, and more episodes of nighttime awakening are accompanied by decreased sleep efficiency (i.e., time asleep/time in bed) and shorter periods of deep sleep in older adults (1). These changes are both common and more severe in nursing home residents than in community-dwelling older adults. Sleep disruption in nursing home residents likely results from multiple factors, including increased time spent in bed, reduced daytime bright light exposure, lack of structured daytime activities, and few regular social and environmental cues to regulate sleeping and waking patterns (2). In addition to age-related changes and environmental factors, certain sleep disorders, such as sleep-disordered breathing (SDB), are more common in older persons (1). SDB is defined as repeated pauses in breathing during sleep lasting 10 seconds or more (3,4). These respiratory events often lead to decreased blood oxygen saturation levels, which can precipitate awakenings followed by resumption of breathing. In addition, SDB contributes to daytime sleepiness and nighttime sleep fragmentation, which can impair daytime functioning. Several adverse outcomes in older adults have been associated with SDB, including cardiovascular consequences (hypertension, cardiac arrhythmias, myocardial infarction, and stroke) and cognitive impairment (memory problems, attention difficulties, and difficulty concentrating) (1,5–8). Some studies have shown that older adults with severe SDB have higher mortality rates compared with those without SDB (9). Associations between SDB and increased age, higher body mass index (BMI), male sex, and dementia have been reported (4,10,11). An association between SDB and increased risk for mortality has been reported in both community-dwelling older adults (9) and female nursing home residents (12). Traditionally, SDB is studied with polysomnography in a sleep laboratory. To definitively diagnose SDB, polysomnography uses multiple channels including the electroencephalogram, electrooculogram, electromyogram, electrocardiogram, pulse oximetry, chest movement, and airflow (3). Persons undergoing polysomnography are then assigned a respiratory disturbance index (mean respiratory events per hour of sleep). A respiratory disturbance index less than 5 is typically considered normal, whereas a respiratory disturbance index of 15 or more suggests moderate to severe SDB worthy of treatment (4,13). Some persons cannot be examined in the traditional sleep laboratory setting. Less intensive ambulatory methods can be useful in detecting sleep disorders in such persons. Pulse oximetry alone has been investigated as an alternative way to screen for SDB because it is simple, allows the patient to sleep in their usual environment, and is less expensive than traditional polysomnography. Pulse oximetry is particularly attractive for use in nursing home residents with dementia because it is more easily tolerated than multichannel polysomnography. Pulse oximetry alone yields an oxygen desaturation index (ODI) rather than a respiratory disturbance index. The ODI is the average number of oxygen desaturations at least 4% below baseline level per hour. Several studies have tried to determine the sensitivity and specificity of this method, with results ranging from sensitivity of 31% to 98% and specificity of 41% to 100%, depending on the specific devices and ODI cutoffs used (14,15). Using a cutoff of ODI of 5 or more appears to optimize the sensitivity of identification of patients with SDB compared with patients without SDB in sleep clinic samples (15); nonetheless, pulse oximetry has been shown to be more specific than sensitive in most studies. As a result, pulse oximetry is unlikely to falsely identify a person as having SDB, but it is more likely to fail to identify some persons who, in fact, have SDB (14,16). To date, no validation studies have been conducted in nursing home settings. One caveat against using pulse oximetry alone is that the indices can be computed based only on the duration of the recording, whereas in the laboratory setting, the indices are computed based on the total amount of time the patient is asleep (i.e., time awake is excluded). This may lead to an underestimation of SDB severity when pulse oximetry is used alone. Estimation of total sleep time (e.g., with wrist actigraphy) may improve the accuracy of pulse oximetry-based estimation of SDB severity. The cardinal symptoms of SDB are snoring and excessive daytime sleepiness. When both symptoms are observed together, SDB is often suspected and further evaluation is indicated. Structured observations of nighttime sleep and symptoms of SDB have been studied in the nursing home setting (17,18), but only limited objective data exist on nighttime oxygen desaturation among nursing home residents. In this study, we assessed the frequency of nocturnal oxygen desaturation in nursing home residents with daytime sleepiness and nighttime sleep disruption. We hypothesized that observations of disturbed breathing during sleep would be related to ODI. We also hypothesized that persons with more cognitive impairment would have higher ODI and more observed breathing disturbances during sleep. We also expected to confirm findings of previous studies showing that male sex and BMI were correlated with SDB severity. Methods Participants Participants were residents in four Los Angeles-area community nursing homes with daytime sleepiness and nighttime sleep disruption enrolled in a randomized controlled trial of nonpharmacologic interventions to improve sleep. Pulse oximetry data for the current study were collected under usual-care conditions. Daytime sleepiness was assessed using behavioral observations of sleep versus wakefulness performed by research staff every 15 minutes from 9:00 am to 5:00 pm for 2 days for all residents at each nursing home. Only those who were bed-bound, in contact isolation, or who left the facility before screening were excluded. Daytime sleepiness was defined as being asleep on 15% or more of observations, where “sleep” was defined as “eyes closed with no purposeful movement.” Residents who met criteria for observed daytime sleepiness and who gave consent to participate then had 2 nights of wrist actigraphy. Those who were scored asleep 80% or less of the time between 10:00 pm and 6:00 am were enrolled in the study. Apparatus Enrolled participants had 1 night (approximately 9:00 pm to 6:00 am) of attended pulse oximetry monitoring with a fingertip sensor connected via cable to an Ohmeda Biox 3700 pulse oximeter (Louisville, CO), which recorded oxygen saturation, heart rate, and time continuously on a laptop computer. Observations during sleep were performed simultaneously by research staff using the Observational Sleep Assessment Instrument (OSAI). The OSAI was developed to assess symptoms of SDB in nursing home residents (18). This instrument involves hourly 3-minute observations during which snoring, breathing rate, loudness and continuity, and chest movements are recorded. To estimate total sleep time during oximetry recordings, nighttime wrist actigraphy was also performed using a Mini-motionlogger (Ambulatory Monitoring, Inc., Ardsley, NY). The actigraph was placed on the dominant wrist, and activity levels were recorded in 60-second epochs. Sleep/awake was scored using a validated algorithm within the ActionW software (Ambulatory Monitoring, Inc.). Wrist actigraphy has been shown to accurately reflect total sleep time in nursing home residents (19). The Mini-Mental Status Examination (MMSE) was used to measure cognitive functioning (20). A measure of comorbidity, the Cumulative Illness Rating Scale–Geriatrics was completed by a study physician using data from a structured medical record review and a brief physical examination (21). In our previous work, the Cumulative Illness Rating Scale–Geriatrics predicted acute illness episodes and death in long-stay nursing home residents (22). Procedures Participants underwent comprehensive screening including the MMSE and Cumulative Illness Rating Scale–Geriatrics. Demographic and medical record information (i.e., diagnoses and medications received during the study) were noted. Pulse oximetry recordings, OSAI observations, and wrist actigraphy were collected or performed while residents slept in their own rooms receiving usual care. Data Analysis For each participant, baseline oxygen saturation was determined using the mean oxygen saturation rate during the first 30 minutes of recording or the first 30-minute period with sufficient data to assign a baseline. An oxygen desaturation event was defined as a decrease to 4% or more below the baseline level. Oxygen desaturation indices were computed for each participant with at least 3 hours of pulse oximetry (mean [SD] recording length = 7.4 [1.2] hours). The ODI was defined as the average number of oxygen desaturations per hour of recording. We used actigraphically estimated total sleep time to calculate the number of desaturations per hour of sleep for 71 participants with acceptable actigraphy recordings. Although there were no differences in ODI between patients with (n = 71) and without (n = 38) wrist actigraphy recordings (t107 = −.319, p =.75), patients with wrist actigraph recordings had lower MMSE scores (9.6 vs 14.5; p =.003), fewer medical diagnoses (9.3 vs 11.0; p =.040), and took fewer routine (8.3 vs 10.6, p =.025) and as-needed (1.5 vs 2.6, p =.002) medications compared with participants without actigraphic recordings. As a result of these differences, we report analyses based on calculations of ODI with and without adjustment for total sleep time. We used Student's t tests and regression analyses to test for associations between ODI and descriptive and OSAI variables. We anticipated that higher ODI would be associated with older age, higher BMI, lower MMSE scores, more medical diagnoses, use of more medications, and higher (more severe) Cumulative Illness Rating Scale–Geriatrics scores. For variables that were highly skewed, descriptives are presented as raw values and we used a normal scores transformation for statistical analyses (23). Two-tailed probability values less than.05 were considered significant. Results Participant Screening Of the 492 nursing home residents screened with daytime behavioral observations, 339 met criteria for daytime sleepiness. Of those, 194 gave informed consent to participate and 133 met criteria for nighttime sleep disturbance. One hundred twenty-one nursing home residents were randomized into the larger controlled trial of which 109 had 3 or more hours of pulse oximetry and were included in the current study. Table 1<--?1--><--?2-->shows descriptive characteristics. No participants had a documented diagnosis of SDB in medical records. Of those participants with complete oximetry data, 77 had concurrent wrist actigraphy recordings during oximetry testing. Actigraphy recordings were considered inadequate if the patient was not asleep for at least 1 hour during the recording (n = 6). Analyses were based on the 109 participants with pulse oximetry and were repeated for the 71 participants with adequate concurrent actigraphic recordings. Table 2 shows the results of the simultaneous pulse oximetry, OSAI, and wrist actigraphy. The mean ODI for the group was 7.0 (SD, 9.3). Forty percent of patients had ODI ≥5, 23% had ODI ≥10, and 13% had ODI ≥15 (Figure 1). The ODI was not significantly correlated with age, sex, ethnicity, BMI, MMSE score, number of diagnoses, or number of medications taken. Loudness of breathing was the only OSAI variable significantly correlated with ODI (r =.28, p =.003; Table 3). Louder breathing was also related to higher BMI (r =.28 p =.004). When BMI was included in the regression model, loudness of breathing remained a significant predictor of ODI (F2,104 = 5.23, p =.007). Oxygen Desaturation Index Based on Actigraphically Estimated Total Sleep Time The mean total sleep time during the oximetry recording period for patients with acceptable actigraphy recordings (n = 71) was 4.3 hours. Patients with higher ODI had shorter total sleep time (r = −.52, p <.0005). When the ODI was computed based on total sleep time, the severity of SDB appeared substantially worse (mean [SD] adjusted ODI = 20.6 [40.6]). Fifty-one percent of patients had adjusted ODI ≥5, 62% had adjusted ODI ≥10, and 25% had adjusted ODI ≥15. The association between the standard and adjusted ODI was strong (r =.94, p <.0005). The relationship between adjusted ODI and BMI was significant such that patients with higher BMI had higher adjusted ODI (r =.30, p =.012). Adjusted ODI was not significantly associated with age, sex, ethnicity, MMSE, number of diagnoses, or number of medications, and adjusted ODI was not related to OSAI variables (i.e., snoring, breathing rate, continuity of breathing, loudness of breathing, or continuity of chest movements). Discussion We found that at least 40% of long-stay nursing home residents with evidence of daytime sleepiness and nighttime sleep disturbance had an abnormal ODI (≥5), and 13% had moderate-to-severe oxygen desaturation (ODI ≥15). When ODI was adjusted for actigraphically estimated total sleep time, the prevalence of severe oxygen desaturation appeared even higher (25% of participants had adjusted ODI ≥15). The ODI was not associated with age, sex, or comorbidity. Although we found that observed loud breathing was significantly related to oxygen desaturation, we did not find a relationship between other observed symptoms of SDB such as snoring or discontinuous chest movement. It is possible that these events are simply more difficult to detect with nighttime observations than is loud breathing. One previous study of SDB in a nursing home population using thoracic and abdominal effort sensors plus wrist actigraphy (12) found 43% of residents had at least five apneas per hour of sleep with no statistical difference between men and women. This is comparable to our results showing that 40% of nursing home residents with daytime sleepiness had ODI ≥5, with no statistical difference between men and women, although our methods of measurement were different. Although BMI was not significantly associated with ODI (based on total recording time), it was related to the adjusted ODI (based on total sleep time). Given that patients with actigraphy recordings were medically healthier (i.e., took fewer medications and had fewer diagnoses), the relationship between ODI and BMI may be stronger in these persons. Nursing home residents with multiple medical comorbid conditions and who take more medications may have SDB as the result of these factors rather than because of high BMI. When BMI and loudness of breathing were evaluated simultaneously, both were related to ODI. Although BMI is strongly predictive of SDB in community-based populations, obesity by itself may not be sufficient to predict SDB in the nursing home setting. Further study is needed to determine whether the combination of high BMI and loud breathing during sleep are predictive of SDB. Although previous studies have shown a link between cognitive impairment and SDB, we did not find an association between MMSE score and ODI. Twenty-nine participants (27%) received a score of 0 out of 30 on the MMSE, suggesting that a floor effect may have influenced our findings. Nursing home residents with intact cognitive abilities but poor physical health may have more SDB than residents who are cognitively impaired but physically healthy. The main limitation of this study is that pulse oximetry is an indirect method of estimating SDB because it only detects respiratory events sufficient to cause oxygen desaturation, and the true number of events occurring during sleep cannot be precisely determined. We tried to address this concern with concurrent wrist actigraphy; however, we could not obtain actigraph recordings on all 109 participants. The participants with actigraphic recordings tended to be more cognitively impaired but more medically healthy. Although full polysomnography would have been more accurate, many nursing home residents will not tolerate such cumbersome recording equipment, and sleep electroencephalography can be difficult to interpret in persons with dementia. The strength of this study is that we were able to complete pulse oximetry recordings for 90% of enrolled participants with daytime sleepiness and nighttime sleep disturbance, which enhances the generalizability of our findings. In addition, based on validation studies conducted in sleep clinic samples, pulse oximetry has been shown to be more specific than sensitive in the identification of patients with SDB (14), and we may have underestimated the true prevalence of SDB in our study sample as a result of using pulse oximetry rather than complete polysomnography. Conclusion We found a high prevalence of abnormal oxygen desaturation among nursing home residents with daytime sleepiness and nighttime sleep disruption. The clinical significance of this desaturation is unclear. Future research should focus on potential implications of oxygen desaturation and SDB in this setting. Decision Editor: John E. Morley, MB, BCh Figure 1. The distribution of oxygen desaturation index (ODI) scores (unadjusted for total sleep time) is shown for all participants. Sixty percent of participants fell within the normal range (ODI <5), whereas the remaining 40% had ODI of 5 or more, suggesting at least mild sleep-disordered breathing Table 1. Descriptive Characteristics (N = 109) and Associations With Oxygen Desaturation Index (ODI)*. Mean (SD) or % r p value Age 86.2 (9.2) −0.044 .65 Years in nursing home 2.9 (2.9) −.042 .68 Body mass index, Kg/m2 25.3 (5.6) 0.186 .053 MMSE score< 11.3 (9.6) −0.089 .36 No. of medical diagnoses 9.9 (3.9) −0.088 .37 No. of routine medications 9.1 (5.2) −0.042 .67 No. of PRN medications 1.9 (1.8) −0.144 .14 CIRS-G score 24.6 (5.0) 0.063 .52 t p value Gender, % female 74.3% −0.62 .54 Ethnicity, % non-Hispanic white 92.7% 1.36 .18 Sedative medications, % taking* 17.4% 0.30 .76 Cardiac disease, % of patients with† 10.1% 0.67 .52 Pulmonary disease, % patients with† 56.9% 0.40 .69 Neurologic disease, % patients with† 62.4% −0.35 .73 * SD = standard deviation; MMSE = Mini Mental Status Examination; PRN = as needed; CIRS-G = Cumulative Illness Rating Scale-Geriatrics. † Includes sedating antipsychotics, benzodiazepines, and benzodiazepine-like agents. ‡ Based on CIRS-G ratings. Table 2. Summary of Nighttime Variables From Simultaneous Pulse Oximetry, Observational Sleep Assessment Instrument (OSAI), and Wrist Actigraphy. Mean (SD) Range Pulse oximetry (N = 109) Total recording time (minutes) 445.4 (71.87) 182.7–610.9 Baseline oxygen saturation (%) 93.7% (2.3%) 84.8%–97.8% Mean heart rate (beats per minute) 67.8 (9.8) 31.2–96.2 Mean oxygen desaturation index (ODI) 7.0 (9.3) 0–52.9 OSAI (N = 109) Number of observations per night* 6.5 (1.7) 2–12 Discontinuity of breathing† 0.27 (0.47) 0–1.83 Discontinuity of chest movements† 0.27 (0.46) 0–2.4 Loudness of breathing‡ 1.4 (0.51) 1.0–3.0 % of observations with snoring 13.3% (25.1%) 0%–100% Breathing rate (per minute) 17.6 (3.5) 11.1–26.0 Wrist actigraphy (N = 71) Total sleep time (TST; in hours) 4.3 (2.1) 1.1–9.0 Percent sleep (TST/total monitoring time) 47.9% (25.4) 5%–100% Number of awakenings during recording 18.2 (8.4) 0–43 * Observations completed only when patient was asleep. † Total number of events during the 3-minute observation period, reported as mean number per hour. ‡ Scored as 1 (low) to 3 (high). Table 3. Relationship Between Observed Symptoms of Sleep Apnea (Observational Sleep Assessment Instrument) and Oxygen Desaturation Index (ODI). Correlation p value Number of observations per night* 0.036 .708 Discontinuity of breathing† −0.007§ .939 Discontinuity of chest movement† 0.140§ .148 Loudness of breathing‡ 0.284 .003 % of observations with snoring 0.126 .191 Breathing rate (per minute) −0.011 .914 * Observations completed only when patient asleep. † Total number of events during the 3-minute observation period, mean per hour. ‡ Scored as 1 (low) to 3 (high). § Non-parametric correlation coefficient (Spearman's rho). Supported by National Institute on Aging grant AG13885 (to Dr. Alessi), VA Greater Los Angeles Geriatric Research Education and Clinical Center, Veterans Administration Health Services Research & Development Associate Investigator Award (to Dr. Martin), and the Hartford/American Federation for Aging Research Medical Student Geriatrics Scholars Program (to Mr. Mory). Presented in part at the Annual Meeting of the American Geriatrics Society, May 2003, Baltimore, Maryland. REFERENCES 1 Bliwise DL. Sleep in normal aging and dementia [Review]. Sleep.1993;16:40-81. 2 Binkley SA, Mosher K. Prior light alters the circadian clock in the chick pineal gland. J Exp Zool.1984;232:551-556. 3 Phillips B, Ancoli-Israel S. Sleep disorders in the elderly. Sleep Med.2001;2:99-114. 4 Ancoli-Israel S. Epidemiology of sleep disorders. In: Roth T, Roehrs TA, eds. Clinics in Geriatric Medicine. Philadelphia: WB Saunders; 1989:347–362. 5 Silverberg DS, Oksenberg A, Iaina A. Sleep-related breathing disorders as a major cause of essential hypertension: fact or fiction? Curr Opin Psychiatry.1998;7:353-357. 6 Palomaki H, Partinen M, Juvela S, Kaste M. Snoring as a risk factor for sleep-related brain infarction. Stroke.1989;20:1311-1315. 7 Bliwise DL. Cognitive function and sleep disordered breathing in aging adults. In: Kuna ST, Remmers JE, Suratt PM, eds. Sleep and Respiration in Aging Adults. New York: Elsevier; 1991:237–244. 8 Ancoli-Israel S, Klauber MR, Butters N, Parker L, Kripke DF. Dementia in institutionalized elderly: relation to sleep apnea. J Am Geriatr Soc.1991;39:258-263. 9 Ancoli-Israel S, Kripke DF, Klauber MR, et al. Morbidity, mortality and sleep disordered breathing in community dwelling elderly. Sleep.1996;19:277-282. 10 Bader GG, Turesson K, Wallin A. Sleep-related breathing and movement disorders in healthy elderly and demented subjects. Dementia.1996;7:279-287. 11 Ancoli-Israel S, Kripke DF, Klauber MR, Mason WJ, Fell R, Kaplan O. Sleep disordered breathing in community-dwelling elderly. Sleep.1991;14:486-495. 12 Ancoli-Israel S, Klauber MR, Kripke DF, Parker L, Cobarrubias M. Sleep apnea in female patients in a nursing home: increased risk of mortality. Chest.1989;96:1054-1058. 13 American Academy of Sleep Medicine Task Force. Sleep-related breathing disorders in adults: recommendations for syndrome definition and measurement techniques in clinical research. Sleep.1999;22:667-689. 14 Netzer N, Eliasson AH, Netzer CM, Kristo DA. Overnight pulse oximetry for sleep-disordered breathing in adults: a review. Chest.2001;120:625-633. 15 Chesson AL, Ferber R, Fry JM, et al. The indications for polysomnography and related procedures. Sleep.1997;20:423-487. 16 Golpe R, Jimenez A, Carpizo R, Cifrian JM. Utility of home oximetry as a screening test for patients with moderate to severe symptoms of obstructive sleep apnea. Sleep.1999;22:932-937. 17 Cohen-Mansfield J, Waldhorn R, Werner P, Billig N. Validation of sleep observations in a nursing home. Sleep.1990;13:512-525. 18 Cohen-Mansfield J, Werner P, Marx MS. An observational study of agitation in agitated nursing home residents. Int Psychogeriatr.1989;1:153-165. 19 Ancoli-Israel S, Clopton P, Klauber MR, Fell R, Mason WJ. Use of wrist activity for monitoring sleep/wake in demented nursing home patients. Sleep.1997;20:24-27. 20 Folstein MF, Folstein SE, McHugh PR. Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res.1975;12:189-198. 21 Miller MD, Paradis CF, Houck PR, et al. Rating chronic medical illness burden in geropsychiatric practice and research: application of the Cumulative Illness Rating Scale. Psychiatry Res.1992;41:237-248. 22 Alessi CA, Schnelle JF, Maldague S, et al. Total incidence and costs of acute medical conditions in long-stay, incontinent nursing home residents. J Am Med Dir Assoc.2002;3:229-242. 23 SPSS for Windows. Version 10.1. Chicago: SPSS, 2000.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0AE88CADBDCDFCE37B0DEF85CBC138F76052ADB1.txt b/test/dataset/in/resources/corpus/Clean_0AE88CADBDCDFCE37B0DEF85CBC138F76052ADB1.txt
deleted file mode 100644
index 6c8e7b6..0000000
--- a/test/dataset/in/resources/corpus/Clean_0AE88CADBDCDFCE37B0DEF85CBC138F76052ADB1.txt
+++ /dev/null
@@ -1 +0,0 @@
-frenchfrenchFrench History1477-45420269-1191Oxford University Press10.1093/fh/crp072ArticlesThe Parlement of Paris and the Ordinances of Blois (1579)DaubresseSylvie**Sylvie Daubresse is ingénieur de recherche at the CNRS in Paris. She may be contacted at sylvie.daubresse@culture.gouv.fr. This article has been translated by Mark Greengrass12200927102009234446466© The Author 2009. Published by Oxford University Press on behalf of the Society for the Study of French History. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org2009The Parlement of Paris invested considerable time and energy in detailed consideration of the substantial reforming legislation that had emerged from the first Estates General of Blois (1576–77) and which formed the eventual Ordinances of Blois (1579). Using the registers of the Parlement of Paris and hitherto unexamined copies of associated remonstrances, this article assesses why they did so, focusing on the issues of ecclesiastical and judicial reform. By placing their intervention in the context of the response from the Parisian magistrates to the holding of other Estates General in the sixteenth and early seventeenth century, it concludes that the sovereign court sought as much to uphold its exalted view of the ‘law of the realm’ and its own conception of reform, as to assert its independence from the Estates General or become a decisive intermediary in the dialogue between the crown and its subjects.Could the Parlement of Paris ever collaborate with the Estates General in the framing of reform legislation?1 On 23 November 1560, when two Parisian échevins asked the Parlement to despatch councillors to the Estates General shortly to be held at Orléans ‘pour adviser à deputer ceux qui seront et debvront estre envoyez de la part de lad. Ville’, the court replied in no uncertain terms that ‘[elle] n'en deputera point pour ce qu'il pourroit advenir que lad. Court congnoistroit par appel ou aultrement des choses appartenans aux doleances desd. Estats’. Sixteen years later, the Parlement continued to stand by its view and declared that ‘la forme ancienne accoustumée [doibt] estre gardée pour le faict des Estatz’.2 It refused to send deputies to the Estates on the grounds that it could not act as both judge and plaintiff.3On 2 August 1561, the Parlement was asked to validate the cahier that contained remonstrances from the Estates General held at Orléans from 13 December 1560 to 31 January 1561.4 This was the first time that doléances had been submitted for registration by the sovereign court. Where did such an initiative originate? It is clear that it arose within the Estates General itself, whose deputies thus subjected their acquiescence in royal demands for finance to the registration without modifications of the Ordinance of Orléans.5 They sought the Parlement's approbation for its measures, its juridical guarantee, not its opinion.The gens du roi, that is, the king's law officers, did nothing to hide their astonishment at this unusual way of proceeding. The written conclusions of the procureur général were drafted around the conviction that the judges of the Parlement had nothing to do with the Estates General and that they were ‘à part par consentement du roi’.6 This was also the view of Chancellor Michel de L'Hospital who, in his opening speech to the Estates of Orléans on 13 November 1560, established a clear distinction between the Parlements, which judge the private suits of individuals, and the Estates, which deliberate on the ‘plaintes generales qui concernoyent l’universel’.7 On 9 August 1561, Antoine de Navarre came before the court to explain that it was not required to consider the proposed ordinances article by article since they had already been deliberated upon by the Privy Council and by the Estates, of which the Parlement was a part. The judges were only required to ‘voir’ the proposed ordinances and ‘l'expedier’ rapidly.8 The premier president, Gilles Le Maistre, replied icily that the Parlement of Paris was not part of the Estates General, either as a corporate body or in any particular. Although two présidents and ten to twelve conseillers had been required to attend the Estates General of Tours in 1467 and 1484, its outcomes had not been forwarded to the Parlement of Paris.9 The premier président reminded Navarre that they would deliberate on the current Ordinances and agree their opinions on them in ‘la maniere accoustumee’. There would doubtless be some articles that would cause them no difficulty, but there might be others which ‘pourront ester trouver scismatiques et grandement doubteux’. The premier président Le Maistre was banned from exercising his office for four months as a reward for his pains, a highly unusual event in the history of the court.10 So the leading sovereign court of the realm was not prepared to rubber-stamp the ordinances, but expected to deliberate upon them and express its views. Herein lay the origins of the misunderstandings between the Estates General, the royal government and the Parlement of Paris, which would be replicated during the Estates General at Blois.Almost four years after the massacre of Saint Bartholomew, the edict of Beaulieu brought the fifth war of religion to a close, embodying the concessions obtained by the Malcontents, a coalition of princes and protestants under the leadership of Henri III's brother, François d'Alençon. Article 58 of the edict confirmed the summoning of the Estates General within six months. They were duly convened at Blois, meeting from 6 December 1576 to 5 March 1577, resulting in an ordinance whose articles were submitted to the Parlement for verification and registration two months after the Estates had been dismissed.11 From 28 February to 26 March 1579 the Parlement devoted seventeen sessions to examining the cahiers of Blois, and then a whole month to the drafting of remonstrances on those articles they regarded as controversial. The debate between the Parlement and the king was marked by various remonstances, reinforced by speeches from distinguished members of the court. The whole process of verification lasted almost twelve months despite all the efforts of the royal council to smooth over the various difficulties.12 Indeed, their bitter deliberations continued for several months after eventual registration in January 1580.Jacques-Auguste de Thou, a contemporary witness, was less than forthcoming, however, on what had transpired.13 Edouard Maugis, historian, furnishes no details whatsoever of these discussions.14 However, in contrast to the situation following the Estates General of Orléans of 1560, when there was next to no discussion because the ordinances were submitted so late, the registers of the Parlement of Paris enable us to follow those for the ordinances of Blois in detail.15 Even though we do not know precisely the text of the initial draft legislation submitted to the Parlement, this article demonstrates how the premier sovereign court of the realm attempted to intervene in the legislative process, following the efforts of the great estates of the kingdom to grapple with all the major issues of the day. These were so numerous, however, that we shall concentrate our attention upon the articles regarding the Church and those concerning judicial affairs.IOn the last day of February 1579, the officers of the king presented Henri III's replies to the cahiers of the Estates General of Blois before the Paris magistrates. The chambers assembled in joint session for the first reading of the edict and the nomination of a rapporteur. On 2 March, the first clause concerning the affairs of the Church was considered and, the following day (as was customary) the greffier Jean Du Tillet presented the resolutions of the court at the time of the verification of the ordinances of Orléans in 1561 and entrusted them to the rapporteur, conseiller Jacques Viole. Previous deliberations such as these always served as points of reference and the starting-point for subsequent discussion, constituting a kind of ‘juridical’ dossier upon which the magistrates could rely for precedents so that they could retain a degree of consistency of approach. On 26 March 1579, they decided to prepare remonstrances on certain articles, which were read out in draft on 6 May in the Grand'Chambre of the Parlement.16 The président, de Harlay, and three conseillers were nominated to present them to the king. They excluded comment on the proposed regulations concerning the substituts of the procureur général in article 157 of the draft Ordinances, because these were aimed exclusively at gens du roi.17These remonstrances of May 1579 exist in two extant copies.18 The first is a copy of the draft of the initial deliberations and it includes a note of the views expressed in the Parlement from which it is possible to reconstruct the spectrum of opinion.19 These varied from straightforward approval (noted with ‘bon’) for almost 60 per cent of the articles to various forms of reserve such as the petition (‘le roi sera supplié de …’), the explication (‘sera fait entendre au roi …’), or the request for further explanation or precision.20 The Parlement could propose to enlarge the scope of an article, limit it, or even completely redraft it.21 Or, again, it could decide upon formal remonstrances. Sometimes, it maintained a hard-line position, especially when it regarded a proposed article as impossible to carry out, in which case it expressed its view in a tone that brooked no dissent: ‘les anciennes ordonnances seront gardees’. The second document, also dated May 1579, is apparently a fair copy of the first, but with some notable differences in the opinions recorded.22 These manuscripts are entirely exceptional because the Parlement of Paris was normally scrupulous about maintaining the secrecy of its deliberations.23Throughout April 1579, the Parlement discussed articles concerning the Church.24 Whilst avoiding overt criticism of the Concordat of Bologna, it presented the Pragmatic Sanction of Bourges as a means of appeasing ‘the wrath of God’, rooting out heresies and bringing the troubles to a close.25 From the onset of the civil wars, the Parlement had tenaciously pressed for the residence of bishops in their dioceses and the confiscation of absentees’ revenues in favour of the poor.26 The court opposed the appointment of lay abbots and priors, and condemned the practice of laymen using custodi nos or prête-noms to acquire the revenues of benefices. Their position was, however, in contradiction with their own privilege of an indult by which they could become the titular holders of certain benefices and enjoy the proceeds.27 The magistrates of the Parlement equally pressed for ecclesiastical benefices to be denied to foreigners, fearing that the revenues from ecclesiastical offices would leave the kingdom.28 To reinforce the reestablishment of discipline in monasteries, the court sought to abolish their possibility of an appeal to the king in litigation involving regular clergy.29 The Parlement also targeted the suppression of économats.30 It set the minimum age for taking monastic vows at twenty years old and required female vows to be monitored to ensure that they involved no improper pressure.31 Finally, the judges wanted churchwardens to have to present their accounts before their parishioners and in the presence of the priest in charge.32 The Parlement clearly accepted the case for a measure of reform in the French Church.33The sovereign court was sympathetic to the idea of a general law prohibiting clandestine marriages and explicitly asked that there should be no further dispensations from publishing bans of marriage—something which had also figured in the canons of the Council of Trent. It associated itself, therefore, with one of the key decision of the Ordinances of Blois, by which marriages became public acts, to be celebrated following the publication of bans and in the presence of a priest in charge and four witnesses, upon pain of being declared null and void. The Parlement continued, however, to oppose the disinheritance of children who entered into marriage without the consent of their parents.On judicial affairs, the cahiers de doléances incessantly repeated the same message: justice was in poor shape, too slow, too expensive and open to improper influence among judges with family connections. Not altogether consonant with reality, this image allowed the king to introduce reforms into the judicial sphere, and especially into its procedures. It was also a good opportunity to emphasize the ideal of the king as source of justice. In reality, the first article on justice projected the image of a king ready to hold court in public and dispense justice in person, without judicial representation or procedures.34 The priority of the Parlement, for its part, was to safeguard the authority of its decisions. It did not take kindly to the fact that members of the royal council ‘prennent connaissance par dessus lesdicts arrests’. The basis for the rescinding (‘cassation’) or withdrawal (‘retractation’) of the court's decisions should follow the normal legal procedures, that is, by means of a civil suit and a legal error.35 Only in the second draft remonstrances, and then for the first time, did the Parlement refer to its decisions having been ‘cassez et retractez’, no doubt a reflection of a certain amount of irritation among judges regarding a practice which they deplored.36 Article 92 of the Ordinances of Blois sought to give them satisfaction on this issue and strictly limited the recourse to extraordinary legal appeal, over and above the ordinary civil suits and legal error.The cahiers of the three orders had been unanimous on the issue of cases removed from the jurisdiction of Parlement upon the authority of the king (évocations). They were a stain on justice.37 It allowed the king to interfere in a judicial process before a tribunal, remove it from its competence and either judge it himself or reassign it to another jurisdiction.38 The Parlement of Paris sought to prevent suitors from seeking évocations whose significance the Chancellor, René de Birague, had minimized in his speech to the Estates General of Blois, claiming that they had mostly occurred because of family connections and ‘menées’ among the magistrates themselves.39 It was in fact the case that these évocations were often connected to cases where there were legitimate grounds for suspicion that one of the parties to the case was related to the magistrates in the tribunal judging the case. An edict of François I had established a strict procedure that these évocations had to follow, according to which they were initially ‘deliberated upon’ by the Chancellor following a report from maîtres des requêtes.40 The Parlement concentrated its efforts therefore on ensuring that the Royal Council clarified its procedures and established clear rules.41 The Parlement of Paris wanted some limited participation from magistrates of the sovereign court in the sessions of the Royal Council. Those summoned to attend would only do so at the express command of the king and only to consider matters directly concerning the Parlement. If a magistrate proferred his advice to the Council he would not have the right to participate in the Parlement's deliberations on the same matter since, to do so, would make him judge and plaintiff in the same case. However, the king's law officers would not be invited to give their views before the Council. The Parlement set its face against the establishment of a corps or ‘college’ of magistrates from which the king might select those to serve on his Council, since he should be free to choose whomsoever he wished and without creating offices for them, ‘forme qui n’a jamais fin’.42 The Parlement had no desire to see its magistrates lured away from their ordinary jurisdiction on the pretext of prolonged service at court. On the contrary, they wanted magistrates to serve in their posts, a plea which fell upon deaf ears.The remonstrances of May 1579 reached the king at a date that we cannot determine precisely, but which coincided with the moment when the ordinance was ‘delivered’ by the king. On 20 June 1579, the sovereign court was still working its way through verifying the articles of the cahiers of the Estates General.43 On 23 June the king's law officers asked for remonstrances specifically in order to resolve the relationship between the Parlement and the Privy Council.44 On 27 June, the judges learnt that Henri III had reacted badly to their criticism.45 On 25 and 26 June the king had declared ‘with bitterness’ that he had no intention of surrendering the regalian right accorded him by the Concordat of Bologna to nominate to ecclesiastical benefices.46 For the deputies from the Parlement, however, royal authority was diminished rather than enhanced by the Concordat and they were utterly convinced that a return to the Pragmatic Sanction was desired by all his kingdom's subjects.47 For the king, by contrast, ‘ce ne seroit approuver les estats si l'on en parloit’.48 Henri III knew only too well that the issue of the right of clerical nomination had been debated at the Estates General of Orléans ‘pour les diversités qui estoit au royaume’, an allusion to the intense debates between Catholics and Protestants at that point on the reform of the Church.49 He refused to compromise on the right of nomination, which he did not want to sacrifice, but he offered assurances that he would not nominate anyone unworthy of the post whilst letting it be known that elections would provoke ‘menees’.50 The declaration indicates the degree to which Henri III had been caught unawares by a debate that must have seemed to him like a rearguard action from a far-off battle. The apparent collective amnesia of the Parlement had led its members to forget what had once been a fundamental reality: the long judicial battles at the Parlement between candidates provided by Rome and those elected locally, battles that masked the essence of the issue, which was to choose the best candidate.51 Such incoherence barely disguised the discomfort of the Parlement towards the Concordat of Bologna. It was difficult for the judges to accuse the monarchy of being partly responsible for the abuses in clerical nominations. Was not their unthinking advocacy of the Pragmatic Sanction in reality nothing other than the only means they could find to exhort the king to choose prelates worthy of their office?On the issue of the Privy Council, Henri III never regarded it as constituting an entirely separate ‘corps’. With notable firmness, he added that he ‘ne se vouloit brider et n'estoit a la cour de limiter sa puissance’. Throughout this bitter exchange of views, the Parlement could claim no victory, but equally it showed no sign of bringing the process of verifying the articles of the proposed ordinance to an early close. On the contrary, it sought to reinforce its stance, and on 3 July the magistrates decided to present their remonstrances once more before the king without substantial modifications.IIThe second set of remonstrances on the cahier of the Estates General, dated 6 July, was more incisive.52 If the Parlement was exigent it was because its members accorded the matter grave importance and sought to react to remarks made by the king on 26 June, to several deputies of the Parlement, to the effect that he wanted to know the names of those who stubbornly maintained views opposed to his own. The king threatened to come before the Parlement, and ‘sçauroit bien ceux qui sont en cause pour en faire ce qu'il faut’. The Parlement sought to defend the freedom of its deliberations which had always been respected by his predecessors. The sovereign court referred to an ordinance of December 1566 authorizing the Parlements to make reiterated remonstrances, a measure that suppressed article 2 of the Ordinance of Moulins (1566) which aimed at forbidding successive remonstrances. It was in utter ‘sincerity’ that the Parlement presented the king with what it thought to be for the good of his service and the reestablishment of his state ‘que l'on voit etre en mauvais ordre’.On the issue of the Church, the Parlement became more direct and precise. Its remonstrances insisted on the importance of the choice of good prelates, the only means as they saw it of staunching heresy and bringing civil war to an end. This was their justification for earnestly seeking a return to the system of clerical election, albeit admitting that the king could make recommendations in certain cases.53 The Concordat of Bologna had resulted inevitably in the triumph of favour over merit. They noted that the payment of annates to Rome, regarded as contrary to the ‘droits et constitutions canoniques et ordonnances anciennes’ and to the councils of the Church to which the pope ought to adhere, was nowhere to be found in the Concordat.54 Deploying the financial argument, they argued that this was money drained out of the kingdom.On justice, the second remonstrances returned to the respect that was owed to the decisions of the Parlement, the ‘guarantee’ of all property. The ‘law of the realm’—the term was employed by the judges for the first time in the process of verifying the articles of the cahier—required that judicial cases be heard in first instance before prévôts, baillis and seneschals, and in further instance before the Parlement, ‘and not before others’. It was impossible ‘d'attribuer par souveraineté la justice aux Parlements et par concurrence de cette souveraineté au Conseil privé du roy’.55 Such a ‘law of the realm’ could not be changed without provoking confusion and disorder in the ‘orders of justice’. The président Séguier, the Parlement's spokesman, went even further and asserted that the king had promised to maintain the old order of the judicial system at his coronation.56 Such rules could only be modified if there was a clear advantage to be derived from doing so, he added. It is tempting to interpret this as a desire to establish a rule of law over and above the king, but, in reality, it was more the application of the fruits of proven historical experience that respect for judicial hierarchy was the best guarantee of good justice. Séguier's argument was one from reason and experience.The Privy Council, which had instituted the greffiers criminels, should refuse to hear the suits of private individuals because ‘il n'y a loix, ny ordonnance en ce royaume qui donne puissance et autorité aux gens de sondit Conseil de connoitre des causes des personnes privées et entreprendre jurisdiction contentieuse sur les sujets du roy …’.57 The decisions of the Parlement were issued in the name of the king. If they were set aside, it was tantamount to the king ‘weakening himself’ the remonstrances deduced with impeccable logic. Upon the maintenance of the ‘choses jugées’ depended the ‘security’ of all the wealth in the kingdom, its peace and the tranquillity of the state. The risks from such initiatives were clear: the ruin and subversion of the state. The position was entirely coherent from a juridical and institutional point of view, although it conveniently ignored the fact that the king was the source of all justice.The remonstrances reminded the king that the Grand Conseil had been instigated by Charles VIII and that its role was to hear cases concerning the ‘limites des différends des Parlements’. Thereafter, however, its competence had increased with the result that it judged all sorts of suits upon évocations from the Parlements, as well as judicial tribunals in first instance, which resulted in intolerable travelling expenses for suitors.58 The remonstrances of 5 May 1579 asked the king to see that article 37 of the Ordinance of Orléans was upheld: ‘Les gens tenans nostre Grand Conseil ne connoistront desormais et ne pourront entreprendre la jurisdiction d'autres matieres et causes que de celles qui leur sont attribuées par leur creation et institution.’59 Furthermore, they recalled :Il n'y a eu lois, ny ordonnance concernant le pouvoir dudit Grand Conseil qui ayant été publiez et veriffiez audit Parlement et qui etoit et est requis par les anciennes ordonnances, et ne suffit la publication faitte par devant eux qui n'ont aucun teritoire, ne juridiction sinon limitte comme dit est. Toutefois il se trouve qu'aud. Grand Conseil les arrestz donnez esd. Parlemens, juges ordinaires et naturels y sont facilement cassez et renversez en quoy ils sont soutenu et appuyez de ceux qui voudroient s'ils pouvoient aneantir les jurisdictions ordinaires pour avoir juges a leurs volontez dont souvent a esté fait plainte au roy et a ses predecesseurs.60As in the earlier remonstrances of May 1579, the Parlement sought the suppression of the recently established commercial magistrates (juge-consuls), reformulating an argument that had already been deployed when they were created, to the effect that the king had commuted his justice to individuals without either the necessary knowledge or experience. In addition, this innovation would encourage gentlemen and other subjects of the king to seek their own judicial tribunals, not to mention the risks of collusion among merchants. In sum, the initiative risked bringing ‘schisme et division’ among the king's subjects. The views of the Parlement (and the third estate) on this subject were almost completely ignored since articles 239 and 241 of the Ordinances of Blois maintained consular justice in commercial towns. Elsewhere, however, commercial suits between merchants were assigned to the ordinary judicial institutions with the intention of limiting the proliferation of lawsuits.The increased authority of the prévôts des maréchaux was also called into question by the Parlement. These provosts had power and jurisdiction over certain crimes and misdemeanours in last resort and without appeal, a situation that the sovereign court magistrates regarded as a source of abuses and injustice. The provosts being charged with apprehending suspected criminals, it was dangerous that they should also be their judges without possibility of appeal. Their cases should be despatched by the ordinary local judges and, if the baillis and seneschals were in residence, then the provosts, along with the vice-baillis and lieutenants de robe courte served no purpose. Finally, the Parlement also opposed the pardons obtained by the prévôt de l'hôtel, acting again in final instance, for individuals ‘craignant la severité et rigueur des justices ordinaires’. The provost of the household had ‘pris un nom bien haut de grand prevot, qui n'a esté recue, ne publié au Parlement …’.61The Parlement approved, however, the reinstitution of the ordinances of François I, known as the Ordinances of Bourdaisière (May 1529) and Chanteloup (March 1546). These had laid down the regulations to be followed in respect of évocations, with the aim of preventing improper collaborations between relatives and clients in one and the same Parlement.62 However, article 118 in the proposed Ordinances of Blois required conseillers and présidents, who had a lawsuit involving even their distant relatives, to bring the case before another Parlement with the resulting absence from court and ‘punitive’ travelling costs. The magistrates of the sovereign court reminded the king that their privilege of commitimus (du petit scel, i.e. their right to bring actions in first instance before the chamber of requests of the Parlement) was ‘le plus ancient privilege qu'ils ayent’. If constrained to defend their case before another Parlement this would be completely contrary to their privileges.In the eyes of the Parlement, all these proposed changes risked compromising respect for justice: ‘Si tout le peuple de ce royaume pouvoir parler par une voix, il feroit cette requeste au roy que la justice fut remise en son premier etat et naturel et que les membres tirés et separés du corps fussent réunis et raportez a iceluy.’ At the end of the sixteenth century, to reform was not to innovate. This stubborn defence of sovereign justice was based on attachment to the old order in a period when justice was the preserve of the king, or his representative, the Parlement. Each jurisdiction should, here as elsewhere, keep to its limits and, as with royal power itself, the jurisdiction of the Parlement should itself be better defined. However, in highlighting the various encroachments upon its authority, the Parlement was also demonstrating its own weakness and the fact that it felt under threat. To present itself as the ‘voice’ of the people was perhaps a better means of making its voice heard in the dialogue between king and subjects.Despite these various different views, royal letters of constraint (lettres de jussion) of 15 July 1579 commanded the Parlement to register the cahiers ‘sans rien changer ny innover’. On 18 July, président Pierre I Séguier delivered a speech before the king which summarized the remonstrances of the Parlement. Remonstrances, when reinforced by the rhetoric of a président, might serve more effectively to persuade the king to change his mind.IIIPiety and Justice ‘are the twin columns which sustain the weight of your crown’, declared Séguier to Henri III. If the king maintained them both, God would be favourably disposed towards him: ‘Vous sçavez, Sire, les exemples si frequens en l'Escripture sur la plaincte et clameur des subjectz, Dieu se courrouce aux roys, mais ce ne sera pas à vous, si luy plaist, car nous tenons certain que vous avez souvenance du support et soulaigement de vostre pauvre peuple.’63 He insisted on the necessity of the church being made up of ‘gens bien suffisans’, fervent in piety and models of virtue for the king's subjects. Séguier then summarized the history of the provision of benefices, starting with the primitive Church, a period in which there had been no interference from kings or popes. It had been precisely because Charles VII had introduced the decrees of the Council of Basel (1431–49) into the realm by the ordinance known as the Pragmatic Sanction (1438) that God had favoured the king of France in his war against the English. Only a general council of the Church, according to the Parlement, was in a position to resolve the differences between Catholics and Protestants. The pope's authority was not above that of a general council of the Church, and he therefore lacked the authority to conclude a ‘concordat’ with the French king. On the question of the in commendam royal nomination to abbeys, Séguier made somewhat curious use of the canon of the Council of Trent condemning the choice of laymen as heads of monastic houses.64 It was a matter on which he invited the comments of the members of the Royal Council but, he continued, they could not change the fact that the Pragmatic Sanction had brought nothing but good to the kingdom, and the Concordat the reverse.The Parlement, taking careful note of the report of this encounter, categorically refused to publish the Ordinances and explicitly decided to lodge a copy of these remonstrances in its registers as a ‘perpetuelle memoyre à ce que la posterité congnoisse que le Parlement s'est mis en son debvoir’.65 This was its familiar way of protesting when it was urged. The remonstrances do not, in fact, feature in its registers although, as we have seen, we have the manuscript copies of them surviving from elsewhere. On 3 September 1579, Henri III (acting through the gens du roi and the présidents) renewed his demand that the Parlement publish those articles which they had already considered and approved before they broke for their annual recess. The Parlement noted, however, the large number of absentees when it reconvened to continue its deliberations on 5 September and refused to ‘divide’ the publication of the articles in that fashion. It was remarkably hard to stop the legal machinery from taking its course once it had begun, and on 7 September the sovereign court was instructed to proceed with its work of verifying the articles, even though it was close to the recess. That same day, the duc de Nevers appeared before the Parlement to confirm the king's will in the matter, the court replying that it ‘a faict ce qu'elle a peu et deu durant la sceance du Parlement …’. In the days that followed, a chambre des vacations (a nominated bench of judges to sit through a judicial recess) was constituted despite the king's orders. Whether Henri III had changed his mind or not, we do not know.Returning from its vacation in November, the Parlement went back to work verifying the cahiers of Blois. On 11 December 1579, conseiller Philippe de Lenoncourt came before the Parlement to remind the judges of the ‘dommage indicible’ caused by the delay in their publication and to read out to them articles 2 and 5 from the cahiers of the provincial estates of Normandy.66 The king's law officers added that the estates had the impression that the delay was caused by the king, but the discussion ended inconclusively: ‘la cour a vacqué sur ce qu'elle avoit a faire sur ce sujet jusqu'a la sonnerie de l'heure’. The following day, it was finally decided to conclude the reading of the cahiers and publish them ‘aprés plusieurs remontrances faites tant de vive voix que par escript’, but without signalling approval for the Concordat (of Bologna) and without accepting the encroachment on the authority of the baillis and the Parlements that the Ordinances implied.67Henri III rejected these reservations, as he explicitly made clear to the premier président de Thou when he was summoned for an audience with the king on 13 January 1580. The king accepted that the reservations be noted in the court's register, but not that they be included in the legal registration, whose publication had to be ‘pure et simple’.68 The king required the court to publish the cahiers otherwise he would be obliged to ‘faire chose dont il seroyt marry’ – in other words, to impose his will by force. The président Bellièvre intervened in order to defend the king's position, describing the objections made to the Privy Council as ‘chicanery’. The discussions continued on 16 January and, on 19 January, the king once more pressed the court to register the ordinances without qualification through the offices of président de Harlay. On 20 January, however, the Parlement decided to prepare further remonstrances.69 Then, three days later, after hearing about personal missives (lettres closes) sent to the premier président, whose content was never revealed, the sovereign court decided that the cahiers should be published without reservation on the following Monday, but that remonstrances be presented on the need to choose the right people to head monastic houses, and that their deliberations be registered along with their remonstrances and proposed modifications of some of the articles.70 Was this a purely formal disposition? Apparently resigned to an inevitable outcome, the Parlement's real concern was the judgement of posterity.71 Yet this was not the end of the matter because, at least until the measure was printed, discussions on the Ordinances of Blois continued.On 28 January 1580, Christophe de Thou was summoned to an audience with the king.72 Three royal valets waylaid him en route for the Louvre to tell him that the king was in no mood to compromise. De Thou outlined the different stages in the verification of an edict to Henri III, emphasizing the necessity of respecting its processes and deliberations. The Parlement never decided anything without considering it at great length. If the king did not want to take account of the remonstrances prepared by the Parlement, the royal letters had to include the customary formula: ‘fait sur le très commandement du roi’. The attempt by Michel de L'Hospital to curtail reiterated remonstrances had never been enacted and the Parlement was free to make as many protests as it saw fit. De Thou expressed the regret that the Chancellor de L'Hospital would have experienced in seeing the Parlement's rights of remonstrance restricted in that manner. By way of conclusion, the premier président returned to the main bone of contention between the Parlement and the king, namely the nomination to ecclesiastical benefices, and expressed the wish that those in future which did not conform to established norms be declared null and void. At this point, the king expressed his approval for de Thou's suggestion. The cahiers, de Thou concluded, contained some good measures and they needed to be put into practice. The Parlement took no pleasure in opposing the king's will. It acted in accordance with his conscience. De Thou reported back to the Parlement that the king had not interrupted him but listened without saying a word. The royal response was ‘pleine de modestie’. He took all that de Thou said in good part. His objective was to see the Ordinances which had been placed before Parlement obeyed, especially those concerning religion and justice. He would not tolerate any infringement of the jurisdictions of baillis, seneschals and parlements and he was firmly resolved to restrict as much as possible the granting of évocations. The king could afford to appear conciliatory since the act had been registered ‘pur et simple’, but he conceded nothing but vague promises by way of response to the Parlement.Thanks to the nuncio Dandino we know that, during the following month (February 1580), the king instructed the Royal Council to examine the Parlement's concerns on ecclesiastical benefices.73 On 1 March 1580, the président Prévost and several conseillers were once more despatched with further remonstrances on the cahiers of the Estates General, arising from the decision to register them on 23 January. It was a purely symbolic act: the Parlement continued to proffer its advice, even though its views were ignored. On 9 March, some further ‘modifications’ were proposed to some of the articles, even though they had just been published without reservation.74 In reality, it was a matter of yet more remonstrances in which the court returned to almost all the points it had previously raised. On the ‘regulation’ of justice, these ‘modifications’ concerned two essential points: to define the role of the Royal Council vis-à-vis the Parlement; and to ensure that the ordinances of Chanteloup and La Bourdaisière were applied and, with them, that the privileges of members of the Parlement were respected when they were in legal contention with members of their own family.75It was with considerable satisfaction, then, that magistrates of the Parlement received article 91 of the eventual Ordinances of Blois limiting the processes of évocation, returning to the Parlements appeals pending before the Royal Council and preventing it from judging such cases in contentious litigation in the future.76 In March 1581, however, they demanded still more, namely the complete annulment of all such contentious litigation still pending before the Council.77 The problem of the privileges of parlementaires who were in litigation with their relatives had clearly still not been settled by then since the court decided to draft new remonstrances about it. Article 97 of the Ordinances of Blois, which forbad actions of évocation de propre mouvement instigated by the king himself, seems to have been a response to one of the demands of the Estates General.78 In one of his letters, the lawyer Etienne Pasquier praised the king's decision on the matter.79 His view was that évocations were unknown in France before the time of Charles VI, which was when the duc de Bourgogne sought to divert some cases to satisfy his clients. Speaking from personal experience, he remembered that, at the beginning of his legal practice in 1549, someone who had the audacity to ask for the diversion of a case from the Parlement on the grounds of letters of evocation was cut short in his tracks and condemned to a fine for his pains. In the Pourparlers du Prince, Pasquier wrote that the sovereign court always reserved to itself the right to remonstrate to the king against évocations de propre mouvement, since all legal actions ‘doivent ‘s'accorder à raison’.80 Otherwise, ‘favourites’ would construct law from their own passions out of self-interest.The issue of évocations de propre mouvement did not disappear with the Ordinances of Blois and, even in the eighteenth century, their use remained one that was strictly controlled. What was the real impact of the Ordinances upon legal practice? The decisions of the Privy Council in Henri III's reign are difficult to classify but there are singularly few that rescind those of the Paris sovereign court and very large numbers which return cases to the Parlement of Paris, a sign of their sensitivity to the concerns of the Parisian senior magistrates as much as their respect for judicial procedure. This observation should, however, be qualified by the fact that there are substantial gaps in our evidence.81 In the immediate aftermath of the publication of the Ordinances various reactions emerged, including the evident disapproval of the papacy.82 In its remonstrances, the Parlement spoke up for the stability of an order willed by God in which everyone implicitly had their place. Its enduring concern was to reinforce royal authority. Since royalty and justice were one and the same thing, to defend the unity of justice was to defend royal sovereignty, it argued. The king's desire, by contrast, was to rectify the abuses of judicial style and practice by ever more technical legislation with the objective of speeding up the delivery of justice. The sovereign court wanted the king to lay down the details of legal procedure, but it also wanted to preserve its power of remonstrance.83 In a general sense, this led to a certain consensus of views.84 So, for example, the Parlement could only welcome the prohibition upon extraordinary commissions (article 98), and equally approve the articles which laid down a summary procedures in justice so that petty cases involving sums under three écus could be despatched quickly (articles 153 and 154). On 25 January, Pierre de l'Estoile noted ironically in his diary:fust publié en la Cour de Parlement l'Edit fait et arresté après longue deliberation de la Cour sur les cahiers des Estats tenus à Blois, en l'an 1577, auquel y a beaucoup de belles et bonnes ordonnances: lesquelles, s'il plaisoit à Dieu et au roy qu'elles fussent bien observées, tous les Estats et peuple de France en seroient grandement soulagés et satisfaits, mais est à craindre qu'on en die, comme de l'Edit des Estats d'Orleans et de toutes autres bonnes ordonnances faites en France: Apres trois jours, non vallables’.85For Jacques-Auguste de Thou it was ‘le malheur des tems’ that prevented numerous laws from being enacted.IVAccording to Edouard Maugis, behind a possible secret complicity between the Estates General and the Parlement to secure their ascendancy there lay an eventual victory for the Parlement, distinct from the Estates General and with a jurisdiction superior to it, as demonstrated in its right to revise and verify the ordinances that resulted from the cahiers de doléances of the Estates.86 It would be unwise, of course, to underestimate the rivalry between these bodies (permanently constituted or not), each wedded to its own specificity and role.87 But if we put to one side this notion of a perpetual confrontation, what do we find? On the one hand, the deputies of the Estates General petitioned the crown to respect the rights of verification of laws in the Parlements, but on the other to reject meddling in the cahiers that resulted from their petitions.88 In these circumstances, why did the crown ask for the advice of the Parlement when it was a body from which juridical cautiousness was only to be expected? On the other hand, the Parlement customarily proferred its opinions when asked to comment on all sorts of royal edicts. By the end of the sixteenth century, the verifying of the cahiers of the Estates by the Parlement had become commonly accepted if we accept the view of conseiller Guillaume Du Vair, who vigorously asserted that the decisions of the Estates General would not have the force of law unless they were verified by the Parlement.89 Yet, as we have seen from the preceding analysis, verification was a mere formality when it was a matter of purely and simply publishing the measure as it was presented to them. In reality, the Parlement of Paris made its voice heard, but it did not seek to intervene decisively in the dialogue between the crown and its subjects. Or, at least, it only sought to do so in the more labyrinthine secret discussions that took place between its procureur général and the king.90In 1588, during the session of a subsequent Estates General, the magistrates of the Parlement were asked to present their own particular cahier of remonstrances.91 The king wanted them to establish the main lines of discussion to be treated in the Estates. Following custom, however, the Parlement submitted its remonstrances only after the Estates General had met. No doubt Henri III wanted to use the Parlement as a means of moderating the views of the Estates General, whose deputies were in large majority members of the catholic League. Not surprisingly, the remonstrances which it drew up in August 1588 were based on those prepared for preceding assemblies of the Estates General. Then, in 1593, in a completely different political context, the Parlement solemnly refused to depute any of its members to attend the Estates General of the League.92 The Parlement verified edicts only following the king's instruction to do so. In that respect the court was a faithful instrument of royal authority. Unlike the situation in 1560–61, when his brother had been forced by financial necessity to do so, nothing required Henri III to ask his sovereign court for its views on the cahiers. In inviting the Parlement to express an opinion, was not the French monarchy providing it with the opportunity to seize the initiative? Little by little, the court was inclined to become a more independent agent in its relationships with both the Estates General and the king. So, to look ahead, to 1615, four days after the closing session of the Estates General on 23 February, the various tribunals of the Parlement met in general session along with the princes and the peers of the realm to solemnly inform the king of the wretched state of the government of the realm. To justify this bold and unprecedented initiative, the Parlement of Paris used the promise that the king had made not to respond to the cahiers of the Estates General without taking account of its views.93 The Parlement no longer wanted to be treated separately and it no longer patiently waited for the royal command before it intervened. This change in attitude represented a profound shift in its relations with royal authority.1A[rchives] N[ationales] X1A 1596, fo. 20v (23 Nov. 1560).2AN X1A 1653, fo. 239v (6 Sept. 1576).3This had been its view since 1413: G. F. Denault, ‘The legitimation of the Parlement of Paris and the Estates General of France, 1560–1614’ (PhD Thesis, Washington University, 1975), 243, cited by A. E. Bakos, ‘Meddling chaperons: the involvement of the Parlement of Paris in the Estates General of 1593’, in idem, Politics, Ideology and the Law in Early Modern Europe (Rochester, 1994), 93.4AN X1A 1598, fo. 117v (2 Aug. 1561).5On 4 Aug. 1561, the sieur de Selve, a conseiller in the Privy Council, came before the Parlement to explain that the king was awaiting a definitive response from the Estates General, currently meeting at Pontoise, to his demands for money. However, this response would only be forthcoming when the publication of the Ordinance had been completed. De Selve was charged ‘d'alléguer [to the Parlement] que comme comprinse ausd. estatz, lad. Court en deust estre’.6Lalourcé and Duval (eds), Recueil de pièces originales et authentiques concernant la tenue des Etats généraux, 9 vols (Paris, 1789), i. 330. Guillaume de Taix, clerical deputy at the Estates of Blois in 1576–7, expressed the same conviction: ibid., ii. 124.7L. Petris, La Plume et la tribune. Michel de l'Hospital et ses discours (1559–1562) (Geneva, 2002), 386. Among the jurists, opinions were more divided. Jean Papon considered that the king could not proceed to alienate the royal domain without the consent of the Estates General. If they were not in session, that duty fell to the Parlement of Paris which, unlike the Estates, sat in permanent session and regularly offered counsel to the king: Secrets du troisième et dernier notaire (Lyon, 1578), 318. Philibert Bugnyon, however, thought that ordinances which had been deliberated in plenary sessions of the Estates General did not need to be submitted to the Parlement for registration: Commentaires sur les ordonnances faictes par le roy Charles neufiesme en sa ville de Moulins au mois de fevrier mil cinq cens soixante six (Lyon, 1567), 17, cited in A. Rousselet-Pimon, Le Chancelier et la loi au XVIe siècle d'après l'œuvre d'Antoine Duprat, de Guillaume Poyet et de Michel de L'Hospital (Paris, 2005), 166.8AN X1A 1598, fo. 173r–v (9 Aug. 1561).9In 1467, the Parlement had deputed the premier président Boulanger and a dozen councillors to the Estates, but this group was sent ‘non comme un corps mais afin de donner conseil’. In 1484, following royal instructions, the Parlement elected to send a deputation of four of its number but in fact none of those chosen ever left Paris: E. Maugis, Histoire du Parlement de Paris de l'avènement des Valois à la mort d'Henri IV, 3 vols (Paris, 1913–16), i. 656.10In reality, it was less the propositions of the premier président which had attracted this censure, but rather the Parlement's delay in ratifying the cahier of the Estates General of Orléans just at the moment when the royal government was in discussion with the Estates’ representatives at Pontoise in August 1561: S. Daubresse, Le Parlement de Paris ou la voix de la Raison (Geneva, 2005), 253.11According to Maugis, the delay was part of the monarchy's strategy of playing a waiting game to obliterate the memory of the Estates General of Blois: Histoire du Parlement de Paris, i. 671. In reality, Henri III was above all preoccupied to reestablish peace, because the edict of Beaulieu had been the prelude to renewed conflict. It was only thanks to pressure from the provincial estates and efficiency of the new chancellor, Hurault de Cheverny, that the new Ordinances were eventually promulgated.12M. Greengrass, Governing Passions: Peace and Reform in the French Kingdom, 1576–1585 (Oxford, 2007), 267.13J.-A. de Thou, Histoire universelle de Jacques-Auguste de Thou, depuis 1543 jusqu'en 1607, traduite sur l'édition latine de Londres (London, 1734), vol. 7, 74–5. He devoted only a few lines to the subject.14Maugis, Histoire du Parlement de Paris, i. 671: ‘Inutile d'entrer dans le détail qui remplirait un chapitre’.15On 2 Aug. 1561, the Parlement received the text of the ordinances of Orléans. On 30 Aug. their draft remonstrances to the king were read out: AN X1A 1598, fo. 295r–v. Thereafter nothing is recorded until 13 Sept. 1561, the date when the ordinances were registered.16AN X1A 1663 – sessions of 5, 6, 7, 9, 10–14, 16, 17, 19, 24 and 26 Mar. 1579. On 2 Mar. it was decided that two hours per day would be devoted to the processes of verification.17AN X1A 1664 fos 32v and 33, (6 May 1579). On article 157, the magistrates noted: ‘Sera fait registre a part et ne sera employé en la remonstrance que l'on fera au roy’. This article required ‘a nos procureurs et avocats generaux de prendre le moindre nombre de substituts qu'il leur sera possible’. It prohibited these substituts from accepting payments from suitors in return for furnishing requêtes, informations, interrogatoires or other trial documents. Guy Coquille regarded the proposal as inconsistent since it could not be right to ask substituts to serve without any payments at all: Oeuvres de Me Guy Coquille, sieur de Romenay (Paris, 1646), 512.18The two texts also exist in copies to be found in AN U 768 fos 265–327.19B[ibliothèque] N[ationale] MS Fr 2703 fos 204v–209, ‘Arrestez de la cour sur le cahier des Estats generaux faits avant les remonstrances de l'Estat ecclesiastique en l'an 1579’. The assembly of the clergy, meeting with the king's consent at Melun, began its deliberations on 26 June 1579 with an examination of the complaints about the abuses which had made their appearance in the French church.20In the latter, a straightforward ‘bon’ is inscribed against 202 of the 363 articles. We should note that a third of these concerned judicial affairs, and that there is a mismatch between the number of articles that figure in the remonstrances and those that appear in the eventual ordinances, which makes following the remonstrances somewhat difficult. This was because six articles were eventually struck out at the king's command and removed from the final published legislation.21Concerning article 97: ‘au lieu de l'article ce qui ensuit sera mis: Nous voulons ….’: ibid., fos 210r–v.).22BN MS Fr 4398 fos 337–339v, ‘Remonstrances faictes au roy par sa cour de Parlement sur l’édit fait sur leurs cahiers généraux du 5 may 1579’. Concerning article 5, the court ‘persiste aux remonstrances qu’elle a faictes sur le premier et deuxieme article ….’ (my emphasis). In other respects, this copy begins with a more developed and formal preamble in praise of the king's bountiful goodness and his desire to reform corruptions and abuses, and includes a reference to the imperative necessity of securing obedience to the law. This would be the version of the remonstrances presented to the king: ibid., fo. 337v.23The deliberations of the Parlement are never transcribed into its parchment registers: F. Hildesheimer, ‘Exemplaire Parlement’, in ‘Fabrique des archives, fabrique de l’histoire’, Revue de Synthèse, 125 (2004), 49–51. The deliberations of governmental organizations and courts of justice always took place in secret.24The papal nuncio Dandino, abreast of the latest news, told the Cardinal de Côme in a letter of 12 Apr. 1579 that the Parlement was in the course of deliberating on its responses to the clauses from the Estates of Blois on ecclesiastical matters: I. Cloulas (ed.), Correspondance du nonce en France, Anselmo Dandino, 1579–1581 (Rome, 1970), 367. From his letters we know that it was the Keeper of the Seals René de Birague who kept the nuncio informed on the discussions in the Parlement: ibid., 424, 15 June 1579.25BN MS Fr 2703 fo. 204r article 1.26Article 14 of the Ordinances of Blois enforced this obligation upon absentees on pain of their being deprived of the revenues of their sees.27The indult allowed conseillers of the Parlement to hold ecclesiastical benefices (if they were conseillers-clercs) or present a candidate of their choice (if they were laymen). Conseillers-clercs were given dispensation from residing in their benefices.28This was a demand that had already been voiced by the procureur général in his conclusions upon the cahiers of the Estates General of Jan. 1561: Lalourcé and Duval, Recueil, i. 333.29This issue was eventually dropped and did not figure in article 30 of the eventual Ordinances.30An économat was instituted in the six months of interregnum in a benefice prior to the issuing of a bull of institution to the next incumbent, during which period its revenues were handed over to an économe appointed by the king.31BN MS Fr 2703, fo. 206, article 28, in which those entering religious orders could dispose of their possessions up to three months after their final vow. The clause seems to reflect almost exactly the draft proposed by the Parlement save that the minimum age was set at sixteen years, a choice reflecting the demands of the clerical deputies. Cf. the clause in the same article concerning female vows: ‘et outre les abbesses et superieurs auparavant que faire bailler aux filles les habits de professes pour les recevoir a profession seront tenus un mois devant avertir l’evesque, son vicaire ou superieur de l’ordre pour s’enquerir par eux et informer de la volonté des dictes filles, s’il y a eu contraincte et induction de leur faire entendre la qualité du veu qu’ils vont faire auparavant qu’elles s’obligent’.32Their proposal was that they should be required to do so within three months of leaving office (article 53) but this was eventually dropped from the final version of the Ordinances.33According to Jean-Marie Le Gall, the Parlement had received numerous appeals early in the sixteenth century from those in the religious orders hostile to reform (‘déformés’), but its decisions were generally in favour of those who sought reform: Les Moines au temps des réformes, 1480–1560 (Seyssel, 2001), 116. Megan Armstrong's study indicates that the Parlement sought to limit the number of appeals from ecclesiastical to royal justice (appels comme d’abus), often pursued by those in monastic orders who regarded themselves as victims of decisions that had gone against them: ‘Spiritual reform, mendicant autonomy, and state formation: French Franciscan disputes before the Parlement of Paris, 1500–1600’, Fr Hist Stud, 25 (2002), 505–30.34See article 89 of the Ordinances of Blois. Jacques Krynen has recently drawn our attention to Jean Bodin's notable reticence about this aspect of royal justice, expressed in the Six livres de la République which had appeared three years earlier in 1576 (Book IV, ch. 6). Bodin invited the king not to seek to deprive magistrates of their authority on the grounds that he (the king) was not a jurist. Coquille, however, gave the article of the Ordinances of Blois his unqualified approval: Oeuvres, 493.35BN MS Fr 2703 fo. 209v. Suitors who had not received satisfaction from legal decisions arrived at by a parlement sought to demonstrate that the court had committed an error and to adduce the grounds for their view. If the latter were found to have weight, the king rescinded the decision of the court without questioning the substantive issues of the case. The Royal Council only considered the legal grounds for an appeal and referred the case back to the Parlement for a new hearing of the evidence. The point is an important one since the rescinding of judicial decisions on such grounds was also used by the magistrates themselves, admittedly in the exceptional circumstances of the split of the Parlement of Paris into two tribunals in the League between 1589 and 1594, when it was a device used by each court against the other: S. Daubresse, ‘De Paris à Tours, le Parlement “du roi” face au Parlement ”de la Ligue” (1589–1590)’, in Le Parlement en exil ou Histoire politique et judiciaire des translations du parlement de Paris (XVe–XVIIIe siècle), ed. S. Daubresse, M. Morgat-Bonnet and I. Storez-Brancourt (Paris, 2007), 466–70.36BN MS Fr 4398, fo. 344r.37G. M. R. Picot, Histoire des Etats Généraux considérés au point de vue de leur influence sur le gouvernement de la France de 1355 à 1614 (Paris, 1872), vol. 2, 568.38A. Rigaudière, Introduction historique à l’étude du droit et des institutions (Paris, 2006), 581. An évocation could occur following a decision by the king himself, or it could be the result of an action instigated by one of the parties to the suit.39Lalourcé and Duval , Recueil, ii. 63–4.40The report of the maîtres des requêtes was made a requirement by the edict of La Bourdaisière of May 1529. The Parlement derived the oversight of evocations back to an ordinance of Charles VI of 15 May 1389 as well as another of Louis XII of 22 Dec. 1499: BN MS Fr 2703 fo. 210v and, above all, the entry on ‘évocation’ in the Encyclopédie of Diderot and d’Alembert.41Even under the reign of Louis XIV, the Parlement was still trying to win this battle, as Albert N. Hamscher demonstrates in The Conseil Privé and the Parlements in the Age of Louis XIV: A Study in French Absolutism (Philadelphia, 1987), 80 and 107.42It is notable that the Parlement of Paris never offered a word of advice on the overall composition of the Royal Council.43According to the nuncio Dandino's correspondence with the Cardinal de Côme, their deliberations on ecclesiastical matters continued throughout May: Correspondance, 383, letter of 4 May 1579; and 386–7, letter of 10 May 1579. On 13 May, Dandino had word of the initial responses of the Parlement on ecclesiastical affairs.44AN X1A 1664 fo. 292, (23 June 1579). There were some discussions between the king and the members of the prosecutor’s office, the latter declaring that the king would enter into discussion with the papacy over the issue of papal annates.45AN X1A 1664, fos 338v–339v, (27 June 1579).46On 21 June, nuncio Dandino reported the identical view of the king, only too well aware of how much the authority to nominate to benefices mattered to him, in a letter to the Cardinal de Côme, a letter in which he totally rejected the magistrates’ views. According to the nuncio, the judges of the Parlement were seeking to poison the good relationship between the king and the papacy.47The Parlement relied on the fifteenth-century councils of the Church, which forbad the obtaining of benefices for money (‘c’estoit trafiquer’) as well as the ordinances of St Louis and Philip the Fair. We should note that the Parlement conflated the right of nomination with the problem of benefice provision. The return to election to benefices was one of the demands of the peasants studied by J.-M. Constant, ‘Le langage politique paysans en 1576: les cahiers de doléances des bailliages de Chartres et de Troyes’, in Représentation et vouloir politiques autour des Etats généraux de 1614, ed. R. Chartier and D. Richet (Paris, 1982), 31, although, as the author explains, they hoped to be able to elect their own clergy.48In their cahiers the third estate and the clergy had demanded a return to clerical elections. The nobility, however, proposed a commission of inquiry before the election of a new prelate, a suggestion that was seized upon by the Royal Council, since it incarnated the notions of the Concordat of 1516: Picot, Histoire des Etats Généraux, 393–4.49O. Christin, Les Réformes: Luther, Calvin et les protestants (Paris, 2007), 23–4.50According to nuncio Dandino, the clergy had obtained assurances from the king that he would not in future allow benefices in commendam and that he would pay no more annates to Rome: Correspondance, 469, Dandino to Côme, 8 Aug. 1579.51V. Julerot, ‘Y a un grant desordre’: élections épiscopales et schismes diocésains en France sous Charles VIII (Paris, 2006), 429–30. She notes the inefficacy of elections, the mismatch between theory and practice, and emphasizes the problem of confirming the result of a clerical election.52BN MS Fr 4398 fos 367–394v, ‘Les secondes remontrances sur les cahiers des Etats generaux, du 6 juillet 1579’, signed de Thou and Viole. The articles are not enumerated.53BN MS Fr 4398 fo. 373. Ever since the Pragmatic Sanction, the king made his recommendation to the ‘elisans’ and ‘celuy qu’il plaisoit au roi etoit ordinairement gratifié et preféré s’il etoit capable et digne’. In his reply to the cahiers of the Estates General of Orléans of 1561, the procureur général then in post had proposed that episcopal elections should involve the archbishop, the cathedral chapter and ‘douze des principaux habitans et bourgeois desdites villes qui seront élus en l’hostel d’icelle par les maire et echevins, consuls et conseillers desdites villes’. The candidate then elected would, of course, be presented to the king.54For président Pierre Séguier, annates were contrary to the word of God and human law: AN X1A 1665 fo. 13, (18 July 1579).55BN MS Fr 4398 fo. 377v. This was to create an ‘incompatible’ jurisidictional conflict and confusion, declared président Séguier on 18 July 1579: AN X1A 1665 fo. 14.56AN X1A 1665 fo. 14, (18 July 1579): ‘Sire, les loyx civilles et politicques dient qu’il ne fault poinct changer ne innover les reigles par lesquelles par le passé ont esté trouvees bonne synon que l’on trouvast au change ung bien et commodité publicq … Sire, l’ordre que dessus est l’ancien de vostre royaulme que vous avez promis a vostre sacre garder et entretenir.’57To Guy Coquille it was evident that the Privy Council was instituted to ‘connoître des affaires generales, non pour les affaires des particuliers’. It was made up of princes and other seigneurs who lacked knowledge of laws, customs and legal conventions: Oeuvres, i. 493.58The third estate equally complained that the Grand Conseil encroached upon the Parlements: Picot, Histoire des Etats Généraux, 567.59The Parlement noted that the letters patent creating offices were addressed to the Grand Conseil ‘chose directement contraire a l’institution et establissement dudit Parlement, ce qui apporte toute confusion a la justice’.60BN MS Fr 4398 fo. 388.61BN MS Fr 4398 fo. 385v, ‘Le Parlement suplie le roy puisqu’il luy plaist remettre la justice en sa splandeur et premier estat pourvoir sur ce fait que les appellations dud. Prevost de l’Hostel retourneront au Parlement.’62F. A. Isambert et al. (eds), Recueil général des anciennes lois françaises depuis 420 jusqu’à la Révolution de 1789, 29 vols (Paris, 1822–33), xii. 312–15 and 908–9.63AN X1A 1665 fos 12–15, (18 July 1579). The président also presented the king with a ‘plus ample’ memorandum from the Parlement.64AN X1A 1665 fo. 13v, ‘La commende en terme de droict commun ne vault rien, et mesme par le concile de Trente, elle est damnee et resprouvee’. The in commendam involved the collation of an ecclesiastical benefice to a cleric who was not in regular orders, or even to a layman, and the Concordat had conceded to the papacy the in commendam right: D. Richet, ‘Une famille de robe: les Séguier avant le chancelier’, in idem, De la Réforme à la Révolution: études sur la France moderne (Paris, 1991), 230.65AN X1A 1665 fo. 14v (18 July 1579).66AN X1A 1666 fo. 110v (11 Dec. 1579).67AN X1A 1666 fo. 114v–115 (12 Dec. 1579).68AN X1A 1666 fo. 237 (15 Jan. 1580). Nuncio Dandino reported the king's insistence on the matter to the Cardinal de Côme although the Parlement, he added, continued to press for the restoration of elections to ecclesiastical benefices and a return to the Pragmatic Sanction: Correspondance, 564.69AN X1A 1666 fo. 274, (20 Jan. 1580).70AN X1A 1666 fo. 281v, (23 Jan. 1580). The injunction on the registration simply reads: ‘Leues, publiees et registrees oy le procureur general du roy après plusieurs deliberations et remonstrances tres humbles faictes audict seigneur’: AN X1A 8635 fo. 184v.71As Maugis noted, Histoire du Parlement, i. 671, the Parlement was much preoccupied by ‘opinion’.72AN X1A 1666 fos 312–314v, (28 Jan. 1580).73Letter of 17 Feb. 1580, in Correspondance, 601.74BN MS Fr 4398 fos 466–471, ‘Modifications sur le cahier des estats de Blois du neuf mars mil cinq cens quatre vingt’, signed de Thou and Viole. After the Estates General of Orléans, equally, the printing of the cahier was delayed because the Parlement sought to make yet further modifications: AN X1A 1598 fo. 412, (27 Sept. 1561).75Article 117. The Parlement wanted this article modified.76Isambert et al., Recueil général, xiv. 404. According to Guy Coquille, it was an issue pursued by the Paris deputies, who ‘firent grande instance pour l’obtention de cet article’: Oeuvres, i. 493.77BN MS Fr 4398 fos 467r–v, ‘Sur le quatre vingt unziesme, le roy sera suplié de declarer toutes et chacunes les procedures qui seront faites en son Conseil nulles és causes de jurisdiction contentieuse et que les parlemens n’auront esgard a cela’.78If a case was to be evoked out of a jurisdiction, it had to be with good cause and the subject of a report prepared by the masters of requests (the Parlement also wanted oversight by the chancellor or Keeper of the Seals, assisted by the masters of requests). A signature of the secretary of state was also required by article 70 of the Ordinances of Moulins (1566).79D. Thickett (ed.), Lettres familières (Geneva, 1974), 71–5, to René Choppin, avocat at the Parlement of Paris.80E. Pasquier, Pourparlers, ed. B. Sayhi-Périgot (Geneva, 1995), 102.81F. Dumont (ed.), Inventaire des arrêts du Conseil privé (règnes de Henri III et de Henri IV) (Paris, 1969–71), vol. 2. The inventory only covers the July quarter of each year, beginning in 1579 and ending in August 1588. I am grateful to Solange Bertheau for allowing me to consult her subject catalogue before it becomes publicly available. Of the 1774 entries for the reign of Henri III in the inventory, there are three arrêts of the Parlement rescinded (‘cassés’), three others which forbad the Parlement from considering a suit, five which order the closure of proceedings in a suit, three that countermanded a decision of the court on the grounds of legal error, and only one resulting from a civil action against its decisions. There are eight evocations of suits to the Conseil privé, Grand Conseil or other jurisdictions. By contrast, seventy-six cases were sent back to the Parlement of Paris, sometimes after an attempt at an évocation (even a case sent before the Grand Conseil was returned to the Parlement: ibid., i. 34, no. 240 [12 Aug. 1580]).82Commentaries of Coquille in Oeuvres, 462–9; also ‘l’Advis de M. P. Pithou, advocat en Parlement sur l’ordonnance de Blois de 1576’, in A. Loisel, Divers opuscules des mémoires de M. Antoine Loisel, advocat en Parlement (Paris, 1652), 345–52. Cf. V. Martin, Le Gallicanisme et la Réforme catholique: essai historique sur l’introduction en France des décrets du Concile de Trente, 1563–1615 (Paris, 1919), 170. The papacy had great difficulty in accepting a measure in which the king was placed in a position of absolute superiority over bishops. It demanded its complete revision, sending a new nuncio to put its case.83L. de Carbonnières, ‘Le style du Parlement de Paris et la législation royale (XIVe–XVe siècles)’, in Modèles français: enjeux politiques et élaboration des grands textes de procédure en Europe: les enjeux politiques de la cofidication de la procédure (Paris, 2008), vol. 2, 171–93. According to Carbonnières, neither in the preamble, nor in the articles of the Ordinances of Montils-lès-Tours (1446), nor those of 1454, was there any reference made to the ‘style’ of the Parlement (i.e. the legal rules governing its procedures), not even in terms of ways of modifying it. The same was true of the Ordinances of Blois. Jean-Louis Thireau also noted that the legislative initiatives of the sixteenth century ‘fait entrer pleinement la procédure dans le domaine de la loi …’: ibid., 211.84They came up against difficulties on article 149, which forbad appeals on grounds other than procedural ‘si ce n’est pour le vuider et juger sur le champ’. In its comments on the majority of the other articles, the Parlement wanted to alter the wording, which had some juridical significance but did not change the overall legislative intent.85P. de l’Estoile, Registre-journal du règne de Henri III, ed. M. Lazard and G. Schrenck (Geneva, 1992–), vol. 3, 91–2.86Maugis, Histoire du Parlement, i. 670. Rousselet-Pimon thinks that the pretensions of the Parlement to interfere in the royal legislation, proposed in cooperation with the estates general, represent a much greater menace to royal power in the sixteenth century than that of the estates general themselves: Le Chancelier et la loi au XVIe siècle, 166.87Pierre de La Place, premier président in the Cour des Aides, thought the Parlement was not required to verify the decisions of the Estates concerning royal impositions: Commentaires de l’Estat de la Religion et Republique soubs les rois Henryy et Francois seconds, et Charles neufiesme (n.p., 1565), 157. In this respect, he was protected by the prerogatives of his own jurisdiction.88Picot, Histoire des Etats Généraux, 561–2: the third estate reminded the crown that ‘de tout temps et par l’institution de la France, nul édit ne doit être reconnu pour édit au préjudice des anciennes lois et ordonnances de France, s’il n’est premièrement vérifié par les cours souveraines’. In Nov. 1576, Blanchefort, the noble deputy expressed his desire to see the king nominate those members of his Council who would negotiate the ordinances with the Estates and then submit them to Parlement, simply for their approbation: Greengrass, Governing Passions, 90. In Dec. 1576, the deputies of the clergy wondered, in the course of a meeting with the other order, whether the reforms demanded by the three orders would be automatically adopted by the king and submitted to the Parlement of Paris as a ‘loi inviolable’: ibid., p. 92. The Parlement was here viewed above all as a guardian of the laws.89In 1593, Guillaume du Vair wrote in his Suasion: ‘veu que ce qui a accoustumé de se résoudre aux Etats généraux de la France bien et légitimement assemblez n’a force ny vigueur qu’après qu’il a esté verifié par vous séans au throsne des rois, au lict de leur justice, en la cour des pairs’, cited in Maugis, Histoire du Parlement, i. 673.90In Dec. 1576, the procureur général Jean de La Guesle presented two documents on ecclesiastical and judicial affairs before the Estates General in the name of the king. These were a project for reform whose origin had lain in a text prepared by the duc d’Anjou (the future Henri III) for the Council of Charles IX on the eve of Anjou's departure for Poland: Greengrass, Governing Passions, 93–6. Such ‘particular’ remonstrances have not been the focus of this article. In a letter of 22 Feb. 1580, nuncio Dandino reported that the responses to the Estates of Blois were in the hands of the procureur général for final modifications before being printed: Correspondance, 605. It was surely at that point that the six articles, struck out on the king's orders (four of which were about judicial matters), were removed: Isambert et al. Recueil général, xiv. 462–3). They concerned issues of legal procedures and were perhaps ‘ostés par commandement du roy’ at the request of the procureur général. They were not the object of any remonstrances by the Parlement itself.91AN X1A 1711 fos 81r–v, (9 Aug. 1588). In a letter of 6 Aug. 1588, Henri III asked the Parlement to submit a memorandum that he intended to present to the Estates General.92Maugis, Histoire du Parlement, ii. 100. Five parlementaires attended the Estates General in 1593, but that was because they were deputed from the third estate of Paris.93E. Glasson, Le Parlement de Paris, son role politique depuis Charles VII jusqu’à la Révolution (Paris, 1901), 123–4. Marie de Médicis issued an arrêt de cassation to rescind the arrêt of the Parlement which had summoned the tribunals and peers to meet in common assembly.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0AEBE99971E44E4AA329E1B8B8EF05C82EAB3BC7.txt b/test/dataset/in/resources/corpus/Clean_0AEBE99971E44E4AA329E1B8B8EF05C82EAB3BC7.txt
deleted file mode 100644
index 56a27a9..0000000
--- a/test/dataset/in/resources/corpus/Clean_0AEBE99971E44E4AA329E1B8B8EF05C82EAB3BC7.txt
+++ /dev/null
@@ -1 +0,0 @@
-ohrohrThe Oral History Review1533-85920094-0798Oxford University Press10.1093/ohr/ohn058ARTICLESSerbian Gypsy Narrative: Between Preferred and True IdentityČvorovićJelenaJelena Čvorović (Ph.D., 2001, Anthropology, Arizona State University, AZ, US) is a long-time student of Gypsies in Serbia. During the past several years, she has authored many articles dealing with various aspects of Gypsies and their traditions. Jelena's strong interests lie in the area of human evolution and behaviorWinter-Spring200917320093614570© The Author 2009. Published by Oxford University Press on behalf of the Oral History Association. All rights reserved. For permissions, Please e-mail: journals.permissions@oxfordjournals.org2009This paper discusses the narration of a Serbian Gypsy who adopted Serbian ethnic identity. Still today in Serbia, Gypsy culture remains in the oral form. Their narratives tell as much about their present as about their past. Several themes underlying the discussion about Gypsy ethnic flexibility are explored: their position in relation to non-Gypsies, the way they perceive it; varying attitudes about ethnicity within Gypsy communities in contrast with personal experiences of a Gypsy who adopted Serbian identity; and the evolving nature of Gypsy identities.ethnicitycommunitiesoral historypreferred ethnic identitySerbian GypsiesIntroductionThis paper presents the voice of a Romanian Gypsy who declares himself to be “a true Serb.” The oral history presented comes from a Gypsy community in the village of Drenovac, in western Serbia's county of Mačva, and makes a contribution to understanding the flexibility of Gypsy ethnic identity. The account reveals that many Gypsy behaviors can be better appreciated by placing together the narration of local people with anthropological objective data.Gypsy communities in Serbia still rely on oral tradition as a source of obtaining knowledge and information on the history of their ancestors, proper kin behavior, economic life, the societies in which Gypsies live, relationships with non-Gypsies, and the everyday world.1 Many times, there are no written documents or historical records that enlarge our understanding of the Gypsy past and the choices they made. Even today, many Gypsies in Serbia are illiterate or semi-literate, but their oral tradition remains transmitted, as a form of cultural heritage, from one generation of kin to the next. Recording their oral histories provides one way of breaking this silence and allows Gypsies, otherwise marginalized on all levels, to speak for themselves about issues important to them.The following interview with a Romanian Gypsy, narratives, and interview quotes I collected in several Mačva villages provide an opportunity to explore several themes related to Gypsy ethnic flexibility: (1) the position of Gypsies in relation to non-Gypsies, and how they perceive this relationship; 2) varying attitudes about ethnicity within Gypsy communities in contrast to the personal experiences of a Gypsy who adopted a Serbian identity; and 3) the evolving nature of Gypsy identity.BackgroundRoma/Gypsies are a diverse ethnic group probably of northern Indian origin, scattered throughout Europe since arriving west of the Balkans in the fifteenth century.2 It is unclear why they left India, and there are no explanatory written documents. From linguistic influences preserved in all Romani dialects, it is most probable that the major Gypsy migration routes passed through Persia, Armenia, Greece, and the Slavic-speaking parts of the Balkans.3 A number of early European historical sources refer to the Gypsies as Egyptians, and the term “Gypsy” is thought to mirror that assumption. According to historical sources, the Gypsies themselves spread the legend about their Egyptian origin; they represented themselves as dukes, kings, and princes from Egypt.4 The inventive process of Gypsy traditions and favored identity in Europe had begun. Roma/Gypsies had begun their sojourn in Europe by taking advantage of the Christian piety of the age. Since it was a Christian obligation to help pilgrims, especially the ones with documents of recommendation from rulers, the Gypsies created letters of passage from high government officials such as King Sigismund of Hungary, representing themselves as penitents for the sins of their ancestors who had rejected Christian teachings in Egypt. As a result of the sins of their ancestors, they were required to wander the earth as pilgrims seeking charity from others.5Today linguistic and historical data, supported by new genetic studies, suggest that the European Gypsies, embracing a large number of socially different endogamous groups, may be a complex conglomerate of founder populations that originated in India.6 Genetic results suggest a limited number of related founders, compatible with a small group of migrants splitting from a distinct caste or tribal group. Individual Gypsy groups can be classified into major metagroups: the Gypsies of East European origin; the Sinti in Germany and Manouches in France and Catalonia; the Kaló in Spain, Ciganos in Portugal and Gitans of southern France; and the Romanichals of Britain.7 Out of these, the greatest variety is found in the Balkans, where numerous groups with well-defined social boundaries exist.8The present study of favored ethnic identity cannot be separated from the wider study of the history of society. In spite of their presence in Europe for centuries, Gypsy integration into European society is poor, and their history is a true story of exclusion and discrimination. Gypsy hunting and other such persecutions have occurred almost from the beginning of the Gypsy presence in Europe, with a peak during World War II.9Today the position of Gypsies in Central and Eastern Europe is more or less the same.10 In Central and Eastern Europe, the policy of socialist governments was to assimilate Gypsies and for decades they were targeted for low-skilled employment within a centrally planned economy.11 Efforts were made to help Gypsies to settle down and to improve their economic and cultural position in the postwar years. This resulted in the banning of nomadism in most of the countries in the region and the destruction of traditional Gypsy occupations.The collapse of the Eastern European socialist system in 1989 was followed by a long and still incomplete period of transition in Serbia. In Serbia, the transition started at a time when the economy was near or at the point of collapse, and as a result impoverishment and unemployment rose significantly.12 Protected to some extent by the socialist regime, Gypsies have been perhaps the most affected by the transition to a market economy because of their lack of a qualified labor force. Thus they have high levels of unemployment, substandard housing, a lack of education and skills, and a deepening dependence on state benefits and services.13 The general situation in Serbia contributed to the present day Gypsy condition. Serbia's economy has been degraded by a decade of war, isolations, and sanctions. On the other hand, due to the assimilation processes and manipulation of their ethnicity by the Roma themselves, many Roma today deny their Roma ancestry.Firstly, not all Roma people refer to themselves as Roma. Some prefer the European term cigani or zigeuner. Secondly, some European groups/tribes of Roma reject altogether the connection with Roma people in general, like the Sinti and Manouches.14 This is perhaps particularly true for Serbian Roma, where many Roma refer to themselves as “smoked Serbs” or simply “Serbs.”Today, Gypsies are perhaps the most segregated ethnic group in Europe. The same is probably true for Serbia.15 The first written document referring to Gypsies in Serbia dates from 1348.16 In Serbia, as in other South Slavic countries under Turkish rule, Gypsies constituted a separate ethnic group; they lived apart in mahalas, in towns or in isolated village areas. In the Balkans, through centuries of Turkish rule, Gypsies were strictly endogamous: even the best men at their weddings were always fellow Gypsies. In the past the Gypsy social unit was the extended family and the occupational niches they filled contributed to the Serbian economy. The traditional Gypsy occupations include making crafts such as troughs, baskets, spoons, blacksmithing, ironsmithing, and music. Despite their contribution to the agriculture-based Serbian economy, Gypsies were despised by the rural Serbian population.17 In Serbia, Gypsies form a complex mixture of groups.18 In fact, there are quite a few subgroups of Gypsies in Serbia.Some of them lost the Romani language and their mother tongue is now Serbian. There are different forms of Romani depending on which group the Gypsy belongs to. Interaction between different groups is limited, and the form of Romani spoken has become an important means of distinguishing between groups. Furthermore, Gypsies have always come under several appellations and names. Each group represents a historical and, for the most part, originally localized entity. Many times Gypsies adopted their hosts’ culture in response to the different requirements of their social and environmental surroundings. The result is a great diversity of Gypsy tribes and a lack of identity as an integrated ethnic group. Therefore, Gypsy culture in general is extremely diverse and difficult to pinpoint. Their ethnicity is also a disputed and complex issue due to the fact that most Gypsies do not consider themselves members of a unified group but identify instead with the subgroup to which they belong.19 Within these subgroups, language and religion also remain diverse; the religion which a Gypsy tribe or ethnicity might hold on to depends on location and circumstances: they adopt whatever religion is prevalent in the local area. A characteristic of Gypsies as a group is their adaptability to religious and political changes. Their religious and political standings depended always on the current political climate. For example, some contemporary Christian Gypsies are ex-Muslims. Until very recently, the characteristic of all groups was that they did not mix with each other—there appeared to be a strongly emphasized antagonism among the groups. A system that divides these groups, much like a caste system, still exists in some parts of Serbia today. In the past, with respect to other Gypsies, allowable marriage choices were largely restricted. Females in particular were expected to marry someone within their particular tribe and most obeyed the rule by marrying within their group.Another characteristic of Gypsy ethnic identity in Serbia is the so-called “ethnic mimicry” or “favored identity” by which Gypsies declare themselves to be members of a majority community. According to a 2005 census, there were 114 000 Roma/Gypsy in Serbia, which is less than 1.4%. However, an official estimate of Serbia's true Gypsies range between 360 000 and 500 000 compared to 7 478 820 Serbs.20Yet another variant of the Gypsy “favored identity” is the construction of the new ethnic identities and new traditions. Among Gypsies, these new identities and traditions assume to restore an ancient origin in spite of the complexity to attest its legality and even obvious contradictions to historical records. A notorious example of the invented tradition and change in the identity of the ethnic group usually called Gypsies is the emergence of the Egyptian community in Serbia and Kosovo.21Alternatively, some Gypsies always retained a strong ethnic attitude. For example, the president of one local Gypsy party in the village of Macvanski Pricinovici claims, “We just got a status of national minority … Some asked for it, and got it. I haven't. We were Cigani before here, now we are Roma. I don't care. I'm a Serbian Ciganin and I'll stay Ciganin forever …”22 One informant from the same village, who declares himself as Roma, argues, “The word Rom means a man, but for me it actually means that you should not be ashamed to say that you are a Rom/cigani. On the other hand, we say that we are Roma, but we live together without peasants [Serbs], we have the same customs and the same religion, and the only thing that differentiates us is our face color!” Another male informant expressed it this way: “What makes us the Roma here? We still have our tradition, that we are Roma, and our name. It's our blood that makes us Roma: if my mother and father are Roma, I'm Roma too. We've stayed ethnically pure because we've held on to our tradition and managed to maintain it and we still stick to it.” These varying attitudes about ethnicity within Gypsy communities themselves reveal distinctions between the groups.FieldworkFieldwork and the interview were conducted in the village of Drenovac, the second largest village in Mačva. Approximately 10 000 Gypsies inhabit Mačva, an agricultural area in western Serbia. There are at least seven tribes/subgroups of Gypsies in Mačva; most tribes were Muslims in the past, except for Christian Romanian Gypsies.23 The Gypsy permanent settlement in Mačva was established in two stages, during the late nineteenth century and the ‘50s, when their sedentarism was enforced by the Serbian government. Usually Gypsy villages in Mačva are relatively poor, and typically there is no running water or a sewage system. The majority of the Gypsy population in Mačva is semieducated, averaging 5.36 years of schooling; few own land and most make a living as hired field workers. Many also work in a black-market economy, engaging in trades of various sorts. Marriages are made at an early age, and having a large number of children is desirable.Drenovac is inhabited by Karavlax Gypsies (Black Vlax), or Romanian Gypsies, who arrived in Serbia during the eighteenth and nineteenth centuries.24 These Gypsies consider themselves to be “natives” of Serbia, rationalizing this statement by the fact that their ancestors were born and raised in Drenovac. They identify themselves as Lejaši, from the Romanian word laiesi. The majority of Gypsies in Romania had been slaves in the middle ages25 and were divided into groups, serving the Crown, noblemen, or households. The laiesi were the slave group, comprised mostly of musicians. After the abolition of slavery in 1864, many migrated to Serbia.Approximately 10% of the houses in Drenovac are occupied by Gypsies who are bilingual. They speak Serbian as their mother tongue, and Romanian as a second language, and Romani, or Gypsy language, is not spoken at all. They belong to the Orthodox Christian Church and all celebrate ancient, traditional Serbian family holidays. Music is their traditional occupation and they were never employed, as were other Gypsy groups, as craftsmen. Although they came from Romania to Serbia as part of the larger Karavlax Gypsies tribe, they deny any connection with Karavlaxs, based on the difference in occupation: Karavlax Gypsies in the area used to be spoon makers. They have a somewhat darker skin color than the rest of the Gypsies.26 During the 1970s and 1980s, the Lejaši traveled throughout former Yugoslavia playing their music. Today they cannot travel so easily due to the political isolation of Serbia born out of the postwar (1991–99) consequences and a strict visa regime. At present, most males work in a village factory owned by native Serbs, while the majority of females labor as housewives.The Lejaši regard themselves as more sophisticated than other Gypsies in the area due to the fact that they do not marry or form social relationships with other groups. Lasting social relationships are created with Serbs of their village. All enter into mutual godparenting relationships with Serbs, and all have traditional Serbian names which indicate Serbian ethnic origin. Most consider themselves to be “true Serbs.” Census data from 1948 to 1991 reveals that no Gypsy in Drenovac has ever declared that he/she was a Roma/Gypsy.Dragan Vasiljkovic, the president of the Roma Association of Western Serbia and a member of the National Roma Council, is a native of Drenovac and exceptional in many ways. He is one of the rare Gypsies who managed to finish his university education (only 2% of Gypsies in Serbia have higher education according to Statistics of Serbia, 2002). He lives in Sabac, the main town in Mačva, with his wife and two children who are both college students. Mr. Vasiljkovic works in a general hospital as a dentist and also operates a successful private practice. He credits his success to his wife's efforts and his upbringing, which was no different than “any Serbian child.” On the other hand, he fully accepts his Gypsy/Roma background and sees the Gypsy ethnic mimicry as a separatism that weakens the already loose unity of their national minority:This is a calculated manipulation in order to avoid and deny their Roma roots, combined with the personal interests. Even today, the Roma from Drenovac don't mix with other Roma groups. They never marry Roma from other groups and this is common to all Roma here, on Mačva territory. The data indicate that the first Roma in Drenovac were Roma immigrants from Romania. In the beginning, in the nineteenth century, there were only a few tribes living in the territory of Drenovac; they were named Jankovic, Vasiljkovic, Stankovic, and Jovanovic. They spoke only Romanian, not Gypsy Romany and they had strong solidarity and cooperation within the group. The solidarity was reflected in helping each other and they got along well. Their main characteristic was that they didn't mix with the other Roma tribes which surrounded the village of Drenovac, such as the Gypsies from Priciniovici, Sevarice and so on. It appears that language was the barrier which prevented them from mixing; and we also know that the other Roma groups didn't accept them as their own. However, there is no proof that they represented a different ethnic entity.27A totally different opinion is expressed by Dragan Buric, a 46-year-old musician, who states that he is a “true” Serb. Dragan Buric remained one of few who managed to engage in the Lejaši's traditional occupation as a source of living. Dragan has a band in which he plays with his son, while his wife sings. Hence, music remains their primary source of income. He has a comfortable house, a car, and a small bar that he operates with his wife, all of which separates him from the rest of his community. The family speaks Serbian as their first language, and Dragan does not know Romanian, or Romani, the Gypsy language. He says that he always felt as a Serb and has declared himself as such. Dragan's interview was conducted in his home during the summer months of 2002. His family was present most of the time and occasionally they would perform some of their music. Interviewer and interviewee engaged in “sharing food”—“social eating” and drinking local brandy, with frequent toasting. The story that Dragan tells testifies to the difficulty of establishing Gypsy identity and his attempt to rationalize the present social order. It provides important data for our understanding of the flexibility of Gypsy ethnic identity. Dragan's account is presented in four parts: childhood; occupation/music; ethnicity; and marriage and family. Through his narrative we can trace how he pursued a strategy worked out by his ancestors to negotiate ethnic and social identity.This is his story:ChildhoodI was born in Skela, near Obrenovac. We lived there until 1978; then we came here, to Drenovac, and we live here since then. My mom was born and raised in Drenovac; she was only married in Skela, to my father Sreten. My childhood was a happy one although our family was very poor. My mom and dad had three children: my sister, my brother and me. We are all approximately two years apart. My parents got along very well; they stayed together until death took them apart. They had what I called “an ideal” marriage; although, I have to say, if it weren't for my mom we wouldn't do anything with our lives. My mom, Nadezda, stayed with us most of the time and taught us how to behave and taught us to love school. My father was away most of the day, working. At night, he performed music. My old man was a musician, he played violin, and actually, my grandfather and his brothers were musicians, too. It runs in the family—I am a musician; I play violin, and my son inherited the talent for music, too. This is our tradition, and not only that, but we all have a good ear for music. At that time, when I was born, playing music didn't pay very well. I remember that we used to live in a poor house made of mud, 4x4m. We had only two beds. I slept with my mom because I was her youngest and her favorite child.My mom was a great mother to us and a hard worker. She used to go around the village and do all kinds of jobs—cleaning, washing, cooking. She would also go out to the fields from time to time. But she never begged, ever. Gypsy women do that—beg for food or money, but my mom never did that. She earned every dinar. She would go out into the fields to work for a whole day. Landlords usually gave workers some food, a breakfast, or something, a piece of meat … but my mom would never eat that. No, she brought home every piece of food for us kids to share. That's how she was. She also never allowed me to accompany her when she worked in the fields—I remember that I cried every time she would go out, but she never let me, for it was too hot for a child to sit in the sun on the open field the whole day.My father, also, used to take every job he could find, for he couldn't support us only with his violin. He was a strong, big, healthy man, and knew how to do a lot of things. He died two years ago, and my mom passed away in 1995. I am very glad that I was able to give them a different life—they sacrificed their whole lives for us kids, and when we moved here and I started to play around Yugoslavia and the neighboring countries—I provided for them, took care of them financially. Both of my parents deserved that, after what they'd been through, when we were little kids.I was born in 1957. The late ‘50s and 1960s were very difficult times. There was never enough food no matter how hard you worked or tried. Both of my parents worked, but I still remember the poverty around us, and that little house where we used to live. On the other hand, maybe those difficult life conditions made us all very attached to each other. My brother and sister and myself, we were always very close. We still can't live without each other: my brother lives in this same house, just has the entrance on the other side, and my sister is married into a house only a 100 meters away from us.We respected our parents—that's how we are brought up, not like the kids today. You give them one little finger, and that's not enough, no, they want the whole hand! We, the kids, all went to school, even if we were so poor. My mom insisted on our education. The nearest school was about 7 kilometers away from the place where we lived, and I had to take a bus every day to get to school. There was one part of the road that I needed to travel by foot, by some forest, and my mom would always wait for me there—during winter especially, to make sure I was safe. We also didn't have appropriate clothes—no winter shoes or gloves. But we were very fortunate, we had the best schoolteachers. My schoolteacher was especially fond of me. She didn't make any distinction between me and other kids—like I'm a Gypsy kid or something. On the contrary, she did all she could so we could enroll in school like every other kid. I'm still very grateful to her. I was a very good student, always. I never needed any special help to learn school things. I know there is a huge problem with Roma children today, in schools, but that's due to the Roma mentality; a lot depends on parents, on their attitudes, and how they advise their children. That's a real pity, a shame, for all those Roma children today, but that's the irresponsibility of their parents, nothing else. I realize that we are not the same, and we don't have the same capabilities. I know now that I have Roma origin, but I have always tried to behave properly. I raised my children, especially my daughter, like God said we all should. I tried to give my kids everything and to bring them up to be good people. I thank God that they didn't have to follow my footsteps and suffer.I have the primary education, I think it's like eight grades of elementary school. I am very sorry, even today, that I couldn't continue my schooling further. I still regret that today. But, after my eighth grade, I started to play with my father. When I was ten years old, he bought me my first harmonica; after six months, I mastered it completely, that's how talented I was for music and playing. That was a very different time than it is today. It was expected from me and my brother to learn how to play and to become musicians. After all, that's what our family, our grandfathers used to do. I have to say that I loved music from a very early age. My father taught me how to play, and I loved it. I was particularly good in playing kolo, our national [Serb] dance: I would play it and the whole room would get up to dance! That was exciting, especially since I was just a little boy. Some guests at weddings used to lift me up on their shoulders and I would play like that. I loved the attention I was getting and I knew I was good. On the other hand, I should have pursued schooling, but the circumstances were the way they were, and I neglected further schooling and became a musician. Who knows what could have happened if I continued my school? I was very good in school, always good with numbers and especially drawing. I could have been an architect if I had different life circumstances. Later on, when I grew up, I bought myself some private music lessons in Belgrade, in a music school, to specialize. I learned a lot, and today there is no melody or a song that I cannot pick up and play. Music is still what I make a living off today. Maybe I'm lucky in that sense—I did and still do what I like the most. There is something about the rhythm that can carry you away, and that's what has happened to me. My brother also dropped out of school, to play, although he wasn't such a talent as I was. My sister, on the other hand, sang like a nightingale; it was such a joy to listen to her! Well, the saddest story of my childhood is connected with her. I was always very attached to her; even though I was her younger brother, I felt a need to protect her, always, for she was the only girl in our family. That's how brothers are toward their sisters. And she was a good person, too. I remember this as if it happened yesterday. It was 1968. I was ten and a half years old, my sister was around thirteen. My father's brother took her with him to a shitty little place called Potoci, in Bosnia, to sing with his band. That place was behind God's back [remote, far away and forgotten by people], with one little furniture factory and a train that goes around once a week. And nothing else. When my uncle took her away, that broke my heart. I can still feel the pain of that separation today. She was just a little girl at the time; she even looked younger than her age. I was a kind of a sensation at ten, playing harmonica so good, and with my sister it was the same. I guess that's why my uncle took her. People would pay you more. He is my uncle, but I dislike what he did—he was using her. I don't know why my parents allowed that, maybe because we were always short on money, don't know. I think we were very good kids, respecting our parents and school and our schoolteacher. We had to obey, that's the key. I think that everything depends on where you were born; if I was born in Belgrade, who knows what I would become. I felt always that I could do whatever I wanted—no trouble in school at all, so talented and how, for music … it was just the life's conditions that determined which path to take, nothing else.My sister came home after three months; she didn't say much about the place, and the restaurant where she had a job, but she told us that she managed to arrange a job for all of us—my father, my brother and me, to come back with her and perform in that same place. It turned out that the landlady, the owner of that restaurant, came to like my sister very much, and when she heard that her whole family was into music she made the arrangement for all of us to come. That meant we were all going to make some money, thanks to my little sister! My father decided we should go, and bought us two new harmonicas, one for me, and one for my brother. Actually now when we were all back together, I was very excited about this trip. This was supposed to be my first trip, my first visit anywhere outside my village, and God knows what I expected. I had great expectations. I was just a ten-year-old kid with a harmonica, but also a musician who's going to make some serious money! I have to admit that it felt good, especially that my sister was back with us.So we left. We only had money to get to that damned place, nothing more, not even for a piece of bread. We traveled by train the whole day and my god it was slow! It had those old-fashioned narrow rail tracks and I thought we would never get to our destination. We had to take a connecting train in Mliniste, around 100 kilometers away from our final destination. It was already dark and getting cold. I remember I was very hungry and I started asking for some food, at least some bread. But my father explained that we would have to wait until we got to that restaurant where we were supposed to play—maybe the landlady would give us some food. He didn't have any money, not even a dinar. So, we were sitting in a waiting room at the station and it felt like we were beggars or something, and I couldn't handle that. I cried loud, asking for my mom and some food. That railway station was a horrible place: dirty, with some dim lights, and somehow everything looked old and beat up. Then my poor darling sister took off her sock, and there, in her elastics, she had two dinars that she had saved for “rainy days”, just like this one was. She bought us one kilogram of bread and I was very grateful; her gesture touched the bottom of my soul. She used that coin to tie/twist the elastics of her sock, and without it her sock kept falling down. My darling sis! I tell this story to my kids all the time. They can learn a lot from my life experience; don't know if they will ever use it.Finally we reached the place. I didn't like it. Who knows what I had expected to find, but this was one lousy, dirty place to be. At least our family was together. I have to say, the owner, the lady who liked my sister, she was nice. She was really nice to us. Her little restaurant was very primitive—no electric power, no heating, very modest all in all. She had put one little kerosene light on each table, and polyvinyl tablecloths. The guests were mostly factory workers; at that time, people were still uncivilized, and a restaurant with a live band was not a good place to be. But we were there and we started to perform every night. I also accompanied my sister and sang with her—my voice still hadn't changed so I had a voice like a little girl and our duets became, so to say, famous. People gathered to listen to us—and my old man was also very good. The thing that I didn't like at all was that my sister was exposed to all kinds of indecency. All female singers go through that sooner or later. Some guests like to touch them or to call them names or to behave too familiar with them. And my sister was just a little girl and very embarrassed. The first time a male guest touched her—I think he placed his hand on her—I started to cry from anger and shame. And I was supposed to be a “real musician”, but this was my sister and it was too difficult to handle. Later on I learned to deal with all kinds of situations and with people; that's what this profession is all about. We managed to earn some money and after a month we got back home, to our mother.Until the day he died, my father was always working on something. Even when he was old he wanted to help. He died in my son's arms, literally. He died from throat cancer, in a hospital. He was 67 years old when he died. I used to tell him that he should slow down, and enjoy his retirement days, but he couldn't stop. He was just very diligent and couldn't sit still. Always a fighter—that's what he was. He fought for us, to provide for us the best he could. I think he passed that down to me because I'm the same with my kids. I would do anything for them.MusicThe life of a musician can be very difficult. For the past 30 years, I go to bed in the morning, around five or seven in the morning, for I play the whole night. I sleep during the day—I get up around noon or so. The basic thing is, you depend on other people, your audience. Whatever they say or want, you learn that they have the right, always. That's the only way to survive in this profession. It took me a while to get that, but I mastered it and now I'm fine. I first played with my father, for years. Then I got a band of my own, something more modern. The band is called “Demons”; don't ask me why, my son named it. We had to adjust to current fashion, or to learn fashionable songs, always. My career was mostly good, in the sense that I don't remember that people harassed me because of my color—because I look like Roma. Maybe just a couple of times but it wasn't personal … people would say cigani to us or something. It was a different time than it is today, people didn't pay so much attention to nationalities then. Now it's different. Luckily, even when there was a potential conflict, like guests would get drunk and offend our female singer or us, I always managed to calm things down. I know how to behave and how to soothe things. But sometimes I would get personally offended when guests called me ciganin/Gypsy—then my soul hurt and I would think: Why am I not a Hungarian? Why call me a Gypsy? People love to say: “Play, Gypsies!” Or they would curse our Gypsy mother. That always hurt me the most. I don't know if there is such a thing as a “Gypsy man”—for me it looks like just one bad, ugly word. I don't like that word, and I don't like to be called one. It's a pejorative term—anybody, even an American or a German could be called a Gypsy if that person behaves bad. I feel very bad when I hear that word. I managed to stay calm, always, and learn to put up with different people, even if they are mafia. If I had behaved differently I would have never survived as a musician.I've played for politicians, for mafia, and for the greatest gentlemen. For one job, I would make around 15,000 German marks. I used to fish in Plitvicka Jezera [Lakes].My music took me to Austria, Linc, in 1989. We had a very good job over there. That was still a very good time for making money, for all of us. However, when the war started in 1991 I came back home. I still don't know why I came back, why I brought my family here. Now we can't travel—if we go to Belgrade to play, that's something. I wish that I could go around the country and play the way I used to. But that's impossible now. I was in Bosnia last year, for two months, to play. They also got it bad. Today it is very difficult to survive from playing music. It's especially difficult for me since my children are used to a good life, and now I can't give them everything the way they are used to. This is a great burden for me. I still feel responsible for them—even though my son is 21 years old; my daughter is 15 now. I don't know what will happen, I'm very concerned for the future. I have some 20 years of service as a musician; I paid for my retirement fund always, but that's not going to come up with very much when I retire. I am a member of the Independent Artists Organization of Serbia, but that's still no help. Before the war this organization used to arrange jobs for us, musicians, in Yugoslavia and abroad. Now they won't do anything; they just ask for more money and more money and we don't have an audience to play anymore. I think that our profession is dying out and I'm sad because of it. This was a nice tradition and a good life, especially when you are young.I keep a few chickens and a pig today so my children won't be hungry. After all, I live in a village and it is normal to have animals.EthnicityI am a Serb. I am a 100% original Serb. My father and his father and my great-grand father were all born in Serbia. They always behaved like Serbs. I'm the same way. I love my country very much. This is where I was born. I don't know where my ancestors came from, but I feel I belong here. Never in my life have I felt like a Gypsy, not even for a moment. I have the same soul [like the Serbs], the same blood, the same upbringing, the same everything. That's how I was raised. We were always Orthodox, always. My saint day, my slava is my Saint Arhandjel Mihajlo, we celebrate that. We have an icon, too. I always declared myself as a Serb, on all censuses. The same is true for my family; we all feel the same. My son, for example, he is a Serb too. We don't speak the Gypsy language, we don't even know one word of it. We only speak Serbian, and we know some Romanian, but it's a dialect with around 30% of Serbian words. We are different than the people in Macvancki Pricinovic, for example. They all speak the Gypsy language and we don't understand them. They are Roma, Gurbeti. The others are Roma, not us. They say they are Gypsies, always did. I'm always afraid that I will offend someone by calling him Gurbet or Roma; I almost apologize every time I have to say, “You, from Gurbeti or Roma tribe.” That's funny; they apologize to me sometimes, too, when they address me.We always had very good relations with our neighbors; my next door neighbor [a Serb] is my godparent. The godparenting runs in our families for about 60 years. His grandfather baptized my mother, and his son did the same with me, and now his children are godparents to my kids. My son now started a new godparenting, with his best friend [a Serb], but I respected our family tradition—my godfather is over 70 years old today. My godfather is very important to me; he is like my father, in the sense that I always respected him and he took care of me the best he could. His wife recently died and he's very old, but he managed to be at my grandson's baptism, and he was the most welcomed person there. That's just our tradition. It is a great honor to be a godfather. The funny thing is, they are all named the same. My godfather is named Sreten, just like my father and my son, and his wife's name was Natasa, or Nadezda, just like my mom's and my daughter's. They gave us their family name and we accepted it. My children are named after our godparents, but sometimes a confusion arises when there is something official to sing or to do since we live next to each other and they are all named the same.My family has very good relations with almost everybody in this village. On the other hand, I can see today some difficulties arising—these modern kids are different than we were. Kids in the village are a little insolent today. Before, we never had any trouble at all. My mom and her family are from Drenovac—that's why we all came back here. When I first moved here, some 25 years ago, everybody greeted me very warmly. There was no difference between me and my next door neighbor at all. I think we all had a different upbringing in those times. My first neighbor is my closest family, and we love each other still. Sometimes he was closer to me than my own brother or a father. I respect my neighbor very much; he is the first one to help me when I need help, and I'm the same way to him and his whole family. As kids, when I used to come to visit my grandparents here in Drenovac we played Cowboys and Indians—I was of course, the Indian. We go a long way back, like a true family. My son grew up with his [the neighbor's] kids; they are still best friends today, and tied with godparenting.On the other hand, there is a kind of trouble in the air lately. This all started with the beginning of the war [1991]—people became nervous or something, started paying attention to nationalities. And we had a lot of refugees coming here from Bosnia. They stayed in my village, in some empty houses, or people just took them in. Anyway, a year and a half ago the worst incident happened: one Gypsy boy got severely beaten by ten or more villagers [Serbs]. I'm very sorry that such a thing could have happened here. It happened in our village center, where young people go out. In situations like this, I always think “Why am I not a Hungarian; why they call me a Gypsy when I'm not?” Then my soul hurts; from where did they get that I'm a Gypsy? Why call me that? I'm not a Gypsy. I never stole anything. I would help everybody. I'm brought up that way; my kids are the same. We all act like our dear God told us to—by God's laws. We don't steal, we don't kill. We just play our music, that's all. That poor Gypsy boy didn't deserve what happened to him. My son, on the other hand, never had any trouble, ever. That is because he behaves well. My son was involved in only one incident. He was out with his girlfriend, now his wife, and some female friends in a restaurant here in the village. They were still teenagers. And a couple of older men started to say things to them, to the girls. They even threatened that they will move out all Gypsies from this village. They cursed them: “I'll screw your Gypsy mother.” When I think of it, it was always like that. Most arguments and fights were about girls. Usually everything ended up with some bad words and humiliation. But before it was I against that one person; now it's like we are taking opposite sides, the whole village. I don't see anything good coming out of it.These bad events, humiliations I had to take, didn't prevent me from feeling like a true Serb. I love my country; this is where I was born. If I was born in Zanzibar, I would feel the same. I don't hate anybody. I don't like to be humiliated. Why me? So what if I have dark skin? I have the same soul as anybody. Maybe I have greater Serbian soul than some Serbs. My kids were always an example in school: good students, always dressed well, better maybe than some people's kids who had more money. My children always had good pocket-money. They were clean; their mother always looked after that. I've always tried hard. My son is a Serb too; he is born here. I don't want to see him humiliated, ever, because someone calls him names. I sent him to high school in Sabac; he was a good student always. He wanted to become an economist. His third day, kids tried to harm him with a screwdriver because he is like a Gypsy. They tried to take his leather jacket. We came to visit him, my wife and daughter and myself, to see his school and how he's doing. I was very proud of him. We waited for him in the school yard. Some kids were near us, saying very bad, vulgar things about us. My daughter was there; I didn't want her to hear any more of that crap. I pulled my son out of that school.Maybe it's a mistake that we are here, in Drenovac; so nonviolent, but that's our characteristic. Roma, in Macvancki Pricinovic, or even worse, in Draginje, they don't get humiliated or beaten up. No, they fight back. If something like the event with that poor boy had happened over there, only an ambulance crew could pick up parts from Serbs. They are Gurbeti, much more violent. On the other hand, we do the only way we know. We have a different mentality than the Roma. We are much more softer, and we never sold our girls. This is maybe not a nice thing to say, but we do have more culture. We don't have differences with peasants, not at all. Whatever they do, we do it, whatever they have, we have it. Actually, we understand ourselves as Serbs. Gypsies in Macvancki Pricinovic, they declare themselves to be Roma, always; we never do. That's because we feel and see ourselves as Serbs. Gurbeti would speak their Gypsy language even in front of Tito. We all speak Serbian. I can understand some Romanian, but I can't speak it. We never used Romanian in this house. Not even in the village.I was never into politics. However, I think that it would have been much better for all of us from ex-Yugoslavia if we made a split right at the beginning. Serbia is Serbia, and it should stand alone. But God didn't let us split without any incident. In 1991, many males [Gypsy] from this village got drafted, and sent to the Croatian war line. I could have stayed in Austria; we were there just at the outset of war. At that time it was still relatively easy to get “papers” to stay and work there. But I couldn't do it; I could have applied to get an asylum status, but I just couldn't do it; that was kind of shameful for me. Maybe I made a huge mistake; if we did that, my children would be much better off today. Instead, I packed my family and we went home, to Serbia. We traveled by car and passed nearby Karlovci just one day before the Croatian's HDZ slaughtered those 12 poor soldiers. This story followed us on our way home. But we were heading back home and I thought that if it is my destiny to die then I will. If God decided that we shall all be killed, that's it. Now I think I did the foolish thing. I regret the most because of my kids; if we stayed, both of my kids would have a secure future. I'm very emotional when it comes to my home and my hearth. I never approved of Milosevic's politics. In school I was taught to love Tito and partisans, and to hate cetnik formations. Maybe that was wrong too. I'm not afraid in front of God. If only God would give us another Tito! If only God gave us another politician of that caliber, so my son can build his life! People say this and that about Tito, but that's beside the point—whatever he was, he was a master of politics, and I respect him for that. Although I was never a member of a Communist Party. I didn't care about politics. We lived like in paradise when he ruled. He was the authority for the whole world, in those black African countries, everybody [foreign politicians] kneeled before him—I don't care if he was a hooligan. When he was alive I was free to travel everywhere, to stop my car in the middle of the night, in the middle of nowhere, in Bosnia's mountains, among Muslims, and I was safe. I only wish that my son can live through a time like that—to be and feel safe in his own country. In Tito's time, and with our Yugoslav passport, there were no restrictions or visas for all of us. I was so proud to have a Yugoslav passport. That is so totally different than today. In 1990, I took a job in Austria; I was in front of one gas station, loading my car, when an elderly gentleman asked me where was I from. When I said, “Serbia, Yugoslavia” he made a facial expression and said “Serbische Dreck” [Serbian shit/worthless Serb]. That was terrible; it made me feel ashamed. I was also very much ashamed and humiliated when Milosevic surrendered Kosovo. My son served in the Army, in Prizren. We used to go and visit him many times. I'm very sorry about Kosovo; that's the old Serbia, our land. But I think that we should blame not only our politicians, but America, the most. If it weren't for the Americans we would still live together, side by side, in our beautiful Yugoslavia. I'm sorry that my children won't be able to live and travel the way we used to. Whatever happens in the future, it will never be the same.Marriage and familyWe didn't have mixed marriages with Roma. In the past it was a big deal. Now things are mixed up all together. Lately, a few peasant women got married to Romanians. They are all young couples, got married during the war years. Actually it's still a big deal, not among young people so much but their parents don't approve most of the time. They always looked at us like we are at some low position. I never minded. I would approve my son's marriage to any girl, as long as she is nice and from a good family. Also, a few males married Muslim women, the refugees from Bosnia. There are now six or seven sisters, all Muslims that are married into this village. After 1991, when the war started, many young people got married, including my son. I thought that he was too young, and he was; he was 17 at that time. But that was a special time, a war, a crisis. When the refugees started to come, first from Petrinje, Croatia, then from Bosnia, it was scary. Then I thought, “These people [the refugees] lost everything; many of their relatives are killed. Who knows what will happen tomorrow?” My son got married in 1999, during the bombing. We had his wedding while the bombs were falling. He actually got married because his girlfriend got pregnant. And in a time like that a baby seemed like a blessing. We didn't know where the war zone would end, and if we would survive.In this village, people marry late—I got married when I was 25 years of age. I'm married to Biljana. She's a great wife and a good mother to my kids. She was my first true love. My wife is from Tabanovic, another Romanian village. My aunt is married there. I met my future wife when she was just a kid. I'm older than she is, five years. I dated many girls from her village and she knew about it. I remember her—she had nice braids, and I used to pull her by her hair. She was very beautiful; even today she has the most wonderful green eyes. But she was too young for me at that time. We had some kind of a sympathy for each other, but I didn't want to get serious until I served in the Army. I was 20 when I went into the Army, in 1977. After that, when I came to visit my aunt and saw Biljana, I knew she was the one. However I was still not good financially at that time and I didn't want to drag her with me into poverty. I first wanted to create something so I could provide a decent life for my wife. Also, I wanted to become known as a good musician, as a good, respectable and stable man before I got married. So we dated for about five years. Finally, I was able to marry her. We got married on January 8th. She had a good voice, and we started to perform together. I took her to Italy on our honeymoon. We played there at some Roma wedding. Those were the Cergari Roma who escaped from Yugoslavia before 1941 to Italy. They were very rich. I don't know what they did, but they had a lot of money. These Gypsies were from the old kind—even their tzar appeared. He was their judge and their tzar. That old man, he was around 70 years of age, was so respected; I've never seen something like that before. He was really the authority for all of them. That's their Roma tradition. We earned so much money then. I remember I bought a car in Italy. Those were the days! Actually, I ended up in an Italian hospital after that wedding. I played for three consecutive days; harmonica can be heavy after a while, and my foot capillary broke. The hospital was great—they gave us a menu every day to choose from like five meals! I enjoyed their pasta very much. We traveled to Italy, Germany, Austria and all over former Yugoslavia. I used to make 4000-5000 Germans marks per night! In 1980, I bought my first Mercedes. But those days are gone now.Now we all live here in Drenovac. My son is married and recently got a little baby boy. I'm very proud to be a grandfather. My daughter is still in school. My brother is still in Austria; he used to work with his two sons in a plaster shop. Actually, my godparent employed him at first. But he got very sick: a brain cancer, but he is doing better now, thank God. My brother and I were like twins, never apart, always there for each other. Our godparent lent us money, 500 German marks, but I don't have to give it back to him. That's how close we are. So my brother and I succeeded in getting out of the poverty of our childhood and built a nice life. Most important, we managed to raise our children well and to keep our family together. The only thing I'm sorry about is that my mother died. I wish she could see us today; she would enjoy her grandchildren and great-grandson. She would enjoy a better life.DiscussionDragan's account tells us about his personal experience. Often he mentions historical events and particular crises in either his life or the life of his community. Mostly, his story is not about how things really were but about how he preferred to remember the past. Jay Ruby28 has argued that what people say about themselves is data to be interpreted, not a simple factual truth. Dragan favors “the old days” of former Yugoslavia because Gypsies, among other things, were able to benefit from Tito's communism in full: guaranteed income in spite of there being little or no work. These are typical Gypsy feelings in Serbia today, as one Gypsy informant explained: “In other countries, if you don't work, you don't have anything to eat, but in [former] Yugoslavia, you can eat and drink and don't have to work at all!”29 Even under communism, Gypsy circumstances were not bright: those employed Gypsy workers were in unskilled jobs that required no, or minimal, skills. In the former Yugoslavia, in 1986, 80% of Gypsy children did not complete primary school and the employment rate among Gypsy workers stood at 80%;30 in 2000, just 20% of the Gypsy labor force held regular jobs, and only 5% of these workers were employed in state companies. Some 58% of Gypsy men and 89% of women have no profession, either traditional or modern.31 The low education rates are a result of both the apathy of the Gypsy community and parents’ disinterest in the education of their children.32 Also, the two main reasons for the high unemployment rate are the Gypsies’ lack of interest in long-term, institutional employment combined with their traditional distrust of authority and their lack of professional skills.33At the same time, the former communist regime in Yugoslavia propagated the idea of a “classless” and ethnicity-less society for the past 60 years, which supposedly promoted general acceptance of all ethnic groups. With his Yugoslav passport, Dragan, like so many Gypsies at that time, was able to travel without restrictions in search for work. One of the preferred destinations for many Mačva Gypsies was Austria, where they have established some sort of residence and benefited from the Austrian welfare system.34Although Dragan has tried hard to trade his “gypsyness” for success in life, unlike some other Gypsies with a strong sense of ethnicity, his account implies a sense of acceptance of the Gypsy low social position and even stereotypes, no matter how much he dislikes it. In fact, it appears that Dragan himself associates Gypsy identity with theft and dubious honesty when he says: “… I'm not a Gypsy. I never stole anything. I would help everybody …” Throughout Europe, the common stereotypes associated Gypsies with tricks and deceiving non-Gypsies: theft, lock-picking, purse-stealing, horse-stealing, casting spells, general witchcraft, and trickery.35 Because of these stereotypes, Dragan seems to be engaged in the task of creating a space of exception for himself and his family, constantly defining himself in explicit opposition to anything with Gypsy connotations. Due to their distinctive phenotype and widespread stereotypes Gypsies often have less agency in choosing/creating identity than other ethnic groups. It is a great irony then, that Dragan, in all of his efforts to define himself in opposition to Gypsy ethnic stereotypes, is faced by another common ethnic stereotype, “Serbian shit,” aimed as a slur for Serbs.In Dragan's oral account, the recurring theme is one of connection—to his family, his “people,” including his neighbors, the Serbs, and his village. These connections are explored through ties of kinship and his attachment to land/place emphasizes his sense of belonging. Both kinship and land offer more than a mere background for Dragan's narrative. They actually structure and shape his story. Like so many Gypsies, Dragan married endogamously, within his own tribe, just as his parents did and, recently, his son. In Serbia, Gypsies traditionally encouraged endogamy: so cooperative in trade, they restrained from marriage with outsiders and even with different Gypsy groups, so maintaining their group identity and separation.36 The Gypsy endogamy has helped their sons get wives—it was always difficult for a Gypsy to marry into a Serbian family—so it was self-maintained as much as it was imposed. In a case of a mixed marriage, not only would a Gypsy's status improve, but his or her potential children would acquire wider kinship ties and better cooperation with Serbs. Especially in Serbian rural areas kinship still remains a very important foundation for cooperation and a lot of what passes for ethnicity at the local, micro level is, in fact, kinship. In the absence of marriage kinship ties, Dragan emphasizes instead traditional godparenting that created good cooperation with the Serbs from his village for generations.When Dragan uses traditional elements of his culture to speak about the past, his life history becomes visible cultural practice and behavior rather than merely a complementary ethnographic account.37 The social stratification and limited marriage choices of rural Gypsies in Serbia preserved their local, village traditions, including their occupations. A key socializing mechanism for Gypsies that promoted their survival and reproduction in the past was an emphasis on the particular occupations employed by various Gypsy tribes. Like his ancestors before him, Dragan chose music/entertainment as his life occupation, and his son followed the family tradition. The behavior Dragan and his family employed for many decades, and the “preferred ethnicity” is better understood by comparing his narration to available historical and anthropological data. Throughout their history, Gypsies have chosen only particular occupations in which they were self-employed and which did not require education.38 Even with modernization, when they abandoned old occupations in favor of new, they did not compromise their freedom. In light of local history and the social environment, Dragan's specific life course and decisions become more recognizable.At the time when Dragan's ancestors first appeared in Serbia, during the second half of the nineteenth century, Mačva was largely uninhabited territory due to the centuries of wars, famine, and disease. Serbia was liberated from the Turkish rule but its territories were almost deserted. At that time, the village of Drenovac had only a few inhabitants, and later on a few villages fused into the present day village.39 Since the Serbian government supported immigrants from Herzegovina, the majority of inhabitants in present day Drenovac are descendants of those immigrants. Reports relate that the natives, especially in the beginning, refused to accept the newcomers even though Mačva was largely uninhabited. Conflict was mitigated not only by various orders of the Serbian government but by the fact that the newcomers from Herzegovina belonged to the same ethnic entity. They were Orthodox Serbs who carried the same cultural models. The government also supported the colonization of various Gypsies. Workers and craftsmen were needed in rural Serbia, and the demand for their goods led to their acceptance in the villages. There are documents noting Gypsy presence in Mačva in 1718 and 1834, but the majority of Romanian Gypsies came after the abolition of slavery in Romania, after 1864. The Lejaši Gypsies, Dragan's ancestors, however, were never craftsmen like other Gypsies. They were musicians, and musicians were never needed by the rural Serbian economy. Certainly, the Lejaši music was very much prized,40 and they were the only group to perform professionally, but they could make a living from their music only occasionally, at weddings, village fairs, and special events. Their payment was in the form of a tip, which made them dependant on the kindness and mood of their audience. Unlike the Serbs from Herzegovina, Gypsies also had highly visible and diverse phenotypes like skin color, hair texture, dress, and a way of life which surely made them appear as outsiders and intruders by the natives. All these factors probably posed a potential danger of ethnic conflict in the form of territorial defense. Gypsy success in surviving has always depended on a favorable social climate. In their effort to survive and become more accepted, many Lejaši must have made a calculated decision to adopt Serbian identity and culture, as did Dragan's family. This required losing their own traditions, language, dress, and wider kinship ties to other Roma in the area. Through godparenting ties, they encouraged behavior that usually occurs between close relatives to be directed towards non-kin. The nineteenth-century Serbian government was interested in forming larger villages that would have more stability, and cooperation was greatly encouraged among ethnic groups. Therefore, the decision of Dragan's ancestors to become “true Serbs” seems justified, especially so since the Lejaši were not liked by other Gypsy groups in the area due to the different language and religion. By transmitting a particular successful behavior to their own descendants, Dragan's ancestors may not have only increased their number of descendants but also sanctified particular behavior.41 These behavioral models influenced the choices that individual Gypsies were forced to make in the context of external circumstances. The preferred identity so became the end product of their life experience, producing patterns of emotional and behavioral traits.ConclusionFor narrators from marginalized groups, oral history presents a chance to tell their own story, in their own words. Their narrative can deepen our understanding of the cultural behavior of individuals by revealing personal versions of events and individuals’ understanding of the past.Dragan's oral history reveals his love of his family and his nostalgia for the past, when he was able to travel freely and enjoy the fruits of Tito's communism, especially acceptance of all ethnic groups. From his account, we learn about his personal experiences, hopes, setbacks, and his family tradition. He remains proud of himself and of his family heritage. They managed to succeed in life despite many obstacles, including their initial poverty and background. Dragan is especially pleased by his godparenting and friendship ties with the Serbs in his village and emphasizes many times that there are no true differences between them. At the same time, he is also aware of his background and the position of Gypsies in the society. Yet, throughout his life, like his parents before him and their parents before them, he has consciously traded his ethnicity for social success in life. The local life-strategy of “the preferred ethnic” identity adopted by his ancestors has created an ongoing dialogue with the social environment and political circumstances.When I came back to the village, in the winter of 2006, many things had changed: many of the Karavlax “true Serbs” now declared themselves “true Roma,” emphasizing their “ethnic revival” and the distinctiveness of the Roma culture. This change was brought about by the general rise of nationalism in Serbia and, among other things relevant for Roma in the area, by the presence of several non-governmental, humanitarian organizations that operate at the territory of western Serbia, and whose main task is to provide humanitarian and material help for the Mačva Roma. The brick factory where the Lejaši males used to work has been sold as part of the privatization process in the transition to a market economy. For most villagers of Drenovac, the only source of income today is social help/welfare and humanitarian help. In order to become recipients of humanitarian help, these “true” Serbs had to declare themselves as Roma. The Gypsies with whom I spoke said that they had to “convert” in order to survive, but also, according to their own words, they finally see their chance to obtain their own “culture and identity,” and the fact that they are sponsored and paid in order to do so helps a lot. With the help of these organizations, the Roma from Drenovac have opened a “Gypsy club” in the village where they gather daily to discuss local news, politics, and their activities. Many of these Roma are now members of the local Roma political parties, financed by the same sources, and they claim to fight for Roma rights and social dignity.Dragan's life changed too: his wife got sick and is unable to perform in the band so he stays home most of the time trying to breed pigs and chickens. He complains that he cannot make enough money for his family anymore. On the other hand, he stays committed to his “preferred ethnicity,” although he clearly sees the need of the Roma people to “finally do something for themselves” as he says, even if that means to reject the previous identity. He supports his fellow villagers in their efforts, but warns that this change has already provoked many sparks among the local villagers and that verbal and sometimes even physical conflicts with the local Serbs are on the rise. He did not apply for any sort of humanitarian or social help.It is too early to conclude how this ethnicity change will affect the Roma from Drenovac. Gypsies always depended on the needs and contacts with their host countries as a source of their livelihood. The Lejaši may possibly lose their relatively good cooperation with their Serbian neighbors but gain more on the other side, in their alliances with the other Roma and political parties from the area. In Serbia, Gypsy ethnic flexibility comes from their readiness to adjust their behavior in order to create the most favorable strategy for making ends meet. Gypsy culture, as well as identity, responds in an ongoing dialogue with their social environment, and so far they have managed to survive because of the social strategies they used, trading their objective identity for the preferred one. Humans do not passively accept models of behavior from their social network. They accept the models that work, change those that do not, and share and transmit those changes back into their social group in a continually evolving creative process.42 Such processes are open-ended and able to strategize adaptively in response to novel environmental situations.431Jelena Čvorović, “Gypsies Drown In Shallow Water: Oral Narratives among Mačva Gypsies,” Journal of Folklore Research 43, no. 2 (2006): 129–48.2Angus Fraser, The Gypsies (Cambridge: Blackwell, 1992).3Ian Hancock, The Pariah Syndrome: An Account of Gypsy Slavery (Ann Arbor, MI: Karoma Publishers, 1987).4Fraser, The Gypsies.5Fraser, The Gypsies; Čvorović, “Gypsies Drown In Shallow Water”; Tihomir Djordjevic, Naš narodni zivot I običaji (Beograd: Knjizevna Zadruga, 1932).6Luba Kalaydjieva, Francesc Calafell, Mark A. Jobling, Dora Angelicheva, Peter de Knijff, Zoë H. Rosser, Matthew E. Hurles, Peter Underhill, Ivailo Tournev, Elena Marushiakova, and Vesselin Popov, “Patterns of Inter- and Intra-group Genetic Diversity in the Vlax Roma as Revealed by Y Chromosome and Mitochondrial DNA Lineages,” European Journal of Human Genetics 9 (2001): 97–104.7Fraser, The Gypsies.8Elena Marushiakova and Veselin Popov, Gypsies (Roma) in Bulgaria (Peter Lang: Frankfurt am Main, 1997); Jelena Čvorović, “Caste Behaviors among Serbian Gypsies,” Collection of Papers EI Serbian Academy of Sciences and Arts 23 (2007): 151–68.9Hancock, The Pariah Syndrome.10Anzej Mirga and Mruz Leh, Romi, razlike I tolerancije (Beograd: Akarit, 1997).11David M. Crowe, A History of the Gypsies of Eastern Europe and Russia (New York: St Martin's Press, 1996).12United Nations Development Programme, Srbija I Crna Gora: izvestaj 2002 (Beograd: UNDP, 2002).13Katherine Pinnock, Denied A Future? The Right to Education of Roma/Gypsy and Traveler Children in Europe (United Kingdom: Save the Children Fund, 2001).14Mirga and Leh, Romi, razlike I tolerancije.15Aleksandra Mitrović, Na dnu. Romi na granicama siromaštva (Beograd: Naučna Knjiga, 1990); Jelena Čvorović, “Sexual and Reproductive Strategies among Serbian Gypsies,” Population & Environment 25 (2004): 217–42.16Tatomir Vukanovic, Romi (Cigani) u Jugoslaviji (Vranje: Nova Jugoslavija, 1983).17Tihomir Djordjevic, Iz Srbije Kneza Miloša. Stanovništvo-naselja (Beograd: Geca Kon, 1924).18Vukanovic, Romi (Cigani) u Jugoslaviji; Čvorović, “Caste Behaviors among Serbian Gypsies.”19Jelena Čvorović, “The Making of Gypsies: Invention of Tradition,” Issues of Ethnology and Anthropology 1, no. 1 (2006): 47–59.20Pinnock, Denied A Future?21Čvorović, “The Making of Gypsies.”22Jelena Čvorović, Gypsy Narratives: From Poverty to Culture (Belgrade: The Institute of Ethnography, Serbian Academy of Sciences and Arts, 2004).23Čvorović, “Gypsies Drown In Shallow Water.”24Čvorović, “Sexual and Reproductive Strategies among Serbian Gypsies.”25Hancock, The Pariah Syndrome.26Čvorović, “Sexual and Reproductive Strategies among Serbian Gypsies.”27Dragan Vasiljkovic, interviewed by Jelena Čvorović, May 10, 2003.28Jay Ruby, “Speaking for, Speaking about, Speaking with, or Speaking Alongside: An Anthropological and Documentary Dilemma,” Visual Anthropology Review 7, no. 2 (1991): 50–67.29Jelena Čvorović, “Gypsy Oral History in Serbia,” Oral History Journal 33, no. 1 (2005): 57–68.30Children of Minorities: Gypsies, ed. Sandro Costarelli (Florence: United Nations Children's Fund, International Child Development Centre, 1993).31Pinnock, Denied A Future?32Jelena Čvorović, “Gypsy Ethnic Socialization in Serbia,” Bulletin of the Institute of Ethnography Serbian Academy of Sciences and Arts 53 (2005): 35–48.33Children of Minorities: Gypsies, 44–7.34Čvorović, Gypsy Narratives.35Fraser, The Gypsies, 62.36Čvorović, “Caste Behaviors among Serbian Gypsies.”37Julie Cruikshank, Life Lived Like a Story (Lincoln and London: University of Nebraska Press, 1992).38Fraser, The Gypsies, 62; Jelena Čvorović, “A Gypsy, Butterfly and Gadje: Narrative as Instruction for Behaviour,” Folklore 119, no. 1 (2008): 29–44.39Tihomir Djordjevic, Iz Srbije Kneza Miloša. Stanovništvo-naselja (Beograd: Geca Kon, 1924).40Djordjevic, Naš narodni zivot I običaji.41Lyle Steadman and Craig Palmer, “Myths as Instructions from Ancestors: The Example of Oedipus,” Zygon 32 (1997): 341–50.42Paul Di Maggio, “Culture and Cognition,” Annual Review of Sociology 23 (1997): 263–87.43Kevin B. MacDonald, Preface to the First Paperback Edition, Separation and Its Discontents: Toward an Evolutionary Theory of Anti-Semitism (Bloomington, IN: Firstbooks Library, 1998/2003).
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B08BB7BF94AD90CF939841DDA5DA30FD14EDA7C.txt b/test/dataset/in/resources/corpus/Clean_0B08BB7BF94AD90CF939841DDA5DA30FD14EDA7C.txt
deleted file mode 100644
index 89a3cbb..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B08BB7BF94AD90CF939841DDA5DA30FD14EDA7C.txt
+++ /dev/null
@@ -1 +0,0 @@
-]>EXG5145S0531-5565(98)00048-510.1016/S0531-5565(98)00048-5Elsevier Science Inc.Fig. 1(A) Discovery site of the Ice Man from the Hauslabjoch in the Ötztaler Alps, September 19, 1991. (B) Preservation unit containing the Ice Man. (C) Face of the Ice Man. Note the high degree of dental abrasion and the diastema between the upper central incisors. (D) Three-dimensional polyacrylic model of the osseous skull (stereolithographic reconstruction). (E) Set of tattoos (vertical lines) in the lumbar region. (F) CT-plane of the thorax showing the lungs and heart for computer assisted endoscopic navigation, investigation, and sample withdrawal. (G) Endoscopic inspection of the lower middle mediastinum. The heart (covered by pericardium) and left lung can be seen. The coronary vessels are also visible. (H) Endoscopic view of the surface of the left lung covered by the pleura. Anthracotic (coal-like) pigment is clearly visible in the lung tissue. (I) Osteon method for determining the age at death of the Ice Man.Original Articles“‘Man from the HauslabjochOthmarGabera*othmar.gaber@uibk.ac.atKunzelKarl-HeinzKünzelaaInstitute für Anatomie, Universitas Leopoldino-Franciscea Oenipontana, Müllerstasse 59, A-6010 Innsbruck, Austria*Correspondence to: O. Gaber; Fax +43-512-507-3061Abstract—On September 19, 1991, at an altitude of 3,200 m (10,498 feet), the remains of a well-preserved, freeze-dried mummified body of a 3,000 bc, Late Stone Age man, was found on the Austrian–Italian border. Studies conducted at the University of Innsbruck have revealed that these are the remains of a 45–46-year-old male, 45 kg (99 lbs) (living weight) and 160 cm (5 feet, 2 inches) tall. Morphometric, pathophysiologic, paleobotanical, photogrammetric, anthropological, computer tomographic, and other studies have been conducted to determine the living conditions and possible causes of death of this Late Stone Age man.KeywordsHauslabjochTyrolmorphometrypathophysiologypaleobotanyphotogrammetryanthropologycomputer tomographyIntroductionOn September 19, 1991, a body emerging from the surrounding ice was observed by two German mountain hikers at an altitude of 3,200 m (10,498 feet), within a glacier area at the border between Austria (Tyrol) and Italy (South Tyrol) (Fig. 1A). After reporting the find, it was quickly realized that these remains were not of some unfortunate contemporary climber, but that of an individual of far greater antiquity. Within a week of the discovery, the news was known worldwide.InvestigationThe find consisted of a fairly well-preserved, freeze-dried, mummified body of a man who also carried with him a number of implements. Dating (14C) of both soft tissues and bone revealed that the individual was approximately 5,300 years old, i.e., from 3,000 bc or the Late Stone Age.To preserve the remains and allow for thorough preservation and investigation, the corpse was transferred to the Institute for Forensic Medicine, Institute for Anatomy at the University of Innsbruck in Austria.Since his discovery, the “Ice Man” has been stored within a specially constructed cooling and conservation system (maintained at −6°C, with a relative humidity of 98%). These parameters are constantly monitored by electronic sensing and warning devices. Anatomical scientists have been in charge of the “Ice Man” since he was placed in this glacier-like environment (Fig. 1B) (Gaber et al., 1990).Sixty-four scientific teams in the fields of archaeology and natural sciences (e.g., botanists, glacerologists, geologists, metallurgists, and physicians) have compiled data to determine the conditions under which the “Ice Man” lived and eventually died.A few facts regarding this Late Stone Age Man: Names: Ice Man; Man from the Hauslabjoch; Man from the Tisenjoch; Similaun Man; Homo tirolensis; Frozen Fritz, or “Ötzi.”Sex: Male.Body height: 160 cm (5 feet, 2 inches).Body weight: 13.3 kg (29.3 lbs) [living weight: 45 kg (99 lbs)].Age: 45–46 years.Historical age: 5,300 years (determined by modern dating [14C] techniques).Genetic ancestry: European (DNA determined from bone and muscle fragments).Special features: high degree of dental abrasion, lack of wisdom teeth, diastema between upper central incisors (Fig. 1C), lack of the 12 set of ribs, rib fracture on left side (healed), rib fracture on the right (recent injury), deep lesion in the left hip, gluteal–femoral region, and tattoos (stripes and crosses) at strategic locations on the body.Nationality: unfortunately, no passport was found with the Man from the Hauslabjoch, but if he was alive today he would be considered Italian. Based upon resurvey of the border between Austria and Italy (Bolzano Province), the site of the discovery was determined to lie 92.56 m (303.7 feet) within Italy. By contractural agreement, all initial investigations were conducted in Innsbruck.Items that the ice man carriedAround the Ice Man were found many objects that were transferred to the Römisch-Germanisches Zentralmuseum in Mainz for preservation and restoration. The items consisted of: 14 arrows—partly feathered and equipped with flint heads; a leather pouch containing objects of flint, sinew, bone, and pitch that appear as some sort of “repair kit;” a flint stone dagger made up of an oakwood handle covered by a braided material; an ax with a blade of nearly pure copper; pierced stone beads to which were attached a number of leather straps; a birch-bark container; two lumps of tree fungi with a leather strap; a backpack frame made of hazelwood; and clothing—footwear leggings, leather loin cloth, bear skin cap, and grass wrapping material.FindingsAnthropological and paleobotanical findingsUpon further investigation, the conditions under which the Ice Man lived were reconstructed. Although his origin and destination are forever lost in history, we have learned a great deal about his living conditions.The anthropological data and the parallel skeletal finds from other Neolithic graves led us to assign the Ice Man to the so-called “Remedello group,” which had settled at the southern part of the Lake Garda region in Northern Italy. Moreover, the paleobotanical investigations, based upon different types of grasses, and remains of grain and pollen all support the theory that the Ice Man’s origin was south of the Alps. Additional support of this theory is that the flints that he carried could have been found or mined from the Trentino region of Italy.Computer-tomographic numerical dataThese data have allowed us to reconstruct the skull three dimensionally, and a plastic model was produced using stereolithography. Once completed, further anthropological measurements and other anatomical investigations could be conducted without transferring the Ice Man from the cooling unit (Fig. 1D) (Künzel et al., 1992; Zur Nedden and Wicke, 1992).PhotogrammetryThis geographical screening method has been used to achieve a reconstruction of the Ice Man’s body surface. Within days of the discovery, some dark-gray alterations of the skin were observed (Fig. 1E) (Sjovold et al., 1995; Van D Velden et al., 1995). Other markings became visible, these consisted of stripes and crosses. Further inspection revealed more cutaneous markings or “tattoos” on both calves.Infrared photographyThis technique allowed additional visualization of the tattoos. The newly discovered tattoos, not visible to the unaided eye, revealed 57 stripes arranged in 16 groups, and in addition, one large and small cross-like marking. These tattoos were localized in certain regions where, due to physical stress, degenerative changes occur that cause chronic pain. Upon further x-ray investigation, discrete bone damage beneath these skin areas was observed.Because of its biomechanical, static, and dynamic functions, the lumbar region of the spine carries the greatest structural weight of the axial skeleton. Manifestations of joint and skeletal stress, with resulting symptoms and signs, are well known to be found in muscle origins and in inserts as well as in the ligaments apparatus and in the vertebral joints. Hence, the markings previously described appear to be localized above these sensitive areas.The position of the tattoos in both lower legs is striking because the marks project towards the functionally important muscle–tendon–ligament structures. The above markings, on both the lumbar spine and lower legs, appear to reflect the results of strain, probably due to moving on rough ground, and would identify areas where pain would originate. Another example of such a marking is the positioning of a large cross on the medial side of the right knee where the marking projects exactly to the medial capsular ligament structures and the medial meniscus, another well-known site of pain resulting from chronic stress.Hence, the tattoos could have been either a form of medical treatment (e.g., acupuncture) or could have been the result of a religious or tribal ceremony. However, the findings above lead us to hypothesize that these markings or tattoos denoted areas of pain due to overuse.Bacteriological evaluationThese studies detected 20 eggs of the parasite (Trichirus trichiura, or whip worm) in the small intestine (Aspöck et al., 1996). The colon also contained a large amount of quartz-like sand, signifying the eating of milled grains. This abrasive material also helps explain the high abrasions found on the Ice Man’s dentition.EndoscopyComputerized tomography is the basis for drawing specimens from intracorporeal regions (Fig. 1F). However, in contrast to living material, and through the process of mummification, all internal organs of the Ice Man had dried out and become much smaller, so newer methods were required for accurate sampling. Hence, a new device to enable minimal invasive endoscopical sampling was developed. This newer technique was a computer-supported “virtual reality,” so exact placement on specific organs could be obtained (Gunkel et al., 1997).The endoscopic precision probes were composed of pure titanium, which prevents any tissue contamination of trace elements that might be detected in future experiments (Mersdorf and Tessadri, 1995).As of this writing, samples have been obtained from the upper respiratory passage, the lungs, liver, spleen, rib, cartilage, as well as bone, muscle, and tissue from the diaphragm, intercostal muscles, brain, and peripheral nerves. These 100 specimens were transferred to 80 team for further investigation. These samples are to be tested for histopathology, DNA testing and determination, parasitology, and trace elements (Fig. 1G and H) (Hess et al., in press, Platzer et al., in press).Recent discoveriesThe age of the Ice Man had been considered to be between 25 and 35 years at the time of his death. However, we used two different methods of age determination based on the physiological changes in the development and transformation of bone tissue during aging. By screening a thin section, the number and the areas claimed by secondary osteons and osteon fragments were recorded, and the definitive age was determined by regression analysis and micromorphological methods. Applying macromorphological methods, we used spongious structure and its changes during aging, at the proximal ends of the humerus and femur. Combination of these two methods yielded results that now age the Ice Man to 45 to 46 years (Fig. 1I) (Gaber, 1998)(Hess et al., 1989)ConclusionThe Ice Man provides a unique view into the past, and new techniques are allowing us to reconstruct the life of an individual in the Late Stone Age. Recently, the Ice Man was transferred to the Archaeological Museum in Bolzano, possibly near his ancestral home, where he will remain as a symbol of European ancestry.ReferencesAspock et al 1996HAspöckHAuerOPicherTrichirus trichiura eggs in the neolithic glacier mummy from the AlpsParasitol. Today1271996255256Gaber et al 1990Gaber, O., Künzel, K.H., Maurer, H., and Platzer, W., Konservierung und Lagerung der Gletschermumie. In: Der Mann im Eis 1, Veröffentlichungen der Universität Innsbruck, vol. 187, Höpfel R., Platzer W., and Spindler K. (Editors), pp. 92–99, Eigenverlag, Innsbruck, 1990Gaber 1998Gaber, O., Altersabhängige Veränderungen an Humerus, Femur und Tibia. Identifikation Unbekannter Toter Tote, Kap. 6, Schmidt-Römhild (Ed.), Lübeck, 1998Gunkel et al 1997A.RGunkelWFreysingerWThumfartM.ITruppeOGaberK.HKünzelWPlatzerFTiefenbrunerOtorhinolaryngologie computer-assisted biopsies of the IcemanArch. Otolaryngol Head Neck Surg.1231997253256Hess et al 1989M.WHessGKlimaKPfallerK.HKünzelOGaberHistological investigations on the Tyrolean Ice ManAm. J. Phys. Anthrophol.1061989521532Kunzel et al 1992Künzel, K.H., Steinlechner, M., Gaber, O., and Platzer, W., Morphologische Vergleichsstudie an Schädeln zur Schädel-CT-Rekonstruktion des Eismannes. Der Mann im Eis 1, Veröffentlichungen der Universitüt Innsbruck, Vol. 198, Höpfel R., Platzer W., and Spindler K. (Editors), pp. 117–130, Eigenverlag, Innsbruck, 1992Mersdorf and Tessadri 1995Mersdorf, E. and Tessadri, R., Materialzusammensetzungen von Werkzeugen zur Untersuchung interner Teile der Hauslabjoch-Mumie. Der Mann im Eis 2, Neue Funde und Eregebnisse. Veröffentlichungen des Forschungsinstitutes für Alpine Vorzeit der Universitüt Innsbruck, The man in the ice, Vol. 2, Spindler, K., Rastbichler-Zissernig, E., Wilfling, H., Nedden, D.Z., and Nothdurfter, H. (Editors), pp. 287–290, Springer, New York, 1995Platzer et al in pressPlatzer, W., Künzel, K.H., Gaber, O., and Maurer, H.,Endoskopische Untersuchungstechnik und Probenentnahmen beim Eismann. Radiological and Anatomic Investigations on the Iceman. Veröfentlichungen des Forschungsinstituts für Alpine Vorzeit der Universität Innsbruck, The Man in the Ice 6, Springer, New York (in press).Sjovold et al 1995Sjovold, T., Bernhard, W., Gaber, O., Künzel, K.H., Platzer, W., and Unterdorfer, H., Verteilung und Größe der Tätowierungen am Eismann vom Hauslabjoch. Der Mann im Eis 2, Neue Funde und Eregebnisse. Veröffentlichungen des Forschungsinstitutes für Alpine Vorzeit der Universität Innsbruck, The Man in the Ice, Vol. 2, Spindler K., Rastbichler-Zissernig, Wilfling, H., Nedden, D.Z., Nothdurfter, H. (Editors), pp. 279–286, Springer, New York, 1995Van D Velden et al 1995Van D. Velden, E., Den Dulk, L., Leenders, H., Dingemans, K., Van D. Bergh Weerman, M., Van D. Putte, S., Vuzevski, V., and Naafs, B., The decorated body of the man from Hauslabjoch. Der Mann im Eis 2, Neue Funde und Eregebnisse. Veröffentlichungen des Forschungsinstitutes für Alpine Vorzeit der Universität Innsbruck, The Man in the Ice, Vol. 2, Spindler, K., Rastbichler-Zissernig, E., Wilfling H., Nedden, D.Z., and Nothdurfter, H. (Editors), pp. 275–278, Springer, New York, 1995Zur Nedden and Wicke 1992Zur Nedden, D. and Wicke, K., Der Eismann aus der Sicht der radiologischen und computertomographischen Daten. Der Mann im Eis 1, Veröffentlichungen der Universität Innsbruck, vol. 187, Höpfel, R., Platzer, W., and Spindler, K. (Editors), pp. 131–148, Eigenverlag, Innsbruck, 1992
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B08F0CB648E93125EAA550E394D6B967D877DAB.txt b/test/dataset/in/resources/corpus/Clean_0B08F0CB648E93125EAA550E394D6B967D877DAB.txt
deleted file mode 100644
index 0934784..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B08F0CB648E93125EAA550E394D6B967D877DAB.txt
+++ /dev/null
@@ -1 +0,0 @@
- gerona J Gerontol A Biol Sci Med Scigerona The Journals of Gerontology Series A: Biological Sciences and Medical Sciences J Gerontol A Biol Sci Med Sci 1079-5006 1758-535X Oxford University Press 9310.1093/gerona/62.1.93 Journal of Gerontology: Medical Sciences Smoking Is a Risk Factor for Decreased Physical Performance in Elderly Women Rapuri Prema B. Gallagher J. Christopher Smith Lynette M. 1Bone Metabolism Unit, Creighton University, School of Medicine, Omaha, Nebraska. 2Department of Preventive and Societal Medicine, University of Nebraska Medical Center, Omaha. Address correspondence to Prema B. Rapuri, PhD. E-mail: premarapuri@yahoo.com 1 2007 62 1 93 99 1 5 2006 18 8 2005 Copyright 2007 by The Gerontological Society of America 2007 Background. In 487 elderly women aged 65–77 years, we examined the association of smoking with physical performance measures of muscle function and whether the effect of smoking on physical performance measures is mediated through its effect on vitamin D or estrogen metabolism. Methods. Timed rise, timed walk at normal and fast speed, grip strength, and serum biochemical measurements were compared between smokers, past smokers, and nonsmokers. Analysis of covariance was used to compare physical performance variables while adjusting for confounding variables. Results. Compared to nonsmokers and past smokers, current smokers were significantly (p <.05) slower on timed rise and timed walk tests and had decreased grip strength. From multivariate analysis, smoking, age, total body fat, and serum 1,25(OH)2D examined as quartiles were predictors of physical performance measures. The effect of current smoking on physical performance was equivalent to a normal age-related decline in physical performance tests of 7–11 years depending on the test. Conclusions. The results of this study suggest that current smoking is a risk factor for decreased muscle strength leading to decreased physical performance in elderly women. The effect of smoking on physical performance is in part mediated by its effect on 1,25(OH)2D metabolism. Smoking may also have an independent effect on physical performance possibly by a direct effect on muscle or through an effect on vascular function. hwp-legacy-fpage 93 hwp-legacy-dochead RESEARCH ARTICLE A decline in muscle mass and function is a well known consequence of aging. Between the ages of 20 and 80 years, humans lose approximately 20%–30% of their skeletal muscle mass. The effects of age-related changes in muscle are declining physical function and loss of muscle strength, leading to increased incidence of falls in elderly people (1). The loss of muscle mass and function with age is the result of complex multifactorial processes, and the risk factors are multiple. Evidence suggests that increased muscle weakness and decreased functional ability with aging can be attributed in part to decreased vitamin D levels, and it has been shown that vitamin D supplementation improves physical performance in older people (2–5). Increased muscle strength has also been suggested to be associated with higher estrogen levels or estrogen therapy in some studies (6,7) but not all (8,9). The health consequences of smoking are well established. There is ample evidence indicating that cigarette smoking contributes to a number of diseases, including cancer, vascular diseases, respiratory diseases, and osteoporosis, among others (10). Smoking adversely affects the vitamin D-parathyroid hormone (PTH) system (11) and has anti-estrogenic effect in women (12). Smoking is associated with lower total body fat mass, which limits the production of estradiol in postmenopausal women (12). Increased metabolism of estradiol by smoking is another possible reason proposed for decreased bioavailability of estradiol (12). Limited cross-sectional and longitudinal studies suggest that smoking also contributes to physical function decline (13–15). In a cross-sectional study of elderly postmenopausal women, we examined the association between current smoking and physical performance measures. In addition, we also explored whether the effect of smoking on physical performance measures in these women could be mediated through its effect on vitamin D and/or estrogen metabolism. Materials and Methods The study population comprised 489 women aged 65–77 years who entered a multicenter osteoporosis intervention trial (STOP IT: Sites Testing Osteoporosis Prevention/or Intervention). The baseline data of this study population were used for the analysis. The study participants and the study design have been described in detail in an earlier publication (16). In brief, participants between the ages of 65 and 77 years were recruited through advertisements in local newspapers or by mass mailing of letters inviting them to participate in a 3-year study. The letter was followed-up with a phone call. About 8005 women were contacted by mail, and 1905 women agreed to come for a preliminary screening. Exclusion criteria have been described in an earlier osteoporosis publication (16); in general, women with any disease known to affect calcium metabolism were excluded from the study. To complete eligibility, femoral neck density had to be within the normal range (± 2 standard deviations [SD]). All participants had normal liver and kidney function for entry into the study. Four hundred eighty-nine women satisfied the eligibility criteria and were enrolled into the study. Participants using estrogen were put on a 6-month washout period before the start of the study. The volunteers recruited for the study were free living, in good health, ambulatory, and willing to give written informed consent. One participant with suspected Paget's disease and one participant with undetermined smoking status were excluded from this analysis. The analysis was performed on the remaining 487 women. The study was approved by Creighton University Institutional Review Board. All participants signed an informed consent form. Smoking and Alcohol and Dietary Intake At baseline recruitment, participants were provided with a questionnaire to report their smoking and alcohol history. They were asked to furnish information on smoking status (current, past, and/or never), number of years smoked, and number of years stopped smoking. Using the above information, we categorized the women into current smokers, past smokers, or nonsmokers. Furthermore, the smokers were classified into light (<1 pack/day) or heavy (>1 pack/day) smokers. The number of pack-years smoked was also calculated for smokers and past smokers (packs/day × number of years smoked). Comparisons were made between nonsmokers, past smokers, and current smokers and between nonsmokers, past smokers, light smokers, and heavy smokers. Dietary intake data were collected using 7-day food dairies. The average daily calcium, vitamin D, and caffeine intakes were calculated by a dietician, using the Food Processor II Plus nutrition and diet analysis system (version 5.1; ESHA Research, Salem, OR). The information on alcohol consumption was also obtained by a self-administered questionnaire. Drinking and nondrinking status was noted. Reproductive history, medical history, and present use of medications and vitamin and mineral supplements were also assessed by a questionnaire. Biochemical Analysis Fasting blood was collected from each participant. Blood samples were allowed to clot and were then centrifuged at 4°C for 15 minutes at 2056 × g to separate serum. All samples were stored frozen at −70°C until analysis. Serum 25-hydroxyvitamin D (calcidiol, 25OHD) was assayed by a competitive protein binding assay (17). The procedure involves initial chromatographic extraction and purification of serum on Sep-Pak cartridges (Waters Associates, Milford, MA) (18). The limit of detection was 12.5 mmol/L (5 ng/mL), and the interassay variation was 5%. Serum 1,25(OH)2 D was measured by a nonequilibrium radioreceptor assay (Incstar Corp., Stillwater, MN) using the calf thymus receptor, after extraction and purification of the serum on nonpolar C18OH octadecysilanol silica cartridge as described (11,16). The limit of detection for the assay was 12 pmol/L, and our interassay variation was 10%. Serum intact PTH was measured with the Allegro immunoradiometric assay (Nichols Institute, San Juan Capistrano, CA) (19). The limit of detection for the assay was 1 ng/L, and our interassay variation was 3.5%. Serum total estradiol, testosterone, and sex hormone binding globulin (SHBG) were measured in fasting baseline serum samples by radioimmunoassay kits obtained from Diagnostic Systems Laboratories (Webster, TX). Serum total estradiol was measured using an ultrasensitive assay. The minimum detection limit for these assays was 2.2 pg/mL for serum estradiol, 0.05 ng/mL for serum testosterone, and 3 nmol/L for serum SHBG. The intraassay coefficient was less than 5% for these assays. The interassay coefficient for the last 10 assays of serum total estradiol for the lowest standard used (5 pg/mL) was 8.4%, and the interassay coefficient for the in-house serum control (mean ± SD, 11.8 ± 1.35 pg/mL) used was 11.4%. The interassay coefficient for the low kit control (mean ± SD, 0.55 ± 0.03 ng/mL) and in-house control (mean ± SD, 0.33 ± 0.029 ng/mL) tested for serum total testosterone were 6% and 9%, respectively. The interassay coefficient for the lowest kit control tested for SHBG was 8.5%. The bioavailable (non-SHBG-bound) serum estradiol and serum testosterone were measured as described by Khosla and colleagues (20). Briefly, tracer amounts of 3H-estradiol or 3H-testosterone were added to aliquots of serum (200 μL), made to 500 μL with saline. To this, an equal volume of saturated solution of ammonium sulfate (final concentration, 50%) was added to precipitate the SHBG with its bound steroid. The SHBG-bound and -unbound steroid were separated by centrifugation at 1100 × g for 30 minutes at 4°C. The percentage of labeled estradiol or testosterone remaining in the supernatant (free and albumin-bound fractions) was then calculated. The bioavailable estradiol or testosterone concentration was obtained by multiplying the total estradiol or testosterone concentrations, as determined by radioimmunoassay, by the fraction that was non-SHBG-bound. All women except one had measurable levels of serum estradiol. Serum testosterone levels were undetectable in about 22 women. Physical Performance Measurements Physical performance measures of muscle function were evaluated by tests of muscle strength (grip strength, right hand), agility and coordination (timed rising), and gait and balance (timed walk, normal and fast). These tests measure physical tasks important for performing the common activities of daily living, and are easily performed in a clinical setting with simple equipment. Grip strength was measured in the right hand by a handheld Jamar dynamometer (Jackson, MI) as described, with some modification (21) and used by others (22). The participant was asked to stand with the right arm flexed at 90° (anterior–posterior plane) and parallel to the floor. The elbow was fully extended, and the wrist and forearm were in neutral anatomical position. In this position, the participants were instructed to squeeze the dynamometer (setting 2) as hard as possible. The average of three trials was recorded (measured in kilograms). Timed walk at normal and brisk pace was tested as described, with some modifications (23,24). Walking speed was assessed as the time taken to walk a 15-foot walkway at the participant's normal pace and then as quickly as possible. Timed rise test involved the time taken by the participant to rise from a chair 3 times as quickly as possible, following the protocol described, with some changes (24,25). The participants sat on a chair with no arm rests. They were asked to stand up and sit down 3 times and were timed with a stopwatch. In addition to these measurements, Physical Activity Scale for the Elderly (PASE) score was calculated based on questionnaire items assessing various domains of physical activity, including walking, sitting activities, light sport and recreational activities, moderate sport and recreational activities, strenuous sport and recreational activities, exercise, and light housework (26,27). Total Body Fat and Lean Muscle Mass Total body fat and lean muscle mass (defined as the weight of the muscle and organs and excluding total body fat and bone weight) were obtained from total body dual-energy absorptiometry scans of the study participants. Statistical Analysis All analyses were performed by the SAS statistical package (version 9.1; SAS Inc., Chicago, IL). Baseline patient characteristics (age; height; weight; dietary calcium, caffeine, and vitamin D intake; alcohol use; total body fat; lean muscle mass; medication use; and comorbid conditions) were compared by smoking status (nonsmokers vs past smokers vs current smokers) with one-way ANOVA for continuous variables and chi-square tests for categorical variables. The biochemical variables in Table 1 were compared by smoking status with ANCOVA while adjusting for significant baseline characteristics. If the overall F tests from the ANOVA or ANCOVA models were significant at a.05 level, then pair-wise comparisons were conducted and Tukey's method was used to adjust p values for multiple comparisons. Univariate comparisons of the physical performance variables were conducted with ANOVA. Multivariate analyses of the physical performance variables were conducted by ANCOVA, examining the effect of smoking status on timed rise, timed walk (normal), timed walk (fast), and grip strength, while adjusting for significant confounding variables including age, body mass index, calcium intake, caffeine intake, and total body fat. Backward selection was used to select significant confounders to adjust for while retaining smoking status in the model. An alpha level of 0.05 was used for model selection. The biochemical variables [serum 25OHD, 1,25(OH)2D, total estradiol, bioavailable estradiol, testosterone, and SHBG] were examined in the ANCOVA models to see if the effect of smoking is mediated through these biochemical variables. The biochemical variable was considered a mediator when it was a significant predictor of the physical performance tests and affected the association of smoking and physical performance measures (β coefficient and significance of smoking). The interaction of smoking status with serum 25OHD, 1,25(OH)2D, total estradiol, bioavailable estradiol, testosterone, and SHBG was also examined. An age related decline was calculated for each of the physical performance variables, using the slope of the regression coefficient for age (β) in Table 2, a 5-year increase in age corresponds to a 5 × β change in physical performance. This age-related decline was then compared to the decline in physical performance seen in smokers (as determined by the regression coefficient for smoking status) as a percentage change, using the change in physical performance tests of nonsmokers as the reference value. Results Characteristics of Study Population A comparison of the characteristics among nonsmokers, past smokers, and current smokers is given in Table 1. Twelve percent of the study population were current smokers, 33% were past smokers, and 55% constituted the nonsmokers. The median pack-years for smokers was 25.9 years (0.15–174 years), and for past smokers was 12 years (0.02–84.0 years). The median time that past smokers stopped smoking was 16 years (0.5–55 years). The median time that current smokers smoked was 50 years (1–64 years), and the median time since past smokers smoked was 20 years (0.6–55 years). On average, smokers weighed 8 kg less than nonsmokers (p <.001) and had lower dietary calcium intake compared to both nonsmokers and past smokers. Smokers had significantly (p <.001) lower total body fat compared to nonsmokers and past smokers. Amount of lean muscle mass was not significantly different between the groups. Past and current smokers had significantly higher caffeine intakes (p <.001) compared to nonsmokers. The percentage of alcohol consumers was slightly higher in smokers compared to nonsmokers (Table 1). Smokers were similar to nonsmokers with respect to age, height, and dietary vitamin D intake. The occurrence of comorbidities classified as 0–2, 3–4, 5–6, and ≥7 was not significantly different between the nonsmokers, past smokers, and current smokers. Medication use in the study population classified as 0, 1, and > 1 was not different between nonsmokers, past smokers, and current smokers (Table 1). Smoking and Biochemical Variables As seen in Table 1, serum 25OHD, 1,25(OH)2D, PTH, total estradiol, and bioavailable estradiol were lower in current smokers than in nonsmokers or past smokers, although the differences were not statistically significant. When the data were examined according to quartiles of 1,25(OH)2D, the percentage of smokers was marginally higher in the lowest quartile (Q1 = 16%) compared to higher quartiles (Q2 = 7%, Q3 = 12%, and Q4 = 13%). In univariate analyses, serum bioavailable estradiol was significantly lower (p =.034) in smokers than in nonsmokers and past smokers; adjustment for total body fat made the differences nonsignificant. Serum total testosterone levels were significantly higher in current smokers than in nonsmokers and past smokers. Serum bioavailable testosterone levels were lower in smokers than in nonsmokers and past smokers. Serum SHBG levels were significantly higher in current smokers than in nonsmokers or past smokers. Smoking and Physical Performance Measures As shown in Table 1 and Figure 1, on average, current smokers were slower on timed rise and timed walk tests and had decreased grip strength compared to past and nonsmokers after adjusting for significant confounders: age, body mass index, calcium intake, caffeine intake, and total body fat. Mean PASE scores (± standard error [SE]) were not different between the current (116.3 ± 7.4), past (110.9 ± 4.1), and nonsmokers (117.7 ± 3.5). From multivariate analysis models, the effect of current smoking on physical performance measures was much higher than that observed with a 1-year age-related decline in physical function in nonsmokers (Table 2). The effect of smoking (smokers vs nonsmokers) compared to a 5-year increase in age was higher by 39% on the timed rise test, 136% on the timed walk (normal) test, 95% on the timed walk (fast) test, and 90% on the grip strength test. Past smokers had physical performance measures similar to those of nonsmokers (Table 1 and Figure 1), except in those participants with a long pack-year history who were slower in the timed walk test (data not given). In contrast, no such effect was seen when the physical performance measures were compared according to pack-years of smoking in current smokers (data not given). There was no significant difference in physical performance measures between the light (<1 pack/day) and heavy (>1 pack/day) smokers (data not given). The interactions of smoking status with serum 25OHD, serum 1,25(OH)2D, serum SHBG, serum total estradiol, and serum bioavailable estradiol on timed rising, timed walk (normal), timed walk (fast), and grip strength were not statistically significant (all p >.05). Serum 25OHD, 1, 25(OH)2D, total estradiol, bioavailable estradiol, testosterone, and SHBG were examined in the ANCOVA models with smoking status while adjusting for age and total body fat (for serum total and bioavailable estradiol, the models were examined without total body fat) to see if the effect of smoking is mediated through these biochemical variables. When serum 1,25(OH)2D was categorized into quartiles, it was a significant predictor of grip strength (p =.026) and a marginally significant predictor of timed rise (p =.06) and timed walk normal (p =.07) (Table 2). The effect of smoking was still significant in models with serum 1,25(OH)2D, though slightly lower than that seen in models without 1,25(OH)2D. Serum 25OHD, total estradiol, bioavailable estradiol (Table 2), and SHBG were not significant predictors of physical performance measures. However, serum total testosterone was found to be a significant predictor of grip strength. Discussion The results presented in this cross-sectional study demonstrate that smoking reduces physical performance in elderly postmenopausal women. Current smokers were slower on timed rise and timed walk tests, and had decreased grip strength. The effect of smoking is equivalent to the normal age-related decline in physical performance tests of 7–11 years. The grip strength measurement in nonsmokers (65–77 years) in the present study (25.2 ± 4.41 kg) is close to the normative data reported in the literature for women between the ages of 60–79 years (24.5 ± 5.0 kg) (28). In the present study, past smokers had physical performance measures similar to those of nonsmokers, apart from the timed walk test, which was slower in those women with more pack-years of smoking. There was no significant association between the quantity of smoking and physical performance measures. There are few studies in the literature that report that smoking is associated with a decrease in physical performance. The results of the present study are in agreement with those of Nelson and colleagues (13), who demonstrated that current smoking is associated with impaired neuromuscular performance. Also, Guralnik and Kaplan (14) reported a higher physical function in nonsmokers than in smokers after 19 years of follow-up. In the elderly and disabled Medicare population, Arday and colleagues (29) reported that current smokers have worse physical and mental functional status than do never smokers. Stovring and colleagues (30) reported that the cumulative effect of smoking from age 50–70 years adversely affects the functional ability at age 75 years. None of these studies suggested the mechanism of how smoking affects physical function. Changes in vitamin D metabolism may play a role in decreased physical performance. 1,25 dihydroxyvitamin D receptor (VDR) is expressed in skeletal muscle tissue, and evidence indicates that 1,25(OH)2D plays a key role in muscle contractility, cell calcium and phosphate transport, and protein and phospholipid synthesis in muscle cells (31). VDR in muscle tissue has been reported to decrease with age (32). Several cross-sectional studies have shown that low serum 1,25(OH)2D and low serum 25OHD levels are related to lower muscle strength and falls in older men and women (2,3). In healthy elderly women, we have shown decreased physical performance in the lower quartiles of serum 1,25(OH)2D and 25OHD compared to the corresponding higher quartiles (33). Vitamin D and calcium supplementation has been demonstrated to improve musculoskeletal function and decrease fall incidence in the elderly population (5,34). In the present study, smokers have marginally lower serum 25OHD and serum 1,25(OH)2D levels. In multivariate analysis, serum 1,25(OH)2D examined as quartiles was a predictor of physical performance measures and lessened the effect of smoking, indicating that some of the effect of smoking is mediated through changes in serum 1,25(OH)2D. Although serum 25OHD levels were lower in smokers, adjustment for serum 25OHD did not affect the association of smoking and physical performance measures. In the present study, smokers had lower serum total and bioavailable estradiol, lower serum bioavailable testosterone, and higher serum SHBG levels. In univariate analyses, serum bioavailable estradiol was significantly lower in smokers compared to nonsmokers and past smokers. It is known that adipose tissue produces estrogen by aromatase conversion of androstenedione in postmenopausal women (12), so the lower body weight in smokers is a probable cause of lower circulating serum estradiol; smoking has also been suggested to increase the metabolism of estrogen (12). However, after adjustment for total body fat, the significance of estradiol was lost. Furthermore, in multivariate analyses, total estradiol and bioavailable estradiol (in absence of total body fat in the models) were not significant predictors of physical performance tests, thus the effect of smoking may not be mediated through changes in estrogen metabolism. Total body fat was a significant independent predictor of physical performance tests other than grip strength; heavier participants were slower on timed walking and chair rise tests. There was no difference in lean muscle mass in smokers compared to nonsmokers, so the loss of muscle mass is not a factor in decreased physical performance. The reason for an effect of smoking on total and bioavailable testosterone is not very clear, but it has been suggested that a decrease in adrenal production of androgens may be responsible (12). Serum total testosterone was a significant predictor of grip strength in the present study, but the mechanism of how it affects physical performance is not very clear. Decreased testosterone levels in men were reported to be associated with lower appendicular skeletal muscle mass and increased risk of falls, impairment of balance, and impaired ability to perform the tandem walk (35). Testosterone supplementation in elderly men has been reported to improve muscle mass and strength (35). Testosterone administration in men was associated with a dose-dependent increase in leg-press strength and leg power, which correlated with testosterone dose and circulating testosterone concentrations (36). However, Snyder and colleagues (37) reported no effect of testosterone on muscle strength in older men. In one controlled study of low weight HIV-infected young women, testosterone supplementation was associated with increase in shoulder flexion, elbow flexion, knee extension, and knee flexion, but not in grip strength (38); however, no corresponding studies have been reported in elderly women. In the present study, the interactions of smoking with serum 25OHD, 1,25(OH)2D, total estradiol, bioavailable estradiol, testosterone, and SHBG on physical performance measures were not significant. This could be because our study was not adequately powered to assess these interactions. The present study has some limitations. The study is cross-sectional in nature. The women enrolled into the study were healthy volunteers, and our findings may not apply to the general population. The smoking status of women in the present study is based on self-report and possibly underestimates the true use. Overall, smoking affects physical performance in part through its effect on 1,25(OH)2D status. However, an independent effect of smoking on physical performance by affecting vascular function (39) or by a direct effect on muscle (40,41) is a possibility. In summary, our results suggest that current smoking is a risk factor for decreased muscle strength leading to decreased physical performance in elderly women. Decision Editor: Luigi Ferrucci, MD, PhD Figure 1. Mean baseline physical performance measures according to smoking status. The values are unadjusted mean ± standard error of the mean. Means were compared by analysis of covariance adjusting for significant confounding variables that were significant in the models and included the covariates given in Table 2. Significance was derived from adjusted data using Tukey's post hoc multiple comparison test. *p <.05 compared to nonsmokers; †p <.05 compared to past smokers Table 1. Characteristics and Biochemical Variables According to Smoking Status. Parameter Nonsmokers Past Smokers Current Smokers N 268 163 56 Age, y 71.7 ± 0.22 71.4 ± 0.29 70.6 ± 0.42 Height, cm 159.2 ± 0.40 159.2 ± 0.53 159.5 ± 0.67 Weight, kg 69.5 ± 0.74 69.5 ± 1.02 61.3 ± 1.60* Total body fat, kg 29.0 ± 5.55 28.9 ± 7.47 22.2 ± 11.3*† Lean muscle mass, kg 37.4 ± 2.54 37.3 ± 3.34 36.2 ± 6.02 Dietary calcium intake, mg/d 755.1 ± 17.3 727.9 ± 25.7 703.5 ± 49.8 Dietary caffeine intake, mg/d 225.4 ± 10.6 294.9 ± 15.4* 420.6 ± 54.6* Dietary vitamin D intake, μg/d 3.68 ± 0.13 3.30 ± 0.16 3.08 ± 0.31 Alcohol drinkers, N (%) 66 (25) 73 (45) 24 (43) Medication use 0 108 (40%) 60 (37%) 19 (34%) 1 78 (29%) 52 (32%) 21 (38%) >1 82 (31%) 51 (31%) 16 (29%) Comorbidities 0–2 44 (16%) 25 (15%) 8 (14%) 3–4 103 (38%) 62 (38%) 26 (46%) 5–6 80 (30%) 45 (28%) 11 (20%) ≥ 7 41 (15%) 31 (19%) 11 (20%) Serum 25OHD, ng/mL 31.7 ± 0.61 31.6 ± 0.75 28.6 ± 1.91 Serum 1,25(OH)2D, pg/mL 34.8 ± 0.51 34.0 ± 0.59 33.9 ± 1.05 Serum PTH, pg/mL 37.9 ± 0.94 36.2 ± 1.09 35.8 ± 1.71 Serum total estradiol, pg/mL 12.34 ± 0.34 11.63 ± 0.46 10.96 ± 0.65 Serum bioavailable estradiol, pg/mL 3.71 ± 0.13 3.50 ± 0.20 2.82 ± 0.21 Serum SHBG, nmol/L 139.4 ± 3.9 143.0 ± 5.8 197.4 ± 12.2*† Serum total testosterone,ng/mL 0.24 ± 0.007 0.23 ± 0.011 0.27 ± 0.018† Serum bioavailable testosterone, pg/mL 27.08 ± 1.03 26.45 ± 1.83 24.88 ± 1.94 Notes: Values are unadjusted means and standard error of means. Characteristics were compared by ANOVA with Tukey's method for post hoc analysis. Categorical variables were compared by chi-square tests. Biochemical variables were compared by ANCOVA adjusting for significant confounding variables including age, body mass index, calcium intake, and total body fat. Significance was derived from adjusted data using Tukey's post hoc multiple comparison test. *p <.05 compared to nonsmokers. †p <.05 compared to past smokers. Table 2. Multivariate ANCOVA Models With Physical Performance Measure Outcomes, Looking at Possible Mediator Variables. Model 1 Model 2 Model 3 Model 4 Outcome Predictor Variable β ± SE p Value β ± SE p Value β ± SE p Value β ± SE p Value Timed rise, s Serum 1,25(OH)2D Quartile 1 NI 0.37 ± 0.43 .062 NI NI Quartile 2 –0.45 ± 0.43 Quartile 3 –0.70 ± 0.43 Quartile 4 Ref Serum total estradiol (1 pg/mL increase) NI 0.003 ± 0.031 .92 NI Serum bioavailable estradiol (1 pg/mL increase) NI NI –0.008 ± 0.076 .92 Total body fat (1 kg increase) 0.000051 ± 0.000017 .0011 0.000056 ± 0.000017 .0008 NI NI Age (1 y increase) 0.20 ± 0.042 <.0001 0.19 ± 0.042 <.0001 0.16 ± 0.044 .0005 0.16 ± 0.044 .0005 Smoking status Nonsmoker Ref Ref Ref Ref Past smoker –0.047 ± 0.33 .016 –0.082 ± 0.33 .025 –0.001 ± 0.35 .3 –0.005 ± 0.35 .32 Smoker 1.39 ± 0.50* 1.30 ± 0.50* 0.80 ± 0.53 0.78 ± 0.54 Timed walk (normal), s Serum 1,25(OH)2D Quartile 1 NI 0.035 ± 0.13 .072 NI NI Quartile 2 –0.060 ± 0.13 Quartile 3 −0.29 ± 0.13 Quartile 4 Ref Serum total estradiol (1 pg/mL increase) NI –0.005 ± 0.010 .63 NI Serum bioavailable estradiol (1 pg/mL increase) NI NI –0.019 ± 0.025 .46 Total body fat (1 kg increase) 0.000021 ± 0.0000052 <.0001 0.000021 ± 0.0000052 <.0001 NI NI Age (1 y increase) 0.055 ± 0.013 <.0001 0.055 ± 0.013 <.0001 0.049 ± 0.015 .0008 0.049 ± 0.015 .001 Smoking status Nonsmoker Ref Ref Ref Ref Past smoker 0.075 ± 0.10 .0002 0.063 ± 0.10 .0003 0.069 ± 0.11 .037 0.069 ± 0.11 .045 Smoker 0.65 ± 0.16** 0.64 ± 0.16** 0.45 ± 0.17* 0.44 ± 0.18* Timed walk (fast), s Serum 1,25(OH)2D Quartile 1 NI 0.13 ± 0.10 .17 NI NI Quartile 2 0.083 ± 0.10 Quartile 3 –0.085 ± 0.10 Quartile 4 Ref Serum total estradiol (1 pg/mL increase) NI NI 0.0024 ± 0.008 .75 NI Serum bioavailable estradiol (1 pg/mL increase) NI NI NI –0.010 ± 0.019 .6 Total body fat (1 kg increase) 0.000011 ± 0.0000040 .0058 0.000011 ± 0.0000040 .0071 NI NI Age (1 y increase) 0.039 ± 0.01 .0002 0.039 ± 0.01 .0002 0.036 ± 0.011 .0012 0.036 ± 0.011 .0013 Smoking status Nonsmoker Ref Ref Ref Ref Past smoker 0.077 ± 0.080 .0076 0.067 ± 0.080 .011 0.098 ± 0.086 .066 0.095 ± 0.086 .084 Smoker 0.38 ± 0.12* 0.37 ± 0.12* 0.30 ± 0.09 0.28 ± 0.13 Grip strength, kg Serum 1,25(OH)2D* Quartile 1 NI –1.20 ± 0.60 .026 NI NI Quartile 2 –0.30 ± 0.60 Quartile 3 0.57 ± 0.59 Quartile 4 Ref Serum total estradiol (1 pg/mL increase) NI NI 0.012 ± 0.048 .81 NI Serum bioavailable estradiol (1 pg/mL increase) NI NI NI 0.095 ± 0.12 .42 Age (1 y increase) –0.20 ± 0.060 .0008 –0.20 ± 0.060 .0007 –0.20 ± 0.060 .001 –0.20 ± 0.060 .0011 Serum testosterone (1 ng/mL increase) 5.39 ± 1.87 .0042 5.11 ± 1.87 .0066 5.57 ± 2.18 .011 5.03 ± 2.15 .02 Smoking status Nonsmoker Ref Ref Ref Ref Past smoker 0.50 ± 0.47 .0046 0.51 ± 0.46 .0073 0.54 ± 0.47 .0045 0.54 ± 0.47 .0067 Smoker –1.99 ± 0.71* –1.87 ± 0.71* –2.0 ± 0.73* –1.92 ± 0.73* Notes: β coefficient values are from multivariate models including significant predictors out of age, body mass index, total calcium intake, caffeine intake, total body fat, sex hormone binding globulin (SHBG), serum 25 hydroxy vitamin D (25OHD), serum 1,25 –dihydroxyvitamin D (1,25(OH)2D), serum total estradiol, serum bioavailable estradiol, serum total and bioavailable testosterone, and smoking status (smoker vs nonsmoker or past smoker). Model 1 looks at only significant predictors; Model 2 looks at significant predictors + serum 1,25(OH)2D; Model 3 looks at significant predictors + serum total estradiol without total body fat; Model 4 looks at significant predictors + serum bioavailable estradiol without total body fat. * p <.01 compared to nonsmokers. ** p <.0001 compared to nonsmokers. ANCOVA = Analysis of covariance; SE = standard error; NI = Not included in the model. Table 3. Co-morbidities in the Study Population. Non-smokers (n = 268) Past Smokers (n = 163) Smokers (n = 56) p value Hypertension 107 (40%) 52 (32%) 14 (25%) 0.052 Anxiety 13 (5%) 10 (6%) 6 (11%) 0.21 Depression 9 (3%) 8 (5%) 1 (2%) 0.61 Glaucoma 17 (6%) 7 (4%) 1 (2%) 0.4 Asthma 6 (2%) 6 (4%) 1 (2%) 0.61 Headaches 16 (6%) 8 (5%) 2 (4%) 0.88 Osteoporosis 17 (6%) 20 (13%) 5 (9%) 0.078 Scoliosis 11 (4%) 8 (5%) 3 (6%) 0.86 Diabetes 13 (5%) 11 (7%) 1 (2%) 0.38 Hypothyroid 50 (19%) 30 (18%) 9 (16%) 0.94 Cancer 10 (4%) 10 (6%) 6 (11%) 0.095 Joint Disease (Arthritis) 101 (38%) 53 (33%) 15 (27%) 0.24 Chronic GI Disease 50 (19%) 36 (22%) 13 (24%) 0.54 Heart Disease 46 (17%) 38 (24%) 11 (20%) 0.25 Lung Disease 11 (4%) 17 (11%) 5 (9%) 0.026 Hip fracture 4 (2%) 6 (4%) 3 (5%) 0.14 The values given are absolute numbers with the percentages in the parentheses. Chi-square test was used to determine the differences between the groups. Table 4. Medication use in Study Population. Non-smoker (n = 268) Past Smoker (n = 163) Smoker (n = 56) p value Thiazide Diuretics 42 (16%) 18 (11%) 5 (9%) 0.26 Loop Diuretics 41 (15%) 20 (12%) 5 (9%) 0.42 Thyroid Preparations 47 (18%) 28 (17%) 7 (13%) 0.72 Sedatives/Anticonvulsants/Narcotics 10 (4%) 8 (5%) 7 (12%) 0.14 Antidepressants 9 (3%) 10 (6%) — 0.099 Tranquilizer 6 (2%) 5 (3%) 4 (7%) 0.13 NSAIDs 45 (17%) 21 (13%) 6 (11%) 0.39 Aspirin 49 (18%) 35 (21%) 10 (18%) 0.7 Antihistamine 14 (5%) 10 (6%) 7 (13%) 0.14 Cholesterol Med 23 (9%) 14 (9%) 1 (2%) 0.2 The values given are absolute numbers with the percentages in the parentheses. Chi-square test was used to determine the differences between the groups. This work was supported by National Institutes of Health Research Grants UO1-AG10373 and RO1-AG10358. We thank Karen A. Rafferty for her help in food diary data collection and analysis. We also thank Kurt E. Balhorn for the laboratory analysis. This work was presented as an abstract at 27th Annual Meeting of the American Society for Bone and Mineral Research, Nashville, Tennessee, September 23–27, 2005. References 1 Moreland JD, Richardson JA, Goldsmith CH, Clase CM. Muscle weakness and falls in older adults: a systematic review and meta-analysis. J Am Geriatr Soc.2004;52:1121-1129. 2 Bischoff HA, Stahelin HB, Urscheler N, et al. Muscle strength in the elderly: its relation to vitamin D metabolites. Arch Phys Med Rehabil.1999;80:54-58. 3 Dhesi JK, Bearne LM, Moniz C, et al. Neuromuscular and psychomotor function in elderly subjects who fall and the relationship with vitamin D status. J Bone Miner Res.2002;17:891-897. 4 Dhesi JK, Jackson SH, Bearne LM, et al. Vitamin D supplementation improves neuromuscular function in older people who fall. Age Ageing.2004;33:589-595. 5 Pfeifer M, Begerow B, Minne HW, Abrams C, Nachtigall D, Hansen C. Effects of a short-term vitamin D and calcium supplementation on body sway and secondary hyperparathyroidism in elderly women. J Bone Miner Res.2000;15:1113-1118. 6 Cauley JA, Petrini AM, LaPorte RE, et al. The decline of grip strength in the menopause: relationship to physical activity, estrogen use and anthropometric factors. J Chronic Dis.1987;40:115-120. 7 Skelton DA, Phillips SK, Bruce SA, Naylor CH, Woledge RC. Hormone replacement therapy increases isometric muscle strength of adductor pollicis in post-menopausal women. Clin Sci (Lond).1999;96:357-364. 8 Kenny AM, Kleppinger A, Wang Y, Prestwood KM. Effects of ultra-low-dose estrogen therapy on muscle and physical function in older women. J Am Geriatr Soc.2005;53:1973-1977. 9 Taaffe DR, Luz VM, Delay R, Marcus R. Maximal muscle strength of elderly women is not influenced by oestrogen status. Age Ageing.1995;24:329-333. 10 Kapoor D, Jones TH. Smoking and hormones in health and endocrine disorders. Eur J Endocrinol.2005;152:491-499. 11 Rapuri PB, Gallagher JC, Balhorn KE, Ryschon KL. Smoking and bone metabolism in elderly women. Bone.2000;27:429-436. 12 Tanko LB, Christiansen C. An update on the antiestrogenic effect of smoking: a literature review with implications for researchers and practitioners. Menopause.2004;11:104-109. 13 Nelson HD, Nevitt MC, Scott JC, Stone KL, Cummings SR. Smoking, alcohol, and neuromuscular and physical function of older women. Study of Osteoporotic Fractures Research Group. JAMA.1994;272:1825-1831. 14 Guralnik JM, Kaplan GA. Predictors of healthy aging: prospective evidence from the Alameda County study. Am J Public Health.1989;79:703-708. 15 LaCroix AZ, Guralnik JM, Berkman LF, Wallace RB, Satterfield S. Maintaining mobility in late life. II. Smoking, alcohol consumption, physical activity, and body mass index. Am J Epidemiol.1993;137:858-869. 16 Gallagher JC, Fowler SE, Detter JR, Sherman SS. Combination treatment with estrogen and calcitriol in the prevention of age-related bone loss. J Clin Endocrinol Metab.2001;86:3618-3628. 17 Haddad JG, Chyu KJ. Competitive protein-binding radioassay for 25-hydroxycholecalciferol. J Clin Endocrinol Metab.1971;33:992-995. 18 Reinhardt TA, Horst RL. Simplified assays for the determination of 25-OHD, 24,25-(OH)2D and 1,25-(OH)2D. In: Norman AW, Schaefer K, Grigoleit HG, Herrath DV, eds. Vitamin D, Molecular, Cellular and Clinical Endocrinology. Berlin and New York: Walter de Gruyter & Co.; 1988:720–726. 19 Nussbaum SR, Zahradnik RJ, Lavigne JR, et al. Highly sensitive two-site immunoradiometric assay of parathyrin, and its clinical utility in evaluating patients with hypercalcemia. Clin Chem.1987;33:1364-1367. 20 Khosla S, Melton LJ, III, Atkinson EJ, O'Fallon WM, Klee GG, Riggs BL. Relationship of serum sex steroid levels and bone turnover markers with bone mineral density in men and women: a key role for bioavailable estrogen. J Clin Endocrinol Metab.1998;83:2266-2274. 21 Fess EE. Grip Strength. Clinical Assesment Recommendations. Chicago, IL: American Society of Hand Therapists; 1992:41–45. 22 Lindsey C, Brownbill RA, Bohannon RA, Ilich JZ. Association of physical performance measures with bone mineral density in postmenopausal women. Arch Phys Med Rehabil.2005;86:1102-1107. 23 Bohannon RW. Comfortable and maximum walking speed of adults aged 20-79 years: reference values and determinants. Age Ageing.1997;26:15-19. 24 Judge JO, Lindsey C, Underwood M, Winsemius D. Balance improvements in older women: effects of exercise training. Phys Ther.1993;73:254-262. 25 Bohannon RW. Sit-to-stand test for measuring performance of lower extremity muscles. Percept Mot Skills.1995;80:163-166. 26 Washburn RA, Smith KW, Jette AM, Janney CA. The Physical Activity Scale for the Elderly (PASE): development and evaluation. J Clin Epidemiol.1993;46:153-162. 27 Washburn RA, McAuley E, Katula J, Mihalko SL, Boileau RA. The physical activity scale for the elderly (PASE): evidence for validity. J Clin Epidemiol.1999;52:643-651. 28 Desrosiers J, Bravo G, Hebert R, Dutil E. Normative data for grip strength of elderly men and women. Am J Occup Ther.1995;49:637-644. 29 Arday DR, Milton MH, Husten CG, et al. Smoking and functional status among Medicare managed care enrollees. Am J Prev Med.2003;24:234-241. 30 Stovring N, Avlund K, Schultz-Larsen K, Schroll M. The cumulative effect of smoking at age 50, 60, and 70 on functional ability at age 75. Scand J Public Health.2004;32:296-302. 31 Boland R, de Boland AR, Marinissen MJ, Santillan G, Vazquez G, Zanello S. Avian muscle cells as targets for the secosteroid hormone 1,25-dihydroxy-vitamin D3. Mol Cell Endocrinol.1995;114:1-8. 32 Bischoff-Ferrari HA, Borchers M, Gudat F, Durmuller U, Stahelin HB, Dick W. Vitamin D receptor expression in human muscle tissue decreases with age. J Bone Miner Res.2004;19:265-269. 33 Rapuri PB, Gallagher JC, Smith LM. Higher serum 1,25-dihydroxy vitamin D and 25-hydroxy vitamin D levels are associated with better physical performance in elderly postmenopausal women (abstract). J Bone Miner Res.2005;20:(Suppl 1): S386. 34 Bischoff HA, Stahelin HB, Dick W, et al. Effects of vitamin D and calcium supplementation on falls: a randomized controlled trial. J Bone Miner Res.2003;18:343-351. 35 Herbst KL, Bhasin S. Testosterone action on skeletal muscle. Curr Opin Clin Nutr Metab Care.2004;7:271-277. 36 Storer TW, Magliano L, Woodhouse L, et al. Testosterone dose-dependently increases maximal voluntary strength and leg power, but does not affect fatigability or specific tension. J Clin Endocrinol Metab.2003;88:1478-1485. 37 Snyder PJ, Peachey H, Hannoush P, et al. Effect of testosterone treatment on body composition and muscle strength in men over 65 years of age. J Clin Endocrinol Metab.1999;84:2647-2653. 38 Dolan S, Wilkie S, Aliabadi N, et al. Effects of testosterone administration in human immunodeficiency virus-infected women with low weight: a randomized placebo-controlled study. Arch Intern Med.2004;164:897-904. 39 Kilaru S, Frangos SG, Chen AH, et al. Nicotine: a review of its role in atherosclerosis. J Am Coll Surg.2001;193:538-546. 40 Larsson L, Orlander J. Skeletal muscle morphology, metabolism and function in smokers and non-smokers. A study on smoking-discordant monozygous twins. Acta Physiol Scand.1984;120:343-352. 41 Orlander J, Kiessling KH, Larsson L. Skeletal muscle metabolism, morphology and function in sedentary smokers and nonsmokers. Acta Physiol Scand.1979;107:39-46.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B09EA51EB71D2482A6296AB120DD0363B40F44F.txt b/test/dataset/in/resources/corpus/Clean_0B09EA51EB71D2482A6296AB120DD0363B40F44F.txt
deleted file mode 100644
index f5a1719..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B09EA51EB71D2482A6296AB120DD0363B40F44F.txt
+++ /dev/null
@@ -1 +0,0 @@
-amjepidajeAmerican Journal of Epidemiology1476-62560002-9262Oxford University Press10.1093/aje/kwp032ORIGINAL CONTRIBUTIONSAre the Short-term Effects of Air Pollution Restricted to Cardiorespiratory Diseases?LarrieuSophieLefrancAgnèsGaultGaëlleChatignouxEdouardCouvyFranckJouvesBernardFilleulLaurentCorrespondence to Dr. Sophie Larrieu, French Institute of Public Health Surveillance (InVS), Cire Aquitaine, Espace Rodesse, 103 bis rue Belleville, BP952, 33063 Bordeaux Cedex, France (e-mail: sophie.larrieu@sante.gouv.fr).1552009242009169101201120818720082612009American Journal of Epidemiology © The Author 2009. Published by the Johns Hopkins Bloomberg School of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org.2009Short-term effects of air pollution on common morbidity are largely unknown. The authors explored links between daily levels of air pollution (nitrogen dioxide, ozone, and particulate matter less than 10 μm in diameter (PM10)) and medical home visits made for diverse reasons in Bordeaux, France, during 2000–2006. Daily numbers of visits were obtained from a network of general practitioners. The excess relative risk (ERR) of a visit for each indicator associated with increased pollutant levels was estimated by fitting a Poisson regression model, controlling for well-known confounding factors and temporal trends. Positive and significant associations were found between air pollution and most health indicators. A 10-μg/m3 increase in PM10 levels was associated with increases in visits for upper and lower respiratory diseases (ERRs were 1.5% (95% confidence interval (CI): 0.3, 2.7) and 2.5% (95% CI: 0.5, 4.4), respectively), headache and asthenia (ERR = 3.5%, 95% CI: 1.3, 5.9), and skin rash and conjunctivitis (ERR = 3.2%, 95% CI: −0.2, 6.8). Significant associations were also found between nitrogen dioxide and ozone and several health indicators. Distributed-lag models showed no harvesting effect, and some effects persisted up to 15 days after exposure increased. These results suggest that considering only the most severe effects of air pollution leads to underestimation of its impact on public health.air pollutionastheniaconjunctivitisexanthemaheadacheprimary health carerespiratory tract diseasesThe health effects of air pollution have been subject to intense epidemiologic and experimental study during the past 2 decades. It is now well-documented that day-to-day variations in air pollutant levels are associated with cardiorespiratory morbidity and mortality (1), and several biologic hypotheses have been proposed to explain air pollutant effects, mainly involving acute pulmonary inflammation and oxidative stress (2). These effects have been widely studied for particulate air pollution, but there is also substantial evidence concerning gaseous pollutants such as ozone and nitrogen dioxide. The exposure-risk relations observed are usually described as linear without any threshold (3, 4), or results suggest that the threshold above which no effect would be observed is far below the concentrations observed in urban areas (5). Furthermore, the severity of these effects is widely variable, and they can affect more or fewer people; indeed, short-term effects of air pollution are often represented as a pyramid, with the mildest but not uncommon effects at the bottom and the least common but more severe effects at the top (Figure 1).Figure 1.Pyramidal representation of the hypothesized short-term effects of air pollution on health.Most epidemiologic studies conducted to date have focused on severe events such as mortality, hospitalizations, and emergency room visits. On the other hand, more common effects have been less well described, mainly because of the unavailability of reliable indicators of primary-care setting, symptoms, or medication use. A few studies showed a significant association between air pollutant levels and general practitioner home visits in Paris, France (6), London, United Kingdom (7–9), and Hong Kong, China (10), but those studies focused on respiratory diseases; other symptoms have rarely been investigated.In Bordeaux, a mid-sized metropolitan area in southwestern France, a network of 60 general practitioners has been collecting data on the practitioners’ activities for many years, allowing us to link their activity with air pollutant levels measured over the Bordeaux area. Therefore, our aim in this study was to explore the links between daily levels of air pollution indicators (nitrogen dioxide, ozone, and particulate matter less than 10 μm in diameter (PM10)) and daily numbers of medical home visits made for nonsevere reasons, some of which are known to be associated with air pollution (respiratory diseases) and others of which have a less well described association with air pollution (headache, asthenia, conjunctivitis, and skin rash) in this area.MATERIALS AND METHODSStudy population and periodThis study was conducted during the period 2000–2006 in the area of Bordeaux, France. It included Bordeaux and 21 other neighboring cities where urbanization and background air pollutant levels could be considered homogeneous, representing a total population of approximately 600,000 inhabitants.Estimates of pollution exposureThe local air quality monitoring network, AIRAQ (http://www.airaq.asso.fr/), provided data on daily levels of ambient air pollutants measured within the study area: nitrogen dioxide, PM10, and ozone. We used measurements from the 4 background monitoring stations available in the study area, because they are not influenced by occasional sources of pollution, since they are located some distance away from major roads and industries.For each air pollutant, we calculated correlation coefficients for correlations between the different stations in order to ensure that they all measured the same type of pollution (r > 0.60); the daily level of each air pollution indicator was then computed as the arithmetic mean of the daily levels recorded by the stations (for PM10 and nitrogen dioxide) or the arithmetic mean of the daily maximum of 8-hour moving averages (for ozone).Health outcomesSOS Médecins is the primary emergency and health-care network in France, providing home medical visits by general practitioners in response to private house calls 24 hours a day, 7 days a week. In the urban area of Bordeaux, SOS Médecins comprises 60 general practitioners who make more than 400 visits per day, operating in an area of approximately 800,000 inhabitants. Characteristics concerning each visit are logged into a local database; all symptoms and/or complaints reported by the patients are coded and recorded according to the International Classification of Primary Care, Second Edition (11), as well as the final diagnosis. From this database, we extracted daily numbers of visits made to people living in the study area for the following reasons: upper and lower respiratory diseases, asthma, asthenia, headache, conjunctivitis, and skin rash (Table 1). In order to evaluate any bias in the analysis, we also extracted and modeled daily numbers of visits made for lumbago, which is unrelated to air pollution a priori, according to the same method.Table 1.Health Indicators and Corresponding ICPC-2 Codes Used in a Study of Air Pollutant Effects, Bordeaux, France, 2000–2006Health IndicatorCorresponding Syndrome(s) and/or Symptom(s)ICPC-2 Code(s)Upper respiratory diseasesTonsillitis, sinusitis, rhinitis, nasopharyngitis, pharyngitis, laryngitis, tracheitisR75, R76, R77, R83, R97Lower respiratory diseasesBronchitis, bronchiolitis, chronic obstructive pulmonary disease, coughR78, R79, R95, R05AsthmaAsthmaR96AstheniaAstheniaA04HeadacheHeadacheN01, N89ConjunctivitisConjunctivitisF70, F71Skin rashDermatitis, eczema, urticaria, skin rashS07, S87, S98Abbreviation: ICPC-2, International Classification of Primary Care, Second Edition (11).Potentially confounding factorsData on the following potentially confounding factors were collected: ambient temperature measured at the Bordeaux meteorologic station, obtained from the national meteorologic institute (Météo-France); periods of influenza epidemics, obtained from the SOS Médecins database (according to a local influenza epidemic threshold that usually allows monitoring of the disease in the area); pollen counts measured at the Bordeaux pollen monitoring station, obtained from the French surveillance system for pollen counts (Réseau National de Surveillance Aérobiologique); and dates of holidays, obtained from the French Ministry of Education.Statistical analysisDaily numbers of medical home visits were analyzed with time-series methods, using generalized additive Poisson regression models allowing for overdispersion (12). Each air pollution indicator was included in the model as a linear term, and different lags were tested: pollutant levels on the current day and up to 3 days before (lags of 0, 1, 2, and 3 days) and mean levels during the current day and the previous 1, 2, or 3 days (lags of 0–1, 0–2, and 0–3 days). The lag that minimized Akaike's Information Criterion (13) was retained. Adjustments were made for possible confounders—including long-term trends, seasonality, days of the week, holidays, minimum temperature on the current day and maximum temperature on the previous day, and influenza epidemics—following the methods of the APHEA-2 [Air Pollution And Health: A European Approach] Study (14). Long-term trends and seasonality were modeled using penalized cubic regression splines. The degree of smoothing of the spline function was chosen to remove seasonal and long-term temporal trends by minimizing the autocorrelation in the residuals. Dummy variables for days of the week and holidays were included as other independent variables. Temperature and influenza epidemics were modeled using natural splines with 3 degrees of freedom for each. The lack of residual autocorrelation was checked through the partial autocorrelation, and the Bartlett test was used to check that white noise was obtained.A term for interaction between ozone and season was introduced into the regression models to specifically assess the effect of ozone during the warmer months (April–September).All results are presented as the excess relative risk (ERR) of a medical home visit (expressed as a percentage) associated with a 10-μg/m3 increase in air pollution indicator level. To determine whether there were any effects in subgroups of the population, we also performed analyses on specific age groups, when the number of events was large enough, notably to estimate specific ERRs for frail populations such as children under age 15 years (15) and people aged 65 years or more (16).Since the health effects of air pollution may persist for several days after exposure or, on the contrary, potential morbidity displacement may exist (17), we also used distributed-lag models to better describe pollutant effects that were delayed by up to 15 days, as follows:with γ being constrained to follow a cubic polynomial function. As in the generalized additive Poisson regression models, the number of degrees of freedom was chosen to minimize residual autocorrelation.This model could not be used for ozone, because the term for interaction between ozone and season led to many breaks in the pollutant series.RESULTSThe study area comprised 22 cities and more than 600,000 inhabitants, of whom 15.5% were less than 15 years of age and 15.7% were aged 65 years or older. Descriptive results for daily air pollutant levels and health indicators are shown in Table 2.Table 2.Daily Air Pollutant Levels and Daily Numbers of General Practitioners’ Visits Made for Different Health Indicators, Bordeaux, France, 2000–2006Air Pollution or Health IndicatorMeanMinimum5th Percentile50th Percentile95th PercentileMaximumPollutant Level, μg/m3Nitrogen dioxide21.93.88.020.341.380.3Ozone69.32.732.568.2111.8142.7Ozone in spring and summera84.837.548.781.5117.1142.7PM1021.15.010.318.838.588.2No. of VisitsUpper respiratory diseases57.68255696149Lower respiratory diseases16.404143877Asthma3.5003819Asthenia2.2002511Headache4.1014818Conjunctivitis0.600028Skin rash3.5003815Lumbago (control)5.70251120Abbreviation: PM10, particulate matter less than 10 μm in diameter.aApril 1–September 30.Bordeaux is a moderately polluted area, with annual mean levels of nitrogen dioxide and PM10 that are largely below the current French annual guidelines (40 μg/m3) (18), but PM10 levels are slightly above the 2005 World Health Organization guidelines (20 μg/m3) (19). Daily maximum levels of 8-hour moving averages of ozone overtook the threshold for health protection (120 μg/m3) during less than 5% of the study period.During the 7-year study period, a total of 895,710 medical home visits were made by SOS Médecins Bordeaux, corresponding to a daily mean of 350 visits. Visits for respiratory diseases represented approximately 20% of the activity of the general practitioners. Since the numbers of visits for the 4 other indicators were quite low, we grouped together the indicators asthenia and headache, which correspond to a general health impairment without another diagnosed disease, and conjunctivitis and skin rash, which can be related to irritation and atopy.As Figure 2 shows, the general practitioners’ activity showed important day-of-the-week and seasonal variations. The daily number of visits increased on weekends and during the winter.Figure 2.Seasonal (main graph) and daily (inset) numbers of medical home visits (gray line) made to persons in the Bordeaux metropolitan area and 7-day moving averages (black line), Bordeaux, France, 2000–2006.We constructed specific models for each health indicator and each air pollution indicator. The mean of the lag 0–3 days led to the best model for all indicators except headache/asthenia, for which a lag of 0 was considered for every air pollution indicator. Table 3 presents the ERR of a medical home visit associated with a 10-μg/m3 increase in air pollutant levels for each health indicator.Table 3.Excess Relative Risk (%) of a Medical Home Visit Associated With a 10-μg/m3 Increase in Air Pollutant Levels, Bordeaux, France, 2000–2006DiagnosisNitrogen DioxidePM10OzoneERR95% CIERR95% CIERR95% CIUpper respiratory diseasesa0.8−0.7, 2.31.50.3, 2.7−0.6−1.7, 0.5Lower respiratory diseasesa2.60.2, 4.92.50.5, 4.4−0.4−2.5, 1.7Asthmaa1.1−3.0, 5.20.5−3.1, 4.1−0.8−3.9, 2.3Headache or astheniab2.80.4, 5.33.51.3, 5.91.70.2, 3.3Skin rash or conjunctivitisa0.3−3.3, 4.23.2−0.2, 6.83.00.4, 5.7Lumbago (control)a−0.3−3.4, 2.90.5−2.3, 3.50.5−1.6, 2.6Abbreviations: CI, confidence interval; ERR, excess relative risk; PM10, particulate matter less than 10 μm in diameter.aLag of 0–3 days.bLag of 0 days.The risk of a visit for upper respiratory diseases was significantly increased by 1.5% (95% confidence interval (CI): 0.3, 2.7) during the 3 days following a 10-μg/m3 increase in PM10 levels; a similar but nonsignificant trend was also observed for an increase in nitrogen dioxide levels. Similarly, the risks of consulting a general practitioner for lower respiratory diseases increased by 2.6% (95% CI: 0.2, 4.9) and 2.5% (95% CI: 0.5, 4.4) following 10-μg/m3 increases in nitrogen dioxide and PM10 levels, respectively. Asthma was not associated with any of the indicators considered, and no trend was found, nor was an association found between visits made for respiratory diseases and ozone.The daily number of visits made for headache or asthenia was significantly associated with the 3 air pollution indicators considered; same-day ERRs were 2.8% (95% CI: 0.4, 5.3), 3.5% (95% CI: 1.3, 5.9), and 1.7% (95% CI: 0.2, 3.3) for 10-μg/m3 increases in nitrogen dioxide, PM10, and ozone levels, respectively. The risk of a visit for skin rash or conjunctivitis was also increased during the 3 days following increases in PM10 (ERR = 3.2%, 95% CI: −0.2, 6.8 (close to significance)) and ozone (ERR = 3.0%, 95% CI: 0.4, 5.7) levels.Lastly, the daily number of visits made for lumbago, which we chose as the control outcome in order to evaluate any bias in the analyses, was not associated with any of the pollutant indicators considered.Results from complementary analyses of specific age groups are illustrated in Figure 3. Although the confidence intervals overlapped, central estimates of ERR were much higher in the elderly for respiratory diseases (upper and lower), regardless of the type of pollutant, suggesting a higher effect in this subgroup. This difference was particularly important for upper respiratory diseases, with ERRs of 12.3% (95% CI: 4.9, 19.7) and 8.3% (95% CI: 2.0, 14.7) being associated with 10-μg/m3 increases in nitrogen dioxide and PM10 levels, respectively, in this subgroup. In the other age groups, effects were close to each other.Figure 3.Excess relative risk (%) of a medical home visit for upper and lower respiratory diseases associated with a 10-μg/m3 increase in air pollutants, Bordeaux, France, 2000–2006. NO2, nitrogen dioxide; O3, ozone; PM10, particulate matter less than 10 μm in diameter. Bars, 95% confidence interval.For asthma, in addition to the whole population, no association was observed in children.When distributed-lag models were used, no harvesting effect was observed, regardless of which health indicator was considered. Figure 4 compares the results from initial models and from distributed-lag models, in which up to 15 days of lag time between the exposure and health effects were considered. Taking into account lags over 15 days did not lead to larger associations between PM10 and numbers of home visits, whatever the diagnosis considered, whereas estimates for nitrogen dioxide were much higher when the effects of lags of up to 15 days on upper respiratory diseases (global ERR = 5.7%, 95% CI: 2.7, 8.8) and lower respiratory diseases (global ERR = 9.4%, 95% CI: 4.7, 14.3) were considered. It is also noteworthy that there was a much higher effect of nitrogen dioxide on visits for skin rash and conjunctivitis when delayed effects were considered, even though the excess risk was not significant because of a large confidence interval.Figure 4.Comparison of excess relative risks (%) of a medical home visit obtained using a lag time of 0–3 days with excess relative risks obtained using distributed-lag models (15-day cumulative effect), Bordeaux, France, 2000–2006. LRD, lower respiratory diseases; NO2, nitrogen dioxide; PM10, particulate matter less than 10 μm in diameter; URD, upper respiratory diseases. Bars, 95% confidence interval.The persistence of nitrogen dioxide effects over 15 days is illustrated in Figure 5, which represents the ERR of a visit for upper respiratory diseases estimated for each lag up to 14 days through distributed-lag models. This ERR started to decrease progressively about 10 days after exposure and was significant up to a lag of 11 days.Figure 5.Excess relative risk (%) of a medical home visit for upper respiratory diseases associated with a 10-μg/m3 increase in nitrogen dioxide levels, by lag time, Bordeaux, France, 2000–2006. Bars, 95% confidence interval.DISCUSSIONThis study is one of the few to have investigated short-term relations between air pollution and morbidity through diagnoses made by general practitioners. It underlines that the health effects of air pollution are not restricted to the cardiorespiratory system but can also increase the risks of other conditions that have rarely been described as being associated with air pollution, such as headache, asthenia, conjunctivitis, and skin rash.The main strength of this study was the availability of reliable health data reflecting the activity of general practitioners in the general population, which is usually hard to monitor. Indeed, anyone in need of a medical visit can call SOS Médecins, which therefore makes visits in the general population by definition. The medical diagnoses are recorded by the physicians themselves after every visit, using standardized codes for each disease. This is not the case for similar organizations in France, which have only been collecting information on the complaints, symptoms, and infections reported by the patients. That is the reason for the main limitation of this study: It was a unicentric study because of the lack of similar available data from other cities covering the past several years. The population served by SOS Médecins might have changed during the study period, which was quite long, and it might also have changed according to season, since the study area has a high level of tourism. However, the adjustment for season and long-term trends allowed us to take into account those variations, as well as variations in the general practitioners’ activity. Concerning exposure, all of the monitoring stations were selected in accordance with local experts who considered them representative of the background exposure within the study area. Although levels measured by these stations do not accurately reflect individual levels of exposure, published research has shown that their day-to-day variations are well correlated with variations in individual exposure (20).Our results for respiratory diseases are globally in agreement with those of the few studies that have investigated relations between air pollution and general practitioner activity, and are supported by biologic hypotheses. Indeed, almost all of the studies that investigated the association between PM10 and general practitioners’ visits for upper (6, 9, 10) and/or lower (6, 21) respiratory diseases observed such a relation; the exception was 1 study focusing on children, for which the study period was very short (22). Concerning nitrogen dioxide, most of the studies found an association between this indicator and consultations for respiratory diseases (10, 21–23); this was not found in the latest period of a study based on visits made by SOS Médecins in the Greater Paris area (6). The associations between PM10 and nitrogen dioxide, on the one hand, and respiratory diseases, on the other hand, were higher in the elderly. It is probable that air pollution exacerbates preexisting conditions that are more likely to affect elderly people. Furthermore, aging is characterized by a decrease in antioxidant defenses, and the elderly may constitute a group at high risk of suffering from oxidation phenomena induced by air pollution (24). Like other investigators (6, 21), we did not find any association between respiratory diseases and ozone, although this pollutant is well-known for its oxidative properties and was significantly associated with respiratory medical visits in London and Hong Kong (8, 10).The most innovative results of this study concern the association between air pollution and conjunctivitis and skin rash, in relation to a potential atopic etiology of these 2 diseases, as well as the association between air pollution and headache and asthenia, which reflects general health impairment without a clearly diagnosed disease. To our knowledge, no experimental study has shown such an association, because these symptoms cannot be objectively measured and therefore are diagnosed on a declarative basis. However, a link between air pollution and physicians’ visits for headache was observed in the Evaluation des Risques de la Pollution Urbaine sur la Santé (ERPURS) program (23), with ERRs slightly lower than but in the same range as those in the current study. Szyszkowicz (25) also found a link between several air pollutant indicators and emergency visits for headache in Montreal, Canada; several lag times were tested, and the highest ERRs associated with pollutant level increases were observed for the same day, just as in the present study, and were within the same range. Thus, air pollutants could have a very short-term effect on unspecific syndromes like asthenia and/or headache—conditions which are not very severe but can be painful and affect many people.Conjunctivitis and skin rash were also significantly associated with PM10 and ozone levels, confirming the potential link between air pollution and allergic diseases other than respiratory diseases. Indeed, exposure to air pollutants is able to trigger an immunoglobulin E response and to induce oxidative protein damage in the stratum corneum, leading to the disruption of barrier function and exacerbation of atopic dermatitis (26). In the ERPURS program (22), eye conditions were found to be exclusively related to ozone levels; the concordance of results from both our study and the ERPURS study strongly favors an association between eye diseases and ozone, which could be explained by its well-known irritant and oxidant properties. Ocular inflammation and dryness were also shown to be related to high concentrations of atmospheric pollutants in patients suffering from eye discomfort syndrome (27). Lastly, 2 other studies comparing prevalences of atopic eczema (28) or atopic dermatitis and allergic rhinoconjunctivitis (29) showed that prevalences were higher in the most polluted areas. Our results are therefore in accordance with those of the few published studies available, which all suggest the existence of a link between air pollutant exposure and dermatitis and/or eye diseases. Furthermore, results found for more commonly studied indicators are totally in accordance with results from the literature, which also favors a robustness of our analyses with regard to these innovative health indicators.Also favoring the robustness of our results is the lack of any significant relation between air pollution and visits made for lumbago, the outcome chosen as the control health indicator, since it is not related a priori to air pollution.In addition, we used distributed-lag models which have been recently used in environmental epidemiology in order to quantify the so-called “mortality displacement effect” (17). In our study, distributed-lag models produced effect estimates similar to or higher than those of models using 3-day moving averages, suggesting that 1) the effect of air pollution is not simply advanced by a few days, since no obvious harvesting effect is observed, and 2) effects associated with increases in nitrogen dioxide levels persist for 2 weeks after exposure for both upper and lower respiratory diseases. The fact that the use of distributed-lag models led to higher effect estimates for nitrogen dioxide and several health indicators (respiratory diseases, conjunctivitis, and skin rash) but not for other pollution and health indicators suggests that different pollutants can have more or less delayed effects. However, it is difficult to know whether these different types of effects are really linked to the pollutant itself or whether the pollutant acts as a surrogate for other exposures.The lack of association between air pollution and asthma can be considered surprising in the context of results existing in the literature (30–32). However, studies using data on general practitioner activity are more controversial: In the London study, visits for asthma were associated with several air pollution indicators (7); in the ERPURS program in Paris, this was also the case during the 1991–1995 period (23), but such an association was no longer observed during 2000–2003 (6). Thus, the lack of association could be explained, on the one hand, by the lower number of daily events in comparison with the previous period (representing only 1% of all visits) and, on the other hand, by the fact that asthma treatment has changed, and the context of crisis that motivates general practitioners’ visits might have been consequently modified. Indeed, asthma patients often treat themselves if a crisis occurs and might go directly to the emergency room (without calling a general practitioner) if the crisis cannot be controlled using their usual medication.In conclusion, we found evidence of an association between air pollutant levels and daily numbers of general practitioners’ visits for various syndromes. This study proves the relevance of such data for epidemiologic research in the field of environmental health. Furthermore, the results of this study will sensitize doctors to the importance of giving as precise a diagnosis as possible, and this will probably help to further increase data quality; better precision in diagnosis will allow us to obtain more specific data on health indicators in the future. The links observed with conjunctivitis, skin rash, headache, and asthenia are very innovative, since few studies have shown relations with them because of the lack of suitable data. They strongly suggest that cardiorespiratory diseases are not the only conditions that can be induced or exacerbated by air pollution and show that focusing on very severe events leads to underestimation of air pollution effects. In terms of public health, this study suggests that a large number of medical visits are attributable to air pollution in Bordeaux, where current levels of air pollutants are globally close to European air quality guidelines for health protection. This is one more convincing argument for promoting all measures aimed at reducing pollutant emissions, on both the individual and collective levels, even in moderately polluted areas.AbbreviationsCIconfidence intervalERPURSEvaluation des Risques de la Pollution Urbaine sur la SantéERRexcess relative riskPM10particulate matter less than 10 μm in diameterAuthor affiliations: French Institute of Public Health Surveillance, Bordeaux, France (Sophie Larrieu, Gaëlle Gault, Laurent Filleul); French Institute of Public Health Surveillance, Saint Maurice, France (Agnès Lefranc); SOS Médecins, Bordeaux, France (Franck Couvy, Bernard Jouves); and Regional Observatory of Health Ile-de-France, Paris, France (Edouard Chatignoux).The authors thank SOS Médecins Bordeaux and AIRAQ for their collaboration in providing data and their very useful participation.Conflict of interest: none declared.1.BrunekreefBHolgateSTAir pollution and healthLancet20023609341123312422.DanielsMJDominiciFZegerSLThe National Morbidity, Mortality, and Air Pollution Study. Part III: PM10 concentration-response curves and thresholds for the 20 largest US citiesRes Rep Health Eff Inst2004941213.SamoliEAnalitisATouloumiGEstimating the exposure-response relationships between particulate matter and mortality within the APHEA multicity projectEnviron Health Perspect2005113188954.BellMLPengRDDominiciFThe exposure-response curve for ozone and risk of mortality and the adequacy of current ozone regulationsEnviron Health Perspect200611445325365.BrookRDBrookJRRajagopalanSAir pollution: the “heart” of the problemCurr Hypertens Rep20035132396.ChardonBLefrancAGranadosDAir pollution and doctors’ house calls for respiratory diseases in the Greater Paris area (2000–3)Occup Environ Med20076453203247.HajatSHainesAGoubetSAAssociation of air pollution with daily GP consultations for asthma and other lower respiratory conditions in LondonThorax19995475976058.HajatSHainesAAtkinsonRWAssociation between air pollution and daily consultations with general practitioners for allergic rhinitis in London, United KingdomAm J Epidemiol200115377047149.HajatSAndersonHRAtkinsonRWEffects of air pollution on general practitioner consultations for upper respiratory diseases in LondonOccup Environ Med200259529429910.WongTWTamWTakSYAssociation between air pollution and general practitioner visits for respiratory diseases in Hong KongThorax200661758559111.World Organization of Family DoctorsICPC-2: International Classification of Primary Care. Second Edition1998New York, NYOxford University Press12.WoodSNGeneralized Additive Models: An Introduction With R2006Boca Raton, FLChapman & Hall/CRC13.AkaikeHPetrovVNCsakiFInformation theory and an extension of the maximum likelihood principleSecond International Symposium on Information Theory1973Budapest, HungaryAkailseoniai-Kiudo26728114.TouloumiGAtkinsonRLe TertreAAnalysis of health outcome time series data in epidemiological studiesEnvironmetrics200415210111715.SunHLChouMCLueKHThe relationship of air pollution to ED visits for asthma differ between children and adultsAm J Emerg Med200624670971316.SunyerJBallesterFTertreALThe association of daily sulfur dioxide air pollution levels with hospital admissions for cardiovascular diseases in Europe (the Aphea-II study)Eur Heart J200324875276017.ZanobettiAWandMPSchwartzJGeneralized additive distributed lag models: quantifying mortality displacementBiostatistics20001327929218.Council of the European UnionDirective 1999/30/CE du Conseil du 22 Avril 1999 Relative à la Fixation de Valeurs Limites pour l'Anhydride Sulfureux, le Dioxyde d'Azote et les Oxydes d'Azote, les Particules et le Plomb dans l'Air Ambiant1999Brussels, BelgiumEuropean Union19.World Health Organization Regional Office for EuropeWHO Air Quality Guidelines for Particulate Matter, Ozone, Nitrogen Dioxide and Sulfur Dioxide. Global Update 2005. Summary of Risk Assessment2006Geneva, SwitzerlandWorld Health Organization20.JanssenNAHoekGBrunekreefBPersonal sampling of particles in adults: relation among personal, indoor, and outdoor air concentrationsAm J Epidemiol1998147653754721.HwangJSChanCCEffects of air pollution on daily clinic visits for lower respiratory tract illnessAm J Epidemiol2002155111022.KeidingLMRindelAKKronborgDRespiratory illnesses in children and air pollution in CopenhagenArch Environ Health199550320020623.MedinaSLe TertreAQuénelPAir pollution and doctors’ house calls: results from the ERPURS system for monitoring the effects of air pollution on public health in Greater Paris, France, 1991–1995Environ Res.1997751738424.KellyFJDunsterCMudwayIAir pollution and the elderly: oxidant/antioxidant issues worth considerationEur Respir J Suppl.20034070s75s25.SzyszkowiczMAir pollution and daily emergency department visits for headache in Montreal, CanadaHeadache200848341742326.NiwaYSumiHKawahiraKProtein oxidative damage in the stratum corneum: evidence for a link between environmental oxidants and the changing prevalence and nature of atopic dermatitis in JapanBr J Dermatol2003149224825427.VersuraPProfazioVCelliniMEye discomfort and air pollutionOphthalmologica1999213210310928.Martin Fernandez-MayoralasDMartin CaballeroJMGarcia-MarcosALPrevalence of atopic dermatitis in schoolchildren from Cartagena (Spain) and relationship with sex and pollution [in Spanish]An Pediatr (Barc)200460655556029.DotterudLKOdlandJØFalkESAtopic dermatitis and respiratory symptoms in Russian and northern Norwegian school children: a comparison study in two arctic areas and the impact of environmental factorsJ Eur Acad Dermatol Venereol200418213113630.AtkinsonRWAndersonHRSunyerJAcute effects of particulate air pollution on respiratory admissions: results from APHEA 2 projectAm J Respir Crit Care Med2001164101860186631.VilleneuvePJChenLRoweBHOutdoor air pollution and emergency department visits for asthma among children and adults: a case-crossover study in northern Alberta, Canada [electronic article]Environ Health200764032.BarnettAGWilliamsGMSchwartzJAir pollution and child respiratory health: a case-crossover study in Australia and New ZealandAm J Respir Crit Care Med20051711112721278
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B15075F176E5D810C9C1627B6698F6727D80CC4.txt b/test/dataset/in/resources/corpus/Clean_0B15075F176E5D810C9C1627B6698F6727D80CC4.txt
deleted file mode 100644
index e9eccdb..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B15075F176E5D810C9C1627B6698F6727D80CC4.txt
+++ /dev/null
@@ -1 +0,0 @@
-MAD2033S0047-6374(97)00174-710.1016/S0047-6374(97)00174-7Elsevier Science Ireland LtdTable 1Analysis of the expected distribution based on the total number of events found in 21 younger DS individuals between 0 to 25 years old (105 lesions), number of identifiable bands per haploid set (400) and number of analyzed metaphases (1050), according to Mariani (1989)ha (h)E(h)p (h)F (h)00.7691264307.6506140010.201895780.758270.230873692.3494322.649881E-0210.599522.897792E-0211.5911632.318645E-030.92745822.47911E-030.991633441.52161 lE-046.0864445E-021.604646E-046.417519E-0257.988458E-063.195383E-038.303454E-063.310744E-03**63.49495E-071.39798E-043.149962E-071.153606E-04***Column h, events per band; Column a (h), probability of finding h events at a given band; Column E (h), expected number of bands showing h events; Column p (h), probability of finding h or more events at a given band; Column F (h), expected number of bands that would show h or more events.*Significant (P<0.05); **Highly significant (P<0.01); ***Very highly significant (P<0.001).Table 2Analysis of the expected distribution based on the total number of events found in 17 older DS individuals between 29 to 48 years old (81 lesions), number of identifiable bands per haploid set (400) and number of analyzed metaphases (850), according to Mariani (1989)ha (h)E (h)p (h)F (h)00.8166865326.6746140010.16537966.15160.183313573.3254121.674462E-026.697851.793447E-027.17380531.130262E-030.45210481.189847E-030.475955545.721952E-052.288781E-025.958439E-052.385065E-02*52.317391E-069.269562E-042.364868E-069.628423E-04**Column h, events per band; Column a (h), probability of finding h events at a given band; Column E (h), expected number of bands showing h events; Column p (h), probability of finding h or more events at a given band; Column F (h), expected number of bands that would show h or more events.*Significant (P<0.05); **Highly significant (P<0.01); ***Very highly significant (P<0.001).Table 3Frequency of lesions in bands characterized as fragile sites, minimum expected number of lesions according to Poisson distribution (Mariani, 1989) and total number of lesions in lymphocytes from younger DS individuals and older DS individualsGroups of individualsNumber of individualsAge/years (±S.D.)Total number of analyzed cellsTotal number of lesionsMinimum expected no. of lesions (Mariani, 1989)Frequency (%) of lesions in bands characterized as fragile sites2q113p145q316p219q12Younger (DS)2117.99 (±6.32)1050105≥54.87.6———Older (DS)1736.39 (±10.15)85081≥44.912.346.27.46.2Down’s syndrome, ageing and fragile sitesMariliade Arruda Cardoso Smith*BiancaBorsattoDepartamento de Morfologia, Disciplina de Genética, UNIFESP Escola Paulista de Medicina, Rua Botucatu, No. 740, 04023 900 CEP São Paulo, Brazil*Corresponding author. Fax: +55 011 5492127/5708378; e-mail: macsmith.morf@epm.brAbstractFragile sites have been interesting for mapping chromosomal regions involved in disease and ageing. The chromosomal fragile site expression from 38 Down’s syndrome (DS) individuals aged 0–48 years was investigated in blood peripheral lymphocytes. Fragile sites were statistically characterized as the minimum expected number of lesions per band based on a Poisson distribution. The results showed that the fragile site 2q11 was associated with the DS condition and fragile sites 5q31, 6p21 and 9q12 with ageing in DS subjects. Fragility in 6p21 has also been associated with Alzheimer’s disease patients.KeywordsDown’s syndromeAgeingFragile sites1IntroductionDown’s syndrome (DS) or trisomy of chromosome 21 is the most common congenital chromosomal aberration in human beings. The majority of individuals (95%) with trisomy 21 have three free copies of this chromosome and in a small percentage of patients (5%), one copy is translocated to another acrocentric chromosome or is present in mosaic form. DS in marked by particular phenotypes that include mental retardation, characteristic physical features and reduced life expectancy. This syndrome generally originates in a meiotic nondisjunction, with an estimated frequency of 1:600–1:800 births. In adults with DS, most of the clinical signs associated with normal ageing occur prematurely. These individuals show an early cognitive decline with age and generally develop the neuropathological changes of Alzheimer’s disease (AD) (Wisniewski et al., 1992). Fragile sites are points in the chromosomes where they preferentially show lesions such as breaks or gaps, either spontaneously or after exposure to specific tissue culture conditions (Sutherland 1979, 1985). The fragile sites can be used as cytogenetic markers in association and linkage studies (Sutherland 1988, Sutherland and Hecht 1995). The use of chromosomal fragile sites has been useful for mapping chromosomal regions of the genome that contain genetic loci important for the causation of diseases and/or ageing. In our laboratory, we observed the chromosomal fragile site 6p21 associated with elderly healthy people and AD patients (Kormann-Bortolotto et al., 1992). Mariani (1989)proposed a characterization of fragile sites according to a minimum expected frequency of lesions (gaps, breaks and rearrangements) per band based on a Poisson distribution. Those sites in which the probability of chance recurrence was less than 5% are considered fragile sites. Up to now no approaches similar to ours have been described which investigated the occurrence of ‘spontaneous’ chromosome lesions in DS subjects. In the present study we investigated the expression of fragile sites under low folate culture conditions in Down’s syndrome patients in order to verify whether some fragile sites can be correlated with the clinical condition of the patients or with their ageing process.2Material and methods2.1Selection of subjectsThe sample of individuals with DS was made up of 38 patients, of whom 21 were between 0 and 25 years old and 17 between 29 and 48 years old. The individuals were selected and supervised within a co-operative program between the Universidade Federal de São Paulo-Escola Paulista de Medicina (UNIFESP-EPM) and Associação de Pais e Amigos dos Excepcionais (APAE-São Paulo). The average age (±S.D.) of younger DS individuals was 17.99 (±6.32) and the average age (±S.D.) of older individuals was 36.39 (±10.15). All the DS patients had free trisomy. The younger group was made up of 11 females and ten males. The older group had one female and 16 males. The DS patient groups were divided arbitrarily based on evidence that persons with DS develop neuropathological changes similar to those of AD after the 4th decade of life (Malamud, 1972). The older group was almost entirely composed of people in the 4th and 5th decade of life. The younger group was almost entirely composed of people in the 2nd and 3rd decade of life.2.2Cytogenetic analysisThe cytogenetic analysis was done on the phytohaemagglutinin-stimulated lymphocytes from these individuals. These cells were cultured for 72 h in TC 199 medium, which is a low folic acid medium, known to enhance the appearance of folate sensitive fragile sites, with 7% fetal calf serum. The investigation of fragile sites was carried out in a blind test using at least 50 cells (all containing 47 chromosomes) per individual. All cells were initially analyzed by conventional staining. Chromosomal lesions (gaps, breaks or rearrangements) were recorded and the slides were destained with fixative solution (methanol:acetic acid 3:1) and restained with G-banding technique for analysis of the same metaphases afterwards. The number of lesions in each band site were computed and only those sites were considered fragile in which the probability of random recurrence was less than 5% (Mariani, 1989)3Results and discussionThere was a total of 105 lesions in the younger DS individuals group and there were 81 among older DS individuals group. These differences were not statistically significant according to the Mann–Whitney test (zcalc=0.75 and zcrit=1.64). Thus, ageing in DS did not increase the number of chromosomal lesions. The older DS sample is almost formed by men, however some findings from the literature have not shown differences between sexes related to the incidence of autosomal fragile sites (Sutherland 1982). A band was considered a fragile site based on a Poisson distribution that indicated the minimum expected frequency of events (lesions) per band (Mariani, 1989). Table 1Table 2Table 3 show the frequency of bands considered fragile sites in younger DS individuals group and older DS individuals group. Fragile site 3pl4 was common to both groups. The occurrence of this fragility was similar in younger and older DS groups. Band 3pl4 is widely cited as the most common fragile site in humans (Simonic and Gericke, 1996). Recently a gene called FHIT has been reported that is located in this chromosomal region. This gene is either completely or partially missing in many neoplasias such as colon, breast and lung cancer which suggests it may be a tumour suppressor gene. Fragile site 6p21 was observed only in the older DS group. The gene of a microtubule associated tau like protein (MTBT2) is located in this band (Kidd et al., 1989). Thus, this gene may be related to chromosomal alterations which involve defects at the level of microtubules and may also be a primary mechanism of the ageing process of DS syndrome. Fragile site 2q11 was observed both in older DS group and younger DS group. This chromosomal region has been described as an unstable secondary constriction on the long arm of this chromosome in patients with different clinical features (Lejeune et al., 1968). Other cases with this secondary constriction show that clinical features can vary considerably, with phenotypes that include mental retardation, congenital heart malformation and growth retardation, as well as some healthy individuals (Ferrante et al., 1968Buhler et al., 1970Williams and Howell, 1976). Whereas this chromosomal alteration might be considered a normal variant, Anneren and Gutstavson (1981)suggest that cell lines with it might have a pathological effect if they occur at critical stages in the organogenesis. The fragility in chromosome 9ql2 was exclusively observed in older DS patients. Fragile site 9ql2 is a chromosomal region formed by a constitutive heterochromatin, with a still unclear biological function. Previous studies have also shown a possible increase of chromosome 9 pericentric inversion among DS subjects and it has been suggested that this inv(9)(qh) predisposes the carrier parent to nondisjunction. Cytogenetic studies show an interaction between the heterochromatin and the acrocentric chromosomes during the cell division process that can interfere in disjunction and chromosome distribution (Natarajan and Ahnstrom, 1969Gagné et al., 1974Miklos and John, 1979Hultén et al., 1987). Fragile site 5q31 was observed only in older DS individuals. In this region, for instance the gene for a lymphokine, interleukin-13, is located which is expressed in activated human T-lymphocytes. Minty et al. (1993)suggested that interleukin-13 may be critical in regulating inflammatory and immune responses. It is well known that DS individuals exhibit an increased frequency of infections and malignancies, especially those involving the head, neck and respiratory tracts. Finally, fragile site 2q11 has been observed in both DS groups whereas has not been observed in young and/or elderly control groups previously analyzed in our laboratory (Kormann-Bortolotto et al., 1992). Thus, we believe that this fragile site is associated with DS condition. Fragile site 6p21 was observed in the older DS syndrome group, in the elderly control group and in AD analyzed previously in our laboratory (Kormann-Bortolotto et al., 1992). Thus, we believe that fragile site 6p21 is associated with the elderly condition.Some studies have suggested that the fragile site could be a cytogenetic representation of gene expression (Yunis et al., 1987Hecht, 1988). Austin et al. (1992)have shown a clustering of spontaneous aberrations at fragile sites that may indicate that these regions are physiologically active during the S-phase of the cell cycle. Fragile sites FRAXA, FRAXE, FRAXF and FRA16A are formed by the (CGG)n trinucleotide repeat unit expansion with subsequent methylation of a adjacent CpG island that can introduce a recognition site for de novo imprinting (Nancarrow et al., 1994). Therefore fragile sites could represent adjacent gene expression. Finally, our findings have shown fragile site 2q11 associated with DS condition and fragile sites 5q31, 6p21 and 9q12 associated with older DS individuals. These fragile sites could indicate chromosomal regions or genes to be further investigated in DS ageing or in this syndrome.AcknowledgementsThe authors are grateful to Prof. Eduardo Katchburian, Prof. Antonio dos Santos Clemente and Dr Elisa Maria Doceu Moreira Garcez and to Associação de Pais e Amigos dos Excepcionais (APAE-São Paulo) for allowing this co-operative program. We would also like to thank Elenice Rosana Salas for technical assistance and Dr Tulio Mariani who provided the statistical computer program. This work was supported by Sandoz Foundation of Gerontological Research, Fundação de Amparo à Pesquisa do Estado de São Paulo, Brazil (FAPESP), Coordenação de Aperfeiçoamento de Nı́vel Superior, Brazil (CAPES) and Conselho Nacional de Desenvolvimento Cientı́fco e Tecnológico, Brazil (CNPq).ReferencesAnneren and Gutstavson, 1981G.AnnerenK.GustavsonA fragile secondary constriction on chromosome 2 in five patients with different clinical featuresHereditas9519816367Austin et al., 1992M.J.F.AustinJ.M.CollinsL.A.CoreyW.E.NanceM.C.NealeR.M.SchiekenJ.A.BrownAphidicolin-inducible common fragile sites expression: results from a population survey of twinsAm. J. Hum. Genet.5019927683Buhler et al., 1970E.M.BuhlerU.LuchsingerU.K.BuhlerK.MéhesG.R.StalderNoncondensation of one segment of a chromosome no 2 in a male with otherwise normal karyotype (and severe hypospadias)Humangenetik9197097104Ferrante et al., 1968E.FerranteL.BruniS.GrimaldiRottura paracentromerica del braccio lungo di un chromosome 2 in una bambina di 9 mesi affetta da cistationuriaritardo psicomotorio e cataratta.. Riv. Clin. Ped.811968612619Gagné et al., 1974R.GagneJ.M.LucianiM.Devictoc-VuilletA.StahlC9 heterochromatin during the first meiotic prophase in human fetal oocyteExp. Cell. Res.851974111116Hecht, 1988F.HechtFragile sites, cancer, chromosome break points and oncogenes all cluster in light G-bandsCancer Genet. Cytogenet.331988110Hultén et al., 1987M.HulténN.SaadallahJ.BatanianMeiotic chromosome pairing in the human male: Experience from surface spread synaptonemal complexesChromosomes Today91987218229Kidd et al., 1989K.K.KiddA.M.BowcockJ.SchmidtkeReport of the DNA committee and catalogs of cloned and mapped genes and DNA polymorphismsCytogenet. Cell Genet.511989622947Kormann-Bortolotto et al., 1992M.H.Kormann-BortolottoM.de A.C.SmithJ.Toniolo NetoFragile sites, Alzheimer’s disease and ageingMech. Ageing Dev.651992915Lejeune et al., 1968J.LejeuneB.DutrillauxJ.LafourcadeR.BergerD.AbonylM.O.RethoreEndoreduplication selective du bras long du chromosome 2 chez une femme et sa filleAcad. Sci. Paris26619682426Malamud, 1972Malamud, D.N., 1972. Ageing and the Brain, Gaitz, C.M., Plenum, New York.Mariani, 1989T.MarianiFragile sites and statisticsHum. Genet.811989319322Miklos and John, 1979G.L.G.MiklosB.JohnHeterochromatin and satellite DNA in man: properties and prospectsAm. J. Hum. Genet.311979264280Minty et al., 1993A.MintyP.ChalonJ.M.DerocqX.DumontInterleukin-13 is a new human lymphokine regulating inflammatory and immune responsesNature3621993248250Nancarrow et al., 1994J.K.NancarrowE.KremerK.HolmanH.EyreN.A.DoggetD.Le PaslierD.F.CallenG.R.SutherlandR.I.RichardsImplications of FRAl6A structure for the mechanism of chromosomal fragile site genesisScience264199419381941Natarajan and Ahnstrom, 1969A.T.NatarajanG.AhnstromHeterocromatin and chromosome aberrationsChromosoma2819694861Simonic and Gericke, 1996I.SimonicG.S.GerickeThe enigma of common fragile sitesHum. Genet.971996524531Sutherland, 1979G.R.SutherlandHeritable fragile sites on human chromosomes I Factors affecting expression in lymphocyte cultureAm. J. Hum. Genet.311979125135Sutherland, 1982G.R.SutherlandHeritable fragile sites on human chromosomes VIII, Preliminary population cytogenetic data on the folic-acid-sensitive fragile sitesAm. J. Hum. Genet.341982452458Sutherland, 1985G.R.SutherlandHeritable fragile sites on human chromosomes XII Population CytogeneticsAnnu. Hum. Genet.491985153161Sutherland, 1988G.R.SutherlandThe role of nucleotides in human fragile site expressionMutat. Res.2001988207213Sutherland and Hecht, 1995Sutherland, G.R., Hecht, F., 1995. Fragile Sites on Human Chromosomes, Oxford University Press, New York.Williams and Howell, 1976A.J.WilliamsR.T.HowellA fragile secondary constriction on chromosome 2 in a severely mentally retarded patientJ. Ment. Defic. Res.211976227230Wisniewski et al., 1992Wisniewski, K.E., Hill, A.L., Wisniewski, H.M., 1992. Down’s Syndrome-Advances in Medical Care, Willey-Liss, New York.Yunis et al., 1987J.J.YunisA.L.SorengA.E.BoweFragile sites are targets of diverse mutagens and carcinogensOncogene11987248250
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B15F4F7680CE01F26C1996D1F74278719AEE02C.txt b/test/dataset/in/resources/corpus/Clean_0B15F4F7680CE01F26C1996D1F74278719AEE02C.txt
deleted file mode 100644
index 698bd57..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B15F4F7680CE01F26C1996D1F74278719AEE02C.txt
+++ /dev/null
@@ -1 +0,0 @@
-geronageronaJournals of Gerontology Series A: Biomedical Sciences and Medical Sciences1758-535X1079-5006Oxford University Press10.1093/gerona/glp105Journal of Gerontology: Medical SciencesRapid CommunicationCardiac Autonomic Dysfunction Is Associated With White Matter Lesions in Patients With Mild Cognitive ImpairmentGalluzziSamantha1NicosiaFranco2GeroldiCristina13AlicandriAlberto2BonettiMatteo4RomanelliGiuseppe2ZulliRoberto2FrisoniGiovanni B.1351Laboratory of Epidemiology, Neuroimaging and Telemedicine (LENITEM), IRCCS San Giovanni di Dio-Fatebenefratelli, Brescia, Italy2Institute of Internal Medicine, Department of Medical Sciences, University of Brescia, Italy3Psychogeriatric Unit, IRCCS San Giovanni di Dio-Fatebenefratelli, Brescia, Italy4Service of Neuroradiology, Istituto Clinico Città di Brescia, Italy5Associazione Fatebenefratelli per la Ricerca (AFaR), Rome, ItalyAddress correspondence to Giovanni B. Frisoni, MD, Laboratory of Epidemiology Neuroimaging & Telemedicine (LENITEM), IRCCS San Giovanni di Dio FBF—The National Center for Research and Care of Alzheimer’s Disease, Via Pilastroni 4, 25125, Brescia, Italy. Email: gfrisoni@fatebenefratelli.itDecision Editor: Luigi Ferrucci, MD, PhD122009307200964A121312131529820082032009© The Author 2009. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org.2009BackgroundCardiac autonomic dysfunction has been associated with cognitive impairment, but the underlying pathogenesis is complex and cerebral white matter lesions (WMLs) might be implicated.MethodsTime and frequency heart rate variability (HRV) and visual rating of WMLs were carried out in 42 patients with mild cognitive impairment.ResultsAfter adjustment for relevant demographic and clinical characteristics, including left ventricular mass, reduced HRV indices of parasympathetic (root mean square of successive difference of RR intervals, RMSSD) and sympathetic modulation (low-frequency [LF] power) were associated with increased WML score (RMSSD: B −0.30, 95% CI −0.52 to −0.08, p = .01; LF: B −0.24, 95% CI −0.46 to −0.02, p = .05). In a multiple-adjusted model, RMSSD was the major independent predictor of WMLs (B −0.35, 95% CI −0.57 to −0.13, p = .002).ConclusionThe evidence for an independent association of cardiac autonomic dysfunction with WMLs might suggest its role in the pathogenesis of WMLs.Heart rate variabilityCardiac autonomic dysfunctionWhite matter lesionsMild cognitive impairmentHEART rate variability (HRV) is a marker of cardiovascular autonomic function (1). Reduced HRV has recently gained attention in the dementia research field because of its association with cerebrovascular risk factors (2) and Alzheimer’s disease (3). Moreover, in a sample of older disabled women from the community, root mean square of successive difference of RR intervals (RMSSD) and high-frequency (HF) power, indicators of parasympathetic-mediated short-term variations in heart rate, were predictors of cognitive impairment development, being associated with 3.4 and 6.7 times greater odds of cognitive impairment, respectively (4).Understanding pathophysiological mechanisms leading to cognitive impairment is of great interest to develop preventive therapies. However, the pathway underlying the association between reduced HRV and cognitive impairment is unclear. Cerebral white matter lesions (WMLs) might have a mediating role (4), but brain imaging studies are lacking.The aim of this study was to evaluate the independent association of HRV with WMLs in patients with mild cognitive impairment (MCI). Given the higher prevalence of WMLs in MCI patients relative to cognitively normal elders, we focused this study on the former group in order to improve the power to detect an association with HRV.METHODSMCI PatientsPatients were recruited from a prospective study on the natural history of MCI (“Mild Cognitive Impairment in Brescia” study) carried out in the outpatients’ section of the Alzheimer’s Unit, IRCCS San Giovanni di Dio-Fatebenefratelli, Brescia, Italy. Inclusion criteria included the presence of subjective and objective memory or other cognitive domain disturbance in the absence of functional impairment. More details were published elsewhere (5).One hundred thirty-six patients were enrolled from April 2002 to March 2005. For the present study, only those 94 patients who performed both 24-hour electrocardiogram (ECG) monitoring and magnetic resonance (MR) imaging were selected. Twelve patients were excluded because of inadequate quality of the exams. Those 82 patients finally included in the study were not different from 54 excluded as education, sex, and vascular risk factors but were younger (69.3 + 7.3 vs 72.5 + 8.8 years, p = .03).Clinical AssessmentGlobal cognition and depressive symptoms were assessed using the Mini-Mental State Examination (6) and the depression subscale of the Brief Symptom Inventory (7), respectively, the latter ranging from 0 to 24 (higher scores indicated poorer depressive symptoms). Hypertension and diabetes mellitus and heart diseases (coronary artery disease, primary arrhythmias, and heart failure) were clinically diagnosed and defined present if currently treated. Echocardiography was made in 67 of 82 patients. The remaining 15 patients refused the examination because of logistic difficulties. The left ventricular (LV) mass was calculated from Devereux’s formula and indexed by height.HRV MeasurementTwenty-four-hour ECG recording (Accuplus 363; Del Mar Avionics, Irvine, CA) was used to analyze HRV using implemented software. In the time domain, standard deviation of the RR intervals, standard deviation of the 5-minute mean values of RRs for each 5-minute interval, average of standard deviations of RR for each 5-minute interval, and the RMSSD were calculated. In the frequency domain, low frequency (LF) 0.04–0.15 Hz, HF 0.15–0.40 Hz, and low- or high-frequency ratio were calculated. A detailed description of ECG recording procedures was available on http://www.centroalzheimer.it/public/Methods_HRV.MR ImagingMR images were acquired at the Service of Neuroradiology, Istituto Clinico Città di Brescia, Brescia, using a 1.0 Tesla Philips Gyroscan. Axial T2 weighted, proton density, and fluid-attenuated inversion recovery images were acquired. WMLs were assessed using the rating scale for age-related white matter changes (ARWMC) (8). Total score (range 0–30) was the sum of subscores for frontal, parieto-occipital, temporal, infratentorial areas, and basal ganglia areas (range 0–3) in the left and right hemispheres. In the present study, WMLs were considered as present when ARWMC scale total score was 4 or more or when beginning confluence of lesions (subscore 2) was observed in at least one area. Based on this criteria, patients were divided into 32 with and 50 without WMLs.Statistical AnalysisThe data were analyzed using SPSS version 13.0 (SPSS, Chicago, IL). Sociodemographic and clinical differences between MCI groups with and without WMLs were assessed with t test for normally distributed continuous variables, Mann–Whitney test for nonnormally distributed continuous variables, and chi-square for categorical variables. In the whole-MCI group, the association between each HRV index and WMLs (i.e., the ARWMC scale total score) was assessed with linear regression models and the power of HRV indices to predict the extent of WMLs with a stepwise multiple regression model. Confounders were chosen among the main demographic and clinical characteristics that, based on current knowledge, have been associated with WMLs (age, diabetes, heart disease, hypertension) or have been showed to affect cardiac autonomic function (beta-blockers use, LV mass).RESULTSTable 1 shows that MCI patients with WMLs were older than those without WMLs while cognition and depression were similar. All the vascular disease and risk factors were more prevalent in MCI with WMLs, only hypertension reaching statistical significance. Among HRV indices, the RMSSD and LF were reduced in MCI patients with WMLs relative to those without WMLs (unadjusted p < .001).Table 1.Demographic and Clinical Characteristics of 82 Patients With MCI by Presence of WMLsCharacteristicsWMLs Absent (n = 50)WMLs Present (n = 32)p*Demographics Age67.6 ± 7.672.2 ± 5.9.005 Gender, female31 (62%)20 (63%).96 Education7.9 ± 4.56.6 ± 2.9.28Cognition and mood Mini-Mental State Examination27.0 ± 1.527.3 ± 1.5.43 Depression, Brief Symptom Inventory4.1 ± 4.83.4 ± 4.2.72Vascular diseases and risk factors Hypertension12 (24%)17 (53%).007 Heart disease8 (16%)7 (22%).50 Diabetes mellitus4 (8%)4 (13%).50 Beta-blocker use7 (14%)9 (28%).12 Left ventricular mass†105.1 ± 31.4116.1 ± 31.0.16Heart rate variability features SD of the RR intervals, ms118.9 ± 36.8111.7 ± 36.9.39 SD of the 5-minute mean values of RR for each 5-minute interval, ms68.3 ± 18.262.6 ± 16.9.16 Average of SDs of RR for each 5-minute interval, ms95.0 ± 38.589.0 ± 35.9.49 Root mean square of successive difference of RR intervals, ms26.5 ± 8.720.2 ± 5.5<.001 Low-frequency power, ms2212.1 ± 152.5151.1 ± 118.8.05 High-frequency power, ms288.0 ± 65.570.4 ± 73.7.06Notes: *Significance on t test or Mann–Whitney test for normally or nonnormally distributed continuous variables and chi-square for categorical variables.†Available in 41 and 26 MCI patients, respectively. Values denote mean ± SD or number (%). MCI = mild cognitive impairment; WMLs = white matter lesions.Table 2 shows that in the whole-MCI group, the RMSSD, LF, and HF were inversely associated with the extent of WMLs in the unadjusted models (p < .001), whereas only the association of decreased RMSSD with increased WML score retained significance in the adjusted models (p = .01). When also the confounding effect of LV mass was taken in account, the association was even significant for RMSSD (p = .01) and reached significance also for LF (p = .05).Table 2.Association of HRV Indices With White Matter Lesions in 82 Patients With Mild Cognitive ImpairmentHRV IndexUnadjustedAdjusted for Age, Diabetes Mellitus, Heart Disease, Hypertension, Beta-blockersAdjusted for the Previous Plus LV MassB95% CIpB95% CIpB95% CIpSDNN−0.08−0.30 to 0.14.480.10−0.12 to 0.32.350.06−0.16 to 0.28.61SDANN−0.18−0.40 to 0.04.11−0.03−0.25 to 0.19.78−0.08−0.30 to 0.14.51SDNN idx−0.05−0.27 to 0.17.680.13−0.09 to 0.35.230.10−0.12 to 0.32.43RMSSD−0.37−0.59 to −0.15.001−0.27−0.49 to −0.05.01−0.30−0.52 to −0.08.01LF−0.30−0.52 to −0.08.006−0.15−0.37 to 0.07.15−0.24−0.46 to −0.02.05HF−0.23−0.45 to −0.01.04−0.11−0.33 to 0.11.31−0.18−0.40 to 0.04.14Notes: Regression coefficients were computed in 18 separate linear regression models and denote the increase of the dependent for a 1-Unit increase of the independent variable. Models adjusted for LV mass included 67 patients.HF = high-frequency power; HRV = heart rate variability; LF = low-frequency power; LV = left ventricular; RMSSD = root mean square of successive difference of RR intervals; SDANN = standard deviation of the 5-minute mean values of RR for each 5-minute interval; SDNN = standard deviation of the RR intervals; SDNN idx = average of standard deviations of RR for each 5-minute interval.Stepwise multiple regression analysis including HRV indices significantly associated with WMLs in univariate models (RMSSD, LF, and HF) and confounders showed that the RMSSD and age were significant predictors of WMLs (B −0.30, 95% CI −0.52 to −0.08, p = .002; B 0.38, 95% CI 0.16–0.60, p < .0005, respectively), also in the model adjusted for LV mass (RMSSD: B −0.35, 95% CI −0.47 to −0.22, p = .002).DISCUSSIONThis study showed that in MCI patients, reduced RMSSD, an index of parasympathetic-mediated short-term component of HRV, was associated with extent of WMLs, suggesting a possible role of cardiac autonomic dysfunction in the WML pathogenesis.Studies assessing the relationship between HRV and WMLs are scanty and have used different methods to measure HRV, making results difficult to compare (9–11). To our best knowledge, the only study assessing HRV using power spectral analysis analyzed only frequency-domain measures and found that LF was significantly lower in hypertensive patients with lacunar infarctions than in those without lacunar infarctions (12).In the present study, RMSSD was the only HRV index to demonstrate an association with WMLs in multivariate models. This was in line with previous finding that proved RMSSD to be more closely associated with cognitive impairment than indicators of long-term variations in heart rate (4). The pathological mechanisms possibly linking cardiac autonomic dysfunction to WMLs might include abnormalities in heart rate control and vascular dynamics (13) or the role of intermediate factors (neurally mediated syncope, orthostatic hypotension, or increased blood pressure variability) that were found to be associated with both HRV (4,13,14) and WML development (15).A possible alternative explanation is that reduced HRV is a consequence and not a cause of WMLs. Although cerebrovascular diseases frequently cause disturbances to cardiovascular autonomic functions, this has been reported only in patients with clinically overt stroke due to large-vessel disease (16,17). Only longitudinal studies assessing causality might clarify this issue.There were some limitations to this study. First, this was a cross-sectional association study that prevents to determine the causal relationships and definitively indicate pathophysiological pathways leading from reduced HRV to WMLs. Second, the possible confounding role of neurodegeneration that, in addition to WMLs, might contribute to cardiac autonomic dysfunction and of vascular risk factors, such as dyslipidemia, has not been taken into account. Lastly, the clinical setting where the study has been implemented prevents generalizability before the results are replicated in a large population-based sample of older persons.In conclusion, we showed that reduced short-term component of HRV was associated with WMLs in patients with MCI. Further studies in larger cohorts of patients are warranted to confirm the role of HRV in the WML pathogenesis.FUNDINGThis study was funded in part by an ad hoc grant of the Italian Ministry of Health “Ricerca Finalizzata RA 00. 6A”.1.Circulation199693104310652.SinghJPLarsonMGTsujiHEvansJCO’DonnellCJLevyDReduced heart rate variability and new-onset hypertension: insights into pathogenesis of hypertension: the Framingham Heart StudyHypertension1998322932973.ZulliRNicosiaFBorroniBQT dispersion and heart rate variability abnormalities in Alzheimer’s disease and in mild cognitive impairmentJ Am Geriatr Soc200553213521394.KimDHLipsitzLAFerrucciLAssociation between reduced heart rate variability and cognitive impairment in older disabled women in the community: Women’s Health and Aging Study IJ Am Geriatr Soc200654175117575.GeroldiCRossiRCalvagnaCMedial temporal atrophy but not memory deficit predicts progression to dementia in patients with mild cognitive impairmentJ Neurol Neurosurg Psychiatry200677121912226.FolsteinMFFolsteinSEMcHughPR“Mini-mental State”. A practical method for grading the cognitive state of patients for the clinicianJ Psychiatr Res1975121891987.De LeoDFrisoniGBRozziniRTrabucchiMItalian community norms for the Brief Symptom Inventory in the elderlyBr J Clin Psychol1993322092138.WahlundLOBarkhofFFazekasFA new rating scale for age-related white matter changes applicable to MRI and CTStroke200132131813229.GoldsteinIBBartzokisGHanceDBShapiroDRelationship between blood pressure and subcortical lesions in healthy elderly peopleStroke19982976577210.SierraCde La SierraAMercaderJGómez-AngelatsEUrbano-MárquezACocaASilent cerebral white matter lesions in middle-aged essential hypertensive patientsJ Hypertens20022051952411.ToyryJPNiskanenLKLansimiesEAPartanenKPUusitupaMIAutonomic neuropathy predicts the development of stroke in patients with non-insulin-dependent diabetes mellitusStroke1996271316131812.KarioKMotaiKMitsuhashiTAutonomic nervous system dysfunction in elderly hypertensive patients with abnormal diurnal blood pressure variation: relation to silent cerebrovascular diseaseHypertension1997301504151013.VinikAIMaserREMitchellBDFreemanRDiabetic autonomic neuropathyDiabetes Care2003261553157914.GrimmWWirthsAHoffmannJMenzVMaischBHeart rate variability during head-up tilt testing in patients with suspected neurally mediated syncopePacing Clin Electrophysiol1998212411241515.BallardCO’BrienJBarberBNeurocardiovascular instability, hypotensive episodes, and MRI lesions in neurodegenerative dementiaAnn N Y Acad Sci200090344244516.KorpelainenJTSotaniemiKAMäkikallioAHuikuriHVMyllyläVVDynamic behavior of heart rate in ischemic strokeStroke1999301008101317.BassiAColivicchiFSantiniMCaltagironeCCardiac autonomic dysfunction and functional outcome after ischaemic strokeEur J Neurol200714917922
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B160E385BFB1BF9AFB7B2E9EFC39E1661E58F77.txt b/test/dataset/in/resources/corpus/Clean_0B160E385BFB1BF9AFB7B2E9EFC39E1661E58F77.txt
deleted file mode 100644
index 6482ac8..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B160E385BFB1BF9AFB7B2E9EFC39E1661E58F77.txt
+++ /dev/null
@@ -1 +0,0 @@
-geronbgeronbJournals of Gerontology Series B: Psychological Sciences and Social Sciences1758-53681079-5014Oxford University Press10.1093/geronb/gbq036Journal of Gerontology: Psychological SciencesArticlesDual-Task Performance Reveals Increased Involvement of Executive Control in Fine Motor Sequencing in Healthy AgingFraserSarah A.LiKaren Z. H.PenhuneVirginia B.Center for Research in Human Development, Department of Psychology, Concordia University, Montreal, Quebec, CanadaAddress correspondence to Sarah Fraser, PhD, Department of Psychology, UQAM et le Centre de recherche institut universitaire de gériatrie de Montréal, 4545 Queen Mary, Montréal, Québec, Canada H3W 1W4. Email: sfraser@live.concordia.ca.Decision Editor: Elizabeth Stine-Morrow, PhD0920101705201065B5526535211020092042010© The Author 2010. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org.2010The purpose of the current study was to examine the role of executive control in fine motor sequencing using a motor-cognitive dual-task paradigm. Younger and older adults performed a sequential tapping task separately and concurrently with a semantic judgment task (Experiment 1) and a mental arithmetic task (Experiment 2). Experiment 1 established that under low cognitive load, older adults were slower and less accurate in sequential tapping than younger adults. Load was manipulated in Experiment 2, and across mental arithmetic difficulty levels, older adults were less accurate in sequential tapping when performing mental arithmetic than younger adults. At the highest difficulty level, both groups suffered performance costs. In line with gross motor research, these findings suggest a role for executive functions in fine motor performance in old age.Dual taskExecutive FunctionFine motor sequencingHEALTHY aging is associated with declines in both motor (Ketcham & Stelmach, 2001; Krampe, 2002) and cognitive control functions (Kramer & Madden, 2008; Verhaeghen & Cerella, 2002). In addition, motor and cognitive functions appear to become more strongly correlated with increasing age, suggesting an increased interdependence between the two domains (Baltes & Lindenberger, 1997; Li & Lindenberger, 2002). Other evidence of this interdependence comes from motor-cognitive dual-task research. The majority of such evidence involves simultaneous gross motor and cognitive task performance, showing in many cases greater dual-task costs (DTCs) for older adults compared with younger adults (for review, Woollacott & Shumway-Cook, 2002). This pattern has been interpreted to mean that motor performance requires more cognitive resources in old age. The coordination of simultaneous task performance has been considered a component of the executive system (Baddeley, 2002), which shows age-related decline (Kramer & Madden). The involvement of executive control in gross motor performance has recently been shown using measures of gait and balance (Mendelson, Redfern, Nebes, & Jennings, 2010; Yogev-Seligmann, Hausdorff, & Giladi, 2008). Fewer studies have explored the possibility of age-related increases in executive involvement during fine motor performance (e.g., Albinet, Tomporowski, & Beasman, 2006). Therefore, our goal was to examine the role of executive control in fine motor performance using a motor-cognitive dual-task paradigm.In both gross and fine motor dual-task research, several factors have been suggested to account for age differences in dual-task performance (Krampe, 2002; Woollacott & Shumway-Cook, 2002). Some of the factors that have been implicated include a general slowing, declines in executive function, type of tasks combined, and physiological arousal. In the case of executive function, it is well documented that executive control processes may be invoked during motor tasks when adaptive online control is needed (Ble et al., 2005; Krampe; Woollacott & Shumway-Cook; Yogev-Seligmann et al., 2008). Kahneman (1973) maintained that all individuals have a limited capacity to process information and that they should be able to process two tasks at once as long as the two tasks do not exceed the individual’s limited capacity or processing resources. If the tasks demands exceed an individual’s capacity, then performance on one or both tasks can deteriorate (Kahneman). Given what is known about declines in executive and motor processes, it is not surprising that age differences are predicted in cognitive-motor dual tasks.Despite this prediction, a growing number of walking and postural control studies have found that results vary depending on the tasks combined and the cognitive load of the component tasks (i.e., Huxhold, Li, Schmiedek, & Lindenberger, 2006; Li, Lindenberger, Freund, & Baltes, 2001; Lövdén, Schäefer, Pohlmeyer, & Lindenberger, 2008). For example, in a study of mildly challenging dual-task treadmill walking, younger and older adults showed cognitive dual-task facilitation and motor DTCs, which were more pronounced in older adults (Fraser, Li, DeMont, & Penhune, 2007). A follow-up experiment that included a cognitive load manipulation demonstrated that both age groups incurred costs in both domains and were negatively affected by the increase of cognitive difficulty (Li, DeMont, Penhune, Fraser, & Abbud, 2008). Interestingly, the younger adults were able to adjust their stride length to accommodate the increase in cognitive demands (Abbud, Li, & DeMont, 2009). The changing pattern of DTCs across experiments suggests that the choice of tasks and the cognitive load of the tasks chosen can have a large impact on the resulting pattern of performance. An added dimension of walking dual-task research is the influence of postural threat (Brown, Shumway-Cook, & Woollacott, 1999). It has been argued that older adults might adopt a “posture-first” principle, prioritizing walking and balance above all other tasks in order to avoid a fall (Woollacott & Shumway-Cook, 2002).The potential confound of postural threat influencing age differences in dual-task performances is removed in fine motor dual-task research. In addition, motor measures (particularly fine and complex motor measures) have been shown to be as accurate as standard cognitive measures in delineating cognitively normal versus cognitively impaired older adults (mild cognitive impairment and mild Alzheimer’s disease; Kluger et al., 1997). This close relationship between cognitive tasks and fine motor tasks in aging has been explored with the dual-task paradigm (Crossley & Hiscock, 1992). Using a within-subjects manipulation of cognitive load, Crossley and Hiscock compared young, middle-aged, and older adults on their performance of a simple tapping task with a concurrent cognitive load. At the highest level of cognitive difficulty, there were no age differences in cognitive performance, but older adults had larger decrements in simple tapping rates in comparison with younger and middle-aged adults. This simple tapping study demonstrates age differences in fine motor dual-task performance that increase with cognitive load. Would the same be true in the dual-task performance that involves a fine motor sequence? Or would the increased complexity of sequential tapping increase the overall cognitive load and increase age differences? One study that directly contrasted simple and sequential tapping with a cognitive load (speech production) found age group differences in DTCs only for sequential tapping (Kemper, Herman, & Lian, 2003). This finding suggests that sequential tapping places an added load on older adults in comparison with simple tapping.The few published studies on aging and dual-task fine motor performance suggest that increasing the complexity of the motor task is more detrimental to older adults than young. However, the literature does not indicate if a similar pattern will emerge when cognitive complexity is varied. The current study was designed to address this gap in the literature. Our approach was to manipulate cognitive complexity in cognitive-sequential tapping task pairings in order to examine the possibility of increasing executive involvement in fine motor performance. Given the evidence for increased age-related involvement of executive functions in the gross motor literature, we began with the prediction that there would be age-related increases in motor DTCs. Such a finding would extend the existing body of research on the increasing role of cognition in motor performance.EXPERIMENT 1MethodParticipants.—Twenty younger adults (20–31 years) and 21 older adults (60–75 years) participated in the experiment. Younger adults were recruited through Concordia Psychology’s undergraduate participant pool and the older adults were recruited from a preexisting participant database. Younger adults received class credits for their participation, and older adults received a small honorarium. All participants were right handed, fluent in English, had normal or corrected vision, had never suffered a stroke, and were screened for medical conditions (e.g., Parkinson’s disease, severe arthritis) and medications that would affect their movement. Individuals who reported hearing difficulties or who wore a hearing aid were excluded. The Forward Digit Span and the Digit Symbol Substitution Test of Wechsler Adult Intelligence Scale III (WAIS; Weschler, 1997), as well as the Trail Making Test (A & B; Spreen & Strauss, 1998), were administered to assess short-term memory, processing speed, and task switching, respectively. All participants were within a normal range for their age on these tests. Descriptive statistics for each group are presented in Table 1. All procedures were approved by the Concordia University Human Research Ethics Committee.Table 1.Descriptive Statistics of the SamplesExperiment 1Experiment 2YoungerOlderYoungerOlderAge23.10 (3.16)67.67 (4.33)21.10 (2.15)70.37 (4.96)Years of Education15.95 (2.09)14.81 (4.18)14.95 (0.89)15.11 (3.26)Digit Symbol88.60 (13.82)*73.85 (18.23)*69.90 (19.10)*56.42 (14.19)*Trails B-A24.42 (12.68)*59.81 (32.92)*27.50 (15.81)*46.47 (30.75)*Digits Forward7.35 (1.04)*6.48 (1.08)*7.15 (1.09)6.68 (1.11)ERVT——7.93 (4.46)*13.03 (4.94)*WAIS math-raw——13.30 (2.92)13.95 (2.90)WAIS math-scaled——10.35 (2.08)10.63 (2.81)Notes: Mean values and standard deviations (in brackets) presented. Years of education = total number of years of formal education; Digit Symbol value based on the total number of symbols correctly completed in 120 s; Trails B−A = time to complete Trails test A minus the time to complete Trails test B; Digits forward value based on the total number of items recalled. ERVT and WAIS math subtest were administered in Experiment 2. ERVT = extended range vocabulary test; WAIS = Wechsler Adult Intelligence Scale III.*p < .05 for age group comparisons.Materials.—Fine motor task.The fine motor task was a modified version of the multifinger sequence task (MFST) used in Fraser, Li, and Penhune (2009). The MFST is a serial reaction time (RT) task, in which a visual stimulus presented in one of four squares on a computer screen and participants tap in response to the stimulus with the four fingers of their right hand on four keys of a piano-like keyboard. The visual stimuli were presented repetitively in fixed 10-tap sequence (4-1-3-4-2-3-1-2-4-3) or in random 10-tap sequences. For the purposes of the current dual-task experiment, only the repeating sequence type was used. For each tap in the repeating sequence, the intertap interval was set at 1,000 ms, in which the stimulus stayed on the screen for 600 ms and disappeared for 400 ms. Therefore, the duration of a motor trial was 10 s. In the previous experiment (Fraser et al., 2009), age equivalence in the performance of the sequence was achieved after 10 presentations of the sequence; therefore, for the current experiment, 14 trials were presented during practice to ensure age equivalence prior to the test phase. Thirty trials were presented in each of the four test runs. For both the practice and the test sessions, participants completed half of the motor trials in isolation (single-task block) and half with the semantic task (dual-task block). An example of each trial type (single motor, single cognitive, and dual task) is presented in Figure 1.Figure 1.Graphic of the trials: single motor, single cognitive, and dual task. Dashed lines represent taps. Numbers under the dashed lines represent the key the participant had to tap. The fingers that corresponded to the keys were index = 1, middle = 2, ring = 3, and pinkie = 4. The solid line represents the time line of each trial (10 s). Arrows represent the word stimuli that were presented auditorially (i.e., mother, tractor, hammer). Note. Word stimuli were presented at random intervals during the trial and a trial could contain one, two, or three words.The visual stimulus in the sequence consisted of a 4.5-cm2 cartoon animal (i.e., “Rolly the Hamster”) that was programmed in C-Sharp and shown on a 19-inch Dell desktop monitor. Each stimulus was displayed in one of four horizontally presented colored 5-cm2 frames that stayed on the screen for the total duration of each trial. The participants responded to the stimuli on an M-Audio O2 Midi Controller piano keyboard. Participants were instructed to “catch the animal” by placing the four fingers of their right hand (i.e., index, middle, ring, and pinkie) on four marked keys, and the keyboard recorded the accuracy and RT of each key press.Cognitive task: semantic judgments.For this task, participants were presented auditorially with word stimuli at random time intervals and they were asked to judge if the word they heard was living (e.g., mother) or nonliving (e.g., chair). Word stimuli used in the current experiment were the same as those presented in Fraser and colleagues (2007). The trial time structure mimicked the motor trials, such that each trial lasted 10 s (see Figure 1). Furthermore, all participants had a practice session in which they judged 30 words and four test sessions that contained 60 words each. Half of the words were presented in isolation (single-task block) and half were presented with the fine motor task (dual-task block). Each list included an equal number of living and nonliving words to judge. The digitized words consisted of two-syllable high-frequency distinct nouns (written frequency less than or equal to one word per million; Kuçera & Francis, 1967) and were spoken in a female voice. To minimize the predictability of the presentation of the words, a trial could contain one, two, or three words. The minimum interstimulus interval (ISI) for each word presentation was 1,500 ms and the maximum was 7,000 ms. An algorithm programmed with Matlab software (The MathWorks, Inc., Natick, MA) produced ISIs that would result an equal distribution of the words across each 10-s trial (equal numbers of words presented at the beginning, middle, or end of the trial). The words presented in the practice lists were not reused in the test lists. All test words were presented twice with a minimum separation of two lists. The word stimuli were randomly ordered within each list and presented with customized software, C-Sharp, through a Dell Inspiron 1300 laptop. Participants heard the words through a Plantronics (Santa Cruz, CA) DSP-300 headset that also recorded vocal RTs. Speech recognition software (Microsoft Speech API) identified participants’ responses (“yes” for living words or “no” for nonliving words), and they were subsequently scored as correct or incorrect with Matlab software.Procedure.—The testing took place in the Adult Development and Aging laboratory at Concordia University. After informed consent, all participants underwent a task familiarization session. For the motor task, participants imitated simple forward (1-2-3-4-1-2-3-4-1-2-3-4) or backward (4-3-2-1-4-3-2-1-4-3-2-1) 12-element sequences to familiarize them with the keyboard and visual stimuli. For the semantic task, participants performed the word repetition baseline where they had to repeat 30 words that were presented auditorially. To ensure adequate hearing for the test phase, participants needed to score 90% or more on the word repetition baseline. All participants met this criterion.Participants then had practice in each of the conditions: single task (semantic), single task (motor), and dual task (semantic and motor). They completed seven trials per condition. Once they practiced the tasks, they completed four counterbalanced test runs of single motor, single semantic, and dual task. For each test run, there were 15 trials per condition. For both the practice and the test sessions, participants completed half of the motor trials in isolation (single-task block) and half with the semantic task (dual-task block). For both the practice and the test runs, participants were instructed that both tasks were equally important and that they should try to respond quickly and accurately. After the test session, participants completed the Digit Symbol, the Trail Making tests, the Digits Forward test, and a demographics questionnaire. Participants were debriefed and received course credit (younger) or an honorarium (older) for their time. The entire session lasted approximately 90 min.Statistical analyses.—Four dependent variables were calculated: accuracy and RT for the cognitive task and accuracy and RT for the motor task. For both the cognitive and the motor data, the mean correct RT (ms) for each trial type was calculated for each participant. The time window for valid motor responses had a 1,000-ms duration, which started 100 ms prior to the presentation of each stimulus, to allow for anticipated responses. Any correct tap within this time window was considered part of the mean RT. For the vocal RT data, RTs were calculated from the offset of the verbal stimuli and responses were excluded if they were ±3 SD from an individual’s overall mean RT. Only a small proportion of the responses were considered outliers (MOlder = 0.02, SE = 0.001; MYounger = 0.01, SE = 0.001). Accuracy for the cognitive task and the motor task were based on the number of correct responses (i.e., correct semantic judgments, correct taps) in all possible responses for each trial type (single and dual). For the cognitive accuracy, motor accuracy, and motor RT, the data were checked for outliers based on the group mean. No such outliers were found.DTCs were calculated for each of the four dependent variables. In the case of RT, dual-task RTs were subtracted from single-task RTs for each individual. For accuracy, single-task accuracy was subtracted from dual-task accuracy on an individual basis. The resulting difference scores represent four DTC scores: DTC motor accuracy, DTC motor RT, DTC semantic accuracy, and DTC semantic RT. For each variable, planned contrasts (α = .05) were conducted to assess age differences in DTCs. All post hoc analyses used a Bonferroni corrected p value (p = .025).Results and DiscussionMean values for single- and dual-task performances are reported in Table 2.Table 2.Mean Single- and Dual-Task Performance Values for Younger and Older AdultsExperiment 1Experiment 2Cognitive taskSemantic judgmentsMinus-1Minus-7YoungerOlderYoungerOlderYoungerOlderAccuracy Single89.2592.9099.5099.7475.0882.37 Dual91.0093.1498.3399.0467.5869.47Reaction times Single662.42762.63567.09600.641584.961462.89 Dual690.52786.61665.46641.201647.081439.82Fine motor taskSequential tappingYoungerOlderYoungerOlderYoungerOlderAccuracy Single96.9094.7697.8293.1396.9791.30 Dual96.3589.9097.4386.8684.2869.96Reaction times Single281.89405.48290.54413.03294.92391.59 Dual307.28466.98313.59447.24377.82486.62Notes: Accuracy values = percentage points (a value of 100 = all responses correct). Reaction time values in milliseconds. Although single-task sequential tapping requires only a response to the visual stimulus, results for Minus-1 and Minus-7 are reported separately in this table because they were presented separately in a Minus-1 or Minus-7 test run. There were no significant differences in single-task sequential tapping accuracy between Minus-1 and Minus-7, t(38) = 1.59, p = .120.Fine motor: MFST.—Accuracy.The t test revealed significant age differences in motor accuracy DTCs, t(39) = 2.23, p = .032, such that older adults had higher motor accuracy DTCs (M = 4.71%, SE = 1.73) than younger adults (M = 0.60%, SE = 0.53). Only older adults’ accuracy DTCs were significantly different from zero, t(20) = 2.72, p = .013.Reaction times.—The t test for motor RT DTCs resulted in a significant age difference, t(39) = 2.39, p = .022. Again, older adults had higher motor RT DTCs (M = 61.50 ms, SE = 11.47) than younger adults (M = 25.39 ms, SE = 10.20). After Bonferroni correction, both younger, t(19) = 2.49, p = .022, and older adults’, t(20) = 5.36, p < .001, motor RT DTCs were significantly different from zero.Cognitive: semantic judgment task.—Accuracy.The t test comparing younger and older adults on their cognitive accuracy DTCs was nonsignificant, t(39) = 0.92, p = .362. An additional analysis on the full sample revealed that the DTCs in accuracy were not significantly different from zero, t(40) = 1.24, p = .222. When split by age, neither younger, t(19) = 1.28, p = .215, nor older adults’, t(20) = 0.31, p = .760, accuracy DTCs were significantly different from zero.Reaction times.—In line with the accuracy results, the t test comparing younger and older adults’ vocal RTs DTCs was nonsignificant, t(39) = 0.35, p = .728. In this case, the t test comparing vocal RT DTCs to zero was significant for the whole sample, t(40) = 2.65, p = .012. However, when split by age, neither younger, t(19) = 1.96, p = .065, nor older, t(20) = 1.73, p = .098, adults’ vocal RT DTCs were significantly different from zero.Testing for trade-offs: within and across domains.—Within each domain (cognitive and motor), bivariate correlations between mean dual-task accuracy and speed (reciprocal of RT: 1/RT) were computed to test for speed-accuracy trade-offs. A negative correlation between speed and accuracy measures would be expected if participants were slowing to maintain accuracy levels or making more mistakes to maintain speed. For younger adults, the correlation between vocal accuracy and speed, r(18) = −.01, p = .973, was nonsignificant. The correlation between and motor accuracy and speed, r(18) = −.46, p = .044, was significant for younger adults; however, a close examination of the scatterplot for this correlation revealed that one participant was driving this finding. When this participant was removed, the negative correlation between motor accuracy and speed was no longer significant, r(17) = .05, p = .831. Older adults had significant positive correlations between motor accuracy and speed, r(19) = .53, p = .015, and vocal accuracy and speed, r(19) = .43, p = .050. This positive correlation suggests a relationship between speed and accuracy such that older individuals who were quick to respond also had high-accuracy scores and those who were slower to respond had lower accuracy scores.To rule out cross-domain trade-offs, bivariate correlations were conducted between motor and cognitive accuracy DTCs, as well as motor and cognitive RT DTCs. A negative correlation between these DTCs would suggest that lower costs in one domain (i.e., cognitive) are associated with greater costs in the other domain (i.e., motor). For correlations between domains in accuracy, no significant trade-offs were found for younger, r(18) = −.07, p = .773, or older adults, r(19) = −.03, p = .910. Similarly, there were no significant correlations between domains in RT for younger, r(18) = .22, p = .353, or older adults, r(19) = −.11, p = .648.Summary.—The results of Experiment 1 replicate the general findings of Crossley and Hiscock (1992) using a sequential tapping task. The lack of an age difference in DTCs in combination with the age differences in motor DTCs aligns well with the notion of an increasing role for cognition in fine motor performance. Beyond age differences in fine motor performances, Crossley and Hiscock demonstrated that these age differences increased when cognitive load increased. In the case of sequential tapping, additional support for age-related cognitive-motor interdependence should be found in conditions with greater cognitive load. In keeping with previous findings, we hypothesized that in Experiment 2, a high concurrent cognitive load would produce greater costs to sequential tapping than a lower cognitive load (for both age groups) and that this difficulty manipulation would have a greater impact on the older adults’ dual-task performances than the young.EXPERIMENT 2MethodParticipants.—Twenty younger adults (18–27 years) and 20 older adults (60–78 years) participated in the experiment. Recruitment and exclusion criteria were the same as in Experiment 1. In addition to the standardized tests administered in Experiment 1, all participants completed the Extended Range Vocabulary Test (ERVT; Educational Testing Service, 1976), and the Math subtest of the WAIS III, to assess vocabulary and math abilities, respectively. Descriptive statistics of the sample are presented in Table 1. All procedures were approved by the Concordia University Human Research Ethics Committee.Materials.—Fine motor task.The motor task was identical to that used in Experiment 1.Cognitive task: mental arithmetic.The cognitive task in this experiment had two levels of difficulty. For the Minus-1 level, participants subtracted one from randomly ordered two-digit numbers presented over headphones. For the Minus-7 level, participants subtracted seven from each stimulus. Stimuli consisted of two-digit numbers ranging from 11 to 99, not including numbers ending with seven (e.g., 17, 27, 37 …) or zero (e.g., 10, 20, 30 …). Two lists composed of 30 stimuli were used during the practice session. Sixty new stimuli were randomly arranged into four lists to be used in the four conditions (single Minus-1, single Minus-7, dual Minus-1, and dual Minus-7). The ISI range used in the current study (ISIs: minimum 2,300 ms and maximum 5,500 ms) was based on the average response times found in Abbud and colleagues (2009) for Minus-7. As compared with Experiment 1, the ISIs were lengthened here to accommodate the more complex cognitive tasks. In all other respects, the delivery of cognitive stimuli was the same as in Experiment 1.Procedure.—The testing took place in the Adult Development and Aging laboratory at Concordia University. After informed consent, all participants underwent the motor familiarization session described in Experiment 1. After the motor familiarization, all participants completed two practice blocks (15 trials each). In the first block, they completed a fixed order of Minus-1, single-task motor, and dual Minus-1; in the second block, they completed a fixed order of Minus-7, single-task motor, and dual Minus-7. Participants were instructed that both tasks were equally important and that they should try to respond quickly and accurately. Prior to the test runs, participants were asked to complete the Digit Symbol test.Once they had practiced the component tasks, they completed four test runs of the single motor, single cognitive, and dual-task trials. Runs 1 and 2 were always the Minus-1 difficulty level and Runs 3 and 4 were always the Minus-7 difficulty level. Within each run, the order was fixed: for Runs 1 and 3, single cognitive was always presented first, and for Runs 2 and 4, single motor was always presented first. The dual-task condition was always at the end of a run. These four runs were counterbalanced (i.e., 1-2-3-4, 4-1-2-3, 3-4-1-2, etc.) so that the difficulty manipulation was evenly distributed across the test session (i.e., with some participants having Minus-1, Minus-1, Minus-7, Minus-7; others Minus-7, Minus-1, Minus-1, Minus-7, etc.). For each of the four test sessions, there were 30 fine motor trials, 15 performed alone (single-task motor) and 15 performed concurrently with mental arithmetic (sequential tapping & Minus-sequential tapping & Minus-7). After the first two test runs, participants completed the Trail Making Test (A & B) and the ERVT, followed by the two remaining test runs. Finally, the participants completed the Digits Forward and the arithmetic subtest of the WAIS. Participants were debriefed and received course credit (younger) or an honorarium (older) for their time. The entire session lasted 90–120 min.Statistical analyses.—Accuracy and RTs were derived in the same way as Experiment 1. DTCs were calculated for each dependent variable in each domain (motor and cognitive) and difficulty level (Minus-1 and Minus-7). For the vocal RT data, responses were excluded if they were ±3 SD from each individual’s overall mean RT. Only a small proportion of the responses were considered outliers (MOlder = 0.01, SE = 0.001; MYounger = 0.01, SE = 0.002). For cognitive accuracy, motor accuracy, and motor RT, the data were checked for outliers ±3 SD from the group mean (younger and older) on single-task performances. One older adult was removed based on this criterion. Consequently, analyses were conducted on 20 younger and 19 older adults. Mixed factorial analyses of variance (ANOVAs; α = .05) were carried out using the four dependent variables (DTCs) with difficulty level (Minus-1 and Minus-7) as the within-subjects factor and age group (younger and older) as the between-subjects factor. All post hoc analyses used a Bonferroni corrected p value (.025).Results and DiscussionMean values for single- and dual-task performances are reported in Table 2, and DTCs for each domain are presented in Figure 2.Figure 2.Experiment 2: (A) Mean dual-task costs (DTCs) in motor accuracy by difficulty level (Minus-1 and Minus-7). (B) Mean DTCs in motor reaction times (RTs) by difficulty level. (C) Mean DTCs in cognitive accuracy by difficulty level. (D) Mean DTCs in cognitive RTs by difficulty level. Error bars are ±1 SE of the mean. Note. * = significant age difference in DTCs; + = DTCs are significantly greater than zero.Fine motor: MFST.—Accuracy.Figure 2A depicts the motor accuracy DTCs for both difficulty levels. The analysis revealed a main effect of difficulty level, F(1,37) = 44.33, p < .001, η2 = .545, such that the Minus-7 had higher costs (M = 17.00%, SE = 2.20) than Minus-1 (M = 3.30%, SE = 1.00). In addition, there was a main effect of age group, F(1,37) = 7.11, p = .011, η2 = .161, where older adults had higher DTCs in motor accuracy (M = 13.80%, SE = 2.00) than younger adults (M = 6.50%, SE = 1.90). The interaction was not significant, F(1,37) = 0.46, p = .503. The younger adults’ motor accuracy DTCs were not significantly different from zero, t(19) = 0.55, p = .590, for Minus-1 but were significantly different from zero for Minus-7, t(19) = 4.93, p < .001. For both difficulty levels, older adults’ motor accuracy DTCs were significantly different from zero, Minus-1: t(18) = 3.41, p = .003; Minus-7: t(18) = 5.85, p < .001.Reaction times.Figure 2B displays the motor RT DTCs. The ANOVA for motor RT DTCs resulted in a significant main effect of difficulty, F(1,37) = 39.48, p < .001, η2 = .516, where Minus-7 resulted in higher DTCs (M = 88.96 ms, SE = 9.55) than Minus-1 (M = 28.63 ms, SE = 5.56). The main effect of age, F(1,37) = 0.89, p = .351, and the interaction, F(1,37) = 0.01, p = .960, were nonsignificant. Analyses of the RT DTCs for the full sample confirmed that the DTCs for both difficulty levels were significantly different from zero, Minus-1: t(38) = 5.13, p < .001; Minus-7: t(38) = 9.38, p < .001.Cognitive: mental arithmetic.—Accuracy.Figure 2C depicts the cognitive DTCs in accuracy for both difficulty levels. The mixed factorial ANOVA revealed a main effect of difficulty, F(1,37) = 23.33, p < .001, η2 = .387, on the accuracy DTCs, such that DTCs were higher on Minus-7 trials (M = 10.20%, SE = 1.90) than on Minus-1 trials (M = 0.94%, SE = 0.40). Both the main effect of age, F(1,37) = 1.49, p = .230, and the interaction of difficulty and group, F(1,37) = 2.33, p = .135, were nonsignificant. Additional analyses on the full sample revealed that the DTCs in accuracy were significantly different from zero for both the Minus-1, t(38) = 2.13, p = .039, and the Minus-7, t(38) = 5.22, p < .001, conditions. When split by age, younger adults’ DTCs in Minus-1 were not significantly different from zero, t(19)= 1.47, p = .158, but they were significantly different from zero in Minus-7, t(19) = 3.10, p = .006. Similarly, the older adults’ DTCs were not significantly different from zero in the Minus-1 condition, t(18) = 1.91, p = .072, but were significantly different from zero in the Minus-7 condition, t(18) = 4.30, p < .001.Reaction times.Figure 2D depicts the cognitive DTCs in RTs for both difficulty levels. There were no significant effects in the cognitive RT data. The DTCs in vocal RTs did not differ by difficulty level, F(1,37) = 0.73, p = .398, nor by age group, F(1,37) = 1.32, p = .258, and the interaction between difficulty level and group did not reach significance, F(1,37) = 0.06, p = .816. Pooling together both age groups, the DTCs in the Minus-1 condition were significantly different from zero, t(38) = 3.42, p = .002, but the DTCs in the Minus-7 condition were not, t(38) = 0.37, p = .717. When split by age, only younger adults’ DTCs were significantly different from zero in the Minus-1 condition, t(19) = 3.97, p = .001, and neither age group had DTCs that were different from zero in the Minus-7 condition, Younger: t(19) = 1.77, p = .092; Older: t(18) = 0.21, p = .837.Testing for trade-offs: within and across domains.—At each level of difficulty, the mean dual-task scores within each domain were tested for a speed-accuracy trade-off. A negative correlation between speed and accuracy measures would indicate a trade-off. Correlations between mean dual-task motor accuracy and speed in the Minus-1 condition were nonsignificant for both age groups, Younger: r(18) = −.04, p = .884; Older: r(17) = .39, p = .104. Both younger and older adults had a significant positive correlation between dual-task motor accuracy and speed in the Minus-7 condition, Younger: r(18) = .52, p = .018; Older: r(17) = .62, p = .005. For both age groups, the cognitive dual-task correlations between accuracy and speed were nonsignificant for all conditions, Minus-1: r(37) = .26, p = .105; Minus-7: r(37) = .21, p = .202. Across both age groups, the lack of a significant negative correlation indicates that there was no speed-accuracy trade-off within domain.Using DTCs, cross-domain trade-offs (i.e., responding quickly in motor task but slowing in cognitive) were tested with bivariate correlations between the cognitive and motor DTCs. For correlations between motor and cognitive accuracy DTCs, neither younger, Minus-1: r(18) = .15, p = .540; Minus-7: r(18) = .20, p = .398, nor older adults, Minus-1: r(19) = .03, p = .913; Minus-7: r(19) = −.24, p = .334, demonstrated any significant cross-domain trade-offs. Similarly, there were no significant cross-domain trade-offs for correlations between motor and cognitive RTs for younger, Minus-1: r(18) = −.22, p = .349; Minus-7: r(18) = .22, p = .354, or older adults, Minus-1: r(19) = .05, p = .846; Minus-7: r(19) = −.07, p = .774.Summary.—Similar to the simple tapping findings of Crossley and Hiscock (1992) and the sequential tapping findings of Kemper and colleagues (2003), tapping sequentially while performing a cognitive task had a greater impact on older adults’ motor performances than younger adults. The younger adults were able to maintain their motor accuracy in the Minus-1 condition, whereas older adults demonstrated significant accuracy costs in both difficulty levels. Both groups slowed when sequentially tapping with a cognitive task but there was no age difference in the degree of slowing. In the cognitive measures, the pattern of results is similar for younger and older adults with the only exception being significant cognitive RT DTCs in the Minus-1 condition for the younger adults. The lack of speed-accuracy trade-offs and cross-domain trade-offs suggests that younger adults were not slowing to maintain performance on another measure. In the Minus-7 condition, the lack of significant cognitive RT DTCs in combination with significant motor DTCs for both age groups in both measures might indicate a prioritization of cognitive task under the highest cognitive load.GENERAL DISCUSSIONThe primary goal of this study was to examine the role of executive control in fine motor performance using a motor-cognitive dual-task paradigm. This study extends previous work on dual-task simple tapping (Crossley & Hiscock, 1992) and complex tapping (Kemper et al., 2003) with age differences found primarily in fine motor performances. The first experiment combined a low-load semantic judgment task with sequential tapping and older adults were slower and less accurate than younger adults on the sequential tapping task. In the second experiment, in which cognitive load was manipulated, there were age differences in motor accuracy, with older adults demonstrating costs in both difficulty levels, whereas motor accuracy costs only emerged in the harder condition for younger adults. Because older adults demonstrate costs in sequential tapping even in the conditions of lowest load and these costs reliably emerge in motor accuracy performance, we propose that older adults require greater executive control processes in order to perform the sequential tapping task.In both experiments, there was an asymmetry in the pattern of results, such that DTCs occurred mainly in the motor domain. This occurred despite differences in temporal predictability across the cognitive and motor tasks. Indeed, one might hypothesize that the less predictable cognitive task (that occurred at different time points during the trial) would be more affected by dual-task interference than the motor task that was presented in a more predictable fashion (one stimulus each second). Ultimately, across the different levels of cognitive load, the cognitive tasks interfered with the sequential nature of the tapping task and older adults were more affected by this interference than younger adults. The interference from the cognitive task affected the older adults’ fine motor performance even in the easiest condition, and younger adults only faltered when task demands were too great. With the exception of the Minus-1 condition RT measure (cognitive and motor), all performance costs for younger adults were found in the harder Minus-7 condition. Perhaps mild cognitive loads taxed younger adults’ coordinative processes (i.e., coordinating the performance of the two tasks). Whereas for older adults, all cognitive loads were sufficiently challenging that key press accuracy or response selection in the motor task was affected.Given that the sequence we presented was repeated throughout the each block of trials, younger and older adults may have encoded the sequence of key presses into a single action plan (Tubau, Hommel, & Moliner, 2007). Findings with younger adults have demonstrated that execution of an action plan can be disrupted by visual and auditory verbal distracters. In addition, sequence learning and action plans have both been shown to involve the prefrontal cortex (Tubau et al.). The prefrontal cortex and the executive control processes it subserves are known to decline with normative aging (Verhaeghen & Cerella, 2002). Therefore, in the current experiment, because older adults relied more heavily on executive control functions for sequential tapping, they demonstrated greater performance costs than their younger counterparts. In support of this proposal, existing sequence learning research (Aizenstein et al., 2006) has found age differences in frontal activity during concurrent sequence learning, such that older adults show greater activity than younger adults in the left dorsolateral prefrontal cortex.The results of the current experiment are also consistent with our previous findings in dual-task walking experiments, which used the same cognitive tasks (Fraser et al., 2007; Li et al., 2008). In particular, Fraser and colleagues (2007) found age differences only in walking performance when performing the semantic task, and Li and colleagues (2008) demonstrated maintenance of walking performance during the Minus-1 condition for younger adults but costs similar to older adults in the harder condition. Similarly, in Experiment 1, there was age equivalence in performance of the semantic task, and age differences emerged in sequential tapping. Furthermore, in Experiment 2, although younger adults slowed their sequential tapping in the Minus-1 condition, they maintained their accuracy when older adults demonstrated accuracy costs and both groups had similar costs in the Minus-7 condition. These similarities suggest that gait and sequential tapping may draw on similar executive control functions. The age-related dual-task effects reported by Crossley and Hiscock (1992) may have been a reflection of age-related reductions in general dual-task coordination processes rather than an indication that simple tapping requires executive control. Indeed, previous research suggests that simple tapping does not rely on executive functions (Hausdorff, Yogev, Springer, Simon, & Giladi, 2005). The similarity of the current pattern of results with that of previous walking research (Fraser et al., 2007; Li et al., 2008) suggests that executive functions might play a role in both walking and fine motor DTCs and that the previous walking findings were not primarily driven by postural threat. Although it is clear that postural threat influences attentional allocation (Brown et al., 1999), it has also been found that different degrees or levels of difficulty of postural threat (Lajoie, Teasdale, Bard, & Fleury, 1996) can modulate DTCs in older adults.CONCLUSIONSTaken together, the findings extend the research on aging and dual-task fine motor performance in demonstrating that concurrent sequential tapping costs are greater in older adults due to the disruption of a planned execution of taps at the executive processing level. Under low cognitive load, younger adults have a more proceduralized or automatic approach to the sequential tapping task that does not require executive control. In contrast, older adults demonstrate costs at every load level demonstrating the cognitive penetration of motor task performance (Teasdale, Bard, LaRue, & Fleury, 1993).FUNDINGThis project was funded by a Natural Sciences and Engineering Research Council of Canada PhD scholarship awarded to S.A.F. and a Canadian Institutes of Health Research operating grant awarded to K.Z.H.L. and V.B.P. (MOP-67757, 68807).Many thanks to Alejandro Endo, Ricco Boma, Madeleine Ward, Laura Fontil, Stephanie Torok, Monique Leblanc, Hilary Greenstone, Simon Desmarais-Zalob, and Vanessa Raccio who all helped with this project.AbbudGACLiKZHDeMontRGAttentional requirements of walking according to the gait phase and onset of auditory stimuliGait and Posture200930227232AizensteinHJButtersMAClarkKAFigurskiJLStengerVANebesRDReynoldsCFIIICarterCSPrefrontal and striatal activation in elderly subjects during concurrent implicit and explicit sequence learningNeurobiology of Aging200627741751AlbinetCTomporowskiPDBeasmanKAging and concurrent task performance: Cognitive demand and motor controlEducational Gerontology200632689706BaddeleyAStussDTKnightRTFractionating the central executivePrinciples of frontal lobe function2002New YorkOxford University Press246277BaltesPBLindenbergerUEmergence of a powerful connection between sensory and cognitive functions across the adult lifespan: A new window to the study of cognitive aging?Psychology and Aging1997121221BleAVolpatoSZulianiGGuralnikJMBandinelliSLauretaniFBartaliBMaraldiCFellinRFerrucciLExecutive function correlates with walking speed in older persons: The InCHIANTI studyJournal of the American Geriatrics Society200553410415BrownLAShumway-CookAWoollacottMHAttentional demands and postural recovery: The effects of agingJournal of Gerontology: Series A: Biological and Medical Sciences199954M165M171CrossleyMHiscockMAge-related differences in concurrent-task performance of normal adults: Evidence for a decline in processing resourcesPsychology and Aging19927499506Educational Testing ServiceExtended Range Vocabulary Test: Kit of factor-referenced cognitive tests1976Princeton, NJAuthorFraserSALiKZHDeMontRGPenhuneVBEffects of balance status and age on muscle activation while walking under divided attentionJournal of Gerontology: Psychological Sciences200762B171178FraserSALiKZHPenhuneVBA comparison of motor skill learning and retention in younger and older adultsExperimental Brain Research2009195419427HausdorffJMYogevGSpringerSSimonESGiladiNWalking is more like catching than tapping: Gait in the elderly as a complex cognitive taskExperimental Brain Research2005164541548HuxholdOLiSSchmiedekFLindenbergerUDual-tasking postural control: Aging and the effects of cognitive demand in conjunction with focus of attentionBrain Research Bulletin200669294305KahnemanDAttention and effort1973Englewood Cliffs, NJPrentice-HallKemperSHermanRELianCHTThe costs of doing two things at once for young and older adults: Talking while walking, finger tapping, and ignoring speech or noisePsychology and Aging200318181192KetchamCJStelmachGEBirrenJSchaieKWAge-related declines in motor controlHandbook of psychology of aging20015th ed.San Diego, CAAcademic Press313348KlugerAGianutsosJGGolombJFerrisSHGeorgeAEFranssenEReisbergBPatterns of motor impairment in normal aging, mild cognitive decline, and early Alzheimer’s diseaseJournal of Gerontology: Series B: Psychological Sciences and Social Sciences199752BP28P39KramerAFMaddenDJCraikFIMSalthouseTAAttentionThe handbook of aging and cognition20083rd ed.New YorkPsychology Press189249KrampeRTAging, expertise and fine motor movementNeuroscience and Behavioral Reviews200226769776KuçeraHFrancisWNComputational analysis of present day American English1967Providence, RIBrown University PressLajoieYTeasdaleNBardCFleuryMUpright standing and gait: Are there changes in attentional requirements related to normal aging?Experimental Aging Research199622185198LiKZHDeMontRPenhuneVFraserSAAbbudGThe dynamic relationship between cognition and walking under dual-task conditions in healthy agingInternational Journal of Psychology200843366LiKZHLindenbergerURelations between aging sensory/sensorimotor and cognitive functionsNeuroscience and Biobehavioral Reviews200226777783LiKZHLindenbergerUFreundAMBaltesPBWalking while memorizing: Age-related differences in compensatory behaviorPsychological Science200112230237LövdénMSchäeferSPohlmeyerAELindenbergerUWalking variability and working memory load in aging: A dual process account relating cognitive control to motor performanceJournal of Gerontology: Psychological Sciences200863BP121P128MendelsonDNRedfernMSNebesRDJenningsJRInhibitory processes relate differently to balance/reaction time dual tasks in young and older adultsAging, Neuropsychology, and Cognition201017118SpreenOStraussEA compendium of neuropsychological tests: Administration, norms, and commentary1998New YorkOxford University PressTeasdaleNBardCLaRueJFleuryMOn the cognitive penetrability of postural controlExperimental Aging Research199319113TubauEHommelBMolinerJLModes of executive control in sequence learning: From stimulus-based to plan-based controlJournal of Experimental Psychology: General20071364363VerhaeghenPCerellaJAging, executive control, and attention: A review of meta-analysesNeuroscience & Biobehavioral Reviews200226849857WechslerDManual of the Wechsler Adult Intelligence Scale—III1997New YorkPsychological CorporationWoollacottMShumway-CookAAttention and the control of posture and gait: A review of an emerging area of researchGait and Posture200216114Yogev-SeligmannGHausdorffJMGiladiNThe role of executive function and attention in gaitMovement Disorders200823329342
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B19D36BBF10A2648F2CD63FC5EF4968BE5075C9.txt b/test/dataset/in/resources/corpus/Clean_0B19D36BBF10A2648F2CD63FC5EF4968BE5075C9.txt
deleted file mode 100644
index 520bd48..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B19D36BBF10A2648F2CD63FC5EF4968BE5075C9.txt
+++ /dev/null
@@ -1 +0,0 @@
- geront Gerontologistgeront The Gerontologist The Gerontologist 0016-9013 1758-5341 Oxford University Press 33510.1093/geront/43.3.335 USE OF TECHNOLOGY A Comparison of Assistive Technology and Personal Care in Alleviating Disability and Unmet Need Agree Emily M. PhD 1 Freedman Vicki A. PhD 2 Address correspondence to Emily M. Agree, PhD, Department of Population and Family Health Sciences, Johns Hopkins Bloomberg School of Public Health, 615 North Wolfe Street, Room E4014, Baltimore, MD 21205. E-mail: eagree@jhsph.edu 6 2003 43 3 335 344 16 9 2002 25 4 2002 The Gerontological Society of America 2003 Purpose:The authors examine differences in reports of residual disability and unmet need by type of long-term care arrangement (assistive technology or personal care).Design and Methods:This study compares three specific dimensions of residual difficulty (pain, fatigue, and time intensity) and reports of unmet need across care arrangements. Samples from the U. S. 1994–1995 National Health Interview Survey Phase 2 Disability Supplements include adults with limitations in bathing, transferring, walking, and getting outside.Results:Even when differences in underlying disability are accounted for, assistive technology (AT) confers no additional benefit in the three dimensions of residual difficulty analyzed here. AT users equally or more often report that tasks are tiring, time consuming, or painful, even when they use assistance. Though this would appear to indicate unmet needs for care, fewer AT users report a desire for hands-on personal care.Implications:Though disability alleviation by technology is no better on specific dimensions of difficulty, technology users report less unmet need for personal care. Designing appropriate and cost-effective home care for adults with disabilities requires a better understanding of the ways in which technology users may differ from others and the circumstances under which technology can be most effective. Long-term care Effectiveness Activities of daily living hwp-legacy-fpage 335 hwp-legacy-dochead RESEARCH ARTICLE Designing appropriate and cost-effective home care for adults with disabilities is a priority in aging societies. The number of older adults with disabilities is large and continues to grow, despite reductions in disability rates over the past two decades (Schoeni, Freedman, & Wallace, 2001). Currently, approximately 15% of the adult U. S. population, or 40 million people, are limited in activities as a result of a chronic health condition (Kaye, LaPlante, Carlson, & Wenger, 1996). The vast majority of adults with disabilities live in the community, and facilitating their ability to live independently is an increasingly important public health concern. More than 75% of older adults with disabilities use some kind of assistive device, most often to assist with mobility (Russell, Hendershot, LeClere, Howie, & Adler, 1997), and the number has been growing over time (Manton, Corder, & Stallard, 1993; Russell et al., 1997). Devices are used both independently and in conjunction with personal caregiving services. However, the relative advantages of assistive technology and personal care are not well understood, in part because most studies evaluate one or the other, but rarely the two together. Most research on assistive technology evaluates the efficacy and costs of specific devices (e.g., Chen, Mann, Tomita, & Burford, 1998; Gitlin & Levine, 1992; Kohn, LeBlanc, & Mortola, 1994; Mann, Hurren, Charvat, & Tomita, 1996; Sanford, Arch, & Megrew, 1995; Steinfeld & Shea, 1995), or it broadly examines the use of any assistive devices (Manton et al., 1993; Norburn et al., 1995; Zimmer & Chappell, 1994) or the number of devices used (Hartke, Prohaska, & Furner, 1998; Heide et al., 1993; Mann, Hurren, & Tomita, 1993). Studies of assistive technology emphasize the extent to which the acceptability and effectiveness of home-based technologies vary, depending on the training provided (Gitlin & Levine, 1992), the amount of stigma perceived by the user (Arras, 1995), and how much the introduction of assistive technologies transforms the home from a personal space to a health care delivery location (Tamm, 1999). Similarly, population-based efforts to understand the effectiveness of personal care arrangements often altogether exclude a consideration of self-care through assistive technology (Allen & Mor, 1997; Desai, Lentzner, & Weeks, 2001; Tennstedt, McKinlay, & Kasten, 1994). These studies instead focus on the trade-offs between formal and informal care or on the extent to which formal care can defer institutionalization (Soldo & Freedman, 1994). Attempts to measure outcomes that reflect the effectiveness of interventions for disability are relatively new and differ depending on whether the goal is to evaluate rehabilitation services specifically or more general effects on the ability to live independently (Andresen, Lollar, & Meyers, 2000). A recent national study examined the effectiveness of assistive technology in terms of its impact on hours of personal care for older adults with disabilities, assuming that the decision to use devices is independent of the acquisition of caregivers (Allen, Foster, & Berg, 2001). This study found that those who use assistive technology also use fewer hours of personal care. Two additional small-scale experiments have evaluated the effectiveness of interventions to increase the use of assistive devices (Mann, Ottenbacher, Fraas, Tomita, & Granger, 1999) and environmental modifications for dementia (Gitlin, Corcoran, Winter, Boyce, & Hauck, 2001). Although sample sizes were small, both studies suggested that increased use of technology may slow functional declines, lower health care costs, and increase efficacy among some caregivers. However, none of these studies directly compare the effect of assistive technology and personal care on functional health. Two other nationally representative studies have evaluated the relative effectiveness of assistive technology and personal care in terms of reported reduction of disability (Agree, 1999; Verbrugge, Rennert, & Madans, 1997). Both studies found that the use of assistive technology was associated with greater reduction of overall levels of difficulty in either activities of daily living (ADLs) or mobility impairments when compared with personal care. Though each used distinct measures and different data sets, these studies both relied on broad judgments about levels of difficulty with tasks to determine effectiveness. In sum, much of the prior research on this topic either has failed to directly compare outcomes related to assistive technology and personal care, begging the question as to when and whether each is most appropriately used (Allen et al., 2001; Gitlin et al., 2001; Mann et al., 1999), or has relied on global assessments of residual difficulty that do not offer insight into the mechanisms by which different care arrangements alleviate needs (Agree, 1999; Verbrugge et al., 1997). The purpose of this study is to further investigate the relative effectiveness of assistive technology and personal care in alleviating disability among adults with difficulty in ADLs. Using nationally representative disability survey data, we compare three dimensions of disability for which both underlying and residual difficulty can be assessed—pain, fatigue, and time intensity—across different combinations of assistive technology and personal care use. We also examine reports of the need for hands-on assistance across care arrangements. Framework The conceptual framework of the disablement process (e.g., Pope & Tarlov, 1991; Verbrugge & Jette, 1994) gives us some foundation for understanding how different care arrangements may alleviate different dimensions of disability. In this framework, disability is socially defined, being the product of the individual's functional capacity, the demands of the physical and social environment, and his or her own expectations about daily life. Verbrugge and Jette (1994) make the additional important distinction between underlying difficulty—the level of difficulty without help or special equipment—and residual difficulty—that is, with whatever assistance is generally used. Both underlying and residual disability encompass various levels of difficulty. At one extreme, an individual may be completely unable to carry out an activity without assistance; alternatively, an individual may be able to carry out an activity without assistance but with some or a lot of difficulty. The most commonly reported dimensions of difficulty in this context are pain and fatigue (Albrecht & Devlieger, 1999); tasks that are time consuming to carry out (or more time consuming than under usual or optimally functioning circumstances) may also be identified as being difficult. Closely related to residual difficulty is the concept of unmet need, that is, whether an individual thinks he or she requires more assistance. The disabling effects of an impairment can be reduced or potentially eliminated in a number of ways (see Figure 1). Individuals may compensate for their disability by using personal care, they may alter their ability to carry out the task (through rehabilitation or device use), or they may alter the demands of the environment (through, e.g., a home modification). Each of these approaches may be adopted in isolation or may be combined into a more complex care arrangement. Compensation, because it involves the cooperation of one or more helpers, creates a state of dependency. In contrast, assistive technologies and related environmental modifications enhance independence either by expanding the capacity of the individual or reducing the demands of the environment. In this article we hypothesize that different approaches to bridging disability may alleviate difficulty in different ways. We expect that assistive devices may alleviate difficulty by minimizing the pain associated with a task, whereas personal care may lessen the time and energy necessary to accomplish a task. Because devices increase an individual's capacity for self-care, we expect devices to be more effective than personal care in reducing unmet need. Methods Data Data are from the 1994 and 1995 Phase 2 Disability Supplements to the U. S. National Health Interview Survey, hereafter referred to as the NHIS-D2 (National Center for Health Statistics, 1998a, 1998b). These data provide detailed information on the use of assistive devices and environmental modifications for limitations in ADLs. The U. S. NHIS is a nationally representative health survey of approximately 48,000 households (122,000 persons) conducted annually by the U. S. National Center for Health Statistics. The survey consists of a core questionnaire that collects health information and demographic background on every person in the household. Each year supplemental surveys on specific health topics are included. In 1994 and 1995, a supplement on disability was administered in two phases over approximately 3 years. Phase 1, administered at the same time as the NHIS core interview, collected basic measures of impairment and functional health from a designated individual about all household members. Household members who were identified as disabled in Phase 1, using a broad definition of disability (the presence of any impairment, functional limitation, or disability), were interviewed in person 7–17 months later to obtain more detailed information (Phase 2). Separate questionnaires were administered to children (under 18) and adults (age 18 and older). Response rates were approximately 95% for the core and 87% for the supplements. Sample Selection To examine the effectiveness of care arrangements in alleviating disability, we focus on the adult population with underlying limitations in ADLs—those who report that they have difficulty carrying out an activity by themselves and without special equipment. This definition includes both persons currently using some form of care as well as a substantial number who report difficulty but make use of no personal care or assistive devices. Because assistive devices tend to be task specific in response to disabling conditions (Agree & Freedman, 2000), analyses are conducted separately for samples reporting underlying difficulty with each of four daily activities: bathing or showering (n = 3,493), getting in or out of bed or chairs (n = 3,834), walking (n = 7,051), and getting outside (n = 3,542). Although respondents also were asked about difficulty with dressing, eating, and getting to or using the toilet, sample sizes were too small, and the use of assistive technology too rare in the case of dressing and eating, to analyze care arrangements for these activities. Respondents with multiple ADL difficulties may be in more than one sample. Measures of Assistive Technology and Personal Care Questions are asked about assistive technology in the ADL assessment section of the NHIS-D2 instrument. For each ADL, participants who answer that they use a special device or piece of equipment to perform a task are asked to name the type of technology that they use. Interviewers use a checklist to simplify the recording of answers. For the four activities we analyze here, the specific equipment includes hand bars or rails, and bath stools, seats, or chairs (bathing); canes, walkers, special cushions, lift chairs, a hospital bed, and a trapeze or sling (transferring); and canes, walkers, crutches, and wheelchairs (getting around indoors and going outside). An analysis of specific items was not possible because devices other than canes and walkers were not mentioned frequently enough to allow separate analyses. Grouping all assistive technologies for a given activity together also allowed us to include sample persons who reported using special equipment without providing a specific type. Respondents also were asked in the same section whether they received hands-on help from another person with that activity. For each ADL, information was obtained on the type and amount of personal care provided to the respondent. On the basis of this information, we classified respondents into four groups: no care, assistive technology only, hands-on personal care only, and both personal care and assistive technology. Level of Underlying Difficulty Comparing the effectiveness of care arrangements requires some ability to control for the differential selection of these arrangements based on the severity of underlying disability. For each activity, respondents were classified as having some difficulty, a lot of difficulty, or being completely unable to carry out the activity without assistance. We also present analyses for all respondents and for the subgroup reporting a lot of underlying difficulty with a given task. Measures of Effectiveness The NHIS-D2 provides task-specific measures of residual difficulty and unmet need. The survey included information on three explicit dimensions of residual difficulty. For each activity, respondents (with the exception of those who were completely unable to perform the task) were asked to assess whether the activity is “very tiring,” “takes a long time,” or is “very painful” under two circumstances: first, if it is performed without personal care or assistive technology, and second, when it is performed with the assistance of personal care or devices (if used). This information was used to classify whether respondents' difficulty with an activity was “eliminated” when using personal care or assistive technology or whether they report residual difficulty even when using assistance. In addition, all respondents who reported underlying difficulty with a given task were asked whether they needed (more) hands-on help with that activity, which was used to derive a measure of unmet need. Limitations of the Data Although the NHIS-D is a rich source of information on the health and care arrangements of persons with disabilities in the United States, there are some limitations of the data that merit attention here. First, as already noted, the NHIS-D2 excludes the most severely disabled (those who report that they are unable to perform a task without assistance) from the questions about specific dimensions of underlying and residual difficulty, thus limiting our conclusions to a more moderately disabled population for those particular analyses. Second, the sensitivity of a disability measure to clinically detectable differences in functional status is an important factor in comparing across types of care (Cohen & Marino, 2000). In this study, the dichotomous indicators for each specific dimension of difficulty made it possible to examine effectiveness only in terms of complete elimination of difficulty. We are therefore limited in our ability to examine more refined gradations of effectiveness in disability reduction as has been done in previous studies (Agree, 1999; Verbrugge et al., 1997). Finally, the cross-sectional nature of the data make it impossible to assess the role that the timing of acquisition of assistive technology and personal care plays in the effectiveness of such arrangements. As already noted, prior research has indicated that assistive technologies may be more effective than personal care in alleviating disability. Consequently, we expect in the following analyses to find that, when we control for the level of underlying need, those who use assistive devices are less likely than those using personal care to report (a) residual difficulty on all dimensions of disability and (b) unmet need for personal care. Methods Because the NHIS-D2 is not a simple random sample but instead has a complex design with geographic clustering, statistical tests have been adjusted to take into account the complex survey design. All estimates are weighted with analytic weights normed to the appropriate sample size. Results discussed in the text are statistically significant at the p <.05 level unless otherwise noted. Results Assistive device use is quite common among persons who have underlying difficulty with the four activities considered here. Between 16% (of persons with underlying difficulty transferring) and 39% (of those who have underlying difficulty walking) use assistive technology alone to accommodate their difficulty (see Table 1). An additional 13–33% use assistive technology in combination with personal care. Taken together, 29–64% use one or more devices, with devices dominating most clearly in basic mobility activities (walking and going outside). Those with underlying difficulty bathing or getting outside are slightly older on average (67 vs. 63 years). However, the distributions by sex, minority status, and education are similar across the samples, with a majority of people being female; approximately 16–17% are non-White, approximately 7% are Hispanic, and the majority have at least completed high school (∼55%). Table 2 shows the distribution of the severity of underlying disability stratified by the type of care used for each of the four ADLs. This table confirms that the combination of care adopted is clearly related to the overall amount of underlying difficulty with the task. Those who use no care of any kind are the least disabled, with between 66% and 79% reporting some underlying difficulty and 1–7% reporting they are completely unable to carry out the activity without assistance (11–20% report some difficulty and 48–60% report being completely unable to carry out the activity). Among those who do use some type of assistance, those who use assistive devices alone report more moderate levels of underlying difficulty compared with those who use hands-on help (alone or in combination with assistive technology). Only 13–21% of those who use technology by itself state that they would be unable to perform the task without assistance, compared with 26–43% of those depending on personal care alone. As already noted, those who rely on both personal care and assistive technology report the greatest amount of difficulty, with 51–59% reporting that they would be completely unable to perform the task without assistance. Figure 2 shows for those with underlying difficulty in each of the four activities the proportion who also report underlying difficulty on three specific dimensions: whether the activity is very tiring, very painful, or takes a long time. The top panel shows estimates for all levels of difficulty, and the bottom panel is limited to those reporting a lot of underlying difficulty. For example, 63% of those with difficulty bathing say that they find the activity to be tiring when they do not have help or equipment. We find that substantial proportions of persons who report underlying difficulty with a given task find the activity very tiring, time consuming, or very painful when they do not have hands-on help or use equipment (47–76%, depending on the specific task). Not surprisingly, the proportions reporting underlying difficulty on each of the specific dimensions are consistently higher among those who report a lot of underlying difficulty on the task, ranging 59–88% on average for all activities. We also investigated (not shown) whether the proportion who reported underlying difficulty on each of the three specific dimensions varied across care arrangements. We found that those using both personal care and assistive technology were generally more likely than those using either one alone to report that a task is very tiring, takes a long time, or is very painful when carried out without assistance. When we stratified by the level of underlying difficulty, we found a much weaker relationship between care arrangement and specific dimensions of underlying difficulty. We next examine reports of residual difficulty, using the same three dimensions (whether the task is tiring, time consuming, or painful) and controlling again for the level of underlying difficulty. As shown in the top panel of Table 3, among persons with underlying difficulty, those who use assistive technology alone report residual difficulty more often than those using personal care (with or without devices). This relationship is significant, however, only for two dimensions (whether the task is tiring or time consuming) and three of the tasks (bathing, transferring, and going outside). For example, among those who report underlying difficulty with bathing, 75% of assistive technology users report that the task is very tiring when they use their equipment, compared with only 55% of those who use personal care alone (p <.05). Pain varies the least across care arrangements and is not significantly related to the type of care for any of the activities. The same general pattern is evident when the investigation is restricted to those with more severe activity limitations (bottom panel of Table 3), although the differences are statistically significant only for bathing. For many activities and dimensions of difficulty, especially pain, care arrangements are not significantly related to residual difficulty. However, in those cases in which a significant association is found, it is opposite to the expected direction—assistive technology users more often report that a task is tiring or takes a long time even when they use assistance. Figure 3 shows the proportion of people reporting a need for hands-on help by activity and level of underlying severity of disability. The proportions shown here are consistent with those of other studies (e.g., Desai et al., 2001) and generally low in comparison with the reports of residual difficulty shown in Table 3. The proportions reporting a need for (more) hands-on help are consistent across levels of underlying disability and types of care (the highest levels are 12–18% reported by those with a lot of underlying difficulty and who use personal care alone). However, those using assistive technology alone are consistently less likely to state that they need any hands-on help, at levels as low or lower than for those who use no care. For bathing, levels of unmet need are highest, but the differences across types of care are minimal and not significant (significance tests not shown). For the three mobility-related ADLs, the differences between those who use personal care (either alone or in combination with assistive technology) and those who do not (using either assistive devices alone or no care at all) are substantial and consistently significant across levels of disability. Discussion Facilitating the ability to live and work independently in the community is a central goal for gerontologists, policymakers, and public health practitioners. Although it is difficult to measure optimal care arrangements for adults with disabilities (Jette & Keysor, 2002), the results presented here further our understanding of the relative effectiveness of assistive technology and personal care in alleviating difficulty with daily activities. They also provide insights into the conceptualization and measurement of unmet need. Consistent with previous studies (Agree & Freedman, 2000; Manton et al., 1993; Verbrugge et al., 1997), we find that assistive devices are the most common means of managing day-to-day tasks for older adults and that the extent to which assistive technology and personal care are used in combination is closely related to the amount of difficulty reported. Unlike previous researchers, however, we compare across care arrangements reports of residual disability on specific dimensions such as fatigue, duration, and pain. We find that although adults with disabilities using only assistive technology tend to be less disabled than users of personal care, they are more likely to report that a task is tiring or time consuming when they use their equipment. We also find that users of assistive technology are less likely to report a need for any hands-on help than those already using personal care. Our analysis is limited in several ways. We were unable to compare the effectiveness of care arrangements on specific dimensions for the most severely disabled; thus we cannot draw any conclusions about the relative efficacy of care arrangements for those unable to perform a task without assistance. Moreover, we were able to explore only the complete elimination—rather than the reduction—of disability. There are many reasons that devices, which require some physical and cognitive effort to use, might reduce overall difficulty more easily than personal care, but not completely eliminate it. Future research should consider the relative effectiveness of personal care and assistive technology both on specific dimensions and in finer gradations. Finally, we were unable to explore with these cross-sectional data the more dynamic acquisition process underlying the disablement process. For example, a long-term user of assistive technology may become more proficient and consequently more satisfied than one who has newly acquired the device. Surely this is a fruitful area for further exploration. Despite these limitations, our analyses provide several new insights into the relative effectiveness of personal care and technology. First, the present study suggests a need to probe more deeply into the salient dimensions of disability that assistive technology and personal care alleviate. Using global measures of difficulty, prior research found that assistive technology appeared to be more effective than personal care in reducing or eliminating disability (Agree, 1999; Verbrugge et al., 1997). We find instead that users of such technology report similar or greater amounts of residual difficulty on the three specific dimensions of residual difficulty with ADL tasks in these data. These findings do not contradict earlier work, but rather inform it, by expanding our definition of effectiveness. Qualitative research may help identify the most important dimensions related to the effectiveness of different care arrangements. In one of the few studies to address this question, Albrecht and Devlieger (1999) report that pain and fatigue are among the most common problems reported, but further work illuminating the nature and meaning of these dimensions is warranted. Second, we find that persons with more severe disabilities are more likely to use personal care and less likely to use assistive technology exclusively. At the same time, like Agree (1999), we show the relative effectiveness of personal care and assistive technology on specific dimensions appears to diminish at higher levels of disability. These findings suggest that selection according to underlying disability is critical to take into account in any comparisons of effectiveness across care arrangements. In prior work this selection may have artificially increased the benefit attributed to assistive technology relative to personal care. In this study, we find the disadvantage of such technology on specific dimensions of disability does not appear to be explained by the underlying severity of disability, which is lower among assistive technology users. Third, the finding that assistive technology users less often report a need for hands-on help despite greater residual difficulty illustrates the likelihood that factors other than the severity of underlying disability influence the choice of care arrangements and must be attended to in both research and clinical practice. In particular, the use of assistive technology may be related to psychological factors such as receptivity, self-efficacy, and motivation, all of which have been shown to be related to the success of rehabilitative efforts (Arnstein, 2000; Grahn, Ekdahl, & Borquist, 2000; Zimmer & Chappell, 1999). Values related to autonomy and privacy also are important aspects of the acceptance and use of technology in rehabilitation and home care (Tamm, 1999). Our findings also have implications for the conceptualization and measurement of unmet needs for care. Studies of unmet need have traditionally focused on personal care (Allen et al., 2001; Gitlin et al., 2001; Mann et al., 1999), finding, as we did here, quite low reported levels of unmet need, even among adults with severe disabilities. Such research may be missing a substantial amount of unmet need for assistance, were the concept more broadly defined to include needs for and use of assistive technology. Not only may there be unmet needs that can be better fulfilled by technology rather than personal care, but the findings from this study suggest that some proportion of the disabled population may have unmet needs but would not report a desire for personal care. Mann, Hurren, and Tomita (1995) reported that persons with arthritis expressed a need for additional devices, even describing new technologies to be invented. Including specific questions about the need for additional devices to patient assessments and survey instruments would contribute a great deal to our understanding of and interventions to reduce unmet needs. In sum, indicators such as residual disability and unmet need do not in and of themselves allow us to identify the reasons that a particular care arrangement is more or less satisfactory. They do, however, make it possible to uncover circumstances in which types of care are less than optimal (Eldar, 2000). To meet the goal of facilitating the ability of adults to live and work independently in the community, it is essential for future research to explore the role of self-care more thoroughly. In particular, greater attention should be paid to identification of the different dimensions by which assistive technology and personal care alleviate difficulty and enhance independence. Prior versions of this article were presented in Syracuse, NY, at the Center for Policy Research Seminar Series in Aging, Labor, and Public Finance in October 2000 and in Washington, DC, at the Annual Meetings of the Population Association of America in March 2001. We acknowledge the able assistance of Jonas Marainen and Hakan Aykan in preparing the data and the tables, and Mary Alice Ernish for bibliographic assistance. We also thank Donna Strobino, Douglas Wolf, and Timothy Smeeding for their helpful comments. This work was supported by Grant R01-AG15135 from The National Institute on Aging. 1 Department of Population and Family Health Sciences, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD. 2 Polisher Research Institute, Madlyn and Leonard Abramson Center for Jewish Life (formerly Philadelphia Geriatric Center), North Wales, PA. Decision Editor: Laurence G. Branch, PhD Figure 1. Alleviating underlying disability through personal care and assistive technology Figure 2. Percentage of those with difficulty on each activity of daily living who report that a task is tiring, time consuming, or painful without assistance (those who are completely unable to perform an activity are not included) Figure 3. Percentage of those reporting an unmet need for hands-on care by type of assistance and activity. AT = assistive technology. Table 1. Care Arrangements and Sample Characteristics. Characteristic Bathing (n = 3,493) Transferring (n = 3,834) Walking (n = 7,051) Getting Outside (n = 3,542) Care arrangement (%) No care 25.8 50.1 40.0 23.3 AT only 20.5 15.9 39.2 31.0 Personal care only 24.6 20.6 6.3 13.1 Personal care and AT 29.2 13.4 14.5 32.7 Total 100.0 100.0 100.0 100.0 Age (mean; years) 67.7 62.6 63.8 66.9 % Female 63.5 61.2 60.4 64.2 % Non-White 16.5 16.2 16.7 17.9 % Hispanic 6.7 7.1 6.7 6.8 Education (%) Primary only (0–7) 17.5 14.5 14.7 17.1 Some HS (8–11) 27.6 27.4 28.6 28.6 HS graduate (12) 32.6 33.5 33.4 32.5 Some college (13+) 22.3 24.7 23.2 21.8 Total 100.0 100.0 100.0 100.0 Note: AT = assistive technology; HS = high school. Table 2. Underlying Global Difficulty by Type of Assistance and Activity. Type of Assistance Activity None AT Only PC Only PC and AT Bathing (n = 3,493)* Some difficulty 71.1 46.9 34.3 17.0 A lot of difficulty 23.4 32.4 22.5 23.8 Completely unable 5.6 20.7 43.2 59.2 Total 100.0 100.0 100.0 100.0 Transferring (n = 3,834)* Some difficulty 78.8 50.0 48.8 20.1 A lot of difficulty 19.9 37.1 24.4 29.2 Completely unable 1.3 13.0 26.8 50.7 Total 100.0 100.0 100.0 100.0 Walking (n = 7,051)* Some difficulty 66.0 39.7 39.3 15.7 A lot of difficulty 29.7 41.3 34.7 36.7 Completely unable 4.2 19.1 26.0 47.6 Total 100.0 100.0 100.0 100.0 Getting outside (n = 3,542)* Some difficulty 66.0 37.5 28.1 11.3 A lot of difficulty 27.1 37.1 30.7 30.8 Completely unable 6.9 25.4 41.3 58.0 Total 100.0 100.0 100.0 100.0 Notes: Denominators are in parentheses. AT = assistive technology; PC = personal care. *χ2 test significant at p <.05. Table 3. Residual Difficulty on Specific Dimensions by Global Level of Underlying Difficulty and Type of Assistance. Type of Assistance Activity n AT Only PC Only PC and AT All With Underlying Difficulty Bathing Very tiring* (695) 74.9 55.1 66.0 Long time* (709) 83.0 49.4 62.8 Very painful, ns (490) 70.2 62.5 67.0 Transferring Very tiring* (694) 70.1 56.0 59.3 Long time* (785) 76.2 57.0 57.3 Very painful, ns (801) 82.4 79.1 76.9 Walking Very tiring, ns (1877) 80.1 77.0 76.6 Long time, ns (1892) 83.0 78.9 78.5 Very painful, ns (1531) 82.1 79.7 83.0 Getting outside Very tiring* (760) 86.7 78.2 77.2 Long time* (766) 88.9 76.0 79.3 Very painful, ns (617) 86.9 85.7 79.7 Those With a Lot of Underlying Difficulty Bathing Very tiring* (317) 76.3 60.2 70.3 Long time* (332) 85.7 60.9 64.0 Very painful, ns (222) 70.7 71.7 68.8 Transferring Very tiring, ns (326) 75.1 68.6 66.5 Long time, ns (350) 78.7 72.1 63.7 Very painful, ns (336) 87.6 88.4 84.9 Walking Very tiring, ns (1009) 85.4 87.1 82.7 Long time, ns (1029) 87.2 86.6 85.8 Very painful, ns (870) 85.1 87.0 86.1 Getting outside Very tiring, ns (410) 90.7 87.2 84.3 Long time, ns (419) 87.5 85.3 85.1 Very painful, ns (346) 86.3 85.6 83.5 Notes: Reported information is for those who use some type of assistance. Respondents who report they are completely unable to perform an activity are not included. AT = assistive technology; PC = personal care; ns = χ2 test not significant (p >.10). *χ2 test significant at p <.05. References Agree, E. M., 1999;. The influence of personal care and assistive technology on the measurement of disability. Social Science & Medicine,. 48:427-443. Agree, E. M., Freedman, V. A., 2000;. Incorporating assistive devices into community-based long-term care: An analysis of the potential for substitution and supplementation. Journal of Aging and Health,. 12:426-450. Albrecht, G. L., Devlieger, P. J., 1999;. The disability paradox: High quality of life against all odds. Social Science & Medicine,. 48:977-988. Allen, S. M., Foster, A., Berg, K., 2001;. Receiving help at home: The interplay of human and technological assistance. Journal of Gerontology: Social Sciences,. 56B:S374-S382. Allen, S., Mor, V., 1997;. The prevalence and consequences of unmet need—Contrasts between older and younger adults with disability. Medical Care,. 35:1132-1148. Andresen, E. M., Lollar, D. J., Meyers, A. R., 2000;. Disability outcomes research: Why this supplement, on this topic, at this time? Archives of Physical Medicine and Rehabilitation,. 81:(Suppl. 2), S1-S4. Arnstein, P., 2000;. The mediation of disability by self-efficacy in different samples of chronic pain patients. Disability and Rehabilitation,. 22:794-801. Arras, J., 1995;. Bringing the hospital home: Ethical and social implications of high tech home care. Baltimore: Johns Hopkins University Press. Chen, L. K., Mann, W. C., Tomita, M. R., Burford, T. E., 1998;. An evaluation of reachers for use by older persons with disabilities. Assistive Technology,. 10:(2), 113-125. Cohen, M. E., Marino, R. J., 2000;. The tools of disability outcomes research: Functional status measures. Archives of Physical Medicine and Rehabilitation,. 81:(Suppl. 2), S21-S29. Desai, M. M., Lentzner, H. R., Weeks, J. D., 2001;. Unmet need for personal assistance with activities of daily living among older adults. The Gerontologist,. 41:82-88. Eldar, R., 2000;. A conceptual proposal for the study of the quality of rehabilitation care. Disability and Rehabilitation,. 22:(4): 163-169. Gitlin, L. N., Corcoran, M., Winter, L., Boyce, A., Hauck, W. W., 2001;. A randomized, controlled trial of a home environmental intervention: Effect on efficacy and upset in caregivers and on daily function of persons with dementia. The Gerontologist,. 41:4-14. Gitlin, L. N., Levine, R., 1992;. Prescribing adaptive devices to the elderly: Principles for treatment in the home. International Journal of Technology and Aging,. 5:(1), 107-118. Grahn, B., Ekdahl, C., Borquist, L., 2000;. Motivation as a predictor of changes in quality of life and working ability in multidisciplinary rehabilitation. Disability and Rehabilitation,. 22:(15), 639-654. Hartke, R. J., Prohaska, T. R., Furner, S., 1998;. Older adults and assistive devices: Use, multiple device use, and need. Journal of Aging and Health,. 10:99-116. Heide, A. van der, Jacobs, J. W. G., Albada-Kuipers, G. A., van, Kraaimat, F. W., Geenen, R., Bijlsma, J. W. J., 1993;. Self-report functional disability scores and the use of devices: Two distinct aspects of physical function in rheumatoid arthritis. Annals of Rheumatic Diseases,. 52:497-502. Jette, A. M., Keysor, J. J., 2002;. Uses of evidence in disability outcomes and effectiveness research. The Milbank Quarterly,. 80:325-345. Kaye, S., LaPlante, M. P., Carlson, D., Wenger, B. L., 1996;. Trends in disability rates in the United States, 1970–1994 (National Institute on Disability and Rehabilitation Research Report). Washington, DC: U. S. Department of Health and Human Services. Kohn, J. G., LeBlanc, M., Mortola, P., 1994;. Measuring quality and performance of assistive technology. Assistive Technology,. 6:(2), 120-125. Mann, W. C., Hurren, D., Charvat, B., Tomita, M., 1996;. The use of phones by elders with disabilities: Problems, interventions, costs. Assistive Technology,. 8:23-33. Mann, W. C., Hurren, D., Tomita, M., 1993;. Comparison of assistive device use and needs of home-based older persons with different impairments. American Journal of Occupational Therapy,. 47:980-987. Mann, W. C., Hurren, D., Tomita, M., 1995;. Assistive devices used by home-based elderly persons with arthritis. American Journal of Occupational Therapy,. 49:810-820. Mann, W. C., Ottenbacher, K. J., Fraas, L., Tomita, M., Granger, C. V., 1999;. Effectiveness of assistive technology and environmental interventions in maintaining independence and reducing home care costs for frail elderly: A randomized controlled trial. Archives of Family Medicine,. 8:210-217. Manton, K., Corder, L., Stallard, E., 1993;. Changes in the use of personal assistance and special equipment from 1982 to 1989: Results from the 1982 and 1989 NLTCS. The Gerontologist,. 33:168-176. National Center for Health Statistics., 1998;. 1994 National Health Interview Survey on Disability, Phase I and II [CD-ROM Series 10, No. 8A]. Hyattsville, MD: U. S. Department of Health and Human Services. National Center for Health Statistics., 1998;. 1995 National Health Interview Survey on Disability, Phase I and II [CD-ROM Series 10, No. 10A]. Hyattsville, MD: U. S. Department of Health and Human Services. Norburn, J. E. K., Bernard, S., Konrad, T., Woomert, A., DeFriese, G., Kalsbeek, W., et al 1995;. Self-care and assistance from others in coping with functional status limitations among a national sample of older adults. Journal of Gerontology: Social Sciences,. 50B:S101-S109. Pope, A. M., Tarlov, A., (Eds.) 1991;. Disability in America: Toward a national agenda for prevention. Washington, DC: National Academy Press. Russell, J. N., Hendershot, G. E., LeClere, F., Howie, L. J., Adler, M., 1997;. Trends and differential use of assistive technology devices: United States, 1994 (Advance Data No. 292). Washington, DC: National Center for Health Statistics. Sanford, J. A., Arch, M., Megrew, M. B., 1995;. An evaluation of grab bars to meet the needs of elderly people. Assistive Technology,. 7:36-47. Schoeni, R., Freedman, V. A., Wallace, R., 2001;. Persistent, consistent, widespread, and robust? Another look at recent trends in old-age disability. Journal of Gerontology: Social Sciences,. 56B:S206-S218. Soldo, B. J., Freedman, V. A., 1994;. Care of the elderly: Division of labor among the family, the market, and the state. In S. Preston & L. G. Martin (Eds.), The demography of aging. Washington, DC: National Academy Press. Steinfeld, E., Shea, S. M., 1995;. Enabling home environments: Strategies for aging in place. Paper presented at the 48th Annual Scientific Meeting of The Gerontological Society of America, Los Angeles, CA. Tamm, M., 1999;. What does a home mean and when does it cease to be a home? Home as a setting for rehabilitation and care. Disability and Rehabilitation,. 21:49-55. Tennstedt, S., McKinlay, J., Kasten, L., 1994;. Unmet need among disabled elders: A problem in access to community long term care? Social Science and Medicine,. 38:915-924. Verbrugge, L. M., Jette, A. M., 1994;. The disablement process. Social Science and Medicine,. 38:1-14. Verbrugge, L., Rennert, C., Madans, J., 1997;. The great efficacy of personal and equipment assistance in reducing disability. American Journal of Public Health,. 87:384-392. Zimmer, Z., Chappell, N., 1994;. Mobility restriction and the use of devices among seniors. Journal of Aging and Health,. 6:(2), 185-208. Zimmer, Z., Chappell, N., 1999;. Receptivity to new technology among older adults. Disability and Rehabilitation,. 21:(5/6), 222-230.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B1F3500F68FAC8AFFECC0A2A4E4F4E5326FEED7.txt b/test/dataset/in/resources/corpus/Clean_0B1F3500F68FAC8AFFECC0A2A4E4F4E5326FEED7.txt
deleted file mode 100644
index 0789cc3..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B1F3500F68FAC8AFFECC0A2A4E4F4E5326FEED7.txt
+++ /dev/null
@@ -1 +0,0 @@
-]>EXG5379S0531-5565(00)00189-310.1016/S0531-5565(00)00189-3Elsevier Science Inc.Fig. 1Net percent gain in muscle weights of young and old rats after 4weeks of remobilisation. Data are expressed as means ±SD (n=5). ∗p<0.05 for muscles of old rats versus young rats.Fig. 2Net percent change of biochemical parameters in gastrocnemius muscles of young and old rats after 4weeks of remobilisation. CPK, creatine phosphokinase; ACP, acid phosphatase; LPX, lipid peroxidation. Data are expressed as means ±SD (n=5). ∗p<0.05 for biochemical data of old versus young muscles.Fig. 3Transverse sections of gastrocnemius muscle (frozen sections stained Hematoxylin and Eosin). (a) left control (contralateral) leg of young (6months old) rat; (b) left control (contralateral) leg of old (26months old) rat; (c) right immobilised leg of old rat (4weeks of external fixation); (d) right leg of old rat (4weeks of external fixation, 4weeks of remobilisation). The external fixation causes major atrophic changes in myofibres, which only manage partial recovery on remobilisation.Fig. 4Details of myofibres of gastrocnemius muscle of old rats after 4weeks of external fixation (a, b); and after 4weeks of external fixation followed by 4weeks of remobilisation (c, d). (a) Transverse section showing myofibre degeneration. Many myofibres develop small vacuoles. (b) Ultrastructural details of myofibre showing sarcomere disruption and loss of myofilaments. (c) Myopathic damage is still clearly evident and there is only limited recovery 4weeks after removal of the external fixation. (d) In some cases lysosome-like bodies are prominent in myofibers, which in many cases remain vacuolated. (a, c, d, 1μm epon sections stained alkaline toluidine blue; b, transmission electron micrograph)Table 1The mean (±SD) body-weight of old rats before and after 4weeks of external fixation and before and after an additional 4weeks of remobilisation (n=5)External fixationExternal fixation plus remobilisationBeforeAfterBeforeAfterBody-weight (g)398±69.8373.8±62.7428.3±63.3450±57Mean difference (g)−2521.795% confidence interval (CI) of difference−127.76 to 71.76−66.15 to 109.55Change (%)−6.1%5.14%p ValueNSNSTable 2The mean (±SD) change in the weight (mg) of the hindlimb muscles in old rats after 4weeks of external fixation and 4weeks of external fixation followed by 4weeks of remobilisation (n=5)External fixationExternal fixation plus remobilisationMuscleContralateral legImmobilised legMean difference95% CI of differenceChange (%)Contralateral legImmobilised legMean difference95% CI of differenceChange (%)Gastrocnemius1556±204914±129−642−891 to −393−41.27*1510±1671073±118−437−648 to −226−28.9*Quadriceps2172±1001221±88−951−1088 to −814−43.77**2314±1261541±139−773−996 to −579−33.4*Plantaris231±11121±11−110−126 to −94−46.6**250±17144±9>−106−126 to −86−42.3*Soleus171±16103±20−6.8−94 to −42−39.5*166±15104±1−62−77 to −46−37.3**p<0.001.**p<0.01, contralateral control v immobilised legs.Table 3Mean (±SD) values for the activity of ACP, CPK and lipid peroxidation in the gastrocnemius muscle of old rats after 4weeks of external fixation followed by 4weeks of remobilisation (n=5)External fixationExternal fixation plus remobilisationMuscleContralateral legImmobilised legMean difference95% CI of difference (%)ChangeContralateral legImmobilised legMean difference95% CI of differenceChange (%)ACP activity (mU/mg)a28.6±1.9538.1±69.472.96 to 15.97+33.229.5±1.8732.6±1.763.10.45 to 5.75+10.7p valuep<0.02p<0.02CPK activity (U/mg)a4.5±0.793.73±0.78−0.77−1.91 to 0.37−17.24.17±0.413.63±0.31−0.54−1.07 to −0.01−12.6p valuep<0.01Lipid peroxidation (nmol MDA/mg protein)b4.13±16.61±0.352.481.38 to 3.57+60.04.04±1.75.26±1.71.22−1.26 to 3.70+30p valuep<0.02NSaActivity of enzymes is expressed per mg of soluble proteins in the supernatant.bMDA, malondialdehyde.Table 4Comparison between young and old rats in their capacities for recovery of hindlimb muscle weights after 4weeks of external fixation followed by 4weeks of remobilisation (n=5)External fixationExternal fixation plus remobilisationYounga % weight lossOld % weight lossYounga % weight lossOld % weight lossYoung net % gainOld net % gainYoung/Old relative % gainGastrocnemius−58−41.27−28.9−28.929.912.42.41Quadriceps−62−43.77−42.7−33.419.310.31.87Plantaris−46.5−46.6−35.7−42.310.94.32.53Soleus−46.6−39.5−41.0−37.35.62.22.54aThe data for the young animals are taken from a recent publication (Zarzhevsky et al., 1999).Table 5Comparison between young and old rats in their capacities for recovery of biochemical parameters (% change) after 4weeks of external fixation followed by 4weeks of remobilisation (n=5)External fixationExternal fixation plus remobilisationYoungaOldYoungaOldYoung net % gainOld net % gainYoung/Old relative % gainACP activity+82.4+33.2−7.3−10.789.745.91.95CPK−36.5−17.2+2.9−12.639.44.68.56Lipid peroxidation+132.5+60.0+32.4+30.0100.130.03.33aThe data for the young animals are taken from a recent publication (Zarzhevsky et al., 1999).Recovery of muscles of old rats after hindlimb immobilisation by external fixation is impaired compared with those of young ratsNZarzhevskyaECarmelibDFuchscRColemanaHSteindA.ZReznicka*reznick@tx.technion.ac.ilaDepartment of Anatomy and Cell Biology, Bruce Rappaport Faculty of Medicine, Technion-Israel Institute of Technology, P.O. Box 9649, Haifa 31096, IsraelbPhysical Therapy Program, Sackler Faculty of Medicine, Tel Aviv University, Ramat Aviv 69978, IsraelcDepartment of Orthopaedics, Bnai Zion Medical Center, Haifa 31408, IsraeldDeparment of Orthopaedics A, Rambam Medical Center, Haifa 31096, Israel*Corresponding author. Tel.: +972-4-829-5388; fax: +972-4-829-5392AbstractThe right hindlimbs of 24-month-old female Wistar rats were immobilised for 4weeks using external fixation of the knee joint. In a further group, after the external fixation was removed, the rats were allowed to remobilise for an additional 4weeks. Hindlimb immobilisation for 4weeks caused a 32–42% reduction in wet weights of the hindlimb muscles of the rats as compared to those of the contralateral non-immobilised legs. After 4weeks of remobilisation the hindlimb muscles had not returned to the “control” weights. Biochemical changes in the gastrocnemius muscle resulting from the external fixation showed greatly elevated acid phosphatase activities (33.2%) and markedly reduced creatine phosphokinase activities (17.2%), which did not recover to preimmobilisation values after 4weeks of remobilisation. Light and transmission electron microscopy showed that remobilisation for 4weeks (after external fixation) resulted in only partial morphological restoration of the damage to the muscles in these aged rats. A comparison of similar hindlimb external fixation and remobilisation in young (6months old) rats showed that remobilisation caused a substantial recovery in biochemical parameters in both age groups, with the muscles of the young group (but not the old group) often reaching almost complete recovery accompanied by morphological restoration. We conclude that the net gain in the recovery period of biochemical and morphological parameters is significantly greater in the young rats compared to the old rats indicating that muscle metabolism and capacity for recovery from disuse atrophy is impaired with ageing.KeywordsDisuse atrophySkeletal muscleExternal fixationRemobilisationAgeingRats1IntroductionAs early as 1944, the cause-and-effect relationship of limb immobilisation and disuse atrophy was recognised (Eccles, 1944; Thomsen and Luco, 1944). Inactivity, bed rest, non-weight bearing, external casts or external fixation all result in loss of muscle mass or muscular atrophy. Atrophy is associated with a net loss of muscle protein synthesis initiated within hours of immobilisation (Goldspink, 1977; Booth and Seider, 1979). The initial atrophy is very rapid, especially within the first week of immobilisation (Booth, 1977; Maxwell et al., 1992; Appell, 1990). Recently a comparison of two different modes of hindlimb immobilisation (Plaster of Paris casts and external fixation) was reported (Reznick et al., 1995). In these studies several biochemical and morphological parameters were investigated as a measure of damage to muscles of immobilised limbs. It was shown that external fixation, with its more rigid immobilisation, resulted in more drastic changes and damage to the hindlimb muscles (Reznick et al., 1995).The capacity for recovery after immobilisation has been little investigated, though it is of the utmost clinical importance as the orthopaedic surgeon needs to decide at what stage the cast or external fixation should be removed in order to achieve the best outcome of the operation. Common clinical experience reveals that muscle recovery following limb immobilisation is more rapid and complete in young individuals rather than in the elderly. Relatively little is known about the effects of muscle hypoactivity in old age (Ansved, 1995). Our working hypothesis was that muscles of old animals will be affected more severely by immobilisation and will be slower to recover compared with muscles of young animals. We have recently reported experimental studies on the capacity for recovery of hindlimb muscles of young (6month old) rats after external fixation (Zarzhevsky et al., 1999). In these studies, 4weeks of remobilisation resulted in the restoration of biochemical and morphological parameters to preimmobilisation values, whereas muscle mass was still lower than preimmobilisation values. In the present study we report similar experiments conducted in aged rats (24months old) and evaluate various biochemical and morphological parameters as indicators of the capacity for recovery from disuse atrophy in these old animals.2Materials and methods2.1Animals15 female Wistar rats (24months old) were housed in standard plastic breeding cages, in groups of three, at 20°C with a 12-h light–dark cycle. They were fed a standard diet and water ad libitum. The animals were maintained in conformity with the Guiding Principles in the Care and Use of Animals of the American Physiological Society and the experimental protocol received approval by the Animal Welfare and Ethics Committee of the Technion Faculty of Medicine.The animals were divided into three groups each of five rats. Group 1 consisted of rats in which the right knee was immobilised by external fixation in a position of 40–50° flexion for 4weeks (Reznick et al., 1995). Group 2 had similar hindlimb immobilisation for 4weeks after which the external fixation was removed and the limb remobilised for a further 4weeks. Group 3 consisted of untreated age-matched animals.2.2Immobilisation by external fixation (Reznick et al., 1995)Rigid immobilisation was achieved by the insertion of two Kirschner wires (0.8mm diameter) driven in pairs through the lateral plane of the femur and tibia. Then, two threaded brass rods fitted with nuts, connected them in order to construct a rigid frame. The brass rods were 4.8mm diameter and 33mm long. Each rod was cut longitudinally from both ends to an equal length of 13mm. These cuts were 1.0–1.2mm in width in order to contain the Kirschner wires. The overall weight of the above device was 12g.2.3Biochemical studiesAnimals were sacrificed by ether anesthesia at 4weeks for Group 1 and 8weeks for Group 2. Group 3 (untreated controls) were sacrificed at 24months of age prior to the start of the experimental hindlimb immobilisation. Muscle specimens for biochemical studies (200mg each) were taken from the belly of the gastrocnemius and quadriceps muscles of the right (immobilised) and left (contralateral) hindlimbs. (The soleus and plantaris muscles weigh less than 200mg and are insufficient for biochemical analysis). The muscles were immersed in 1.4ml of 50mM tris(hydroxymethyl) aminomethane buffer, pH 7.4. The mixture was homogenised in a Polytron homogeniser (Kinematica GmbH, Lucerne, Switzerland) three times for 15s. The homogenate was centrifuged for 30min at 14,000g. The supernatant was separated and utilised for enzymatic assays. Acid phosphatase activities were determined as previously described (Reznick et al., 1995). Creatine phosphokinase assay was performed according to the method of Tanzar and Gilvard (1959) using 0.05M glycine buffer, pH 7.4. Lipid peroxidation measurements were performed using thiobarbituric acid (TBA) according to the procedure of Ohkawa et al. (1979). Protein concentration in muscle extracts was determined according to the method of Lowry et al. (1951).2.4Light and electron microscopyAt sacrifice, the gastrocnemius muscle was rapidly removed. The belly of the gastrocnemius muscle was rapidly frozen in isobutane cooled by liquid nitrogen. 6μm cryostat transverse sections were stained with Hematoxylin and Eosin (H and E). For higher resolution sections, a 2mm-thick transverse slice of the fresh gastrocnemius muscle was cut with a razor blade from a site approximately 3mm distant from the tendon of origin and immersion fixed in 3% glutaraldehyde in 0.1M sodium cacodylate buffer, pH 7.4 containing 0.01% calcium chloride at room temperature. Blocks of tissue approximately 1mm3 were cut and left in the fixative for 3h prior to rinsing and storage overnight in 0.1M sodium cacodylate buffer containing 7.5% (wt/vol.) sucrose. This was followed by post-fixation in 1% unbuffered aqueous osmium tetroxide for 2h, dehydration in ascending ethanols, treatment with propylene oxide and epoxy embedding in Pelco Eponate 12 (Pelco International, Redding, CA, USA). After heat polymerisation (60°C for 18h), 1μm thick sections were cut on an ultramicrotome and stained with 0.1% toluidine blue in 1% borax for light microscopy. Sections (60–90nm thick) were cut with a diamond knife on the ultramicrotome, collected on uncoated copper grids, and contrast-stained with 1% uranyl acetate in 70% ethanol followed by 1% lead citrate prior to examination in a JEOL 100SX transmission electron microscope at 80kV.2.5Statistical analysisStatistical analysis was performed using either paired or independent Student's t-test. Statistical significance was set at p<0.05. Results are reported as ±SD.3Results3.1Body weightsThe body weights of animals with hindlimb immobilisation by external fixation for 4weeks and remobilisation are shown in Table 1. After 4weeks of immobilisation (Group 1), the old animals lost 6.1%, which was not significant. On remobilisation for 4weeks (Group 2) the rats had made a complete recovery of body weights and even showed a slight increase (5.14%). The differences between body weights at the start of the experiment and following external fixation and remobilisation were not significant.3.2Muscle weightsTable 2 shows the weights of the four muscles studied in the old animals after 4weeks of external fixation and after further 4weeks of remobilisation. The weights of the gastrocnemius, quadriceps, plantaris and soleus muscles were reduced by 41.2, 43.7, 46.6 and 39.5%, respectively. Following remobilisation the weights of these muscles were still much reduced (28.0, 33.4, 42.3 and 37.3%, respectively) compared with those of the contralateral legs, which was still statistically significant (p<0.001).3.3Biochemical studiesThe changes in enzymatic activities of acid phosphatase (ACP), creatine phosphokinase (CPK) and lipid peroxidation of gastrocnemius muscle in old rats after external fixation and remobilisation are shown in Table 3. Immobilisation for 4weeks caused a 33.2% increase in acid phosphatase activities, whereas there was a decrease of 17.2% in CPK activities. Lipid peroxidation increased by 60% as measured by the TBA assay. These changes were significant for acid phosphatase (p<0.02), lipid peroxidation (p<0.02) and CPK activities (p<0.01). On remobilisation for 4weeks, activities of acid phosphatase and CPK remained significantly lower than the control values indicating an incomplete recovery of enzymatic activities after 4weeks of remobilisation. Lipid peroxidation values recovered on remobilisation and were no longer significantly different from the contralateral control values.3.4Comparative age-associated studiesTable 4 presents a comparison of changes in hindlimb muscle weights of young and aged rats after external fixation and remobilisation. As a result of hindlimb immobilisation in the young rats the larger muscles, quadriceps and gastrocnemius, lose about 60% of their muscle weight, but in the aged rats these muscles lose only 41–43%. The weight loss in the smaller muscles, plantaris and soleus, was similar in both age groups. After 4weeks of remobilisation, none of the muscles of both age groups recovered to preimmobilisation weights. If values are expressed as net % gain (the difference between maximum loss in % after 4weeks of EF and the percentage weight loss after 4weeks of remobilisation) the values for the young animals are much higher than those of the aged animals. The relative % gain as calculated from the ratio young: old in the net % gain for the gastrocnemius, quadriceps, plantaris and soleus muscles was 2.41, 1.87, 2.53 and 2.54, respectively. The net % gain of both young and old animals is illustrated graphically in Fig. 1.Table 5 presents a comparison of the capacity of enzyme activities (acid phosphatase, creatine phosphokinase) and lipid peroxidation in the gastrocnemius muscle of young and old rats to recover after hindlimb immobilisation and remobilisation. After 4weeks of remobilisation in young rats the % change of acid phosphatase, CPK and lipid peroxidation was −7.3, +2.9 and +32.4%, respectively, which was not significantly different from that of the contralateral control leg. The % change for these same parameters in the old rats was −10.7, −12.6 and +30.0, respectively, which was significant for both acid phosphatase and CPK.Hence, it was possible to conclude that the net % change for young animals was much more pronounced for all the biochemical parameters studied compared with the values for the old animals. The net % change in biochemical parameters is illustrated graphically in Fig. 2. It is clear that in all categories of muscles, net % weight gain and biochemical differences, the changes of muscles of the young animals were much more pronounced than in the muscle of old animals.3.5Morphological studiesThe morphological studies were performed on gastrocnemius muscle of the aged rats. Frozen sections (Fig. 3a and b) show that the morphological appearance of myofibres is fairly similar in gastrocnemius muscles of control legs in both the young (6months old) and old (26months old) rats. Fig. 3c shows that 4weeks of external fixation in the old rats causes marked myopathic changes with the myofibres becoming markedly shrunken and distorted. On remobilisation after external fixation (Fig. 3d) there is only a partial structural recovery of the fibres. Fig. 4a shows greater detail of the myopathic changes after external fixation. Many myofibres show marked degeneration and numerous small vacuoles are seen in myofibres. At the ultrastructural level (Fig. 4b) the myopathic changes are clearly visible including sarcomere disruption and myofilament loss. After remobilisation (Fig. 4c and d), myopathic damage is still clearly evident and there is only limited recovery of myofibre structure. Many myofibres remain swollen and distorted. Sarcomeres in many myofibres remain irregular and distorted. Many myofibres still are highly vacuolated and others contain large numbers of lipid droplets. Some myofibres develop large numbers of lysosome-like bodies (Fig. 4d). Rounded cells with large regular nuclei and pronounced nucleoli (possibly activated satellite cells), as well as inflammatory cells and adipose cells are common in the connective tissue between the myofibres.4DiscussionExternal fixation is a highly effective and widely used method for bone and joint immobilisation that requires the invasive use of pins through bone and muscle. In the present study we showed that the body weights of aged rats subjected to unilateral external fixation of a hindlimb were not changed significantly as a result of the limb immobilisation or after subsequent remobilisation. We also demonstrated that none of the muscles studied from the immobilised limb were capable of returning to preimmobilisation weights or to the weights of the non-immobilised contralateral muscles after 4weeks of remobilisation. This interesting observation confirms that of Maeda et al. (1993), who wrote: “Both muscle and bone require a time period longer than the period of immobilisation in order to make a complete recovery from temporary deterioration”. Similar observations were reached by Booth (1978), who showed that gastrocnemius muscle of rats immobilised in plaster-of-Paris casts for 28days returned to preimmobilisation weight only 50days following the cast removal. However, in another study of 90 day immobilisation of rat hindlimbs, the soleus muscle weight was shown to return to control values 14days after the start of the recovery period (Booth and Seider, 1979). It should be noted that the loss of muscle weight due to immobilisation does not occur as a result of changes in muscle water content, as it was previously shown that the ratio of muscle dry weight to muscle wet weight remains the same for immobilised and control legs (Reznick et al., 1995).Alkaline phosphatase, acid phosphatase and CPK activities have been shown to change considerably after limb immobilisation (Witzmann et al., 1982). Using acid phosphatase and CPK activities as criteria for normal levels of muscle function and metabolism, we have shown that hindlimb immobilisation in old rats altered the level of activities of these enzymes in gastrocnemius muscles, however, after 4weeks of remobilisation, the levels of acid phosphatase and CPK have still not returned to pre-immobilisation levels. Lipid peroxidation was shown to increase considerably after 4weeks of limb immobilisation but on remobilisation recovered by 30% to values that were not statistically significantly different from the contralateral controls.In comparing the recovery responses of young rats to aged rats after external fixation of hindlimbs we have shown that the relative % change of muscle weights and of biochemical parameters is by far greater in the young animals. This may indicate that the capacity for active metabolism and turnover of muscle protein is indeed far greater in young animals than in the aged animals. It would appear that the older animals need much longer times to recover from the stress and disuse atrophy of limb immobilisation than the younger animals and this reflects common clinical experience on traumatic healing of muscle and bone in the elderly.These conclusions are confirmed by the morphological studies. We recently showed that the gastrocnemius muscle of young rats subjected to hindlimb immobilisation by external fixation shows an excellent structural recovery after 4weeks of remobilisation (Zarzhevsky et al., 1999). In contrast, the present study has shown that under similar experimental circumstances, the gastrocnemius muscle of old rats shows only a very limited structural recovery in this time period. We are led to conclude that skeletal muscles of hindlimbs of aged rats show very limited recovery processes after disuse atrophy compared with their younger counterparts.As can be seen from Table 4, % weight loss of gastrocnemius and quadriceps muscles of immobilised hindlimbs was higher in young animals compared to old ones. However, the % weight loss was similar for plantaris and soleus muscles for both age groups (Table 4). Indeed it appears that the larger muscles, such as the gastrocnemius and quadriceps, are undergoing a faster rate of catabolism in young animals compared to old ones. This phenomenon of reduction in the rate of protein breakdown in tissues of old animals compared to young ones has been observed in the past. A recent paper has shown using 3-methylhistidine excretion, that whole body muscle protein catabolism is slower in elderly humans compared to younger ones, by as much as 30–40% (Morais et al., 1997). Similarly, various subcellular fractions of proteins, as well as total proteins, are degraded more slowly in livers of old mice compared to livers of young mice (Lavie et al., 1982). A very recent study has shown that chaperone-mediated autophagy of lysosomes is reduced in livers of ageing rats as compared with young rats (Cuervo and Dice, 2000). Thus it is not unique that muscles of old animals in our experimental model were degraded to a lesser extent that those of the younger animals.The capacity of muscle recovery from muscle degeneration following bupivacaine injection in aging rats was studied by Marsh et al (1997a,b). The mass of tibialis anterior muscle was restored to control values within 21days in young rats (3months old), whereas in adult (18months old) and old rats (31months old) it remained 40% less than that of the contralateral controls at 21 and 28days of recovery. Furthermore, expression of myogenin and myoD mRNA levels increased in regenerated muscles of young, adult and old rats at 5 and 14days, respectively. Whereas in the young (3months old) rats mRNA levels returned to preinjection control values by 21days, in old animals they remained elevated even at 28days of recovery. It appears that either old animals have diminished capacities to down-regulate myogenin and MyoD mRNAs, or possibly their reduced ability for myogenic effect includes elevated levels of these mRNAs (Marsh et al., 1997a). Similar results were also obtained with IGF-1 mRNA, where impaired regeneration of the tibialis anterior was associated with a prolonged elevation of IGF-1 mRNA expression in the old muscle (Marsh et al., 1997b). Since the turnover of muscle proteins (synthesis and degradation) appears to slow down with age, it will be of great interest in future studies to investigate the molecular events behind the differences between young and old animals. Understanding these mechanisms in old animals may provide some clues as to the general phenomenon of sarcopenia of old age, and ways to alleviate some of the associated clinical problems.AcknowledgementsThis research was supported in part by a grant from the Jan M. and Eugenia Krol Foundation, Lakewood, NJ, USA, and by grant #181-883 from the Technion, Vice-Provost for Research. We would also like to acknowledge the excellent technical help of Mrs Pessiah Shenzer.ReferencesAnsved, 1995T.AnsvedEffects of immobilization on the rat soleus muscle in relation to ageActa Physiol. Scand.1541995291302Appell, 1990H.J.AppellMuscular atrophy following immobilizationSports Med.1019904252Booth, 1977F.W.BoothTime course of muscular atrophy during immobilization of hindlimbs in ratsJ. Appl. Physiol.431977656661Booth, 1978F.W.BoothRegrowth of atrophied skeletal muscle in adult rats after ending immobilizationJ. Appl. Physiol.441978225230Booth and Seider, 1979F.W.BoothM.J.SeiderRecovery of skeletal muscle after 3months of hindlimb immobilization in ratsJ. Appl. Physiol.471979435439Cuervo and Dice, 2000A.M.CuervoJ.F.DiceAge-related decline in chaperone-mediated autophagyJ. Biol. Chem.2000(in press)Eccles, 1944J.C.EcclesInvestigations on muscle atrophies arising from disuse and tenotomyJ. Physiol. (Lond)1031944253256Goldspink, 1977D.F.GoldspinkThe influence of immobilization and stretch on protein turnover of rat skeletal muscleJ. Physiol. (Lond)2641977267282Lavie et al., 1982L.LavieA.Z.ReznickD.GershonDecreased protein and puromycinyl-peptide degradation in livers of senescent miceBiochem. J.20219824751Lowry et al., 1951O.H.LowryN.J.RosebroughA.L.FarrR.J.RandallProtein measurement with the Folin phenol reagentJ. Biol. Chem.1931951265275Maeda et al., 1993H.MaedaD.B.KimmelD.M.RaabN.E.LaneMusculoskeletal recovery following hindlimb immobilization in adult female ratsBone141993153159Marsh et al., 1997aD.R.MarshD.S.CriswellJ.A.CarsonF.W.BoothMyogenic regulatory factors during regeneration of skeletal muscle in young, adult, and old ratsJ. Appl. Physiol.83199712701275Marsh et al., 1997bD.R.MarshD.S.CriswellM.T.HamiltonF.W.BoothAssociation of insulin-like growth factor mRNA expressions with muscle regeneration in young, adult, and old ratsAm. J. Physiol.2731997R353R358Maxwell et al., 1992L.C.MaxwellM.R.MoodyC.S.EnwemekaMuscle atrophy continues after early cast removal following tendon repairAnat. Rec.2331992376386Morais et al., 1997J.A.MoraisR.GougeonP.B.PencharzP.J.H.JonesR.RossE.B.MarlissAm. J. Clin. Nutr.661997880889Ohkawa et al., 1979H.OhkawaN.OhishiK.YagiAssay for lipid peroxides in animal tissues by thiobarbituric acid reactionAnal. Biochem.951979351358Reznick et al., 1995A.Z.ReznickG.VolpinH.Ben-AriM.SilbermannH.SteinBiochemical and morphological studies on rat skeletal muscles following prolonged immobilization of the knee joint by external fixation and plaster cast: a comparative studyEur. J. Exp. Musculosketal Res.419956976Tanzar and Gilvard, 1959M.L.TanzarC.GilvardCreatine kinase measurementJ. Biol. Chem.234195932013204Thomsen and Luco, 1944P.ThomsenJ.V.LucoChanges of weight and neuromuscular transmission in muscles of immobilized jointJ. Neurophysiol.71944241245Witzmann et al., 1982F.A.WitzmannA.J.TroupR.H.FittsAcid phosphatase and protease activities in immobilized rat skeletal musclesCan. J. Physiol. Pharmacol.60198217321736Zarzhevsky et al., 1999N.ZarzhevskyR.ColemanG.VolpinD.FuchsH.SteinA.Z.ReznickMuscle recovery after immobilisation by external fixationJ. Bone Jt. Surg. Br.81-B1999896901
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B2A2F56C3B4690B89D311653C95EAD549C2829B.txt b/test/dataset/in/resources/corpus/Clean_0B2A2F56C3B4690B89D311653C95EAD549C2829B.txt
deleted file mode 100644
index 0bbea45..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B2A2F56C3B4690B89D311653C95EAD549C2829B.txt
+++ /dev/null
@@ -1 +0,0 @@
-]>MAD2077S0047-6374(98)00042-610.1016/S0047-6374(98)00042-6Elsevier Science Ireland LtdFig. 1The inhibition kinetics of DNP-SG transport elicited by various inhibitors and effects of increasing concentrations of inhibitors on CumOOH-induced MDA formation in erythrocytes. Suspensions of erythrocytes from aged (n=5) or young (n=5) subjects were treated with increasing concentrations of Tween 80 (A), fluoride (B) or vanadate (C) as described in Section 2. After centrifugation the supernatants were assayed for DNP-SG transport. MDA levels of erythrocytes treated with 0.1 mmol/l CumOOH after pretreatment with inhibitors were also determined. Percent inhibition was expressed as the percentage decrease of DNP-SG efflux from erythrocytes. The values are means of duplicate analyses from 10 experiments.Fig. 2Effects of various inhibitors of the glutathione conjugate pump on CumOOH-induced lipid peroxidation in erythrocytes from healthy aging (n=18) and young (n=18) adults. DNP-SG inhibitors, Tween 80 (1/1200), fluoride (30 mmol/l) or o-vanadate (0.2 mmol/l) were added before the erythrocyte suspensions were incubated with 0.1 mmol/l CumOOH for 15 min at 37°C. MDA values shown are means with standard deviations.Fig. 3(See left) Effects of the addition of CumOOH to the reaction medium following preincubation with various inhibitors of glutathione conjugate pump on the changes in chemiluminescence. (A) Chemiluminescence response of erythrocytes treated with 0.1 mmol/l CumOOH following the inhibition of DNP-SG transport of erythrocytes from aging group. (B) Comparison of the chemiluminescence response in erythrocytes from aging and young adults. DNP-SG transport inhibitors, Tween 80, fluoride or o-vanadate were added to erythrocyte suspensions before mixing with 0.1 mmol/l CumOOH and chemiluminescence was recorded at indicated time. Data represent the mean of 16 determinations in the elderly group and 15 determinations in the young group.Table 1Transport of 2,4-dinitrophenyl-S-glutathione in erythrocytes from aging and young subjectsTime (h)Erythrocyte efflux (nmol DNP-SG/ml erythrocytes)Aging group (n=18)Young group (n=18)1410±77390±422685±117674±713971±128933±11441174±1651118±157All data expressed as mean±S.D.Transport of glutathione conjugate in erythrocytes from aged subjects and susceptibility to oxidative stress following inhibition of the glutathione S-conjugate pumpİlhanOnarana*AhmetÖzaydinaMustafaGültepebGönülSultuybekaaDivision of Biomedical Sciences, Cerrahpaşa Medical Faculty, Istanbul University, Istanbul, TurkeybGATA, Military Educational Hospital, Department of Biochemistry, Istanbul, Turkey*Corresponding author. Present address: Ortaklar Cd. Bütan Sk., Hyzal Apt. No:2 D:3, Mecidiyeköy 80290, İstanbul, Turkey. Fax: +90 212 5299433.AbstractThe aim of the present study was to investigate the effect of donor aging on the glutathione conjugate transport in erythrocytes and whether it plays a role in the resistance to oxidative stress of the erythrocytes of aging subjects. In our comparative study on intact erythrocytes of healthy aging and young adults, in which 2,4-dinitrophenyl-S-glutathione (DNP-SG) was used as model glutathione S-conjugate, we found that the efflux of DNP-SG remained unchanged in the aged subjects. This result suggests that the detoxification function is maintained against the chemical stress employed in erythrocytes of aging subjects. In the assay conditions used, which were optimized to obtain maximal inhibition of glutathione S-conjugate transport, our results also indicated that the susceptibility of erythrocytes to in vitro lipid peroxidation generated by cumene hydroperoxide was enhanced by pretreatment with DNP-SG inhibitors in both age groups. However, the difference in susceptibility was not a function of aging. Further, the results suggested that inhibition of glutathione S-conjugate pump may impair cellular protection of the erythrocytes against oxidative damage.KeywordsAgingErythrocyteGlutathione S-conjugateGlutathioneTransportOxidative stress1IntroductionTransport of glutathione S-conjugates is an important element of xenobiotic detoxification. Glutathione is enzymatically conjugated to many electrophilic compounds in a reaction catalyzed by glutathione S-transferases (GST). The glutathione S-conjugates formed in these reactions are afterwards exported out of cells by a specific adenosine 5′-triphosphate (ATP)-dependent glutathione S-conjugate pump (for review see Zimniak and Awasthi, 1993). The active transport systems for glutathione S-conjugates have been described for a number of human tissues and this system has also been identified in human erythrocyte membranes (Awasthi et al. 1983, Ishikawa and Sies 1984, La Belle et al. 1986, Kunst et al. 1989). The glutathione S-conjugate pump also plays a significant role in cellular defence under oxidative stress conditions since it is able to transport glutathione conjugates of lipid peroxidation products and extrudes oxidized glutathione (GSSG) (Ishikawa 1989, Grune et al. 1991, Akerboom et al. 1992). Erythrocytes are a good model system for the explanation of such processes and the molecular nature of the defect in the transport of glutathione conjugates since their lack of γ-glutamyltranspeptidase precludes further metabolizing of formed conjugates. Although the transport system exhibited a broad substrate specificity towards different types of glutathione S-conjugate (Ishikawa, 1989), the compound most widely used in such studies is 2,4-dinitrophenyl-S-glutathione (DNP-SG) formed from 1-chloro-2,4-dinitrobenzene (CDNB) in a reaction catalyzed by GST.Different detoxification mechanisms such as glutathione, GST, glutathione reductase, glutathione peroxidase, superoxide dismutase and catalase but not glutathione S-conjugate transport have been tested for aging process (Stohs et al. 1984, Jozwiak and Jasnowska 1985, Stohs and Lawson 1986, Al-Turk et al. 1987, Farooqui et al. 1987, Foldes et al. 1988, Matsubara and Machado 1991, Picot et al. 1992). Results from a number of studies have suggested that a glutathione-deficiency state is a general phenomenon in aging cells, since aging-specific decrease in glutathione levels was observed in all tissues including erythrocytes (Stohs et al. 1984, Stohs and Lawson 1986, Al-Turk et al. 1987, Farooqui et al. 1987, Matsubara and Machado 1991). Furthermore, different results showing decreased or unchanged GST activity in elderly subjects when compared with young subjects were also reported (Stohs et al. 1984, Al-Turk et al. 1987, Picot et al. 1992). On the other hand, results of our previous study indicated that the susceptibility of intact erythrocytes to in vitro oxidative stress by cumene hydroperoxide (CumOOH) in aging subjects is not much greater than young subjects (Onaran et al., 1997). The above observations prompted us to study how the transport of glutathione conjugate in erythrocytes, generated through the reactions of glutathione and GST, is affected by the aging process. We also wished to examine whether glutathione S-conjugate transport has any effect on the resistance to oxidative damage of the intact erythrocytes of aging subjects.2Materials and methods2.1SubjectsSubjects were young (aged 18–31) or elderly (aged 67–85) normal volunteers. All subjects were non-smokers and had normal blood counts, normal blood levels of urea, glucose, creatinine, albumin, alkaline phosphatase, lactate dehydrogenase and bilirubin, and no history of hematologic abnormality, recent infectious disease or significant medical illness. The participants were instructed not to take aspirin-like drugs for 2 weeks prior to blood sampling.2.2ReagentsAll reagents were of analytical quality whenever possible, obtained mainly from Sigma (St. Louis, MO) and Merck (Darmstadt, Germany).2.3Preparation of erythrocytesFresh heparinized fasting blood sample was centrifuged at 1500×g for 10 min. Erythrocytes were separated from plasma and buffy coat and then washed three times with 10 vol of phosphate buffered saline (PBS, pH 7.4). Erythrocytes containing chronologically young population rich with reticulocytes were removed by discontinuous Percoll density gradient essentially as described by Rennie et al. (1979). The gradient was built up in two layers containing Percoll with specific density values between 1.100 and 1.124 g/ml. At the end of centrifugation while the cells were collected at the interface of the Percoll above 1.124 g/ml, enriched fraction of reticulocytes above low density were removed. The erythrocytes were then washed twice with PBS.Hemoglobin (Hb) concentrations in the samples were determined with Drabkin's reagent as described by Beutler (1984). Reticulocyte count in the fraction was performed on glass slides after staining with 0.1% brilliant cresyl blue saline. The number of reticulocytes in the high density cell fractions studied did not show significant differences between elderly and young control groups.2.4Reaction systemsFor inhibition of glutathione conjugate transport, erythrocytes were incubated at hematocrit 10% in PBS (with 8 mmol/l glucose) containing one of the following reagents (15 min at 37°C with continuous shaking at 120 cycles/min): 5–30 mmol/l sodium fluoride; 0.025–0.2 mmol/l sodium o-vanadate; diluted 1/4000–1/1200 Tween 80 (polyoxyethylenesorbitan-monooleate). Controls were incubated with PBS alone. To check whether the glutathione conjugate transport was inhibited in erythrocyte suspensions the export of DNP-SG was measured. Hemolysis was also checked in samples and it was always below 1–2%. After preincubation with inhibitors, lipid peroxidation was induced by addition of freshly prepared CumOOH (final concentration, 0.1 mmol/l). Immediately after CumOOH addition, the tubes were sealed and incubated at 37°C with continuous shaking at 120 cycles/min. In parallel, control samples were incubated under the same conditions but without CumOOH. After 15 min, cell suspensions were centrifuged at 7000×g for 1 min to sediment the erythrocytes. Both pellet and supernatant were used for malondialdehyde (MDA) determination.2.5Measurement of the transport of glutathione S-conjugateTransport of DNP-SG was measured according to the procedure of Board (1981). Washed erythrocytes were resuspended in PBS and incubated with 1 mmol/l CDNB for 15 min at 37°C to form DNP-SG. Then the cells were washed of excess CDNB at 4°C and suspended at a hematocrit of 20% in the PBS (pH 7.4) containing 1 mmol/l MgCl2 and 8 mmol/l glucose. The cell suspensions were then incubated at 37°C and the export of DNP-SG was quantified by withdrawal of aliquots of the cell suspensions after 1, 2, 3, and 4 h. After centrifugation, DNP-SG was detected in the supernatant by spectrophotometric determination at 340 nm using an mmol/l extinction coefficient of 9.6.2.6Malondialdehyde (MDA) measurementThe concentration of MDA in the samples was measured with fluorometric determination improved by synchronous fluorescence (Conti et al. 1991, Onaran et al. 1997). For erythrocytes, 0.2 ml of packed cells were suspended in 0.8 ml PBS, 0.025 ml of 4% (w/v) butylated hydroxytoluene and 0.050 ml of 0.025 mmol/l disodium-ethylenediaminetetraacetic acid, and 0.5 ml of 30% trichloroacetic acid was then added. Tubes were vortexed and allowed to stand in ice for 2 h. After centrifugation, assay conditions were the same in both the supernatant from this step and the supernatant fraction from the erythrocyte suspension (Onaran et al., 1997). The level of MDA was expressed as total mmol MDA/g Hb.2.7Chemiluminescence assayChemiluminescence measurements were carried out in a TriCarb 1500 liquid scintillation analyzer in the single-photon count mode (Videla et al., 1984). Briefly, erythrocytes were resuspended in 3 ml of PBS containing 8 mmol/l glucose and preincubated with indicated concentrations of fluoride, vanadate or Tween 80 for 15 min at 37°C. The erythrocytes were transferred to scintillation vials. After the addition of CumOOH (final concentration, 0.1 mmol/l) the chemiluminescence measurements were taken every 10 min for 60 min. The zero time is the time of CumOOH addition.2.8Glutathione content and glutathione S-transferase (GST) activityThe concentration of glutathione was determined by the method of Tietze (1969). Total GST activity with CDNB as substrate was measured according to Habig et al. (1974).2.9Statistical analysisAll tests were performed in duplicate or triplicate samples, and the results were expressed as means±S.D. Statistical difference was determined by Student's t-test, with significance defined as P<0.05.3ResultsThe aging group (n=18) and young controls (n=18) used in this study showed no difference in their erythrocyte GST activity, 2.1±0.8 and 2.1±0.6 mmol of CDNB conjugated/min per g Hb, respectively (P>0.05). However, we found statistically significant lower glutathione levels in older group. The mean±S.D. values of glutathione were 5.24±0.69 and 6.59±0.71 mmol/g Hb for the elderly and controls, respectively (P<0.05).On incubation with CDNB at 37°C, glutathione is completely depleted in erythrocytes of the two age groups. Under these experimental conditions, the rate of DNP-SG efflux of the erythrocytes from aging subjects during 4 h did not differ from those of the young controls (P>0.05). The rate of transport of DNP-SG in both age groups was linear for a period of at least 4 h (Table 1).In order to find out whether glutathione S-conjugate transport has any effect on the resistance to oxidative stress of the erythrocytes of aging subjects, transport of erythrocytes was inhibited by incubation with various inhibitors, including sodium fluoride, vanadate or Tween 80. The transport process was, more markedly, inhibited as a function of the concentration of these inhibitors. Despite the differences in inhibition profiles of the transport rate of DNP-SG elicited by these agents, the inhibition kinetics of erythrocytes from aged subjects were not different from that of young control subjects. When the erythrocytes were exposed to 0.1 mmol/l CumOOH after preincubation with inhibitors, increasing relative concentration of inhibitors resulted in an enhanced MDA formation. The production of MDA elicited by CumOOH corresponded generally to the degree of inhibition in DNP-SG transport caused by different inhibitors. As shown in Fig. 1, significant differences in MDA formation as a function of the concentrations of inhibitors were not observed in erythrocytes of aged subjects in comparison to that of young subjects. Therefore, the concentrations corresponding to 30 mmol/l fluoride, 0.2 mmol/l vanadate and 1/1200 Tween 80 which were effectively able to inhibit the DNP-SG transport were selected for further experiments. In these experiments, MDA levels of erythrocytes treated with inhibitors of DNP-SG transport were not significantly different from the levels of erythrocytes treated with CumOOH alone (Fig. 2). MDA levels also remained unchanged within each group when compared to basal conditions (data not shown). However, MDA levels of erythrocytes treated with CumOOH after preincubation with inhibitors were significantly different from those treated with CumOOH without the inhibitors (P<0.05) (about 1.6–2-fold increase over controls with CumOOH alone). Also in this case, no differences in MDA overproduction were observed between the aging and the young groups, i.e., the ratios between values of the samples treated with CumOOH alone and values of the samples treated with CumOOH after pretreatment with inhibitors were statistically unchanged (data not shown). The MDA production in erythrocytes which were kept under oxidative stress was very similar for Tween 80 and vanadate. Inhibition by fluoride caused slightly less MDA production than other inhibitors (Fig. 2).Chemiluminescence intensity of erythrocytes treated with inhibitors of DNP-SG transport were not significantly different from the intensity of erythrocytes treated with CumOOH alone or from those in their basal conditions (Fig. 3A); during 60 min, light output in erythrocytes remained constant. Addition of 0.1 mmol/l CumOOH to erythrocytes pretreated with inhibitors enhances chemiluminescence signal in erythrocytes of both age groups. However, no differences in spectral distributions of chemiluminescence were observed between the aging and the young groups (Fig. 3B), i.e., the ratios of increase in chemiluminescence between treated and control (erythrocytes treated with CumOOH alone) values were statistically unchanged (data not shown). In addition, in the erythrocytes treated with Tween 80 or vanadate, onset and maximal chemiluminescence were slightly higher than those of fluoride.4DiscussionIn the present study, we employed human erythrocytes to investigate the effect of aging on the glutathione S-conjugate transport by determining the DNP-SG transport. We used erythrocytes other than the chronologically young population rich with reticulocytes, the reasons for which were explained in our previous article (Onaran et al., 1997). In experimental conditions, in which erythrocyte glutathione is completely depleted by incubating the erythrocytes with CDNB, we did not observed a significant difference in the efflux of DNP-SG in erythrocytes from aged subjects when compared with young subjects. The result indicates that the detoxification function is maintained against the chemical stress employed in their erthyrocytes.It is known that the cells transport the glutathione S-conjugates outward through ATP-dependent efflux process. The transport of glutathione S-conjugate in erythrocytes was first reported by Board (1981)and it was shown that the addition of CDNB to erythrocytes results in the efflux of DNP-SG with rapid irreversible depletion of glutathione. The transport of this conjugate is inhibited by the depletion of intracellular ATP or by known organic anion transport inhibitors such as vanadate, an inhibitor of P type ATPases (La Belle et al., 1986), and fluoride (interaction with the transporter molecule versus cellular energy depletion) (Awasthi et al., 1983). We also observed pronounced inhibition of DNP-SG transport by these inhibitors.Although the effects of aging on different detoxification mechanisms such as glutathione and GST in erythrocytes have been reported (Stohs et al. 1984, Jozwiak and Jasnowska 1985, Stohs and Lawson 1986, Al-Turk et al. 1987, Foldes et al. 1988, Matsubara and Machado 1991, Picot et al. 1992), to our knowledge no data on the effect of donor age on transport of glutathione conjugates of membrane is available. It has been shown that there is an age-related decrease in glutathione content of the erythrocytes (Stohs et al. 1984, Al-Turk et al. 1987, Matsubara and Machado 1991). However, some evidence pointed out that glutathione status, physical health and longevity of life were closely interrelated (Foldes et al. 1988, Calvin et al. 1992). Furthermore contradictory results showing decreased or unchanged GST activity in erythrocytes with donor aging were also reported (Stohs et al. 1984, Al-Turk et al. 1987, Picot et al. 1992). In addition, it has been shown that changes in glutathione concentration in mammalian cells have various effects, influencing cellular radiosensitivity and the cytotoxicity of several chemotherapeutic agents (Bump et al. 1982, Li and Kaminskas 1984, Fernandes and Cotter 1994). In the aging group that provided erythrocytes for our study, glutathione levels were lower for about 20% and GST activities remained unchanged. The present results thus reveal that cellular glutathione status of the aging subject has no adverse effect on glutathione S-conjugate transport and the present amount of glutathione in their erythrocytes might be sufficient to detoxify the employed chemical stress. However, we do not know how the transport of glutathione S-conjugate is affected by different levels of intracellular glutathione.It has been previously reported that transport systems for glutathione S-conjugates play significant roles in the biotransformation of toxic products arising from oxidative stress induced lipid peroxidation (Ishikawa et al. 1986, Grune et al. 1991, Akerboom et al. 1992). On the other hand, results of our previous study indicated that the intact erythrocytes from aged subjects were not more susceptible to oxidative stress than those of young subjects (Onaran et al., 1997). In the second part of our study we therefore decided to determine how glutathione S-conjugate transport plays a role in the resistance to oxidative stress of the erythrocytes of the aged subjects.To investigate this, DNP-SG transport was inhibited by various agents, i.e., fluoride, vanadate or Tween 80. These agents were used for inhibitory action in the studied system as their diverse effects on glutathione conjugate pump have been explained previously (Akerboom et al. 1992, Pulaski and Bartosz 1995). In preliminary studies we observed that under experimental conditions employed the concentrations of the inhibitors did not affect the glutathione and MDA contents, chemiluminescence formation, rate of hemolysis and GST activities in the individual's erythrocytes. The erythrocytes were then treated with 0.1 mmol/l CumOOH in order to induce oxidative stress and then MDA levels, a common end product of oxidative damage, were measured to asses whether the inhibition of glutathione transport is subjected to increased oxidative damage in erythrocytes. Analyses of the effects of inhibitors on DNP-SG transport as a function of the inhibitor concentrations under the in vitro standard oxidative stress conditions, indicate that increasing the relative concentrations of inhibitors to erythrocytes resulted in an enhanced MDA formation and that enhancement of sensitivity caused by each of the concentrations used was unchanged as a function of aging. In addition, the degree of the DNP-SG transport inhibition for each of inhibitors used was very similar in young and aged subjects. Because of these results it was decided that further experiments could be performed in the inhibition conditions which were able effectively to prevent the conjugate transport, in order to obtain an insight into the effects of glutathione S-conjugate transport on the resistance to oxidative stress. At this stage of the study the extent of oxidative damage was evaluated by measuring the chemiluminescence formation as well as MDA formation. The present study reports that following stimulation by CumOOH after the inhibition of DNP-SG, peroxidation- dependent changes of MDA and chemiluminescence formation took place at the same level in aging and young adults. This indicates that the erythrocytes from the elderly are equally capable of withstanding the oxidative stress following the inhibition of glutathione conjugate pump and oxidative effect of lipid peroxidation is not much greater in aging individuals with diminished glutathione levels but normal GST activity than young adults. Furthermore, these results also indicate that the erythrocytes with inhibited DNP-SG transport were shown to be more susceptible to oxidative stress generated by CumOOH than those which were not inhibited. We concluded, therefore, that this result indicates that inhibition of glutathione conjugate transport may impair cellular protection to oxidant stress. Although we are at present not able to explain the mechanisms governing these changes, several possible explanations of the effects observed may be considered. As inhibitory effects of glutathione S-conjugates on GSTs as well as glutathione reductase (Ishikawa et al. 1986, Ishikawa 1989) and the transport of GSSG in various tissues including erythrocytes (Bilzer et al. 1984, Akerboom et al. 1991) have been reported (Akerboom et al. 1982, 1991, Bilzer et al. 1984, Ishikawa et al. 1986, Ishikawa 1989), accumulation of the S-conjugates resulting from radical induced lipid peroxidation in the cell may be crucial for detectable oxidative damage. It is known that GST and glutathione reductase reduce organic hydroperoxides such as CumOOH (Prohaska and Ganther 1977, Awasthi et al. 1980), and it was also shown that the transport of GSSG is an important process for the cell to avoid highly oxidative stress (Adams et al., 1983). Even though such an effect is not known, inhibition of glutathione conjugate transport may also directly compromise the oxidant suppression of the cells by inhibiting key enzymes. However, further investigation is obviously necessary to understand the function of glutathione S-conjugate transport responsible for the protection from oxidative damage.In conclusion, this study indicates that the transport of DNP-SG in erythrocytes of healthy elderly with low glutathione is unaffected and their erythrocytes are not more susceptible to oxidative stress than those of young subjects after in vitro pretreatment with inhibitors which inhibit glutathione conjugate pump. Our findings also suggest that the ability of the erythrocytes to withstand oxidative damage may be diminished with inhibition of glutathione conjugate transport.ReferencesAdams et al., 1983J.DAdamsB.HLauterburgJ.RMitchellPlasma glutathione disulfide as an index of oxidant stress in vivo: effects of carbon tetrachloride, dimethylnitrosamine, nitrofurantoin, metronidazole, doxorubicin and diquatPharmacol. Exp. Ther.2271983749754Akerboom et al., 1982T.P.MAkerboomMBilzerHSiesCompetition between of glutathione disulfide (GSSG) and glutathione S-conjugates from perfused rat liver into bileFEBS Lett.14019827376Akerboom et al., 1991T.P.MAkerboomVNarayanaswamiMKunstHSiesATP-dependent S-(2,4-dinitrophenyl) glutathione transport in canalicular plasma membrane vesicles from rat liverJ. Biol. Chem.26619911314713152Akerboom et al., 1992T.P.MAkerboomGBartoszHSiesLow and high-Km transport of dinitrophenyl glutathione in inside out vesicles from human erythrocytesBiochim. Biophys. Acta11031992115119Al-Turk et al., 1987W.AAl-TurkS.JStohsF.HRashidySOthmanChanges in glutathione and its metabolizing enzymes in human erythrocytes and lymphocytes with ageJ. Pharm. Pharmacol.3919871316Awasthi et al., 1980Y.CAwasthiD.DDaoR.PSanetoInterrelationship between anionic and cationic forms of glutathione S-transferases of human liverBiochem. J.1911980110Awasthi et al., 1983Y.CAwasthiGMisraD.KRassinandS.KSrivastavaDetoxification of xenobiotics of glutathione S-transferases in erythrocytes: the transport of the conjugate of glutathione and 1-choloro-2,4-dinitrobenzeneBr. J. Haematol.551983419425Beutler, 1984Beutler, E., 1984. Red cell metabolism. In: A Manual of Biochemical Methods, 3rd ed. Grune and Stratton, New York, pp. 8–19.Bilzer et al., 1984MBilzerR.LKrauth-SiegelR.HSchirmerT.P.MAkerboomHSiesG.ESchulzInteraction of a glutathione S-conjugate with glutathione reductaseEur. J. Biochem.1381984373378Board, 1981P.GBoardTransport of glutathione S-conjugates from human erythrocytesFEBS Lett.1241981163165Bump et al., 1982E.ABumpN.YYuJ.MBrownRadiosensitization of hypoxic tumor cells by depletion of intracellular glutathioneScience2171982544545Calvin et al., 1992A.LCalvinSNaryshkinD.LSchneiderB.JMillsR.DLindemanLow blood glutathione levels in healthy aging adultJ. Lab. Clin. Med.1201992720725Conti et al., 1991MContiP.CMorandPLevillainALemonnierImproved fluorometric determination of malondialdehydeClin. Chem.37199112731275Farooqui et al., 1987M.Y.HFarooquiW.WDayD.MZamoranoGlutathione and lipid peroxidation in the aging ratComp. Biochem. Physiol.88B1987177180Fernandes and Cotter, 1994R.SFernandesT.GCotterApoptosis or necrosis: intracellular levels of glutathione influence mode of cell deathBiochem. Pharmacol.481994675681Foldes et al., 1988JFoldesT.MAllweisJMenczelOShalevErythrocyte hexose monophosphate shunt is intact in healthy aged humansClin. Physiol. Biochem.619886467Grune et al., 1991TGruneWSiemsJKowalewskiHZollnerHEsterbauerIdentification of metabolic pathways of the lipid peroxidation product 4-hydroxynonenal by enterocytes of rat small intestineBiochem. Int.251991963971Habig et al., 1974W.HHabigM.JPabsW.BJakobGlutathione S-transferasesJ. Biol. Chem.249197471307139Ishikawa, 1989TIshikawaATP/Mg2+-dependent cardiac transport system for glutathione S-conjugatesJ. Biol. Chem.26419891734317348Ishikawa and Sies, 1984TIshikawaHSiesCardiac transport of glutathione disulfide and S-conjugateJ. Biol. Chem.259198438383843Ishikawa et al., 1986TIshikawaHEsterbauerHSiesRole of cardiac glutathione transferase and of the glutathione S-conjugate export system in biotransformation of 4-hydroxynonenal in the heartJ. Biol. Chem.261198615761581Jozwiak and Jasnowska, 1985ZJozwiakBJasnowskaChanges in oxygen-metabolizing enzymes and lipid peroxidation in human erythrocytes as a function of age of donorMech. Ageing Dev.3219857783Kunst et al., 1989MKunstHSiesT.P.MAkerboomS-(4-Azidophenacyl) [35S] glutathione photoaffinity labeling of rat liver plasma membrane-associated proteinsBiochim. Biophys. Acta98219891523La Belle et al., 1986E.FLa BelleS.VSinghS.KSrivastavaY.CAwasthiEvidence for different transport systems for oxidized glutathione and S-dinitrophenyl glutathione in human erythrocytesBiochim. Biophys. Acta1391986538544Li and Kaminskas, 1984JLiEKaminskasAccumulation of DNA strand breaks and methotrexate cytotoxicityProc. Natl. Acad. Sci. USA81198456945698Matsubara and Machado, 1991L.SMatsubaraP.E.AMachadoAge-related of glutathione content, glutathione reductase and glutathione peroxidase activity of human erythrocytesBraz. J. Med. Biol. Res.241991449454Onaran et al., 1997YOnaranA.SYalçynGSultuybekEffect of donor age on the susceptibility of erythrocytes and erythrocytes membranes to cumene hydroperoxide-induced oxidative stressMech. Ageing Dev.981997127138Picot et al., 1992I.CPicotJ.MTrivierANicoleP.MSinetMTheveninAge-correlated modifications of copper-zinc superoxide dismutase and glutathione-related enzyme activities in human erythrocytesClin. Chem.3819926670Prohaska and Ganther, 1977J.RProhaskaH.EGantherThe glutathione peroxidase activity of glutathione S-transferasesBiochem. Biophys. Res. Commun.761977437445Pulaski and Bartosz, 1995LPulaskiGBartoszTransport of bimane-S-glutathione in human erythrocytesBiochim. Biophys. Acta12681995279284Rennie et al., 1979C.MRennieSThompsonA.CParkerAMaddyHuman erythrocyte fractionation in `Percoll' density gradientsClin. Chim. Acta981979119125Stohs and Lawson, 1986Stohs, S.J., Lawson, T., 1986. The role of glutathione and its metabolism in aging. In: Kitani, K. (Ed.), Liver and Aging. Elsevier, Amsterdam, pp. 59–70.Stohs et al., 1984S.JStohsF.HEl-RashidyTLawsonR.HKobayashiB.GWulfJ.FPotterChanges in glutathione and glutathione metabolizing enzymes in human erythrocytes and lymphocytes as a function of age of donorAge7198437Tietze, 1969FTietzeEnzymic method for quantitative determination of nanogram amounts of total and oxidized glutathioneAnal. Biochem.271969502522Videla et al., 1984L.AVidelaM.IVillenaGDonosoJ.DFuenteELissiVisible chemiluminescence induced by t-butyl hydroperoxide in red blood cell suspensionsBiochem. Int.81984821830Zimniak and Awasthi, 1993PZimniakY.CAwasthiATP-dependent transport systems for organic anionsHepatology171993330339
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B2D0D17A02E9D3F4810ED4B2C7748AAF937E04E.txt b/test/dataset/in/resources/corpus/Clean_0B2D0D17A02E9D3F4810ED4B2C7748AAF937E04E.txt
deleted file mode 100644
index 73a4a2f..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B2D0D17A02E9D3F4810ED4B2C7748AAF937E04E.txt
+++ /dev/null
@@ -1 +0,0 @@
-fampractfamprjFamily Practice1460-22290263-2136Oxford University Press10.1093/fampra/cmm012Primary Care EpidemiologyPatterns of pain and consulting behaviour in patients with musculoskeletal disorders in rural Crete, GreeceAntonopoulouMariaabAntonakisNacHadjipavlouAdLionisCaaClinic of Social and Family Medicine, Department of Social Medicine, School of Medicine, University of Crete, Crete, GreecebSpili Health CentrecAnogeia Health Centre, Regional Health and Welfare System of Crete, Crete, GreecedDepartment of Orthopaedics and Traumatology, School of Medicine, University of Crete, Crete, GreeceCorrespondence to: Maria Antonopoulou, Clinic of Social and Family Medicine, University of Crete, PO Box 2208, Heraklion 71003, Crete, Greece; Email: antonopm@uoc.grAntonopoulou M, Antonakis N, Hadjipavlou A and Lionis C. Patterns of pain and consulting behaviour in patients with musculoskeletal disorders in rural Crete, Greece. Family Practice 2007; 24: 209–216.6200715520072432092163620063220071832007© The Author 2007. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org.2007Background. Musculoskeletal disorders (MSDs) account as a reason for frequent consultations in primary care. However, the magnitude of the problem at the GP's clinic, the patterns of pain and the consulting behaviour has not been sufficiently explored.Objectives. The aim of this study was to report on patterns of pain relevant to MSDs and explore the co-morbidities and consulting behaviour in rural primary care settings in Crete.Methods. Three primary care centres (PCCs) of Crete were selected for a study period of 2 weeks. Every visitor, aged 20–75 years, regardless of the reason for visiting the facility was invited to participate. The Greek version of the general Nordic questionnaire for the analysis of musculoskeletal disorders (NMQ) was used for data collection.Results. A total of 455 subjects answered the NMQ. Three hundred and seventy-six (82.6%) of the study population reported having one or more symptoms during the previous year. Low back (56.9%), neck (34.1%), shoulder (29.9%) and knee (27.9%) were the commonest sites of pain. In almost half cases (48.6%), the complaints about pain were accompanied by activity restrictions. Multivariate statistical analysis showed significant correlations with increasing age and female gender (P < 0.05). Common chronic conditions were associated with co-morbidities from the musculoskeletal system. Only one-third of those who reported MSDs had consulted their GPs for the same problems within the previous year.Conclusions. MSDs are highly prevalent among rural population in Crete but fewer patients seek care than those who report symptoms.Greece, musculoskeletal disorderspain patternsprimary careruralIntroductionMusculoskeletal disorders (MSDs) comprise a major health problem for the general population, affecting their quality of life, demanding increased health care and organization.1 According to reports from Canada and the Netherlands,2,3 the prevalence of musculoskeletal problems range from 29% to 74.5%, respectively. However, the annual consultation rate by health care professionals of musculoskeletal problems is about 20%.4 MSDs often cause pain and significant disability, especially in the elderly population, signifying a heavy community burden.5 Furthermore, working conditions correlate well with MSDs.6 In family practice settings, MSDs often co-exist with multiple medical conditions, addressing special care needs.7In Greece, although MSDs have been considered as common reasons for patients' visits to GPs in rural areas,8,9 issues of pain patterns and consulting behaviour are still unexplored. This paper reports on the magnitude of musculoskeletal problems within the primary care setting with emphasis on co-morbidities and consulting behaviour and intends to discuss its implications for GPs in a southern European country.MethodsSettingData were gathered from three PCCs in rural Crete. These PCCs provide around the clock outpatient care for a rural district, serving a population of approximately 37 000 inhabitants10 with most of them being farmers [International Standard Classification of Occupations, 88 (ISCO-COM 6) and craft and related trades sectors (ISCO-COM 7)]. People are also employed in restaurant services and travel-related services (ISCO-COM 5), since a lot of tourists are visiting these areas. The unemployment rate is nearly 12%, slightly higher, as compared to 10% in the rest of the country. The educational status is low with more than 50% of the general population being either illiterate or with primary education only [International Standard Classification of Education 1997 (ISCED 0 + 1)].SamplingAll consecutive visitors (20–75 years of age) in the three PCCs during a period of 10 working days in November 2002 participated in the study. A written consent was obtained and the study protocol was applied by health professionals. The month of November was selected because most agricultural tasks in the fields take place during this period throughout the catchment area of the PCCs. During that period, farmers are engaged in intensive harvesting of grapes and olives, demanding hard manual work with the spine in a bend over position for a prolonged period of time. Other awkward positions during farming can also extract musculoskeletal strain. Moreover, at this time there are no tourists visiting the PCCs, so the study population comes from the permanent inhabitants served by these PCCs. The sample collected was further scrutinized for clinical co-morbidities and the consulting behaviour, using the medical records of one PCC.MethodsThe Greek version of the general form of the standardized Nordic questionnaire for the analysis of musculoskeletal symptoms (NMQ) was used for data collection. NMQ had been previously translated and validated in the Greek language.11 Apart from a few general demographic questions, it contains a picture of the back aspect of the human body divided in nine anatomical areas (neck, shoulders, elbows, wrists/hands, upper back, low back, hips/thighs, knees and ankles/feet) (Fig. 1). The rest of the questions are distributed in three columns. There are nine screening questions in the first column. For example, ‘have you at any time during the last 12 months had trouble i.e. ache, pain or discomfort in this area?’. In case of a positive answer, the respondents answer whether or not they had been prevented from doing their normal work at home or away from home, during the same period because of that ailment. They are also asked if they had experienced that symptom during the previous 7 days (questions in the second and the third column). Therefore, the questionnaire's structure investigates the prevalence of different MSDs and examines which of the particular problems are the more debilitating (i.e. prohibited patients from engaging in their normal work routine within the previous 12 months). Even though, NMQ is self-administered, because of the special features of the study population (mean age of 51.9 years and education level between 0 and 6 years), trained health professionals assisted the subjects with the completion of the questionnaire.FIGURE 1The back aspect of the human body as it appears in the NMQAdditional data concerning age, sex, body mass index (BMI) and working records were extracted by the NMQ. BMI was classified into four categories: underweight (BMI < 20 kg/m2), normal (BMI ≥ 20 and < 25 kg/m2), overweight (BMI ≥ 25 and < 30 kg/m2) and obese (BMI ≥ 30 kg/m2).12 To determine the extent to which NMQ identified musculoskeletal symptoms during primary care visits, we carried out a medical audit in one of the participating PCCs and all medical records of the subjects were reviewed. Co-morbidities (according to International Classification for Primary Care 2), education level (according to ISCED 1997), type of occupation (according to ISCO-COM 88), cohabitation status (any kind of cohabitation or living alone) and consultation patterns were also collected using a pre-tested questionnaire.Statistical analysesStatistical analyses were performed using SPSS version 12.0. A P-value of P < 0.05 was accepted as significant. For the nine anatomical areas, the 1-year period prevalence, the prevalence of daily limitations due to reported symptoms and the point prevalence (musculoskeletal symptoms during last week) were calculated. Those were the dependent variables. Independent variables were age, gender, BMI, working time at present work and working hours per week. Adjusted odds ratios (ORs) were calculated. Multiple logistic regression analysis was performed to evaluate the association between each exposure variable and musculoskeletal symptoms. This method was applied stepwise in an enter approach. The association between education level, type of occupation, cohabitation status, co-morbidities and the patterns of pain was also measured. To study the effects of different occupations to limitations in daily activities, the total score of positive answers for each patient and for each question was calculated. Wilcoxon paired and unpaired (Mann–Whitney) tests were used to compare these scores into the same and among different occupational groups. Chi-square test was used to compare those who had consulted their physicians for MSDs with those who did not.ResultsSample characteristicsOverall, 455 subjects completed the questionnaire at the three PCCs with participation rate at 95.2%. Table 1 shows the general demographic characteristics. Two hundred and sixty (57.2%) of the respondents were females and the mean age was 51.9 [95% confidence interval (CI): 50.3–53.7] years.Table 1Sample characteristicsn = 455, %GenderMen42.8Women57.2Age group20–3929.340–5928.260–7542.5BMIUnderweight (BMI < 20)0.8Acceptable (20 ≤ BMI < 25)20.2Overweight (25 ≤ BMI < 30)55.4Obese (BMI ≥ 30)21.1No records available2.4Residence areaAnogeia15.4Perama45.9Spili38.7Years at present work<1–412.15–1420.415–2413.025–3412.335–4411.645+4.8No records available25.7Weekly working hours5–144.215–245.325–342.935–4417.145–5425.155+17.1No records available28.3Patterns of painThree hundred and seventy-six subjects (82.6%) reported at least one musculoskeletal problem during the previous year, while 219 (48.1%) subjects reported limitations of activities due to their symptoms during the same period of time. Low back pain was the commonest complaint (56.9%), followed by neck (34.1%), shoulder (29.9%) and knee (27.9%) problems. Two musculoskeletal co-morbidities were present in 24% of patients (n = 109); one in 23.3% (n = 105), while 3.3% (n = 15) reported widespread pain in all nine regions of the body. Low back pain was also the most debilitating symptom reported (28.6%), causing limitations in daily activities. Low back and neck pain was the commonest combination reported (n = 124, 27.5%), resulting in the higher disability rate (11.0%). Although the co-existence of more than two symptoms was not frequent, this seemed to be associated with more disability but without statistical significance. The reported patterns of pain by age groups and gender are presented in Table 2.Table 2Patterns of pain by age and genderPain patterns/age groupsPrevalence of pain during the last year/genderMalesFemalesLevel of significancean%n%Neck4121.011443.8<0.0001 20–39 years31.5166.20.025 40–64 years157.74918.80.001 65+ years2211.34818.50.048Shoulder4523.19135.00.008 20–39 years73.6197.30.139 40–64 years157.73413.10.092 65+ years2211.33714.20.441Elbow189.23915.00.087 20–39 years00.031.20.336 40–64 years73.6197.30.139 65+ years105.1176.50.670Wrist/hand2110.86725.8<0.0001 20–39 years31.593.50.307 40–64 years63.13513.50.0002 65+ years126.2238.80.393Upper back2010.37729.6<0.0001 20–39 years31.5114.20.166 40–64 years63.1186.90.113 65+ years115.62911.20.054Low back8644.117366.5<0.0001 20–39 years2412.33513.50.813 40–64 years2914.97328.10.001 65+ years3216.46324.20.056Hip2010.37729.6<0.0001 20–39 years21.083.10.234 40–64 years42.13212.30.0001 65+ years136.73613.80.023Knees4322.18432.30.024 20–39 years63.1114.20.716 40–64 years115.62710.40.096 65+ years2713.83613.80.890Foot/ankle3316.95220.00.926 20–39 years42.193.50.549 40–64 years84.1228.50.093 65+ years2010.3218.10.519a Statistical significant values are printed in italic.The prevalence of the reported symptoms was statistically related to increasing age (P < 0.05) for neck, elbow, low back, hip and knee pain. Multiple musculoskeletal symptoms were also related to aging. Patients with shoulder and low back pain reported also more disability in terms of limitation in their daily activities.Morbidity patterns also differed between women and men. While low back pain was the commonest symptom in both genders, female patients reported higher prevalence for every symptom (P < 0.05) except for elbow and foot (Table 2). Women reported more pain conditions than men 11.9/10.7 (P < 0.05) and more disability.In the questions about working schedules, the majority of the sample reported 20 (95% CI: 18.6–21.8) years of work for 45.6 (95% CI: 43.6–47.6) hours per week. Men worked longer hours but this tended to decrease, as they grew older. Women worked mainly at home or at the farms, fewer hours per week than men (men/women ratio: 1.1). Working for a period of more that 15 years was related to higher morbidity mainly from shoulder and elbow pain (Table 3). Limitations in daily activities due to neck, shoulder, low back and foot complaints were related to longer working schedule per week. Multiple regression analysis for the most frequent patterns of pain is presented in Table 3.Table 3Logistic regression analysis of most frequent patterns of pain by age, gender, BMI and working timeVariableBSEP-valueExp. B95% CILow back pain Age0.0230.0120.0461.0241.000–1.047 Gender0.8580.2390.00032.3591.477–3.768 BMI0.0330.0310.2941.0330.972–1.098 Years in the present work0.0010.0120.9561.0010.977–1.025 Hours per week0.0050.9840.4580.9950.981–1.009Neck pain Age0.0540.0130.000031.0561.029–1.083 Gender1.3230.2740.0000013.7562.193–6.431 BMI0.0090.0330.7920.9910.930–1.057 Years in the present work0.0070.0130.5820.9930.968–1.018 Hours per week0.0100.0080.1910.9900.975–1.005Shoulder pain Age0.0140.0130.2851.0140.989–1.040 Gender0.6310.2670.0181.8791.114–3.171 BMI0.0290.0320.3731.0290.966–1.096 Years in the present work0.0280.0130.0311.0281.033–1.055 Hours per week0.0040.0080.5961.0040.989–1.019Knee pain Age0.0330.0130.0111.0341.008–1.061 Gender1.0060.2800.0032.7351.578–4.738 BMI0.0330.0330.3211.0330.969–1.102 Years in the present work0.0150.0130.2581.0150.989–1.041 Hours per week0.0070.0080.3961.0070.991–1.022Further analysis from the sample in one PCC showed that occupation did not constitute a statistically significant risk factor for the presence of MSDs. Thirty-one per cent of the sample were farmers (ISCO-COM 6), 11% were employed in restaurant services and travel-related services (ISCO-COM 5) and 27%—almost half of all the women—were housewives. Comparing the total scores in the three questions of the NMQ among various occupational groups, it came out that the majority kept their usual everyday activities during the previous year, even though they reported a great number of musculoskeletal symptoms (P < 0.001). Different occupations did not statistically affect this result.Cohabitation and educational status did not significantly differ in the patterns of reported symptoms.Co-morbiditiesHalf of the study population (55.4%) for both sexes were overweight (BMI > 25 kg/m2) with mean BMI = 29.0 (95% CI: 28.6–29.9), especially women in advanced age (BMI > 30 kg/m2 in 24.6% of females) (Pearson R correlation, Rp = 0.328, P < 0.001 of correlation between overweight women and age). Overweight (BMI > 25) and obese (BMI > 30) patients had a tendency to report more musculoskeletal symptoms but not of statistical significance. However, they reported more activity limitations as a result of shoulder, elbow, wrist and foot pain.Review of medical records in one PCC revealed that the most common diagnoses co-existing with MSDs were hypertension (64/176, 36%), osteoarthritis (OA) of hip and/or knee (56/176, 32%), coronary artery disease (CAD) (21/176, 12%) and mental disorders (20/176, 11%). The number of co-morbidities was increasing with age without statistical significance. Mental disorders, mainly depression, ranked as the medical conditions with the highest proportion of musculoskeletal symptoms (95.0%), followed by chronic obstructive pulmonary disease (COPD) (92.8%), CAD (90.4%), OA (84.0%), diabetes (83.3%) and hypertension (82.8%). The multiple logistic regression analysis of the most common co-morbidities in correlation to the reported symptoms highlighted strong correlations, especially with arthritic conditions. Shoulder pain was statistically associated with COPD (adjusted OR = 8.94, 95% CI 1.47–54.17, P = 0.017) and OA (adjusted OR = 4.25, 95% CI 1.67–10.81, P = 0.002), whereas wrist pain was more frequent in subjects with diabetes (adjusted OR = 3.87, 95% CI 1.09–13.68, P = 0.035). The prevalence of elbow pain was correlated to CAD (adjusted OR =5.01, 95% CI 1.01–24.87, P = 0.048). Patients with OA also reported statistically significant activity restriction (P < 0.05) due to neck, shoulder, upper back, hip, knee and foot symptoms, unlike diabetic patients who did not complain for many daily restrictions. Cancer of any form, mental disorders or stomach problems were not statistically correlated to MSDs.Consulting behaviourData from one PCC showed that although the prevalence of reported MSD was measured at 71.4% (125 out of 176), only 56 subjects (32%) had reported the same symptoms to their GP during previous consultations. These symptoms originated mainly from low back, neck, knee and hip areas. Patients with multiple co-morbidities were more likely to have previously consulted their GPs for those problems (adjusted OR = 2.21, 95% CI 1.6–2.9, P < 0.0001). Seventy-four of the participants (41.7%) reported symptoms during the last week (question in the third column of the NMQ), but only 13 (17.5%) visited their GP complaining for musculoskeletal symptoms. When comparing those visiting GPs with those who did not (chi-square test) during the previous year, we found that younger patients of both genders with limited education did not consult their physician due to MSDs. A patient with knee or shoulder pain was seeking care more often regardless of age. Older patients and patients suffering from OA or hypertension reported significantly more consultations. Other factors like gender, BMI, working time or occupation were not found to have statistical significance.DiscussionSummary of resultsThere are several interesting points revealed by this study such as the patterns and distribution of pain, the differences between genders and the consulting behaviour. Low back pain was the commonest reported site among all age groups. Low back pain, followed by neck and shoulder pain, was also the main symptom for daily living limitations. This finding is in accordance with the results of a recent systematic review in general practice of work-related diseases.13 Age was found to be a risk factor for many musculoskeletal symptoms. Younger patients (20–39 years) reported fewer symptoms and less frequently, as compared with older patients (66.2% prevalence of any MSDs versus 90% for the older group over 60 years of age, P < 0.05).Pain patterns also differed between genders. In our study, women reported on average more symptoms and more disability. Women to a greater extent than men blamed musculoskeletal pain in neck, wrist, low back, hip or knee for restrictions in daily activities. The tendency for MSDs to affect more women than men has been shown in other epidemiological studies.3,14 Reasons for gender differences, apart from the apparent biological differences (genetics, physiology, hormones, etc.), include different psychosocial aspects of symptoms and care. Women also seem to visit more promptly health care facilities and recall health problems to a greater extent than men.15 However, in our study, such gender preference was not revealed.In our study, overweight or obesity, a common public health problem in contemporary Crete,16 showed significant correlations only with the reported disabilities due to pain of upper limb and foot. According to other studies, obesity has been related to hip and low back pain in middle-aged men,17 whereas in rural communities, it causes more disability to patients with knee OA.18Working for many years was correlated to shoulder and elbow pain. More comprehensive surveys that included physical and psychosocial work factors, such as work intensity or working postures, have previously shown associations with MSDs,6 but these were not explored in our study.Unlike other epidemiological evidence which relates MSDs to physically demanding professions like farming,19 occupation did not statistically relate to any MSD. However, analysing the positive answers for all occupational groups, we found that the majority of participants carried on with their daily functions, independently of their job, despite the occurrence of multiple musculoskeletal problems. Cultural or local social network variables, not included in this study, could play a role in the perceived limitations of activity for this population.20Other social factors, such as living alone or low education, were not found to affect the reported patterns of pain, which comes in agreement with other studies.21The clinical co-morbidity rate was high. Patients with OA, COPD, diabetes, hypertension or CAD tended to report more musculoskeletal symptoms. Co-morbidities with MSDs were also shown to prevent their daily activities. Co-morbidity for OA in general practice has also been found to be extensive with musculoskeletal as well as non-musculoskeletal conditions in the UK.22 Patterns of co-morbidity in the Dutch population have shown that MSDs were most likely to coincide with lung and heart diseases, neurological disorders, diabetes and cancer.23 The explanation for the perceived co-morbidity lies beyond the present study in shared pathology of diseases or the impact which one condition might have in the occurrence of another. However, the recognition of this large extent of co-morbidity has implications for the way in which primary care should be organized. The burden is on primary care physicians to provide the majority of care, not only for the target condition but also for all co-morbid conditions.Another interesting point was that only 13 out of 74 who reported musculoskeletal symptoms were actually seeking care at the time of the study. On an annual basis, 32% of this sample had contacted a primary care physician because of musculoskeletal pain which can be considered high in comparison with other studies,4 although it corresponds to less than half of the same population reporting MSDs in the NMQ. The more concurrent chronic conditions they had, the more likely they were to have consulted their GP and complain for MSDs, too. Those were mainly older patients with higher education suffering from OA and hypertension that came complaining for knee or shoulder pain. Similarly, another study in a rural Swedish population showed that only one-fifth of those reporting current neck and/or low back pain had a primary care consultation.24 Possible explanations could be either the poor registration rates of primary care physicians who underestimate the symptoms or that symptoms are not severe enough to be reported. A recent report on primary care consultations due to indigestion problems in the same district showed under diagnosis of functional gastrointestinal disorders by physicians.25 However, a recent European study involving eight countries regarding musculoskeletal pain showed that up to 27% of people with pain do not seek medical care in spite of constant or daily pain.26 The disbelief or ignorance of effective treatments for MSDs could prevent patients from consulting a health professional.23 Even when they do seek help, this is often limited to prescription for non-steroidal anti-inflammatory drugs (NSAIDS), which result in non-compliance to their treatment.26 In fact, according to a previous study conducted in the same district, prescriptions for painkillers and NSAIDS ranked second accounting for 19% of prescriptions during 1-year period.27 To overcome the doctors' difficulty in identifying those diseases, a change in doctor–patient communication with an emphasis on patients' perceptions and needs seems as a promising option.26Strengths and weaknessSeveral concerns should be raised when discussing study findings. Overestimating prevalence due to sampling method is one issue: our sample was probably overrepresented by patients with multiple morbidities waiting to see their physicians. High prevalence of MSDs has been documented in many studies, especially when self-administered questionnaires are being used for data collection.28 People tend to recall the most recent or most debilitating painful conditions when asked to complete questionnaires. Whether or not the reported symptoms were major or minor cannot be explored through the NMQ.Although the NMQ has been used in a great number of studies in different countries,29 it still has several limitations. Since NMQ is counting symptoms throughout the human body, the estimated prevalence is expected to be high. Most general population studies focus on special anatomic sites, while others deal with specific occupational groups.30,31 The used questionnaire seems to be appropriate for general screening in occupational groups, without permitting further discrimination. For example, the pain in the hips, considering the area marked as ‘hip/thighs’ in the picture of NMQ is unclear if it is actually a hip problem or originates from the low back area and reflects at the hips. Moreover, the use of a picture of the back aspect of the body may not allow problems in the frontal aspect to be revealed.The timing of data collection did not seem to have an effect on the measured prevalence. Data from one PCC that was collected during 10 working days in spring did not show a seasonal variation for the reported MSDs.Although the sample size reflected the underlying population i.e. the general population of rural primary care in Crete (95% confidence level and 95% CI 7%), limitations should be taken into account interpreting the results for co-morbidities and consulting behaviour since the data came only from one PCC (n = 175).ConclusionsIn conclusion, this study showed that musculoskeletal pain is very common in the rural population of Crete. GPs need to be alert and well skilled in order to prevent or identify early the musculoskeletal problems. The observed patterns of co-morbidity highlight the need for primary care oriented towards patients' overall health care. Patients, especially women and elderly, who suffer from hypertension or diabetes, could also suffer from musculoskeletal problems, even if this does not come up as their major complain, and they should be treated accordingly. Further evaluation of the consequences of MSDs to this population in terms of disability or effect in the quality of life should contribute towards that effort.DeclarationFunding: None.Ethical approval: Approval of a specific scientific committee of the University of Crete.Conflicts of interest: None.1WoolfADPflegerBBurden of major musculoskeletal conditionsBull World Health Organ2003816466562BadleyEMWebsterGKRasoolyIThe impact of musculoskeletal disorders in the population: are they just aches and pains? Findings from the 1990 Ontario Health SurveyJ Rheumatol1995227337393PicavetHSJSchoutenJSAGMusculoskeletal pain in the Netherlands: prevalences, consequences and risk groups, the DMC3-studyPain20031021671784BadleyEMRasoolyIWebsterGKRelative importance of musculoskeletal disorders as a cause of chronic health problems, disability and health care utilization: findings from the 1990 Ontario Health SurveyJ Rheumatol1994215055145MakelaMHeliovaaraMSieversKKnektPMaatelaJAromaaAMusculoskeletal disorders as determinants of disability in Finns aged 30 years or moreJ Clin Epidemiol1993465495596ColeDCIbrahimSAShannonHSScottFEylesJWork correlates of back problems and activity restriction due to musculoskeletal disorders in the Canadian national population health survey (NPHS) 1994–5 dataOccup Environ Med2001587287347MartinFBravoGHudonCVanasseALapointeLPrevalence of multimorbidity among adults seen in family practiceAnn Fam Med200532232288TzimisLKatsantonisNLeledakiAVasilomanolakisKKafatosAAntibiotics prescription for indigent patients in primary careJ Clin Pharm Ther1997222272359KoutisADIsacssonALionisCLindholmLHSvenningerKFioretosMDifferences in the diagnose panorama in health care in Dalby, Sweden and Spili, CreteScand J Soc Med199321515810General Secretariat of National Statistical Service of Greece, 2001http://www.statistics.gr (accessed on 6 January 2005)11AntonopoulouMEkdahlCSgantzosMAntonakisNLionisCTranslation and validation into Greek of the standardised general Nordic questionnaire for the musculoskeletal symptomsEur J Gen Pract200410353612Measuring Obesity: Classification and Description of Anthropometric Data. Report on a WHO Consultation on the Epidemiology of Obesity, Warsaw, 21–23 October, 19871988CopenhagenWHO Regional Office for Europe, (document EUR/ICP/NUT 125)13WeeversHJvan der BeekAJAnemaJRvan der WalGvan MechelenWWork-related disease in general practice: a systematic reviewFam Pract20052219720414LerouxIDionneCEBourbonnaisRBrissonCPrevalence of musculoskeletal pain and associated factors in the Quebec working populationInt Arch Occup Environ Health20057837938615FillingimREdwardsRPowellTSex-dependent effects of reported familial pain history on recent pain complaints and experimental pain responsesPain200086879416MamalakisGKafatosAPrevalence of obesity in GreeceInt J Obes Relat Metab Disord19962048849217BergeruddHNillsonBThe prevalence of locomotor complaints in middle age and their relationship to health and socioeconomic factorsClin Orthop199430826427018JordanJMLutaGRennerJBSelf-reported functional status in osteoarthritis of the knee in a rural southern community: the role of sociodemographic factors, obesity, and knee painArthritis Care Res1996927327819HolmbergSStiernstromELThelinASvardsuddKMusculoskeletal symptoms among farmers and non-farmers: a population-based studyInt J Occup Environ Health2002833934520PincusTBurtonAKVogelSFieldAPA systematic review of psychosocial factors as predictors of chronicity/disability in prospective cohorts of low back painSpine200227E109E12021SaastamoinenPLeino-ArjasPLaaksonenMLahelmaESocio-economic differences in the prevalence of acute, chronic and disabling chronic pain among ageing employeesPain200511436437122KadamUTJordanKCroftPRClinical comorbidity in patients with osteoarthritis: a case control study of general practice consulters in England and WalesAnn Rheum Dis20046340841423WestertGPSatarianoWASchellevisFGVan de BosGAPatterns of comorbidity and the use of health services in the Dutch populationEur J Public Health20011136537224HolmbergSAThelinAGPrimary care consultation, hospital admission, sick leave and disability pension owing to neck and low back pain: a 12-year prospective cohort study in a rural populationBMC Musculoskelet Disord200676625LionisCOlsen-FaresjoAAnastasiouFWallanderMAJohanssonSFaresjoTMeasuring the frequency of functional gastrointestinal disorders in rural Crete: a need for improving primary care physicians' diagnostic skillsRural Remote Health2005540926WoolfADZeidlerHHaglundUMusculoskeletal pain in Europe: its impact and a comparison of population and medical perceptions of treatment in eight European countriesAnn Rheum Dis20046334234727AntonakisNTsoulouSPeiosDSimeonidisPLyrarakiELionisCDrug prescribing in primary health. Part II: drugs for cardiovascular, musculoskeletal and central nervous system diseasesArch Hell Med2001185057(in Greek)28PicavetHSHazesJMPrevalence of self reported musculoskeletal disease is highAnn Rheum Dis20036264465029BaoSWinkelJShahnavazHPrevalence of musculoskeletal disorders at workplaces in the People's Republic of ChinaInt J Occup Saf Ergon2000655757430MansfieldNJMarshallJMSymptoms of musculoskeletal disorders in stage rally drivers and co-driversBr J Sports Med20013531432031HagenKBMagnusPVetlesenKNeck/shoulder and low-back disorders in the forestry industry: relationship to work tasks and perceived psychosocial job stressErgonomics19984115101518
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B2E51E5CC19F61CCD947B081C04BF3592E1CC86.txt b/test/dataset/in/resources/corpus/Clean_0B2E51E5CC19F61CCD947B081C04BF3592E1CC86.txt
deleted file mode 100644
index de2be13..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B2E51E5CC19F61CCD947B081C04BF3592E1CC86.txt
+++ /dev/null
@@ -1 +0,0 @@
-annoncannoncAnnals of Oncology1569-80410923-7534Oxford University Press10.1093/annonc/mdn179original articlespalliative care/quality of lifeUse of palifermin for the prevention of high-dose methotrexate-induced oral mucositisSchmidtE.ThoennissenN. H.RudatA.BiekerR.SchliemannC.MestersR. M.ZühlsdorfM.Müller-TidowC.BerdelW. E.*Department of Medicine, Hematology and Oncology, University Hospital Muenster, Muenster, Germany*Correspondence to: Dr W. E. Berdel, Universitätsklinikum Muenster, Medizinische Klinik, Albert-Schweitzer-Strasse 33, D-48149 Muenster, Germany. Tel: +49-251-8347587; Fax: +49-251-8347588; E-mail: berdel@uni-muenster.de92008252008199164416492811200726320083132008© The Author 2008. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: journals.permissions@oxfordjournals.org2008Background: Oral mucositis is a frequent problem after high-dose methotrexate (HD-MTX), impairing patient's quality of life, leading to higher rates of infections and delaying subsequent chemotherapy. This report describes the effect of palifermin in patients treated within the GMALL-B-ALL 2002 protocol containing HD-MTX who developed a severe mucositis in cycle A1/B1.Patients and methods: Ten patients, all with World Health Organization grades III–IV oral mucositis in cycles A1/B1, obtained palifermin with subsequent similar or identical cycles to reduce mucositis. Thus, patients serve as their own control for efficacy of palifermin.Results: All 10 patients developed grades III–IV mucositis in cycles A1/B1 without palifermin, whereas only two of 10 developed grades III–IV mucositis in corresponding cycles A2/B2 with palifermin. Only four of 10 patients showed infections in the cycles with palifermin compared with 10 of 10 patients without palifermin. The duration of mucositits in patients who acquired a higher grade mucositis despite treatment with palifermin could be reduced from 12.9 days (median) without to 11 days with palifermin. The amount of i.v. opioid analgetics could be significantly reduced.Conclusion: Palifermin might reduce the incidence, severeness and duration of oral mucositis in HD-MTX-based chemotherapy and may influence clinical sequelae such as infection and quality of life.methotrexateoral mucositispaliferminintroductionOral mucositis is a severe and frequent adverse effect of high-dose methotrexate (HD-MTX)-based chemotherapy. It is caused by damage of the mucosal lining of the gastrointestinal tract (GIT) as a result of a dynamic series of biological events involving different cellular and tissue compartments of the GIT mucosa. Understanding of the pathobiology has increased rapidly during the last years [1–5]. Mucosal barrier injury is based on a network of interactions involving the endothelium, extracellular matrix, metalloproteinases, submucosal reactions and connective tissue [1–3]. Sonis and coworkers developed a pathophysiological model dividing the dynamical process of mucositis in five phases: initiation, up-regulation, signal amplification, ulceration and healing [2, 3].Oral mucositis and myelosuppression represent dose-limiting toxicity effects of HD-MTX. Mucositis is associated with an increased risk of life-threatening infections [6, 7], with the need for total parenteral nutrition, i.v. analgetic therapy, and may lengthen hospital stay with increased economic burden and consumption of health care resources [8–11]. Oral mucositis impairs the patient's quality of life by pain, the inability to eat, swallow and talk. For many patients with grades III–IV mucositis, the subsequent cycle of chemotherapy is delayed. A grade IV mucositis is an emergency situation in oncology, ethically demanding to relieve the patient's situation. The recent evidence-based management guidelines from the Multinational Association of Supportive Care in Cancer [12] make several recommendations to provide a better patient comfort including good oral hygiene, oral decontamination with antibacterial and antifungal mouthwash and topical and systemic pain management.Numerous substances have been used in addition to treat oral mucositis such as ice chips [13, 14], antioxidants as glutamine, N-acetylcysteine, benzydamine hydrochloride and antiinflammatory agents like prostaglandine E1 and E2 [15]. Until now none of them has proven an unequivocal clinical benefit.Palifermin is a recombinant human keratinocyte growth factor (KGF) which is known to stimulate growth of epithelial cells in a wide variety of tissues. In murine models, a beneficial effect on mucositis could be shown [16]. In December 2004, the Food and Drug Administration approved palifermin (recombinant KGF from Escherichia coli, Amgen, Thousand Oaks, CA, USA) on the basis of randomized data showing a decrease of incidence and duration of severe oral mucositis in patients with hematologic malignancies receiving myelotoxic therapy requiring hematopoietic stem-cell support [17]. Therefore, the updated guidelines of the Multinational Association of Supportive Care in Cancer recommend the use of palifermin in patients who are receiving high-dose chemotherapy and total body irradiation with autologous stem-cell transplantation (level 1, grade A recommendation) [12].KGF receptors are present on epithelial cells in many tissues including the tongue, buccal mucosa, salivary gland, esophagus, stomach, intestine, lung, liver, pancreas, kidney, bladder, mammary gland and eye lens as well as on tumor cells of different histology [18]. Receptors could not be found on lymphoma and lymphoblastic leukemia cell lines [18].The mechanism of action of palifermin in mucositis is not completely elucidated. It leads to a down-regulation of proinflammatory cytokines [19] and increases antiinflammatory cytokines such as IL-13 [20, 21]. Palifermin protects the epithelium against reactive oxygen molecules by modifying the expression of detoxifying transcription factors and enzymes [22].In this series, we observed patients who were treated in the GMALL-B-ALL/B-NHL 2002 protocol, which is based on HD-MTX and is used for certain aggressive B-cell lymphomas and lymphoblastic leukemias. The incidence of grades III–IV (World Health Organization, WHO, oral toxicity scale, Miller et al. [23]) mucositis upon HD-MTX within this protocol ranges from 56% (cycle A1) to 32% (cycle B1) as described by Hoelzer et al. [24]. Another study shows a mucositis rate of 42% [25]. The data of the B-NHL90 trial (GMALL study group) with 3 g MTX/m2 (>55 years 500 mg MTX/m2) showed a grade III (Donelly scale) mucositis rate of 41%–54% in cycle A1 and 61%–75% in cycle B1. In previous trials (B-NHL 83, B-NHL 86), mucositis grades I–IV were observed in 50%–53% of chemotherapy cycles [26].All patients reported here developed a severe mucositis in cycle A1 or B1. In order to prevent additional mucositis, these patients received palifermin for the corresponding following cycles of similar or identical chemotherapy. Thus, patients in this retrospective series can serve as their own control for efficacy of palifermin in preventing mucositis.patients and methodspatientsWe report on a retrospective series of 10 patients with B-cell acute lymphoblastic leukemia (B-ALL) and aggressive B-cell lymphoma classified according to the WHO classification. Patients' characteristics are given in Table 1. We have included only patients into this report who developed severe mucositis in initial treatment cycle A1 or B1 without palifermin and were then given palifermin in the subsequent treatment cycles.Table 1.Patients treated in the GMALL-B-ALL/B-NHL 2002 protocol (severe infection is defined as body temperature >38.5°C and use of systemic antibiotics)Patient ID No.DiagnosisMaximum grade of mucositis (WHO)Maximum grade of mucositis (WHO)Dosage of paliferminWithout paliferminWith palifermin1B-ALLGrade IV, A1, 12 days; Grade III, B1, 8 days; infection A1, B1Grade 1 A2–C2 10 daysA2 60 μg/kg bodyweight (BW); B2 and C2 30 μg/kg BW2Burkitt-like lymphomaGrade IV, A1, 15 days; grade III, C1, 14 days; infection A1, C1Grade II, B1, 9 days; grade II, B2, 7 daysA2–C2 60 μg/kg BW; B1 30 μg/kg BW3Mediastinal B-NHLGrade IV B1, 15 days; infection B1Grade II C1, 4 daysA2–C2 60 μg/kg BW; C1 30 μg/kg BW d−3 to −1; 60 μg/kg BW d1–3 after chemotherapy4Mediastinal B-NHLGrade IV A1, 10 days; grade III B1, 13 days; infection A1–C1Grade I A2A2–C2 30 μg/kg BW5B-ALLGrade IV A1, 13 days; grade III, B1 7 days; infection A1Grade II cycle B2, 5 daysA2–C2 30 μg/kg BW6Burkitt lymphomaGrade II, A1, 4 days; grade IV, B1 18 days; grade III, C1, 18 days; infection A1, C1Grade I, A2, 6 days; grade I, B2, 7 days; infection A2, B2A2 ∼40 μg/kg BW; B2–C2 ∼50 μg/kg BW7Atypical Burkitt lymphomaGrade IV B1, 12 days; grade II A1, 5 days; infection A1, B1, C1Grade III, A2, 10 days; grade IV, B2, 12 days; infection A2A2-B2, 60 μg/kg BW; protocol could not be finished because of infectious complications8B-ALL, elderly protocolGrade IV A1, 16 days; infection A1Grade II B1/B2, 3 days; infection B2∼40 μg/kg BW B1–B39Burkitt lymphomaGrade IV B1, 12 days; infection B1Grade III C1 13 days; grade III A2, 10 days60 μg/kg BW10B-ALLGrade IV A1, 14 days; grade IV B1, 8 days; infection A1, B1Grade I A2-C2, 7 days; grade III C1, 10 days; infection C160 μg/kg BW C1–C2treatment protocolThe patients were treated from September 2004 to March 2007 on protocol GM-ALL-B-ALL/B-NHL 2002 (upon approval by ethical board and written informed consent) and developed severe mucositis. The protocol outline with age-related dose modifications is depicted in Figure 1. Two of 10 patients were treated in the protocol for patients >55 years of age with a slightly reduced MTX dosage (500 mg/m2 compared with 1500 mg/m2). One patient was treated in the ‘elderly protocol’.Figure 1.Prephase: cyclophosphamide 200 mg/m2 d1–5, prednisone 60 mg/m2 d1–5. Cycle A1/A2: rituximab 375 mg/m2 d7 (A2 d77); dexamethasone 10 mg/m2 d8–12 (A2 d78–82); vincristine 2 mg d8 (A2 d78); MTX 1500 mg/m2 (500 mg/m2 > 55 years) d8 (A2 d78); ifosfamide 800 mg d8–12 (A2 d78–82); cytarabine 2 × 150 mg/m2 d11–12 (A2 d81–82); VP16 100 mg/m2 d11–12 (A2 d81–82); cytarabine 40 mg i.th. d8, 12 (A2 d78, 82); MTX 15 mg i.th. d8, 12 (A2 d78, 82); dexamethasone 4 mg i.th. d8, 12 (A2 d78, 82). Cycle B1/B2: rituximab 375 mg/m2 d28 (B2 d98); dexamethasone 10 mg/m2 p.o. d29–33 (B2 99–103); vincristine 2 mg d29 (B2 d99); MTX 1500 mg/m2 (500 mg/m2 > 55 years) d29 (B2 d99); cyclophosphamide 200 mg/m2 d29–33 (B2 99–103); adriamycine 25 mg/m2 d32–33 (B2 d102–103); cytarabine 40 mg i.th. d29, 33 (B2 d99, 103); MTX 15 mg i.th. d29, 33 (B2 d99, 103); dexamethasone 4 mg i.th. d29, 33 (B2 d99, 103). Cycle C1/C2: rituximab 375 mg/m2 d49 (C2 d119); dexamethasone 10 mg/m2 p.o. d50–54 (C2 120–124); vindesine 3 mg/m2 maximum 5 mg d50 (C2 120); MTX 1500 mg/m2 (500 mg/m2 > 55 years) d50 (C2 d120); VP16 250 mg/m2 d53–54 (C2 d123–124); cytarabine 2 × 2 g/m2 (2 × 1 g/m2 > 55 years) d54 (C2 d124). Protocol for elderly patients: Prephase: cyclophosphamide 200 mg/m2 d1–5, prednisone 60 mg/m2 d1–5. Cycle A1/A2/A3: rituximab 375 mg/m2 d7 (A2 d49, A3 d98); dexamethasone 10 mg/m2 d8–12 (A2 d50–54, A3 99–103); MTX 500 mg/m2 d8 (A2 d50, A3 99); ifosfamide 400 mg d8–12 (A2 d50–54, A3 99–103); cytarabine 2 × 60 mg/m2 d11–12 (A2 d53–54, A3 d102–103); VP16 60 mg/m2 d11–12 (A2 d53–54, A3 d102–103); MTX 12 mg i.th. d8 (A2 d50, A3 99). Cycle B1/B2/B3: rituximab 375 mg/m2 d28 (B2 d77, B3 d119); dexamethasone 10 mg/m2 p.o. d29–33 B2 78–82, B3 d120–124); vincristine 1 mg d29 (B2 d78, B3 d120); MTX 500 mg/m2 d29 (B2 d78, B3 d120); cyclophosphamide 200 mg/m2 d29–33 (B2 78–82, B3 120–124); adriamycine 25 mg/m2 d32–33 (B2 d81–82, B3 123–124); MTX 12 mg i.th. d29 (B2 d78, B3 120).Eight of 10 patients received a ‘run-in’ therapy (prephase) of 5 days with prednisone 60 mg/m2 and cyclophosphamide 200 mg/m2 before the cycle A1. Two patients with mediastinal B-cell non-Hodgkin's lymphoma (B-NHL) did not receive any prephase therapy.Patients with mucositis in cycle A1 or B1 received prophylaxis and supportive treatment including morphine, tetracaine-, panthenole and camomile-containing mouthwashes, oral hygiene and amphotericin B suspensions to prevent intestinal candidosis. All patients received recombinant granulocyte colony-stimulating factor (G-CSF) to shorten the time of grade IV neutropenia. Prophylactic oral hygiene, amphotericin B and G-CSF were not changed during the subsequent cycles.grading of mucositisMucositis grade was judged daily by the four-point WHO scale [23]. The observation was carried out by the authors and, to ensure uniformity of judgment, by the first and the last author during the time of mucositis maximum. All these patients developed severe WHO grades III–IV oral mucositis in the cycle A1 (seven of 10) or B1 (three of 10). One of these patients (number 9, Table 1) did not get MTX in cycle A1 because of a surgical bowel intervention.palifermin applicationBecause of their mucositis, all 10 patients obtained palifermin (Kepivance™, Amgen) for 3 days before and another 3 days after the following corresponding chemotherapy cycle (A2 or B2) to reduce the risk of further higher grade mucositis. Prior written informed consent was obligatory. Palifermin dose was 60 μg/kg/day, which is the full dose recommended or a lower daily dose as outlined in Table 1.resultsFigure 2A and B depicts the maximum grade of mucositis observed in chemotherapy cycles A1/B1 without palifermin versus A2/B2 with palifermin support. Evaluated together, there were 16 episodes of severe grade IV and one grade III mucositis in the chemotherapy cycles without palifermin support. In contrast, we have observed only one episode of grade IV, four episodes of grade III and four episodes of grade II mucositis during the chemotherapy cycles with palifermin support (difference P < 0.05, Mann–Whitney–Wilcoxon test).Figure 2.(A) Maximum grade mucositis with and without palifermin in cycles A1 versus A2 (P < 0.05, Mann–Whitney–Wilcoxon test). Numbers in circles refer to patients as given in Table 1. Patient 9 did not receive MTX during cycle A1. (B) Maximum grade mucositis with and without palifermin in cycles B1 versus B2 (P < 0.05, Mann–Whitney–Wilcoxon test). Numbers in circles refer to patients as given in Table 1. Patient 2 received palifermin in cycle B1.The amount of i.v. opioid analgetics could be reduced from a total of 2403 mg morphine in cycle A1 (3380 mg cycle B1) without palifermin to 640 mg in cycle A2 (480 mg cycle B2) with palifermin (difference P < 0.05, Mann–Whitney–Wilcoxon test). The intraindividual dose changes are presented in Figure 3A and B.Figure 3.(A) Total amount of i.v. given morphine (in milligram) with and without palifermin in cycles A1 versus A2 (P < 0.05, Mann–Whitney–Wilcoxon test). Numbers in diamonds refer to patients as given in Table 1. Patient 9 did not receive MTX during cycle A1. (B) Total amount of i.v. given morphine (in milligram) with and without palifermin in cycles B1 versus B2 (P < 0.05, Mann–Whitney–Wilcoxon test). Numbers in diamonds refer to patients as given in Table 1. Patient 2 received palifermin during cycle B1 and developed no higher grade mucositis.Furthermore, the duration of mucositis in patients who acquired a higher grade mucositis despite treatment with palifermin could be slightly reduced. The median duration of grade III/IV mucositis without palifermin was 12.9 days (16 cycles with mucositis) with a range from 7 to 18 days. The median duration of grades III–IV mucositis with palifermin was 11 days with a range from 10 to 13 days (not significant, Mann–Whitney–Wilcoxon test). In this evaluation, patients not developing severe mucositis under palifermin protection were not included.Ten of 10 patients developed a severe infection in cycle A1 or B1 without palifermin. Severe infection is defined as fever (>38.5°C) in neutropenia which had to be treated with empiric systemic antibiotic therapy. In the palifermin-supported chemotherapy cycles, only four of 10 patients showed a severe infection. Interestingly, infection is associated with a higher grade mucositis.Since cycles A1 and A2 contain identical (with the exception of the run-in phase of corticosteroids and cyclophosphamide given to eight of 10 patients before A1) and cycles B1 and B2 contain identical chemotherapy, patients can serve as their own controls for the antimucositis efficacy of palifermin. Figure 4 documents the maximum appearance of mucositis in three patients at identical times following chemotherapy cycles without (part A, C and E) or with palifermin (part B, D and F) support. Only three individual cycle comparisons for two patients (numbers 7 and 9, Figure 2, Table 1) did not show an improvement of mucositis severity by palifermin. Patient number 10 did receive palifermin in cycle C1, but showed a grade III mucositis. However, in the corresponding cycle C2, a mucositis of only grade I was observed.Figure 4.(A) Grade IV mucositis at day 13 after cycle B1 without palifermin support (patient 3, Table 1). (B) Grade I mucositis and hyperkeratosis at day 13 after cycle B2 with palifermin support (patient 3, Table 1). (C) Grade III mucositis at day 11 after cycle B1 without palifermin support (patient 10, Table 1). (D) Grade I mucositis at day 12 after cycle B2 with palifermin support (patient 10, Table 1). (E) Grade IV mucositis at day 11 after cycle B1 without palifermin support (patient 6, Table 1). (F) Grade I mucositis at day 10 after cycle B2 with palifermin support (patient 6, Table 1).As we report on an individual patient series, there was some variation in the course of supportive palifermin treatment concerning individual cases and these are depicted in Table 1.discussionWe report descriptive observation data from 10 patients who were treated in the trial protocol GM-ALL-B-ALL/NHL 2002 or the respective protocol for elderly patients (one patient) and developed severe mucositis. In summary, we observed a considerable effect of palifermin reducing the dosage of i.v. morphine, severity, duration and clinical sequelae (such as infection) of mucositis when similar (A1/A2) or identical (B1/B2) chemotherapy cycles without palifermin support were compared with those with palifermin support in identical patients.Recently, palifermin has been described to reduce the prevalence and severity of oral mucositis in patients with hematologic malignancies undergoing high-dose therapy with autologous peripheral blood stem-cell transplantation [17]. However, there are also standard chemotherapy regimens not requiring stem-cell support, which can be associated with severe mucositis. In a placebo-controlled phase II study published recently, it was demonstrated that palifermin is effective in reducing the incidence of oral mucositis in patients with metastatic colorectal cancer undergoing 5-fluorouracil/calcium folinate chemotherapy [27]. In contrast to these data, a recently published Cochrane review about ‘Interventions for preventing oral mucositis for patients with cancer receiving treatment’ by Worthington et al. [28] on the basis of three randomized trials [17, 29, 30] shows that there is no statistically significant difference between palifermin/repifermin and placebo. It was concluded that recombinant KGF could not be supported or refuted as more or less effective than placebo due to insufficient evidence. However, HD-MTX-based chemotherapy is well known to be associated with high mucositis rates and MTX-based chemotherapy regimen was not included in the Cochrane review [28]. This opens the possibility that there might be a stronger mucositis-preventive effect of palifermin in MTX-based chemotherapy.The cases reported here receiving HD-MTX provide evidence that palifermin might reduce the risk of oral mucositis and its severe side-effects such as infections. Our data lead to the suggestion that the effect is due to palifermin. In cycle A1, the incidence of mucositis could be increased by the application of the run-in therapy. Therefore, it cannot be completely excluded that the decrease in mucositis rate is partly caused by the absence of the run-in therapy in the subsequent cycles. This, however, is not the case in cycles B1 and B2, where treatment is identical. Furthermore, it seems theoretically possible that single patients may tolerate subsequent cycles of chemotherapy with a less severe mucositis than the first cycle. This, however, is not in line with our experience treating patients in the time before accessability of palifermin. With two exceptions (both obtaining the run-in treatment in the first cycles), every patient had a benefit with a reduction in mucositis grade and also shown by the amount of opioid analgetics given during the mucositis period. Having used palifermin for a while in our department, patients started to actively ask for off-label use of this specific drug for mucositis prevention and refused HD-MTX-containing chemotherapy without. Furthermore, our observations suggest the possibility of a dose reduction for palifermin which should be further studied.However, this is a report on a case series with retrospective evaluation and cannot replace prospective, controlled studies which should be carried out before the use of palifermin in HD-MTX regimen can be recommended as standard of care for mucositis prevention in HD-MTX chemotherapy.1.SonisSTMucositis as a biological process: a new hypothesis for the development of chemotherapy-induced stomatotoxicityOral Oncol19983439432.SonisSTThe pathobiology of mucositisNat Rev Cancer200442772843.SonisSTEltingLSKeefeDPerspectives on cancer therapy-induced mucosal injury: pathogenesis, measurement, epidemiology, and consequences for patientsCancer20041009 Suppl199520254.SonisSTThe biological role of nuclear factor—kappaB in disease and its potential involvement in mucosal injury associated with antineoplastic therapyCrit Rev Oral Biol Med2002133803895.ScullyCSonisSDizPDOral mucositisOral Dis2006122292416.RuescherTJSodeifiAScrivaniSJThe impact of mucositis on alpha-hemolytic streptococcal infection in patients undergoing autologous bone marrow transplantation for hematologic malignanciesCancer199882227522817.CostaSFMiceliMHAnaissieEJMucosa or skin as source of coagulase-negative staphylococcal bacteraemia?Lancet Infect Dis200442782868.SezerOEuckerJMetznerBMucositis is associated with increased rate of documented infections and treatment related mortality after high-dose therapy and autologous peripheral stem-cell transplantationProc Am Soc Clin Oncol20001956a(Abstr 216)9.SonisSTOsterGFuchsHOral mucositis and the clinical and economic outcomes of hematopoietic stem-cell transplantationJ Clin Oncol2001192201220510.EltingLSShihYTStiffPJEconomic impact of palifermin on the costs of hospitalization for autologous hematopoetic stem-cell transplant: analysis of phase 3 trial resultsBiol Blood Marrow Transplant20071380681311.FanningSRRybickiLKalaycioMSevere mucositis is associated with reduced survival after autologous stem cell transplantation for lymphoid malignanciesBr J Haematol200613537438112.KeefeDMSchubertMMEltingLSUpdated clinical practice guidelines for the prevention and treatment of mucositisCancer200710982083113.DumontetCSonnetABastionYPrevention of high-dose L-PAM-induced mucositis by cryotherapyBone Marrow Transplant19941449249414.EdelmanMJGandaraDPerezEAPhase I trial of edatrexate plus carboplatin in advanced solid tumors: amelioration of dose-limiting mucositis by ice chip cryotherapyInvest New Drugs199816697515.KuhrerIKuzmitsRLinkeschWTopical PGE2 enhances healing of chemotherapy-associated mucosal lesionsLancet1986162316.FarrellCLRexKLChenJNThe effects of keratinocyte growth factor in preclinical models of mucositisCell Prolif200235Suppl 1S78S8517.SpielbergerRStiffPBensingerWPalifermin for oral mucositis after intensive therapy for hematologic cancersN Eng J Med2004351252590259818.OelmannEHaghguSKulimovaEInfluence of keratinocyte growth factor on clonal growth of epithelial tumor cells, lymphoma and leukemia cells and on sensitivity of tumor cells towards 5-flourouracil in vitroInt J Oncol2004251001101219.KrijanovskiOLHillGRCookeKRKeratinocyte growth factor separates graft-versus-leukemia effects from graft-versus-host diseaseBlood19999482583120.EllisonCANatulikSAFischerJMEffect of recombinant human keratinocyte growth factor (rHuKGF) on the immunopathogenesis of intestinal graft-vs.-host disease induced without preconditioning regimenJ Clin Immunol20042419721121.BlijlevensNSonisSPalifermin (recombinant keratinocyte growth factor-1): a pleiotropic growth factor with multiple biological activities in preventing chemotherapy- and radiotherapy-induced mucositisAnn Oncol20071881782622.BraunSHanselmannCGassmannMGNrf2 transcription factor, a novel target of keratinocyte growth factor action which regulates gene expression and inflammation in the healing skin woundMol Cell Biol2002225492550523.MillerABHoogstratenBStaquetMReporting results of cancer treatmentCancer19814720721424.HoelzerDBaurKHGiagounidisAShort intensive chemotherapy with rituximab seems successful in Burkitt NHL, mature B-ALL and other high grade B-NHLBlood2003102(Abstr 236)25.ReiterASchrappeMZimmermannMA 4-hour i.v. infusion of methotrexate 5 g/sqm is not as efficacious for treatment of advanced B-cell neoplasms of childhood and adolescence than 24 hour i.v. infusion although less toxic. Interim results of trial NHL-BFM 95Blood200198(11 Part 1)26.HölzerDLudwigWDThielEImproved outcome in adult B-cell lymphoblastic leukemiaBlood19968749550827.RosenLSAbdiEDavisIDPalifermin reduces the incidence of oral mucositis in patients with metastatic colorectal cancer treated with fluorouracil-based chemotherapyJ Clin Oncol2006245194520028.WorthingtonHVClarksonJEEdenOBInterventions for preventing oral mucositis for patients with cancer receiving treatmentThe Cochrane Library2007429.FreytesCORatanatharathornVTaylorCPhase I/II randomized trial evaluating the safety and clinical effects of repifermin administered to reduce mucositis in patients undergoing autologous stem cell transplantationClin Cancer Res200410248318832430.MerepolNJSomerRAGutheilJRandomized phase I trial of recombinant human keratinocyte growth factor plus chemotherapy: potential role as mucosal protectantJ Clin Oncol200321814521458
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B3AAA7EFC7FFCAD2D500C2EE40EEC027617B835.txt b/test/dataset/in/resources/corpus/Clean_0B3AAA7EFC7FFCAD2D500C2EE40EEC027617B835.txt
deleted file mode 100644
index 44a8a22..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B3AAA7EFC7FFCAD2D500C2EE40EEC027617B835.txt
+++ /dev/null
@@ -1 +0,0 @@
- gerona J Gerontol A Biol Sci Med Scigerona The Journals of Gerontology Series A: Biological Sciences and Medical Sciences J Gerontol A Biol Sci Med Sci 1079-5006 1758-535X Oxford University Press 0012710.1093/gerona/56.1.M14 Journal of Gerontology: Medical Sciences Prevalence Rate of Urinary Incontinence in Community-Dwelling Elderly Individuals The Veneto Study Maggi Stefania a Minicuci Nadia a Langlois Jean b Pavan Mara a Enzi Giuliano a Crepaldi Gaetano a aCenter on Aging, National Research Council, Padova, Italy bNational Center for Injury Prevention and Control, Centers for Disease Control, Atlanta, Georgia 1 1 2001 56 1 M14 M18 26 6 2000 20 6 2000 The Gerontological Society of America 2001 Background. Urinary incontinence (UI) is a common problem in elderly people, due mainly to functional impairments and concurrent medical diseases. Few studies, however, have assessed the prevalence of UI in noninstitutionalized individuals. The objectives of the present work were to estimate the prevalence of UI in a community-based population of elderly Italians and to determine the associated physical, social, and psychological factors. Methods. A random sample of noninstitutionalized men (n = 867) and women (n = 1531), aged 65 years and older, from the Veneto region of northeastern Italy, were interviewed at home, using an extensive multidisciplinary questionnaire, to assess their quality of life and social, biological, and psychological correlates. Results. The prevalence rate of UI was of 11.2% among men and of 21.6% among women. Among those reporting the condition, approximately 53% of women and 59% of men reported experiencing incontinence daily or weekly. Association of UI was found for participants older than 70 years in both men (odds ratio [OR] 2.49, 95% confidence interval [CI] 1.45–4.28) and women (OR 1.49, 95% CI 1.11–2.02). Three of the medical conditions investigated were associated with increases in the odds in women, namely chronic obstructive pulmonary disease (OR 1.53, 95% CI 1.11–2.12), Parkinsonism (OR 2.27, 95% CI 1.14–4.54), and hip fracture (OR 1.38, 95% CI 1.02–1.88), whereas chronic diarrhea was the only condition associated with UI in men (OR 6.92, 95% CI 2.22–21.5). Participants with a physical disability were two times more likely to report incontinence, and the odds were increased by 50% in women who had sleep disturbances. Conclusions. Incontinence is highly prevalent in the Italian elderly population, and several common chronic conditions are significantly associated with it. Moreover, very few people with incontinence seek health care or are aware of potential treatments. hwp-legacy-fpage M14 hwp-legacy-dochead RESEARCH ARTICLE Decision Editor: John E. Morley, MB, BCh URINARY incontinence (UI) is a common problem in elderly people, due mainly to functional impairments and concurrent medical diseases (1)(2). Several studies have determined the prevalence of UI in nursing homes (3)(4) and in the community (5)(6). The reported prevalence rates approximate 50% for people in nursing homes and range between 2% and 55% for adults living in the community, depending on the definition of incontinence, the population characteristics, and the methodological approach (7). The burden of UI on elderly people is evidenced not only by the economic costs but also by other adverse effects on health and quality of life. UI has been estimated to account for 2% of health care costs in the United States and Sweden (8) and is one of the main reasons for the permanent institutionalization of elderly people (9). In Italy, cost estimates are available for patients with dementia and UI: An average of 1000 US$ is spent each year for pads alone, the most commonly used remedy for incontinence in elderly persons (10). The physical complications may be serious and include an increased risk of urinary tract infections, pressure sores and other skin problems, and sleep disturbance. People with UI also report significant and often long-lasting social and psychological effects such as decreased social interaction and greater feelings of depression and isolation (11)(12)(13)(14)(15)(16)(17). The purpose of this cross-sectional study was to estimate the prevalence of UI in a community-dwelling population of elderly Italians and to determine the associated physical, social, and psychological factors. Methods Study Population The details of the methods have been described elsewhere (18). Briefly, the population for the Veneto Study consisted of a random sample of 2700 noninstitutionalized individuals aged 65 years and older, residing in the community on May 1, 1989 in five rural and four urban defined geographic areas of the Veneto region of northeastern Italy. Names and addresses of eligible individuals were obtained from the resident lists maintained by the municipalities, in which all births and deaths are registered within one week of their occurrence. A random sample from each of five age strata (65–69, 70–74, 75–79, 80–84, and 85+ years) was taken, with an over-sampling of those aged 85 years and older to obtain 20% of the total sample in this age category. Approximately 8.4% of the total elderly population of the Veneto Region was aged 85 years and older in 1989 (19). Eighty-nine percent (n = 2402) of people identified as eligible participated in the study. The 298 nonrespondents included those who refused to participate, those who were not found at home after three attempts on different days, or those for whom a proxy refused their participation due to severe dementia or terminal illness (approximately 2%). Only basic demographic data for nonrespondents were collected from a proxy; all data presented herein were provided by the respondents themselves. After exclusion of four people with missing or incorrect demographic data, the final number available for analysis was 2398. Assessment of Participants All participants were administered a comprehensive interview and a brief physical examination in the home. The interview obtained information on sociodemographic characteristics, living arrangements, family composition, social support, income (including family contributions), participation in social activities, self-reported history of medical conditions, health status, physical functioning (including activities of daily living [ADLs] (20) and instrumental activities of daily living) (21), health behaviors, self-rated health, and use of health and social services. Low mental status score was defined as a score of <0.8 for the ratio of the number of questions answered correctly to the number of possible answers (i.e., less than 24 correct answers out of 30, if all questions were applicable), or an adjusted score (number of questions answered correctly divided by the number answered, if some of the questions were not applicable) on the Mini-Mental State Examination (22). Participants whose scores on the Center for Epidemiologic Studies–Depression scale (CES-D) (23) were in the upper quintile (CES-D score >8.8 for men and >24.5 for women) were categorized as having depressive symptomatology. The interviewers were physicians enrolled in a postgraduate geriatric medicine program, trained to administer performance measures in a standardized manner and certified by a scientific training committee. The examination included assessments of height, weight, vision, hearing, and a qualitative assessment of physical performance. Assessment of Urinary Incontinence Participants were asked if they had UI problems and how often they occurred: rarely, 1 to 2 times per month, at least once per week, or every day. Participants were asked whether the incontinence occurred only under strain (i.e., during coughing or sneezing), if they used any remedies, and if they were aware that there are remedies. Health Status Indicators and Medical Conditions Participants were classified as having a disability in the basic ADLs if they reported that they needed help from another person or that they were unable to perform one or more of the following activities: bathing, dressing, getting out of bed, or eating. Mobility disability was classified as “difficult” or “impossible” to climb stairs, to walk 800 m without resting, or as the need for help from another person or the inability to walk across a room. Participants were asked how often they (i) have trouble falling asleep, (ii) have trouble with waking up during the night, (iii) have trouble with waking up too early and not being able to fall asleep again, and (iv) feel really rested when they wake up in the morning. Night awakeners were defined as those who reported often or always waking up during the night. History of arthritis, diabetes, hypertension, chronic respiratory disease, stroke, heart disease, Parkinsonism, and chronic diarrhea was ascertained by asking participants if a doctor had ever told them that they had the condition. Data Analysis To account for the over-sampling of persons aged 85 years and older, the prevalence of urine incontinence was calculated using weights determined by the ratio between the population fraction relative to the 1991 census population and the sampling fraction relative to the age stratum. All statistical analyses were conducted using the SAS statistical analysis package (SAS, Inc, Cary, NC) (24). The association of UI with the sociodemographic, behavioral, health, and other factors was assessed in logistic regression models with incontinence as the dependent variable. Results Table 1 shows the distribution of our sample by the main demographic and health-related characteristics. The final sample consisted of 867 men (mean age 75.2 ± 7.1 years) and 1531 women (mean age 76.9 ± 7.6 years). Women had a significantly lower education level and higher prevalence rates of mental and physical health disorders and obesity than males. A greater percentage of women reported fair or poor self-rated health and used sleep medications more often than men. Conversely, more men than women reported drinking alcohol and cigarette smoking. Data regarding the prevalence rates and the frequency of UI are presented in Table 2 and Table 3 . The prevalence rate was 11.5% among men and ranged from 4.6% for those aged 65 to 69 years to 23.6% for participants aged 85 years and older. In women, the prevalence rate was of 21.6%, ranging from 16.4% to 34.7% for the same age groups. Among the 415 participants with this condition, 36.8% reported that they experienced incontinence every day, and 17.7% reported that they experienced incontinence every week. Approximately 59% of men and 53% of women reported that they experienced UI frequently (every week or day). Although the percentages showed a trend with age, the association was not statistically significant (p = .06). Nevertheless, more than 50% of the affected people aged 65 to 69 years faced this problem frequently versus 67% of subjects aged 85 years and older. Approximately 58% of women and 28.5% of men (data not shown) stated that they experienced incontinence only under stress. Table 4 shows the ORs for the association of all sociodemographic and health-related factors with incontinence by sex. After adjustment for other factors, age 70 years and older, ADL disability, and mobility disability were significantly associated with incontinence in both elderly men and women. Diarrhea showed the strongest association in men but was not significant in women. Chronic obstructive pulmonary disease (COPD), Parkinsonism, hip fracture, and night awakening were significant in women but not in men. Self-rated health was borderline significant in men only. Discussion UI is a highly prevalent condition in elderly persons and deeply affects quality of life because of its negative effects mainly on social interactions, but also on physical health. The prevalence rates of UI in our study are high, similar to those reported in previous studies (25)(26)(27), but seem to be higher compared with those reported in a recent study conducted in Italy (28). However, the different methodologies could easily explain the discrepancies. The study by Bortolotti and colleagues (28) was carried out using a telephone interview, and we believe that it might exclude sicker people and therefore those at higher risk for incontinence and other chronic conditions. Moreover, our survey was carried out by physicians, who left the most sensitive questions, such as those related to incontinence, toward the end of the personal interview, when a more relaxed interaction with the participants was established, which would enhance the reliability of their answers. We confirmed a higher prevalence of UI among women than men and noted that it represents a severe problem for most of them, given that about half of those reporting incontinence experience it every day or more than once a week. The increasing prevalence with age has been also found in other studies (7), and this could be associated with age-related conditions, such as prostatic hyperplasia among men (29). Data on the presence of diseases were self-reported by the participants, and therefore we did not ask information about the prevalence of prostatic hyperplasia because of the high probability of underreporting for that condition. Age, parity, and menopausal status with the consequent incompetence of the internal urethral sphincter could be determinants of stress incontinence among elderly women (30), and, indeed, its frequency was twice as much as in men. Physical disability was a positive risk factor for functional incontinence in this population, and problems in mobility could easily explain this association. We also found a strong association with Parkinsonism and hip fracture, easily explained both in terms of the neurological impairments and mobility problems of these patients (31). The higher risk of incontinence in women with COPD could be due to the effect of increased abdominal pressure when coughing and the consequent stress incontinence, as already described in previous reports (32). Night awakening was associated with incontinence in women and not in men. This was surprising, given that prostatic problems and need to void urine usually lead to frequent night awakening and to nocturia among elderly men. It is possible that the higher prevalence of physical disability among women was responsible for their functional incontinence. A recent study (5) has underlined that few community-dwelling elderly persons seek health care for UI in spite of its high prevalence in this group. Given that symptomatic improvement or a cure would be possible in many cases, a better evaluation of this condition is needed. This would require increasing the sensitivity and knowledge of the population about these problems and about potential medical and behavioral interventions. It is certain that most people, particularly women, tend to use pads as the only solution to the problem, ignoring completely that other more appropriate and definite approaches could be very effective. Of course, treatment of incontinence requires attention to the general health status of an individual (1)(33). Treatment of urinary infections or estrogen deficiency or prostatic problems should be the first step when these conditions are present. Pelvic floor exercise could be very effective for patients with stress incontinence, while behavioral interventions, such as decreasing fluid intake, particularly caffeine-rich drinks, could be general measures applied to all patients (34)(35). This study has some limitations. First, the use of cross-sectional data does not allow for the assessment of the cause-and-effect relationship between incontinence and health conditions. Second, information on medical conditions and physical function was self-reported. However, several studies support the reliability of information on health status reported by the elderly population (36)(37). Moreover, the prevalence of diseases investigated was comparable to that reported in other studies (38), and the association with incontinence was consistent with previously reported findings in different populations (1)(2)(3)(4)(5)(6). This study has some peculiar strengths, such as the high response rate and the representativeness of the sample, that allow the generalizability of our results to the elderly population of Italy. In conclusion, incontinence is probably more frequent than generally known among community-dwelling individuals and has a negative impact on their quality of life and general health status. Given that symptomatic improvement or treatment is possible in many cases, physicians should always ask their elderly patients about incontinence and then provide appropriate evaluation and management of it. Table 1. Sociodemographic and Health-Related Characteristics by Sex (Veneto Study, 1989) p Value Men (n = 867) Women (n = 1531) Difference Between Men and Women Age Trend for Men Age Trend for Women Demographics Mean Age, y (SD) 75.2 (7.1) 76.9 (7.6) <.001 Marital Status (%) <.001 <.001 <.001 Married 78.1 32.0 Not married 5.9 12.5 Widowed 15.1 54.1 Divorced 0.9 1.4 Education (%) <.001 .156 <.001 ≤5 y 70.6 82.5 6–8 y 14.8 10.7 9–13 y 9.7 5.7 University 4.9 1.1 Mental Health (%) <.001 <.001 <.001 Low mental status\|[dagger]\| 18.8 27.1 Depressive symptomatology (%) <.001 .522 .028 Low CES-D score\|[Dagger]\| 8.8 24.5 Physical Health (%) Disability status Mobility disability\|[sect ]\| 28.5 41.5 <.001 <.001 <.001 ADL disability\|[Verbar]\| 12.4 17.6 <.001 <.001 <.001 Body mass index\|[para ]\| <.001 <.001 <.001 ≤20 4.1 4.8 >20 and <26 46.1 38.9 ≥26 and <31 37.8 37.4 ≥31 and <35 10.5 15.0 ≥35 1.5 3.9 Self-rated health <.001 .017 0.837 Excellent/good 62.6 50.3 Fair/poor 37.4 49.7 Behavioral Factors (%) Use of sleep medications <.001 .219 .702 Sometimes 6.6 11.3 Often/always 13.9 24.2 Alcohol use <.001 .157 .412 Abstainers 3.6 16.2 Past users 10.5 15.2 Current users 86.0 68.5 Cigarette smoking <.001 <.001 <.001 Never smokers 17.6 81.1 Former smokers 57.9 10.9 Current smokers 24.4 8.0 Note: CES-D = Center for Epidemiologic Studies–Depression Scale; ADL = activities of daily living. \|[dagger]\| Mini Mental State Examination score ratio ≤0.8. \|[Dagger]\| CES-D score >80th percentile. \|[sect ]\| Reported climbing stairs or walking 800 m without resting as difficult or impossible. \|[Verbar]\| Reported needing help from another person or unable to perform one or more of the following activities: bathing, dressing, getting out of bed, eating, or walking across a room. \|[para ]\| Weight (kg)/height (m)2. Table 2. Prevalence of Urinary Incontinence by Age and Sex 65–69 y 70–74 y 75–79 y 80–84 y ≥85 y Total Men 4.6 12.6 12.3 22.2 23.6 11.5 Women 16.4 17.8 24.8 23.9 34.7 21.6 Note: Data are given as percentages. Table 3. Frequency of Urinary Incontinence by Age and Sex 65–69 y 70–74 y 75–79 y 80–84 y ≥85 y Total Men Women Rarely 34.6 44.6 24.7 33.3 20.0 31.6 26.5 33.3 1 to 2 times a mo 13.0 10.8 18.0 14.1 12.6 13.9 14.5 13.7 Every week 20.2 13.4 22.9 15.4 14.1 17.7 13.6 19.0 Every day 32.2 31.2 34.4 37.2 53.3 36.8 45.4 34.0 Note: Data are given as percentages. Table 4. Odds Ratios for the Association of Sociodemographic and Health-Related Factors With Incontinence, by Sex Factor Odds Ratio 95% CI p Value Men Age ≥70 y 2.49 1.45–4.28 .0009 Diarrhea 6.92 2.22–21.50 .0008 Comorbidity 1.52 0.89–2.61 .12 ADL disability 2.31 1.23–4.32 .009 Mobility disability 2.11 1.16–3.85 .01 Self-rated health 1.69 0.997–2.88 .05 Women Age ≥70 y 1.49 1.11–2.02 .008 COPD 1.53 1.11–2.12 .01 Parkinsonism 2.27 1.14–4.54 .02 Hip fracture 1.38 1.02–1.88 .04 ADL disability 1.75 1.22–2.52 .002 Mobility disability 1.81 1.32–2.49 .0003 Night awakening 1.48 1.11–1.97 .007 Note: CI = confidence interval; ADL = activities of daily living; COPD = chronic obstructive pulmonary disease. Address correspondence to Stefania Maggi, MD, MPH, Center on Aging, National Research Council, Clinica Medica 1°, University of Padua, Via Giustiniani, 2, 35128 Padova, Italy. 1 Sirls LT, Rashid T, 1999. Geriatric urinary incontinence. Geriatr Nephrol Urol9:87-99. 2 Koyama W, Koyanagi A, Mihara S, et al. 1998. Prevalence and conditions of urinary incontinence among the elderly. Methods Inf Med37:151-155. 3 Brandeis GH, Baumann MM, Hossain M, et al. 1997. The prevalence of potentially remediable urinary incontinence in frail older people: a study using the Minimum Data Set. J Am Geriatr Soc45:179-184. 4 Ouslander JG, Kane RL, Abrass IB, 1982. Urinary incontinence in the elderly nursing home patients. JAMA248:1194-1198. 5 Roberts RO, Jacobsen SJ, Rhodes T, et al. 1998. Urinary incontinence in a community-based cohort: prevalence and healthcare-seeking. J Am Geriatr Soc46:467-472. 6 Herzog AR, Fultz NH, 1990. Prevalence and incidence of urinary incontinence in community-dwelling population. J Am Geriatr Soc38:273-281. 7 Thom D, 1998. Variation in estimates of urinary incontinence prevalence in the community: effects of differences in definition, population characteristics, and study type. J Am Geriatr Soc46:473-480. 8 Hu TW, 1990. Impact of urinary incontinence on health care costs. J Am Geriatr Soc38:292-295. 9 Baker DI, Bice TW, 1995. The influence of urinary incontinence on publicly financed home care services to low-income elderly people. The Gerontologist35:360-369. 10 Levorato A, Rozzini R, Trabucchi M. I costi della vecchiaia. Ed. Bologna, Italy: Il Mulino;1994. 11 Dugan E, Cohen SJ, Bland DR, et al. 2000. The association of depressive symptoms and urinary incontinence among older adults. J Am Geriatr Soc48:413-416. 12 Naughton MJ, Wyman JF, 1997. Quality of life in geriatric patients with lower urinary tract dysfunction. Am J Med Sci314:219-227. 13 Ouslander JG, 1997. Aging and the lower urinary tract. Am J Med Sci314:214-218. 14 Chutka DS, Fleming KC, Evans MP, Evans JM, Andrews KL, 1996. Urinary incontinence in the elderly population. Mayo Clin Proc71:93-101. 15 Busby Whitehead JM, Johnson TM, 1998. Urinary incontinence. Clin Geriatr Med14:285-296. 16 Fiers SA, 1996. Breaking the cycle: the etiology of incontinence dermatitis and evaluating and using skin care products. Ostomy-Wound-Manage42:32-34. 17 Wyman JF, Harkins SW, Fantl JA, 1990. Psychosocial impact of urinary incontinence in the community-dwelling population. J Am Geriatr Soc38:282-288. 18 Maggi S, Bush TL, Enzi G, Crepaldi G, 1991. Quality of life among the elderly in Veneto, Italy: a cross-national study. Vital Health Stat 56:211-214. 19 National Institute of Statistics. Le regioni in cifre. Rome, Italy: National Institute of Statistics;1990:39. 20 Katz SC, Ford AB, Moskowitz RW, Jackson BA, Jaffe MW, 1963. Studies of illness in the aged. The Index of ADL: a standardized measure of biological and psychosocial function. JAMA185:914-919. 21 Lawton MP, Brody EM, 1982. Assessment of older people: self-maintaining and instrumental activities of daily living. J Gerontol37:91-99. 22 Folstein MF, Folstein SE, McHugh PR, 1975. “Mini-Mental State”: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res12:189-198. 23 Radloff LS, 1977. The CES-D Scale: a self-report depression scale for research in the general population. Appl Psych Measur1:385-401. 24 SAS Institute1989. SAS/STAT User's Guide, Version 6 4th ed. SAS Institute, Inc, Cary, NC. 25 Temml C, Haidinger G, Schmidbauer J, Schatzl G, Madersbacher S, 2000. Urinary incontinence in both sexes: prevalence rates and impact on quality of life and sexual life. Neurourol Urodyn19:259-271. 26 Gavira Iglesias FJ, Caridad y Ocerin JM, Perez del Molino Martin J, et al. 2000. Prevalence and psychosocial impact of urinary incontinence in older people of a Spanish rural population. J Gerontol Med Sci55A:M207-M214. 27 Ushiroyama T, Ikeda A, Ueki M, 1999. Prevalence, incidence, and awareness in the treatment of menopausal urinary incontinence. Maturitas33:127-132. 28 Bortolotti A, Bernardini B, Colli E, et al. 2000. Prevalence and risk factors for urinary incontinence in Italy. Eur Urol37:30-35. 29 Johnson TM, Ouslander JG, 1999. Urinary incontinence in the older man. Med Clin North Am83:1247-1266. 30 Karram MM, Partoll L, Bilotta V, Angel O, 1997. Factors affecting detrusor contraction strength during voiding in women. Obstet Gynecol90:723-726. 31 Nakayama H, Jorgensen HS, Pedersen PM, Raaschou HO, Olsen TS, 1997. Prevalence and risk factors of incontinence after stroke: The Copenhagen Study. Stroke28:58-62. 32 DuBeau CE, 1996. Interpreting the effect of common medical conditions on voiding dysfunction in the elderly. Urol Clin North Am23:11-18. 33 Milsom I, 1996. Rational prescribing for postmenopausal urogenital complaints. Drugs Aging9:78-86. 34 Iqbal P, Castleden CM, 1997. Management of urinary incontinence in the elderly. Gerontology43:151-157. 35 Gallo ML, Fallon PJ, Staski DR, 1997. Urinary incontinence: steps to evaluation, diagnosis, and treatment. Nurse Pract22:21-24. 36 Bush TL, Miller SR, Golden A, 1989. Self-report and medical report of selected chronic conditions in the elderly. Am J Public Health79:1554-1556. 37 Kehoe R, Wu SY, Leske MC, Chylack LT, 1994. Comparing self-reported and physician-reported medical history. Am J Epidemiol139:813-818. 38 National Institute of Statistics. Indagine multiscopo sulle famiglie, IV e V ciclo. Roma, Italy;1994.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B3B0444849AA7F246C15872661128DBD7584D0E.txt b/test/dataset/in/resources/corpus/Clean_0B3B0444849AA7F246C15872661128DBD7584D0E.txt
deleted file mode 100644
index 67c0685..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B3B0444849AA7F246C15872661128DBD7584D0E.txt
+++ /dev/null
@@ -1 +0,0 @@
-]>HEARES3362S0378-5955(99)00194-X10.1016/S0378-5955(99)00194-XElsevier Science B.V.Fig. 1Example of intracellular calcium concentration changes in a single type I sensory cell induced by iontophoretic applications (arrows) of glutamate (Glu), glycine (Gly), NMDA, aspartate (Asp) and NaCl. Significant, rapid and transient increases in the concentration of intracellular calcium were observed for Glu, Gly, NMDA and Asp. No significant change was induced by NaCl.Fig. 2Effect of 7-chlorokynurenate (7CK) on the calcium response to glycine. The presence of 7CK (10 μM) reduced the calcium response induced by glycine application by about 60% (A and B). APV (10 μM), a competitive antagonist of the NMDA receptor, did not affect the calcium response induced by glycine (B). The addition of 7CK (10 μM) and strychnine (50 μM) together resulted in a larger decrease in the calcium response to glycine stimulation (A).Fig. 3Effect of strychnine, a specific antagonist of the glycine Cl−- channel receptor. The presence of strychnine (50 μM) decreased the calcium response to glycine stimulation by about 60%. Strychnine caused no significant decrease in the calcium response induced by glutamate.Fig. 4Effect of Cl−-free gluconate medium. The presence of gluconate (137 mM) decreased the calcium response induced by glycine. In the same condition, there was no significant decrease in the calcium response induced by glutamate application. The addition of gluconate and 7-chlorokynurenate (7CK, 10 μM) together resulted in a larger decrease in the calcium response induced by glycine stimulation.Fig. 5Effect of chelating Ca2+ in the medium by the addition of EGTA. The calcium response induced by glycine was reduced by adding EGTA (1.5 mM) to the medium. When the cells returned to normal medium, the calcium peak level tended toward the typical calcium levels induced by glycine application.Fig. 6Schematic drawing, representing the putative presence of NMDA and glycine receptors on vestibular type I sensory cells and the hypothesis of a peripheral control of the sensory cell activity by the afferent nerve calyx. An activation of these receptors by glycine could induce changes in [Ca2+]i. The increase in [Ca2+]i could also be due to the activation of ionotropic and metabotropic glutamate receptors after release of glutamate.Table 1Analysis of calcium responses to Glu, Gly, NMDA, Asp and NaCl applications in isolated type I sensory cellsGluGlyNMDAAspNaClHair cells(n)2026557Baselineratio 340/3800.77±0.100.77±0.090.76±0.030.78±0.030.76±0.05Peakratio 340/3801.2±0.231.19±0.221.02±0.131.15±0.070.76±0.06Time to peak(s)7.36±1.219.15±2.347.00±1.418.20±1.64Regulation time(s)19.09±3.5920.58±5.0415.08±3.8719.40±3.36Increase/base(%)565534470[Ca2+]i are means of the fluorescence signal of the whole cell soma. The baseline [Ca2+]i values (mean±S.D.) and the response peak values (mean±S.D.) are expressed as the 340 nm/380 nm fluorescence ratio. The time course of the increase [Ca2+]i was rapid, and then returned to basal levels. The estimated time-to-peak and regulation time are given in seconds (mean±S.D.). The nuber of type I sensory cells (n) tested for each agent is indicated.Table 2Calcium levels following glycine stimulation in the presence of 7-chlorokynurenate (7CK, 10 μM), a selective antagonist of the glycine site on NMDA receptor and of APV (10 μM), a competitive antagonist for the glutamate site on NMDA receptorGluGlu+APVGlyGly+APVGly+7CKNMDANMDA+APVHair cells(n)7171711Baselineratio 340/3800.71±0.100.70.71±0.100.70.71±0.100.70.7Peakratio 340/3801.16±0.110.951.09±0.161.280.87±0.101.030.75Time to peak(s)8±157.14±1.0776.83±1.1755Regulation time(s)17.67±3.061019.71±2.431917.67±3.14165Increase/base(%)6333548022455Decrease/specific(%)4806090The [Ca2+]i baseline values (mean±S.D.) and peak response values (mean±S.D.) are expressed as the 340 nm/380 nm fluorescence ratio. The values are expressed as in Table 1. The percentage decrease was evaluated as the ratio of the amplitude of the increase in [Ca2+]i in the presence of the antagonist to that in the presence of glycine alone.Table 3[Ca2+]i changes following glutamate and glycine stimulations in the presence of strychnine (50 μM)GluGlu+strychn.GlyGly+strychn.Gly+strychn.+7CKHair cells(n)62664Baselineratio 340/3800.65±0.060.63±0.030.65±0.060.65±0.060.66±0.06Peakratio 340/3800.98±0.080.96±0.060.94±0.110.76±0.090.72±0.04Time to peak(s)6.33±1.157.5±0.716.83±0.756.83±1.946.25±0.83Regulation time(s)19.33±3.0621±1.4116.17±2.3217±3.5817.25±2.38Increase/base(%)5148451812Decrease/specific(%)66073The values are expressed as in Table 1. The percentage decrease in the calcium response is calculated as the ratio of the peak calcium level in the presence of strychnine to that following glutamate or glycine stimulation alone. The addition of both strychnine (50 μM) and 7CK (10 μM) resulted in a larger decrease, by about 73%, in the calcium response induced by glycine application.Table 4[Ca2+]i changes in Cl−-free gluconate mediumGluGlu+glucoGlyGly+glucoGly+gluco+7CKGly after rinseHair cells(n)336613Baselineratio 340/3800.71±0.060.71±0.060.71±0.060.71±0.060.60.68±0.03Peakratio 340/3801.17±0.100.96±0.251.31±0.160.96±0.140.81.23±0.16Time to peak(s)8±07.33±1.159.5±2.178±1.2677.67±2.08Regulation time(s)21.33±3.2118±4.5823.33±4.520.17±6.052622.33±3.21Increase/base(%)643484351373Decrease/specific(%)19588413Glutamate or glycine were applied in normal or Cl−-free gluconate medium. The values are expressed as in Table 1. After rinsing with normal medium, the peak calcium level tended to be similar to that in control after glycine stimulation. The concomitant absence of Cl− and presence of 7CK resulted in a larger decrease in the calcium response to glycine than the one obtained in the absence of Cl− alone.Table 5[Ca2+]i after glycine application in the presence of EGTA (1.5 mM), a specific calcium chelator, and after rinsing with normal mediumGlyGly+EGTAGly after rinseHair cells(n)444Baselineratio 340/3800.67±0.090.54±0.040.67±0.09Peakratio 340/3800.99±0.280.57±0.180.81±0.12Time to peak(s)7.25±0.965.50±0.716.50±2.12Regulation time(s)17.25±2.0619±1.4117.50±0.71Increase/base(%)47621The values are expressed as in Table 1.Glycine induced calcium concentration changes in vestibular type I sensory cellsGinaDevau*gdevau@univ.montp2.frINSERM U432, Université Montpellier II, place Eugène Bataillon, 34095 Montpellier Cedex 5, France*Tel.: +33 4 67 14 48 30; Fax: +33 4 67 14 36 96AbstractGlutamate is the neurotransmitter of the synapse between vestibular type I hair cells and the afferent nerve calyx. This calyx may also be involved in local feedback, which may modify sensory cell activity via N-methyl-D-aspartate (NMDA) receptors. Glycine is the co-agonist of glutamate in NMDA receptor activation. Both agents have been detected by immunocytochemistry in the nerve calyx. Glutamate and NMDA stimulations cause changes in the intracellular calcium concentration ([Ca2+]i) of isolated type I sensory cells. We investigated the effect of glycine stimulation on [Ca2+]i in guinea pig type I sensory cells by spectrofluorimetry with fura-2. Glycine application to isolated type I sensory cells induced a rapid and transient increase in [Ca2+]i. The fluorescence ratio increased by 55% above the resting level. The peak was reached in 9 s and the return to basal level took about 20 s. A specific antagonist of the glycine site on NMDA receptors, 7-chlorokynurenate (10 μM), decreased the calcium response to glycine by 60%. Glycine may activate NMDA receptors. Glycine may also activate the strychnine-sensitive glycine receptor-gated channel. Strychnine (50 μM) decreased the calcium response to glycine by 60%. Thus, glycine probably induces calcium concentration changes in type I vestibular sensory cells via NMDA receptors and/or glycine receptors.KeywordsN-Methyl-D-aspartateGlycineCalciumVestibular sensory cellGuinea pig1IntroductionIn mammals, vestibular sensory cells are mechano-sensory cells, classified as type I or type II according to morphological criteria (Wersäll, 1956; Wersäll and Bagger-Sjöback, 1974). Type I sensory cells are surrounded by an afferent nerve calyx. Type II sensory cells are contacted only by button-shaped afferent endings. Physiological and pharmacological studies have shown that glutamate is the principal excitatory neurotransmitter at the cytoneuronal synapse between sensory cells and primary afferent nerve fibers, in vestibular hair cells (Annoni et al., 1984; Drescher et al., 1987; Drescher and Drescher, 1992; Guth et al., 1988; Soto and Vega, 1988; Zucca et al., 1992) and in cochlear hair cells (Anson and Ashmore, 1994; Kataoka and Ohmori, 1996). Different glutamate receptor types are involved in vestibular neurotransmission between the sensory cells and the primary afferent neurons. Non-N-methyl-D-aspartate (NMDA) receptors have been detected mostly by electrophysiological and pharmacological studies in the vestibular system of batracians (Soto and Vega, 1988; Prigioni et al., 1990, 1994), although NMDA receptors have also been shown to be present (Zucca et al., 1993; Soto et al., 1994). In mammals, immunocytochemical and in situ hybridization studies have implicated the GluR2/R3 and GluR4 subunits of AMPA receptors (Demêmes et al., 1995; Niedzielski and Wenthold, 1995; Rabejac et al., 1997), and the NR1 and NR2A–D subunits of NMDA receptors in vestibular neurotransmission (Fujita et al., 1994; Niedzielski and Wenthold, 1995). Moreover, presynaptic glutamate receptors may also interact as autoreceptors (Valli et al., 1985; Prigioni et al., 1990; Guth et al., 1991, 1998b). The presence of NMDA receptors has been detected on isolated type I sensory cells from guinea pig by calcium spectrofluorimetry with fura-2 (Devau et al., 1993). Vestibular hair cell neurotransmission has been reviewed by Guth et al. (1998a).Type I sensory cells have the particularity that they are surrounded by an afferent nerve calyx. Various studies support the hypothesis of a local control of type I sensory cell activity by this afferent nerve calyx (Scarfone et al., 1988, 1991). In particular, the calyx contains synaptic-like microvesicles and proteins involved in the synaptic vesicle cycle (synaptophysin, synapsin I and rab3a, Dechesne et al., 1997) suggesting that neuroactive substances may be released. This local control could involve glutamate (Demêmes et al., 1990) and/or a neuropeptide, substance P (Scarfone et al., 1996).The NMDA receptors on isolated type I sensory cells may act as presynaptic receptors in the active zone of neurotransmitter release or may be involved in the local control by the afferent nerve calyx. Glutamate and glycine are the co-agonists that activate NMDA receptors (Johnson and Ascher, 1987; Langosch et al., 1988; Corsi et al., 1996). Glycine and glutamate have both been detected by immunochemistry, co-localized in the thick calyceal afferent fibers of the vestibular epithelium (Reichenberger and Dieringer, 1994; Straka et al., 1996a,b; Bäurle et al., 1997). Glycine may have a modulatory effect: possibly co-released with glutamate from the nerve calyx, glycine may activate NMDA receptors or act independently on putative glycine receptors.To test whether there was local control of sensory cells involving the afferent nerve calyx, we investigated the action of glycine on isolated type I sensory cells of guinea pig by measuring variations in intracellular calcium concentration ([Ca2+]i) by spectrofluorimetry with the sensitive dye fura-2. We show that glycine can act on at least two different sites: the glycine site of the NMDA receptor selectively antagonized by 7-chlorokynurenic acid (Kemp et al., 1988; Kemp and Leeson, 1993), and the glycine receptor, which is strychnine-sensitive and highly permeable to chloride (Van den Pol and Gorcs, 1988; Vandenberg et al., 1992; Zafra et al., 1997; Breitinger and Becker, 1998).2Materials and methods2.1Vestibular type I sensory cell preparationIsolated vestibular sensory cells were prepared as previously described (Devau et al., 1993). Briefly, young pigmented guinea pigs (200–250 g) were anesthetized and then decapitated. The labyrinths were removed, opened and the cristae ampullaris quickly dissected in Hanks’ balanced salt solution (HBSS, Sigma, France). This medium contained 137 mM NaCl, 5.4 mM KCl, 1.7 mM Ca2+, 0.4 mM MgSO4, 0.31 mM Na2HPO4, 0.46 mM KH2PO4 and 5.5 mM D-glucose (Sigma, France). HBSS solution was buffered to pH 7.4 with 5 mM HEPES and adjusted to 290 mOsm. Hair cells were dissociated enzymatically by incubating the sensory epithelium in HBSS containing collagenase (1 mg/ml, Sigma, France) for 10 min at room temperature (20°C). After three rinses in HBSS, hair cells were mechanically dissociated by gentle trituration with fine iron microelectrodes. The isolated hair cells were then transferred onto slides coated with poly-L-lysine (0.5 mg/ml, Sigma, France).Type I hair cells were identified by their characteristic amphora shape, which was preserved after isolation. An elongated neck region separated the spherical basal region (containing the nucleus) from the cuticular plate, into which the intact hair bundle was inserted. The type I sensory cells were selected on the following criteria: smooth plasmalemma and no granules within the cytoplasm. This morphological examination also verified that there was no shape abnormality, for example turgescent swelling or plasmolytic shrinking due to osmotic pressure imbalance. This preparation was used for a patch clamp study (Griguer et al., 1993).The care and the use of animals in this study were approved by the French Ministère de l’Agriculture et de la Forêt (authorization number 04889).2.2Calcium measurements[Ca2+]i was measured using the fluorescent dye fura-2. The cells were loaded with 2.5 μM fura-2/AM (Molecular Probes, Eugene, OR, USA) in HBSS containing 0.02% pluronic-DMSO for 30 min at room temperature. Experiments were performed with an inverted microscope (Axiovert 10, Zeiss, Le Pecq, France) equipped with a fast fluorescence photometer controlled by MSP 69/AIS. Excitation light was provided by a xenon lamp (XBO 75 W) and was passed through filters to select the two excitation wavelengths, 340 and 380 nm. Light at both excitation wavelengths was dimmed by about 90% by a neutral density filter (green UG11) to protect the cells from phototoxicity and bleaching. The emission signals were passed through a 510 nm narrow bandpass filter. A circular 10 μm diaphragm was used to limit fluorescence measurement to a field containing a single sensory cell and centered on the soma. Each fluorescence ratio value was determined by an 80 ms time resolution of the measurement. Changes in fluorescence ratio (free dye 340 nm/bound dye 380 nm) were used to estimate changes in free cytosolic calcium concentration as previously described (Grynkiewicz et al., 1985).2.3Pharmacological applications2.3.1Iontophoretic application of glutamate, NMDA, aspartate and glycineIsolated sensory cells were stimulated by focal iontophoretic application of various agents. Glutamate, NMDA, aspartate and glycine were applied via a micropipette with the tip situated about 2 μm from the basolateral hair cell membrane. Microelectrodes were filled with one of the following solutions: glutamate (1 M, pH 8), NMDA (0.4 M, pH 8), aspartate (1 M, pH 8) and glycine (1 M, pH 8) (Sigma, France). The tip diameter of the micropipettes was less than 1 μm and their resistance when they were filled with any of the various solutions was about 25 MΩ. Electrolytes were ejected using a negative pulse (−3 μA) lasting 5 s. Breaking currents of opposite polarity were used to prevent cation leakage from iontophoretic electrodes. Control tests of the iontophoretic procedure were performed by ejection of Cl− using a negative current or of Na+ using a positive current from microelectrodes filled with sodium chloride (NaCl, 1 M).The electric charge ejected through the pipette was estimated from the Faraday equation Q=It/F where I is the amplitude of the current applied, t the duration of the application and F the Faraday constant (96 500 C). After ejection, the diffusion of the agent depends on several factors including the viscosity of the medium and the size of the molecule.2.3.2Microperfusion of glutamate and glycine antagonistsAntagonist microperfusion was started 30 s before stimulation with glutamate or glycine. Antagonists were dissolved in HBSS at a concentration of 50 μM for strychnine (Sigma, France) and 10 μM for 7-chlorokynurenate (Sigma, France) and DL-2-amino-5-phosphonovaleric acid (APV, Sigma, France). Microperfusion of antagonist alone did not affect the resting calcium concentration.2.4Data analysisCells loaded with fura-2 responded to the application of glutamate agonists or glycine by an increase in fluorescence intensity ratio corresponding to an increase in [Ca2+]i. The fluorescence ratio is expressed in terms of molarity using the equation of Grynkiewicz et al. (1985): [Ca2+]i=(R−Rmin/Rmax−R)αKD, where Rmin=0.4; Rmax=11.3; α=11.9; KD=224 nM.Fluorescence ratio changes are presented rather than changes in absolute [Ca2+]i because many factors (e.g. viscosity of the medium, diffusion, etc.) can affect the absolute [Ca2+]i. The mean basal [Ca2+]i was calculated for all the cells tested in the different experiments. Data are expressed as the peak increase in [Ca2+]i from initial basal levels. The calculated mean amplitude and duration of the calcium responses of each hair cell were averaged for each agonist and antagonist tested. Values are expressed as means±S.D.3Results3.1Calcium concentration changes induced by glutamate, glutamate agonists and glycine in type I vestibular sensory cellsThe calcium fluorescence ratio for type I sensory cells was 0.75±0.09 (n=40) at rest, corresponding to a calcium concentration of 88.4 nM. This value is consistent with the satisfactory maintenance of viable cells in vitro. Applications of glutamate, NMDA and aspartate induced rapid and transient increases in fluorescence ratio of 56, 34 and 47% respectively (Fig. 1). Glycine application induced a similar increase of 55% (Fig. 1). The delay between application and response was about 1 s. The peak value was reached in about 9 s and the return to resting level took about 20 s, indicating that regulatory processes were occurring (Table 1). The shapes of response curves were similar for all cells tested.The specificity of the calcium response induced by iontophoretic applications was checked by ejection of a different anion (Cl−) or cation (Na+). Neither caused any change in fluorescence ratio (Fig. 1).3.2Effect of 7-chlorokynurenic acid on the calcium response to glycine in sensory cellsTwo consecutive applications of glycine, separated by a rinse of 5 min, induced similar calcium responses (Fig. 2A,B), with similar profiles and peak amplitudes.A selective antagonist of the glycine site on NMDA receptors, 7-chlorokynurenate (10 μM), reduced the calcium response evoked by glycine stimulation by about 60% (Fig. 2A,B). The presence of APV (10 μM), a competitive antagonist of the glutamate site on the NMDA receptor, had no significant effect on the calcium response evoked by glycine application (Fig. 2B), whereas the presence of APV decreased the calcium response induced by glutamate by 48% and that induced by NMDA by 90% (Table 2).3.3Effect of strychnine on the calcium response induced by glycine in type I sensory cellGlycine may also activate the ionotropic glycine receptor, which is selectively antagonized by strychnine. The calcium response evoked by glycine was therefore evaluated in the presence of strychnine (50 μM): it was inhibited by 60% (Fig. 3, Table 3). The calcium response induced by glutamate application was not affected by the presence of strychnine (Table 3).Strychnine and 7-chlorokynurenate, applied together, reduced the calcium response induced by glycine by 73% (Fig. 2, Table 3).3.4Effect of extracellular chloride ions on the calcium response induced by glycineThe ionotropic glycine receptor is selectively permeable to chloride ions. We tested whether there was a relationship between chloride influx and changes in [Ca2+]i. The chloride ions of the medium were replaced with gluconate which cannot pass through the glycine receptor channel or the plasmalemma. In the presence of sodium gluconate (137 mM), the calcium resting level was slightly lower (Fig. 4) and the calcium response evoked by glycine was 58% smaller (Fig. 4, Table 4). The addition of both gluconate and 7-chlorokynurenate decreased the calcium response to glycine by 84% (Fig. 4, Table 4).3.5Effect of extracellular calcium ions on the calcium response induced by glycineWe tested whether external calcium affects the changes in calcium concentration induced by glycine by adding a specific calcium chelator EGTA (1.5 mM) to the medium (HBSS). In the presence of EGTA, calcium resting levels in the cell decreased by about 20% and the increase in calcium variation by glycine was only 6% of the basal calcium level measured in presence of EGTA (Fig. 5, Table 5). Therefore, extracellular calcium seemed to be the major source of the calcium concentration changes.4Discussion4.1Calcium concentration changes induced by glycine in type I vestibular sensory cellVestibular sensory cells are mechano-transducers, encoding and sending sensory messages by the afferent nerve to the vestibular nuclei. In isolated type I vestibular sensory cells, the low and stable [Ca2+]i before each stimulation indicated that the cells in our preparations were not altered or excited. The [Ca2+]i in type I sensory cells was similar to that observed in other cells in vitro such as cochlear ganglion neurons (Harada et al., 1994), vestibular ganglion neurons (Rabejac et al., 1997), hippocampal neurons (Murphy and Miller, 1988), and Purkinje neurons (Hockberger et al., 1989).Glycine application induced a rapid and transient increase in [Ca2+]i. Increases in [Ca2+]i were also evoked by glutamate and by two NMDA receptor agonists, aspartate and NMDA. The response curve following glycine stimulation was similar to those obtained after glutamate and aspartate stimulation. After stimulation, [Ca2+]i returned to basal levels. The [Ca2+]i is regulated by processes involving calcium binding proteins, the Ca-ATPase pump in the plasmalemma and intracellular membrane of the endoplasmic reticulum, and the Na/Ca exchanger (Tucker and Fettiplace, 1995; Tucker et al., 1996). Intracellular calcium is a messenger in fast event cascades in various cellular processes. Electrophysiological and confocal microscopy studies have shown that [Ca2+]i increases particularly in microdomains in which the calcium concentration reaches about 100 μM. These microdomains are located near the basolateral membrane in frog vestibular sensory cells (Lenzi and Roberts, 1994; Issa and Hudspeth, 1994; Tucker and Fettiplace, 1995; Tucker et al., 1996). This site corresponds to active zones of the sensory cells characterized by the presence of voltage-activated Ca2+ channels and Ca2+-activated K+ channels (Hudspeth and Lewis, 1988) and synaptic bodies involved in neurotransmitter release, which requires the rapid regulation of Ca2+ levels (Parsons et al., 1994; Fuchs, 1996). It has also been suggested that changes in [Ca2+]i at the apex of sensory cells may affect the adaptation of the mechano-transduction channels in the hair bundle, adjusting the operating range of the transducer (Denk et al., 1995).4.2Effect of glycine on NMDA/glycine receptorsGlycine may activate a glycine site on the NMDA receptor in type I vestibular sensory cells. We found that a selective antagonist of this site, 7-chlorokynurenate, reduced the increase in [Ca2+]i induced by glycine. Glycine is a co-agonist with glutamate to activate the NMDA receptors (Moriyoshi et al., 1991; Hollmann and Heinemann, 1994; Mori and Mishina, 1995; Corsi et al., 1996; Sucher et al., 1996). However, in some NMDA receptor subunit configurations, heteromeric NMDA receptors may be activated by glycine alone. The presence of the NR1 subunit associated with NR2A or NR2B or NR2C subunits in Xenopus oocytes results in the induction of a small but clear inward current induced by 10 μM glycine alone (Meguro et al., 1992; Kutsuwada et al., 1992). In rat neocortical neurons, nerve terminals, which release cholecystokinin and somatostatin, possess NMDA receptors, the channels of which can be operated by glycine or D-serine alone with no apparent activation of the glutamatergic co-agonist site (Paudice et al., 1998). The functional properties of the NMDA receptor channel are determined by the NR2 subunit (Kutsuwada et al., 1992; Molinoff et al., 1994; Paudice et al., 1998). The composition of NMDA subunits, in the vestibular epithelium, is unknown but the NR1 subunit has been detected in vestibular sensory cells (Devau et al., 1997). Glycine may act as a co-agonist with residual glutamate released by the sensory cell near the soma before rapidly spreading (schematic drawing Fig. 6). This contamination by glutamate may account for the NMDA receptor activation. The presence of APV (Watkins and Olverman, 1987), a competitive antagonist of the glutamate site on NMDA receptors, decreased the calcium response to NMDA by 90% and that to glutamate by 48% but did not alter the calcium response to glycine. This suggests that NMDA receptors sensitive to glycine are present on vestibular sensory cell. The almost complete inhibition of NMDA by APV, and only about 50% inhibition by glutamate alone suggest that glutamate acts on both ionotropic and metabotropic receptors (Devau et al., 1993; Guth et al., 1998a,b)However, 7-chlorokynurenate, a selective antagonist of the glycine site of NMDA receptors, contributes to the partial inhibition of the calcium response induced by glycine.4.3Effect of glycine on strychnine-sensitive glycine receptorsGlycine also activates ionotropic glycine receptors, which are specifically permeable to chloride ions and sensitive to strychnine. Strychnine reduced the increase in [Ca2+]i after glycine application by 60%, but decreased by only 6% the calcium response induced by glutamate. Strychnine interacts with the nicotinic receptors of acetylcholine as a potent antagonist. In frog vestibular hair cells, strychnine interacts with nicotinic-like receptors (Guth et al., 1994; Guth and Norris, 1996). The α9 subunit of the nicotinic receptor is very sensitive to strychnine and mRNA encoding the α9 subunit has been detected in the adult rat peripheral vestibular system (Hiel et al., 1996). Acetylcholine is the principal neurotransmitter of the efferent system and cholinergic receptors are present in type II sensory cells. However, the type I sensory cells of mammals are surrounded by afferent nerve calyces and are not in contact with any efferent nerve endings. Thus, the cholinergic receptors are not responsible for the type I sensory cell response to strychnine.The stimulation of glycine receptors resulted in chloride permeability. In the hair cell, chloride ions contribute greatly to the maintenance of cell volume (Rennie et al., 1997). In low chloride medium, the outer hair cells of the cochlear epithelium first shorten, then rapidly increase in length. The cells then collapse due to an efflux of Cl− followed by K+ and water (Cecola and Bobbin, 1992). In Cl−-free gluconate-HBSS, resting calcium levels were low in isolated type I sensory cells. Transferring the cells to normal HBSS resulted in a return to normal calcium resting levels. In the medium, gluconate may be associated with sodium or calcium. In many preparations, gluconate has been used to substitute for chloride ions in experiments with isolated vestibular type I hair cells (Rennie et al., 1997). However, gluconate may chelate external calcium and partially reduce the glycine effect.In other systems, for example, at some stage of development in chick embryo ciliary ganglion cells, activation of the ionotropic glycine receptor triggers an increase in [Ca2+]i, which is inhibited by strychnine (Sorimachi et al., 1997). Activation of the embryonic ionotropic glycine receptor causes depolarization and calcium transients in other neurons such as rat dorsal horn neurons (Wang et al., 1994). In adult neurons and smooth muscle cells, calcium activates chloride channels (Marsh et al., 1997; Wang and Kotlikoff, 1997). We found that in adult vestibular type I sensory cells, activation of the strychnine-sensitive glycine receptor induced an increase in [Ca2+]i, but the biochemical steps linking these events are unknown.The addition of strychnine and 7-chlorokynurenate together strongly decreased the calcium response suggesting that there are two components involved in the calcium transient. In the presence of an agonist, the activity of a ligand-gated channel such as an NMDA or glycine receptors may be modified by desensitization, which may in turn affect cell excitability (Jones and Westbrook, 1996). After washing for 5 min, a second glycine stimulation induced a calcium response with an amplitude similar to that of the first. Thus, we found no evidence of desensitization in type I vestibular sensory cells.4.4Glycine and amino acid transportersGlycine may also directly activate glycine transporters. These transmembrane proteins have 12 transmembrane segments like other members of the transporter family, and their function depends on the cotransport of electrogenic sodium and chloride (Attwell and Mobbs, 1994; Olivares et al., 1997; Zafra et al., 1997). The detection and the cellular localization of glycine transporters in the vestibular sensory epithelium could be determined by an immunocytochemical study, with specific antibodies directed against the glial glycine transporter (GLYT1) and the neuronal glycine transporter (GLYT2). The distribution of GLYT2 is correlated with that of strychnine-sensitive glycine receptors in most of the central nervous system except cerebellum (Nelson, 1998). They may be responsible for the residual response in the presence of 7-chlorokynurenate and strychnine. The sodium and chloride flux induced by glycine uptake in vestibular sensory cells may cause a change in [Ca2+]i contributing to the overall calcium response to glycine stimulation. To measure the sodium concentration changes induced by activation of glycine transporters a specific sodium-sensitive dye, SBFI, could be used.4.5Calcium source: influx of extracellular calcium and mobilization of intracellular calcium storesThe increase in [Ca2+]i induced by glycine may result from the entry of external calcium or from release from internal stores or both. Calcium may enter through the highly sensitive calcium channel, of the NMDA receptor, or may be released from the internal calcium store following signal transduction. Activation of the NMDA receptor may also depolarize the membrane and calcium influx may be increased by opening of the L-type calcium channels present in both types of vestibular sensory cells (Boyer et al., 1998). The presence of nickel and cadmium ions (Ni2+/Cd2+), which block voltage-dependent calcium channels, decreased by 30% the calcium response induced by glycine (personal observation). Calcium channel specific blockers (such as dihydropyridines for L-type or ω-conotoxin-GVIA for N-type) could be used to determine the type of the voltage-dependent calcium channels involved in the calcium concentration changes in vestibular sensory cell after glycine application. However, only L-type calcium channels have been shown in type I vestibular sensory cell (Boyer et al., 1998). In the presence of EGTA, a specific calcium chelator, the calcium response to glycine application was very small indicating that an influx of external calcium was the major component of the response. However, a residual response persisted demonstrating that internal calcium stores were not totally empty.In conclusion, the application of glycine induced an increase in [Ca2+]i in isolated type I vestibular sensory cells from guinea pig. Changes in calcium concentration involved at least two receptors: (i) heteromeric NMDA receptors, which are highly sensitive to glycine, and (ii) the strychnine-sensitive glycine receptor. Changes in calcium concentration may affect neurotransmitter release in the basolateral region of the cell. Glycine may then act in synergy with glutamate or aspartate in local control of the cytoneuronal synapse between the sensory cell and the afferent nerve calyx.AcknowledgementsI would like to thank Claude Dechesne, Danielle Demêmes, Marie-Thérèse Nicolas, Jacqueline Raymond for excellent constructive criticism and Denis Orcel for designing the schematic drawing.ReferencesAnnoni et al., 1984J.M.AnnoniS.L.CochranW.PrechtPharmacology of the vestibular hair cell-afferent fiber synapse in the frogJ. Neurosci.4198421062116Anson and Ashmore, 1994Anson, L.C., Ashmore, J.F., 1994. Evidence for release of an excitatory amino acid from inner hair cells isolated from the guinea pig. First International Symposium Inner Ear Neuropharmacology, Vol. 1, p. 3.Attwell and Mobbs, 1994D.AttwellP.MobbsNeurotransmitter transportersCurr. Opin. Neurobiol.41994353359Bäurle et al., 1997J.BäurleJ.KleineO.-J.GrüsserW.GuldinCo-localization of glycine and calbindin D-28K in the vestibular ganglion of the ratNeuroReport8199724432447Boyer et al., 1998C.BoyerJ.LehouelleurA.SansPotassium depolarization of mammalian vestibular sensory cells increases [Ca2+]i through voltage-sensitive calcium channelsEur. J. Neurosci.101998971975Breitinger and Becker, 1998H.G.BreitingerC.-M.BeckerThe inhibitory glycine receptor: prospects for a therapeutic orphan?Curr. Pharm. Des.41998315334Cecola and Bobbin, 1992R.P.CecolaR.P.BobbinLowering extracellular chloride concentration alters outer hair cell shapeHear. Res.6119926572Corsi et al., 1996M.CorsiP.FinaD.G.TristCo-agonism in drug-receptor interaction: illustrated by the NMDA receptorsTrends Pharmacol. Sci.171996220223Dechesne et al., 1997C.J.DechesneC.KauffO.StettlerB.TavitianRab3A immunolocalization in the mammalian vestibular end-organs during development and comparison with synaptophysin expressionDev. Brain Res.991997103111Demêmes et al., 1990D.DemêmesR.J.WentholdB.MoniotA.SansGlutamate-like immunoreactivity in the peripheral vestibular system of mammalsHear. Res.461990261270Demêmes et al., 1995D.DemêmesA.LleixaC.J.DechesneCellular and subcellular localization of AMPA-selective glutamate receptors in mammalian peripheral vestibular systemBrain Res.67119958394Denk et al., 1995W.DenkJ.R.HoltG.M.G.ShepherdD.P.CoreyCalcium imaging of single stereocilia in hair cells: localization of transduction channels at both ends of tip linksNeuron15199513111321Devau et al., 1993G.DevauJ.LehouelleurA.SansGlutamate receptors on type I vestibular hair cells of guinea pigEur. J. Neurosci.5199312101217Devau et al., 1997Devau, G., Dechesne, C.J., Demêmes, D., 1997. Modulation de l’activité des cellules sensorielles de cobaye par la glycine. 3rd meeting of the French Neuroscience Society, May 25–28, Bordeaux France.Drescher and Drescher, 1992M.J.DrescherD.G.DrescherGlutamate, of the endogenous primary alpha-amino acids, is specifically released from hair cells by elevated extracellular potassiumJ. Neurochem.5919929398Drescher et al., 1987M.J.DrescherD.G.DrescherJ.S.HatfieldPotassium-evoked release of endogenous primary amine-containing compounds from the trout saccular macula and saccular nerve in vitroBrain Res.41719873950Fuchs, 1996P.A.FuchsSynaptic transmission at vertebrate hair cellsCurr. Opin. Neurobiol.61996514519Fujita et al., 1994S.FujitaS.UsamiH.ShinkawaK.SatoH.KiyamaM.TohyamaExpression of NMDA subunit mRNA in the vestibular ganglion of the rat and guinea-pigNeuroReport51994862864Griguer et al., 1993C.GriguerA.SansJ.ValmierJ.LehouelleurInward potassium rectifier current in type I vestibular hair cells isolated from guinea pigNeurosci. Lett.14919935155Grynkiewicz et al., 1985G.GrynkiewiczM.PoenieP.Y.TsienA new generation of Ca2+ indicators with greatly improved fluorescence propertiesJ. Biol. Chem.260198534403450Guth and Norris, 1996P.S.GuthC.H.NorrisThe hair cell acetylcholine receptors: a synthesisHear. Res.98199618Guth et al., 1988P.S.GuthC.H.NorrisS.E.BarronThree tests to the hypothesis that glutamate is the sensory hair cell transmitter in frog semicircular canalHear. Res.331988223228Guth et al., 1991P.S.GuthA.AubertA.J.RicciC.H.NorrisDifferential modulation of spontaneous and evoked neurotransmitter release from hair cells: some novel hypothesesHear. Res.5619916978Guth et al., 1994P.S.GuthA.DunnK.KronomerC.H.NorrisThe cholinergic pharmacology of the frog sacculeHear. Res.751994225232Guth et al., 1998aP.S.GuthP.PerinC.H.NorrisP.ValliThe vestibular hair cells: post-transductional signal processingProg. Neurobiol.541998193247Guth et al., 1998bP.S.GuthJ.C.HoltP.PerinG.AthasM.GarciaA.PuriG.ZuccaL.BottaP.ValliThe metabotropic glutamate receptors of the vestibular organsHear. Res.1251998154162Harada et al., 1994N.HaradaD.Y.HanM.KomedaT.YamashitaGlutamate-induced intracellular Ca2+ elevation in isolated spiral ganglion cells of the guinea-pig cochleaActa Otolaryngol. (Stockh.)1141994609612Hiel et al., 1996H.HielA.ElgoyhenD.G.DrescherB.J.MorlayExpression of nicotinic acetylcholine receptor mRNA in the adult rat peripheral vestibular systemBrain Res.7381996347352Hockberger et al., 1989P.E.HockbergerH.Y.TsengJ.A.ConnorFura-2 measurements of cultured rat Purkinje neurons show dendritic localization of Ca2+ influxJ. Neurosci.9198922722284Hollmann and Heinemann, 1994M.HollmannS.HeinemannCloned glutamate receptorsAnnu. Rev. Neurosci.17199431108Hudspeth and Lewis, 1988A.J.HudspethR.S.LewisKinetic anlysis of voltaged- and ion-dependent conductances in hair cells of the bulllfrog, Rana catesbeianaJ. Physiol.4001988237274Issa and Hudspeth, 1994N.P.IssaA.J.HudspethClustering of Ca2+ and Ca2+-activated K+ channels at fluorescently labeled presynaptic active zones of hair cellsProc. Natl. Acad. Sci. USA91199475787582Jones and Westbrook, 1996M.V.JonesG.L.WestbrookThe impact of receptor desensitization on fast synaptic transmissionTrends Neurosci.19199696101Johnson and Ascher, 1987J.W.JohnsonP.AscherGlycine potentiates the NMDA response in cultured mouse brain neuronsNature3251987529531Kataoka and Ohmori, 1996Y.KataokaH.OhmoriOf known neurotransmitters, glutamate is the most likely to be released from chick cochlear hair cellsJ. Neurophysiol.76199618701879Kemp and Leeson, 1993J.A.KempP.D.LeesonThe glycine site of the NMDA receptor – five years onTrends Pharmacol. Sci.1419932025Kemp et al., 1988J.A.KempA.C.FosterP.D.LeesonT.PriestleyR.TridgettL.L.IversenG.N.Woodruff7-Chlorokynurenic acid is a selective antagonist at the glycine modulatory site of NMDA receptor complexProc. Natl. Acad. Sci. USA85198865476550Kutsuwada et al., 1992T.KutsuwadaN.KashiwabuchiH.MoriK.SakimuraE.KushiyaK.ArakiH.MeguroH.MasakiT.KumanishiM.ArakawaM.MishinaMolecular diversity of the NMDA receptor channelNature35819923641Langosch et al., 1988D.LangoschL.ThomasH.BetzConserved quaternary structure of ligand gated ion channels: the postsynaptic glycine receptor is a pentamerProc. Natl. Acad. Sci. USA85198873947398Lenzi and Roberts, 1994D.LenziW.M.RobertsCalcium signaling in hair cells: multiple roles in compact cellCurr. Opin. Neurobiol.41994496502Marsh et al., 1997S.J.MarshJ.TrouslardJ.L.LeaneyD.A.BrownSynergistic regulation of a neuronal chloride current by intracellular calcium and muscarinic receptor activation: a role for proteine kinase CNeuron151997729737Meguro et al., 1992H.MeguroH.MoriK.ArakiE.KushiyaT.KutsuwadaM.YamazakiT.KumanishiM.ArakawaK.SakimuraM.MishinaFunctional characterization of a heteromeric NMDA receptor channel expressed from cloned cDNAsNature35719927074Molinoff et al., 1994Molinoff, P.B., Williams, K., Pritchett, D.B., Zhong, J., 1994. Molecular pharmacology of NMDA receptors: modulatory role of NR2 subunits. In: Bloom, F. (Ed.), Progress in Brain Research, Vol, 100. Elsevier Science, Amsterdam, pp. 39–45.Mori and Mishina, 1995H.MoriM.MishinaStructure and function of the NMDA receptor channelNeuropharmacology34199512191237Moriyoshi et al., 1991K.MoriyoshiM.MasuT.IshiiR.ShigemotoN.MizunoS.NakanishiMolecular cloning and characterization of the rat NMDA receptorNature35419913137Murphy and Miller, 1988S.N.MurphyR.J.MillerA glutamate receptor regulates Ca2+ mobilization in hippocampal neuronsProc. Natl. Acad. Sci. USA85198887378741Nelson, 1998N.NelsonThe family of Na+/Cl− neurotransmitter transportersJ. Neurochem.71199817851803Niedzielski and Wenthold, 1995A.S.NiedzielskiR.J.WentholdExpression of AMPA, Kainate and NMDA receptor subunits in cochlear and vestibular gangliaJ. Neurosci.15199523382353Olivares et al., 1997L.OlivaresC.AragonC.GimenezF.ZafraAnalysis of the transmembrane topology of the glycine transporter GLYT1Biochem. Mol. Biol.272199712111217Parsons et al., 1994T.D.ParsonsD.LenziW.AlmersW.M.RobertsCalcium-triggered exocytosis and endocytosis in an isolated presynaptic cell: capacitance measurements in saccular hair cellsNeuron131994875883Paudice et al., 1998P.PaudiceA.GemignaniM.RaiteriEvidence for functional native NMDA receptors activated by glycine or D-serine alone in the absence of glutamatergic coagonistEur. J. Neurol.10199829342944Prigioni et al., 1990I.PrigioniG.RussoP.ValliS.MasettoPre- and postsynaptic excitatory action of glutamate agonists on frog vestibular receptorsHear. Res.461990253260Prigioni et al., 1994I.PrigioniG.RussoS.MasettoNon-NMDA receptors mediate glutamate induced depolarization in frog cristae ampullarisNeuroReport51994516518Rabejac et al., 1997D.RabejacG.DevauJ.RaymondAMPA receptors in cultured ganglion neurons: detection and activationEur. J. Neurosci.91997221228Reichenberger and Dieringer, 1994I.ReichenbergerN.DieringerSize-related colocalization of glycine and glutamate immunoreactivity in frog and rat vestibular afferentsJ. Comp. Neurol.3491994603614Rennie et al., 1997Rennie, K.J., Ashmore, J.F., Correia, M.J., 1997. Evidence for an Na+-K+Cl− cotransporter in mammalian type I vestibular hair cells. Am. J. Physiol. 273 (Cell Physiol. 42), C1972–C1980.Scarfone et al., 1988E.ScarfoneD.DemêmesR.JahnP.De CamilliA.SansSecretory function of the vestibular nerve calyx suggested by the presence of vesicles, synapsin I and synaptophysinJ. Neurosci.8198846404645Scarfone et al., 1991E.ScarfoneD.DemêmesA.SansSynapsin I and synaptophysin expression during ontogenesis of the mouse peripheral vestibular systemJ. Neurosci.11199111731181Scarfone et al., 1996E.ScarfoneM.UlfendahlT.LundebergThe cellular localization of the neuropeptides substance P, neurokinin A, gene-related peptide and neuropeptide Y in guinea-pig vestibular sensory organs: a high-resolution confocal microscopy studyNeuroscience751996587600Sorimachi et al., 1997M.SorimachiJ.S.RheeM.ShimuraN.AkaikeMechanisms of GABA- and glycine-induced increases of cytosolic Ca++ concentrations in chick embryo ciliary ganglion cellsJ. Neurochem.691997797805Soto and Vega, 1988E.SotoR.VegaActions of excitatory amino acid agonists and antagonists on the primary afferents of the vestibular system of the axolotl (Ambystoma mexicanum)Brain Res.4621988104111Soto et al., 1994E.SotoA.FloresC.ErosteguiR.VegaEvidence for NMDA receptor in the afferent synaptic transmission of the vestibular systemBrain Res.6331994289296Straka et al., 1996aH.StrakaI.ReichenbergerN.DieringerSize-related properties of vestibular afferent fibers in the frog: uptake of and immunoreactivity for glycine and aspartate/glutamateNeuroscience701996685696Straka et al., 1996bH.StrakaK.DeblerN.DieringerSize-related properties of vestibular afferent fibers in the frog: differential synaptic activation of N-methyl-D-Aspartate and non-N-methyl-D-aspartate receptorsNeuroscience701996697707Sucher et al., 1996N.J.SucherM.AwobuluyiY-B.ChoiS.A.LiptonNMDA receptors: from genes to channelsTrends Pharmacol. Sci.171996348355Tucker and Fettiplace, 1995T.TuckerR.FettiplaceConfocal imaging of calcium microdomains and calcium extrusion in turtle hair cellsNeuron15199513231335Tucker et al., 1996T.TuckerJ.J.ArtR.FettiplaceRoutes of calcium entry and extrusion in turtle hair cells. New directions in vestibular researchAnn. NY Acad. Sci.7811996123137Valli et al., 1985P.ValliG.ZuccaI.PrigioniL.BottaC.CasellaP.GuthThe effect of glutamate on the frog semicircular canalBrain Res.330198519Van den Pol and Gorcs, 1988A.N.Van den PolT.GorcsGlycine and glycine receptor immunoreactivity in brain and spinal cordJ. Neurosci.81988472492Vandenberg et al., 1992R.J.VandenbergC.A.HandfordP.R.SchofieldDistinct agonist- and antagonist-binding sites on the glycine receptorNeuron91992491496Wang et al., 1994J.WangD.B.ReichlingA.KyrosisA.B.MacDermottDevelopmental loss of GABA- and glycine-induced depolarization and Ca2+ transients in embryonic rat dorsal horn in cultureEur. J. Neurosci.6199412751280Wang and Kotlikoff, 1997Wang, Y.-X., Kotlikoff, M., 1997. Muscarinic signaling pathway for calcium release and calcium-activated chloride current in smooth muscle. Am. J. Physiol. Soc. 273 (Cell Physiol. 42), C509–C519.Watkins and Olverman, 1987J.C.WatkinsH.J.OlvermanAgonists and antagonists for excitatory amino acid receptorsTrends Neurosci.1071987265272Wersäll, 1956J.WersällStudies on the structure and innervation of the sensory epithelium of the cristae ampullaris in the guinea-pig. A light and electron microscopic investigationActa Otolaryngol. (Stockh.)126Suppl.1956185Wersäll and Bagger-Sjöback, 1974Wersäll, J., Bagger-Sjöback, D. (1974) Morphology of the vestibular sense organ. In: Kornhuber, H.H. (Ed.), Handbook of Sensory Physiology 6, Springer-Verlag, Berlin, pp. 123–170.Zafra et al., 1997F.ZafraC.AragonC.GimenezMolecular biology and glycinergic neurotransmissionMol. Neurobiol.141997117142Zucca et al., 1992G.ZuccaL.BottaV.MilesiF.DaganiP.ValliEvidence for L-glutamate release in frog vestibular organsHear. Res.6319925256Zucca et al., 1993G.ZuccaG.N.AkoevA.MaracciP.ValliNMDA receptors in frog semicircular canalsNeuroReport41993403404
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B44A118F148947F01791D40DDD487B44E25650B.txt b/test/dataset/in/resources/corpus/Clean_0B44A118F148947F01791D40DDD487B44E25650B.txt
deleted file mode 100644
index 9dd4afe..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B44A118F148947F01791D40DDD487B44E25650B.txt
+++ /dev/null
@@ -1 +0,0 @@
-fampractfamprjFamily Practice1460-22290263-2136Oxford University Press10.1093/fampra/cmq021Primary Care EpidemiologyA pilot study on the validity and reliability of the Patient Enablement Instrument (PEI) in a Chinese populationLamCindy LKaYuenNatalie YKa*MercerStewart WbWongWendyaaFamily Medicine Unit, Department of Medicine, Li Ka Shing Faculty of Medicine, The University of Hong Kong, 3/F, Ap Lei Chau Clinic, 161 Main Street, Ap Lei Chau, Hong Kong, ChinabDepartment of General Practice and Primary Care, Faculty of Medicine, University of Glasgow, 1 Horselethill Road, Glasgow G12 9LX, Scotland, UK*Correspondence to Natalie YK Yuen, Family Medicine Unit, Department of Medicine, Li Ka Shing Faculty of Medicine, The University of Hong Kong, 3/F, Ap Lei Chau Clinic, 161 Main Street, Ap Lei Chau, Hong Kong, China; Email: nykyuen@gmail.com820103042010274395403281020092003201023032010© The Author 2010. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org.2010Background. The Patient Enablement Instrument (PEI) was developed to measure patients’ enablement, which is an indicator of the effectiveness of a primary care consultation; however, to date, the PEI has not been tested in Asian populations.Objectives. The purpose of this study is to test the acceptability, validity, reliability and other psychometric properties of a Chinese [Hong Kong (HK)] translation of the PEI in Chinese patients in Hong Kong and whether these properties would be affected by different timing of administration.Methods. A Chinese (HK) translation of the PEI was developed by iterative forward–backward translations and the content validity was assessed by a cognitive debriefing interview with 10 Chinese patients. It was then administered to 152 adult patients attending a government-funded primary care clinic in Hong Kong both immediately after the consultation and 2–3 weeks later by telephone. Internal construct validity was assessed by item-scale correlations and factor analysis, test–retest reliability was assessed by intraclass correlation (ICC) and sensitivity was assessed by known group comparison.Results. The Chinese (HK) PEI was semantically equivalent to the original PEI for all items. Acceptability of the PEI was high with 83.1% response and 100% completion rates. Statistical analyses showed no difference between test and retest means as well as good reproducibility (ICC 0.75). Internal reliability determined by Cronbach’s alpha was >0.8 irrespective of timing of administration. Scale construct validity was confirmed by strong (r > 0.4) item-scale correlations and resumed to a one-factor hypothesized structure. PEI scores were significantly higher in younger patients supporting sensitivity. There was no significant difference in the psychometric properties or scores between the assessment results from immediately after and 2-weeks post-consultation.Conclusions. A Chinese (HK) translation of the PEI equivalent to the original is now available for application to Chinese populations. Pilot testing supported its acceptability, validity, reliability and sensitivity. Further studies to confirm its construct validity and responsiveness will help to establish the Chinese (HK) PEI as an outcome measure of the effectiveness of primary care consultations in Chinese patients.Chinesecontent validitypatient enablementprimary carereliabilityIntroductionThe clinical encounter between patient and practitioner is the core activity of all systems of health care. As such, the practitioner–patient relationship has been extensively investigated in conventional care, particularly in the primary care setting.1–4 However, while various methods exist to measure the effectiveness of primary care consultations, few exist to measure the effectiveness of interpersonal care in primary care.5 Many primary care consultations are for problems that are self-limiting, for which a favourable biomedical outcome can be anticipated regardless of the activities of the doctor. On the other hand, a large number of consultations are for multidimensional problems, where physical, psychological and social elements as well as the need for health promotion interact in a complex way that traditional clinical outcomes are not sensitive enough to capture.6There is some evidence to suggest that there is a positive link between patient-centred consultations and health outcomes.2,3,7 Two core values of family medicine are patient-centred care and holism,8 upon which is built the concept of patient enablement, first articulated by Howie et al.9,10 ‘Enablement’ describes a consultation outcome that reflects the extent to which patients understand their health problems and feel able to cope with them as a result of the consultation. According to Wensing, patients find it very important to be able to understand the nature of their problem and manage their own illness,11 which supports the concept of enablement as a patient-centred health-related benefit resulting from the consultation.Howie’s theory of quality of care implies that greater enablement is achieved when the patient’s needs are identified, acknowledged and dealt with in the appropriate context.12 It recognizes the importance of themes of patient centeredness and empowerment and of the patient’s ability to understand and cope with their health and illness and recognized these issues as being conceptually distinct from satisfaction.5 The original version of the Patient Enablement Instrument (PEI) was developed by Howie et al.5 based on these concepts to assess the enablement of patients after a consultation with their primary care clinician. The PEI has six questions: able to cope with life, able to understand your illness, able to cope with your illness, able to keep yourself healthy, confident about your health and able to help yourself. Each question has four response options: much better/better (Questions 1–4) or much more/more (Questions 5–6) or same or less or not applicable. The PEI is intended to be a measure of the effectiveness of primary care consultations13 and has been used for this purpose in several studies in Western countries, including the UK,10,14 Croatia,8 the Netherlands11 and Poland.15Previous studies have shown that patient enablement may be increased by doctor’s empathy,14 familiarity of the patient with the practitioner,16 positive patient evaluation of the doctor’s communication,17 as well as longer consultation duration18 and appropriate doctor training.15 In turn, patient enablement results in greater compliance with therapy and improved outcomes,18,19 as well as increased well-being up to 1 month post-consultation.14 A 6-week programme enabling patients with chronic disease in Hong Kong showed that Chinese patients undergoing the programme exhibited significantly higher self-efficacy in managing their illness and felt more energetic, the results of which supported the validity and relevance of the concept of enablement in Chinese culture.20 In contrast, low enablement is associated with a lack of continuity of care, poor self-perceived health, low educational level and low quality of life.8Hong Kong’s health system is a combination of a tax-funded health model combined with a private sector, which account for 56% and 44% of health care spending respectively.21 Total expenditure on health care amounts to ∼5.6% of the gross domestic product. Hong Kong has no mandatory primary care network, which means that patients are free to choose to consult doctors of any specialty, which results in a lack of gate-keeping, increased doctor shopping and increased health care costs.Ninety-five percent of Hong Kong’s population is Chinese and for 91% of the population, the first language is Cantonese.22 The culture of consultation is traditionally seen as paternalistic, with little chance for conversation or joint decision making by patient and doctor. Chinese society often embodies collectivism and power distance23 which means that patients rarely assert individual rights during their consultations and that they often behave deferentially towards doctors. While their dissatisfaction with doctor-centred consultation may not be evident during the consultation, it is exhibited by the high rate of doctor shopping and increased consultation rate. One study found that patients in Hong Kong preferred patient-centred consultations, regardless of differences in their age, income, education, housing and sex.23There is very little research on patient-centred care or enablement in Chinese populations. The PEI would be a very useful tool for the evaluation and enhancement of patient enablement in the consultation in order to shift the doctor-dependent paradigm in this part of the world. For a patient-reported outcome measure to be used in Hong Kong, it must first be translated into Chinese and then tested for cross-cultural equivalence, validity and psychometric adequacy. The PEI has never been tested in an Asian population, and the data on its psychometric properties in any populations are rather limited.The aim of our study was to develop a Chinese [Hong Kong (HK)] equivalent translation of the PEI and to evaluate its acceptability, validity and reliability in Chinese patients. We wanted to confirm that the items of the PEI were valid indicators of the underlying construct and that they share a common principal factor in order to justify direct summation of the item scores. In addition, we wanted to find out whether the Chinese (HK) PEI results would be influenced by patient characteristics and timing of its administration. Results showing the validity and reliability of the PEI in our Chinese population would support its use in other Chinese populations and adaptation of the PEI to other Asian cultures.MethodsDevelopment of the Chinese (HK) PEI and evaluation on content validityThe Chinese (HK) translation of the PEI was developed according to an internationally recommended method.24 Two professional translators, who were native speakers of Chinese, translated the original English PEI into Chinese independently. A panel consisting of the translators and two bilingual authors (CLKL and WW) reviewed the translations to form the first draft of the Chinese (HK) PEI. Another professional translator blind to the original PEI back-translated this first draft to English. The back-translation was assessed for equivalence to the original by one of the authors who are a native English speaker familiar with the original PEI (SM). Discrepancies between the original English language PEI and the back-translation of the first draft Chinese (HK) PEI were reviewed and revised by the panel (Appendix 1). Cognitive debriefing on the second draft was performed with 10 Chinese patients to assess the clarity and interpretation of each item and response option. The cognitive debriefing subjects were a convenient sample balanced for age and sex recruited among attendees of a government primary care clinic in Hong Kong. The final version of the Chinese (HK) PEI (Appendix 2) was then used to collect data for psychometric testing.SubjectsAll patients aged 18 years or more consulting one randomly selected doctor in an average size government-funded primary care clinic in Hong Kong from 20 April to 9 May 2009 were included. The clinic is a teaching practice with three to four doctors consulting in any one session. It is situated in an urban, low socio-economic district of Hong Kong. Patients present with similar problems to other government clinics, namely predominantly chronic conditions such as hypertension and diabetes. All doctors speak Cantonese and 99% of the consultations are conducted in this dialect. To recruit patients, a random number table generated by computer was used to select one doctor (by room number) from all those available for consultation during a particular clinical session. The aim, procedures and nature of the study were explained and written consent was obtained from each subject. Patients were excluded if they were unable to communicate in Cantonese, if they were too ill to complete the interview or if they refused to take part in the study. Patients did not receive any incentives for participating in the study.Data collectionEach subject answered the Chinese (HK) PEI via interview with a trained research assistant, who was not part of usual clinic staff, after the consultation. Information was gathered on purpose of consultation (chronic disease follow-up or acute illness), age, sex, marital status (single, married, divorced and widowed), education level and occupation (according to the British Registrar General’s classification of occupation).25All subjects were followed up by telephone 2 or more weeks after the initial consultation, when the PEI was administered again by a different trained research assistant. Lam et al.26 previously showed that telephone interviews give similar results on health service utilization as those found in face-to-face surveys.Outcome measures and statistical analysesThe data of all subjects who had completed the follow-up (retest) interviews were included in the analysis of the test–retest reliability. In order to evaluate whether the psychometric properties of the PEI would be affected by the timing of administration, we split the subjects in half for this part of the data analysis. Seventy-six subjects were randomly selected from the 131 who completed the retest to form the sample for the evaluation of delayed administration, and the rest of the 152 who completed the baseline interviews were used as the sample for the evaluation of immediate administration of the Chinese (HK) PEI.The Chinese (HK) PEI score was calculated by the mean of the scores of the applicable items multiplied by six. Cases that had more than three ‘not applicable’ items were excluded. Responses of ‘much better/much more’, ‘better/more’ and ‘same or less’ were scored 2, 1 and 0, respectively, giving a summation PEI score ranging from 0 to 12. Descriptive statistics were determined, including mean, SD, percentage of floor and ceiling of scores by assessment time.Test–retest reliability was evaluated by paired t-tests on the difference in mean and also with intraclass correction coefficient (ICC) between test and retest scores. ICC is the recommended test on reproducibility of psychometric measures and the expected standard is ≥0.7.27 Internal reliability was measured by Cronbach’s alpha coefficient, for which ≥0.7 has been suggested as the standard for group comparison.28The internal construct validity of the PEI items was first evaluated by Spearman’s rho correlation, corrected for overlap, between the item and the scale (total PEI) score; an item should be substantially linearly related (r ≥ 0.4) to the underlying concept (total PEI score) being measured. Factor analysis using the varimax rotation method was then performed on the item scores to extract factors that had an eigenvalue >1. As all items should measure the same construct of patient enablement, one principal component factor was hypothesized and they should have similar factor loading and correlation. Sensitivity was also explored by known group comparison of the mean scores by age, sex and consultation reason. It has been suggested in the literature that patient enablement would be higher in patients who are younger, male, and who present for an acute problem.8,17 The sensitivity in detecting a difference between groups was evaluated by two-sample t-tests. All data analyses were conducted by SPSS 17.0. Ethics approval was given by HKU/HA HKW IRB reference number, UW 07-226.ResultsContent validity and translation equivalence of the Chinese (HK) PEIThe back-translation of the first draft Chinese (HK) PEI was equivalent to the original PEI for all questions and responses, with few exceptions. The review panel found that the literal translation of ‘cope with’ in Chinese (e.g. or ) was linguistically inappropriate when collocated with ‘illness’ (); thus, we translated it to the Chinese term that means ‘to face’ (), which is close in meaning to the original. In addition, ‘life’ was translated to ‘daily living’. The native English expert (SM) agreed that the back-translation of the Chinese (HK) PEI was equivalent in meaning to the original. The cognitive debriefing group was composed of 10 patients, 5 male and 5 female, aged 21–74 years (mean 48.4 years), who took 1–6 minutes to complete the test (mean 3.3 minutes). Each item was understood, rated as relevant, appropriate and not difficult and correctly interpreted by >80% of respondents of the cognitive interviews. A few subjects found the title of the instrument redundant and did not clearly interpret that the stem statement was asking about a change. In addition, several suggested that the separation of the response option ‘the same or worse (less)’ into two [‘the same’ and ‘worse (less)’] options. Therefore, revisions were made to simplify the title, reword the stem question as a question and to separate the response options to improve the clarity of the instrument.Acceptability and completion rateThe flow of patient response and follow-up is shown in Figure 1: 203 patients were contacted, of whom 31 patients were excluded because of cognitive or physical impairment (9), inability to speak Cantonese (7), aged <18 years (6) or not accessible by telephone for follow-up (9) and 20 refused to take part. One hundred and fifty-two (83.1% response rate among those eligible) completed the baseline survey with no missing data. One hundred and thirty-one (86% of 152) of the participants completed the telephone follow-up interviews, and the median time interval between the two assessments was 15 days (range 14–22 days). Data from two subjects at baseline and three subjects on follow-up were excluded from further analysis because there were more than three not applicable responses on their Chinese (HK) PEI.FIGURE 1Response and follow-up of patients consulting at Ap Lei Chau ClinicThe characteristics of the study subjects can be seen in Table 1. Most subjects were elderly and of low education and socio-economic class, and more consultations were for chronic disease rather than acute illnesses. Subject characteristics were similar except for the education level between those who completed and defaulted the follow-up interviews, suggesting little systematic bias in the retest results.TABLE 1Characteristics of study subjectsCharacteristicsBaseline (n = 152)Follow-up (n = 131)Default follow-up (n = 21)Mean age (SD)62.1 (12.68)62.1 (12.84)62.2 (11.92)Age group (years, %) 18–5942.142.042.9 60 or above56.657.352.4 Missing1.30.84.8Gender (%) Male38.238.238.1 Female61.861.861.9Education level (%)a Nil23.019.147.6 Primary36.240.59.5 Secondary35.535.138.1 Tertiary5.35.34.8Marital status (%) Married67.167.266.7 Widower17.816.823.8 Single6.66.94.8 Divorced8.69.24.8Occupation (%) Professional3.33.14.8 Associated professional7.27.64.8 Skilled worker19.719.819.0 Semi-skilled worker16.418.34.8 Unskilled worker27.026.033.3 Housewife23.023.719.0 Others3.31.514.3Consultation reason (%) Chronic disease67.167.961.9 Acute illness32.932.138.1aSignificant difference between follow-up and default follow-up subjects by chi-square test (P < 0.05).Psychometric properties of the Chinese (HK) PEITable 2 shows the distribution of baseline and follow-up Chinese (HK) PEI scores. There were few subjects who had the lowest possible (floor) or highest possible (ceiling) scores at baseline, however a relatively high proportion (18.9%) scored at the floor in the follow-up assessment. The paired t-test (P < 0.124) indicated no difference between the test and retest means, and the test–retest reliability ICC was 0.75, indicating good reproducibility. The Cronbach’s alpha coefficient of internal reliability was 0.84 and 0.85 at baseline and at follow-up, respectively, further supporting the reliability of the PEI.TABLE 2Distribution and reliability of baseline and follow-up Chinese (HK) PEI scoresMean (SD) (95% CI)% Floor% CeilingCronbach’s αPaired t-test, P valueaTest–retest reliability (ICC)Baseline (n = 150)4.65 (2.76) (4.21–5.10)6.03.30.84Follow-up (n = 127)4.22 (3.37) (3.63–4.81)18.93.90.850.1240.75PEI score calculated as mean of answered items times 6, excluding cases that were not applicable or missing in more than three items, range 0–12. Two baseline and three follow-up subjects were excluded because cases answered not applicable in more than three items.aOnly 127 subjects who had valid PEI scores in both baseline and follow-up assessments were included in the test–retest analyses.Table 3 shows the item-scale correlation between each item and the total PEI score by administration time. The correlations were all strong (≥0.65) indicating good internal consistency, in support of the internal construct validity.TABLE 3Item-scale correlations of Chinese (HK) PEIBaseline (n = 76)Follow-up (n = 76)Mean (SD)Correlation with total PEI scoreaMean (SD)Correlation with total PEI scoreaAble to copy with life0.61 (0.66)0.680.55 (0.70)0.70Able to understand your illness0.80 (0.67)0.680.74 (0.76)0.77Able to copy with your illness0.77 (0.70)0.830.74 (0.79)0.78Able to keep yourself healthy0.96 (0.56)0.750.92 (0.80)0.76Confident about your health0.86 (0.71)0.840.89 (0.79)0.79Able to help yourself0.66 (0.60)0.670.76 (0.75)0.72aSpearman’s rho correlation, corrected for overlap, were all significant, P < 0.001.Only one factor with eigenvalue >1 was extracted by factor analysis, which explained >70% and 67% of the reliable variance of the item scores at baseline and follow-up, respectively. Table 4 shows the factor coefficients and correlations between principal component and the PEI items at baseline and follow-up. Factor loadings of all the items were similar, and correlation between principle component and the Chinese (HK) PEI were strong (0.69–0.8), further confirming internal construct validity.TABLE 4Factor coefficients and correlations between principal component and items of Chinese (HK) PEIBaseline (n = 76)Follow-up (n = 76)Factor coefficientsFactor correlationsFactor coefficientsFactor correlationsAble to copy with life0.1830.680.2120.73Able to understand your illness0.1930.710.2330.80Able to copy with your illness0.2310.850.2260.78Able to keep yourself healthy0.2220.820.2180.75Confident about your health0.2270.840.2250.78Able to help yourself0.2130.790.2040.70Eigenvalue3.7003.448Total variance explained (%)61.6757.46Table 5 shows the mean PEI scores by age, sex, and consultation reasons for all subjects. Older (aged 60 years or above) subjects had significantly lower mean PEI scores than younger (aged 18–59 years) subjects, both at baseline and at follow-up. There was no significant difference in mean PEI score with respect to sex or consultation reason. The PEI score on follow-up was significantly lower than that at baseline in subjects who consulted for chronic diseases, but there was no difference for those consulting for acute illness.TABLE 5Mean (SD) of PEI summary score by age, gender and consultation reasonBaseline (n = 150)Follow-up (n = 128)nMeanSDESnMeanSDESAge group (years) 18–59635.64a2.76565.20a2.89 60 or above853.84a2.470.65723.47a3.530.60Gender Male584.662.99494.183.57 Female924.642.620.01794.273.240.03Consultation reason Chronic disease1024.85b2.95874.31b3.41 Acute illness484.222.280.21414.083.290.07ES, effect sizeaSignificant difference in mean PEI score between age group by independent t-test (P value < 0.01).bSignificant difference in mean PEI score for chronic disease between baseline and follow-up by paired t-test (P value = 0.047).DiscussionWe believe that this is the first report in the literature on the validity, reliability and other psychometric properties of the PEI in an Asian population. Such evidence is important to support the cross-cultural application of a patient-reported outcome measure. An equivalent translation is an essential although not sufficient step for the application of an instrument to another population with a different culture and language. In our case, an equivalent Chinese translation could be made for all the PEI items and response options. The term most difficult to translate was enablement in the title as this is a relatively new concept in the collective Chinese culture. Nevertheless, our Chinese patients could understand and identify with the indicators (items) of this construct. This further supports the relevance of the concept of enablement in the Chinese culture, as found by an earlier study that enabling programmes led to beneficial outcomes for Chinese patients with chronic disease.20Our results demonstrated good acceptability of the PEI by Chinese patients with a response rate of 83%, completion rate of 100% and analysable data from 98% of the respondents. The cognitive debriefing interviews showed that all subjects, including elderly ones, could complete the instrument in a few minutes. Most of our subjects were elderly and had a low level of education, which is typical of patients attending public primary care clinics in Hong Kong. We would expect the PEI to be even more acceptable to younger better-educated Chinese patients.Our mean PEI scores were higher than those found in studies conducted in English-speaking patients (3.1) and comparable to studies for patients consulting in other languages (4.5).10 This may be explained by cultural differences in illness coping or response to questionnaires. Howie et al.10 found significantly higher enablement scores in ‘other language patients’, while Freeman et al.29 found higher enablement scores when non-English-speaking patients had consultations in their own language. It may not be valid to directly compare the absolute scores between different cultural groups because each may have different life expectations and coping needs. Qualitative exploration on the meaning of the scores from respondents would shed light on any difference in the interpretation and function of the concept of enablement between different cultures.The proportion of subjects scoring the lowest and highest scores were all <20%, which is generally regarded as the threshold for a floor or ceiling effect,30 suggesting that the Chinese (HK) PEI is potentially sensitive in detecting differences in the outcomes of consultations.The results of the item-scale correlation and factor analyses supported the internal construct validity of the PEI as a measure of the core concept of patient enablement. All items were relevant and important as shown by the high and almost equal correlations with the total score and the principal factor. The high item-score correlation indicated that not only were the original and Chinese (HK) versions of the PEI similar in language (semantics) but there were also equivalent in function.The finding of lower PEI scores in the elderly provided some support to the external construct validity of the PEI, which is consistent with existing Western literature.8 However, there was no difference in PEI score according to reason of consultation contrary to findings in a previous study.17 One would expect higher enablement scores in those presenting with acute illness, as chronic conditions entail regular follow-up, and doctors would assume that patients are stable and have acquired knowledge of their condition over time, thereby resulting in less enablement achieved at each subsequent consultation. On the other hand, doctors might see the need to help patients with chronic diseases to cope better. Pawlikowska et al.15 also did not find any effect from case mix on PEI scores.It should be noted that the findings with relation to those presenting for acute versus chronic conditions need to be interpreted with caution since the ‘acute illness’ group was relatively small. Our study patients presenting with ‘acute’ conditions were instructed to focus on their original acute problem only; however, there is a chance that they may have found it difficult to separate acute from chronic illness, had they also had chronic illnesses. In addition, it may also have been possible that they had separate consultations (other than those included in the baseline) that might have affected their responses. Nevertheless, we expected that the number of patients presenting for re-consultation in the 2 weeks of the study would have been low.There was no difference in the scores or psychometric performance of the PEI according to timing of administration (immediately post-consultation compared with delayed for 2 or more weeks). This not only supported test–retest reliability but also the feasibility of measuring patient enablement with the PEI by telephone up to a few weeks after the consultation. The PEI was originally developed to be administered immediately post-consultation.5 Our results indicate that a ‘delayed’ completion of the PEI may be used instead, with beneficial implications for its usage. For patients, they would not have to spend more time at the clinic in addition to their consultation time duration. For clinical practice and research, the option of delayed administration of the PEI over the telephone (or via other means) confers a definite logistical advantage.It was interesting to note that a higher proportion of subjects had a floor PEI score in the assessment at 2 or more weeks, which might imply the enablement perceived immediately after the consultation was lost with time. This was more apparent in those consulting for chronic diseases whose mean PEI score also decreased with time. This more likely represented a true change rather than poor reproducibility. If it were a systematic bias, then the same change should have been found in all subjects. It is not possible to ascertain whether immediate post-consultation, compared with delayed PEI administration, represented a more accurate reflection of the ‘truth’, but conceptually, it would take time before a patient could really tell whether he/she had been more enabled. In addition, some patients might feel obligated to give a more positive response when they are interviewed immediately after the consultation, especially if they have a good relationship with the doctor.LimitationsThis pilot study intended to establish the cross-cultural applicability of the PEI in a Chinese population. The results provide preliminary evidence on its internal construct validity and reliability; however, its construct validity in relation to external criteria has not been firmly established. Our findings on the differences between known groups were inconclusive due to limitations of the sample size and the study setting. There are certain factors documented in the literature to be associated with patient enablement, such as patient age and sex, doctor’s communication, length of consultation, how well the doctor and the patient know each other (proxy for continuity of care) and overall satisfaction of the visit, which were not explored due inherent difficulties in our primary care clinic. For instance, in our setting, patients are usually not seen by the same doctor, and consultation lengths are fairly standard (around 8 minutes). Further studies should be carried out to test the construct validity of the PEI and to further examine determinants of patient enablement.Conclusion(s)The Chinese (HK) version of the PEI is semantically equivalent to the original English language PEI, and our results have supported its acceptability, content and internal construct validity, reliability and sensitivity. As our subjects are of different ages and sex, education levels and social classes, we propose that the PEI is valid for use with the general Hong Kong Chinese population. As all Chinese share the same written language (be it in traditional or simplified form), we hope in the future it can be applied to other Chinese populations worldwide. Further studies should be carried out to evaluate the relationship between PEI scores and other external criteria to confirm its construct validity. Sensitivity in differentiating consultation outcomes in different settings or morbidity groups and responsiveness to changes with intervention will also need to be established.DeclarationFunding: Tung Wah Group of Hospitals.Ethical approval: HKU/HA HKW IRB reference number, UW 07-226.Conflict of interest: none.We would like to thank Professor John Howie in giving us permission to adapt the PEI to our Chinese population. Thanks also go to Mr Carlos Wong and Ms Mandy Tai for helping with the data collection and analysis.TRIAL REGISTRATION: The research was registered with the HKU Clinical Trial Centre Registry (HKCTR-477).Appendix 1. Draft Chinese (HK) PEI with back-translation used in cognitive debriefing with English back-translations in brackets1StewartMBrownJBWestonWWPatient-Centred Medicine. Transforming the Clinical Method1995Thousand Oak, CASage2LittlePEverittHWilliamsonIPreferences of patients for patient-centred approach to consultation in primary care: observational studyBMJ20013224683LittlePEverittHWilliamsonIObservational study of effect of patient centeredness and positive approach to outcomes of general practice consultationsBMJ2001323908114HowieJGRHeaneyDMaxwellMQuality, core values and general practice consultations: issues of definition, measurement and deliveryFam Pract200421458685HowieJGHeaneyDJMaxwellMWalkerJJA comparison of a Patient Enablement Instrument (PEI) against two established satisfaction scales as an outcome measure of primary care consultationsFam Pract199815165716StottNCHDavisRHThe exceptional potential in each primary care consultationJ R Coll Gen Pract19792920157Di BlasiZHarknessEErnstEGeorgiouAKleijnenJInfluence of context effects on health outcomes: a systematic reviewLancet2001357757808AdzicZOKaticMKernJPatient, physician and practice characteristics related to patient enablement in general practice in Croatia: cross-sectional survey studyCroat Med J200848813239HowieJGHeaneyDJMaxwellMMeasuring quality in general practice. Pilot study of needs, process and outcome measureOccas Pap R Coll Gen Pract19977513210HowieJGHeaneyDJMaxwellMQuality at general practice consultations: cross sectional surveyBMJ19993197384311WensingMvan LieshoutJJungHPHermsenJRosemannTThe Patients Assessment Chronic Illness Care (PACIC) questionnaire in The Netherlands: a validation study in rural general practiceBMC Health Serv Res2008818212HowieJGRAddressing the credibility gap in general practice research: better theory; more feeling; less strategyBr J Gen Pract1996464798113McKinstryBWalkerJBlaneyDHeaneyDBeggDDo patients and expert doctors agree on the assessment of consultation skills? A comparison of two patient consultation assessment scales with the video component of the MRCGPFam Pract200421758014MercerDWNeumannMWirtzMFitzpatrickBVojtGGeneral practitioner empathy, patient enablement, and patient-reported outcomes in primary care in an area of high socio-economic deprivation in Scotland—A pilot prospective study using structural equation modellingPatient Educ Couns200873240515PawlikowskaTRBNowakPRSzumilo-GrzesikWWalkerJJPrimary care reform: a pilot study to test the evaluative potential of the Patient Enablement Instrument in PolandFam Pract20021919720116PriceSMercerSMacPhersonHPractitioner empathy, patient enablement and health outcomes: a prospective study of acupuncture patientsPatient Educ Couns2006632384517MeadNBowerPRolandMFactors associated with enablement in general practice: cross-sectional study using routinely-collected dataBr J Gen Pract2008583465218MercerSWFitzpatrickBGourlayGMore time for complex consultations in a high-deprivation practice is associated with increased patient enablementBr J Gen Pract200757960619HaughneyJCottonPRosenJPThe use of a modification of the Patient Enablement Instrument in asthmaPrim Care Respir J200716899220SiuAMChanCCPoonPKChuiDYChanSCEvaluation of the chronic disease self-management program in a Chinese populationPatient Educ Couns2007654250Epub 2006 Jul 2621LeungGMWongIOLChanWSChoiSLoSVHealth Care Financing Study Group. The ecology of health care in Hong KongSoc Sci Med2005615779022Population By-census Office, Census and Statistics Department, The Government of the Hong Kong Special Administrative RegionHong Kong Statistics, Population and Vital Events [Internet]Hong Kong[updated 2006; cited 2009 Jun 23]. Available from http://www.censtatd.gov.hk/23SmithDHDixonASLamCLKLamTPSmithDHHong Kong patients’ preferences for physician communication styleWorking Papers on Health Communication and ChinaHong Kong, ChinaThe Intercultural Communication Research Programme of The David C Lam Institute for East-West Studies24WildDGroveAMartinMPrinciples of Good Practice for the Translation and Cultural Adaptation Process for Patient-Reported Outcomes (PRO) Measures: report of the ISPOR Task Force for Translation and Cultural AdaptationValue Health200589410425BartleyMHealth Inequality: an Introduction to Theories, Concepts and Methods2004Cambridge, UKPolity Press3126LamTHKleevansWLWongCMDoctor consultation in Hong Kong: a comparison between findings of a telephone interview with general household surveyCommunity Med198810175927TerweeCBBotSDde BoerMRQuality criteria were proposed for measurement properties of health status questionnairesJ Clin Epidemiol200760344228NunnallyJCPsychometric Theory19943rd ednNew YorkMcGraw Hill29FreemanGKRaiHWalkerJJNon-English speakers consulting with the GP in their own language: a cross-sectional surveyBr J Gen Pract20025236830McHorneyCAWareJELuRSherboumeCDThe MOS. 36-Item Short-Form health survey (SF-36): III. Tests of data quality, scaling assumptions and reliability across diverse patient groupsMed Care1994324066
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B48C07191826EBBDBADD0B5B2A899A764E75367.txt b/test/dataset/in/resources/corpus/Clean_0B48C07191826EBBDBADD0B5B2A899A764E75367.txt
deleted file mode 100644
index 01a27de..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B48C07191826EBBDBADD0B5B2A899A764E75367.txt
+++ /dev/null
@@ -1 +0,0 @@
-carcincarcinCarcinogenesis1460-21800143-3334Oxford University Press10.1093/carcin/bgp132CARCINOGENESISDisruption of estrogen receptor signaling enhances intestinal neoplasia in ApcMin/+ miceClevelandAlicia G.†OikarinenSeija I.1†BynotéKimberly K.MarttinenMaija1RafterJoseph J.23GustafssonJan-Åke23RoyShyamal K.4PitotHenry C.5KorachKenneth S.6LubahnDennis B.78910MutanenMarja1GouldKaren A.*Department of Genetics, Cell Biology and Anatomy, University of Nebraska Medical Center, Omaha, NE 68198, USA1Department of Applied Chemistry and Microbiology (Nutrition), PO Box 66, University of Helsinki, Helsinki FIN-00014, Finland2Department of BioSciences and Nutrition, Karolinska Institute, NOVUM, S-14186 Huddinge, Sweden3Center for Nuclear Receptors and Cell Signaling, Department of Biology and Biochemistry, University of Houston, Houston, TX 77204, USA4Department of Cellular and Integrative Physiology, University of Nebraska Medical Center, Omaha, NE 68198, USA5McArdle Laboratory for Cancer Research, University of Wisconsin, Madison, WI 53706, USA6Receptor Biology Section, National Institute of Environmental Health Sciences, National Institutes of Health, Research Triangle Park, NC, USA7Department of Biochemistry8Department of Child Health9Department of Animal Sciences10MU Center for Phytonutrient and Phytochemical Studies, University of Missouri, Columbia, MO, USA*To whom correspondence should be addressed. Tel: +1 402 559 2456; Fax: +1 402 559 7328; Email: kagould@unmc.edu†These authors contributed equally to this work.92009116200930915811590611200830420092352009© The Author 2009. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org2009Estrogen receptors (ERs) [ERα (Esr1) and ERβ (Esr2)] are expressed in the human colon, but during the multistep process of colorectal carcinogenesis, expression of both ERα and ERβ is lost, suggesting that loss of ER function might promote colorectal carcinogenesis. Through crosses between an ERα knockout and ApcMin mouse strains, we demonstrate that ERα deficiency is associated with a significant increase in intestinal tumor multiplicity, size and burden in ApcMin/+ mice. Within the normal intestinal epithelium of ApcMin/+ mice, ERα deficiency is associated with an accumulation of nuclear β-catenin, an indicator of activation of the Wnt–β-catenin-signaling pathway, which is known to play a critical role in intestinal cancers. Consistent with the hypothesis that ERα deficiency is associated with activation of Wnt–β-catenin signaling, ERα deficiency in the intestinal epithelium of ApcMin/+ mice also correlated with increased expression of Wnt–β-catenin target genes. Through crosses between an ERβ knockout and ApcMin mouse strains, we observed some evidence that ERβ deficiency is associated with an increased incidence of colon tumors in ApcMin/+ mice. This effect of ERβ deficiency does not involve modulation of Wnt–β-catenin signaling. Our studies suggest that ERα and ERβ signaling modulate colorectal carcinogenesis, and ERα does so, at least in part, by regulating the activity of the Wnt–β-catenin pathway.IntroductionColorectal cancer incidence is 35% lower in women than men (1). Although the basis for this difference is unknown, gonadal steroids are considered a contributing factor. In women, high parity, early age at first pregnancy, oral contraceptive use and estrogen replacement therapy are each associated with decreased colorectal cancer risk (2–4). In the Women's Health Initiative study, hormone replacement therapy conferred a 40% reduction in colorectal cancer risk (5). Altogether, these data suggest that estrogens and/or progestins protect against colorectal cancer.Consistent with the hypothesis that estrogens protect against colorectal cancer, administration of either a synthetic estrogen or the naturally occurring 17β-estradiol (E2) inhibits carcinogen-induced gastrointestinal tumorigenesis in rats (6,7). Similarly, E2 attenuates intestinal tumorigenesis in ApcMin/+ mice (8,9), in which inheritance of the multiple intestinal neoplasia (Min) mutation in the adenomatous polyposis coli (Apc) gene predisposes carriers to spontaneous intestinal tumorigenesis (10,11). The effects of estrogens are mediated by estrogen receptors (ERs) [ERα (Esr1) and ERβ (Esr2)], which function as ligand-dependent transcription factors. Both receptors are expressed in the intestinal epithelium (12–15), but their function in this tissue is not known. Loss of ERα expression due to promoter hypermethylation occurs in colorectal cancers in both men and women (13), suggesting that ERα functions as a tumor suppressor in the colon. ERα promoter hypermethylation occurs in normal colonic epithelium during aging (13,16) and is postulated to create a microenvironment permissive to carcinogenesis. In contrast, loss of ERβ expression correlates with advanced Dukes stages (17).Some support for the hypothesis that ERα and ERβ impact intestinal tumorigenesis has been obtained from studies using ApcMin/+ mice. In one study, heterozygosity for an ERα knockout allele (Esr1Pcn1) was associated with increased incidence of colon tumors in ApcMin/+ mice (18). The impact of homozygosity for this ERα knockout allele (ERα−/−) on tumorigenesis in ApcMin/+ mice could not be evaluated because no ERα−/− mice carrying the ApcMin/+ mutation were obtained from the intercross. Although two studies have examined the impact of an ERβ knockout allele (Esr2Pcn1) on tumorigenesis in ApcMin/+ mice, the results of these studies are strikingly different. In one study, heterozygosity and homozygosity for the ERβ knockout allele were associated with increased incidence of colon tumors (18). In the second study, this same ERβ knockout allele had no impact on colon tumorigenesis (19). Rather, these investigators observed that homozygosity for the ERβ knockout was associated with an increase in small intestinal tumor number in ApcMin/+ female but not male mice (19). Altogether, these results do support the idea that disruption of ERα and ERβ signaling promotes intestinal tumorigenesis in ApcMin/+ mice. However, because a crucial genotypic class was absent in the ERα study and the two ERβ studies yielded disparate results, the nature of the impact of ERα and ERβ signaling on intestinal tumorigenesis remains largely unknown.In the present study, we sought to clarify the role of ER signaling in tumorigenesis in the ApcMin/+ mouse by crossing targeted disruptions of ERα and ERβ that are distinct from those used in the previous studies described above with the ApcMin strain. We intercrossed these ERα and ERβ knockout strains with the ApcMin strain and evaluated their impact on intestinal tumorigenesis. The impact of these knockouts on the activity of the Wnt–β-catenin pathway in ApcMin/+ mice was also evaluated. Our data indicate that loss of ERα is associated with increased adenoma multiplicity in ApcMin/+ mice, whereas loss of ERβ increases the incidence of colon tumors in ApcMin/+ mice without significantly altering tumor multiplicity. Loss of ERα but not ERβ is associated with activation of Wnt–β-catenin signaling in the intestinal epithelium of ApcMin/+ mice. These studies suggest that both ERα and ERβ modulate colorectal carcinogenesis, and ERα does so, at least in part, by regulating the Wnt–β-catenin pathway.Materials and methodsCare and treatment of miceThe ERα cross and the second ERβ cross were conducted at the University of Nebraska Medical Center. The Institutional Animal Care and Use Committee of the University of Nebraska Medical Center approved all procedures involving live animals in these studies. B6 ApcMin/+ mice were obtained from The Jackson Laboratory (Bar Harbor, ME). The ERα knockout strain (B6.129-Esr1tm1Ksk) (20) was provided by D.Lubahn. The ERβ knockout strain (B6.129-Esr2tm1Ksk) (21) was obtained from Taconic (Hudson, NY). Animals were housed under controlled temperature, humidity and 12 h light–12 h dark lighting conditions in a facility accredited by the American Association for Accreditation of Laboratory Animal Care and operated in accordance with the standards outlined in Guide for the Care and Use of Laboratory Animals (The National Academies Press, 1996). Mice were provided Harlan irradiated rodent diet 7904 (Harlan Teklad, Madison, WI), which contains soy, milk and meat-based protein sources and allowed to feed ad libitum.B6 females heterozygous for targeted disruption of the ERα gene (ERα+/−) were crossed to B6 ApcMin/+ males. The resulting ERα+/− ApcMin/+ males were crossed to B6 ERα+/− females to produce ERα−/−, ERα+/− and ERα+/+ ApcMin/+ mice. Similar crosses were performed with the ERβ knockout. The ERα and Apc genotypes were determined by polymerase chain reaction (PCR) as described previously (20,22). ERβ genotype was determined by PCR using neo primers (20) and primers that amplify ERβ exon 3 (forward primer 5′-AGAGACCCTGAAGAGGAAGC-3′ and reverse primer 5′-GCTTCTTTTAAAAAAGGCCT-3′) only in the absence of the neo insert.The first ERβ cross was conducted at the University of Helsinki. The Laboratory Animal Ethics Committee of the Faculty of Agriculture and Forestry, University of Helsinki, approved this study protocol. ApcMin/+ mice on a C57BL/6J (B6) genetic background were obtained from The Jackson Laboratory. Mice from the B6.129-Esr2tm1Ksk strain (21) that was heterozygous for the ERβ knockout (ERβ+/−) mice were serially backcrossed to B6 mice. Mice were genotyped at Apc and ERβ using PCR-based assays as described previously (22,23).After 10 generations of backcrossing to B6 (N10), ERβ+/− females were intercrossed with ApcMin/+ males to generate ApcMin/+ ERβ+/− mice. Male ApcMin/+ ERβ+/− mice were then bred with female ERβ+/− mice to obtain ApcMin/+ ERβ+/+, ApcMin/+ ERβ+/− and ApcMin/+ ERβ−/− offspring. All mice were fed with Teklad Global 14% Protein Rodent Maintenance Diet 2014 (Harlan Teklad, Horst, The Netherlands), which contains 14% protein, 3.5% oil, 5.5% fiber and no alfalfa or soybean meal. In this diet, the level of natural phytoestrogens is negligible. Animals were housed in plastic cages in a temperature- and humidity-controlled animal facility, with 12 h light–dark cycle. Food and water were provided ad libitum.Analysis of tumor multiplicity, size and histologyMice were euthanized by CO2 asphyxiation. Intestinal tumors were counted and measured as described previously (10,24–26) except that the small intestine was cut into four segments of equal length and the entire length of these segments, along with the colon, was examined. In Omaha crosses, segments of intestinal tissue fixed in 10% neutral buffered formalin were processed, sectioned and stained with hematoxylin and eosin. Stained sections were examined by an observer blinded with respect to the ER genotype. For intestinal tissue used for extraction of RNA or protein, adenomas were excised under the stereomicroscope at ×10 magnification. The remaining normal epithelium was scraped from the submucosal smooth muscle, frozen in liquid nitrogen and stored at −70°C.Quantitative reverse transcription–PCRTotal RNA was prepared using the Absolutely RNA miniprep kit (Stratagene Corporation, La Jolla, CA). Reverse transcription reactions were performed as described previously (27). For Cyclin D1 and c-Myc complementary DNA, PCR was performed using 900 nM each primer and 200 nM probe. The primer and probes sequences were as follows: 5′-TCCGCAAGCATGCACAGA-3′ (Cyclin D1 forward primer); 5′-CTTTGTGGCCCTCTGTGCCACAGA-3′ (Cyclin D1 probe); 5′-GGTGGGTTGGAAATGAACTTCA-3′ (Cyclin D1 reverse primer); 5′-TGAGCCCCTAGTGCTGCAT-3′ (c-Myc forward primer); 5′-AGGAGACACCGCCCACCACCAG-3′ (c-Myc probe) and 5′-CCTCATCTTCTTGCTCTTCTTCAGA-3′ (c-Myc reverse primer). PCR was performed and analyzed using the ABI 7500 Real-Time PCR System and Sequence Detection Software version 1.9 (Applied Biosystems, Foster City, CA) as described previously (27). Transcript abundance was normalized to glyceraldehyde 3′-phosphate dehydrogenase (Gapdh) (27). Neither the ApcMin mutation nor the ERα deficiency impacted Gapdh expression (data not shown).Determination of genotype at polymorphic markersOligonucleotide primers specific to simple sequence length polymorphism repeats polymorphic in length between the 129P2 and the B6 strains were used for genotyping as described previously (27,28).RadioimmunoassayE2 and progesterone in unextracted sera collected at killing were measured using a coated tube radioimmunoassay (Diagnostic Products Corp., Los Angeles, CA) as described previously (29). Samples were assayed in duplicate. For ERα+/+ Apc+/+ and ERα+/+ ApcMin/+ mice, E2 measurements were obtained during the follicular phase of estrous and progesterone measurements were obtained during the luteal phase. Phase of estrous was determined based on ovarian histology. ERα−/− females do not cycle normally and are anovulatory; E2 measurements were made in these mice without selection for phase of estrous.ImmunoblottingMucosal scrapings were isolated in ice-cold phosphate-buffered saline supplemented with protease and phosphotase inhibitors. Nuclear and cytoplasmic extracts from fresh tissue were prepared using the Nuclear Extract kit (Active Motif, Carlsbad, CA) according to the manufacturer's instructions. Whole-cell extracts were prepared from frozen tissue and the same kit using manufacturer's instructions. Western blots were performed as described previously (30) except that the Li-Cor Odyssey Imaging system (Li-Cor, Lincoln, NE) was used. The antibodies used were as follows: E-5 β-catenin and MS-3 nucleolin mouse monoclonal antibodies and I-19 actin antibody (Santa Cruz Biotechnology, Santa Cruz, CA); α-tubulin mouse monoclonal antibody, clone B-5-1-2 (Sigma Corporation, St. Louis, MO) and secondary antibodies labeled either with Alexa fluor® 680 (Invitrogen Corporation, Carlsbad, CA) or with IRDye800™ (Rockland Immunochemicals, Gilbertsville, PA). Primary and secondary antibodies were used at a dilution of 1:1000 and 1:2000, respectively.Quantitative assay for Apc locusDNA was isolated from normal intestinal epithelium scrapings. The quantitative PCR assay for the Apc locus was performed and analyzed as described previously (31). For this assay, the Apc+ and ApcMin alleles are amplified and digested with HindIII, resulting in a 123 bp product from the Apc+ allele and a 144 bp product from the ApcMin allele. Radiolabeled products were resolved on a 5% denaturing polyacrylamide gel and quantitated using a Molecular Dynamics PhosphorImager and ImageQuaNT 5.0 software (GE Healthcare, Piscataway, NJ).Statistical analysisAverage tumor multiplicity, tumor diameter and tumor burden were compared using the Wilcoxon rank sum test. Differences in the distribution of tumors along the intestine were evaluated with the chi-square test. Tumor incidence data were compared using Fisher's exact test. Differences in gene expression were evaluated using a t-test. All statistical analyses were conducted using SPSS software, version 13 (SPSS, Chicago, IL). P ≤ 0.05 was considered to be significant.ResultsERα deficiency enhances ApcMin-induced tumorigenesisTo examine the impact of disrupting ERα in ApcMin/+ mice, we intercrossed the ERα knockout and ApcMin mouse strains. All genotypes of F2 progeny were obtained at the expected frequencies (P > 0.05; data not shown). Mice were killed at 80 days of age. Mean tumor multiplicity in ERα+/+ ApcMin/+ mice was 37.9 (Figure 1A). Mean tumor multiplicity in ERα+/− ApcMin/+ mice was 37.1 and did not differ from that in ERα+/+ ApcMin/+ mice (P > 0.05; Figure 1A). Relative to ERα+/+ ApcMin/+ and ERα+/− ApcMin/+ mice, the mean adenoma multiplicity in ERα−/− ApcMin/+ mice was increased by 50% to 56.4 (P = 0.01; Figure 1A). Gender did not impact tumor multiplicity (Figure 1B; P > 0.05). ERα genotype did not alter the distribution of adenomas along the length of the intestine (Figure 1C; P > 0.05).Fig. 1.Disruption of ERα leads to enhanced intestinal tumorigenesis in ApcMin mice. The mean adenoma multiplicity in the intestinal tract of ERα+/+ ApcMin/+ (n = 22), ERα+/− ApcMin/+ (n = 17) and ERα−/− ApcMin/+ mice (n = 16) at 80 days of age is illustrated (A). The same data are illustrated in panel (B), except that the mean multiplicity in females (F) is shown separately from that in males (M). In panel (C), the mean percentage of adenomas located within each of the four defined regions of the small intestine (SI) and the colon (C) is shown. For the SI, the regions are depicted proximal to distal, with region 1 beginning just distal of the stomach and region 4 ending immediately proximal of the cecum. The mean adenoma burden in the intestinal tract of ERα+/+ ApcMin/+ (n = 22), ERα+/− ApcMin/+ (n = 17) and ERα−/− ApcMin/+ mice (n = 16) at 80 days of age is illustrated in panel (D). The ‘1’ indicates a statistical difference from ERα+/+ ApcMin/+ and the ‘2’ indicates a statistical difference from ERα+/− ApcMin/+.To examine the effect of ERα deficiency on tumor size, we focused on small intestinal adenomas because these lesions are flat; therefore, tumor diameter and/or tumor area are used as indicators of tumor size (25). The mean tumor area in ERα+/+ ApcMin/+ mice did not differ significantly from that in ERα+/− ApcMin/+ mice (P > 0.05). In contrast, the average tumor area in ERα−/− ApcMin/+ mice was significantly larger than that in either ERα+/+ or ERα+/− mice (P ≤ 0.001). Combining tumor multiplicity and tumor size data, we calculated tumor burden. The average tumor burden in ERα+/+ ApcMin/+ mice averaged 70 ± 23 mm2. The mean tumor burden in ERα+/− ApcMin/+ mice of 81 ± 14 mm2 did not differ significantly from that in ERα+/+ ApcMin/+ mice (P > 0.05; Figure 1D). In contrast, the mean tumor burden of 143 ± 28 mm2 in ERα−/− ApcMin/+ was significantly greater than that in ERα+/+ and ER+/− mice (P < 0.005; Figure 1D). ERα genotype did not impact adenoma histology in ApcMin/+ mice (Figure 2A and B).Fig. 2.Impact of ERα deficiency on intestinal adenoma and ovarian histology in ApcMin/+ mice. Representative examples of fairly well-differentiated polypoid adenomas from ERα+/+ ApcMin/+ (A) and ERα−/− ApcMin/+ (B) mice. A healthy ovary from an 80-day-old ERα+/+ Apc+/+ female showing multiple functional corpora lutea (indicated by *) and healthy maturing follicles (indicated by arrows) are shown in panel (C). By comparison, a representative ovary from an age-matched ERα+/+ ApcMin/+ female (D) exhibits few corpora lutea (*) and those that are present show signs of diminished function and steroidogenesis. The majority of maturing follicles here show evidence of degeneration (arrows). A quantitative analysis of the mean number (±SD) of corpora lutea (E) and mean serum progesterone (F) is also shown. Corpora lutea were counted in both ovaries from Apc+/+ (n = 6) and ApcMin/+ mice (n = 12) during the luteal phase of the estrous cycle. Serum progesterone was measured by radioimmunoassay during the luteal phase. The ‘1’ indicates a statistical difference from wild-type. The ovary from an ERα−/− ApcMin/+ mouse showing portions of two large, hemorrhagic, cystic follicles is also shown (panel G). The edge of one follicle, showing what appears to be a single layer of granulosa cells and underlying ovarian stromal cells, is defined by the box with the dashed line and is shown at higher magnification in the inset. The arrow indicates the stromal cells. In panel (H), the ovary from an ERα−/− Apc+/+ mouse shows portions of four large, hemorrhagic and cystic follicles. A portion of the hypertrophied thecal layer lining one follicle, defined by the box with the dashed line, is shown at higher magnification in the inset. The arrowhead indicates the thecal layer and the arrow indicates ovarian stromal cells. The scale bar represents 100 or 44 μM (inset panels).Impact of ERα deficiency on E2 levels and ovarian histology in ApcMin/+ miceERα deficiency in female Apc+/+ mice is associated with a 10-fold elevation in serum E2 by 60–120 days of age (32). These observations raised the possibility that ERα deficiency might modulate tumorigenesis via an unknown mechanism driven by high levels of E2. Therefore, we examined serum E2 levels in ERα−/− ApcMin/+ females. We observed that in 80-day-old ERα−/− ApcMin/+ females, the mean serum E2 concentration was 32 ± 7 pg/ml. This value is at the low end of the normal physiological range of E2 in female mice, which varies from 20 to 100 pg/ml depending upon the phase of estrous. Thus, the enhanced intestinal tumorigenesis in female ERα−/− ApcMin/+ mice is not a consequence of high serum E2.We postulated that ERα−/− ApcMin/+ females might exhibit low levels of serum E2 as a result of deleterious effects of the ApcMin mutation on ovarian function. In the estrus phase, when E2 levels are highest, the mean concentration of serum E2 in 80-day-old ERα+/+ ApcMin/+ mice is <20 pg/ml, the lower limit of detection of the assay used. In contrast with ERα+/+ Apc+/+ females (Figure 2C), ERα+/+ ApcMin/+ females at 80 days of age possess ovaries exhibiting a high degree of follicular degeneration (Figure 2D) and few corpora lutea (Figure 2E; P ≤ 0.05). ERα+/+ ApcMin/+ females also exhibit low-serum progesterone levels (Figure 2F; P ≤ 0.05), indicating reduced ovarian function. Although ovaries in ERα−/− ApcMin/+ mice showed the enlarged, hemorrhagic and cystic follicles (Figure 2G) characteristic of ERα−/− Apc+/+ females [Figure 2H; (32)], these follicles in ERα−/− ApcMin/+ females did not show the hypertrophied theca, indicative of leutinizing hormone stimulation, frequently seen in older ERα−/− Apc+/+ females.Potential impact of 129-derived loci in the ERα knockout cross?We genotyped our B6.129-Esr1tm1Ksk founders using markers distributed across the genome (Table I) to identify residual 129P2 alleles even though this strain had been backcrossed for 12 generations. In these founders, the only 129P2 alleles detected were at three markers linked to ERα on proximal chromosome 10. The 129P2 alleles at these markers were segregating in the F2 intercross mice (Figure 3). Although 70% of ERα−/− ApcMin/+ mice retained the 129 allele at D10Mit152, the marker most closely linked to ERα, linkage disequilibrium dropped off sharply distal of this marker. Genotype at D10Mit51, the next marker distal of D10Mit152, exhibited poor concordance with ERα genotype. Only 10% of ERα−/− mice were homozygous for the 129P2 allele at D10Mit51, indicating that the 129P2 allele at this marker is not in disequilibrium with the 129P2-derived ERα knockout allele. All mice evaluated were homozygous for the B6 allele at markers including and distal to D10Mit87. Thus, <13 Mbp from the 129P2 genome on chromosome 10 is retained in association with the ERα knockout.Table I.Markers used in genotypic evaluation of B6.129-Esr1tm1Ksk and B6.129-Esr2tm1Ksk strainsaMarker (location in cMb; Mbpc)D1Mit236 (26; 45)D10Mit95 (51; 92)D1Mit440 (54; 91)D10Mit233 (62; 114)D1Mit159 (82; 161)D11Mit71 (1; 7)D1Mit111 (92; 171)D11Mit4 (37; 68)D1Mit206 (96; 175)D11Mit333 (66; 108)D1Mit426 (101; 182)D12Mit182 (2; 11) (ERβ)dD2Mit61 (34; 61)D12Mit60 (16; 35)D2Mit395 (70; 119)D12Mit91 (29; 73) ERβ)D2Mit148 (105; 175)D12Mit118 (45; 92) (ERβ)D3Mit203 (11; 27)D12Nds2 (59; N.D.e)D3Mi98 (40; 86)D13Mit275 (16; 37)D3Mit320 (72; 143)D13Mit13 (35; 56)D4Mit193 (7; 32)D13Mit151 (71; 116)D4Mit17 (31; 63)D14Mit60 (15; 46)D4Mit308 (57; 124)D14Mit39 (30; 68)D5Mit348 (8; 24)D15Mit252 (15; 23)D5Mit201 (45; 75)D15Mit107 (49; 84)D5Mit95 (68; 125)D15Mit161 (69; 97)D6Mit138 (1; 4)D16Mit131 (4.3;) (ERβ)D6Mit284 (38; 93)D16Mit101 (17; 24)D6Mit373 (74; 147)D16Mit139 (43; 66)D7Mit267 (11; 29)D17Mit245 (3; 12)D7Mit248 (28; N.D.e)D17Mit51 (23; 43)D7Mit323 (50; 100)D17Mit10 (20; N.D.e) [ERα]D8Mit155 (1; 5)D17Mit180 (29; 50) [ERα]D8Mit292 (19; 36)D17Mit93 (45; 74)D8Mit45 (40; 90)D18Mit177 (20; 41)D9Mit90 (9; 32)D18Mit144 (57; 86)D9Mit129 (26; 43)D19Mit96 (15; 22)D9Mit350 (61; 111)D19Mit90 (41; 42)D10Mit152f (4; 12) [ERα]gDXMit64 (31; 73)D10Mit213f (11; 20)DXMit223 (73; 163)D10Mit36 (29; 48) [ERα]aERα (Esr1) is located at 12cM and 5.7 Mbp and ERβ (Esr2) is located at 33cM and 77 Mbp (www.informatics.jax.org).bLocation of marker in centiMorgans obtained from mouse genome informatics (www.informatics.jax.org).cLocation of marker in mega base pairs obtained from mouse genome informatics (www.informatics.jax.org).d(ERβ) indicates that this marker was evaluated in mice derived from B6.129-Esr2tm1Ksk strain.eN.D., location of marker on physical map has not been determined.fThese markers were mapped proximal to Esr1 but the physical mapping places these markers distal to Esr1. Genetic and physical mapping data obtained from mouse genome informatics. 129 alleles at these markers were segregating in N12 mice from the B6.129-Esr1tm1Ksk strain.g(ERα) indicates that this marker was evaluated in mice derived from B6.129-Esr1tm1Ksk strain.Fig. 3.A small region of the 129P2 genome is retained in association with the ERα knockout allele. The genotypes of ERα−/− ApcMin/+ mice (n = 10), ERα+/− ApcMin/+ mice (n = 7) and ERα+/+ ApcMin/+ mice (n = 7) on chromosome 10 are shown. The white bars represent homozygosity for the 129P2 allele, the shaded bars represent heterozygosity for the 129P2 allele and the black bars represent homozygosity for the B6 allele. The position of each marker in Mbp is indicated in parentheses.ERα deficiency is associated with an accumulation of nuclear β-catenin and increased expression of Wnt–β-catenin target genesThe Wnt–β-catenin pathway is constitutively activated in virtually all colon cancers. Although the impact of estrogen on Wnt–β-catenin activity in the intestine is not known, estrogens do influence this pathway in other tissues (33–36). These data, together with our observations that ERα deficiency enhanced intestinal tumorigenesis, led us to postulate that ERα deficiency might activate Wnt–β-catenin signaling in the intestinal epithelium of ApcMin/+ mice. In whole-cell extracts, β-catenin expression was higher in normal intestinal epithelium of ERα−/− ApcMin/+ mice relative to ERα+/+ ApcMin/+ mice (Figure 4A). Nuclear accumulation of β-catenin, a molecular indicator of activation of Wnt–β-catenin signaling, was also evaluated. Quantitative analysis of these western blots indicated that nuclear β-catenin is increased by, on average, 60% in intestinal epithelium of ERα−/− ApcMin/+ mice relative to ERα+/+ ApcMin/+ mice (Figure 4B), consistent with an increased activation of Wnt–β-catenin signaling in the absence of ERα. Expression of Cyclin D1 and c-Myc, two transcriptional targets of the Wnt–β-catenin pathway, was increased on average 2.2-fold and 2.4-fold, respectively, in the intestinal epithelium of ERα−/− ApcMin/+ mice relative to ERα+/+ ApcMin/+ mice (Figure 4C), lending further support to the hypothesis that ERα deficiency is associated with increased Wnt–β-catenin signaling.Fig. 4.ERα deficiency activates the Wnt–β-catenin pathway in the intestinal epithelium of ApcMin/+ mice. β-Catenin expression in whole-cell extracts (A) and nuclear (N) and cytoplasmic (C) protein extracts (B) prepared from normal intestinal epithelium of 80-day-old ERα+/+ ApcMin/+ (n = 6) and ERα−/− ApcMin/+ (n = 6) mice. Actin expression was used as a positive and loading control. Nucleolin was used as a positive control for the nuclear fraction and a negative control for the cytoplasmic fraction, whereas α-tubulin was used as a positive control for the cytoplasmic fraction and a negative control for the nuclear fraction. Cyclin D1 and c-Myc expression was determined using quantitative reverse transcription–PCR (C). Both male and female mice were used. For each sample, expression of Cyclin D1 and c-Myc was normalized to Gapdh expression. The average level of normalized Cyclin D1 and c-Myc expression in the ERα+/+ ApcMin/+ mice (n = 4) was set at 1.0 expression units. The relative expression level in ERα−/− ApcMin/+ mice (n = 7) was expressed as a percentage of that in ERα+/+ ApcMin/+ mice. Each bar represents the mean gene expression ± SEM. The asterisk indicates a statistically significant difference from ERα+/+ ApcMin/+. The mean Apc+/ApcMin ratio values in normal intestinal epithelium of 80-day-old ERα+/+ ApcMin/+ (n = 6) and ERα−/− ApcMin/+ (n = 6) mice is show (D). Each bar represents the mean ratio value ± SD.To verify that the normal intestinal epithelium specimens used in these analyses did not contain microadenomas, small precursors of adenomas that have lost wild-type Apc allele (Apc+) and thus exhibit increased nuclear β-catenin (37), we used a quantitative assay to determine the relative abundance of the Apc+ and ApcMin alleles in genomic DNA isolated from these specimens. It has been reported that in histologically normal intestinal epithelium of B6 ApcMin/+ mice, the mean Apc+:ApcMin ratio value was 0.83 ± 0.18 (31). Because microadenomas and adenomas in B6 ApcMin/+ mice exhibit loss of the Apc+ allele (31), contamination of normal tissue with these lesions would reduce the Apc+:ApcMin ratio. Here, normal intestinal epithelium samples from ERα+/+ ApcMin/+ mice yielded a mean Apc+:ApcMin ratio value of 0.78 ± 0.24 (Figure 4D). A mean Apc+:ApcMin ratio value of 0.90 ± 0.21 was obtained from normal intestinal epithelium samples from ERα−/− ApcMin/+ mice (Figure 4D). The mean Apc+:ApcMin ratio values from ERα+/+ ApcMin/+ in mice did not differ significantly from the value ERα−/− ApcMin/+ mice (P > 0.05), and both means were consistent with the value reported previously (31). This result indicates that the normal epithelium samples used were not contaminated with microadenomas.Effect of deletion of ERβ on tumorigenesis in ApcMin/+ miceTo evaluate the effect of ERβ deletion on tumor formation in ApcMin/+ mice, we intercrossed ERβ+/− mice with ApcMin/+ mice. In this intercross (Helsinki intercross), tumor multiplicity was evaluated at 90 days of age. ERβ genotype did not have a significant impact on the mean number of adenomas in either the entire intestinal tract (data not shown; P ≥ 0.05) or the small intestine alone (Figure 5A; P ≥ 0.05). Because ERβ is thought to play a role during the later stages of colorectal cancer progression, a second set of intercross mice was produced and evaluated at 140 days of age. This second set of intercross mice (Omaha intercross) was performed at a separate institution. However, both intercrosses involved the same ERβ knockout (Esr2Ksk1) and ApcMin/+ alleles, both of which were extensively backcrossed to the B6 strain. Consistent with our observation in the first set of intercross mice, ERβ deficiency did not impact total tumor multiplicity (data not shown; P ≥ 0.05) or small intestinal tumor multiplicity (Figure 5B; P ≥ 0.05) in the second intercross population.Fig. 5.ERβ deficiency increases incidence of colon tumors in ApcMin mice in the Helsinki but not Omaha intercross. Adenoma multiplicity data from the small intestinal tract (A–B) and colon (C–D) are illustrated. Colon adenoma incidence data are shown in panels (E and F). The ‘1’ indicates a value significantly different from that in ERβ+/+ ApcMin/+ mice and the ‘2’ indicates a value significantly different from that in ERβ+/− ApcMin/+ mice. The number of mice from the Helsinki N10 intercross was as follows: ERβ+/+ ApcMin/+ (n = 19), ERβ+/− ApcMin/+ (n = 14) and ERβ−/− ApcMin/+ mice (n = 14). The number of mice from the Omaha N10 intercross was as follows: ERβ+/+ ApcMin/+ (n = 9), ERβ+/− ApcMin/+ (n = 18) and ERβ−/− ApcMin/+ mice (n = 17).Analysis of the tumorigenesis in the colon indicated that ERβ genotype had no significant effect on the mean number of adenomas in the colon at 90 days of age (Figure 5C; P = 0.06–0.20) or 140 days of age (Figure 5D; P ≥ 0.05). In contrast, the 100% (14/14) incidence of colon tumors in 90-day-old ApcMin/+ ERβ−/− mice was greater than the 57% (8/14) incidence in age-matched ApcMin/+ ERβ+/+ mice (Figure 5E; P < 0.01) and the 68% (13/19) incidence in age-matched ApcMin/+ ERβ+/− mice (Figure 5E; P < 0.05). ERβ had no impact on colon adenoma incidence in the Omaha intercross mice evaluated at 140 days of age (Figure 5F; P ≥ 0.05).Potential impact of 129P2-derived loci in the ERβ knockout crosses?To assess the potential impact of residual 129P2 alleles present in the ERβ intercross populations, we genotyped the B6.129-Esr2tm1Ksk mice used in each study with markers distributed across the genome (Table I) to identify 129P2-derived regions. In the Helsinki intercross, 129P2 alleles were detected at only two markers—one unlinked to ERβ (D1Mit206) and one linked to ERβ (D12Mit91) (Table I). At these two markers, ApcMin/+ ERβ+/+, ApcMin/+ ERβ+/− and ApcMin/+ ERβ−/− intercross mice were either homozygous for the B6 allele or heterozygous. Genotype at these markers had no significant impact on total adenoma number in the intercross mice (data not shown; P = 0.3). Likewise, genotype at these markers did not have a significant impact on colon adenoma number (data not shown; P > 0.05).In the Omaha intercross mice, 129P2 alleles were detected at one marker unlinked to ERβ (D3Mit203) and three markers linked to ERβ (D12Mit60, D12Mit91 and D12Mit118) (Table I). Genotype at D3Mit203 had no impact on tumor multiplicity (P > 0.05; data not shown). This analysis is straightforward because D3Mit203 and ERβ are unlinked. In contrast, evaluating the potentially confounding effects of 129P2 alleles at markers linked to the ERβ knockout was complicated by the significant linkage disequilibrium. Genotype at ERβ and the three markers proximal of ERβ on chromosome 12 showed a high degree of concordance as follows: 93% for D12Mit91, 84% for D12Mit60 and 62% for D12Mit182. Thus, as much as 77 Mbp of the 129P2 genome on chromosome 12 is retained in association with the ERβ knockout allele (Table I).Thus, both the Helsinki and the Omaha intercrosses were performed on a congenic B6 background in which only a single marker unlinked to ERβ was still segregating 129P2 alleles. This uniform genetic background facilitates the interpretation of results because the possible impact of modifier loci unlinked to the ERβ mutation can be discounted. However, there were significant differences between the ERβ knockout mice used in these two crosses with respect to markers linked to ERβ. In the Helsinki cross, the degree of linkage disequilibrium between the ERβ knockout and linked 129P2 alleles is quite small, and the potential for linked modifiers is therefore minimal. In contrast, a larger block of 129P2 alleles that was retained in association with the ERβ knockout in the Omaha cross. Although there is no direct evidence supporting the idea that a locus linked to ERβ modifies tumorigenesis in ApcMin/+ mice, our observations prevent us from discounting this possibility.ERβ deficiency does not impact the levels of β-catenin, Cyclin D1, E-cadherin or ERα in the colonic mucosaBecause ERβ deficiency increased colon tumor incidence, we examined the impact of ERβ deficiency on β-catenin expression and localization, the expression of Wnt–β-catenin target genes including Cyclin D1 and the interaction of β-catenin with the cell adhesion molecule E-cadherin. We used western blotting to analyze the levels of β-catenin, E-cadherin and Cyclin D1 in the nuclear and membrane fraction of the colonic mucosa of ApcMin/+ ERβ+/+ and ApcMin/+ ERβ−/− mice. Additionally, we analyzed the impact of ERβ deficiency on the expression of the ERα. Deletion of ERβ had no effect on the expression of the nuclear β-catenin or Cyclin D1 or membrane β-catenin or E-cadherin (supplementary Figure 1 is available at Carcinogenesis Online). Furthermore, the expression of ERα protein was unaffected (supplementary Figure 1 available at Carcinogenesis Online). Thus, we found no evidence for ERβ-dependent modulation of the Wnt pathway.DiscussionIn the present study, ERβ deficiency did not increase adenoma multiplicity in ApcMin/+ mice. This lack of effect of ERβ deficiency on total tumor multiplicity or small intestinal tumor multiplicity was confirmed in a second, independent cross involving the same ERβ knockout allele (Esr2Ksk1) and ApcMin/+ mice. This second cross was conducted at a separate institution using distinct husbandry conditions. A similar lack of effect of ERβ deficiency on small intestinal tumorigenesis in ApcMin/+ mice was also reported by Cho et al. (18) in a study using a distinct ERβ knockout allele (Esr2Pcn1) (38). These three observations differ from that of Giroux et al. (19), in which it is reported that deficiency of the Esr2Pcn1 ERβ knockout allele increased tumor multiplicity and size in female but not male ApcMin/+ mice. However, it should be noted that in the study of Giroux et al., female mice had been ovariectomized and implanted with E2 and progesterone pellets that result in continuous serum levels of these hormones in the high physiological range. In the two intercrosses we performed, as well as the study of Cho et al., ovary intact females were used. Consistent with the ovarian defects described previously in ApcMin/+ females (39), we observe that ApcMin/+ females have reduced serum levels of E2 and progesterone. Thus, there are significant differences in hormone levels between mice in these various studies. These differences may contribute to the disparate results observed.In the present study, we observed that ERβ deficiency was associated with an increased incidence of colon tumors in ApcMin/+ mice. Although our results differ from those described in Giroux et al. (19), in which the Esr2Pcn1 ERβ knockout allele was found to have no impact on colon tumorigenesis in ApcMin/+ mice, our data are consistent with the study of Cho et al. (18), which reported that homozygosity for the Esr2Pcn1 ERβ knockout allele increased colon tumor incidence in ApcMin/+ mice. In our study, we found that the increased incidence in colon tumors in ERβ-deficient ApcMin/+ mice was not associated with enhanced activation of the Wnt–β-catenin pathway. These observations are consistent with a previous report (19) suggesting that ERβ deficiency may promote intestinal tumorigenesis by modulating the transforming growth factor β pathway. Thus, our present studies using ApcMin/+ mice and a distinct ERβ knockout (Esr2Ksk1) help resolve the controversy regarding the impact of ERβ on colon tumorigenesis and provide independent support for the hypothesis that ERβ deficiency may promote tumorigenesis through a Wnt–β-catenin-independent pathway.We also observed that genetic disruption of ERα in ApcMin/+ mice was associated with an increase in intestinal adenoma multiplicity, size and burden. These data are consistent with the hypothesis that loss of ERα expression contributes to intestinal tumorigenesis. Although the ERα knockout used here is a hypomorph, resulting in a truncated receptor (40) with a deleted amino-terminus and no activation function 1 domain but with some residual ligand binding and/or function, it is clear that disruption of ERα, even if incomplete, is associated with enhanced tumorigenesis ApcMin mice. It is noteworthy that when Cho et al. (18) crossed another ERα knockout (B6.129-Esr1Pcn1), one reported to be a more complete disruption of ERα, with ApcMin/+ mice, no ERα−/− ApcMin/+ mice were recovered. One intriguing interpretation of this result is that there is a synthetic lethal interaction between the ERα allele and the ApcMin mutation. In our study, ERα−/− ApcMin/+ mice were recovered at the expected frequency.Based on data from humans implicating ERα as a tumor suppressor in colon, we conclude that ERα deficiency itself is responsible for the enhanced tumorigenesis in ERα−/− ApcMin/+ mice. However, it is impossible to completely exclude the possibility that the enhanced tumorigenesis in ERα−/− ApcMin/+ mice is due to another allele that resides within the 129P2-derived region of <13 Mbp on chromosome 10 that is retained in association with the ERα knockout allele. Less than a dozen genes map to this region, and none of these is known to modulate tumorigenesis in ApcMin/+ mice. The Scc14 locus, which modulates susceptibility to azoxymethane-induced colon cancer in a cross between the BALB/c and the CcS19/Dem strains, is linked to a marker 3.4 Mbp proximal of ERα (41). Nothing is known about the impact of the 129P2 allele of Scc14 on colon cancer. Assuming that ERα and Scc14 represent distinct genes, we cannot exclude the possibility that the 129P2 allele of Scc14 increases tumor multiplicity in our cross. However, our previous studies provide no evidence that 129 substrains carry alleles that enhance tumorigenesis in ApcMin/+ mice. Rather, we have shown that the 129S2/SvPas substrain carries alleles that reduce tumor multiplicity in ApcMin/+ mice (27). Additionally, in Oikarinen et al., in this issue of Carcinogenesis, we report the localization of Mom5, a modifier of tumorigenesis in ApcMin/+ mice; the 129P2 allele of Mom5 is associated with reduced tumor number.The impact of ERα deficiency was equivalent in ApcMin/+ males and females. These results are consistent with the fact that ERα expression is silenced in colorectal cancers in both men and women (13,42). Although ERα signaling reduces intestinal tumorigenesis in both male and female ApcMin/+ mice, it is unclear whether this effect requires endogenous estrogens or is ligand independent. ApcMin/+ females in our colony exhibit diminished ovarian function and low serum concentrations of E2 and progesterone. Although the physiological basis for the impact of the Min mutation of ovarian function is unknown, these observations, together with our observation that tumorigenesis in ApcMin/+ females is not influenced by ovariectomy (K.A.Gould, unpublished data), suggest that in our colony, ovarian estrogens do not play a role in ERα-dependent processes that attenuate tumorigenesis in ApcMin/+ mice.We report that ERα deficiency in ApcMin/+ mice is associated with an accumulation of nuclear β-catenin, suggesting that ERα deficiency activates canonical Wnt–β-catenin signaling. The level of β-catenin in normal intestinal epithelium of ERα−/− ApcMin/+ mice is less than that in ApcMin-induced intestinal adenomas (K.A.Gould, unpublished data), suggesting that further activation of Wnt–β-catenin signaling occurs during tumorigenesis, probably as a consequence of loss of Apc+, which occurs at a frequency of 100% in adenomas of B6 ApcMin/+ mice (11,31). Consistent with the hypothesis that ERα deficiency enhances Wnt–β-catenin signaling, the expression of the Wnt–β-catenin target genes Cyclin D1 and c-Myc was increased in the normal intestinal epithelium of ERα−/− ApcMin/+ mice. The expression of Cyclin D1 and c-Myc is regulated by many factors, including estrogens (43,44). Thus, it is possible that the increased expression of Cyclin D1 and c-Myc in ERα−/− ApcMin/+ epithelium is unrelated to Wnt–β-catenin activation. Nonetheless, the upregulation of these genes, which are known to play a role in colorectal cancer, indicates that ERα deficiency in ApcMin/+ mice is associated with increased expression of genes involved in intestinal tumorigenesis.Although it is known that there is cross talk between the estrogen and the Wnt pathways in estrogen-responsive tissues such as the mammary gland (33), further studies will be required to understand the molecular basis and significance of cross talk between ERα and Wnt signaling in the intestinal epithelium. In contrast to what is observed in the mammary gland, our studies suggest that Wnt and estrogen signaling have opposing effects in intestine, with Wnt signaling promoting tumorigenesis and ERα signaling inhibiting tumorigenesis.Supplementary materialSupplementary Figure 1 can be found at http://carcin.oxfordjournals.org/FundingNebraska Department of Health and Human Services; National Institutes of Health (K01-CA113413 to K.A.G.).AbbreviationsApcadenomatous polyposis coliERestrogen receptorE217β-estradiolMinmultiple intestinal neoplasiaPCRpolymerase chain reactionConflict of Interest Statement: None declared.1.American Cancer SocietyColorectal Cancer Facts and Figures Special Edition 20052005Atlanta, GAAmerican Cancer Society2.CalleEEEstrogen replacement therapy and risk of fatal colon cancer in a prospective cohort of postmenopausal womenJ. Natl Cancer Inst.1995875175233.BroedersMJHistory of childbearing and colorectal cancer risk in women aged less than 60: an analysis of Swedish routine registry data 1960–1984Int. J. Cancer1996661701754.MartinezMEA prospective study of reproductive factors, oral contraceptive use, and risk of colorectal cancerCancer Epidemiol. Biomarkers Prev.19976155.RossouwJERisks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results From the Women's Health Initiative randomized controlled trialJAMA20022883213336.Campbell-ThompsonM17Beta-estradiol modulates gastroduodenal preneoplastic alterations in rats exposed to the carcinogen N-methyl-N′-nitro-nitrosoguanidineEndocrinology1999140488648947.GershbeinLLAction of estrogen and adrenocorticoids on adenocarcinoma induction by 1,2-dimethylhydrazine in male ratsRes. Commun. Chem. Pathol. Pharmacol.1993811171208.WeyantMJReciprocal expression of ERalpha and ERbeta is associated with estrogen-mediated modulation of intestinal tumorigenesisCancer Res.200161254725519.JavidSHModulation of tumor formation and intestinal cell migration by estrogens in the Apc(Min/+) mouse model of colorectal cancerCarcinogenesis20052658759510.MoserARA dominant mutation that predisposes to multiple intestinal neoplasia in the mouseScience199024732232411.ShoemakerARStudies of neoplasia in the Min mouseBiochim. Biophys. Acta19971332F25F4812.MooreJTCloning and characterization of human estrogen receptor beta isoformsBiochem. Biophys. Res. Commun.1998247757813.IssaJPMethylation of the oestrogen receptor CpG island links ageing and neoplasia in human colonNat. Genet.1994753654014.LuBEstrogen receptor-beta mRNA variants in human and murine tissuesMol. Cell. Endocrinol.199813819920315.Campbell-ThompsonMExpression of estrogen receptor (ER) subtypes and ERbeta isoforms in colon cancerCancer Res.20016163264016.BelshawNJMethylation of the ESR1 CpG island in the colorectal mucosa is an ‘all or nothing’ process in healthy human colon, and is accelerated by dietary folate supplementation in the mouseBiochem. Soc. Trans.200533Pt 470971117.JassamNLoss of expression of oestrogen receptor beta in colon cancer and its association with Dukes’ stagingOncol. Rep.200514172118.ChoNLEstrogen receptors alpha and beta are inhibitory modifiers of Apc-dependent tumorigenesis in the proximal colon of Min/+ miceCancer Res.2007672366237219.GirouxVEstrogen receptor beta deficiency enhances small intestinal tumorigenesis in ApcMin/+ miceInt. J. Cancer200812330331120.LubahnDBAlteration of reproductive function but not prenatal sexual development after insertional disruption of the mouse estrogen receptor geneProc. Natl Acad. Sci. USA199390111621116621.KregeJHGeneration and reproductive phenotypes of mice lacking estrogen receptor betaProc. Natl Acad. Sci. USA199895156771568222.DietrichWFGenetic identification of Mom-1, a major modifier locus affecting Min- induced intestinal neoplasia in the mouseCell19937563163923.WindahlSHIncreased cortical bone mineral content but unchanged trabecular bone mineral density in female ERbeta(-/-) miceJ. Clin. Invest.199910489590124.ShoemakerARN-ethyl-N-nitrosourea treatment of multiple intestinal neoplasia (Min) mice: age-related effects on the formation of intestinal adenomas, cystic crypts, and epidermoid cystsCancer Res.1995554479448525.GouldKAMom1 is a semi-dominant modifier of intestinal adenoma size and multiplicity in Min/+ miceGenetics19961441769177626.PajariAMPromotion of intestinal tumor formation by inulin is associated with an accumulation of cytosolic beta-catenin in Min miceInt. J. Cancer200310665366027.BynoteKKEstrogen receptor-alpha deficiency attenuates autoimmune disease in (NZB x NZW)F1 miceGenes Immun.2008913715228.GouldKAGenetic determination of susceptibility to estrogen-induced mammary cancer in the ACI rat: mapping of Emca1 and Emca2 to chromosomes 5 and 18Genetics20041682113212529.ShullJDOvary-intact, but not ovariectomized female ACI rats treated with 17beta-estradiol rapidly develop mammary carcinomaCarcinogenesis1997181595160130.GouldKADES action in the thymus: inhibition of cell proliferation and genetic variationMol. Cell. Endocrinol.2000170313931.LuongoCLoss of Apc+ in intestinal adenomas from Min miceCancer Res.1994545947595232.CouseJFEstrogen receptor null mice: what have we learned and where will they lead us? [published erratum appears in Endocr Rev 1999 Aug;20(4):459]Endocr. Rev.19992035841733.KatohMExpression and regulation of WNT1 in human cancer: up-regulation of WNT1 by beta-estradiol in MCF-7 cellsInt. J. Oncol.20032220921234.KatohMDifferential regulation of WNT2 and WNT2B expression in human cancerInt. J. Mol. Med.2001865766035.El-TananiMDifferential modulation of transcriptional activity of estrogen receptors by direct protein-protein interactions with the T cell factor family of transcription factorsJ. Biol. Chem.2001276416754168236.KouzmenkoAPWnt/beta-catenin and estrogen signaling converge in vivoJ. Biol. Chem.2004279402554025837.PaulsenJEQualitative and quantitative relationship between dysplastic aberrant crypt foci and tumorigenesis in the Min/+ mouse colonCancer Res2001615010501538.DupontSEffect of single and compound knockouts of estrogen receptors alpha (ERalpha) and beta (ERbeta) on mouse reproductive phenotypesDevelopment20001274277429139.BennettLMMammary tumor induction and premature ovarian failure in ApcMin mice are not enhanced by Brca2 deficiencyToxicol. Pathol.20012911712540.CouseJFAnalysis of transcription and estrogen insensitivity in the female mouse after targeted disruption of the estrogen receptor geneMol. Endocrinol.199591441145441.RuivenkampCAFive new mouse susceptibility to colon cancer loci, Scc11-Scc15Oncogene2003227258726042.DawsonPMOestrogen receptors in colorectal carcinomaJ. Clin. Pathol.19904314915143.DubikDMechanism of estrogen activation of c-myc oncogene expressionOncogene199271587159444.SabbahMEstrogen induction of the cyclin D1 promoter: involvement of a cAMP response-like elementProc. Natl Acad. Sci. USA1999961121711222
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B48D236D9F55F298C43CF722F0CFC3D53AD9D5C.txt b/test/dataset/in/resources/corpus/Clean_0B48D236D9F55F298C43CF722F0CFC3D53AD9D5C.txt
deleted file mode 100644
index 6b99d2f..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B48D236D9F55F298C43CF722F0CFC3D53AD9D5C.txt
+++ /dev/null
@@ -1 +0,0 @@
- geront Gerontologistgeront The Gerontologist The Gerontologist 0016-9013 1758-5341 Oxford University Press 0020410.1093/geront/41.4.449 RESEARCH ARTICLE Come Talk With Me Improving Communication Between Nursing Assistants and Nursing Home Residents During Care Routines Burgio Louis D. a b Allen-Burge Rebecca a b Roth David L. c Bourgeois Michelle S. d Dijkstra Katinka d Gerstle John a Jackson Erik a Bankester Leanna a aApplied Gerontology Program, The University of Alabama, Tuscaloosa bDepartment of Psychology, The University of Alabama, Tuscaloosa cDepartment of Psychology, University of Alabama at Birmingham dDepartment of Communication Disorders, Florida State University, Tallahassee Louis D. Burgio, The University of Alabama, Department of Psychology and The Applied Gerontology Program, Box 870315, Tuscaloosa, AL 35487-0315. E-mail: lburgio@sw.ua.edu. 1 8 2001 41 4 449 460 6 4 2001 18 9 2000 The Gerontological Society of America 2001 Purpose: We examined the effects of communication skills training and the use of memory books by certified nursing assistants (CNAs) on verbal interactions between CNAs (n = 64) and nursing home residents (n = 67) during care routines. Design and Methods: CNAs were taught to use communication skills and memory books during their interactions with residents with moderate cognitive impairments and intact communication abilities. A staff motivational system was used to encourage performance and maintenance of these skills. Formal measures of treatment implementation were included. Results:Results were compared with those for participants on no-treatment control units. Trained CNAs talked more, used positive statements more frequently, and tended to increase the number of specific instructions given to residents. Changes in staff behavior did not result in an increase in total time giving care to residents. Maintenance of CNA behavior change was found 2 months after research staff exited the facility. Although an increase was found in positive verbal interactions between CNAs and residents on intervention units, other changes in resident communication were absent. Implications: Nursing staff can be trained to improve and maintain communication skills during care without increasing the amount of time delivering care. The methodological advantages of including measures to assess treatment implementation are discussed. Long-term care Staff training Dementia hwp-legacy-fpage 449 hwp-legacy-dochead RESEARCH ARTICLE Decision Editor: Laurence G. Branch, PhD The social milieu of the nursing home is a critical component of both quality of care and quality of life in this setting. In recent years, researchers have focused their efforts on understanding and improving various aspects of staff–resident and resident–resident social interactions. Direct behavioral observation of residents on a psychogeriatric ward revealed that 21% of residents' time was spent in physical care interactions (Hallberg, Norberg, and Eriksson 1990). Observation of certified nursing assistants' (CNAs') work behavior in nursing homes revealed that 53% of observation time was spent in a broader category of resident care, with 11.8% of observation time engaged in staff–resident verbal interaction (Burgio, Engel, Hawkins, McCormick, and Scheve 1990). Sadly, positive verbal interactions appear to be as rare as negative verbal interactions in nursing homes (Burgio et al. 1990), where neutral interactions appear to be the norm (Carstensen, Fisher, and Malloy 1995). These data suggest that intervention is necessary to structure staff–resident social interactions during care to positively affect resident quality of life. Communication-based interventions in nursing homes have the potential to improve resident behavioral deficits such as excess disability, low-rate social engagement, and behavioral excesses such as disruptive vocalization (Allen-Burge, Stevens, and Burgio 1999; Burgio and Stevens 1999). Through these interventions, staff are taught therapeutic strategies for approaching residents and responding to both adaptive and maladaptive behaviors. Recent small-sample studies of communication-based staff-training interventions to reduce resident problem behaviors have suggested that training staff to communicate more effectively might reduce the incidence of physical aggression (Boehm, Whall, Cosgrove, Locke, and Schlenk 1995; Hoeffner, Radar, McKenzie, Lavelle, and Stewart 1997). Many recent studies have used packaged staff interventions that include both communication skills training and the use of prosthetic memory aids, such as memory books, to compensate for residents' diminished cognitive capacity. Memory books contain images and brief, simple sentences that use the preserved automatic processing abilities of frail older adults to improve the structure and quality of communication with others, hopefully making interactions more pleasant for both older individuals and their partners in communication (Bourgeois 1990, Bourgeois 1992a, Bourgeois 1992b, Bourgeois 1993; Bourgeois and Mason 1996). Memory books can serve two functions. First, they can provide meaningful stimulation to nursing home residents in an environment that can be both physically and socially barren. Second, they can be used as a tool by staff or other residents to facilitate communication with cognitively impaired residents. Through the use of memory books, community-residing dementia patients with a wide range of cognitive deficits have increased the informativeness and accuracy of their conversations and decreased their ambiguity and repetitiveness (Bourgeois 1990, Bourgeois 1992a, Bourgeois 1992b; Bourgeois and Mason 1996). Use of these aids has decreased disruptive verbal behaviors such as repetitive questioning (Bourgeois, Burgio, Schulz, Beach, and Palmer 1997). In nursing homes, content analysis of researcher-initiated, semistructured 5-min conversational interactions between demented residents and their primary CNAs, and social validity ratings of these interactions by unfamiliar judges, revealed modest improvements in resident and CNA conversational behaviors (Hoerster, Hickey, and Bourgeois in press). Ripich, Wykle, and Niles 1995 developed and tested a didactic training program for improving communication between CNAs and residents. In addition to teaching specific verbal and nonverbal communication strategies, training modules included information on language difficulties associated with different stages of dementia, depression, and cultural–ethnic considerations. Results showed that training improved both CNA knowledge and attitudes toward residents with dementia. McCallion, Toseland, Lacey, and Banks 1999 evaluated a Nursing Assistant Communication Skills Program on various staff and resident outcomes in nursing homes. In addition to teaching verbal and nonverbal communication skills, staff were taught to use memory aids. Memory aids included placing written and graphic signs in important locations and the use of memory charts placed on resident bedroom walls. Memory charts contained photographs, brief statements, and conversation topics that were meant to encourage communication with the residents. To facilitate maintenance of therapeutic effort, the trainer made informal monthly visits to the nursing homes to verify continued implementation of the Nursing Assistant Communiation Skills Program techniques. Results showed durable effects (6 months posttreatment) on resident depressive symptomatology. However, the effects on resident disruptive behaviors were mixed and generally transient. McCallion, Toseland, and Freeman 1999 also developed a Family Visit Education Program, which focused on improving communication among residents, nursing staff, and visiting family members. Compared with a control group, the Family Visit Education Program showed benefits for residents and family members, but not for nursing staff. Specifically, family members demonstrated improved communication patterns with residents; residents showed improvements in depression, irritability, and verbal behavior. Staff also reported a reduction in agitation. The success of communication-based interventions designed to improve nursing home care depends heavily on the receptiveness of nursing staff to learn new skills and on the establishment of staff motivational systems to ensure the maintenance of these skills (Burgio and Burgio 1990; Burgio and Scilley 1994; Burgio and Stevens 1999; Stevens et al. 1998). Staff motivational systems such as behavioral supervision (Burgio and Burgio 1990) and total quality management (Schnelle, Ouslander, Osterweil, and Blumenthal 1993) have been used successfully to motivate staff in the implementation of a variety of interventions in nursing homes. Behavioral supervision is designed to motivate management and direct care staff to observe and analyze skill performance in order to identify problems and to induce supervisory staff to provide direct and specific feedback suggesting practical ways of maximizing skills (Burgio and Stevens 1999; Daniels 1994). Using behavioral supervision, Stevens et al. 1998 showed that CNAs could be trained to increase and maintain the rate of task announcements and positive statements made to residents during care over a 46-week assessment period. In a small intra-subject pilot study (N = 8), Allen-Burge, Burgio, Bourgeois, Sims, and Nunnikhoven 2001 examined the effects of communication skills training and the use of memory books. CNAs were taught to use communication skills and memory books during their interactions with residents with mild to moderate cognitive impairment but relatively intact communication abilities. An abbreviated system of behavioral supervision was attempted to encourage performance of these skills on the nursing units. Results showed that, regardless of sporadic implementation of the intervention by nursing staff, the intervention improved communication between staff and residents during care routines, increased the amount of time other residents and visitors spent talking with target residents, and increased the rate of positive statements made by the target residents and others in their immediate environment. The purpose of the present study was to examine the efficacy of communication skills training and the use of individualized memory books in improving communication between CNAs and a larger group of residents (N = 67) during care. In contrast to the pilot study, we used a more comprehensive staff motivation system (behavioral supervision). In addition, formal measures of treatment implementation were included (Lichstein, Riedel, and Grieve 1994). We hypothesized that (a) through the use of behavioral supervision, CNAs' use of therapeutic procedures would increase, compared with their baseline performance, and would maintain at a 2-month follow-up assessment; (b) trained CNAs would increase their frequency of overall verbal interaction and positive verbal interactions compared with their own baseline performance and untrained CNAs; and (c) residents on trained nursing units would show increases in overall amount of coherent verbal interaction (i.e., understandable speech) compared with residents on untrained units. Methods Settings This study was conducted in five nursing homes with an average census of 120 residents on three units. On average, 11% of the beds were Medicare, 50% were Medicaid, and the remainder were private pay. Four of the nursing homes were corporately owned facilities; one was privately owned with a religious affiliation. Special care units were not included in the study. Residents had an average length of stay of 3 years; 93% were female, 75% were white, and 25% were African American. Resident-to-CNA ratios were 9:1 during the day shift and 12:1 during the evening shift. Reported yearly rates of staff turnover were 8% among registered nurses (RNs), 18% among licensed practical nurses (LPNs), and 49% among CNAs. The American Health Care Association reported that the national turnover rate is 97% for CNAs (as cited in Harrington et al. 2000). Participants Residents Residents were entered into the study on the basis of the following criteria: (a) age of at least 55 years; (b) Mini-Mental Status Examination (MMSE; Folstein, Folstein, and McHugh 1975) total score greater than 0 or Short Portable Mental Status Questionnaire (Pfeiffer 1975) with fewer than 10 errors (our purpose for using these criteria were to exclude nonresponsive residents); (c) retention of minimal ability in verbal communication involving spontaneous speech based on criteria established by Bourgeois 1993(criteria available by writing Louis D. Burgio or Rebecca Allen-Burge); (d) absence of major sensory impairment; (e) life expectancy greater than 6 months; and (f) residence within the facility for at least 1 month. Because of the importance in this study of observing staff–resident interactions during care routines, residents who completed activities of daily living (ADLs) independently during baseline were excluded. Residents were considered independent if the care tasks were completed with no nursing assistant present for at least 5 minutes across two observational occasions during baseline. Two hundred sixty-three of 403 residents in the five nursing homes met entry criteria on the basis of nursing staff referral; 111 consented to participate, the majority through proxy. The consent rate was 42%. Nineteen were excluded during the baseline phase because the communication assessment and ADL observation indicated that they did not meet entry criteria; thus, 92 residents meeting entry criteria completed baseline. A decision was made to include in the data analyses only residents who had data available for all three assessment points (baseline, Postintervention Assessments I and II). Twenty-five residents were excluded due to incomplete data, leaving 67 residents for data analysis (33 control and 34 intervention residents). The reasons for exclusion were the following: eight residents died prior to the intervention phase, 9 moved to another nursing unit or different facility, 4 residents or sponsors withdrew consent for participation, and 4 residents refused to be observed during care. There were no significant differences between the surviving (n = 67) and the excluded (n = 25) residents on age, gender, race, dementia diagnosis, MMSE, or Functional Independence Measure (ADL) scores. Table 1 shows selected characteristics of surviving residents in the study sample at baseline. As can be seen, there were no significant differences between residents on the control and treatment units. Staff Ninety-eight CNAs were working on either the intervention or control units during baseline assessment. Observational data were available during all three assessment points for 64 CNAs; thus, 34 CNAs were dropped from the analysis. There were no differences in gender, race, education, age, or length of service between the 98 CNAs available at baseline and the 64 surviving CNAs. Table 2 shows selected characteristics of surviving CNAs in the study sample at baseline; however, demographic data were available for only 37 CNAs in the intervention group. As can be seen, there were no significant differences between CNAs on the control units (n = 25) and treatment units (n = 39; 37 providing demographic data). Design and Procedures We used a two-group comparison design with an intragroup comparison component embedded in each group. The two groups were (a) memory book intervention with a staff motivational system and (b) no-treatment control (NTC). The study was conducted sequentially across nursing homes. Upon entering a nursing home, two units were selected randomly and also randomly assigned to treatment and control conditions. Data were collected on treatment and control units concurrently. We spent the first 3 to 4 weeks in each facility, prior to baseline recording, gathering information on all consenting treatment and control residents and on regular day and evening shift nursing staff. This was followed by a 4-week (Weeks 1 to 4) baseline phase (both units) designed to assess the social environment in the nursing homes before intervention. During this and all phases, direct observational and paper-and-pencil assessments were completed (see Measures section). On treatment units, communication–memory book skill inservice workshops were conducted during a 1-week period immediately following baseline (Week 5; see Intervention Conditions section). This was followed by hands-on training on the intervention units during the next 4 weeks (Weeks 6 to 9). During this phase, CNAs were instructed in the use of communication–memory book skills by research and indigenous nursing staff. Research staff also instructed supervisory nursing staff in the use of the staff motivational system during this time. The next 8 weeks consisted of evaluation of the intervention's effectiveness using the same assessments as at baseline. This assessment period was divided into two phases: Postintervention Assessment I (PIA–I) represented staff and resident performance during the first 4 weeks (Weeks 10 to 13); Postintervention Assessment II (PIA–II) represented their performance during the second 4 weeks of this phase (Weeks 14 to 17). During the first segment of this latter phase, supervisory nursing staff continued to receive assistance from research staff on the use of the staff motivational system. By the end of the phase, indigenous staff controlled all aspects of the system and received no assistance from research staff. To enhance maintenance of the staff motivational system, consultation visits by the project manager were conducted at 3 and 6 weeks after we had departed from the facility (i.e., after PIA–II). Two months after PIA–II, probe follow-up data were collected on staff and resident performance during a 1-week period in four of the five nursing homes. The fifth nursing home declined the follow-up assessment because of a high level of administrative and nursing staff turnover. Observational data were collected using a reduced sampling schedule. Intervention Conditions Communication–Memory Book Skills and Staff Motivational System Participating residents received a personalized 12-page laminated memory book consisting of biographical, orientation, and daily schedule information. Each 5- × 5-in. page included a 6–10 word declarative sentence printed in 14- or 20-point Times New Roman type with a color or black-and-white photograph or line drawing illustrating the statement. Residents' books might contain pictures of their wedding and family, their CNA, other residents, their daily schedule, instructions on bathing, and pages targeting behavior problems such as wandering, aggression, or repetitive questioning. It was the CNA's responsibility to ensure that books were always present in the residents' living space. Residents were provided with individually tailored vests that attached the book to the resident's torso, wheelchair caddies that attached the book to the resident's wheelchair, or a stand to display the book prominently in the resident's room. Additional pages were added to the books every 4 weeks in order to maintain staff and resident interest in this aspect of the intervention. Memory books were replaced by research staff if lost. Replacements were provided for 38% of the residents. All nursing staff on intervention units were trained in the use of memory books and general communication skills. The first component of staff training consisted of a 2-hr inservice session on communication–memory book skills conducted by the project manager (a licensed clinical psychologist). Members of the nursing staff were instructed to use memory books with residents (a) to increase general communication among residents and between nursing staff and residents, (b) to increase the independent functioning of residents during targeted care routines, and (c) as a distraction to decrease resident disruptive behaviors. Inservice materials are available from Louis D. Burgio or Rebecca Allen-Burge on request. CNAs and supervisory staff then attended a 1-hr inservice to introduce the staff motivational system. An additional hour of training was provided to LPNs and RNs on the supervision of this system. The critical roles of the CNA and LPN were emphasized because these members of the nursing staff have the most direct contact with residents (Burgio et al. 1990). In addition to didactic training, active learning techniques including the use of role play, discussion of real-life examples from the nursing units, and discussion of written vignettes were used to engage staff. Notebooks were provided to each member of the nursing staff that included all information from the inservices. Staff were encouraged to refer to these notebooks while working with residents on the unit. To increase the likelihood that staff would use the newly learned skills on the units, behavioral supervision (Burgio and Burgio 1990) was taught during the inservice, implemented by research staff during the 4-week hands-on training phase and trans-ferred to indigenous nursing home staff during the 8-week postintervention assessment period. Initially, the CNAs were trained by research staff in the use of memory books, communication skills, and components of the staff motivational system. CNAs on the day and evening shifts were observed by research staff once per day while providing care to residents. Researchers and LPNs used the Communication Skills Checklist (CSC) during the care routine to record whether CNAs displayed the skills taught in the inservice; feedback was provided to CNAs regarding their use of the memory books, use of specific versus general instructions, one-step instructions, positive statements, responses to behavioral disturbances, and general distraction techniques. CNAs were provided verbal performance feedback immediately following the care routine. The project manager also completed the Observation of the LPN's Supervisory Activities (OLSA) on the LPNs during this period. Only data collected by research staff were used in outcome analysis. Thirty-nine CNAs on the intervention units were taught to monitor and record their own skill performance as a means of tracking their performance and as a daily reminder of the skills needed to be an effective communicator. CNAs were asked to use the self-monitoring form during shifts worked throughout the week. CNAs had the opportunity to use the form an average of four times (range 1–10) per week. To meet job performance goals, CNAs were asked to complete at least an average of 80% of their assigned self-monitoring forms. To reach performance criterion, they were also expected to obtain an average of 80% accuracy on the CSC completed by LPNs or research assistants. Trained CNAs received public recognition for meeting job performance criteria (i.e., 80% completion of forms and 80% accuracy) by having their names posted weekly on a CNA Honor Roll. Those CNAs whose names appeared on the honor roll received an opportunity for a performance incentive. All honor-roll CNAs listed were entered into a performance-based lottery held once each week for day and evening shifts (Reid, Parsons, and Green 1989). For each shift, the individual winning the lottery was provided with his or her choice of incentives from a list of choices determined by each nursing home, including the following: (a) free lunch in the cafeteria every day for 1 week, (b) a "goodie bag" full of inexpensive snacks, beauty products, and knick-knacks, (c) permission to arrive at work 15 min later than scheduled every day for 1 week, (d) permission to leave work 15–30 min earlier than scheduled every day for 1 week, (e) permission to leave work 2.5 hr early on Friday, or (f) 2.5 hr of extra pay. Across nursing homes, the most frequently chosen incentives were the opportunity to leave work earlier than scheduled, extra pay, and goodie bags. LPNs on the intervention unit also received public recognition for meeting job performance criteria (i.e., 80% completion of CNA monitoring and supervisory forms) by having their names posted weekly on an LPN Honor Roll. On average, LPNs were assigned 2–5 CNAs to monitor and supervise in their section on their shift. NTC CNAs and LPNs on the NTC units did not participate in any intervention-related activity. Assessments were conducted on NTC units on the same schedule as on the intervention units. CNAs and LPNs on the intervention units were occasionally pulled to the NTC units. When this occurred, they were asked not to use any of the skills learned on the intervention units. After the 2-month follow-up assessment period, staff on the NTC units were offered instructions and materials on communication–memory book skills. Measurement CNA Communication Skills Checklist The CSC is a direct observation behavioral frequency measure allowing the researcher to assess CNAs' use of specific versus general instructions, one-step instructions, positive statements, biographical statements not included in the memory book (generalization), responses to behavioral disturbances, and use of general distraction techniques (Allen-Burge et al. 2001). Research staff attempted to observe every CNA daily during hands-on training, twice during baseline and each of the two postintervention assessments, and once during follow-up. We were successful in completing this target number of observations for 83% of the CNAs. The CSC focused on communication skills demonstrated during a care interaction. To allow for an adequate sampling of CNAs' use of communication–memory book skills, at least three separate care activities (i.e., bathing, dressing, transfer) were included during the observation in order for the data to be retained. A total of 76.37 hr of observation (range 2–60 min per CSC) were completed across all study phases (i.e., 22.60 hr during baseline, 18.94 hr during PIA–I, 16.20 hr during PIA–II, and 18.63 hr during follow-up). Data were expressed as total number of occurrences of each type of statement divided by the total number of seconds in the observation period. Interobserver reliability was assessed independently among observers during 13% (80/603; 14.03 hr) of the observations across all sites and phases. Observer agreement for each behavioral category measured by the CSC was calculated by using a total occurrence agreement calculation. The average percentage of agreement for this measure across all categories was 84% (range 71%–99%). OLSA The OLSA allows the researcher to measure LPNs' accuracy in observing and recording CNA skill performance during supervision, and LPNs' skill in providing CNAs with verbal performance feedback. As part of the staff motivational system, LPNs observed each of the CNAs under their supervision once per week using a form similar to the CSC during 15-min samples of care routines with residents. During these observations, LPNs recorded the occurrence of CNAs' use of memory books, specific versus general instructions, one-step instructions, positive statements, negative or unhelpful statements, biographical statements not included in the memory book, and CNAs' responses to behavioral disturbances. LPNs' accuracy was checked against the project manager's recording of the same behaviors during the same observation. The project manager was present with LPNs observing CNAs using the CSC during the staff training period. The OLSA also allowed the project manager to rate the LPNs' use of verbal supervisory feedback. Verbal supervisory feedback is rated with regard to beginning and ending a session with a supportive statement to the CNA, the provision of accurate and specific positive and corrective feedback to the CNA, the provision of specific performance scores to the CNA with a statement indicating the CNA's training status, and the provision of an opportunity for the CNA to discuss any of the feedback given by the LPN. CNAs received positive or negative written feedback letters at the end of the hands-on training period and at the end of PIA–I. To receive a positive feedback letter, CNAs had to attain an average performance score of 80% or better in their completion of self-monitoring forms and LPN evaluations for the previous 4-week period. These letters were placed in the CNAs' personnel files, and copies were given to CNAs and to administrative nursing staff. Computer-Assisted Behavioral Observation System (CABOS): Hardware and Software As in Burgio, Scilley, Hardin, Hsu, and Yancey 1996 prior research, the Portable Computer Systems for Observational Research software programs from Communitech International (DeKalb, IL) were used for this project (Repp, Karsh, van Acker, Felce, and Harman 1989). The recording of behavior was synchronized with each computer's internal time clock and controlled through a software routine. Real-time CABOS data were generated by sampling behaviors during care interaction between residents and their primary day- and/or evening-shift CNA. All participating residents were observed twice during baseline, PIA–I and PIA–II, and once during follow-up. Care routines had to be at least 5 min in length for the data to be retained. Domains of behavior divided into mutually exclusive and exhaustive categories relevant to the hypothesized outcomes were identified, and detailed operational definitions were generated for all behavior codes within categories. One category coded the residents' activity, including the amount of ADL care. Four categories coded aspects of the residents' social environment: (a) nonverbal presence of staff or other residents, (b) verbal interaction, (c) verbal content, and (d) disruptive behaviors. Each behavioral category contained codes that allowed specification of type of behavior within the category and whether it was initiated by staff or resident (e.g., content had separate positive and negative statement codes). A total of 70.29 hr of observation (range 5–51 min per care interaction) was completed across all study phases (i.e., 24.24 hr during baseline, 20.31 hr during PIA–I, 18.90 hr during PIA–II, 6.84 hr during follow-up). Interobserver agreement was calculated using Cohen's kappa (Cohen 1968; Hays 1994). Reliability was assessed independently among four observers during 14% (14.03 hr) of the total observation time across all phases. Average kappa reliability across all categories was .79 (range .68–.95). In behavioral research, recommended lower limits for acceptable kappas range from .60 to .75 (Hartmann 1982). Memory Book Checks Research assistants recorded whether residents were in possession of memory books (i.e., memory books were prominently displayed and within arm's reach of the resident) during separate observational checks conducted twice daily, once during the morning and evening shifts. This yielded a total of 282 checks during PIA–I, PIA–II, and the 2-month follow-up assessment. Data were recorded dichotomously (yes/no) at both checks for each resident. Because NTC residents did not possess memory books, checks were conducted only on intervention units. Paper-and-Pencil Measures Assessment instruments were administered either by a clinical psychologist or by research assistants trained and supervised by the clinical psychologist to administer these measures. MMSE (Folstein et al. 1975) The MMSE, a measure of global cognitive ability, measures orientation, immediate and delayed recall for words, attention and concentration, language, and praxis (total score range 0–30). The test–retest and interevaluator reliabilities are .89 and .83, respectively. Functional Independence Measure (FIM)—REACH Version Information for completing this instrument was provided by CNAs familiar with the daily care needs of the participating resident (Hamilton, Laughlin, Fiedler, and Granger 1994; Kidd et al. 1995). The version used in this study was developed for the National Institutes of Health–funded cooperative agreement Resources for Enhancing Alzheimer's Caregiver Health (REACH; Coon, Schulz, and Ory 1999). As in the REACH project, only the motor subscale consisting of self-care, sphincter control, transfer, and locomotion items was used in this study (total score range 13–91). Reliability data indicated an intraclass correlation coefficient of .96 for the motor domain, with unweighted kappas ranging from .53 to .66 (Hamilton et al. 1994). Method of Analysis Analyses were conducted on the CSC and on CABOS data collected during care routines. CNAs were the unit of analysis for CSC data; data were expressed as mean rate of CNA statements per hour. CSC outcome data included (a) the rate of specific, one-step instructions, (b) the rate of CNA positive statements, (c) the rate of biographical statements, (d) the rate of multiple-step instructions, and (e) the total duration of care. CABOS data were expressed as mean total percentage of observation time spent in each activity or mean rate per hour of positive statements, and the unit of analysis was the resident. CABOS outcome data included (a) total percentage of time of resident coherent verbal interaction, (b) total percentage of time of staff speech directed to the resident, and (c) the rate per hour of positive statements made by either residents or CNAs. We did not discriminate between staff and resident positive statements on this variable. Positive statements are brief duration events and were expressed as the rate of statements per hour. Our analytic questions centered on (a) assessment of treatment delivery and enactment (Lichstein et al. 1994), (b) establishing an initial change in behavior for both staff and residents after implementation of the intervention, and (c) assessing the maintenance of any staff or resident behavior change 2 months later. Descriptive statistics were used to address the first question. For the second analytical question, analyses were run by calculating 2 (group) × 3 (time) mixed-factor analysis of variance (ANOVA), where the between-subjects factor Group included intervention and control and the within-subjects factor Time included baseline, PIA–I, and PIA–II phases. The primary outcome of interest in these analyses were significant Group × Time interaction effects, showing that the intervention group improved in the targeted behavioral outcomes for CNAs and residents over and above any improvements seen in the control group as a result of the passage of time or observation. For these interaction effects, Greenhouse-Geisser p values are reported throughout in order to correct for potential violations to the sphericity assumption (Winer 1971). Finally, a 2 (group) × 2 (time) mixed-factor ANOVA was run comparing PIA–II outcomes to follow-up. The primary outcome of interest in these analyses were significant main effects of Group, showing that intervention participants maintained any behavior change. Results Treatment Receipt and Enactment Memory Book Availability Observational data from research staff memory book checks indicated that residents were in possession of their memory books during 72% of the morning checks and 75% of the afternoon checks during the hands-on training phase. This figure remained steady or increased during the PIA–I (70% morning, 73% afternoon), PIA–II (78% morning, 81% afternoon), and 2-month follow-up phases (77% morning, 80% afternoon). CNA/LPN Training and Performance of the Staff Motivational System Thirty-nine CNAs with data at all three assessment times (baseline, PIA–I, PIA–II) completed the hands-on training phase on the intervention units. Of these 39 CNAs, 92% (n = 36) passed a final evaluation with a performance score of 80% or above in the use of communication skills as measured by the CSC. The average final evaluation score for CNAs was 84.77% (range 53%–97%). The average number of training sessions on the unit per CNA was 6 (range 1–13). CNAs completed an average of 64% of their self-monitoring forms during the training phase (SD = 28%; range 5%–100%). The percentage of self-monitoring forms completed increased somewhat during the postintervention phases (PIA–I M = 67%, SD = 33%, range 0%–100%; PIA–II M = 66%, SD = 36%, range 0%–100%). CNAs were not required to self-monitor during the follow-up phase. Across the five nursing homes, a total of 55 written feedback letters regarding CNAs' skill performance were distributed. Fifty-eight percent of these letters were positive; 42% were negative. The number of supervisory training sessions conducted with 20 LPNs ranged between two and four. One hundred percent of LPNs reached a criterion of 80% correct performance in their accuracy of CSC recording and the provision of verbal feedback to CNAs regarding communication skill performance by the end of the hands-on training phase. On average, during the training phase LPNs conducted 62% (SD = 36%, range 0%–100%) of their assigned observations of CNAs. This number remained relatively stable after the hands-on training phase when LPNs conducted CNA observations independently (i.e., not accompanied by the project manager). During PIA–I, LPNs conducted 57% (SD = 44%, range 0%–100%) of their assigned independent observations, and during PIA–II LPNs conducted 61% (SD = 46%, range 0%–100%). CNA Use of Communication Skills There were no baseline differences between intervention and control groups on the CSC-generated overall communication skills score, time spent giving care to residents, specific one-step instructions, multistep instructions, positive statements, or biographical statements. Results of the 2 (group) × 3 (time) ANOVA analyses are shown in Fig. 1Fig. 2Fig. 3. The mixed-factor ANOVA for the overall total correct communication skill percentage score revealed significant main effects of group, F(1,62) = 16.87, p = .0001, and time, F(2,124) = 37.30, p = .0001, and a significant Group × Time interaction, F(2,124) = 17.20, p = .0001 (see Fig. 1). CNAs in the intervention group improved their use of general communication skills with residents during care in comparison with NTC group CNAs at both PIA–I and PIA–II. More specifically, CNAs in the intervention group tended to increase their rate per hour use of specific, one-step instructions during care in comparison with control group CNAs at both PIA–I and PIA–II (see Fig. 2), as shown by a marginally significant Group × Time interaction, F(2,124) = 2.80, p = .06. Intervention CNAs significantly increased their rate of positive statements made to residents during care in comparison with control group CNAs at both PIA–I and PIA–II (see Fig. 3), as shown by a significant Group × Time interaction, F(2,124) = 6.16, p = .004. There was also a significant group effect with intervention CNAs using fewer multistep instructions, F(1,62) = 5.21, p = .03, but no Group × Time interaction. There were no differences between intervention and control group CNAs across time in use of biographical statements. Notably, there was also no difference between intervention and control group CNAs across time in the amount of time spent in daily care with residents. Forty-three residents (21 control, 22 intervention) and 29 CNAs (13 control, 16 intervention) had data available from all assessment points from baseline assessment through the 2-month follow-up (i.e., maintenance sample). Analyses indicated that there were no significant differences in the characteristics of residents or CNAs who survived to follow-up and those who completed baseline. There were also no differences between CNAs on the control and intervention units at the 2-month follow-up assessment. The only significant difference between residents on the control and intervention units at the 2-month follow-up was that residents on intervention units were more independent in self-care as measured by the FIM, F(1,41) = 4.33, p = .04. Results of the ANOVAs show significant main effects of group in overall communication skills score on the CSC, F(1,27) = 12.92, p = .001 (Fig. 1) and the use of positive statements, F(1,27) = 11.91, p = .002 (Fig. 3), indicating maintenance of change at the 2-month follow-up assessment point. CNA–Resident Interactions During Care Results of the 2 (group) × 3 (time) mixed-factor ANOVA of the CABOS care interaction data are shown in Fig. 4 and Fig. 5. There were significant main effects of group, F(1,65) = 5.37, p = .02, and time, F(2,130) = 8.57, p = .0003, and a significant Group × Time interaction, F(2,130) = 4.11, p = .02, for the total amount of staff speech directed toward residents (see Fig. 4). CNAs in the intervention group increased their verbal interaction with residents during care in comparison with control group CNAs at PIA–II. Regarding resident coherent verbal interaction, only the main effects of group, F(1,65) = 6.50, p = .01, and time, F(2,130) = 3.20, p = .05, were significant. However, the rate of positive verbal interactions between CNAs and residents during care routines increased in the intervention group across time, as shown in Fig. 5 by a significant Group × Time interaction, F(2,130) = 4.81, p = .01. Analysis of maintenance data indicated that there were significant main effects of group for total amount of staff speech, F(1,41) = 5.20, p = .03 (Fig. 4), and positive statements by either staff or resident (i.e., positive verbal interactions), F(1,41) = 5.83, p = .02 (Fig. 5). Intervention group CNAs maintained a higher rate of speech with residents during care, and there were more positive statements made during care in CNA–resident dyads in the intervention group. The occurrence of agitation during care among the 67 residents was extremely low during all phases of the study (e.g., on average less than 1% of observation time). Thus, changes in this behavior could not be assessed. Discussion The results of this study suggest that, through the use of communication–memory book skills training and a staff motivational system, CNAs can improve aspects of their communication with nursing home residents. CSC data, derived from direct observation of care interactions, showed that CNAs' overall communication skills level increased significantly as compared with an NTC group. CNAs increased their use of positive statements and showed a marginal increase (p = .06) in single, one-step instructions, which can be less confusing than multiple nonspecific instructions when interacting with individuals with cognitive impairments. Moreover, CABOS data showed a significant increase in the amount of staff speech directed toward the resident during care routines. Finally, a behavioral code measuring the amount of positive statements during CNA–resident verbal interactions also showed a significant increase. Importantly, all of the changes in staff behavior were maintained at an assessment conducted 4 months after the initiation of treatment. Thus, the current data would appear to corroborate pilot data from Allen-Burge and colleagues 2001 and the findings of McCallion and colleagues 1999. These changes in staff behavior were brought about without increasing the amount of time necessary to deliver daily care. It has been argued that even if CNAs can be taught to play a more therapeutic role in the nursing home, it would require a time commitment that might not be feasible in the current nursing home environment (Schnelle and Beck 1999). For example, Rogers and colleagues 1999 reported success in training research therapists to increase independent dressing in residents; however, their therapeutic procedure required more time for staff to complete. Our data suggest that using better communication skills and memory books during care routines does not require more staff time. It is important that researchers delineate which therapeutic routines do and do not require additional time commitments so that staffing adjustments can be made when planning therapeutic routines. Contrary to our hypothesis, residents in the intervention group did not improve their rate of coherent verbal interactions with CNAs during care routines in comparison with control group residents. As discussed above, there was a significant increase in the rate of positive statements during CNA–resident interactions after training. However, due to our definition of this behavioral code, we did not examine how much residents contributed to this increase. Moreover, we were unable to assess any positive effects of the intervention on disruptive behavior due to its extremely low rate of occurrence in this sample. In this study, residents were required to show minimal ability in verbal communication involving spontaneous speech for entry. Consequently, residents were, on average, moderately cognitively impaired (MMSE M = 13.39), and prior research has indicated that moderately cognitively impaired residents are less likely to display serious disruptive behaviors (Burgio et al. 1994; Cohen-Mansfield, Werner, and Marx 1990). A main feature of this study was the careful attention given to the assessment of treatment implementation (Lichstein et al. 1994). Although the assessment of treatment implementation has long been a standard in nongerontological psychosocial intervention research (Cook and Campbell 1979; Moncher and Prinz 1991; Sechrest, West, Phillips, Redner, and Yeaton 1979), it has seldom been used in gerontological intervention research (Burgio, Corcoran, et al. 2001). Aspects of treatment receipt (e.g., CNAs' understanding of and use of the treatment) and treatment enactment (e.g., use of the staff motivational system and memory books by CNAs and LPNs) were investigated. The methodological advantages of directly assessing the implementation of interventions in applied settings are numerous and include greater assurance in the internal and external validity of the study. Internal validity concerns whether observed changes in behavior in CNA and resident outcomes actually coincide with increased use of the communication–memory book skills by CNAs. Measurements of treatment implementation also provide guidelines for the assessment of external validity by indicating what level of enactment of therapeutic skills by indigenous nursing staff was necessary to produce the observed therapeutic gains. Our results show that both CNAs and LPNs received (i.e., learned) the intervention as we intended. Ninety-two percent of CNAs demonstrated through direct observational evaluation that they could perform communication skills at the 80% criterion. One hundred percent of the LPNs reached a criterion of 80% accuracy of completing the CSC as compared with the project manager's completion of the CSC. Demonstrating skill acquisition in no way guarantees that staff will use these skills in the clinical setting. However, our assessment of treatment enactment shows that memory books were available to staff and residents during an average of 76% of observations. Their availability remained constant after the intervention was transferred completely to indigenous staff. CNAs completed an average of 66% of their self-monitoring forms. LPNs sent CNA performance feedback memos as instructed, and they completed an average of 60% of assigned observations of CNAs with the CSC form. This also remained relatively constant throughout all phases of the study. Although the completion of CNA self-monitoring forms and LPN CSC observations were only in the 60% range, considering the workload of these personnel we consider this to indicate successful, though less than ideal, treatment enactment. There are several limitations of this study. First, the reported CNA turnover rate in the nursing homes was 49%, well below the national average of 97% (American Health Care Association, 1997, cited in Harrington et al. 2000). Large staff turnover rates present significant difficulties in implementing psychosocial interventions, and our results may not be replicable in more typical nursing homes with larger turnover rates. In this study, newly hired staff viewed videotapes of the inservice, and hands-on training was also provided. Still, it is possible that continuous training of incoming staff in nursing homes with turnover near 100% may present challenging logistical problems. In all studies using direct observational measurement, participant reaction to observation is a potential problem. Specifically, the CNAs' performance of skills during observation may be influenced by social desirability factors and may not be representative of their actual day-to-day performance. There is no completely satisfactory answer to this problem, and we do not know to what degree social desirability influenced the CNA results. However, in this and our prior studies, we have developed specific procedures for observing staff unobtrusively; thus, and presumably, limiting reactivity (see Burgio 1996, and Burgio, Scilley, Hardin, and Hsu 2001, for a more detailed discussion of this issue). Multiple outcome measures were collected to provide a comprehensive assessment of treatment effects. This was useful for broadly identifying effects on CNA behavior, resident responses, and CNA–resident interactions. However, the Type I error rate is undoubtedly inflated by the fact that no adjustments were made for each individual analysis. A Bonferroni correction might be considered, although this procedure markedly elevates the Type II error rate and decreases power. Given the relatively limited sample size available for this initial efficacy evaluation, we chose to not implement this more conservative approach. Exact probability levels are reported instead so that the reader can make any interpretive adjustments deemed necessary. Another potential limitation was our choice to randomize nursing units within each nursing home into groups instead of randomizing entire nursing homes. Because CNAs were occasionally pulled to work on other nursing units, it is possible that trained CNAs applied communication skills on control units. This is a potential threat to the study's internal validity. However, we believe that this threat was minimized by including only nursing homes using fixed staffing and minimal pulling. The alternative of randomizing groups to nursing homes presents its own methodological problems, including the difficulty of finding nursing homes that are equivalent on all factors that might affect outcome. We have written about this dilemma in more depth elsewhere (Burgio and Stevens 1999). Because this was an evaluation of a multicomponent treatment package, we do not know the separate contribution of communication training and memory book usage to the changes in staff behavior. Perhaps more important, we do not know whether the inclusion of a staff motivational system was necessary for producing and maintaining staff behavior change. Studies that have used similar training techniques, including hands-on training, have generally shown an immediate training effect but no maintenance of behavior change (e.g., Schnelle, Newman, and Fogarty 1990). In a study currently under review (Burgio, Stevens, et al. 2001) we compared a behavior management skills training package with and without a staff motivational system. As hypothesized, results showed an immediate training effect for both groups, but maintenance of CNAs' behavior change was demonstrated more frequently at a 6-month follow-up in the group that received a staff motivational system. In conclusion, the results of this study suggest that CNAs can be trained to use improved communication skills and memory books with their residents during care interactions. With the use of a staff motivational system integrated into the training program, staff will perform these skills and will maintain this performance up to 4 months after the initiation of training. Although increases were observed in the overall amount of staff speech during interactions, use of positive statements, and positive statements during dyadic interactions between staff and residents, there is no direct evidence of changes in the resident communication behaviors targeted in this study. Future studies may want to examine other resident behaviors that may be influenced by improved CNA communication skills. One candidate for investigation is resident affect, which has been shown to be sensitive to change due to intervention in residents with dementia (Lawton, Van Haitsma, and Klapper 1996). Finally, although management of the intervention program was transferred to indigenous staff midway through the intervention phase, this was predominantly an efficacy trial. To optimize the internal validity of the study, training was conducted by a PhD-level licensed clinical psychologist, both inservice and hands-on training were used, and, as discussed above, nursing homes with relatively low rates of CNA turnover were chosen. Although it is our belief that this intervention can be manualized and implemented by typical staff development personnel in most nursing homes, only an effectiveness trial can shed light on the true feasibility of this intervention program. Practice Concepts The Forum Book Reviews Table 1. Surviving Resident Characteristics by Group Variable Intervention (n = 34) No-Treatment Control (n = 33) F /χ2 (df) Agea 81.78 (8.88) 82.42 (7.10) 0.10 (1,65) Mini-Mental Status Examinationa 13.50 (6.72) 12.94 (6.00) 0.13 (1,65) Functional Independence Measurea 46.88 (22.26) 37.30 (18.79) 3.61 (1,65) No. medicationsa 15.59 (9.79) 15.52 (8.49) 0.00 (1,65) Gender (women)b 73.53 75.76 0.04 (1) Race (White)b 76.47 90.91 2.54 (1) Dementiab,c 52.94 72.73 2.80 (1) a Characteristics reported as means (SD). b Characteristics reported as percentages (SD). c Dementia is listed as a diagnosis in the medical chart. Table 2. Surviving Staff Characteristics by Group Variable Intervention (n = 37)a No-Treatment Control (n = 25) F /χ2 (df) Ageb 39.30 (11.46) 35.13 (7.05) 2.43 (1,56) Years of educationb 13.57 (0.99) 14.00 (0.85) 3.01 (1,58) Months at facilityb 46.79 (50.62) 40.41 (53.71) 0.21 (1,56) Months as certified nursing assistantb 111.50 (81.26) 106.78 (64.94) 0.05 (1,57) Gender (women)c 86.49 88.00 0.03 (1) Race (Black)c 86.49 88.00 1.59 (1) a Demographic data were available for only 37 of the 39 CNAs in the intervention group. b Characteristics reported as means (SD). c Characteristics reported as percentages (SD). Figure 1. Mean total percentage correct on the certified nursing assistant Communication Skills Checklist across study phases for control (white bars) and intervention (black bars) groups. Asterisks denote significant differences between groups at p < .05. PIA–I = Postintervention Assessment I; PIA–II = Postintervention Assessment II. Figure 2. Mean rate per hour of specific, one-step instructions by certified nursing assistants across study phases for control (white bars) and intervention (black bars) groups. Asterisks denote significant differences between groups at p < .05. PIA–I = Postintervention Assessment I; PIA–II = Postintervention Assessment II. Figure 3. Mean rate per hour of positive statements by certified nursing assistants during care across study phases for control (white bars) and intervention (black bars) groups. Asterisks denote significant differences between groups at p < .05. PIA–I = Postintervention Assessment I; PIA–II = Postintervention Assessment II. Figure 4. Mean percentage of total observation time across study phases during which certified nursing assistants were speaking to residents for control (white bars) and intervention (black bars) groups. Asterisks denote significant differences between groups at p < .05. PIA–I = Postintervention Assessment I; PIA–II = Postintervention Assessment II. Figure 5. Mean rate per hour of positive statements made by certified nursing assistants or residents during care across study phases for control (white bars) and intervention (black bars) groups. Asterisks denote significant differences between groups at p < .05. PIA–I = Postintervention Assessment I; PIA–II = Postintervention Assessment II. Portions of this article were presented at the 52nd annual scientific meeting of the Gerontological Society of America, San Francisco, CA, November 1999. The research reported in this article was supported by funding from the National Institute on Aging (RO1AG13008) to M. Bourgeois and L. Burgio. We thank the nurses, certified nursing assistants, and administrative staff of Civic Center Nursing Home, Pleasant Grove Health Care Center, St. Martin's in the Pines, Montclair East Nursing Home, and Shelby Ridge Nursing Home for their support and assistance. Special thanks are extended to Christy Fargason for assistance in data entry; to Shermetra DeLaine, Rachael Ragsdale, and Shane Thomas for assistance in data collection; and to Debbie Turpin, Jeff Phillips, and Karen Quarles for assistance in manuscript preparation. Allen-Burge R., Burgio L. D., Bourgeois M. S., Sims R., Nunnikhoven J., 2001. Increasing communication among nursing home residents. Journal of Clinical Geropsychology 7:213-230. Allen-Burge R., Stevens A. B., Burgio L. D., 1999. Effective behavioral interventions for decreasing dementia-related challenging behavior in nursing homes. International Journal of Geriatric Psychiatry 14:213-232. Boehm S., Whall A. L., Cosgrove K. L., Locke J. D., Schlenk E. A., 1995. Behavioral analysis and nursing interventions for reducing disruptive behaviors of patients with dementia. Applied Nursing Research 8: (3) 118-122. Bourgeois M., 1990. Enhancing conversation skills in Alzheimer's disease using a prosthetic memory aid. Journal of Applied Behavior Analysis 23:29-42. Bourgeois M., 1992. Conversing with memory impaired individuals using memory aids Northern Speech Services, Gaylord, MI. Bourgeois M., 1992. Evaluating memory wallets in conversations with patients with dementia. Journal of Speech and Hearing Research 35:1344-1357. Bourgeois M., 1993. Effects of memory aids on the dyadic conversations of individuals with dementia. Journal of Applied Behavior Analysis 26:77-87. Bourgeois M. S., Burgio L. D., Schulz R., Beach S., Palmer B., 1997. Modifying repetitive verbalizations of community-dwelling patients with AD. The Gerontologist 37:30-39. Bourgeois M. S., Mason L. A., 1996. Memory wallet intervention in an adult day-care setting. Behavioral Interventions 11: (1) 3-18. Burgio L. D., 1996. Direct observations of behavioral disturbances of dementia and their environmental context. International Psychogeri-atrics 8: (3) 343-349. Burgio L. D., Burgio K. L., 1990. Institutional staff training and management: A review of the literature and a model for geriatric long-term care facilities. International Journal of Aging and Human Development 30:287-302. Burgio L. D., Corcoran M., Lichstein K. L., Nichols L., Czaja S., Gallagher-Thompson D., Bourgeois M., Stevens A., Ory M., Schulz R., 2001. Judging outcomes in psychosocial interventions for dementia caregivers: The problem of treatment implementation. The Gerontologist 41:481-489. Burgio L. D., Engel B. T., Hawkins A. M., McCormick K., Scheve A. S., 1990. A descriptive analysis of nursing staff behaviors in a teaching nursing home: Differences among CNAs, LPNs, and RNs. The Gerontologist 30:107-112. Burgio L. D., Scilley K., 1994. Caregiver performance in the nursing home: The use of staff training and management procedures. Seminar in Speech and Language 15:313-322. Burgio L. D., Scilley K., Hardin J. M., Hsu C., 2001. Temporal patterns of disruptive vocalization in elderly nursing home residents. International Journal of Geriatric Psychiatry 16:378-386. Burgio L. D., Scilley K., Hardin J. M., Hsu C., Yancey J., 1996. Environmental "white noise": An intervention for verbally agitated nursing home residents. Journal of Gerontology: Psychological Sciences 51B:P354-P373. Burgio L. D., Scilley K., Hardin J. M., Janosky J., Bonino P., Slater S. C., Engberg R., 1994. Studying disruptive vocalization and contextual factors in the nursing home using computer-assisted real-time observation. Journal of Gerontology: Psychological Sciences 49:P230-P239. Burgio L. D., Stevens A. B., 1999. Behavioral interventions and motivational systems in the nursing home. Schulz R., Maddox G., Lawton M. P., , ed.Annual review of gerontology and geriatrics: Vol. 18. Focus on interventions research with older adults 284-320. Springer, New York. Burgio, L. D., Stevens, A., Burgio, K. L., Roth, D. L., Paul, P., & Gerstle, J. (2001). Teaching and maintaining behavior management skills in the nursing home. Manuscript submitted for publication. Carstensen L. L., Fisher J. E., Malloy P. M., 1995. Cognitive and affective characteristics of socially withdrawn nursing home residents. Journal of Clinical Geropsychology 1: (3) 207-218. Cohen J., 1968. Weighted kappa: Nominal scale agreement with provision for scaled disagreement or partial credit. Psychological Bulletin 70:213-220. Cohen-Mansfield J., Werner P., Marx M., 1990. Screaming in nursing home residents. Journal of the American Geriatrics Society 38:785-792. Cook T. D., Campbell D. T., 1979. Quasi-experimentation: Design and analysis issues for field settings Rand-McNally, Chicago. Coon D. W., Schulz R., Ory M. G., 1999. Innovative intervention approaches for Alzheimer's disease caregivers. Beigel D., Blum A., , ed.Innovations in practice and service delivery across the lifespan 295-325. Oxford University Press, New York. Daniels A. C., 1994. Bringing out the best in people: How to apply the astonishing power of positive reinforcement McGraw-Hill, New York. Folstein M., Folstein S., McHugh P., 1975. "Mini-Mental State." A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatry Research 12:189-198. Hallberg I. R., Norberg A., Eriksson S., 1990. A comparison between the care of vocally disruptive patients and that of other residents at psychogeriatric wards. Journal of Advanced Nursing 15:410-416. Hamilton B. B., Laughlin J. A., Fiedler R. C., Granger C. V., 1994. Inter-rater reliability of the 7-level Functional Independence Measure (FIM). Scandinavian Journal of Rehabilitation Medicine 26:115-119. Harrington C., Kovner C., Mezey M., Kayser-Jones J., Burger S., Mohler M., Burke R., Zimmerman D., 2000. Experts recommend minimum staffing standards for nursing facilities in the United States. The Gerontologist 40:5-16. Hartmann D. P., , ed.New directions for methodology of social and behavioral science: Using observers to study behavior 1982Jossey-Bass, San Francisco. Hays W. L., 1994. Statistics 5th ed. Holt, Rinehart, & Winton, New York. Hoeffner B., Radar J., McKenzie D., Lavelle M., Stewart B., 1997. Reducing aggressive behavior during bathing cognitively impaired nursing home residents. Journal of Gerontological Nursing 23: (5) 16-23. Hoerster, L., Hickey, E., & Bourgeois, M. (in press). Effects of memory aids on conversations between nursing home residents with dementia and nursing assistants. Neuropsychological Rehabilitation. Kidd D., Stewart G., Baldry J., Johnson J., Rossiter D., Petruckevitch A., Thompson A. J., 1995. The Functional Independence Measure: A comparative validity and reliability study. Disability and Rehabilitation 17: (1) 10-14. Lawton M. P., Van Haitsma K., Klapper J., 1996. Observed affect in nursing home residents with Alzheimer's disease. Journal of Gerontology: Psychological Sciences 51B:P3-P14. Lichstein K. L., Riedel B. W., Grieve R., 1994. Fair tests to clinical trials: A treatment implementation model. Advances in Behavior Research and Therapy 16:1-29. McCallion P., Toseland R. W., Freeman K., 1999. An evaluation of a family visit education program. Journal of the American Geriatric Society 47: (2) 203-214. McCallion P., Toseland R. W., Lacey D., Banks S., 1999. Educating nursing assistants to communicate more effectively with nursing home residents with dementia. The Gerontologist 39:546-558. Moncher F. J., Prinz R. J., 1991. Treatment fidelity in outcome studies. Clinical Psychology Review 11:247-266. Pfeiffer E., 1975. A Short Portable Mental Status Questionnaire for the assessment of organic brain deficit in elderly patients. Journal of the American Geriatrics Society 23:433-441. Reid D. H., Parsons M. B., Green C. W., 1989. Staff management in human services: Behavioral research and application Charles C Thomas, Springfield, IL. Repp A. C., Karsh K. G., van Acker R., Felce D., Harman M., 1989. A computer-based system for collecting and analyzing observational data. Journal of Special Education Technology 9:207-216. Ripich D. N., Wykle M., Niles S., 1995. Alzheimer's disease caregivers: The focused program. Geriatric Nursing 16: (1) 15-19. Rogers J. C., Holm M. B., Burgio L. D., Granieri E., Hsu C., Hardin J. M., McDowell B. J., 1999. Improving morning care routines of nursing home residents with dementia. Journal of the American Geriatrics Society 47: (9) 1049-1057. Schnelle J. F., Beck C., 1999. Costs of promoting independence. Journal of the American Geriatrics Society 47: (9) 1151-1153. Schnelle J. F., Newman D. R., Fogarty T., 1990. Management of patient continence in long-term care nursing facilities. The Gerontologist 30: (3) 373-376. Schnelle J. F., Ouslander J. G., Osterweil D., Blumenthal S., 1993. Total quality management: Administrative and clinical applications in nursing homes. Journal of the American Geriatrics Society 41:1259-1266. Sechrest L., West S. G., Phillips M. A., Redner R., Yeaton W., 1979. Some neglected problems in evaluation research: Strength and integrity of treatments. Sechrest L., West S. G., Phillips M. A., Redner R., Yeaton W., , ed.Evaluation studies review annual (Vol. 4) 15-35. Sage, Beverly Hills, CA. Stevens A., Burgio L. D., Bailey E., Burgio K. L., Paul P., Capilouto E., Nicovich P., Hale G., 1998. Teaching and maintaining behavior management skills with nursing assistants in a nursing home. The Gerontologist 38:379-384. Winer B. J., 1971. Statistical principles in experimental design 2nd ed. McGraw-Hill, New York.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B49154E14BFAF64F435F6C799AF7E239785764F.txt b/test/dataset/in/resources/corpus/Clean_0B49154E14BFAF64F435F6C799AF7E239785764F.txt
deleted file mode 100644
index 0b7d070..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B49154E14BFAF64F435F6C799AF7E239785764F.txt
+++ /dev/null
@@ -1 +0,0 @@
- gerona J Gerontol A Biol Sci Med Scigerona The Journals of Gerontology Series A: Biological Sciences and Medical Sciences J Gerontol A Biol Sci Med Sci 1079-5006 1758-535X Oxford University Press 020035BS10.1093/gerona/57.11.B400 Journal of Gerontology: Biological Sciences Designer Microarrays From Soup To Nuts Wang Eugenia a Lacelle Chantale a b Xu Suying a Zhao Xuechun a Hou Michael a aDepartment of Biochemistry and Molecular Biology, University of Louisville School of Medicine, Kentucky bDepartment of Anatomy and Cell Biology, McGill University, Montréal, Canada Eugenia Wang, Department of Biochemistry and Molecular Biology, University of Louisville School of Medicine, 570 South Preston Street, Baxter Building Room 304, Louisville, KY 40292 E-mail: eugenia.wang@louisville.edu. 1 11 2002 57 11 B400 B405 7 8 2002 28 5 2002 The Gerontological Society of America 2002 The recognition that multigene mechanisms control the pathways determining the aging process renders gene screening a necessary skill for biogerontologists. In the past few years, this task has become much more accessible, with the advent of DNA chip technology. Most commercially available microarrays are designed with prefixed templates of genes of general interest, allowing investigators little freedom of choice in attempting to focus gene screening on a particular thematic pathway of interest. This report describes our “designer microarray” approach as a next generation of DNA chips, allowing individual investigators to engage in gene screening with a user friendly, do-it-yourself approach, from designing the probe templates to data mining. The end result is the ability to use microarrays as a platform for versatile gene discovery. hwp-legacy-fpage B400 hwp-legacy-dochead RESEARCH ARTICLE Decision Editor: James R. Smith, PhD OVER the past century, as a result of increasing medical knowledge, better sanitation, and improved nutrition, developed countries are experiencing an unprecedented increase in average human life span, along with a higher incidence of multifactorial diseases such as cardiovascular diseases, neurodegenerative disorders, type 2 diabetes, and cancers (1). These age-dependent diseases, plaguing people as young as their mid-50s, are products of the combined influences of genetics and environment (2). Nature and nurture together provide predispositions to cancer, cardiovascular disease, diabetes, and neurodegenerative disorders, presenting a complex picture for the development of these perils in the fast-growing middle- and old-age subpopulations of our society. Although recent advances in medical research have enabled us to diagnose several age-associated diseases, alleviate pain associated with them, and retard the onset of their acute stages, we remain largely incapable of identifying at an early stage those individuals bearing genetic predispositions to these diseases, and thus of administering preventive medicine or treatment. Because the human genome contains some 30,000 genes (3), and modern industrialized society yields increasing environmental complexity, it is an ever-greater challenge to perceive how the integration of our genes and surrounding environment creates disease-predisposed states. For example, why do certain individuals suffer lung cancer at an early age, after a few years of cigarette smoking, whereas some centenarians tolerate lifelong smoking without dying of the same disease? Such questions led to the idea of the need to identify “genetic signatures.” Once genetic signatures are secured, one may develop means to prevent and/or treat diseases in an individual manner, creating individualized medicine for prognostic, diagnostic, and therapeutic purposes. In general, large-sample gene chips, bearing perhaps 10,000 genes, are applicable only to early-stage screening and may yield voluminous lists of potential positive results; here we describe a next-generation, medium-density microarray approach embodying considerable quality control in both chip design and analysis, which results in fewer hits of enhanced accuracy and pertinence. Genetic Signatures and Microarray Technology During the 1980s and early 1990s, biologists were busy dissecting the functions of single genes, by using a reductionist approach, which, although thorough in its exact methodological analysis of genetic impact, was restricted in explaining how each particular single gene functions in the context of many homologous genes or partners to accomplish a biological task. In an attempt to shed light on these biological tasks, “biochips” were introduced in the mid-1990s (4) as a new tool for molecular medicine; they now constitute a rapidly developing field of biomedical research, which combines molecular medicine, nanotechnology, computer science, and engineering. Biochip technology was developed for high-throughput gene screening, capable of simultaneously identifying changes in the expression of hundreds or thousands of genes, and thus genetic signatures defining particular physiological states. Consequently, as a result of its power, the young field of DNA microarray technology has rapidly gathered speed and popularity within the biomedical research community. However, as universities across North America have started to establish microarray core facilities, they are realizing that the next generation of microarrays must be more versatile, user friendly, and inexpensive to ensure that these facilities will meet the divergent needs of their researchers, and ultimately provide practical answers to fundamental biological problems. Thus, 4 years ago, parallel to the development of commercial microarrays, we began devoting significant time and effort to developing pathway-specific “designer biochips”; here we provide an account of our quest for designer arrays, and we discuss some of the challenges that lie ahead for those seeking the perfect array. The Theory Behind Microarrays Unlike Northern blots, in microarrays the probes, not the targets, are immobilized to a physical platform. Microarrays consist of a platform to which are independently attached numerous nucleic acid sequences known as probes, to screen targets of prelabeled nucleic acids obtained from donors (human or animal). The quantification of probe–target complexes formed during hybridization permits measuring gene distribution and intensity, as complementary probe-labeled target sequences bind together. Although the principle behind microarrays is simple, creating and implementing microarray technology is difficult, as several parameters (discussed in the paragraphs that follow) can drastically affect the validity of the results obtained from microarrays. Furthermore, because the number of probes included on each microarray platform is great, the magnitude of results obtained from microarrays is huge, and thus requires powerful computerized image processing and statistical software to classify and analyze the data; without these, little significant gain can be obtained from using microarrays. Thus, microarray core facilities must integrate expertise in biology, computer science, engineering, and statistics. It is with this in mind that we started our quest for designer biochips. Platform and Printing Robots The first consideration when developing microarray technology is the type of platform, commonly either membrane or glass slide (5)(6)(7)(8)(9). Membrane-based microarrays use nylon or nitrocellulose membranes, and they are generally used when radioactive or colorimetric tags are used to label the targets, whereas glass slides are more suitable for microarrays using fluorescent tagging; we chose to use a membrane-based platform for our microarrays. Unlike glass slides, which must be chemically treated to permit the attachment of probes and decrease background fluorescence before use as a platform for microarrays (5)(8)(9)(10)(11)(12)(13), most membranes (whether bearing positive, negative, or neutral charge) require no such pretreatment. Because nucleic acids are negatively charged, the use of positive or neutral membranes greatly improves the signal; however, neutral membranes are generally more suitable, as positive-charged membranes yield higher background readings as well. Negatively charged membranes, although providing very little background, are usually not suitable because they yield poor signal. Thus we chose to use neutral-charged membranes for our arrays. Once a microarray platform has been chosen, probes must be attached to the platform. Two obvious methods exist: synthesis of probes directly on the platform (14)(15)(16)(17)(18), and probe-spotting by use of a contact or noncontact printing robot (4)(19)(20)(21)(22). Although leading biochip companies often synthesize oligos directly on their microarrays by using techniques such as photolithography, this method is not easily mastered, nor accessible to most laboratories. In contrast, probe-spotting can be accomplished using any of several commercially available printing robots (22). Because we use membranes attached to glass slides as our platform instead of glass slides directly, we encountered several problems, including skipped dots and uneven printing, when we first attempted to print arrays. We had to substantially modify the printing heads of the first robot we purchased, and we had to build an enclosure over it that permitted maintenance of constant humidity, to ensure even printing of the probes. Probes Almost any type of nucleic acid can be printed with a printing robot; nucleic acids obtained from complementary DNA (cDNA) libraries, as well as polymerase chain reaction (PCR) products generated by reverse transcription PCR (RT-PCR), are commonly used as probes for microarrays. The use of cDNA libraries is of considerable value for screening previously unknown genes or a great many genes with no predefined preferences for certain gene families. However, mistakes arising from mislabeling of clones or contamination can cause problems (23). We spot PCR products on our chips to generate thematic arrays of particular known genes of interest. Commercially available biochips are often restricted to specific sets of genes contained on each biochip, posing user-unfriendly conditions. In the gene discovery task, users are generally conditioned to screen according to the preset configuration of genes, without the possibility of generating thematic or pathway-specific microarrays covering genes known to belong to a specific family, as demanded by a particular research program. The use of designer biochips can circumvent this problem. The following sections describe the strategy we used to fabricate thematic arrays. Thematic Microarray Design Gene Selection and Primer Design To generate thematic microarrays bearing genes from a particular family or the same pathway, we use several public databases, including Unigene and GenBank (http://www.ncbi.nlm.nih.gov/GenBank, http://www.ncbi.nlm.nih.gov/Unigene), and conduct an extensive literature search (http://www.ncbi.nlm.nih.gov/PubMed) to obtain a repertoire of genes. Once a list of potential genes is constructed, we use GenBank and Unigene to obtain the sequences of all candidate genes. Primers for each gene sequence are designed using Primer3 software (http://www.genome.wi.mit.edu/genome_software/other/primer3.html) to generate a PCR product, or “amplicon,” with a length between 300 and 700 bp and a melting constant that ranges from 75°C to 89°C. For each gene we choose a pair of sense and antisense primers with an annealing temperature of approximately 60°C and an average length of 23 bp, for amplicon production using 96-well PCR plates. These clustered amplicon sizes and melting constants support standardized hybridization conditions for all probes across the microarray platform, including stringent washing, thus decreasing nonspecific signals while maximizing a specific signal. The length of the PCR product, as documented by Stillman and Tonkinson (6), is particularly important in maximizing a specific signal. Primer design is perhaps the most time-consuming step in our microarray production, because once a primer pair is selected, an analysis must be performed with Blast (proprietary software available on the National Center for Biotechnology Information website) to ensure that each primer pair amplifies only the gene of interest. This is crucial, because results obtained from the microarray are dependent on the specificity of the amplicons. However, in some instances, the specificity of the primer may not guarantee the specificity of the generated amplicon, when a conserved or shared domain lurks somewhere within the sequence. It is therefore highly recommended that the entire amplicon sequences themselves be Blasted to identify homologous regions, which can cause nonspecific binding. With probes obtained from cDNA libraries this may become a pitfall, especially when the spotted nucleic acid sequence is unknown; often highly homogenous sequences may result in nonspecific binding between genes of high homology. Stringent hybridization conditions and washing can generally eliminate this nonspecific binding, if the homologous region is not too long. Controls As in any biological experiment, and most importantly for microarrays, controls must be carefully selected. It is important to spot on all microarrays negative and positive controls as well as “housekeeping genes,” used in more traditional experiments such as quantitative RT-PCR, which show little or no physiological change in expression among the subjects or conditions being studied. The inclusion of housekeeping genes is useful for data normalization; for our designer microarrays targeted to mouse models, we selected six mouse genes (glyceraldehyde phosphate dehydrogenase, ribosomal S6, beta-actin, hypoxanthine-guanine phosphoribosyltransferase, phospholipase A2, and ubiquitin) commonly used in the literature as controls. In general, the validity of these controls must be determined a priori by using independent tests, such as Northern blotting assays or quantitative RT-PCR (24). For instance, EF-1α would be a poor choice for a housekeeping gene if the target nucleic acids were obtained from skeletal muscle, as it is not expressed in adult muscle cells; it would, however, be a good control when cDNA from liver is used as a target (25). The use of housekeeping genes permits the measuring of changes in gene expression against a gene whose expression does not vary significantly; in some cases this can be of great value. Negative controls should include buffer, bacterial, and viral DNA, as well as amplicons from genes known not to be expressed in target tissues. Negative controls are used to assess the level of background noise arising from nonspecific nucleic acid binding during probe–target hybridization. Positive controls such as total cDNA or genomic DNA permit the detection of suboptimal conditions of hybridization and staining, which may obscure appropriate signal intensity. Quality Control for Amplicon Production In order to avoid producing the wrong amplicon for printing as a result of contaminated PCR reactions, the use of dedicated equipment and reagents in the PCR setup and reaction areas is recommended. For each PCR reaction with a unique amplification primer pair, a negative control should be used to ensure the absence of reagent contamination, often caused by the presence of exogenous nucleic acids. This control reaction is identical to the regular reaction, except that no template is present. Agarose gel electrophoresis is used to verify the amplicons and ensure that they are of expected size. In instances where multiple bands result from the PCR, products can be resolved on an agarose gel, and the fragment of expected size excised; these amplicons can then be sequenced to confirm their identity. It is our experience that, when a primer pair is well chosen, multiple bands seldom result from the PCR reaction. Printing the Arrays Once amplicons have been produced for all genes of interest as well as housekeeping genes, arrays can be printed. To avoid positional bias, arrays should be printed in a scattered fashion, with several repeats of the same amplicon located in different regions of the chip. It is important to avoid positional bias, as uneven distribution of charges on the membrane can result in regions of increased background. A typical microarray manufactured in this fashion carries arrayed triplicates or quadruplets of amplicons from selected genes, positioned on the array among many control spots. The rationale for triplicate printing is to provide three data points for statistical analysis of significance; ideally, the three could be expanded to four or five repeats, to yield more data points for statistical analysis. This approach of scattered array printing requires considerable careful analytical software design, to enable tracking of amplicon repeats across the platform; however, it approaches an ultimate solution to resolving positional bias. Although the spots of microarrays printed onto membranes affixed to glass slides are usually colorless, it is possible to monitor quality to detect gross errors in printing, such as missing, smeared, or non-uniform spots; immediately before a batch of microarrays are printed, a colored dye can be used to print a test array of dots onto a membrane. Microscopic visual inspection of the spots enables any necessary adjustments to be made to the robot before sample printing begins. While large batches of chips are being printed, quality can be monitored by inserting poly-L-lysine-coated glass slides among the membrane platforms. Unlike membranes, the clear surface of glass slides permits the researcher to see printed spots by breathing on the slide and viewing it through a transmitted light source. Once the probes are printed on the membranes, they are cross-linked to the microarray to permit better attachment of the nucleic acids to the substratum; probes are denatured by boiling the membranes before hybridization. Target Labeling Donor nucleic acids can be labeled by adding a labeled base to the RT reaction used to generate target cDNA. Whereas most commercial arrays use fluorescence-conjugated or radiolabeled bases, we developed our microarrays based on a nonradioactive colorimetric method. The bulkiness of some of the fluorescent tags, fluorescent quenching over time, and the need of specialized scanners to read fluorescent signals dissuaded us from using the fluorescent approach (26). Similarly, we wanted to avoid using radioactive labels for safety reasons, because they often give saturated signals, and to save the time required for exposure of the labeled arrays to x-ray film. Using commercially available digoxigenin (DIG)-labeled dUTP (Roche, Palo Alto, CA) and alkaline phosphatase (AP)-conjugated anti-DIG antibody (27), we have developed a new application for DIG in microarrays (28). In our method, the cDNA to each donor RNA is synthesized with a DIG-labeled base. Following hybridization of the DIG-labeled target with the probes, positive reactions are revealed by incubating with anti-DIG antibody conjugated to AP, and subsequent staining with Nitro-blue-tetrazolium/5-Bromo-4-chloro-3-indolyl phosphate (NBT/BCIP, Roche) to detect AP (29). Taking advantage of the fact that two complementary nucleotide strands can hybridize with each other, we generate microarray results by quantifying the signal obtained from the labeled targets bound to the immobilized probes. Thus the positive loci are visible as bluish spots, easily identified as round deposits for each positive locus. The final detection is revealed as a matrix of many round dots of varying intensity of staining. Microarray Inventory Extensive records should be maintained on the microarray printing process, including logistical parameters of the physical status of the microarrayer (e.g., relative humidity of the chamber during printing, or how often the printing pins are cleaned during the run), the specifics of printing, including the printing sequence, the identity of each amplicon-containing spot, and all other applicable data or comments. To ensure that identical records are made for each printing, we record all data for our microarrays onto a standardized form that we have developed over the past several years. We are also developing a two-dimensional bar-coding system to keep detailed track of our inventory of microarrays. Image Acquisition and Data Processing The work just described supports the process of data generation; the following sections describe the mining of the resulting data. Array Normalization and Background Subtraction As our arrays are based on a colorimetric detection method, a high-resolution scanner is used to scan them into digital images. Before a normal office scanner is used, it is important to ensure that it digitizes accurately without transforming the image (26). If the image is transformed by the scanner, mathematical correction transformation should be applied to the result. Following acquisition, the digitized images can be normalized and subtracted as desired. We have developed a software program, GeneAnalyzer, which accomplishes background subtraction, array normalization, and quantification. When colorimetric microarrays are analyzed, several types of background must be considered; for instance, regional background subtraction is useful when the array shows differential intra-array background expression, whereas global background subtraction is suitable when the background value is constant within arrays but variable between arrays, as a result of experimental conditions. For interarray comparison to be supported, arrays may be normalized by several methods, including reference to housekeeping gene levels and median chip values. However, investigators should think carefully about the effects of performing such background standardization or normalization before they start analyzing their results; they should especially consider the effect of background subtraction on diminishing the signal of low-expression genes. Software for Microarray Data Acquisition In general, image acquisition and data analysis include the following processes: (a) image grabbing and digitizing; (b) image processing; and (c) data mining, including a qualitative and quantitative analysis of all digitized images, and a statistical analysis of data. We developed our software with a user-friendly interface and a limited number of preset functions, to enable researchers to analyze their own data. The main features of our program are as follows. First, we provide users a personal identification number, which allows optimal security of their data and access to the interactive functions of our web server facility. Second, users can upload their electronic images from remote sites over the Web. Third, our system processes the users' initial data to enhance the image profile, through standard computer software such as MatLab. Fourth, our system supports the users' data archiving and database organization for the next stage of data analysis. Statistical Analysis and Data Mining Although it is necessary to use statistical methodologies to analyze voluminous microarray data, it is equally important to generate adequate microarray-derived data to support statistical testing. Thus it is recommended for most studies to use at least three mice (or individuals) chosen at random from each test group (e.g., young vs old mice). RNA extracted from each donor tissue source (e.g., lung) is then subjected to three separate RT reactions, and each of these nine independently obtained cDNAs is used as targets for three different microarrays. Because on our microarray each amplicon is spotted in triplicate, we have a total of 81 (34) spots, or data points, for each gene being analyzed (RNA from three mice multiplied by three RT reactions multiplied by three microarrays multiplied by three spots on each slide). The use of at least triplicates for each step of microarray fabrication enables us to obtain statistically meaningful data. Without this replication, simple variations in the efficiency of the RT reaction, interanimal variation, or misprinting of a spot could all result in falsely perceived changes in gene expression. Because the statistical analysis of microarrays presents a challenge to many biologists, it is recommended that a statistician be consulted as necessary. Statistical consultants can be extremely helpful, not only at the final stage of data analysis, but also at the initial experimental design step; for example, they may provide answers to fundamental questions, such as how many animals are needed to establish a statistically significant data analysis, or whether or not one may pool RNA samples. Once microarray data are processed through statistical analysis, data entry points deemed of true “significant” value, that is, gene expression changes as effects of an experimental physiological change, should be subjected to the next level of data analysis, now popularly termed the data mining process (30). Many established methods have been popularized among microarray users, including principal component analysis (31)(32)(33), hierarchical clustering (34)(35)(36)(37), multidimensional scaling (38)(39)(40), and self-organizing maps (41). In general, the selection of any of these methods is dependent on the individual investigator's preference and expertise. For example, GeneSpring software, sold by Silicon Genetics, Inc. (San Carlos, CA), and Significance Analysis for Microarrays from Stanford University (42), are preferred for many gene screening data mining tasks because they can analyze data generated by several different microarray platforms. These data mining software packages enable researchers to display their data in forms suitable for publication, easily conveying the essence of the results. Following data mining, microarray data that seem to be significant should be validated by using one of two popular methods: Northern blotting or quantitative RT-PCR. In general, it is advisable that microarray data be validated by the selection of four or five randomly designated genes from each of three categories: those showing high, intermediate, and low levels of significant difference. Because we use amplicons to generate our probes, we can easily validate our results, using the same primers used to generate our amplicons by quantitative or semi-quantitative RT-PCR. During data analysis, special consideration must be given to low-expression genes, which generally exhibit the greatest variance in expression levels. On any given microarray, these genes show very weak intensities, and in some cases they are barely visible above the background value. Here, standard global normalization and thresholding are not practical, because the signals are so weak. Often we find that global thresholding is too crude, allowing in one case the gene expression to be quantified as a gain, and in another case allowing the same gene expression to be quantified as a loss. One possible solution for this problem is to use “segmental thresholding,” localized thresholding for each individual weak spot. Then the local background level is calculated against the global background level to obtain confidence level indices. The actual gene expression level for these low-abundance genes is then the “minithresholding level” divided by the confidence level. We realize that this is not a perfect solution; often we have to disregard these data points altogether. Conclusions The notion that the bioinformatics of microarray studies is still in its infancy pertains not only to studies in the aging area, but also to many other biological systems as well. The entire field of microarrays is experiencing volatile changes in methodological approaches, technological design, and bioinformatics development for data interpretation. Part of the growing pains in the use of microarray technology is the constant need for new cutting-edge technology and the reevaluation of methodology. Therefore, for any designer microarray projects, it is necessary to be vigilant for any new methodology and technology developments, to improve the design and fabrication process as well as the bioinformatic aspects of gene screening tasks. As with all technologic advances, the microarray approach is not an end in itself; it is just a beginning. Obviously, one wants to know whether the genes identified as significantly changed at the RNA level are truly manifested at the protein level. For this purpose, the recent explosion of proteomic technology is certainly a testimony to the need for follow-up to microarray data. Ultimately, gene expression microarray studies have to be followed with experiments to examine protein changes, thus permitting a comprehensive examination of gene expression changes from RNA to protein levels. This work was supported by Grant R01 AG07444 from the National Institute on Aging of the National Institutes of Health, and from the Defense Advanced Research Project Agency of the Department of Defense, to E. Wang. We express our sincere gratitude to Ms. Sherry Chen, Ms. Angel Wang, Mr. Keith Liang, Dr. Yih-Jing Tang, Dr. Nagathihalli Nagaraj, Dr. Bo Yu, and Ms. Jane Williams for their excellent technical assistance, and to Mr. Alan N. Bloch for proofreading this manuscript. 1 Hayflick L, 2000. The future of ageing. Nature408:267-269. 2 Perls T, 2002. Genetic and environmental influences on exceptional longevity and the AGE nomogram. Ann NY Acad Sci959:1-13. 3 International Human Genome Sequencing Consortium2001. Initial sequencing and analysis of the human genome. Nature409:860-921. 4 Schena M, Shalon D, Davis RW, Brown PO, 1995. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science270:467-470. 5 Beier M, Hoheisel JD, 1999. Versatile derivatisation of solid support media for covalent bonding on DNA-microchip. Nucleic Acids Res27:1970-1977. 6 Stillman BA, Tonkinson JL, 2001. Expression microarray hybridization kinetics depend on length of the immobilized DNA but are independent of immobilization substrate. Anal Biochem295:149-157. 7 Bertucci F, Bernard K, Loriod B, et al. 1999. Sensitivity issues in DNA array-based expression measurements and performance of nylon microarrays for small samples. Hum Mol Gen8:1715-1722. 8 Stillman BA, Tonkinson JL, 2000. FAST slides: a novel surface for microarrays. Biotechniques29:630-635. 9 Afanassiev V, Hanemann V, Wolf S, 2000. Preparation of DNA and protein microarrays on glass slides coated with an agarose film. Nucleic Acids Res28:e66 10 Linftood K, Lijedahl U, Raitio M, Syvänen A-C, 2001. Minisequencing on oligonucletide microarrays: comparison of immobilization chemistries. Nucleic Acids Res29:e69 11 Bordoni R, Consolandi C, Castiglioni B, et al. 2002. Investigation of the multiple anchors approach in oligonucleotide microarray preparation using linear and stem-loop structured probes. Nucleic Acids Res30:e34 12 Podyminogin MA, Lukhtanow EA, Reed MW, 2001. Attachment of benzaldehyde-modified oligodeoxynucleotide probes to semicarbazide-coated glass. Nucleic Acids Res29:5090-5098. 13 Maskos U, Southern EM, 1992. Oligonucleotide hybridizations on glass supports: a novel linker for oligonucleotide synthesis and hybridization properties of oligonucleotides synthesized in situ. Nucleic Acids Res.20:1679-1684. 14 Beier M, Hoheisel JD, 2000. Production by quantitative photolithographic synthesis of individually quality checked DNA microarrays. Nucleic Acids Res28:e11 15 Beier M, Hoheisel JD, 2002. Analysis of DNA-microarrays produced by inverse in situ oligonucleotide synthesis. J Biotechnol94:15-22. 16 Fodor SPA, Read JL, Pirrung MC, Stryer L, Lu AT, Solas D, 1991. Light-directed spatially addressable parallel chemical synthesis. Science251:767-773. 17 Singh-Gasson S, Green RD, Yue Y, et al. 1999. Maskless fabrication of light-directed oligonucleotide microarrays using a digital micromirror array. Nat Biotechnol17:974-978. 18 LeProust E, Zhang H, Yu P, Zhou X, Gao X, 2001. Characterization of oligonucleotide synthesis on glass plates. Nucleic Acids Res29:2171-2180. 19 O'Donnell-Maloney MJ, Smith CL, Cantor CR, 1996. The development of microfabricated arrays for DNA sequencing and analysis. Trends Biotechnol14:401-407. 20 Hughes TR, Mao M, Jones AR, et al. 2001. Expression profiling using microarrays fabricated by an ink-jet oligonucleotide synthesizer. Nat Biotechnol19:342-347. 21 Ramakrishnan R, Dorris D, Lublinsky A, et al. 2002. An assessment of Motorola CodeLink microarray performance for gene expression profiling applications. Nucleic Acids Res30:e30 22 Bowtel DDL, 1999. Options available—from start to finish—for obtaining expression data by microarray. Nat Genet21:25-32. 23 Knight J, 2001. When the chips are down. Nature410:860-861. 24 Selvey S, Thompson EW, Matthaei K, Lea RA, Irving MG, Griffiths LR, 2001. Beta-actin—an unsuitable internal control for RT-PCR. Mol Cell Probes15:307-311. 25 Khalyfa A, Bourbeau D, Chen E, et al. 2001. Characterization of elongation factor-1A (eEF1A-1) and eEF1A-2/S1 protein expression in normal and wasted mice. J Biol Chem276:22,915-22,922. 26 Ramdas L, Coombes DR, Baggerly K, et al. 2001. Sources of nonlinearity in cDNA microarray expression measurements. Genome Biol2: (11) 0047 27 Hotke JH, Ankenbauer W, Muhlegger K, et al. 1995. The digoxigenin (DIG) system for non-radioactive labeling and detection of nucleic acids—an overview. Cell Mol Biol.41:883-905. 28 Semov A, Marcotte R, Semova N, Ye X, Wang E, 2002. Microarray analysis of E-box binding-related gene expression in young and replicatively senescent human fibroblasts. Anal Biochem.302:38-51. 29 Thompson D, Larson G, 1992. Chromogenic phosphatase substrate producing a blue-colored precipitate at the site of enzymatic activity. Biotechniques12:656 30 Miller RA, Galecki A, Shmookler-Reis RJ, 2001. Interpretation, design, and analysis of gene array expression experiments. J Gerontol Biol Sci56A:B52-B57. 31 Hilsenbeck S, Friedrichs W, Schiff R, et al. 1999. Statistical analysis of array expression data as applied to the problem of tamoxifen resistance. J Natl Cancer Inst.91:453-459. 32 Crescenzi M, Giuliani A, 2001. The main biological determinants of tumor line taxonomy eludicated by a principal component analysis of microarray data. FEBS Lett.19:114-118. 33 Fellenberg K, Hauser NC, Broors B, Neutzner A, Hoheisel JD, Vingron M, 2001. Correspondence analysis applied to microarray data. Proc Natl Acad Sci USA.98:10,781-10,786. 34 Yan PS, Chen CM, Shi H, Rahmaatpanah F, Wei SH, Caldwell CW, 2001. Dissecting complex epigenetic alterations in breast cancer among CpG island microarrays. Cancer Res23:8375-8380. 35 Hoffman KF, McCarty TC, Segal OH, et al. 2001. Disease fingerprinting with cDNA microarrays reveals distinct gene expression profiles in lethal type 1 and type 2 cytokine-mediated inflammatory reactions. FASEB J.15:2545-2547. 36 Huang J, Qi R, Quackenbush J, Dauway E, Lazaridis E, Yeatman T, 2001. Effects of ischemia on gene expression. J Surg Res99:222-227. 37 Dysvik B, Jonassen I, 2001. J-Express: exploring gene expression data using Java. Bioinformatics17:369-370. 38 Khan J, Simon R, Bittner M, et al. 1998. Gene expression profiling of alveolar rhabdomyosarcoma with cDNA microarrays. Cancer Res58:5009-5013. 39 Hess KR, Fuller GN, Rhee CH, Zhang W, 2001. Statistical pattern analysis of gene expression profiles for glioblastoma tissues and cell lines. Int J Mol8:183-188. 40 Helgason A, Hickey E, Goodacre S, et al. 2001. mtDNA and the islands of the North Atlantic: estimating the proportions of Norse and Gaelic ancestry. Am J Hum Genet68:723-737. 41 Tamayo P, Slonim D, Mesirov J, et al. 1999. Interpreting patterns of gene expression with self-organizing maps: methods and application to hematopoietic differentiation. Proc Natl Acad Sci USA.96:2907-2912. 42 Tusher VG, Tibshirani R, Chu G, 2001. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA98:5116-5121.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B4A738BFF1E9CD2E041D19AFD5E6A3AD39D01D9.txt b/test/dataset/in/resources/corpus/Clean_0B4A738BFF1E9CD2E041D19AFD5E6A3AD39D01D9.txt
deleted file mode 100644
index fd60b00..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B4A738BFF1E9CD2E041D19AFD5E6A3AD39D01D9.txt
+++ /dev/null
@@ -1 +0,0 @@
-]>HEARES3104S0378-5955(98)00140-310.1016/S0378-5955(98)00140-3Elsevier Science B.V.Fig. 1An example of click waveform recorded from animal's concha in a sound-attenuating chamber.Fig. 2The responses of a PL/Type I/III AVCN unit to clicks and click pairs. Upper right panel shows the PSTH of the response of this unit to tone bursts at characteristic frequency (CF). The coefficients of variation (CV) for this unit were 0.62 and 0.57 in the 2–14.9 and 20–39.9-ms time windows (called CV2–14.9 and CV20–39.9), respectively. Responses to single click and click pairs at interclick intervals (ICI) between 1 and 16 ms at 65 and 85 dB SPL are shown in the two left columns. Lower right panel shows the second click response as a function of ICI at 45, 65 and 85 dB SPL and represents the percentage of recovery of second-click responses.Fig. 3The responses of a Chop-S/Type III AVCN unit to clicks and click pairs. The format of this figure is the same as Fig. 2. CV2–14.9=0.1; CV20–39.9=0.14; Ronset/Rsust.=1.0.Fig. 4The responses of a Chop-S/Type III AVCN unit to clicks and click pairs. The format of this figure is the same as Fig. 2 except that click levels of 105 dB SPL are included. CV2–14.9=0.27; CV20–39.9=0.23; Ronset/Rsust.=1.7.Fig. 5The responses of an On-C/Type I/III AVCN unit to clicks and click pairs. The format of this figure is the same as Fig. 2. CV2–14.9=0.07; there were no spikes in the 20–39.9-ms window to calculate the CV and Ronset/Rsust..Fig. 6The responses of a Chop-T AVCN unit to clicks and click pairs. Upper right panel shows the PSTH of the response of this unit to tone bursts at CF. The format of this figure is the same as Fig. 2. CV2–14.9=0.34; CV20–39.9=0.61; Ronset/Rsust.=1.3. This unit was classified as Chop-T type based the CV-vs.-time pattern even though chopping was not visible in the PSTH.Fig. 7The responses of an On-L/Type I/III AVCN unit to clicks and click pairs. The format of this figure is the same as Fig. 2. CV2–14.9=0.14; there were insufficient spikes in the 20–39.9-ms window to calculate the CV; Ronset/Rsust.=4.5.Fig. 8The second-click recovery function for primary-like (left column), chopper (middle column) and onset (right column) AVCN units at the four stimulus levels (45, 65, 85 and 105 dB SPL). The thick line in each panel represents the mean second-click response recovery function. The number of units represented, N, is shown in each panel. Only those units with data available for three or more ICIs are included.Fig. 9Mean second-click response±standard error of means (S.E.M.) vs. ICI for the three main PSTH types (primary-like, chopper and onset; panels a–c) and the three AVCN EIA types (Type I, Type I/III and Type III; panels d–f) at four stimulus levels (45, 65, 85, and 105 dB SPL). Mean recovery function for Type I units at 45 and 105 dB SPL are not available due to the small sample size.Fig. 10Comparison of SR distribution of AN fiber and AVCN unit. The number of units represented, N, is shown in each panel. AN data are from Kim et al. (1991).Fig. 11Mean second-click recovery functions (±S.E.M.) for two populations of AVCN neurons at four stimulus levels. AVCN units are divided into a low-SR (SR<20 spikes/s) group and a high-SR (SR>20 spikes/s) group.Fig. 12Comparison of mean second-click recovery functions among low-SR and high-SR AN fibers and AVCN neurons at four stimulus levels (45, 65, 85 and 105 dB SPL). Data from AN fibers at 105 dB SPL are not available.Fig. 13Mean spike counts (a), absolute number of spikes (b) and mean response duration (c) in response to single clicks as a function of stimulus level among low- and high-SR AN fibers and AVCN units at four stimulus levels (45, 65, 85 and 105 dB SPL).Table 1Number of AVCN units in PSTH and EI-area classification schemesPSTH TypeEI-area typeII/IIIIIIII or IVUnclassifiedTotalPercentPL6918174134Chop02223395747On073141512Unusual0410387Total64245523121Percent53537419Responses of anteroventral cochlear nucleus neurons of the unanesthetized decerebrate cat to click pairs as simulated echoesKParhamH.-BZhaoYYeD.OKim*Division of Otolaryngology, Surgical Research Center, Department of Surgery, Neuroscience Program, The University of Connecticut Health Center, Farmington, CT 06030-1110, USA*Corresponding author. Tel.: +1 (860) 679-3690; Fax: +1 (860) 679-2451; E-mail: kim@neuron.uchc.eduAbstractTo elucidate the contribution of the anteroventral cochlear nucleus (AVCN) to `echo' processing, this study documents the responses of AVCN neurons to simulated echoes and compares them to those of auditory nerve (AN) fibers. Single unit discharges were recorded from 121 units in the AVCN of 21 unanesthetized decerebrate cats in response to click pairs with inter-click intervals ranging from 1 to 32 ms between 45 and 105 dB SPL re 20 μPa. Units were classified according to the post-stimulus time histogram (PSTH) and excitatory-inhibitory response area (EI-area) schemes. Based on their spontaneous rates (SR), units were subdivided into low- (<20 spikes/s) and high- (>20 spikes/s) SR groups. A majority of the units exhibited second-click responses whose recovery time courses were similar to those of AN fibers. These units included primary-like, chopper and onset units in the PSTH scheme and Types I, I/III and III units in the EI-area scheme. A minority of the units exhibited responses that were distinct from those of AN fibers, in that they had second-click response recovery times that were either markedly reduced or prolonged. This group of units included those with primary-like, chopper and onset PSTHs and Type I/III and III EI-areas. No significant difference was found in the second-click response among various PSTH or EI-area types. High-SR AVCN units exhibited a decrease in the second-click response with increasing level. In contrast, low-SR AVCN units showed little level-dependent change in the second-click responses. This SR-based difference was similar to that previously found among AN fibers. The present results suggest that, although a majority of AVCN units exhibit similar time courses of second-click response recovery to those of AN fibers, there do exist mechanisms in the cochlear nucleus that can substantially alter this representation. Furthermore, the difference between the second-click response recovery functions of low- and high-SR AVCN units and the consistency of this finding between AVCN and AN suggest that SR represents an important dimension for signal representation in the AVCN neurons.KeywordsAnteroventral cochlear nucleusEcho processingClick pairSpontaneous rate1IntroductionIn a reverberant environment, such as a lecture hall, an acoustic signal originating from one location is typically followed by its echoes that arrive at the ears from multiple directions. Nevertheless, a listener can correctly perceive a single auditory image and its correct direction. This phenomenon of the first wavefront taking precedence over later-arriving wavefronts has been referred to as the `law of the first wavefront' or `precedence effect' (Wallach et al., 1949; Haas, 1951; Von Bekesy, 1960; Blauert, 1997). To gain an understanding of the mechanisms involved in auditory processing of the echoes, it would be useful to study this phenomenon across multiple structures within the ascending auditory pathway.We recently reported on echo processing features of the auditory nerve (AN) fibers in the unanesthetized decerebrate cat (Parham et al., 1996). A main finding was that the recovery of the response to the trailing click of a click pair among AN fibers not only depended on stimulus parameters (i.e., interclick interval and click level) but also on fiber spontaneous rate (SR). High-SR fibers exhibited longer recovery times than their low-SR counterparts.The cochlear nucleus (CN) serves as the obligatory termination site of AN fibers. Anatomically, it can be subdivided into three main divisions: the anterior and posterior ventral cochlear nuclei (AVCN and PVCN, respectively) and the dorsal cochlear nucleus (DCN). Each subdivision is composed of populations of neurons that, by virtue of different combinations of intrinsic membrane properties and synaptic inputs, can express diverse responses to sound. As a result, the CN consists of an array of neurons whose responses fall on a continuum from a faithful preservation to a substantial transformation of the incoming afferent signals. These responses are conveyed to higher auditory centers along parallel pathways arising from the CN subdivisions.Based on their responses to sounds, CN neurons have been characterized according to two schemes: temporal discharge patterns reflected in the post-stimulus time histogram (PSTH) scheme (e.g., Pfeiffer, 1966) and the expressions of excitatory and inhibitory regions and responsiveness to broad-band noise reflected in the excitatory-inhibitory area (EI-area) scheme (e.g., Young, 1984). The response types within the two schemes appear to be distributed differentially within CN subdivisions and are correlated with different morphological classes of neurons in the CN. In the present context, the anatomical and physiological diversity of CN neurons gives rise to three important questions: (i) Is echo representation among CN neurons different from that among AN fibers? (ii) Do various physiological classes of CN neurons differ from one another with respect to echo representation? (iii) Are SR-based differences in echo processing of AN fibers also present in the CN?Recently, Wickesberg (1996)reported that ventral cochlear nucleus (VCN) units showed rapid recovery of the second-click response in the anesthetized chinchilla. Units characterized by primary-like and chopper PSTHs showed 50% recovery by 2 ms interclick interval and a nearly complete recovery by 4 ms. Wickesberg also reported some variability in the responses of VCN neurons, in that a subset exhibited little to no reduction or enhanced response to the second click.In this study we attempt to elucidate the contribution of AVCN to echo processing. We systematically investigated the behaviors of AVCN neurons regarding their response recovery functions under a click-pair stimulus paradigm in decerebrate unanesthetized cats. In this assessment of AVCN behavior in response to click pairs, neurons were characterized according to both PSTH and EI-area schemes and SR-based differences were investigated. Early results of this study were presented at a meeting of the Association for Research in Otolaryngology (Kim et al., 1992).2MethodsThe care and use of animals reported on in this study were approved by the University of Connecticut Health Center Animal Care Committee (protocol title: `Physiology of the Cochlear Nucleus'; protocol number: 91-014-97-2). The details of Section 2are described elsewhere (Parham et al., 1996). Briefly, we used unanesthetized decerebrate cats. We opened the left posterior fossa, aspirated a portion of the cerebellum, and exposed the CN. A micropipette was positioned over the AVCN which was visualized through an operating microscope. Single unit spike discharges were recorded using micropipettes filled with 3 M sodium chloride or potassium acetate that had resistances of 20–50 MΩ. In some penetrations the recording location was marked by iontophoretic injection of markers, 10% biotinylated dextran amine or 10% horseradish peroxidase in 0.5 M potassium acetate in Tris buffer. As the electrode was advanced, we applied a tonal search stimulus whose frequency was swept with a triangular modulating waveform over 0–50 kHz with a modulation repetition period of 2 or 4 s. The level of the swept tone was in the range of 40–60 dB SPL (re 20 μPa). For each AVCN unit encountered, we recorded the following: (i) spontaneous discharge rate based on a 10- or 20-s sample without acoustic stimulation; (ii) response area curve representing discharges induced by 200-ms tone bursts (two repetitions) at various frequencies spaced at 50 frequencies per decade on a logarithmic scale and at a constant dB SPL, typically 10–30 dB SPL; from the response area we estimated the characteristic frequency (CF) (the frequency to which a unit is most sensitive) of the unit; (iii) responses to 60-dB SPL CF tone bursts (50 ms duration, 500 ms repetition interval, 5 ms rise/fall time, 60 repetitions) from which a post-stimulus time histogram was constructed; (iv) responses to 200-ms noise bursts (three repetitions, 5 ms rise/fall; 1000 ms repetition interval) in 5-dB steps between 20–70 dB SPL (measured as rms level with A-weighting); and (v) responses to single rarefaction clicks and equilevel click pairs with inter-click intervals (ICI) ranging from 1 to 32 ms; repetition period of 500 ms, 40 repetitions, at 45, 65, 85 and 105 dB SPL peak. The rarefaction clicks were produced by applying 30-μs pulses to the earphone. The acoustic waveform of a single click recorded in the concha of the cat is shown in Fig. 1.Event times for spike discharges and stimulus markers were stored with 20 μs resolution in the data files such that subsequent `off-line' analyses would have access to full information about spike discharge and stimulus events. Single units were classified according to the PSTH and EI-area schemes. Discharges of all units were examined for regularity based on the behavior of the coefficent of variation (CV) of spike discharges as a function of time (e.g., Young et al., 1988; Parham and Kim, 1992). Chopper units were subdivided into three subtypes: Chop-S (CV<0.35); Chop-T (CV increased from <0.35 to >0.35 within the first 15 ms of the response) and Chop-O (CV>0.35). Based on PSTH of the responses to pure tone at CF, onset units were identified based on a quantitative criterion that the ratio of the discharge rate over the first 5 ms of the response (Ronset) to the last 40 ms of the response (Rsust.) be greater than 2.5.The procedure for analysis of single unit responses to click pairs is described in detail elsewhere (Parham et al., 1996). Briefly, we examined the response of each AVCN unit to single clicks and click pairs in the form of PSTHs with binwidths of 0.08 ms. For graphical illustrations binwidths of 0.2 or 0.4 ms were used. To quantify the results, we defined the time window that included all of the evoked spike discharges to the single click as W1. W1 was defined for each unit and at each click level because response latency and duration varied among units and as a function of stimulus level. W1 served as the reference time window for subsequent computations of the responses to a click pair. The time window of the response to the second click of a pair (W2) was defined as a window identical to W1 except for a delay by the amount of ICI. The second-click response (SCR) was expressed as percentage of the response to the first click. At short ICIs, the quantification of the SCR took into account the tail end of W1 which overlaps with W2 (for details see Parham et al., 1996).At each ICI the level dependent differences between SCRs of various unit types were statistically evaluated using analyses of variance (ANOVAs). Significant interactions were followed up by lower order ANOVAs and significant main effects were further examined with Tukey HSD tests of mean comparisons.3ResultsThis study is based on the responses of 121 AVCN units recorded in 21 cats. Table 1 shows the number of AVCN units subdivided according to the PSTH and EI-area schemes. In the PSTH scheme, the majority of the units (93%) exhibited either primary-like (PL, 34%), chopper (Chop, 47%) or onset (On, 12%) PSTHs. The remaining units (7%) had unusual or non-descript PSTH appearances. The PL group (N=41) included primary-like (98%) and primary-like-with-notch (2%) PSTH types. The Chop group (N=57) was further subdivided into sustained (Chop-S, 33%), transient (Chop-T, 53%) and other (Chop-O, 14%) chopper subtypes. The On group (N=15) consisted of units exhibiting onset without sustained discharges (On-I, 13%), onset chopping (On-C, 47%), onset with low level of sustained activity (On-L, 33%) and onset followed by inhibition of spontaneous activity (On-inh, 7%). In the EI-area scheme, the majority of the units (77%) were characterized as Type I (5%), I/III (35%) or III (37%) with a minority falling into other categories (Type II, 3%; Type IV, 1%). For the remaining 19% of the units, response to broadband noise was unavailable and these units were not classified in the EI-area scheme. Among the present PL units, 18 out of 41 units were EI-area Type III. This is a much greater tendency than a previous report (Shofner and Young, 1985). The reason for this difference is not clear.Fig. 2 shows the responses of a low-SR PL/Type I/III unit to single clicks and click pairs. Responses of this unit to single clicks at 65 and 85 dB SPL peak are shown in the top PSTHs of the two left-most columns. The response was characterized by a large peak followed by a gradual decay of the response over 5–6 ms. Responses to click pairs at short ICIs (1–3 ms), shown in the remaining rows of the two columns, tended to be slightly more prolonged than the response to single clicks. At larger ICIs the response to the second click was evident. As the ICI increased, the second-click response gradually increased in strength such that by 16-ms ICI, the second-click response was similar to the first-click response. The lower right panel provides the magnitude of the second-click response as a function of ICI at 45, 65 and 85 dB SPL. The second-click response gradually recovered from near 25% at 1-ms ICI to above 50% by 4-ms ICI and to near a complete recovery (80–100%) by 16-ms ICI. This PL/Type I/III unit exhibited a slower recovery of the second-click response as click level was increased from 45 to 85 dB SPL.The behavior of the sample PL/Type I/III unit shown in Fig. 2 was typical of a majority of the units recorded from the AVCN in this study (to be shown further below) and resembled the behavior of AN fibers recorded under similar conditions (Parham et al., 1996). However, a minority of the units exhibited responses to click pairs that were distinct from the typical behavior resembling AN fibers. The next five figures (Fig. 3Fig. 4Fig. 5Fig. 6Fig. 7) illustrate units characterized by sustained and transient responses to pure tones that exhibited click pair responses that were distinct from that of the AN fibers.The three examples shown in Figs. 3–5 exhibited faster recovery of the second-click responses than AN fibers. The responses of a Chop-S/Type III AVCN unit to single clicks and click pairs is shown in the two left-most panels of Fig. 3. This unit's responses preserved the representation of both clicks of a pair at low and moderate stimulus levels (45 and 65 dB SPL), even at ICIs as short as 1 ms. At 45 dB SPL, the PSTH of the response of this unit to single clicks consisted of a single narrow peak. At 1-ms ICI the response to the second click is distinct and nearly equal to that of the first click. At 85 dB SPL, the single click response consisted of two peaks, separated by an interval (1.8 ms) which is approximately equal to the interpeak interval of the PSTH of the response to the CF tone (upper right panel). The response to click pairs at 1-ms ICI was similar to that of the single-click response except that the second peak of the PSTH was narrower and taller (i.e., distributed across fewer bins). A distinct response to the second click was present at 2-ms ICI which gradually recovered as ICI was increased to 32 ms. The lower right panel shows the second-click response recovery functions of this unit. At 45 and 65 dB SPL the second-click response remained near 100% for all ICIs, but at 85 dB SPL the second-click response recovered from 0% at 1-ms ICI to 100% at 32 ms.The responses of another Chop-S/Type III AVCN unit are shown in Fig. 4. This unit exhibited a poor recovery of the second-click response at a low stimulus level for ICIs even up to 32-ms ICI, but strong second-click response at high stimulus level even at 1-ms ICI. At 45 dB SPL a weak response to the second click was present by 2-ms ICI. Unlike previous examples, however, second-click response was completely absent by 8-ms ICI and remained below 50% for ICIs up to 32 ms. In contrast, at 105 dB SPL the second-click response was present at 1-ms ICI in the form of a broad peak. By 2-ms ICI the second-click response was nearly the same as that of the first. The second-click recovery functions of this unit are shown in the lower right panel. At 45 dB SPL the second-click response remained below 50%, but at 105 dB SPL the second-click response was recovered to 85–100% across all ICIs examined. At intermediate levels of 65 and 85 dB SPL the second-click response exhibited more rapid recovery than that expected from AN-like recoveries.Fig. 5 shows the responses of an On-C/Type I/III unit. At 65 dB SPL the second-click response was strong (>75%) across all ICIs. At other click levels the recovery of the second-click response was more gradual, although somewhat faster than those exhibited by AN fibers. For example, at 85 dB SPL the response to the second click was greater than 75% at ICIs beyond 3 ms.The next two examples shown in Figs. 6 and 7 exhibited slower recovery of the second-click response recovery function than AN fibers. The responses of a Chop-T AVCN unit is shown in Fig. 6. The second-click response of this unit at 45 dB SPL exhibited a strong response at 1-ms ICI, but weak responses at 3–8-ms ICIs. The second-click response at 65 dB SPL was similar to that of 45 dB SPL (lower right panel). At 85 dB SPL this unit exhibited a more gradual recovery (middle column and lower right panel).The responses of an On-L/Type I/III AVCN unit are shown in Fig. 7. This unit was characterized by a weak second-click response up to 32-ms ICI for all stimulus levels tested. Interestingly, the second-click response increased between 2- and 8-ms ICIs, but decreased for longer ICIs.Fig. 8 shows recovery functions for populations of primary-like (left column), chopper (middle column) and onset (right column) units at 45–105 dB SPL. Individual curves in Fig. 8 represent separate units, while the thick line in each panel represents the mean second-click response recovery function. In general, the majority of the units exhibited recovery functions qualitatively similar to those observed at the level of the AN. However, a small number of non-AN-like recovery functions were observed in all three populations. Examples of fast recovery to the second click can be seen in primary-like (panel d), chopper (panels b, e and k) and onset (panels c and f) units. Examples of weak response to the second click can also be seen in primary-like (panel g), chopper (panels b, e, h and k) and onset (panels f, i and l) units.The mean recovery functions of the three main AVCN PSTH types are shown in Fig. 9a–c. Primary-like units showed little change in the mean second-click response with the click level (Fig. 9a). Only at 16- and 32-ms ICIs were the second-click responses to 45 and 65 dB SPL significantly greater than those at 105 dB SPL stimuli (Tukey HSD, P<0.05). Chopper units exhibited an orderly weakening of the mean second-click response as click level was increased from 45 to 85 dB SPL for 1–3-ms ICIs (Fig. 9b). However, at 105 dB SPL the chopper units showed the strongest mean second-click responses for 1–3-ms ICIs. The mean second-click response of the onset units also decreased with increasing stimulus level (Fig. 9c). However, the level-dependent differences among chopper and onset units were not statistically significant (P>0.05).The mean second-click recovery functions of the three unit types were similar at 65 and 85 dB SPL. At 45 dB SPL the onset units exhibited more rapid recovery of the second-click to above 75% than the primary-like and chopper units (i.e., 3 vs. 4 ms, respectively). However, at 105 dB SPL the onset units exhibited slower recovery of the second-click response than the primary-like and chopper units (75% SCR: 7 vs. 6 ms, respectively). At 45 and 105 dB SPL the mean second-click recovery functions of the primary-like units were intermediate to those of the chopper and onset units. The differences between the three PSTH types were not statistically significant (P>0.05).The mean second-click recovery functions for the three EI-area types are shown in Fig. 9d–f. The mean second-click response of the Type I group increased as the stimulus level increased from 65 to 85 dB SPL (Fig. 9d). Because of the small sample size for Type I units mean second-click response recovery functions are not available at 45 and 105 dB SPL. The mean second-click response decreased from 45 to 85 dB SPL for Type I/III and III units (Fig. 9e,f). However, at 105 dB SPL Type I/III and III units showed an increase in the mean second-click responses relative to those at 85 dB SPL (Fig. 9e,f). The mean second-click response recovery functions of the three EIA types differed little when compared at each click level.Fig. 10 shows the SR distributions for AVCN units recorded in this study and a sample of AN fibers recorded in a previous study (Kim et al., 1991). The AVCN units were divided into low- (<20 spikes/s, N=76, 63%) and high-SR (>20 spikes/s, N=45, 37%) groups, as was previously done for AN fibers (Parham et al., 1996). The mean second-click response recovery functions of these two groups are shown in Fig. 11. Low-SR AVCN units showed little level-dependent changes in the second-click response recovery functions between 45 and 85 dB SPL. Interestingly, at 105 dB SPL the low-SR AVCN units exhibited stronger second-click responses than at lower stimulus levels. One-way ANOVAs at each ICI indicated a significant main effect of level only at the 2-ms ICI (F(3,124)=3.42, P=0.02). In contrast, the high-SR AVCN units exhibited a more noticeable level-dependent decrease in the second-click response between 45 and 105 dB SPL. This decrease was statistically significant at 3-, 4- and 6-ms ICIs (P<0.02). The mean recovery functions of the two AVCN SR groups are compared to one another and their AN counterparts in Fig. 12. At 45 and 65 dB SPL, the four populations exhibited similar second-click response recovery functions. One-way ANOVAs at 45 and 65 dB SPL showed no significant differences in mean second-click responses of the two SR groups. At 85 dB SPL, the high-SR AVCN and AN fibers exhibited weaker second-click response recoveries than their low-SR counterparts. Two-way ANOVAs at each ICI at 85 dB SPL revealed main effects of SR at all ICIs less than 32 ms (P<0.05) without any significant interaction with location (i.e., AN vs. CN).At 105 dB SPL differences in second-click response between the low- and high-SR AVCN units are much more evident, with the high-SR AVCN units having weaker second-click response for ICIs less than 32 ms. The AN data for 105 dB SPL are not available. ANOVAs indicated that the second-click responses of the high-SR AVCN units were significantly weaker than their low-SR counterparts at all ICIs less than 8 ms (P<0.05).Since the second-click response recovery functions presented above are a relative measure, they do not convey information about the absolute neural activities. The latter information can be inferred by combining the relative recovery functions with an absolute measure of neural activities for various stimulus conditions. For this purpose, we examined the number of spikes elicited by single clicks at various levels, as shown in Fig. 13. For all groups, the mean number of spikes elicited by single clicks increased as the stimulus level increased (Fig. 13a). However, this increase was substantially larger for the high-SR AVCN and AN units. Consequently, the low-SR AVCN units and AN fibers consistently exhibited fewer spikes per click than their high-SR counterparts across all levels. The duration of the single-click response is shown in Fig. 13c. Although the response duration increased as stimulus level increased for all four groups, the response durations of high-SR AVCN units and AN fibers were consistently longer than those of their low-SR counterparts.4DiscussionA majority of AVCN units recorded in this study exhibited second-click response recovery time courses that were similar to those of AN fibers. These units had primary-like, chopper and onset PSTHs and Types I, I/III and III EI-areas. A minority of AVCN units, however, exhibited second-click responses whose recovery time courses were markedly reduced or prolonged relative to those of AN fibers. The latter group of units had primary-like, chopper and onset PSTHs and Type I/III and III EI-areas. The mean second-click response recovery functions showed little difference among the three main PSTH and EI-area types. When AVCN units were subdivided into low- and high-SR groups, however, a clear difference in the second-click response recovery functions was found between the two SR groups.In this study, the responses of many AVCN neurons to single click stimuli were often prolonged. For example, multipeaked PSTHs of click responses were relatively common in low-CF (i.e., below 2 kHz) primary-like units. Similar to previous observations in the AN, the PSTH of the responses of these low-CF primary-like units to clicks had a multi-peaked appearance, with the interpeak interval being equal to the 1/CF interval. Some high-CF CN units also had PSTHs of the response to click stimuli that were multi-peaked (e.g., Figs. 3, 5 and 6). However, as previously noted by Kiang (1965), the inter-peak intervals in these high-CF units did not correspond to 1/CF. Kiang's report did not classify CN units according to the PSTH type. In the current study, multi-peaked PSTHs among high-CF units were associated with chopper and onset-chopper type units. The inter-peak intervals in the PSTHs of the response to the clicks in these units approximated those in the PSTHs of the responses to pure-tone bursts at their CFs.The level-dependent weakening of the second-click responses of high-SR AVCN units and their slower recovery relative to low-SR AVCN units are analogous to our earlier observations in the AN (Parham et al., 1996). Spontaneous activities of AVCN neurons appear to be mainly dependent on their AN inputs. Koerber et al. (1966)observed that, after cochlear destruction, spontaneous activity of AVCN neurons was abolished suggesting that SR of AVCN neurons mainly arises from their AN afferent inputs. Therefore, the differences between the two SR groups in the AVCN for the most part appear to be a reflection of their AN afferent inputs.The results presented in this report offer a possible reason for the difference in the dependence of the second-click response on the level between the two SR populations of the AN and AVCN, i.e., neuronal refractoriness. Gaumond et al. (1982)characterized refractoriness of AN fibers in terms of a conditional discharge probability conditioned on the time since the last spike, called tau, for a steady stimulus. The conditional discharge probability of an AN fiber typically was zero (i.e., completely refractory) for tau<0.75 ms, rapidly increased from zero to about 60% of a fully recovered value at tau=4 ms, and more gradually increased to the fully recovered value (i.e., zero refractoriness) between 4 and about 40 ms. Gaumond et al., 1982(Figs. 12 and 13) also observed that, when there were successive short (e.g., <10 ms) interspike intervals, the refractoriness of AN fibers was accumulated, i.e., the conditional discharge probability further decreased. The fact that the time courses (0 to about 40 ms) and general shapes of the two measures, i.e., second-click response vs. inter-click interval and refractoriness vs. time since the last spike, are quite comparable to each other, supports the possibility that neuronal refractoriness is the main mechanism underlying the reduction and a gradual recovery of the second-click response being described in the present report and in Parham et al. (1996). In Fig. 13a, it is seen that, as the stimulus level was increased from 45 to 105 dB SPL, the mean absolute number of spikes increased from about 1 to 3 spikes per click for high-SR AN and AVCN units whereas they changed little for low-SR AN and AVCN units. It is also seen in Fig. 13c that the multiple spikes at high stimulus levels occurred over a period of about 4–10 ms. This corresponds to interspike intervals of about 2–4 ms.In view of the Gaumond et al. study discussed above, the refractoriness of a high-SR unit should be substantially greater during a period following the clicks presented at high levels than that at low levels because additional spikes were discharged with short interspike intervals at high stimulus levels. This expected change in the amount of refractoriness of high-SR units with stimulus level is consistent with the decrease of second-click response observed in high-SR AVCN units (Fig. 11) and AN fibers (Parham et al., 1996, Fig. 7). Likewise, this reasoning as applied to low-SR units regarding their absolute number of spikes per click being nearly independent of stimulus level (Fig. 13a) predicts that their second-click response recovery functions should be nearly independent of stimulus level. Low-SR AVCN units (Fig. 11) and AN fibers (Parham et al., 1996, Fig. 7) indeed exhibited much less dependence of their second-click recovery functions on stimulus level than their high-SR counterparts, largely consistent with the prediction. There is one discrepancy between the refractoriness-based prediction and the observation in this regard. That is, the second-click response of low-SR AVCN units increased somewhat at 105 dB SPL compared with those at lower levels (Fig. 11, left panel) whereas, based on Fig. 13, one predicts that it should remain nearly constant with a slight decrease. The reason for this discrepancy is not clear.When one considers that the discharge behavior of an AVCN neuron is affected by neuronal refractoriness, two components of refractoriness can be considered: (i) refractoriness of AN fibers acting as an input to the AVCN neuron, and (ii) refractoriness of the AVCN neuron. The present observations (Fig. 12) indicate that the second-click response recovery functions are rather similar between the AN fibers and AVCN neurons. In interpreting the second-click response recovery functions in terms of neuronal refractoriness, this similarity implies that the combined total refractoriness observed in AVCN units is similar to that in AN fibers. To our knowledge, direct examinations of neuronal refractoriness (analogous to the Gaumond et al. study) for AVCN neurons have not been reported. We believe that such a study should be useful.The faster recovery of the second-click response of low-SR AN fibers and AVCN neurons than their high-SR counterparts contrasts with findings of forward masking studies using tone bursts at the level of the AN (Relkin and Pelli, 1987; Relkin and Doucet, 1991) and CN (Boettcher et al., 1990; Shore, 1995). Forward masking studies noted that low-SR units required longer times for recovery of responses to tone bursts than high-SR units, opposite of the present observation of the second-click response. Previously, we considered that possible reasons for the discrepancy might include longer signal duration (tone burst) and a large level difference between the masker and probe in the forward masking studies compared with click pairs of the present studies (Parham et al., 1996). Even when the masker and probe levels were nearly equal in the CN forward masking studies, low-SR neurons showed slower recovery than high-SR neurons (Boettcher et al., 1990; Shore, 1995). Therefore, the difference between masker and probe levels does not explain the discrepancy between click-pair and forward masking studies. Different mechanisms may underlie the two phenomena. Adaptation at the inner hair cell and its afferent synapse and suppression of cochlear responses mediated by the olivocochlear efferent system (Bonfils and Puel, 1987) may mainly underlie the forward masking phenomenon whereas neuronal refractoriness may mainly underlie the second-click response reduction phenomenon. We do not expect that the second-click response recovery function is mainly mediated by the efferent system because the click stimulus had a brief duration (Fig. 1) presented with a long (500 ms) repetition period, whereas efferent-mediated effect required a sufficiently long (greater than 35 ms) stimulus (Bonfils and Puel, 1987).The similarity of the recovery rates of low- and high-SR AVCN neurons to their AN counterparts supports the notion that SR may be an important axis in signal representation at the level of the CN. This conclusion is consistent with that of Shore (1995)in her study of forward masking in the VCN. The suggestion of an SR-based segregation of AN fiber projections to the CN, first made by anatomical studies (Leake and Snyder, 1989; Kawase and Liberman, 1992; Liberman, 1991), has received additional neurophysiological support. Recent recordings from the AVCN showed that neurons in the marginal shell were predominantly low SR, whereas the central core included high- and low-SR neurons (Ghoshal and Kim, 1996Ghoshal and Kim, 1997).Since a majority of the AVCN units display AN-like behavior with respect to second-click responses, these neurons do not modify the AN afferent inputs but act to relay AN signals to higher auditory centers. In a forward masking study of CN neurons, Boettcher et al. (1990)arrived at a similar conclusion. A minority of AVCN units, however, showed echo processing behaviors that were distinct from those of the AN fibers. These units had primary-like, chopper and onset PSTH types. Forward masking studies in the CN have also identified neurons whose recovery from short-term adaptation is distinct from those of AN fibers (Boettcher et al., 1990; Shore, 1995). In vivo intracellular recording and labeling experiments have shown that PSTH types tend to be associated with morphological cell types in the AVCN (Rhode et al., 1983; Rouiller and Ryugo, 1984; Smith and Rhode 1987, 1989; Ostapoff et al., 1994). For example, the primary-like responses are generally associated with bushy cells, whereas chopper responses are generally associated with multipolar cells. Since the unusual second-click recovery behaviors of some AVCN neurons were not associated with a distinct response type in the AVCN, the present observation suggests that these novel echo processing strategies may not be associated with a unique cell type in the AVCN.The unusual echo processing patterns of AVCN neurons consisted of neurons with either enhanced or suppressed representation of the echo relative to that of the AN fibers. The enhancement of the second-click response was observed at short ICIs (i.e., less than 4 ms) where individual AN fibers show typically weak responses (less than 50%; Parham et al., 1996). The present study is the first to demonstrate the existence of neurons with very fast second-click response recovery times in the AVCN. One possible way to extract adequate information regarding the temporal position of an echo, is for some AVCN neurons to sample and integrate the responses of many AN fibers. In other words, convergence of several AN fiber inputs onto an AVCN neuron could lead to summation of temporally coincident excitatory post-synaptic potentials thus eliciting an action potential. This is analogous to previously proposed mechanisms for certain VCN neurons (e.g., Smith and Rhode, 1986; Winter and Palmer, 1995; Palmer et al., 1996). Responses of spherical bushy cells in the AVCN are dominated by their AN fiber inputs which are believed to be form one or two calyx-type endings on one bushy cell (e.g., Sento and Ryugo, 1989). Therefore, the spherical bushy cells are a poor candidate for this type of integration. More likely candidates are multipolar cells receiving bouton-type endings from the AN which are distributed over their soma and proximal dendrites (Cant, 1981).The suppression of the second-click responses beyond that exhibited by the AN fibers is likely mediated by circuits releasing inhibitory neurotransmitters such as glycine and/or GABA in some CN neurons (Backoff et al., 1997). Suppression of the second-click response of AVCN neurons could arise partly from intranuclear and/or descending inhibitory projections to AVCN neurons. The intranuclear mechanisms that may underlie such suppression have been discussed in detail elsewhere (Wickesberg and Oertel, 1990; Wickesberg, 1996). Briefly, one hypothesized source is the tuberculoventral neurons of the DCN which provide glycinergic input to the AVCN. In vitro excitation of these neurons through stimulation of AN fibers which project to tuberculoventral neurons, as well as AVCN neurons, showed inhibitory postsynaptic potentials in AVCN neurons about 2 ms after AN induced excitatory postsynaptic potentials. Wickesberg (1996)provided further support for this circuit in a pharmacological perturbation study. Wickesberg showed that, following application of lidocaine in the DCN, responses of 19% of the VCN units (primary-like and chopper type) to the trailing click increased. Such inputs could account for the trough-like recovery functions observed in this study (e.g., Fig. 6). For 62% of the VCN units, however, Wickesberg found that the response to the first click decreased after lidocaine administration. The reason for the latter unexpected finding is not clear.In the AVCN, ICI at 50% recovery of the second-click response ranged from 1.8 ms at 45 dB SPL to 3 ms at 85 dB SPL without any substantial differences between the two AVCN SR groups (Fig. 12). This compares with 1.3–3 ms over the same stimulus level range in the AN (Parham et al., 1996). Thus, there was only a very slight increase in the average 50% recovery time in AVCN relative to the AN. The recovery functions of AVCN units are similar to those reported for units in the SOC by Fitzpatrick et al. (1995). Monaural units in the SOC exhibited second-click recovery functions that were analogous to those of AVCN neurons. The average 50% recovery time of monaural units in the SOC was ∼2 ms (Fitzpatrick et al., 1995). The fact that recovery functions in the SOC are comparable to those in the AVCN with the average 50% recovery times being similar in the two nuclei suggests that no significant modification of the echo processing occurs between these two structures. Substantial alterations apparently occur higher up at the level of the IC as the average 50% recovery time increases to ∼6–20 ms (Yin, 1994; Fitzpatrick et al., 1995). Additional suppression mechanisms are interpreted to be present in the cortex, since in the auditory cortex the average 50% recovery times ranged up to 300–500 ms (Fitzpatrick et al. 1997, 1998).Some of the auditory cortical neurons were observed to exhibit rapid recovery times of <2 ms (Fitzpatrick et al., 1997). Considering that the AVCN also contains neurons exhibiting comparably rapid recovery times (e.g., Figs. 3–5), the projections of such AVCN neurons could form the beginning of a channel that may convey information about the presence and nature of echoes. Such information may underlie the quality of the sound heard, i.e., its timbre (e.g., Blauert, 1997; Harris et al., 1963; Sayers and Toole, 1964). The projections from the majority of AVCN neurons, whose second-click responses recovery times are 2–10 ms, may represent the beginning of a separate channel that may underlie the perception of the sound location, thus mediating the precedence effect which has a time window of about 2–10 ms (Wallach et al., 1949). In addition, certain neurons in the AVCN (present study) and higher auditory centers (e.g., Fitzpatrick et al. 1995, 1997), whose second-click recovery times are in the range of about ten to a few hundred milliseconds, may mediate a time window (`echo window') where a lagging sound (i.e., echo) is heard but its perception is still affected by a leading sound (Fitzpatrick et al., 1998).Based on the findings of our studies in the AN and AVCN, it is important in future studies to keep in mind two important variables: SR and stimulus level. In both AN and AVCN, the 50% recovery time of the second-click response increased from 1.3 and 1.8 ms, respectively, to 3 ms between 45 and 85 dB SPL. At 105 dB SPL, however, the 50% recovery time for low-SR AVCN neurons was less than 1.5 ms, while that of high-SR AVCN neurons was ∼5 ms. Results for the AN at 105 dB SPL are not available for comparison. Given that both SR and stimulus level exert a substantial influence on echo processing, greater attention needs to be paid to these variables when making comparisons across structures.Regarding psychophysical studies of precedence effect which examined the influence of the overall stimulus level, we are aware of only one, Shinn-Cunningham et al. (1993). Using two stimulus levels (80 and 110 dB SPL 1-ms noise bursts), they found: (i) that the stimulus level had relatively small influence on the strength of the precedence effect; (ii) that the strength of the precedence effect was little affected by the stimulus level for 1-ms ICI (or `lag' of Shinn-Cunningham et al.); and (iii) that the strength of the precedence effect tended to decrease somewhat at the higher stimulus level for 10-ms ICI. This psychophysical observation of a decrease of the precedence effect strength with increasing stimulus level for 10-ms ICI appears to be in opposite direction to what one may anticipate from the neurophysiological observations of an increase of the recovery time with increasing stimulus level (Fig. 11 of the present paper; Figs. 7 and 11 of Parham et al., 1996). The reason for this apparent discrepancy is not clear.To summarize, a survey of the available results throughout the auditory pathway suggests that the response to a trailing sound or echo can be either weakened (a characteristic of the AN fibers), enhanced or suppressed. As we ascend the auditory pathway, suppression appears to become more robust or prolonged, but an enhanced representation of a simulated echo continues to persist. These results suggest that, with respect to echo processing, the CN may be the origin of several parallel channels within which the AN signal is either relayed or altered on its way to higher centers.AcknowledgementsThis study was supported in part by the National Institute on Deafness and Other Communication Disorders, NIH, Grant P01-DC-01366. We thank Drs. S. Kuwada, D. Fitzpatrick, and A. Palmer for their helpful comments on this manuscript.ReferencesBackoff et al., 1997P.M.BackoffP.S.PalombiD.M.CasparyGlycinergic and GABAergic inputs affect short-term suppression in the cochlear nucleusHear. Res.1101997155163Von Bekesy, 1960Von Bekesy, G., 1960. Experiments in Hearing. McGraw-Hill, New York, NY.Blauert, 1997Blauert, J., 1997. Spatial Hearing: The Psychophysics of Human Sound Localization. MIT Press, Cambridge, MA.Boettcher et al., 1990F.A.BoettcherR.J.SalviS.S.SaundersRecovery from short-term adaptation in single neurons in the cochlear nucleusHear. Res.481990125144Bonfils and Puel, 1987P.BonfilsJ.-L.PuelFunctional properties of the crossed part of the medial olivo-cochlear bundleHear. Res.281987125130Cant, 1981N.B.CantThe fine structure of two types of stellate cells in the anterior division of the anteroventral cochlear nucleus of the catNeuroscience6198126432655Fitzpatrick et al., 1995D.C.FitzpatrickS.KuwadaR.BatraC.TrahiotisNeural responses to simple simulated echoes in the auditory brain stem of the unanesthetized rabbitJ. Neurophysiol.74199524692486Fitzpatrick et al., 1998Fitzpatrick, D.C., Kuwada, S., Kim, D.O., 1998. Responses of neurons to click-pair stimuli: auditory nerve to auditory cortex. Assoc. Res. Otolaryngol. Abstr. 21.Fitzpatrick et al., 1997D.C.FitzpatrickS.KuwadaK.ParhamResponses of neurons in the auditory cortex to click-pair stimuliAssoc. Res. Otolaryngol. Abstr.201997207Gaumond et al., 1982R.P.GaumondC.E.MolnarD.O.KimStimulus and recovery dependence of cat cochlear nerve fiber spike discharge probabilityJ. Neurophysiol.481982856873Ghoshal and Kim, 1996S.GhoshalD.O.KimMarginal shell of the anteroventral cochlear nucleus: intensity coding in single units of the unanesthetized, decerebrate catNeurosci. Lett.20519967174Ghoshal and Kim, 1997S.GhoshalD.O.KimMarginal shell of the anteroventral cochlear nucleus: single-unit response properties in the unanesthetized decerebrate catJ. Neurophysiol.77199720832097Haas, 1951H.HaasThe influence of a single echo on the audibility of speechAcoustica119514958Harris et al., 1963G.G.HarrisJ.L.FlanaganB.J.BernardBinaural interaction of a click with a click pairJ. Acoust. Soc. Am.351963672678Kawase and Liberman, 1992T.KawaseM.C.LibermanSpatial organization of the auditory nerve according to spontaneous discharge rateJ. Comp. Neurol.3191992312318Kiang, 1965N.Y.S.KiangStimulus coding in the auditory nerve and cochlear nucleusActa Oto-Laryng.591965186200Kim et al., 1991D.O.KimK.ParhamJ.G.SirianniS.O.ChangSpatial response profiles of the posteroventral cochlear nucleus neurons and auditory-nerve fibers in unanesthetized decerebrate cats: responses to pure tonesJ. Acoust. Soc. Am.89199128042817Kim et al., 1992D.O.KimK.ParhamT.TracyResponses of auditory nerve fibers and anteroventral cochlear nucleus units to click pairs in unanesthetized decerebrate catsAssoc. Res. Otolaryngol. Abstr.15199216Koerber et al., 1966K.C.KoerberR.R.PfeifferW.B.WarrN.Y.KiangSpontaneous spike discharges from single units in the cochlear nucleus after destruction of the cochleaExp. Neurol.161966119130Leake and Snyder, 1989P.A.LeakeR.L.SnyderTopographic organization of the central projections of the spiral ganglion in catsJ. Comp. Neurol.2811989612629Liberman, 1991M.C.LibermanCentral projections of auditory-nerve fibers of differing spontaneous rate. I. Anteroventral cochlear nucleusJ. Comp. Neurol.3131991240258Ostapoff et al., 1994E.M.OstapoffJ.J.FengD.K.MorestA physiological and structural study of neuron types in the cochlear nucleus. II. Neuron types and their structural correlation with response propertiesJ. Comp. Neurol.34619941942Palmer et al., 1996A.R.PalmerD.JiangD.H.MarshallResponses of ventral cochlear nucleus onset and chopper units as a function of signal bandwidthJ. Neurophysiol.751996780794Parham and Kim, 1992K.ParhamD.O.KimAnalysis of temporal characteristics of dorsal cochlear nucleus neurons of unanesthetized decerebrate catsJ. Neurophysiol.67199212471263Parham et al., 1996K.ParhamH.B.ZhaoD.O.KimResponses of auditory nerve fibers of the unanesthetized decerebrate cat to the click pairs as simulated echoesJ. Neurophysiol.7619961729Pfeiffer, 1966R.R.PfeifferClassification of response patterns of spike discharges for units in the cochlear nucleus: tone-burst stimulationExp. Brain Res.11966220235Relkin and Pelli, 1987E.M.RelkinD.G.PelliProbe tone thresholds in the auditory nerve measured by two-interval forced-choice proceduresJ. Acoust. Soc. Am.82198716791691Relkin and Doucet, 1991E.M.RelkinJ.R.DoucetRecovery from prior stimulation. I: Relationship to spontaneous firing rates of primary auditory neuronsHear. Res.551991215222Rhode et al., 1983W.S.RhodeD.OertelP.H.SmithPhysiological response properties of cells labeled intracellularly with horseradish peroxidase in cat ventral cochlear nucleusJ. Comp. Neurol.2131983448463Rouiller and Ryugo, 1984E.M.RouillerD.K.RyugoIntracellular marking of physiologically characterized cells in the ventral cochlear nucleus of the catJ. Comp. Neurol.2251984167186Sayers and Toole, 1964B.McA.SayersF.E.TooleAcoustic-image lateralization judgments with binaural transientsJ. Acoust. Soc. Am.36196411991205Sento and Ryugo, 1989S.SentoD.K.RyugoEndbulbs of Held and spherical bushy cells in cats: morphological correlates with physiological propertiesJ. Comp. Neurol.280198955562Shinn-Cunningham et al., 1993B.G.Shinn-CunninghamP.M.ZurekN.DurlachAdjustment and discrimination measurements of the precedence effectJ. Acoust. Soc. Am.93199329232932Shofner and Young, 1985W.P.ShofnerE.D.YoungExcitatory/inhibitory response types in the cochlear nucleus: relationships to discharge patterns and responses to electrical stimulation of the auditory nerveJ. Neurophysiol.541985917939Shore, 1995S.E.ShoreRecovery of forward-masking responses in ventral cochlear nucleus neuronsHear. Res.8219953143Smith and Rhode, 1986P.H.SmithW.S.RhodeEncoding timing and intensity in the ventral cochlear nucleus of the catJ. Neurophysiol.561986261286Smith and Rhode, 1987P.H.SmithW.S.RhodeCharacterization of HRP-labeled globular bushy cells in the cat anteroventral cochlear nucleusJ. Comp. Neurol.2661987360375Smith and Rhode, 1989P.H.SmithW.S.RhodeStructural and functional properties distinguish two types of multipolar cells in the ventral cochlear nucleusJ. Comp. Neurol.2821989595616Wallach et al., 1949H.WallachE.V.NewmanM.R.RosenzwiegThe precedence effect in sound localizationAm. J. Psychol.571949315336Wickesberg, 1996R.E.WickesbergRapid inhibition in the cochlear nuclear complex of the chinchillaJ. Acoust. Soc. Am.100199616911702Wickesberg and Oertel, 1990R.E.WickesbergD.OertelDelayed, frequency-specific inhibition in the cochlear nuclei of mice: a mechanism for monaural echo suppressionJ. Neurosci.10199017621768Winter and Palmer, 1995I.M.WinterA.R.PalmerLevel dependence of cochlear nucleus onset unit responses and facilitation by second tones and broadband noiseJ. Neurophysiol.731995141159Yin, 1994T.C.YinPhysiological correlates of the precedence effect and summing localization in the inferior colliculus of the catJ. Neurosci.14199451705186Young, 1984Young, E.D., 1984. Response characteristics of neurons of the cochlear nuclei. In: Berlin, C.I. (Ed.), Hearing Science, Recent Advances. College-Hill Press, San Diego, CA, pp. 423–460.Young et al., 1988E.D.YoungJ.M.RobertW.P.ShofnerRegularity and latency of units in ventral cochlear nucleus: implications for unit classification and generation of response propertiesJ. Neurophysiol.601988129
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B4C81E27458178FDFF60AA067D04A887C57B142.txt b/test/dataset/in/resources/corpus/Clean_0B4C81E27458178FDFF60AA067D04A887C57B142.txt
deleted file mode 100644
index c8d76d9..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B4C81E27458178FDFF60AA067D04A887C57B142.txt
+++ /dev/null
@@ -1 +0,0 @@
-]>HEARES2986S0378-5955(98)00018-510.1016/S0378-5955(98)00018-5Elsevier Science B.V.Fig. 1Drawing of a stage 52 larva of the newt Cynops pyrrhogaster, including a three-dimensional reconstruction of the inner ear and endolymphatic system, which are shown at higher magnification below. Abbreviations: UTO, utricular otolith; ES, endolymphatic sac; EO, endolymphatic otolith; ED, endolymphatic duct; SO, saccular otolith.Fig. 2Light micrographs of gross changes in the endolymphatic sac, otic vesicle, midbrain, during development and maturation in the newt Cynops pyrrhogaster. The sections were cut transversely through the middle portion of the endolymphatic sac (1 μm thickness) and stained with 1% toluidine blue. Bar=50 μm. a: Stage 39, an oval shaped endolymphatic sac first appears at the dorsomedial tip of the otic vesicle. The endolymphatic sac is adjacent to the midbrain. b: Stage 41, several tiny vacuoles (arrow in enlarged insert) appear in the sac, indicating that the endolymphatic sac is starting to expand. Scale bar in magnified insert is 10 μm. c: Stage 46, the endolymphatic sac elongates significantly. A slit-like lumen starts forming in the middle portion of the saccule. The endolymphatic duct opens into the otic vesicle. The epithelial lining becomes distinguishable. d: Stage 48, the lumen of the endolymphatic sac expands, and is clearly distinguished from the outer epithelial lining. This section shows that the proximal portion of the endolymphatic sac has extended to the upper end of the medial wall of the otic vesicle. e: Stage 50, the endolymphatic sac lumen enlarges and a number of bright otoconia are present. f: Stage 54, the endolymphatic sac expands and extends markedly forming a large lumen. The lumen of the endolymphatic sac accumulates numerous otoconia nearest the duct entering the otic vesicle (OC and arrow in enlarged insert, scale bar=10 μm). Abbreviations: OT, otic vesicle; MB, midbrain; LM, lumen; EL, epithelial lining; OC, otoconia. Other abbreviations as in Fig. 1.Fig. 3TEM of epithelial cells of the ES at different developmental stages. Bar=1 μm. a: Stage 39, the epithelial cell is of polymorphous type and contains a minimum of cytoplasm. Note the intercellular space is narrow and has a smooth appearance. b: Stage 44, the epithelial cell has a columnar shape. The cytoplasm is not dense with only a few mitochondria seen. The intercellular space contains several vacuoles and primitive microvilli on its surface. The basement membrane blends with the fibrils of the connective tissue (unlabeled arrow). c: Stage 48, the epithelial cell contains dense cytoplasm in which mitochondria are evident. The dilated intercellular space has an irregular surface with finger-like projections where cell junctions form. d: Stage 54, the epithelial cell contains a very dense cytoplasmic matrix and has a fibrous appearance. Abundant cell organelles such as mitochondria, smooth endoplasmic reticulum and Golgi complexes are seen in the cytoplasm. Note many granules and vesicles are also located in the apical and lateral cytoplasm. The intercellular space is loose and has a labyrinthine appearance. Abbreviations: EP, epithelial cell; IS, intercellular space; LV, lipid vesicle; VC, vacuole; M, mitochondria; MV, microvilli; N, nucleus; CJ, cell junction; GR, granules; V, vesicles. Other abbreviations as in Fig. 2.Fig. 4TEM of the lumen of the ES at different stages during development. Bar=5 μm. a: Stage 46, a slit-like lumen goes longitudinally through the middle part of the sac. The intercellular spaces are perpendicular to the surface of the lumen. The vacuoles in the space are fused, which makes the space have a wave-like surface. Note that there is a widened intercellular space between and beneath the epithelial cells. b: Stage 48, the ES lumen expands. This shows an uneven lining formed by microvilli. Floccular material appears in the lumen. c: Stage 50, the lumen dilates markedly. Otoconia are present in the lumen. The spaces where the otoconia were appear as holes, indicating artifacts of processing. The floccular material becomes more dense nearest to the otoconia. The intercellular spaces have a honeycomb appearance. d: Stage 54, the lumen expands its width and length significantly with dense floccular material inside. The relatively thick and long microvilli arise from the luminal surface of the flattened epithelial cells. Abbreviations: FM, floccular material. Other abbreviations as in Figs. 2 and 3.Fig. 5TEM of two morphologies of the ES appearing at stage 54. a: High resolution TEM of a matured microvillus on the luminal surface of an epithelial cell. Note that the lipid bilayer is visible. A small vesicle is also found in the cytoplasm. Bar=0.1 μm. b: TEM of a free floating cell. The cytoplasm of this cell is relatively light with numerous vacuoles and empty mitochondria inside. Long microvilli are present around all the free surfaces of the cell. Bar=1 μm. Abbreviations: LB, lipid bilayer; FF, free floating cell. Other abbreviations as in Figs. 2–4.Fig. 6TEM of the endolymphatic duct during development and maturation. The sections were cut longitudinally through the duct. a: Stage 46, the primitive endolymphatic duct with a narrow lumen within the medial wall of the otic vesicle. Bar=5 μm. b: Stage 50, the lumen of the endolymphatic duct expands at the intermediate portion. Some floccular material is present in the lumen. Note that the luminal surface is relatively smooth with only a few microvilli seen. Bar=2 μm. c: Stage 54, the intermediate portion of the endolymphatic duct. Note that the epithelial cells contain dense cytoplasm. The luminal surface bears numerous microvilli that display an intermeshing appearance. Also note that there are fibrils in the basement membrane. Bar=1 μm. Abbreviations: GC, Golgi complex; F, fibrils. Other abbreviations as in Figs. 2–4.Fig. 7TEM of junctional complex between the epithelial cells of the ES during development. a: Stage 46, tight junctions are formed close to the lumen. The outer layers of the adjacent plasma membrane fuse in the wall of the junction. Bar=0.1 μm. b: Stage 50, a tight junction and an adherent junction are present between the epithelial cells, indicating that junctional complex has developed. Note some granular substance is adhering along the outer wall of the tight junction and cytoplasmic filaments are associated with the adherent junction. Bar=0.2 μm. c: Stage 54, the walls of the tight junction are dense and parallel with a narrow interspace. The cytoplasmic filaments crossing the adherent junction are longer than those at stage 50. Below the junctional complex, both of the cell membranes are closely apposed at some parts. Bar=0.2 μm. Abbreviations: AJ, adherent junction; TJ, tight junction; CF, cytoplasmic filaments. Other abbreviations as in Figs. 2–4.Table 1Abbreviations used in figuresAJadherens junctionFMfloccular materialNnucleusCFcytoplasmic filamentsGCGolgi complexOCotoconiaCJcell junctionGRgranuleOTotic vesicleEDendolymphatic ductISintercellular spaceSOsaccular otolithELepithelial liningLBlipid bilayerTJtight junctionEOendolymphatic otolithLMlumenUTOutricular otolithEPepithelial cellLVlipid vesicleVvesicleESendolymphatic sacMmitochondriaVCvacuoleFfibrilsMBmidbrainFFfree floating cellMVmicrovilliTable 2Quantitative and qualitative characteristics of the developing ESStageDSizeType of cellCytoplasm and cell organellesMicrovilli (μm)Intercellular spaceLumen (μm)Material in lumen391165×25×15polymorphousminimumnarrow, smooth4415columnarnot dense, only a few Mvacuoles4618120×30×25cylindricalincrease in M0.3–0.5fused vacuoles, expanded, wave-like39×3.3×44820150×35×40cylindrical, cuboidalrelatively dense M, scattered R and G0.4–0.6dilated, finger-like protrusions, expandedfloccular material5026160×35×50cuboidaldense, packed M, R, and G vesicles0.5–0.7honeycomb, intercellular digitations120×20×20floccular materialotoconia5448190×40×55cuboidal, flatvery dense with fibrous appearance; abundant M, R, and G vesicles and granules1.0labyrinthine, intercellular digitations, floccular material140×30×25floccular materialotoconiacell debrisfloating cellsD, days after eggs are laid; M, mitochondria; R, ribosome; G, Golgi complex.Table 3Quantitative and qualitative characteristics of the developing EDStageLength (μm)Diameter of lumen (intermediate) (μm)Diameter of lumen (low end) (μm)Cell typeMicrovilli (intermediate)Microvilli (lower end)Material in lumen39250046450.50.8duct-likea few501000.81.0lower cuboidalonly a fewa fewfloccular material542001.22.0low cuboidal and flatintermeshinga fewfloccular materialotoconiaDevelopment of the endolymphatic sac and duct in the Japanese red-bellied newt, Cynops pyrrhogasterWenyuanGaoMichael LWiederhold*Jeffery LHarrisonDepartment of Otolaryngology-Head and Neck Surgery, University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78284-7777, USA*Corresponding author. Tel.: +1 (210) 567 5655; Fax: +1 (210) 567 3617.AbstractThe development and maturation of the endolymphatic sac (ES) and duct (ED) were studied in the newt Cynops pyrrhogaster. The ES first appears as an oval capsule at the dorsal-medial tip of the otic vesicle at stage 39, about 11 days after oviposition. The ES consists of polymorphous epithelial cells with a minimum of cytoplasm. The intercellular space (IS) between the epithelial cells is narrow and has a smooth surface. At stage 44, the size of the ES increases as many vacuoles in the IS become filled. At stage 46, 18 days after oviposition, the ES elongates markedly and a slit-like lumen is found in the ES. The epithelium contains a few cell organelles which are scattered in the cytoplasm. The vacuoles in the IS are fused, which expands the IS. Two days later (stage 48), floccular material (endolymph) is present in the expanded lumen. The IS dilates and has a wide and irregular appearance. At stage 50, approximately 26 days after oviposition, the ES extends and expands significantly and crystals (otoconia) can now be seen in the widened lumen of the ES. The cytoplasm of the cuboidal epithelial cells contains an abundance of vesicles surrounded by ribosomes and Golgi complexes. Intercellular digitations are formed in the expanded IS. At stage 54, the ES forms a large bellow-like pouch. Numerous otoconia accumulate in the lumen. Free floating cells and cell debris can be seen in the lumen at this stage. The epithelial cells contain numerous cytoplasmic organelles which are evenly distributed in the cytoplasm. Granules are found in the apical and lateral cytoplasm. The IS is loose and displays a labyrinthine appearance. The primitive ED first appears as a connection between the ES and the saccule but no lumen is present inside at stage 39. At stage 46, a narrow lumen is formed in the ED, which corresponds to the formation of the ES lumen. At stage 50, as the ED extends, floccular material is seen in the lumen. At stage 54, the ED bears numerous microvilli on its luminal surface. Otoconia and endolymph are present in the ED. Tight junctions between the epithelial cells are formed at stage 46. A fully developed intercellular junctional complex is produced at stage 54. Based on the development of the ES and ED, the maturation of function of the ES and ED are discussed.KeywordsEndolymphatic sac and ductDevelopmentOtoconiaIntercellular space1IntroductionThe endolymphatic sac (ES) and duct (ED), which are part of the membranous labyrinth, exist in most vertebrates. The biological function of this portion of the inner ear has received much attention in mammals. Absorption and endocytosis of endolymph (Fukazawa et al. 1990, 1991, 1995; Hoshikawa et al., 1994), secretion of endolymph and other substances (Friberg et al., 1986; Barbara, 1989; Rask-Andersen et al., 1991), pressure regulation of the endolymphatic compartment (Bagger-Sjöbäck and Rask-Andersen, 1986; Friberg et al., 1986; Barbara et al., 1987; Takumida et al., 1988) and immunoreactive function in the inner ear (Rask-Andersen and Stahle, 1980; Tomiyama and Harris, 1986) have been suggested for the entire endolymphatic system. In amphibians, however, the only function of the ES which is known so far is to store calcium. This stored calcium is used to make otoconia and later to mineralize the skeleton and in egg production (Guardabassi, 1960; Marmo et al., 1986).One way to obtain further knowledge about the ES and ED is to study the embryology of the system. Bast and Anson (1949)found that the human ES forms early in a 7-mm embryo by an anterior-posterior crease in the otic vesicle. Watske and Bast (1950)showed changes in size and position of the ES during its maturation. Apart from the studies by Hultcrantz et al. (1987, 1988)in which the ES and ED were investigated at the light and electron microscopic level in CBA/CBA mouse, few studies have been devoted to the development of this system in mammals. In amphibians, only a few early studies of the embryology of the ED and ES have been reported. Whiteside (1922)demonstrated in Rana temporaria linne, that the ED is a well-differentiated canal, situated at the medial side of the otic vesicle; its upper end expands into a small vesicle to form the ES during the first stage (4 mm long). Dempster (1930)reported that the formation of the ES starts as a dorsal evagination of the closed otic vesicle in 15 mm long cryptobranchidae alleganiensis. Unfortunately, these studies were based on light microscopic observations and did not describe the whole developmental process.The Japanese newt Cynops pyrrhogaster is a favorable species in which to study development, since its vestibular system is similar to those of mammals but develops much more rapidly. The development of the otolith organs and semicircular canals, the gross appearance of the ES and ED and the morphogenic features of otoconia in the utricle and saccule have been investigated in this laboratory (Wiederhold et al. 1992, 1995, 1997; Steyger et al., 1995). However, there is a lack of detailed information concerning the development of the ES and ED. The present study was performed to elucidate the pattern and sequence of events during organogenesis and cytodifferentiation of the ES and ED at the light and electron microscope level. It is hoped that this study will shed light on the function of this system in amphibians.2Materials and methodsNewt embryos of the desired developmental stage were obtained as reported previously (Wiederhold et al., 1995; Steyger et al., 1995). Development stages were determined by observation through a dissecting microscope and comparison with the sequence and stage description reported by Okada and Ichikawa (1947)and Okada (1989).Three embryos or larvae at each stage (from stage 25 to 56) were used as subjects. The specimens were fixed by immersion in 3% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) for 2 h. After a triple-rinse in 0.1 M cacodylate buffer, the specimens were post-fixed in 1% osmium tetroxide (pH 7.4) for 1 h, and again rinsed in buffer. Specimens were dehydrated with increasing concentrations of ethanol (35% to 100%, 15 min each). A transition fluid of propylene oxide was used (45 min) before infiltration and embedding in Epon. The polymerization process required three days in an oven (37, 45, 60°C for 24 h each).Serial transverse sections (1 μm thickness) were cut through the ES and ED. The sections were stained with 1% toluidine blue, mounted on slides and cover slipped. The sections were examined under an Olympus BH-2 light microscope. Ultrathin sections (90 nm) were cut transversely at the middle portion of the ES and ED. The sections were stained with uranyl acetate and Reynold's lead citrate (Reynolds, 1963), and examined with a Philips 301 electron microscope.The care and use of the animals reported in this study were approved by the University of Texas Health Science Center at San Antonio's Institutional Animal Care and Use Committee.3Results3.1Development of the endolymphatic sac (ES)3.1.1Gross structureThe endolymphatic system is much larger, relatively, in adult and developing amphibians, than in mammals. Fig. 1 (in Table 1 are the abbreviations used in the figures) shows a drawing of a stage 52 Cynops larva and a three-dimensional reconstruction of the inner ear and ES and ED. Stage 52 is approximately 28 days after oviposition and 16 days after the larvae usually hatch from the egg. At this stage, the two ESs are separate but represent large projections from the dorsomedial aspect of the saccule. As early as stage 31, the precursor of the ES can be recognized as a slight extension of the otic vesicle (see Wiederhold et al., 1995) and both the ES and ED expand greatly from stages 50 to 56. In an adult newt, the ES from the two sides join to form a continuum across the top of the brainstem (see Koike et al., 1995).Before stage 38, serial sections did not reveal an ES. At stage 39, the ES first appears as an oval capsule at the dorsomedial tip of the otic vesicle (Fig. 2a). The ES is about 65 μm long, 25 μm wide and 15 μm thick. Serial sections did not show a lumen inside the ES. At stage 41, several tiny vacuoles (arrow in insert of Fig. 2b) appear in the sac, indicating the ES is ready to expand. At stage 46, the ES elongates markedly, to approximately 120 μm long and 30 μm wide. The capsule boundary appears in 25 serial sections. Its proximal portion extends to the upper part of the medial wall of the otic vesicle. A narrow slit-like lumen can barely be seen in the middle of the ES (LM in Fig. 2c). At stage 48, the ES increases in size with a long and short diameter of 150 μm and 35 μm, respectively. The whole ES continued to display its outline in about 40 serial sections with the expanded lumen clearly distinguished from the wall of the ES (LM in Fig. 2d). At stage 50 the ES is 160 μm long, 35 μm wide and 50 μm thick. The lumen becomes much wider than that at stage 48. In the lumen, a number of crystals (otoconia) can now be seen (OC in Fig. 2e). The birefringence of these otoconia display the same colors in polarized light microscopy as do the saccular and utricular otoconia, indicating that they are, indeed, crystalline. At stage 54, the ES forms a large bellow-like pouch, which is 190 μm long, 40 μm wide and 55 μm thick. The sac extends from the middle portion of the medial wall of the otic vesicle to the dural membrane. The large lumen cavity contains numerous otoconia piled together (arrow in magnified insert in Fig. 2f).3.1.2Epithelial cell and intercellular spaceAt stage 39, the epithelial cells are of a polymorphous type and contain a minimum of cytoplasm with very few organelles (Fig. 3a). At stage 44, the epithelial cells have a columnar shape. The cell cytoplasm has increased in size but is not dense, with only a few mitochondria inside (Fig. 3b). At stage 46, the epithelium of the ES contains a single layer of cylindrical cells. The content of mitochondria appears increased. A few microvilli appear on the epithelial cell luminal surface. At stage 48, the epithelial cells are of both cuboidal and cylindrical types. The epithelial cells contain relatively dense cytoplasm in which many mitochondria, as well as scattered ribosomes and Golgi complexes can be found (Fig. 3c). At stage 50, the epithelial cells are cuboidal, with a height of about 8 μm and a width of 7 μm. The dense cytoplasm contains an abundance of vesicles surrounded by Golgi complexes and ribosomes. Mitochondria are packed together in the cytoplasm of some epithelial cells (see Fig. 4c). At stage 54, cells are cuboidal at the proximal portion but more flattened at the intermediate and distal portion. The cytoplasm is of high density with a fibrous appearance. Abundant cytoplasmic organelles such as mitochondria, ribosomes, Golgi complexes, endoplasmic reticulum are distributed in the cytoplasm. Numerous granules are found in the apical and lateral cytoplasm (GR in Fig. 3d) at this stage. These granules, 0.2–0.4 μm in diameter, are usually round or oval and surrounded by a limiting membrane.The epithelial cells are separated by intercellular spaces (ISs). At stage 39, these spaces are narrow and have a smooth surface, with nothing inside (Fig. 3a). At stage 44, the ISs are radially distributed from the center of the ES with some vacuoles filled in the space (see Fig. 3b). Primitive microvilli are occasionally found on the lateral surface of the intercellular spaces (MV in Fig. 3b). At stage 46, the vacuoles in the IS are fused, which makes the space expand and have a wave-like surface (see Fig. 4a). At stage 48, the ISs dilate irregularly, in which finger-like protrusions arise from adjacent surfaces (CJ in Fig. 3c). The ISs expand markedly and have a honeycomb-like appearance at stage 50 (see Fig. 4c). Numerous protrusions arise from the surface of the space to form intercellular digitations. At stage 54, floccular material is found in the widened irregular IS (not illustrated). The IS is loose and contains many digitating projections, which display a labyrinthine appearance (Fig. 3d).3.1.3LumenIn Fig. 4a, the slit-like lumen of the ES lies longitudinally in the middle of the ES at stage 46. The lumen is approximately 39 μm long and 3.3 μm wide and displays its outline in only four serial sections. The luminal surfaces of the epithelial cells bear a few short microvilli, which are about 0.3–0.5 μm long. At stage 48, the ES lumen becomes wider and longer than before. The epithelial surface is uneven with several short protrusions and microvilli in the lumen (MV in Fig. 4b). The microvilli are longer than before (0.4–0.6 μm). Floccular material (endolymph) appears in the lumen (FM in Fig. 4b). This material is not dense and is evenly distributed within the lumen. Crystals were not found in the lumen in serial sections at this stage. At stage 50, the ES lumen dilates markedly and is about 120 μm long, 20 μm wide and 20 μm thick. The luminal surface is irregular with several protrusions into the lumen (not illustrated). The microvilli on the luminal surface are relatively long (0.5–0.7 μm) and curved. Otoconia are first found in the lumen at stage 50. The spaces where the crystals were appear bright under TEM, indicating that these spaces are ghosts of dislodged or dissolved otoconia. The endolymph becomes dense at places adjacent to the otoconia. At stage 54, the ES lumen forms a large cavity (Fig. 4d) that is about 140 μm long, 30 μm wide and 25 μm thick. In the lumen, numerous otoconia pile together. The size of individual otoconia is similar to that at stage 50. The floccular material is dense and contains occasional cell debris. Thick and long microvilli (1 μm long and 0.1–0.2 μm wide) arise from the luminal surface of the epithelial cell. At high magnification, these microvilli have a triple layered unit membrane with the middle layer more dense, indicating a lipid bilayer (LB in Fig. 5a). The floccular material close to the microvilli and the otoconia sometimes becomes very condensed as seen in Fig. 5a. Free floating cells first appear in the ES lumen at this stage. The cytoplasm of this cell is relatively light with numerous vacuoles inside. Long microvilli (2 μm) are present around all the free surface of the cell (Fig. 5b).3.2Development of the endolymphatic duct (ED)Before stage 38, no ED was found in serial sections. At stage 39, a primitive ED is present as a connection, about 25 μm long, between the ES capsule and the medial side of the otic vesicle (ES in Fig. 2a). Serial sections did not show a lumen inside this connection. At stage 41, the ED extends and expands, but still no lumen was found. A true short and thin tube (ED lumen), which joins the ES lumen and passes into the otic vesicle lumen at the upper end of the saccule (SC), is present at stage 46 (LM in Fig. 2c,Fig. 6a). The ED is about 55 μm long and has a lumen diameter of 0.5 μm. The ED is lined with a single layer of duct-like epithelial cells which rest on a smooth basement membrane. A large nucleus with its long axis parallel to the duct can be seen in the cytoplasm. Intercellular spaces are present between the epithelial cells. At the low end of the ED, near the saccule, the duct dilates to 0.8 μm in diameter. Microvilli are seen only on the luminal surface of this dilated end of the ED (ED in Fig. 6a).At stage 50, the ED extends its length to about 100 μm in length with a lumen diameter of 0.8 μm. At the low end of the ED, the diameter of the lumen extends to 1.0 μm. The lower cubical epithelial cells of the ED contain more cytoplasm than before. The nucleus usually lies at the base. Some floccular material (FM in Fig. 6b) can now be seen in the lumen. The luminal surface of the ED is relatively smooth and bears only a few microvilli (Fig. 6b).At stage 54, the ED is about 200 μm long with a lumen diameter of 1.2 μm. At the low end of the ED, close to the saccule, the diameter of the lumen extends to 2 μm. In the low portion of the ED, the epithelial cells bear a few short microvilli which rise perpendicularly from the apical surface. In the intermediate and transitional portion, the epithelial cells bear numerous relatively long and thick microvilli. These microvilli protrude into the lumen at different directions and touch or overlap each other, which makes the lumen have an intermeshing appearance (MV in Fig. 6c). The epithelial cell becomes low cubical. The cytoplasm is rather dense and has many mitochondria, Golgi complexes, vesicles and vacuoles. The nucleus is irregular in shape with condensed nucleoplasm inside. Floccular material and otoconia are occasionally found in the lumen of the ED. A well-defined and dilated IS can be seen between the epithelial cells.3.3Development of junctional complexes between the epithelial cellsBefore stage 45, there is no sign of junctions between the adjacent epithelial cells in the ES and ED. The epithelial cells are connected to each other with a smooth or rugose IS. At stage 46 (Fig. 7a), tight junctions first appear in the lateral space between the epithelial cells close to the luminal surface of the ES. The outer layers of the adjacent plasma membrane fuse in the wall of the junction. The junction is rather short (0.2 μm long). Very fine filamentous material fills the interspace of the junction. Below the junction, the IS expands and has an irregular surface.At stage 50 (Fig. 7b), the length of the tight junction increases (0.4 μm). In the cytoplasm of the adjacent cells close to the junction, some granular substance adheres to the outer wall of the junction. Below the tight junction, an adherent junction 0.15 μm long is formed, indicating that a junctional complex has developed. The adherent junction consists of two dark cytoplasmic plaques very close to the innermost layer of the plasma membrane (AJ in Fig. 7b). Several parallel filaments traverse across the junction and extend to the cytoplasm of both cells (CF in Fig. 7b). Below the junctional complex, cellular digitations can be seen in the lateral intercellular space.At stage 54 (Fig. 7c), the cell walls of the tight junction are parallel with a narrow intercellular space of about 200 Å. In the adhering junction, there are two oval plaques of dense material in the cytoplasm close to the inner layer of the plasma membrane. The crossing parallel filaments become longer and have a microtubule-like appearance (CF in Fig. 7c). Below the junctional complex, both of the cell membranes are closely apposed at some point (TJ in Fig. 7c) indicating that multiple cell junctional complexes are formed.4DiscussionThe membranous labyrinth of the inner ear derives from ectoderm. In C. pyrrhogaster, a small protuberance of the otic vesicle appears as a precursor of the ES, on the dorsal-medial aspect of the saccule at stage 31, about 7 days after the eggs are laid (Wiederhold et al., 1995). In the present study, a true egg-like capsule, which is composed of epithelial cells of the ES and a connection between the capsule and the saccule, which forms the primitive ED, are seen at stage 39, approximately 11 days after oviposition. It is believed that the protuberance of the otic vesicle extends to produce the ED and its upper end expands to form the ES during development.Before the 1980's, investigations on the development of the inner ear only briefly mentioned or failed to describe the process of organogenesis of the ES and ED in mammals (Ruben, 1967; Sher, 1971). In amphibians, earlier descriptions of the development of C. pyrrhogaster showing histological sections through the brain of embryos up to stage 42 did not mention the ES and ED (e.g. Okada and Ichikawa, 1947). In this study, the whole sequence of maturation of the ES and ED has been demonstrated, both concerning organogenesis and epithelial cell cytogenesis (Table 2Table 3).From Tables 2 and 3, it is known that a slit-like lumen of the ES is formed at stage 46. At the same time, a true lumen of the ED, which joins the ES lumen at its upper end and the lumen of the otic vesicle at the low end, first appears. Several days later (stages 48 and 50), floccular material (endolymph) is found in the expanded ES lumen and the ED. It is believed that the entire endolymphatic system is in communication with the other parts of the inner ear at this period. The floccular material flows into the ES from the saccule, or out to the otic vesicle lumen from the ES, by passing through the ED at stage 46. Apparently, in Cynops, the ES and ED play their biological function only from this stage.The cytodifferentiation of the epithelial cells in Cynops occurs from stage 39 to stage 54. During the process, the height of the cell reduces but the cytoplasm becomes larger with increasing number of cytoplasmic organelles. This indicates that the metabolic activity within the epithelial cells of the ES increases during development. At stage 54, the cytoplasm of the epithelial cells is of high density with a fibrous appearance. Abundant cytoplasmic organelles such as mitochondria, ribosomes, Golgi complexes and endoplasmic reticulum are distributed in the cytoplasm. The fine structure of the epithelial cells at stage 54 is similar to that of the epithelial cells at stages 56 and 58 (a nearly adult newt). Thus, it is concluded that the epithelial cells in the newt C. pyrrhogaster have reached maturation at stage 54.In the present study, the cytoplasm of the epithelial cells of the ES contains numerous vesicles at stage 50 and many granules at stage 54. These vesicles and granules are surrounded by a limiting membrane and located close to the apical cell membrane. Salamat et al. (1980)have also observed vesicles in the cells of the labyrinthine wall of the rat. They believed that the content within the vesicles would be expelled into the labyrinthine lumen. In amphibians, Kawamata et al. (1987)found that most epithelial cells of the ES in the tree frog, Hyla arborea japonica, contain granules. These granules, varying in size, are surrounded by a limiting membrane and apposed to the apical membrane of the cell. Kawamata suggested that these granules are released into the lumen by the epithelial cells. In a histological and ultrastructural study, Dahlmann and Düring (1995) found many ribosomes in some epithelial cells of the ES of the rat. They suggested that supranuclear accumulation of organelles, especially ribosomes, Golgi complexes and vesicles in the cytoplasm are interpreted as signs of secretory activity. In this study, numerous ribosomes and Golgi complexes are richly distributed around the vesicles and granules in the apical cytoplasm of the epithelial cells at stage 50. These findings strongly suggest that, in the newt, the epithelial cells of the ES are secretory after the animals reach stage 50.The intercellular spaces (ISs) between epithelial cells of the ES have received much attention (Friberg et al., 1985; Takumida et al., 1988). This study demonstrates the developmental process of this structure. At stage 39, the primitive IS is narrow and has a smooth surface. Vacuoles are found in the IS at stage 41. At stage 46, about 18 days after the eggs are laid, the IS starts to expand. The intercellular spaces dilate evidently and have an irregular appearance at stage 48. After stage 50, intercellular digitations arise from the surface of the IS. The IS is honeycomb-like between the epithelial cells. At stage 54, the IS contains many intercellular digitations and displays a loose labyrinth appearance.The function of the intercellular spaces of the ES is still not known. Many morphological and experimental studies indicate that the lining of the ES and ED may represent a fluid transporting epithelium with an ability to transfer water and solutes to and from the ES and ED (Lundquist, 1965; Adlington, 1967; Rask-Andersen et al., 1981; Bagger-Sjöbäck and Rask-Andersen, 1986). Friberg et al. (1985)and Takumida et al. (1988)suggested that the intercellular spaces of the ES may form a pathway for transepithelial water flow from the ES to blood vessels. According to their assumption, the epithelial cells may actively transfer salt from the ES lumen to the IS, and then water is drawn through local osmotic force from the lumen to the intercellular space. Based on the `longitudinal theory' (Guild, 1927), endolymph flows from the vestibule out to the ED and ES. Friberg et al. (1985)confirmed that cessation of longitudinal flow to the ES by surgery would make the intercellular spaces collapse. They considered that this cessation of flow also arrests transepithelial movement of fluid. In the present study, the intercellular spaces are not dilated before stage 45, when the lumen of the ED and ES have not yet been formed. After the lumen of the ED and ES have developed at stage 46, the intercellular spaces start to expand. The IS dilates evidently at stage 48 when abundant floccular material appears in the lumen. This finding suggests that fluid transport across the sac epithelium may be initiated at the same time the IS starts to dilate.Tight junctions first appear in the epithelium at stage 46, about the same time the lumen is formed. At stage 50, adherent junctions are present below the tight junctions, indicating that a junctional complex has developed. At stage 54, fully developed multi-junctional complexes are present between the epithelial cells. Tight junctions act as a permeability barrier to fluids and ions (Claude and Goodenough, 1973; Bagger-Sjöbäck and Anniko, 1984). Anniko and Bagger-Sjöbäck (1982) reported that the tight junctions of the mouse have developed before formation of the high potassium concentration in the endolymph. In the present study, tight junctions are formed at stage 46, at which time a slit lumen of the ES and very narrow ED have just developed. Apparently, formation of the tight junctions occurs before endolymph is available in the ES lumen. These tight junctions probably play an important role in maintaining ion gradients between the endolymphatic lumen and lateral and extralabyrinthine space, and the elevated Ca2+ concentration in endolymph which is necessary for formation of otoconia, a major function of the ES in amphibians.An interesting finding in this study is that so many microvilli touch and overlap each other, which gives an intermeshing appearance to the lumen when the ED reaches maturity. The functional significance of this organization of the microvilli is not known, but would suggest an active secretory function.AcknowledgementsThese studies were supported by the NASA Space Biology Program (Grants NAG 2-952 and NAG 10-0180) and the National Science Foundation (Grant BIN-95-29136). We are very grateful to Dr. Bagger-Sjöbäck for his helpful comments on an earlier version of this manuscript.ReferencesAdlington, 1967P.AdlingtonThe ultrastructure and the functions of the saccus endolymphaticus and its decompression in Ménières diseaseJ. Laryngol. Otol.811967759776Anniko and Bagger-Sjöbäck, 1982M.AnnikoD.Bagger-SjöbäckMaturation of junctional complexes during embryonic and early postnatal development of inner ear secretory epitheliaAm. J. Otolaryngol.31982242253Bagger-Sjöbäck and Anniko, 1984D.Bagger-SjöbäckM.AnnikoDevelopment of intercellular junctions in the vestibular end-organ. A freeze-fracture study in the mouseAnn. Otol. Rhinol. Laryngol.9319848995Bagger-Sjöbäck and Rask-Andersen, 1986D.Bagger-SjöbäckH.Rask-AndersenThe permeability barrier of the endolymphatic sac. A hypothesis of fluid and electrolyte exchange based on freeze fracturingAm. J. Otol.71986134140Barbara et al., 1987M.BarbaraH.Rask-AndersenD.Bagger-SjöbäckUltrastructure of the endolymphatic sac in the Mongolian gerbilArch. Otorhinolaryngol.2441987284287Barbara, 1989M.BarbaraCarbohydrate content of the endolymphatic sac. A histochemical and lectin labelling study in the Mongolian gerbilJ. Laryngol. Otol.1031989137142Bast and Anson, 1949Bast, T.H. and Anson, B.J. (1949) The Temporal Bone and the Ear. Ch.C. Thomas, Springfield, IL, 3–100.Claude and Goodenough, 1973P.ClaudeD.A.GoodenoughFracture faces of zonulae occludentes from ``tight'' and ``leak'' epitheliaJ. Cell Biol.581973390400Dahlmann and Düring, 1995A.DahlmannM.V.DüringThe endolymphatic duct and sac of the rat: a histological, ultrastructural, and immunocytochemical investigationCell Tissue Res.2821995277289Dempster, 1930W.T.DempsterThe morphology of the amphibian endolymphatic organJ. Morphol. Physiol.501193071126Friberg et al., 1985U.FribergD.Bagger-SjöbäckH.Rask-AndersenThe lateral intercellular spaces in the endolymphatic sac. A pathway for fluid transport?Acta Otolaryngol. (Stockh.) Suppl.4261985117Friberg et al., 1986U.FribergP.WackymD.Bagger-SjöbäckH.Rask-AndersenEffect of labyrinthectomy on the endolymphatic sac. A histological, ultrastructural and computer-aided morphometric investigation in mouseActa Otolaryngol. (Stockh.)1011986172182Fukazawa et al., 1990K.FukazawaT.MatsunagaH.FujitaUltrastructure of the endolymphatic sac in the guinea pig: with special regards to classification of cell types of the epithelium and uptake of India ink particles into free floating cells and epithelial cells of the sacJ. Clin. Electron. Microsc.231990135147Fukazawa et al., 1991K.FukazawaM.SakagamiT.MatsunagaH.FujitaEndocytotic activity of the free floating cells and epithelial cells in the endolymphatic sac: an electron microscopic studyAnat. Rec.2301991425433Fukazawa et al., 1995K.FukazawaM.SakagamiM.UmemotoT.KuboEndocytosis and transepithelial transport of endolymph in the endolymphatic sacHear. Res.8619958288Guardabassi, 1960A.GuardabassiThe utilization of the calcareous deposits of the endolymphatic sacs of Bufo bufo bufo in the mineralization of the skeleton. Investigations by means of Ca45Z. Zellforsch.511960278282Guild, 1927S.R.GuildObservation upon the structure and normal contents of the ductus and saccus endolymphaticus in the guinea-pig (cavia cobaya)Am. J. Anat.391927166Hoshikawa et al., 1994H.HoshikawaH.FurutaN.MoriS.SakaiAbsorption activity and barrier properties in the endolymphatic sac. Ultrastructural and morphometric analysisActa Otolaryngol. (Stockh.)11419944047Hultcrantz et al., 1987M.HultcrantzD.Bagger-SjöbäckH.Rask-AndersenThe development of the endolymphatic duct and sac. A light microscopical studyActa Otolaryngol. (Stockh.)1041987406416Hultcrantz et al., 1988M.HultcrantzD.Bagger-SjöbäckH.Rask-AndersenThe pre- and postnatal maturation of the epithelium in the endolymphatic sacActa Otolaryngol. (Stockh.)1051988303311Kawamata et al., 1987S.KawamataK.TakayaT.YoshidaLight and electron-microscopic study of the endolymphatic sac of the tree frog, Hyla arborea japonicaCell Tissue Res.24919875762Koike et al., 1995A.KoikeK.NakamuraK.NishimuraI.KashimaM.L.WiederholdM.AsashimaNon-invasive assessment of otolith formation during development of the Japanese red-bellied newt Cynops pyrrhogasterHear. Res.881995206214Lundquist, 1965P.G.LundquistThe endolymphatic duct and sac in the guinea pig. An electron microscopic and experimental investigationActa Otolaryngol. (Stockh.) Suppl.2011965Marmo et al., 1986F.MarmoG.BalsamoP.CrispinoUltrastructural aspects of the endolymphatic organ in the frog Rana esculentaActa Zool.6719865361Okada and Ichikawa, 1947Y.K.OkadaM.IchikawaNormal table of Triturus pyrrhogasterJpn. J. Exp. Morphol.31947165Okada, 1989Okada, T. (1989) Development of Vertebrates. Baifukan, Tokyo.Rask-Andersen and Stahle, 1980H.Rask-AndersenJ.StahleImmunodefense of the inner ear? Lymphocyte-macrophage interaction in the endolymphatic sacActa Otolaryngol. (Stockh.)891980283294Rask-Andersen et al., 1981H.Rask-AndersenG.BredbergL.LyttkensG.LööfThe function of the endolymphatic duct. An experimental study using ionic lanthanum as a tracerAnn. N. Y. Acad. Sci.37419811119Rask-Andersen et al., 1991H.Rask-AndersenN.Danckwardt-LilliestromF.H.LinthicumW.F.HouseUltrastructural evidence of a merocrine secretion in the human endolymphatic sacAnn. Otol. Rhinol. Laryngol.1001991148156Reynolds, 1963E.S.ReynoldsThe use of lead citrate at high pH as an electron opaque stain in electron microscopyJ. Cell Biol.171963208212Ruben, 1967R.J.RubenDevelopment of the inner ear of the mouse: An autoradiographic study of terminal mitosisActa Otolaryngol. (Stockh.) Suppl.2201967Salamat et al., 1980M.S.SalamatM.D.RossD.R.PeacorOtoconial formation in the fetal ratAnn. Otol. Rhinol. Laryngol.891980229238Sher, 1971A.E.SherThe embryonic and postnatal development of the inner ear of the mouseActa Otolaryngol. (Stockh.) Suppl.2851971Steyger et al., 1995P.S.SteygerM.L.WiederholdJ.BattenThe morphogenic features of otoconia during larval development of Cynops pyrrhogaster, the Japanese red-bellied newtHear. Res.8419956171Tomiyama and Harris, 1986S.TomiyamaJ.P.HarrisThe endolymphatic sac: Its importance in inner ear immune responsesLaryngoscope961986685691Takumida et al., 1988M.TakumidaD.Bagger-SjöbäckH.Rask-AndersenThe basement membrane and associated structures in the murine endolymphatic sacArch. Otorhinolaryngol.2451988266272Watske and Bast, 1950D.WatskeT.H.BastThe development of and structure of the otic (endolymphatic) sacAnat. Res.1061950361379Whiteside, 1922B.WhitesideThe development of the saccus endolymphaticus in Rana temporaria linneAm. J. Anat.301922231266Wiederhold et al., 1992M.L.WiederholdM.YamashitaM.AsashimaDevelopment of the gravity-sensing organs in the Japanese red-bellied newt, Cynops pyrrhogasterProc. Int. Symp. Space Technol. Sci.18199221032108Wiederhold et al., 1995M.L.WiederholdM.YamashitaK.LarsenJ.BattenH.KoikeM.AsashimaDevelopment of the otolith organs and semicircular canals in the Japanese red- bellied newt, Cynops pyrrhogasterHear. Res.8419954151Wiederhold et al., 1997M.L.WiederholdW.Y.GaoJ.L.HarrisonR.HejlDevelopment of gravity-sensing organs in altered gravityGravit. Space Biol. Bull.1019979196
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B55479AB153C8E2AA83528EEF0D29CD88601242.txt b/test/dataset/in/resources/corpus/Clean_0B55479AB153C8E2AA83528EEF0D29CD88601242.txt
deleted file mode 100644
index a3b0d63..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B55479AB153C8E2AA83528EEF0D29CD88601242.txt
+++ /dev/null
@@ -1 +0,0 @@
- geronb J Gerontol B Psychol Sci Soc Scigeronb The Journals of Gerontology Series B: Psychological Sciences and Social Sciences J Gerontol B Psychol Sci Soc Sci 1079-5014 1758-5368 Oxford University Press P4110.1093/geronb/63.1.P41 Journal of Gerontology: Psychological Sciences Spousal Social Activity Trajectories in the Australian Longitudinal Study of Ageing in the Context of Cognitive, Physical, and Affective Resources Hoppmann Christiane A. Gerstorf Denis Luszcz Mary 1School of Psychology, Georgia Institute of Technology, Atlanta. 2Department of Human Development and Family Studies, The Pennsylvania State University, University Park. 3Centre for Ageing Studies, Flinders University, Adelaide, Australia. Address correspondence to Christiane A. Hoppmann, Georgia Institute of Technology, School of Psychology, 654 Cherry Street, Atlanta, GA 30332. E-mail: ch295@mail.gatech.edu 1 2008 63 1 P41 P50 6 9 2007 21 12 2006 Copyright 2008 by The Gerontological Society of America 2008 We examined the dyadic interdependence of spousal social activity trajectories over 11 years by using longitudinal data on 565 couples from the Australian Longitudinal Study of Ageing (Mage = 76 years at Time 1). Social activity trajectories were interrelated in elderly couples, and they depended not only on individual but also on spousal cognitive, physical, and affective resources at baseline. Most associations examined were similar in husbands and wives. However, wives performed more social activities and displayed different depression–social activity associations than did husbands. We found stronger within-couple associations in the domain of social activities than for cognition. Our findings illustrate the important role of social relationships for late-life development and suggest that the mechanisms involved in dyadic interdependencies may be domain and gender specific. Successful aging Social activities Couples Growth curve modeling hwp-legacy-fpage P41 hwp-legacy-dochead RESEARCH ARTICLE ENGAGEMENT in social activities is key to successful aging and has attracted considerable interest regarding its association with cognition, physical functioning, and affect (Andrews, Clark, & Luszcz, 2002; Bath & Deeg, 2005; Ghisletta, Bickel, & Lövdén, 2006; Hultsch, Hertzog, Small, & Dixon, 1999; Lang & Baltes, 1997; Lövdén, Ghisletta, & Lindenberger, 2005; Rowe & Kahn, 1997; Ryff & Singer, 2000). Social activities comprise a person's activities that are performed for the purpose of direct interaction with others (Herzog, Ofstedal, & Wheeler, 2002). An important benefit of social activities is that they can be implemented by older adults despite physical limitations that may restrict other activities. Hence, engagement in social activities may continue to contribute to a person's sense of competence in a way not afforded by other types of activities (Herzog et al.). Studies on social activities have primarily been based on unrelated people. Because social activities inherently depend on others, marriage partners are prime social targets, particularly in late life. Couples comprise a unique social unit in old age that is often characterized by a long relational history, frequent interaction, and considerable closeness (Antonucci & Akiyama, 1991; M. M. Baltes & Carstensen, 1998; Dixon, 1999; Lang, 2001). Simply being married may be socially advantageous, given the high rates of bereavement in old age. However, past research shows that the positive potential of marriage has to be qualified by spousal behaviors (Coombs, 1991; Hagedoorn et al., 2006). Investigating social activities in couples therefore addresses one specific process that occurs within marital contexts. Changes in, and entrainment of, social activity trajectories over 11 years in couples from the Australian Longitudinal Study of Ageing (ALSA; Luszcz, 1998) is the central focus of this article. Spousal development is interdependent in many central life domains (Carstensen, Gottman, & Levenson, 1995; Gruber-Baldini, Schaie, & Willis, 1995; Townsend, Miller, & Guo, 2001). The social activities of elderly couples may be interrelated for at least two reasons. First, social networks are dominated by joint kin and become smaller but increasingly interdependent throughout adulthood and into old age (M. M. Baltes & Carstensen, 1998; Kahn & Antonucci, 1980; Lang, 2001; Milardo & Helms-Erikson, 2000). Second, elderly couples typically have long marital histories with many shared experiences. Because each partner has developed an in-depth knowledge of the other's strengths and weaknesses (Dixon, 1999; Meegan & Berg, 2002), each partner's expertise and skills can be invoked for the sake of both partners. For example, if one spouse initiates a social activity (e.g., inviting friends for dinner), the other spouse also will have a social engagement opportunity. We therefore expect social activity trajectories of elderly couples to be closely interrelated. Our second major objective is to examine associations among social engagement by spouses and resources in other domains, namely cognitive, physical, and affective resources. Associations between social activities and cognitive resources may in part be due to the complex nature of social activities that place high demands on cognitive skills, thereby potentially protecting against cognitive decline (Fratiglioni, Paillard-Borg, & Winblad, 2004; Ghisletta et al., 2006; Lövdén et al., 2005; Pushkar et al., 1999). Alternatively, cognitive fitness may be a prerequisite for social activities (Hultsch et al., 1999). We also expect social activities to be positively associated with physical resources. Social activities modulate stress reactivity and promote psychobiological recovery processes that are associated with age-related illnesses (Seeman & McEwen, 1996). Finally, social activities may be linked to affective resources (e.g., resistance to depression) because they fulfill an inner need for affiliation (Cantor & Sanderson, 1999; Herzog et al., 2002). We therefore expect social activities to be positively associated with cognitive, physical, and affective resources. Some of these links with resources may be shared with other activity categories. Cognitive resources in particular are likely to shape a broad range of activities, for example, crossword puzzling (Ghisletta et al., 2006). In contrast to crossword puzzling, we expect that additional resource domains impinge on social activity. This makes social activities particularly interesting for researchers interested in successful aging, and it also highlights the special role of cognitive resources for activities. Hence we also explore dyadic associations in longitudinal trajectories of cognitive functioning (cf. Anstey, Hofer & Luszcz, 2003). This enables us to investigate the domain specificity of entrainment of partner's functioning and to assess whether social activity and cognitive trajectories wax and wane together. Thoroughly examining the relative contributions of different categories of activities or domains of functioning is beyond the scope of this article; rather, we target entrainment and change in social activities as particularly salient in the context of long-term marriage. One feature of this study is the capacity to identify associations of social activities to functioning in other domains, not only at the individual level but also at the couple level. For example, if one spouse experiences serious cognitive decline, resources of the partner may compensate, at the expense of social activities (Clark & Bond, 2000; Tremont, Davis, & Bishop, 2006). Hence, we assume that one spouse's cognition, physical functioning, and affect not only relate to his or her own but also to spousal social activities. Following the terminology of the dyadic literature (Cook & Kenny, 2005), we refer to associations of an individual's characteristics with his or her own score on a given variable as actor effects and with the partner's score on a given variable as partner effects. Applying this nomenclature, we hypothesize that a spouse's cognitive, physical, or affective resources relate not only to his or her own social activities (actor effects) but also to those of the other spouse (partner effects). Finally, the expected relationships between social activities and cognitive, physical, and affective resources may not be symmetric in husbands and wives. A large body of literature indicates that women take a more active part in their social networks than do men, playing an important role in initiating and organizing social activities (Antonucci, 1994, 2001; Moen, 2001). Additionally, relationship maintenance is often seen as obligatory by women whereas men's social activities are more voluntary (Antonucci, 1994). Hence, husbands may not enter all social engagement opportunities that are provided by their wives. We therefore expect wives to report more social activities than husbands. Being more socially invested than men (Moen, 2001) may have both positive and negative implications for wives, because they may not only benefit more from their social activities but also be more vulnerable when circumstances necessitate a change in them (Rook, 1998). We therefore expect that social activities are more strongly linked with both wives' own and their spouses' resources. Hence, wives' social activities are expected to be more closely related to both actor and partner resources than husbands' social activities are. In summary, we examine three major questions by using 11-year longitudinal couple data. First, we examine dyadic interdependencies in level and overall change in social activities in couples. Second, we examine how social activity trajectories are related to individual and spousal resources. We further explore dyadic interdependencies of longitudinal changes in social activities and cognition. Third, we investigate potential gender-based (a) symmetries of the associations examined. Methods The ALSA is a broad biopsychosocial and behavioral panel study. On the advice of the third author (M. Luszcz), we selected variables on an a priori basis as the best available markers of the constructs targeted for investigation. Detailed descriptions of variables and procedures are published elsewhere (Andrews et al., 2002; Anstey et al., 2003; Luszcz, Bryan, & Kent, 1997). Participants and Procedure The baseline ALSA sample (N = 2,087) was stratified by age and sex into four cohorts (70–74 years, 75–79 years, 80–84 years, and 85+ years); men and those over the age of 85 years were oversampled. The sample (55% response rate) was obtained from the South Australian Electoral Roll that identified households with residents over 70 years of age (Hugo, Healy, & Luszcz, 1987). Randomly sampled individuals within these households were invited to volunteer for ALSA; spouses over age 65 were also invited to participate. We included all couples with valid data on the variables of interest (N = 565 or 1,130 people). Average marriage duration was 46.33 years (SD = 11.23; range = 0–70 years); number of children was 2.7 (SD = 1.63). Husbands were about 3 years older than their wives, at 77.7 versus 74.3 years; F (1, 1128) = 101.1, p <.001. Responding to a 5-point Likert-scale, couples were “extremely” to “very” satisfied with their marriage (M = 1.69, SD = 0.78) and family life (M = 1.97, SD = 0.81), but slightly less satisfied with their friendships (M = 2.37, SD = 0.76). As compared with the remaining 957 participants from the total ALSA sample, participants in the ALSA couple subsample did not differ in number of social activities, education, or health constraints (all p >.10). However, they were younger [M = 76.66, SD = 5.87 vs M = 80.72, SD = 6.70; F (1, 2085) = 295.6, p <.001], reported fewer depressive symptoms [M = 7.27, SD = 6.90 vs M = 9.36, SD = 7.85; F (1, 1991) = 40.0, p <.001], and performed better on the Digit Symbol Substitution task [M = 30.59, SD = 10.89 vs M = 27.20, SD = 11.20; F (1, 1241) = 28.6, p <.001]. Data were collected in two 1.5- to 2-hour sessions at participants' home (98.4% private household or 1.6% institution). Session 1 assessed demographics, social activities, depression, and health. The Session 2 functional assessment occurred approximately 2 weeks later and included tests of perceptual speed for a subsample of, on average, 70% of Session 1 participants (see Table 1). We use four waves of longitudinal data spanning 11 years: baseline in 1992–1993 (Time 1 or T1; n = 1,130), 1994–1995 (Time 3 or T3; n = 929), 2000–2001 (Time 6 or T6; n = 487), and 2003–2004 (Time 7 or T7; n = 315). Assessments at Times 2, 4, and 5 did not provide data on the targeted domains; we excluded them. On average, T3 occurred 1.53 years (SD = 0.50), T6 occurred 7.43 years (SD = 0.50), and T7 occurred 10.57 years (SD = 0.50), respectively, after T1. By T3, T6, and T7, 9.9%, 38.4%, and 51.7%, respectively, of participants had become deceased (this corresponds closely to the incidence of widowhood, i.e., 10%, 37%, and 52%); whereas in the residual sample of 987 participants, there were 12.9%, 52.9%, and 65.1% of the sample who became deceased across the same period. Measures Social activities We selected four items from the Adelaide Activity Profile (Clark & Bond, 1995) that specifically assessed social-life aspects in which both partners can potentially participate. These concerned the frequency of (a) having invited other people to one's home, (b) making phone calls to friends or family, (c) attending social activities at a center such as a club, a church, or a community center, and (d) participating in outdoor social activities. Participants responded by means of a 4-point Likert-scale, with higher scores indicating more activity. A structural model at T1 indicated that all factor loadings for husbands and wives were above.60 (M =.79), suggesting a reliable social activity factor. Covariates We considered education, T1 health constraints and depression, and processing speed at T1 and over time. We dichotomized years of education by contrasting participants who left school at age 15 plus (higher education) with those who left under age 15. Health constraints were self-reports of the number of chronic medical conditions from a comprehensive list of 61 (e.g., stroke, diabetes, arthritis). For depressive symptoms, we used the Center for Epidemiological Studies–Depression scale (Radloff, 1977); 20 items asked participants how often over the past week they felt symptoms such as lack of energy or sad feelings. For a parsimonious indicator of cognition, we selected perceptual speed because it is highly reliable and sensitive to change throughout adulthood (Anstey et al., 2003; Salthouse & Madden, in press), and, as a cognitive primitive (Luszcz & Bryan, 1999), is conceptually closer to a resource than other abilities. In contrast, decline in verbal ability, for example, occurs very late in life (Anstey et al., 2003), when it may be too late to engage in compensatory behaviors (Ghisletta et al., 2006), and memory change is fundamentally limited by change in perceptual speed. We assessed perceptual speed by using the Digit Symbol Substitution subscale of the revised Wechsler Adult Intelligence Scale (Wechsler, 1981; for details, see Luszcz et al., 1997), at T1, T3, T6, and T7. Participants substituted symbols corresponding to the numbers 1–9 as rapidly as possible into a randomly ordered array of 93 digits. Symbols were presented throughout the task. We used the number of correct substitutions in 90 seconds. Data Preparation We standardized social activities and perceptual speed to a T metric (M = 50; SD = 10), with the T1 ALSA couples providing the reference. This transformation ensured a common metric while maintaining the psychometric properties of the scores and the longitudinal changes in means and variances. Missing data for social activities amounted to 7% (206 out of 2,861 attainable data points over four occasions). Missing data for perceptual speed amounted to 6% of Session 2 participants and were primarily due to poor vision.1 No data imputation procedure was applied. The average longitudinal observation interval was 3.68 years (SD = 3.63). Table 1 presents, separately for husbands and wives, the age at assessment as well as means and standard deviations for the variables under study. Participants were tested, on average, in their late seventies and early eighties. Husbands and wives did not differ in education or health problems (p >.10), but wives reported more depressive symptoms than did husbands, F (1, 1084) = 4.5, p <.05. The effects of social activities and perceptual speed were our main focus and are presented in the Results section. Statistical Procedures We applied two-variable and four-variable extensions of the latent growth model over time (McArdle, 1988) that estimates fixed effects (average level and change) and random effects (interindividual differences in level and change). The two-variable growth curve is a straightforward extension of a univariate growth curve. Specifically, we consider the couple as the unit of analysis and model separate growth processes for social activities of husbands and wives. A graphical representation of the model is given in Figure 1. The diagram shows observed variables as squares, latent variables as circles, and the required constant as a triangle as well as fixed model parameters as one-headed arrows and random parameters as two-headed arrows. Unlabeled paths are fixed to 1. The four repeated measures of X (husband) and Y (wife) have three sources: the latent intercept, X0, Y0, with unit loadings; the latent slope, Xs, Ys, with linear loadings (0, 1.5, 7.4, 10.6); and the time-specific residual, ex[t], ey[t]. Intercepts and slopes are estimated at the population level and are allowed to vary and to covary. Time-specific residuals ex[t], ey[t] have a mean of zero, a single variance (σ2ex, σ2ey), and are allowed to covary with each other within occasion (σ2ex,ey). Because of the age-heterogeneous sample, all models reported include husbands' and wives' ages as covariates to adjust otherwise inflated covariances among variables that are due to their common association with chronological age. In Figure 1, covariates are represented by variables Zx and Zy with their own mean (μZx, μZy), variance (σ2Zx, σ2Zy), and covariance (σ2Zx, Zy), as well as regression effects on intercepts and slopes both as actor effects (βZx,X0, βZx,Xs, βZy,Y0, βZy,Ys) and partner effects (βZx,Y0, βZx,Ys, βZy,X0, βZy,Xs). We used the full-information maximum likelihood estimation algorithm and included all data points available, which allowed us to accommodate for unbalanced data structures and incomplete data under the missing-at-random assumption (McArdle, 1994). We estimated the models by using the Mplus program, Version 4 (Muthén & Muthén, 1998–2006). We carried out three sets of analyses. First, we examined level and overall change of social activities among husbands and wives as well as associations among these factors at the couple level. Second, we investigated whether level and overall change in social activities were related to T1 measures of education, perceptual speed, health constraints, and depression, both at the individual and couple level. We did this by including additional Z variables for each covariate in the model shown in Figure 1, allowing for intercorrelations among all covariates. In follow-up analyses, we explored covariations among longitudinal changes in social activities and perceptual speed, both at the individual and couple level. To do so, we extended the two-variable growth model shown in Figure 1 to a four-variable model that also included linear and quadratic slopes of perceptual speed for both partners. Third, we evaluated the (a) symmetry of covariate–growth factor associations between spouses by means of statistically nested model comparisons of models that either freely estimated the respective parameters or set it invariant across spouses. Results Level and Change of Social Activities in Older Couples Results for spousal social activity trajectories (Table 2) indicate that the model fit the data well (e.g., root mean square error of approximation or RMSEA =.023). Consistent with our expectation, fixed effects for the levels indicate that wives reported more activities than did husbands, which was corroborated by a substantial loss of model fit when we set the levels invariant (Δχ2/df = 34.1/1, p <.001). Fixed effects for both slopes were not significantly different from zero, suggesting that social activities did not show much change over time on average. Significant estimates for all random effects among husbands as well as wives revealed, however, that individuals reliably differed from one another in both their activity levels and change over time. We also note the significant and negative level–slope covariance among wives, but not among husbands, indicating that wives who initially reported more social activities tended to show an overall subsequent decline. Of particular interest were the three significant partner effects. Social activities between partners were strongly correlated, and more activities among wives were associated with activity decline for husbands. In addition, associations between the time-specific residuals indicate that, after we accounted for individual activity changes, activities of both partners were still related with one another. In line with our expectations, these findings suggest that social activity trajectories were interrelated not only at the individual level but also at the couple level. Relations of Level and Change in Social Activities to Individual and Spousal Correlates To examine associations between level and change in social activities with individual and spousal correlates (Table 3), we introduced husbands' and wives' T1 measures of education, perceptual speed, health constraints, and depressive symptoms as covariates. Regarding actor effects, husbands' activity levels were negatively associated with husbands' age and health constraints and positively with husbands' perceptual speed. Similarly, wives' activity levels related negatively to wives' age and depression and positively to wives' education and perceptual speed. The only significant correlate of change in activity was husbands' depression, and the negative sign indicates that more baseline depression related to more activity decline. Again, partner effects were of primary interest. The only statistically significant association at the partner level was that depression of wives related positively to husbands' activity change, suggesting that husbands of depressed wives showed an increase in activities. In follow-up analyses, we examined a four-variable growth model of longitudinal changes in social activities and perceptual speed among husbands and wives.2 This model again provided reasonable fit to the data (e.g., RMSEA =.040). Fixed effects for social activities were very similar to the aforementioned two-variable model (e.g., wives reported more activities than husbands); for perceptual speed, we found no spousal differences at any time, but both husbands and wives showed significant decline over time. The fixed effects from this model are shown in Figure 2. <--CO?1-->What is most important for our question is that both husbands and wives showed significant interindividual differences in their linear change components on social activities and perceptual speed (whereas no interindividual differences were found in the quadratic change component on perceptual speed). This allowed for a close inspection of the covariances across level and linear slope factors at the individual and couple level (Table 4). These intercorrelations did not provide evidence that activity changes either at the individual or couple level were accompanied by cognitive changes, but showed three significant actor effects. Analogous to the two-variable model, level–slope intercorrelations for social activities were negative among wives, but not among husbands. Linking domains of functioning, we found that levels of activities and perceptual speed were associated among both husbands and wives. We also found four significant partner effects: Levels of activities and perceptual speed were positively related, as were husbands' activity levels and wives' level of perceptual speed. In addition, level of activity for wives was strongly associated with activity decline for husbands. Finally, examining asymmetrical effects in the four-variable model revealed that activity levels showed stronger associations among both husbands and wives (r =.59) than did levels of perceptual speed (r =.13; Δχ2/df = 5.4/1, p <.01). This suggests that mutual engagement in activity is stronger than entrainment of cognitive capacity. In sum, our results suggest that social activity trajectories were not only associated with one's own resources but also with the resources of one's partner.3 Asymmetry of Associations Between Husbands and Wives Finally, we conducted a series of statistically nested model comparisons to evaluate whether social activity trajectories show differential associations among husbands and wives. To begin with, the nominal difference in level–slope associations among spouses (−.26 vs −.41; see Table 2) was not statistically different, given that a model with these parameters set invariant across spouses did not result in loss of model fit (Δχ2/df = 0.2/1, p >.10). Regarding asymmetrical actor effects in covariate–growth factor associations (see Table 3), we found no effects for age, education, perceptual speed, and health (all p >.10). However, level and slope effects of depression differed between spouses, suggesting that depression was associated with fewer activities among wives but not among husbands (Δχ2/df = 6.5/1, p <.01); and in husbands, but not wives, depression related to activity decline (Δχ2/df = 5.7/1, p <.01). Asymmetrical partner effects indicated that the effects of wives' depression on husbands' activity slope were not mirrored in the effects of husbands' depression on wives' activity slope (Δχ2/df = 3.2/1, p <.05).4 In sum, affective resources of husbands and wives are differentially associated with social activity trajectories of the spouse. Discussion On the basis of 11-year longitudinal couple data, we demonstrated that level of, and change in, social activities are closely interrelated among elderly spouses and that wives perform more social activities than do husbands. We also showed that spousal social activity trajectories were related to both individual cognitive, physical, and affective resources, and to spousal resources. Specifically, significant partner effects emerged linking depression with social activity changes and levels of cognition and social activities. Longitudinally, significant actor effects concerning cognition–social activity associations emerged. Partner effects were stronger in social activities than perceptual speed. Finally, several social activity–depression associations were gender specific. Level and Change of Social Activities in Older Couples Our findings concerning dyadic interdependencies in social activities show that developmental trajectories of spouses are closely intertwined. As expected (Antonucci, 2001; Moen, 2001), wives reported more social activities than husbands. We can only speculate as to the mechanisms linking social activity trajectories in couples. In keeping with socioemotional selectivity theory, our findings may reflect an increased focus on emotionally meaningful relationships in old age as manifested in high mutual engagement in social activities among spouses (Carstensen, 1995). More work is needed, however, to differentiate whether such interdependencies are due to similar background characteristics, overlapping networks, or a reciprocal activity involvement (Milardo & Helms-Erikson, 2000). Relations of Level and Change in Social Activities to Individual and Spousal Resources Our findings that social activity trajectories are related to individual cognitive, physical, and affective resources are in line with past research on unrelated individuals. Within our sample, poorer physical functioning at baseline was related to fewer social activities (Zunzunegui et al., 2005), and spouses with more educational and cognitive resources reported more social activities (Hultsch et al., 1999). Although we found correlation coefficients of different sizes across spouses (e.g., linking physical functioning and social activities), we note that statistically nested model comparisons indicated that these relationships were similar among husbands and wives. Over and above these actor affects, we also found partner effects linking depression and social activities. These findings are discussed in the context of gender-specific associations. We also explored longitudinal associations between social activities and perceptual speed. In line with previous research, participants living socially active lifestyles showed better cognitive performance than did socially inactive participants (Ghisletta et al., 2006; Lövdén et al., 2005). We also found partner effects, and these were much stronger in social activities relative to cognition, thereby highlighting the important role of spouses for social lifestyles. This difference in relative strength of dyadic associations illustrates the increased need for cultural support at advanced ages (P. B. Baltes, 1997). Culture is likely to be more efficacious in the social domain than in the cognitive domain, because spouses can provide input that directly helps their partner to remain socially engaged. Perceptual speed is a fluid ability and more reliant on biological propensities than on cultural resources. Hence, older adults are likely to encounter loss-based cognitive limitations that may be harder for marriage partners to influence. When conducting analyses of cognition and social activities including both partners' age, education, health, and depression as covariates, we found that only partner effects in social activities remained. In sum, our findings suggest that taking into account long-term relationships as an important contextual factor may help researchers to better understand the processes underlying successful aging. Asymmetry of Associations Between Husbands and Wives Our findings of gender-specific actor and partner effects in depression–social activity associations illustrate that husbands and wives are both vulnerable to the effects of depression on social activities. However, the underlying processes may arise at different times. Whereas for wives the negative depression–social activity associations may result from processes that occurred earlier in life, husbands' negative social activity slopes suggest that higher depression scores are followed by social activity decline over the study period. This interpretation is in line with research showing that women report more depression than men, but the relative difference is larger in midlife than in old age (Anstey, van Sanden, Cox, & Luszcz, in press; Barefoot, Mortensen, Helms, Avlund, & Schroll, 2001). Hence, the study period may coincide with a point in time when husbands' depression-related processes unfold, whereas those processes for wives have already emerged. Over and above these actor effects, we also found evidence that wives' depressive symptoms were associated with an increase in husbands' social activities. These results can be interpreted in the context of substantial positive correlations between spousal depression scores (Townsend et al., 2001) and suggest that husbands assume responsibility for social activities when their wives are emotionally challenged. These findings demonstrate that associations of depression and social activity trajectories in elderly couples are best studied in dyads, rather than unrelated individuals. Strengths and Limitations We examined a unique longitudinal couple data set that allowed us to explore actor and partner effects in the relationship between social activities and cognitive, physical, and affective resources. Despite the longitudinal data, most findings refer to mean level results. Like other studies (Ghisletta et al., 2006; Hultsch et al., 1999; but see Lövdén et al., 2005), our study indicated that social activities showed relatively little average change over time (but reliable individual differences therein), which may impose limits on detecting relationships of longitudinal changes to other individual and partner variables. One promising route for future research could thus be to explore whether other (social) activity indicators that are more sensitive to normative change (e.g., loss of bridge partners or closure of a favorite social club) reveal different patterns of longitudinal associations. In addition, our focus on only genuinely social activities, such as inviting friends, contrasts with previous reports from the ALSA (Luszcz et al., 1997; Newson & Kemps, 2005) and other studies (Ghisletta et al., 2006; Hultsch et al., 1999; Lövdén et al., 2005; Pushkar et al., 1999), which used much broader activity measures covering a wide range of physical, leisure, and work activities. We selected social activities that both partners could potentially participate in, but we cannot differentiate whether social activities were performed jointly or individually. Follow-up analyses found that accounting for variability in marriage duration did not change our results, but we could not determine how other relationship characteristics such as mutual affection shape social activities. Together, our results highlight the necessity to distinguish the nature of “activity” in order to accurately understand interrelationships between activities and resources. We acknowledge that our growth curve models produced estimates of average within-person change independent of whether or not an individual (or couple) stayed in the sample over time (i.e., missing-at-random assumption). Nonetheless, we corroborated our main results in a restricted sample of longitudinal participants.<5 Furthermore, these models consider concurrent associations between level and overall change on a given set of variables but not their causal ordering, such as whether one spouse's social activities affect subsequent change in the other spouse's perceptual speed (for an overview of alternative approaches, see Ghisletta et al., 2006). Future research may apply “dynamic” modeling techniques such as bivariate dual change score models (McArdle & Hamagami, 2001) to examine potential lead–lag associations. Conclusions We showed that social activity trajectories are interrelated in elderly married couples and associated with not only individual, but also spousal, characteristics. This attests to the importance of social relationships for late-life development and some domain specificity in underlying dynamics. Together, our results suggest that although physical and cognitive resources provide an environment that is conducive to social activities, the affective resources of one marital partner are crucial to variations in both level and change in social activities of the other and hence to that of the couple as a unit. Our findings are heuristic in posing new questions about the dynamics of long-term spousal relationships for successful aging. For example, is the codependence in social activities due to a priori similarities or does it develop over time (Luo & Klohnen, 2005)? Does the codependence originate in shared environmental characteristics or is it similar across different environmental contexts? Is the relationship between social activities and resources driven by social activity performance or by social embeddedness? Future research may substantiate these findings by going beyond individual differences variables and addressing the underlying dyadic processes. This may get at issues concerning social reserve capacities and vulnerabilities in late life. Decision Editor: Thomas M. Hess, PhD 1Because of the specific study design, sample attrition between Session 1 and Session 2 over the 2-week period was substantial. As discussed extensively by >nstey and Luszcz (2002), attrition primarily reflected self-selection as a result of poor health or physical or cognitive functioning. Using the sample in Session 1 as a reference, we found that missing data for the Digit Symbol task amounted to 31%. 2Results from a preliminary two-variable latent growth model for perceptual speed revealed that both husbands and wives showed significant decline over time (husbands, −0.74, SE = 0.09; wives, −0.55, SE = 0.09; both p < .001) as well as interindividual differences in decline (husbands, 0.25, SE = 0.08, p < .01; wives, 0.21, SE = 0.09, p < .05). 3To examine whether or not our main results concerning partner effects for social activity trajectories were contingent upon physical activities, we included a covariate in our models that comprised a unit-weighted composite of the following items: doing heavy housework, light and heavy gardening, or household or car maintenance; engaging in sports; and walking outside for more than 15 minutes. These analyses revealed substantively identical results to those reported here. In addition, we examined the relationship between physical activities and spousal cognitive, physical, and affective resources. Results showed that the pattern of relationships concerning physical activities differs from the reported findings concerning social activities. In addition, we found no association between physical activity performances of both spouses. Hence, our findings appear to be specific to social activities and suggest that social activities do not represent a proxy for activities in general. 4When introducing the covariates, we found that the reported individual (actor) and spousal (partner) covariances among level and change factors of social activities remained virtually the same. 5For our major findings at the couple level (husband's activity decline relates to wives' level of activities and depression at T1; correlated residuals), analyses of a restricted sample of those who remained in the study up to T6 or longer (n = 164) revealed substantively identical results to those reported here. Figure 1. Graphical representation of the two-variable latent growth curve model (McArdle, 1988) as applied in the present study. Observed variables are represented by squares, latent variables by circles, regression weights by one-headed arrows, and variances and covariances by two-headed arrows; the triangle represents a constant indicating means and intercepts. All unlabeled paths are set to 1. There are four repeated measures of social activities for husbands (X) and wives (Y), and covariates are represented by variables Zx (husbands) and Zy (wives) Figure 2. Model-implied means over time in study from a four-variable latent growth curve model of social activities (solid lines) and perceptual speed (dashed lines) among husbands and wives. We standardized scores to the T metric by using the Time 1 Australian Longitudinal Study of Ageing couple sample (N = 1,130), M = 50, SD = 10 Table 1. Age at Assessment and Descriptive Statistics for Measures Used in the Present Study. Husbands Wives Measure n Age M SD n Age M SD Social activities T1 563 77.71 48.37 9.88 561 74.30 51.63 9.86 T3 415 78.79 49.11 8.89 452 75.88 52.89 9.19 T6 164 83.38 50.00 9.89 242 80.70 52.87 9.20 T7 96 85.08 52.50 9.71 170 82.85 55.01 9.72 Perceptual speed T1 371 77.23 49.40 9.79 357 74.03 50.64 10.19 T3 339 78.45 49.69 10.06 350 75.55 52.43 10.03 T6 128 82.60 48.88 9.70 180 79.92 51.45 9.20 T7 85 84.94 47.17 9.66 158 82.60 50.01 9.66 Health constraints T1 508 77.44 5.43 3.00 508 74.11 5.33 3.13 Depression T1 489 77.39 6.83 6.51 497 74.05 7.71 7.25 Percent Percent High education T1 242 77.85 43% 267 74.35 47% Note: T1, T3, T6, and T7 = Times 1, 3, 6, and 7, respectively. We standardized scores of social activities and perceptual speed to the T metric by using the T1 Australian Longitudinal Study of Ageing couple sample (N = 1,130), M = 50, SD = 10. We collected data on perceptual speed in Session 2, which was attended by two thirds to three fourths of Session 1 participants. The average age of Session 1 and Session 2 participants is highly similar. High education signifies that the participants were 15 years of age or older when they left school. For depression, scores >16 correspond to the cutoff for depression according to the Center for Epidemiological Studies–Depression scale. Table 2. Estimates of a Two-Variable Latent Growth Model for Social Activities of Husbands and Wives. Actor Effect Husbands Wives Partner Effect Social Activities Estimate SE Estimate SE Estimate Fixed effects Level μ0 49.12*** 0.44 51.54*** 0.45 Slope μs 0.01 0.10 0.03 0.08 Random effects Variance of level σ20 39.02*** 4.71 44.14*** 4.69 Variance of slope σ2s 0.28* 0.13 0.21* 0.09 Correlation level σ20, slope σ2s −0.26 −.41* Residual variance σ2e 47.33*** 3.01 46.35*** 2.62 Correlations Husband level σ20 − Wife level σ20 .60*** Husband slope σ2s − Wife slope σ2s .14 Husband level σ20 − Wife slope σ2s −.23 Wife level σ20 − Husband slope σ2s −.43* Residual variance σ2e Husband − Wife .26*** Notes: All noncorrelation estimates are unstandardized. The significance tests assigned to the correlations refer to the corresponding covariances. Model includes age of husbands and wives as covariates. We standardized social activities to the T metric by using the Time 1 Australian Longitudinal Study of Ageing couple sample (N = 1,130), M = 50, SD = 10. Model fit statistics: χ2 (df) = 45.0 (35); Comparative Fit Index, CFI = 0.988; and root mean square error of approximation, RMSEA = 0.023. * p <.05, **p <.01, ***p <.001. Table 3. Estimates of a Two-Variable Latent Growth Model for Social Activities of Husbands and Wives, Including the Effects of Individual Difference Variables Assessed at Time 1. Husbands Wives Parameter or Covariate Estimate SE Actor Effect Partner Effect Estimate SE Actor Effect Partner Effect Parameter Level μ0 48.51*** 0.59 51.44*** 0.60 Slope μs −0.10 0.13 −0.07 0.11 Variance of level σ20 35.45*** 4.52 38.69*** 4.43 Variance of slope σ2s 0.12 0.12 0.19 0.09 Residual variance σ2e 47.51*** 3.00 46.06*** 2.60 Covariate Age −.23** .01 −.15* −.06 Age × Slope −.30 −.22 −.16 −.12 Education .04 .08 .14* −.09 Education × Slope −.18 .22 −.15 .24 Perceptual speeda .14 .06 .21** −.02 Perceptual Speed × Slope −.02 .04 −.16 −.17 Health constraints −.18** −.07 −.06 −.06 Health Constraints × Slope .11 .27 −.11 .11 Depression .11 −.04 −.16* .13 Depression × Slope −.50** .39* .07 −.05 Notes: All noncorrelation estimates are unstandardized. The significance tests assigned to the correlations refer to the corresponding covariances. We standardized social activities to the T metric by using the Time 1 Australian Longitudinal Study of Ageing couple sample (N = 1,130), M = 50, SD = 10. Age, health constraints, and depression were centered at zero. Education: 0 = 0–14 years of age when the participant left school; 1 = 15 years of age or older when the participant left school. Model fit statistics: χ2 (df) = 84.1 (67); Comparative Fit Index, CFI = 0.981; root mean square error of approximation, RMSEA = 0.021. Variance in social activities accounted for level of husbands (R2 =.170), linear slope of husbands (R2 =.601), level of wives (R2 =.190), and linear slope of wives (R2 =.131). a For perceptual speed, p =.06. * p <.05, **p <.01, ***p <.001. Table 4. Intercorrelations Among Level and Slope Factors of Social Activities and Perceptual Speed Among Husbands and Wives (Obtained From a Four-Variable Latent Growth Model). 1 2 3 4 5 6 7 Husbands Actor effect 1. Activity level σ20 — 2. Activity slope σ2s −.25 — 3. Speed level σ20 .16** −.03 — 4. Speed slope σ2s −.06 .39 −.10 — Wives Partner effect Actor effect 5. Activity level σ20 .59*** −.43** .00 .07 — 6. Activity slope σ2s −.22 .15 −.07 −.16 −.39* — 7. Speed level σ20 .16* −.07 .13* −.07 .28*** −.14 — 8. Speed slope σ2s .17 −.09 −.02 −.11 .01 .36 −.22 Notes: Model includes age of husbands and wives as covariates. The significance tests assigned to the correlations refer to the corresponding covariances. Model fit statistics: χ2 (df) = 226.5 (116); Comparative Fit Index, CFI = 0.940; root mean square error of approximation, RMSEA = 0.041. * p <.05, **p <.01, ***p <.001. Preparation for this manuscript was supported by a Research Fellowship awarded by the German Research Foundation (DFG) to Christiane Hoppmann. Large parts of this article were prepared while Denis Gerstorf was at the Department of Psychology, University of Virginia on a Research Fellowship awarded by the German Research Foundation (DFG). The ALSA was conducted by the Centre for Ageing Studies, Flinders University, Adelaide, Australia. ALSA was financially supported by grants from the U.S. National Institute on Aging (Grant AG08523), the South Australian Health Commission, Australian Rotary Health Research Fund, National Health and Medical Research Council (Grant 229922), the Australian Research Council, the South Australian Department of Human Services, Flinders Medical Centre Foundation, Elderly Citizens Homes PL (ECH), and the Flinders Research Grants Scheme. We thank the ALSA participants, who have given their time over many years, and without whom the present study would not have been possible. We dedicate this article to the memory of the late Professor Gary Andrews (May 2, 1938–May 18, 2006), Director of the Centre for Ageing Studies and founding Chief Investigator. References Andrews, G., Clark, M., Luszcz, M. (2002). Successful aging in the Australian Longitudinal Study of Ageing: Appling the MacArthur Model cross-nationally. Journal of Social Issues, 58,749-765. Anstey, K. J., Hofer, S. M., Luszcz, M. A. (2003). A latent growth curve analysis of late-life sensory and cognitive function over 8 years: Evidence for specific and common factors underlying change. Psychology and Aging, 18,714-726. Anstey, K. J., Luszcz, M. A. (2002). Selective non-response to clinical assessment in the Longitudinal Study of Ageing: Implications for estimating population levels of cognitive function and dementia. International Journal of Geriatric Psychiatry, 17,704-709. Anstey, K. J., van Sanden, C., Cox, K. S., Luszcz, M. A., in press Prevalence and risk factors for depression in a longitudinal, population-based study including individuals in the community and residential care. Australian Journal of Geriatric Psychiatry.. Antonucci, T. C. (1994). A life-span view of women's social relations. In B. F. Turner & L. Troll (Eds.), Women growing older: Psychological perspectives (pp. 239–269). Thousand Oaks, CA: Sage. Antonucci, T. C. (2001). Social relations: An examination of social networks, social support, and sense of control. In J. E. Birren (Ed.), Handbook of the psychology of aging (pp. 427–453). San Diego, CA: Academic Press. Antonucci, T. C., Akiyama, H. (1991). Convoys of social support: Generational issues. Marriage & Family Review, 16,103-123. Baltes, P. B. (1997). On the incomplete architecture of human ontogeny: Selection, optimization, and compensation as foundation of developmental theory. American Psychologist, 52,366-380. Baltes, M. M., Carstensen, L. L. (1998). Social-psychological theories and their applications to aging: From individual to collective. In V. L. Bengtson & K. W. Schaie (Eds.), Handbook of theories of aging (pp. 209–226). New York: Springer. Baltes, P. B., Lindenberger, U. (1997). Emergence of a powerful connection between sensory and cognitive functions across the adult life span: A new window to the study of cognitive aging? Psychology and Aging, 12,12-21. Baltes, P. B., Staudinger, U. M. (1996). Interactive minds in a life-span prologue. In P. B. Baltes & U. M. Staudinger (Eds.), Interactive minds: Life-span perspectives on the social foundation of cognition (pp. 1–32). New York: Cambridge University Press. Barefoot, J. C., Mortensen, E. L., Helms, M. J., Avlund, K., Schroll, M. (2001). A longitudinal study of gender differences in depressive symptoms from age 50 to 80. Psychology and Aging, 16,342-345. Bath, P. A., Deeg, D. (2005). Social engagement and health outcomes among older people: Introduction to a special section. European Journal of Ageing, 2,24-30. Bryk, A. S., Raudenbush, S. W. (1992). Hierarchical linear models in social and behavioral research: Applications and data analysis methods. Newbury Park, CA: Sage. Cantor, N., Sanderson, C. A. (1999). Life task participation and well-being: The importance of taking part in daily life. In D. Kahneman & E. Diener (Eds.), Well-being: The foundations of hedonic psychology (pp. 230–243). New York: Russell Sage Foundation. Carstensen, L. L. (1995). Evidence for a life-span theory of socioemotional selectivity. Current Directions in Psychological Science, 4,151-156. Carstensen, L. L., Gottman, J. M., Levenson, R. W. (1995). Emotional behavior in long-term marriage. Psychology & Aging, 10,140-149. Clark, M. S., Bond, M. J. (1995). The Adelaide Activities Profile: A measure of the lifestyle activities of elderly people. Aging Clinical and Experimental Research, 7,174-184. Clark, M. S., Bond, M. J. (2000). The effect on lifestyle activities of caring for a person with dementia. Psychology, Health & Medicine, 5,13-27. Cook, W. L., Kenny, D. A. (2005). The Actor–Partner Interdependence Model: A model of bidirectional effects in developmental studies. International Journal of Behavioral Development, 29,101-109. Coombs, R. H. (1991). Marital status and personal well-being: A literature review. Family Relations, 40,97-102. Dixon, R. A. (1999). Exploring cognition in interactive situations: The aging of N + 1 minds. In T. M. Hess & F. Blanchard-Fields (Eds.), Social cognition and aging (pp. 267–290). San Diego, CA: Academic Press. Fratiglioni, L., Paillard-Borg, S., Winblad, B. (2004). An active and socially integrated lifestyle in late life might protect against dementia. The Lancet, 3,343-353. Ghisletta, P., Bickel, J.-F., Lövdén, M. (2006). Does activity engagement protect against cognitive decline in old age? Methodological and analytical considerations. Journal of Gerontology: Psychological Sciences, 61B,P253-P261. Gruber-Baldini, A. L., Schaie, K. W., Willis, S. L. (1995). Similarity in married couples: A longitudinal study of mental abilities and rigidity–flexibility. Journal of Personality and Social Psychology, 69,191-203. Hagedoorn, M., Yperen, N. V., Coyne, J. C., Jaarsfeld, C. H. M., Ranchor, A. V., Sonderen, E., et al. (2006). Does marriage protect older people from distress? The role of equity and recency of bereavement. Psychology and Aging, 21,611-620. Herzog, A. R., Ofstedal, M. B., Wheeler. L. M. (2002). Social engagement and its relationship to health. Clinics in Geriatric Medicine, 18,593-609. Hertzog, C., Hultsch, D. F., Dixon, R. A. (1999). On the problem of detecting effects of lifestyle on cognitive change in adulthood: Reply to Pushkar et al. (1999). Psychology and Aging, 14,528-534. Hugo, G., Healy, J., Luszcz, M. (1987). Surveying the aged in Australia. Australian Journal of Ageing, 6,14-19. Hultsch, D. F., Hertzog, C., Small, B. J., Dixon, R. A. (1999). Use it or lose it: Engaged lifestyle as a buffer of cognitive decline in aging? Psychology and Aging, 14,245-263. Jang, Y., Mortimer, J. A., Haley, W. E., Small, B. J., Chisolm, T. E. H., Graves, A. B. (2003). The role of vision and hearing in physical, social, and emotional functioning among older adults. Research on Aging, 25,172-191. Kahn, R. L., Antonucci, T. C. (1980). Convoys over the life course: Attachment, roles, and social support. In P. B. Baltes & O. G. Brim (Eds.), Life-span development and behavior (Vol. 3, pp. 254–283). New York: Academic Press. Lang, F. R. (2001). Regulation of social relationships in later adulthood. Journal of Gerontology: Psychological Sciences, 56B,P321-P326. Lang, F. R., Baltes, M. M. (1997). Being with people and being alone in late life: Costs and benefits for everyday functioning. International Journal of Behavioral Development, 21,729-746. Lövdén, M., Ghisletta, P., Lindenberger, U. (2005). Social participation attenuates decline in perceptual speed in old and very old age. Psychology and Aging, 20,423-434. Luo, S., Klohnen, E. C. (2005). Assortative mating and marital quality in newlyweds: A couple-centered approach. Journal of Personality and Social Psychology, 88,304-326. Luszcz, M. A. (1998). A longitudinal study of psychological changes in cognition and self in late life. The Australian Developmental and Educational Psychologist, 15,39-61. Luszcz, M. A., Bryan, J. (1999). Toward understanding age-related memory loss in late adulthood. Gerontology, 45,2-9. Luszcz, M., Bryan, J., Kent, P. (1997). Predicting memory performance of very old men and women: Contributions from age, depression, activity, cognitive ability and perceptual speed. Psychology and Aging, 12,340-351. McArdle, J. J. (1988). Dynamic but structural equation modeling of repeated measures data. In J. R. Nesselroade & R. B. Cattell (Eds.), Handbook of multivariate experimental psychology (Vol. 2, pp. 561–614). New York: Plenum Press. McArdle, J. J. (1994). Structural factor analysis experiments with incomplete data. Multivariate Behavioral Research, 29,409-454. McArdle, J. J., Hamagami, F. (2001). Latent difference score structural models for linear dynamic analyses with incomplete longitudinal data. In L. M. Collins & A. G. Sayer (Eds.), New methods for the analysis of change (pp. 137–176). Washington, DC: American Psychological Association. Meegan, S. P., Berg, C. A. (2002). Contexts, functions, forms, and processes of collaborative everyday problem solving in older adulthood. International Journal of Behavioral Development, 26,6-15. Milardo, R. M., Helms-Erikson, H. (2000). Network overlap and third-party influence in close relationships. In C. Hendrick & S. S. Hendrick (Eds.), Close relationships: A sourcebook (pp. 33–45). Thousand Oaks, CA: Sage. Moen, P. (2001). The gendered life course. In R. H. Binstock & L. K. George (Eds.), Handbook of aging and the social sciences (Vol. 5, pp. 179–196). San Diego, CA: Academic Press. Muthén, L. K., Muthén, B. O. (1998). Mplus user's guide, the comprehensive modeling program for applied researchers (4th ed.). Los Angeles, CA: Author. Newson, R., Kemps, E. (2005). General lifestyle activities as a predictor of current cognition and cognitive change in older adults. Journal of Gerontology: Psychological Sciences, 60B,P113-P120. Pushkar, D., Etezadi, J., Andres, D., Arbuckle, T., Schwartzman, A. E., Chaikelson, J. (1999). Models of intelligence in late life: Comment on Hultsch et al. (1999). Psychology and Aging, 14,520-527. Radloff, L. W. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1,385-401. Rook, K. S. (1998). Investigating the positive and negative sides of personal relationships: Through a lens darkly? In B. H. Spitzberg & W. R. Cupach (Eds.), The dark side of close relationships (pp. 369–393). Mahwah, NJ: Erlbaum. Rowe, J. W., Kahn, R. L. (1997). Successful aging. The Gerontologist, 37,433-440. Ryff, C. D., Singer, B. (2000). Interpersonal flourishing: A positive health agenda for the new millennium. Personality and Social Psychology Review, 4,30-44. Salthouse, T. A., Madden, D. J., in press Information processing speed and aging. In J. DeLuca & J. Kalmar (Eds.), Information processing speed in clinical populations. New York: Psychology Press. Schwarzer, R., Leppin, A. (1989). Social support and health: Meta-analysis. Psychology and Health, 3,1-15. Seeman, T. E., McEwen, B. S. (1996). Impact of social environment characteristics on neuroendocrine regulation. Psychosomatic Medicine, 53,459-471. Townsend, A. L., Miller, B., Guo, S. (2001). Depressive symptomatology in middle-aged and older married couples: A dyadic analysis. Journal of Gerontology: Social Sciences, 56B,S352-S364. Tremont, G., Davis, J. D., Bishop, D. S. (2006). Unique contribution of family functioning in caregivers of patients with mild to moderate dementia. Dementia and Geriatric Cognitive Disorders, 21,170-174. Wechsler, D. (1981). Wechsler Adult Intelligence Scale—Revised manual. New York: Psychological Corporation. Zunzunegui, M. V., Rodriguez-Laso, A., Otero, A., Pluijm, S. M. F., Nikula, S., Blumstein, T., Jylhae, M, Minicuci, N., Deeg, D. J. H. (2005). Disability and social ties: Comparative findings of the CLESA study. European Journal of Ageing, 2,40-47.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B55C433ABD4232EAEDE4027FD5FA0733BAF644A.txt b/test/dataset/in/resources/corpus/Clean_0B55C433ABD4232EAEDE4027FD5FA0733BAF644A.txt
deleted file mode 100644
index 2ed1804..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B55C433ABD4232EAEDE4027FD5FA0733BAF644A.txt
+++ /dev/null
@@ -1 +0,0 @@
-eltjeltjELT Journal1477-45260951-0893Oxford University Press10.1093/elt/ccn068ArticlesThe collaborative development of teacher training skillsStillwellChristopherChristopher Stillwell has conducted workshops on teacher observation and collaborative professional development throughout Asia and the USA, and has contributed several chapters to TESOL's Classroom Practices series. Currently a senior lecturer and research coordinator at Kanda University of International Studies in Chiba, Japan, he has also worked as a teacher trainer at Teachers College Columbia University, where he received his MAEmail: stillwellc@aol.comFinal revised version received June 200810200926112008634353362© The Author 2008. Published by Oxford University Press; all rights reserved.2009This paper describes ‘mentor development’, a means of collaborative professional development through peer observation that was initiated by the author with 18 peers, all native English speaker EFL teachers at Kanda University of International Studies in Chiba, Japan. It shows how such a programme allows teachers to learn from one another through classroom observations and peer mentoring, where observers practise teacher-educator skills by taking on the role of ‘mentor’ in post-observation conferences. A third colleague attends the post-observation conference with the aim of helping both the mentor and observed teacher reflect on and learn from their interaction during the conference, and to explore the implications these discoveries may have for effective teaching and mentoring.Professional development through peer observationPeer observation is a powerful means by which language teachers can become aware of a broad range of possibilities for conducting classes effectively. While many language teaching professionals will already have experienced peer observation in their training, a recent review of the state-of-the-art in ELT professional development indicates that the practice continues beyond graduate school as an important component of such undertakings as lesson study, team teaching, and peer coaching (Mann 2005). Of these three, peer coaching makes perhaps the greatest use of peer observation as colleagues visit one another's classrooms to help refine practices, gather information related to persistent problems, or get and give feedback on the implementation of new teaching methods (Benedetti 1997).A great deal of the literature on peer coaching, and teacher observation in general, focuses on the post-observation conversation, going to great lengths to warn participants of the risks involved with providing evaluative feedback (Bailey, Curtis, and Nunan 2001; Richards and Farrell 2005). In ‘An etiquette for nonsupervisory observation of L2 classrooms’, Murphy (1992: 225) includes the statement that ‘even well-intentioned feedback to classroom teachers often misfires’. For peer coaching, Showers and Joyce (1996: 15) therefore state that it is ‘necessary and important to omit verbal feedback as a coaching component’. Sheal (1989: 93) notes that feedback issues extend to ‘official’ observations by supervisors and administrators as well, for much observation ‘… is unsystematic and subjective. Administrators and teachers generally have not been trained in observation … Consequently they tend to use themselves as a standard, and they observe impressionistically’.Still, many of these same writers concede that teachers typically ask their observers for feedback, and Cosh (1999) notes that teachers are unlikely to get the full reflective value of being observed if they do not receive any comments. Therefore, it seems unrealistic to expect participants in peer observation to refrain from sharing their opinions on what they have seen. If, on the contrary, the opportunity to exchange insights is embraced as a chance to develop skill and experience with feedback, participants can gain heightened awareness of the power of their positive and negative comments, finding optimal ways to offer guidance that inspires rather than disheartens. This article describes mentor development, an approach to peer observation that taps into teachers’ ability to teach themselves, as well as their potential to be valuable resources to one another, by providing frameworks through which colleagues can work together to train themselves in the best practices of observation and mentoring.Mentor developmentThe inspiration for mentor development came from a desire to help teachers enhance their skills and potentially make a transition to teacher training by providing opportunities for colleagues to practise training techniques with, and on, one another. The first participants began by taking part in a number of reading discussions on teacher observation literature, coming to consensus on ground rules for visits to peers’ classrooms and the conferences that follow. In ensuing years these reading discussions have been replaced by workshops that give participants hands-on experience with various observation and conferencing techniques recommended by the literature. Once the participants have had some exposure to the key issues involved in peer observation, they are divided into groups of three to begin the actual process.First, the colleagues meet in pairs to make the necessary preparations for a series of three classroom visits, in which Teacher A observes Teacher B, B observes C, and C observes A (Figure 1).FIGURE 1The first cycle of observationsAfter this set of classroom visits has been completed, all three colleagues come together for a post-observation conference that takes place in three two-part stages, with each of the participants playing a different role at each stage (Table 1). Each stage focuses on a different classroom observation, with the teacher and classroom observer (‘mentor’) discussing predetermined areas of interest regarding the execution of the lesson while the third party silently listens in as the conference observer. At the end of the discussion, this third party develops teacher training skills by guiding the other two teachers through a conference review in which they reflect and share their perspectives on the post-observation conference, identifying those mentoring behaviours that promoted reflection as well as those that had the potential to have negative consequences (Table 2). Through this debriefing process, all parties gain deeper understanding of the challenges involved with giving feedback and develop greater self-awareness with regard to their own skill at sharing and receiving it.TABLE 1Roles during post-observation conferenceFirst stageSecond stageThird stageTeacher ATeacherThird partyMentorTeacher BMentorTeacherThird partyTeacher CThird partyMentorTeacherTABLE 2Three-stage sequence of post-observation conferenceFirst stageSecond stageThird stageTime0:00–20:0020:00–30:0030:00–50:0050:00–60:0060:00–80:0080:00–90:00FocusConferenceReviewConferenceReviewConferenceReviewRoleTeacher (A) and Mentor (B) discuss the observed lesson. Third party (C) listens.Third party (C) guides Teacher (A) and Mentor's (B) reflection on the conference.Teacher (B) and Mentor (C) discuss the observed lesson. Third party (A) listens.Third party (A) guides Teacher (B) and Mentor's (C) reflection on the conference.Teacher (C) and Mentor (A) discuss the observed lesson. Third party (B) listens.Third party (B) guides Teacher (C) and Mentor's (A) reflection on the conference.At the start of each new stage, a different classroom observation becomes the focus, and all participants change roles accordingly. In this fashion, the three stages of mentor development afford every participant experience giving and receiving feedback at the levels of teacher, mentor, and even as a mentor educator (‘third party’). The cyclical nature of the process and the focus on reflection can lead to great insight, with additional self-awareness and sensitivity resulting from the fact that every feedback giver will have a recent or impending experience in the ‘hot seat’ to consider.Ground rules for peer observationThe execution of the pre-observation and observation stages of mentor development closely resemble other typical practices of peer observation. Key principles include the following:The pre-observation conference1 The observed teacher should set the agenda, determining what the goals of the observation are (Mann 2005). The more specific the observed teacher can be, the more the mentor is likely to provide useful information (Sheal op. cit.).2 The mentor can gain valuable insight and promote better preparation prior to the observation by asking such questions as:▪ What are the goals of this lesson?▪ How does this material fit with previous learning?▪ What problems might be anticipated? (Zuck 1984).3 The more time that is invested in this meeting, the more likely the mentor and teacher are to have a fruitful post-observation discussion. As a frame of reference, Richards and Lockhart (1991) state that in their work with peer coaching, the meetings typically last no more than an hour.During the observation1 The observer is often an invasive presence in the classroom. It is up to the observer to take whatever measures possible to minimize the potential ill-effects of the visit (Master 1983).2 The teacher can put the students at ease by telling them that the observer is not there to watch them but rather to help the teacher develop more skills.3 At the end, the observer should thank the teacher for the opportunity to visit, but should not make any substantive comments about the class. Comments of that nature can provoke premature discussion and circumvent the teacher's own reflection process.The post-observation conferenceOnce each teacher has observed one other, the three come together for the post-observation conference, where each participant takes a turn as a mentor holding a conference with the observed teacher while a third party watches and prepares for the conference review. The following section outlines best practices of mentor development post-observation conferences, by supporting ideas from relevant literature with data collected from teachers from the USA, Canada, the UK, Australia, and New Zealand who chose to participate in the mentor development programme at Kanda University of International Studies (KUIS), an EFL-focused university in Chiba, Japan. The participants had between 2 and 14 years of classroom teaching experience and all hold Masters degrees. At the time of the data collection, the length of their participation in the mentor development programme ranged from one semester to two years or more. Data were collected in two forms:1 anonymous responses given by 16 of the 18 participants to a short online survey;2 excerpts from personal teaching portfolios and journals that were volunteered by six participants.A few technical points were clarified in the survey data. For one thing, it is important to schedule the post-observation conference shortly after the classroom observations. For those nine participants who held their conferences within a week of the observations, only two reported difficulties sharing feedback as a result of delay, while four others reported that the delay was actually helpful for organizing notes and fostering reflection. On the other hand, four of the seven who put off their conference for a week or more found the delay problematic.To make sure that everyone gets equal attention and practice at the conference, time allotments for each part of the discussion can be agreed upon at the outset. KUIS participants settled on 20 minutes for each mentor–teacher discussion and ten minutes for each third party-guided conference review, with a full session thus taking 90 minutes in total. While many of the teachers initially expressed doubt that the sessions should be so long, survey responses show that 75 per cent of the participants ultimately found this amount of time was either ‘adequate’ or even ‘too short’.Mentor and observed teacherZuck (op. cit.: 340) states that it is not advisable for a mentor to begin a post-observation conference by telling a teacher that a class was good or bad because it ‘denies the teacher's role as self-evaluator and gets the discussion off to a bad start’. In order for any conclusions to be internalized by the teacher, the work of the mentor and teacher needs to be fully collaborative. The mentor ‘will need to listen to the concerns of the teacher and signal this to the teacher’ (Randall and Thornton 2005: 87). Indeed, for most cases, it is the teacher who is uniquely capable of accounting for why things went the way they did (Nunan 1996).If the teacher is given a large degree of control over the direction that the conversation takes, the door will be open for the teacher to ‘own’ the discoveries that are made. KUIS participants reported a variety of techniques for making this happen. A mentor may begin by inviting the teacher to share his/her own perspective on the lesson, perhaps by talking through the lesson from beginning to end. Typically, the result of this start is that ‘the points that the observer planned to raise are brought up first by the teacher’. Another reported approach was to encourage reflection on the positive aspects of the class:I knew if I launched into ‘So tell me about your lesson’, she would outline negatives only … I decided to start with a question for her to find one of the positives of her lesson. This was commented on by the other group members, as they said it set the tone for the peer observation feedback.As the conversation proceeds, a mentor's sensitivity can create a safe atmosphere:Throughout the interview, I wanted to make sure I phrased questions that didn't ask ‘why?’ as they can often sound like one is asking the other to justify their rationale for an activity or so on.Allowing the teacher control over the direction of the conversation may at times mean that certain issues remain overlooked, but this should not be taken as an indication that the process is a failure, for it must be remembered that participants in mentor development are peers voluntarily engaging in collaborative professional development. Mentors should resist the urge to stamp out any and all imperfect teaching behaviours, for this approach will only put a teacher on the defensive and demoralize. Nor does this mean that participants should contain themselves to positive feedback, for then ‘the whole exercise becomes a pointless act of mutual back-patting’ (Cosh op. cit.: 24). On this point, some KUIS teachers agreed, saying ‘You don't just want positive feedback—you do it for the negative feedback’ and ‘It can be frustrating for a teacher who wants some honest feedback to hear that everything was “great”’.Generally speaking, it is when the teacher's need to be heard and understood is satisfied that the path is best laid for the teacher to comfortably invite feedback. At that point, the mentor can make appropriate substantive comments within the context of what the teacher has already shared, ideally keeping in mind Gottesman's (2000: 8) advice that feedback should be… specific in nature, about items the teacher can control, … descriptive rather than evaluative, tactful, well-timed, checked for clarity and simplicity, dealing with behaviours rather than personalities (of either teacher or students), … and well organized.Criteria such as these can also be useful to guide the participants’ discussion in the following conference review, when the third party shifts the focus to the appropriateness and effectiveness of the feedback given.The conference reviewThe conference review:… enables us (all in turn) to explore the extremely important aspect of how we provide feedback that is affirming, non-threatening, and, at the same time, effective. Asking a colleague into the ‘sacred space’ of our classroom is a potentially threatening proposition, as is soliciting feedback. It is therefore important that we monitor ourselves (both givers and receivers of feedback) so that we are getting feedback that we want without needing to have a crying session afterwards. (Survey response)The inclusion of a conference review guided by a third party allows for exploration of the way teachers talk to one another and the sort of feedback they view as acceptable, and how this varies across cultures. Collaboration between two colleagues alone would of course be simpler in many ways, but survey responses showed that▪ 69 per cent of the participants would not have preferred to work in a pair▪ 75 per cent felt that they had benefited from taking the role of third party▪ 81 per cent felt that they had benefited from having a third party present when they were giving and/or receiving feedback about teaching.The third party faces a challenging but rewarding dance, having to both lead and demonstrate a discussion on good mentoring behaviours—proverbially ‘walking the walk’ while ‘talking the talk’. Although third parties have little or no time to prepare for the discussion immediately following the mentor–teacher conference, they may have the most dynamic material to work with of all, having the opportunity to generate instantaneous reflection on a mentoring situation witnessed from point blank. One simple way to begin is to follow the template of a typical mentor–teacher discussion by putting the observee in control of the discussion, having the mentor talk through his/her perspective of how each stage of the conference went. Throughout this post-conference discussion, the third party should ask probing questions to promote deeper reflection, making sure that the conversation does not dry up at a superficial level. For instance, the mentor may be asked how the conference did or did not go according to plan and how things might be handled differently if they come up again in the future.At times, the observed teacher's presence can complicate the dynamics of the discussion, making it difficult for the mentor to be completely open about challenges that arose. This can be especially troublesome for mentors who tend to make long lists of faulty teaching behaviours to ‘correct’. Mentors who instead concentrate on helping others to reflect and come to conclusions for themselves will likely have more fulfilling post-conference discussions, as they can more candidly explore the successes and failures of the various techniques employed. Though the presence of the observed teacher may still lend an air of artificiality to the proceedings at times, a touch of awkwardness is not necessarily a bad thing, as it can heighten the senses, making everyone more keenly aware of how well the impact of their words matches their intentions. When the time feels right, the observed teacher can also be brought into the conversation to provide the perspective of what it was like to be mentored in this instance and how the mentor's intended messages were received. It can be instructive to consider whether the teacher felt the mentor's language came across as neutral or judgemental, how much of the mentor's feedback fell into the areas the teacher had previously requested, and how the mentor made the teacher feel safe to reflect candidly on the lesson and come to conclusions for himself/herself.The third party's lack of prior information about the observed class can facilitate complete focus on teacher training issues such as the mentor's techniques for setting the tone and creating a safe environment, asking questions effectively and helping the teacher reflect, and responding appropriately to what the teacher shares. Still, a bit of context can be of great help. Just as a pre-observation conference can lay the groundwork for a good post-observation discussion, the mentor and third party can benefit from a conversation prior to the conference as well. In this meeting, the mentor can make choices about what sort of feedback would be desirable from the third party. A mentor may tell the third party that he/she plans to use a particular style of mentoring or that he/she wants to talk as little as possible during the discussion (much like a teacher of a student-centred class may choose to do) or simply to avoid closed questions, and thus the third party will be invited to comment on the mentor's successfulness in these particular areas. The mentor may even choose to share a ‘conference plan’ (the mentoring equivalent of a lesson plan) with the third party to further guide the observation of the conference.At the end of the conference review, the mentor and observed teacher should take a moment to share reflections on the mentor educator's work as well, closing the circle and taking full advantage of opportunities to increase awareness of the way feedback is given and received. This process is not fully completed until a second cycle is carried out in the opposite direction, so everyone can once again have the experience of the shoe being on the other foot. In other words, instead of Teacher A visiting Teacher B, this time Teacher A will visit Teacher C, B will visit A, and C will visit B (Figure 2), and this will once again be followed by a round of post-observation conferences and conference reviews. Scheduling can be tricky: working in these trios for two full cycles of observation and conference typically occupied KUIS participants for as long as a semester or more. After that point, the group members can decide whether to disperse into new groups to gain further experience with new colleagues in different classrooms or to repeat the process with the same partners so as to reap the benefits of the increased level of trust that will have developed over the course of the two rounds of observation and conference.FIGURE 2The second cycle of observationsFuture directionsThere is much that could be learnt from additional research into mentor development in practice, particularly from a conversation analysis of post-observation conferences to see exactly what sort of feedback tends to be shared, in what manner, and the effects various techniques tend to have. It would also be valuable to have concrete information about the adaptations that groups make as they tailor the model to their own needs and interests. Recordings are now being collected, and it is hoped that additional research can be shared in the future.In the meantime, some ideas for adapting the practices outlined in this article to meet various teaching circumstances and interests may be of use. In cases where it is not possible to find three colleagues to collaborate, pairs of teachers can work together and do the ‘policing’ for themselves. A video camera can serve as a substitute for the third party, and the pair can later watch the video (together or apart) to observe and reflect on their performances as mentors. An additional layer of reflection and feedback can be brought into the process if the participants write journals on the experience, which they can then exchange and respond to, and a distant third party can even participate asynchronously if the tape and journals are shared electronically. A completely different option practised by some KUIS participants was to diminish the focus on the feedback and instead turn the post-observation conference into a sort of collective brainstorming session on solutions to typical classroom problems. In this approach, the participation of a third party in the conversation is welcome but not absolutely essential.Teachers are notoriously busy people, so the idea of setting aside 90 minutes for a post-observation conference may be unrealistic for some. Some teachers at KUIS dealt with this problem by instead having short sessions after each observation, perhaps over lunch. Others simply agreed to tighter time limits, squeezing the whole process into 45 minutes or an hour. Another approach is to make the third-party conference review an optional step that is only taken when the participants feel there is a need.The multilayered practices of mentor development can lead to a variety of rewards. Insofar as it is a form of collaborative professional development, it brings peers together to talk shop and tap into one another's experience, breaking down barriers and giving novice teachers a chance to learn from the pros, and vice versa. The peer observation component adds the opportunity for participants to become more self-aware as they recognize their own behaviours in the practices undertaken by the observed teacher. At times, these observations can lead to renewed enthusiasm for teaching: peers may get new ideas for things to try out, or they may take comfort in the knowledge that they are not alone in facing certain challenges. Students ultimately stand to benefit from mentor development as their teachers find fresh approaches to instruction and improve their skill at delivering feedback. And for those interested in making a career transition into teacher education, mentor development provides practical hands-on experience and training.BaileyKCurtisANunanDPursuing Professional Development: The Self as Source2001Boston, MAHeinle and HeinleBenedettiT‘Enhancing teaching and teacher education with peer coaching’TESOL Journal19977/1412CoshJ‘Peer observation: a reflective model’ELT Journal199953/1227GottesmanBPeer Coaching for Educators2000Second EditionLondonScarecrow PressMannS‘The language teacher's development’Language Teaching20053810318MasterP‘The etiquette of observing’TESOL Quarterly198317/3497501MurphyJ‘An etiquette for the nonsupervisory observation of L2 classrooms’Foreign Language Annals199225/321525NunanDBaileyKNunanD‘Hidden voices: insiders’ perspectives on classroom interaction’Voices from the Language Classroom1996CambridgeCambridge University PressRandallMThorntonBAdvising and Supporting Teachers2005CambridgeCambridge University PressRichardsJFarrellTProfessional Development for Language Teachers: Strategies for Teacher Learning2005CambridgeCambridge University PressRichardsJLockhartC‘Teacher development through peer observation’TESOL Journal19911/2710ShealP‘Classroom observation: training the observers’ELT Journal198943/292103ShowersBJoyceB‘The evolution of peer coaching’Educational Leadership199653/61216ZuckJ‘The dynamics of classroom observation: evening the odds through information’TESOL Quarterly198418/233741
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B5A24388E2BC52A6A3774D010A856296E90D769.txt b/test/dataset/in/resources/corpus/Clean_0B5A24388E2BC52A6A3774D010A856296E90D769.txt
deleted file mode 100644
index bc43d6e..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B5A24388E2BC52A6A3774D010A856296E90D769.txt
+++ /dev/null
@@ -1 +0,0 @@
-jexbotexbotjJournal of Experimental Botany1460-24310022-0957Oxford University Press10.1093/jxb/ern137Research PapersOver-expression of a zeatin O-glucosylation gene in maize leads to growth retardation and tasselseed formationPineda RodóAlbert1BrugièreNorbert2VankovaRadomira3MalbeckJiri3OlsonJaleh M.1HainesSara C.1MartinRuth C.4HabbenJeffrey E.2MokDavid W. S.1MokMachteld C.1*1Department of Horticulture and Center for Genome Research and Biocomputing, Oregon State University, Corvallis, OR 97331-7304, USA2Agronomic Traits/Discovery Group, Pioneer Hi-Bred International, Inc., Johnston, IA 50131-1004, USA3Institute of Experimental Botany v.v.i., Academy of Sciences of the Czech Republic, Rozvojová 263, 165 02 Prague 6, Czech Republic4United States Department of Agriculture, Agricultural Research Service, National Forage Seed Production Research Center, Corvallis, OR 97331, USA*To whom correspondence should be addressed. E-mail: mokm@hort.oregonstate.edu72008315200859102673268610120083132008742008© 2008 The Author(s).2008This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.This paper is available online free of all access charges (see http://jxb.oxfordjournals.org/open_access.html for further details)To study the effects of cytokinin O-glucosylation in monocots, maize (Zea mays L.) transformants harbouring the ZOG1 gene (encoding a zeatin O-glucosyltransferase from Phaseolus lunatus L.) under the control of the constitutive ubiquitin (Ubi) promoter were generated. The roots and leaves of the transformants had greatly increased levels of zeatin-O-glucoside. The vegetative characteristics of hemizygous and homozygous Ubi:ZOG1 plants resembled those of cytokinin-deficient plants, including shorter stature, thinner stems, narrower leaves, smaller meristems, and increased root mass and branching. Transformant leaves had a higher chlorophyll content and increased levels of active cytokinins compared with those of non-transformed sibs. The Ubi:ZOG1 plants exhibited delayed senescence when grown in the spring/summer. While hemizygous transformants had reduced tassels with fewer spikelets and normal viable pollen, homozygotes had very small tassels and feminized tassel florets, resembling tasselseed phenotypes. Such modifications of the reproductive phase were unexpected and demonstrate a link between cytokinins and sex-specific floral development in monocots.Corncytokininplant developmenttasselseedZea mayszeatin O-glucosyltransferaseIntroductionCytokinins are plant hormones that are essential for cell division and plant development (Mok and Mok, 1994). Natural cytokinins are adenine derivatives, generally with an isoprenoid side chain at the N6 position, although cytokinins with aromatic side chains are also known to occur (Strnad, 1997). Zeatin is considered to be an essential cytokinin in higher plants due to its ubiquitous nature and high activity. Other free bases with cytokinin activity, cis-zeatin, dihydrozeatin, and N6-(Δ2-isopentenyl) adenine, are also present in most plant tissues. Derivatives of these bases include the corresponding ribosides and nucleotides, as well as glucosides with the sugar moiety at the O of the side chain or at the N7, N9, or N3 of the adenine ring.Until recently, the true activities of the various cytokinin metabolites were difficult to assess. The activity of natural as well as synthetic compounds have been traditionally determined by bioassays such as the tobacco callus, bean callus, or radish cotyledon bioassays (Murashige and Skoog, 1962; Letham, 1971; Mok et al., 1978). The results of these bioassays indicated that free bases and ribosides had high cytokinin activity. However, apparent cytokinin activities may not necessarily reflect their true activities in planta due to rapid metabolism or catabolism. For instance, while the N7 and N9 glucosides of zeatin are mostly inactive in bioassays (Letham et al., 1983), the O-glucoside is very active in bioassays, sometimes more than zeatin itself (Mok et al., 1992). The high activity of O-glucosylzeatin was assumed to result from conversion to the aglycone since the side chain far exceeds the optimal molecule size for cytokinin activity (Skoog and Armstrong, 1970).The identification of cytokinin receptors provided an alternative evaluation of cytokinin activity. For example, studies with the CRE1/AHK4 and AHK3 receptors of Arabidopsis showed that indeed, zeatin O-glucoside is not an active cytokinin (Spíchal et al., 2004). In general, these studies indicate that the highest cytokinin activity is conferred by free bases. Relatively lower activity was displayed by ribosides, while no activity was detected with glucosides (Spíchal et al., 2004; Yonekura-Sakakibara et al., 2004). There are, however, differences between receptors in cytokinin recognition as illustrated by the binding of only the trans isomer of zeatin by the Arabidopsis CRE1/AHK4 receptor, but both the cis and trans isomers by the maize ZmHK1 receptor (Yonekura-Sakakibara et al., 2004; Mok et al., 2005).Cytokinin O-glucosides are generally assumed to be storage products. While they lack activity per se, they can be activated through hydrolysis by β-glucosidases. Although conversion of zeatin to its O-glucoside by O-glucosyltransferases temporarily inactivates zeatin, this process also protects zeatin from cytokinin oxidases/dehydrogenases, which can degrade zeatin but not its O-glucoside (McGaw and Horgan, 1983; Armstrong, 1994). Complicating this picture is the distribution of these metabolites in the cell, resulting possibly in differential localization of substrates and enzymes. For instance, dihydrozeatin O-glucoside and the glucosides of a number of other plant growth regulators were found in the vacuoles (Garcia-Martinez et al., 1981; Schmitt and Sandermann, 1982; Bray and Zeevaart, 1985; Lehmann and Glund, 1986; Fusseder and Ziegler, 1988; Dean et al., 2003) whereas the maize β-glucosidase is targeted to plastids (Kristoffersen et al., 2000). Thus, while the high activity of O-glucosylzeatin in bioassays indicates that it can be converted to zeatin, the extent of this conversion in plants is unclear. It is possible that it is limited to certain tissues or particular stages of plant development or that it only occurs under specific conditions.Enzymes and genes involved in zeatin glycosylation and glucoside hydrolysis have been identified. The first zeatin O-glucosyltransferase (EC 2.4.1.203) was isolated from immature P. lunatus seeds (Dixon et al., 1989) while a variant of this enzyme, a zeatin O-xylosyltransferase (EC 2.4.2.40), was obtained from those of P. vulgaris (Turner et al., 1987). Subsequently, the genes encoding these enzymes, ZOG1 and ZOX1, were cloned (Martin et al., 1999a, b). In addition, two genes encoding O-glucosyltransferases with a preference for cis-zeatin were isolated from maize (Martin et al., 2001b; Veach et al., 2003). Other zeatin O-glucosyltransferase genes as well as two cytokinin N-glucosyltransferase (EC 2.4.1.118) genes were identified in the Arabidopsis genome (Hou et al., 2004). These enzymes all belong to Family 1 of the UDP-sugar requiring glycosyltransferases (http://www.cazy.org) and contain the signature sequence for this family in the C-proximal portion.An enzyme converting zeatin O-glucoside to zeatin was first identified in maize (Brzobohatý et al, 1993). This β-glucosidase (EC 3.2.1.21) has somewhat broader substrate specificity than the O-glucosyltransferases since it cleaves kinetin-N3-glucoside as well as some other substrates. The corresponding gene was cloned (Zm-p60.1) and found to be highly expressed in root meristems (Brzobohatý et al., 1993). A similar gene was isolated from Brassica napus (Falk and Rask, 1995).We are interested in the effects of over-expressing as well as repressing ZOG-type genes to determine the regulatory properties of O-glucosylation on plant development. The developmental modifications of maize transformants harbouring ZOG1 driven by the constitutive ubiquitin (Ubi) promoter is reported here. Our data indicate that zeatin O-glucosylation clearly affects root formation, leaf development, chlorophyll content, senescence, and male flower differentiation. Most interesting are the effects of the ZOG1 transgene on tassel development, with the formation of tasselseed in the homozygous transformants.Materials and methodsGeneration and selection of transgenic linesTransgenic Zea mays plants were obtained using an Agrobacterium-mediated transformation procedure of GS3×HC69 hybrid embryos as previously described (Zhao et al., 1998). The construct that was used contained the Ubi (ubiquitin) promoter (Christensen et al., 1992) upstream of the ZOG1 gene (accession no. AF101972) and a bialaphos resistance gene (Block et al., 1987) as the selectable marker. For further details on the construct see Fig. S1 in Supplementary data at JXB online. T0 plants were backcrossed four times to HC69.Five replications of four BC4 lines (P1, P3, P6, and P7) were planted from February through April and grown in the greenhouse at 25/20 °C (day/night) under natural light. Plants were identified as bialaphos-resistant or bialaphos-sensitive by painting part of the lower leaf with 200 mg l−1 bialaphos. For each replication, data were obtained from four plants per BC4 line.Crosses between sibs of the BC4 lines produced progeny equivalent to F2s. These progenies segregated at the expected 3:1 ratio of bialaphos-resistant to bialaphos-sensitive plants. Three repeats of two F2 lines (P3 and P7) with 30 plants per line were planted in early October to study the F2 generation. The photoperiod was gradually extended by using high-pressure sodium lamps (Hortilux LU1000B/Ht1/En) to simulate daylength conditions in spring and summer (April planting). Two additional 30 plant replications of P3 and P7 were planted in the spring and grown under natural light.Distribution of materialsNovel materials described in this publication may be available for non-commercial research purposes upon acceptance and signing of a material transfer agreement. In some cases such materials may contain or be derived from materials obtained from a third party. In such cases, distribution of material will be subject to the requisite permission from any third-party owners, licensors or controllers of all or parts of the material. Obtaining any permission will be the sole responsibility of the requestor. Plant germplasm and transgenic material will not be made available except at the discretion of the owner and then only in accordance with all applicable governmental regulations.PCRDNA was isolated from 100 mg tissue samples by sequential treatment with 500 μl extraction buffer (200 mM TRIS-HCl, 250 mM NaCl, 25 mM EDTA, 0.5% SDS), 500 μl saturated phenol, and 500 μl chloroform, followed by precipitation with 400 μl isopropanol. Internal primers for ZOG1, ZOG411F (5′-CATCTCAAATGTTGAA-AACTAC-3′) and ZOG930B (5′-CTTCACTTCCGGCAAAGATGTC-3′), were used for PCR. An initial denaturation cycle at 95 °C for 4 min was followed by 30 cycles of 95 °C for 1 min, 52 °C for 1 min, and 72 °C for 1 min. After an additional cycle at 72 °C for 10 min, samples were stored at 4 °C.RT-PCRTotal RNA was isolated from leaves, roots, and tassels at various stages using TRIzol® Reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. Polysaccharide-rich samples were treated with 20% polyvinylpyrrolidine (PVP) and 8 M lithium chloride and incubated overnight at –20 °C. Synthesis of cDNA was achieved using the SuperScript™ II reverse transcriptase kit (Invitrogen, Carlsbad, CA, USA). The resulting cDNA product was subjected to semi-quantitative PCR using actin to normalize samples. Primers for actin PCR were mzACTIN-32F (5′-GTGACAATGGCACTGGAATG-3′) and mzACTIN-741B (5′-GACCTGACCATCAGGCATCTC-3′). The internal primers ZOG411F and ZOG930B were used to amplify ZOG1. The following conditions were used for PCR: initial denaturation at 94 °C for 4 min followed by 30 cycles of 94 °C for 1 min, 57 °C for 45 s, and 72 °C for 1 min. The last cycle was followed by incubation at 72 °C for 10 min.Western analysesProtein was extracted from leaves with buffer containing 55 mM TRIS, 50 μM EDTA, 5 mM DTT, pH 7.4 (800 μl extraction buffer for 200 mg leaf tissue) followed by 30–75% NH4SO4 fractionation. Samples were desalted and then purified by Blue Sepharose 6B affinity chromatography (Dixon et al., 1989). A fraction equivalent to 7 mg fresh leaf tissue was separated on a 12% SDS polyacrylamide gel and blotted to Immobilon-P transfer membrane (Millipore, Billerica, MA, USA). Western blots were developed as described in Martin et al. (1990).Cytokinin analysesCytokinin levels were determined in roots and leaves of 28-d-old and in leaves of 110-d-old non-transformed and hemizygous plants. For extraction and purification of cytokinins as well as determination of cytokinin levels, the procedures described by Veach et al. (2003) were followed. The HPLC/MS system consisted of a HTS-Pal auto-sampler with a cooled sample stack (CTC Analytics, Zwingen, Switzerland), quaternary HPLC pump Rheos 2200 (Flux Instruments, Basel, Switzerland), Delta Chrom CTC 100 Column oven (Watrex, Prague, Czech Republic) and TSQ Quantum Ultra AM triple-quad high resolution mass spectrometer (Thermo Electron, San Jose, USA). For HPLC a Synergi Hydro column (Phenomenex, Torrance, USA) was used. Data were obtained from three samples per genotype, with two HPLC injections and analyses per sample.Phenotypic characterizationsPhenotypic characterization of transformed lines included measurements on vegetative growth rate, plant height, root length, chlorophyll content, leaf width, stomatal distribution, meristem structure, development of reproductive organs (time of appearance, size), pollen viability, and kernel development.Foliar data presented were taken from the 10th leaf, which was found to be representative of the whole plant by measuring all leaves from fewer plants (data not shown). Foliar chlorophyll levels were monitored with a chlorophyll content meter CCM-200 (Optisciences Inc., Tyngsboro, MA, USA) over a 175-day period. Four measurement points were taken from each leaf. To convert these measurements to chlorophyll units per leaf fresh weight, chlorophyll was extracted by incubating leaf tissues in N,N′-dimethylformamide (7% leaf weight/solvent volume) at 4 °C for 48 h, as described in Inskeep and Bloom (1985). Sample absorbances were measured at 647 nm and 664.5 nm (maximum for chlorophyll b and a, respectively) and absolute chlorophyll contents calculated according to the equation:Optimeter CCI units were then transformed into μg of chlorophyll per mg fresh weight with restriction curves. Random leaf samples were examined with a light microscope to determine stomatal cell distribution patterns.Apical meristem samples were collected from 30-d-old plants, fixed in FAA (50% ethanol, 10% acetic acid, 5% formalin, by vol.), dehydrated in a graded ethanol series, and embedded in Technovit 7100 plastic (Heraeus Kulzer, Wehrheim, Germany). Sections were cut on an AO 820 rotary microtome to obtain 46-μm-thick serial sections. Samples were stained with Toluidine Blue-O, and analysed with Image-Pro Plus software (Media Cybernetics, Silver Spring, MD, USA).Pollen viability was tested prior to pollination assays. Fresh pollen grains were collected at 09.00 h and kept for 2 h on liquid germination media containing (l−1) 236.2 mg Ca(NO3)2, 24.8 mg H3BO3, and 30% (w/v) of polyethyleneglycol (PEG). Silk filaments from the female flower were added to enhance pollen tube growth. A pollen cryo-preservation protocol was optimized to ensure a constant supply of viable pollen during pollination assays. Pollen grains were collected, dried at 20 °C and 20–40% humidity for 2 h and shock-frozen by immersion in liquid nitrogen as described by Barnabas (1994). A saturated MgCl2 solution was used to stabilize the humidity in the desiccator. Prior to use, pollen was thawed in a water bath at 40 °C for 2 min. The combination of fresh and cryopreserved pollen allowed uniform pollination coverage during the receptive period of the maize ear.The experimental design for all phenotypic traits was fully randomized. Data were analysed statistically by multivariate analysis of variance (MANOVA) with repeat, line, and type as the main sources of variation, followed by a Tukey test at 5% significance level. The analyses aimed to identify the interactions among the independent variables as well as their effect on the response variables. For those traits where no relevant variable interactions were observed, two-way and one-way ANOVA were performed.ResultsTransgenic Ubi:ZOG1 maize lines were generated using Agrobacterium-mediated transformation. Bialaphos resistance was the selectable marker and initial transformants were backcrossed to the parental inbred HC69 for four generations to generate transformants with genetic backgrounds similar to this inbred. Four BC4 lines (P1, P3, P6, and P7), derived from independent transformation events and containing a single insert each, were selected for further study. The BC4 lines segregated into bialaphos-resistant hemizygous (P+) and bialaphos-sensitive non-transformed (NT) plants (1:1). To obtain homozygous transformants (P++), hemizygous BC4 plants were selfed. For simplicity, the progeny obtained after selfing the hemizygous BC4 is referred to as F2.Bialaphos-resistant plants transcribe and translate ZOG1RT-PCR (Fig. 1a) and western (Fig. 1b) blots of BC4 plants indicated that transcription and translation of Ubi:ZOG1 occurs in transgenic maize tissues. Tissue samples from bialaphos-resistant plants of all four lines produced clear RT-PCR and protein bands, but those from bialaphos-sensitive plants did not.Fig. 1.Ubi::ZOG1 expression. (a) RT-PCR products with ZOG1-specific primers and actin as a control showing expression of the transgene in roots, leaves, and tassels. P1+, P3+, P6+, and P7+ are independent hemizygous lines. NT is a non-transformed control. Plasmid I-1124 contains the Ubi::ZOG1 insert. (b) Western blot of ZOG1 product in leaves (4 mg equivalent) from hemizygous transformants and non-transformed sibs, developed with monoclonal antibodies specific to ZOG1.Over-expression of ZOG1 leads to large increases in zeatin O-glucosideCytokinins were analysed in roots and leaves of young plants and in leaves of mature plants (Fig. 2). As expected, the zeatin O-glucoside levels were increased significantly in the hemizygous transformants: 100-fold in roots, and 38-fold and 88-fold higher in leaves of young and older plants, respectively (Fig. 2a, d, g). By contrast, the cis-zeatin O-glucoside levels differed only up to 2-fold and glucosylation of ribosides was low. These levels are both in agreement with the substrate preference of the ZOG1 enzyme (Martin et al., 1999a). Overall, O-glucosides constituted the main pool of cytokinin metabolites, even in the non-transformed plants. The levels of N-glucosides were negligible compared with those of O-glucosides (data not shown).Fig. 2.Cytokinin concentrations in roots and leaves. (a, b, c) Cytokinin concentrations in roots of 28-d-old non-transformed (NT) and hemizygous Ubi::ZOG1 (+) maize plants. (d, e, f) Cytokinin concentrations in leaves of 28-d-old maize plants. (g, h, i) Cytokinin concentrations in leaves of 95-d-old maize plants. (a, d, g) O-glucosides; (b, e, h) free bases; (c, f, i) ribosides and nucleotides. Z, trans-zeatin; cZ, cis-zeatin; iP, N6-(Δ2-ispentenyl)adenine; ZR, trans-zeatin riboside; cZR, cis-zeatin riboside; iPR, N6-(Δ2-ispentenyl)adenosine; ZOG, trans-zeatin O-glucoside; cZOG, cis-zeatin O-glucoside. Values are means (±SE) of three samples with two HPLC analyses for each. Significant differences between transformants and controls are indicated by *P <0.05, **P <0.01, or ***P <0.001.In roots, levels of free bases were extremely low, both in transformed and control plants (Fig. 2b). Root riboside levels were higher than free bases but did not differ between transformants and controls (Fig. 2c). In leaves of both young and old plants, the levels of zeatin were higher in transformants than in controls (Fig. 2e, h). Although this increase in zeatin could be accounted for by hydrolysis of only 1% of the zeatin O-glucosides in transformants during storage and extraction, the ribosides and nucleotides were also elevated in transformants (Fig. 2f, i), which could not stem from glycoside hydrolysis. The levels of dihydrozeatin and its derivatives were generally below detection levels. Altogether, the data indicate that the levels of active cytokinins (free bases and ribosides) are slightly lower in transformed roots but higher in transformed leaves.Increased cytokinin conjugation leads to pronounced changes in maize plant architecture and leaf widthThe hemizygous Ubi:ZOG1 BC4 plants were morphologically distinguishable from non-transformed sibs. They grew slower, were shorter, and had narrower leaves. There were three types of F2 plants: bialaphos-sensitive non-transformed plants, plants resembling the hemizygous BC4 transformants, and a class of extremely small plants (Fig. 3a). The segregation ratios were 16:36:10 and 15:27:22 in the P3 and P7 lines, respectively, fitting a 1:2:1 ratio and thus supporting the hypothesis that the very small plants were homozygous for the transgene. There were significantly fewer leaves on the hemizygous transformants than on the non-transformed plants and even fewer leaves on the homozygous transformants (Fig. 3b). At the time of tassel emergence, plant height was reduced by 22–41% in the hemizygous and by 48–60% in the homozygous Ubi:ZOG1 plants compared with non-transformed plants (Fig. 3c). There was a greater reduction in plant height than in the number of leaves, indicating that internodes were shortened in response to transgene activity. Stems were slender in hemizygous and very slender in homozygous Ubi:ZOG1 plants. The leaves of hemizygous plants were significantly narrower than those of non-transformed sibs (Fig. 3d, e), but their length was the same (data not shown). The leaves of homozygous transformants were extremely narrow, even at plant maturity (Fig. 3e), and also shorter than control and hemizygous plants. The thinner stems and leaves of transformants suggested smaller meristematic regions and microtome sections confirmed that hemizygous and homozygous meristems were indeed significantly smaller (Fig. 3f-i). These characteristics were not influenced by environmental fluctuations (light intensity or day length).Fig. 3.Vegetative development. (a) Non-transformed (left), hemizygous (centre) and homozygous (right) 97-d-old plants. (b) Number of leaves over time. P+ and P++ represent hemizygous and homozygous lines, respectively. (c) Plant height at tassel emergence. (d) Non-transformed (top), hemizygous (centre) and homozygous (bottom) leaves. (e) Average width of the tenth leaf. (f) Shoot meristem size of 40-d-old plants. (g) Non-transformed meristem. (h) Hemizygous meristem. (i) Homozygous meristem. Values are means (±SE), obtained from two experiments with a total of 60 plants each. Values for (f) are means (±SD) of 10 meristems.Chlorophyll formation and distribution of stomates are altered in transformantsDifferences in foliar chlorophyll content were noticeable as early as 23 d after germination. The transformed plants were darker green than the non-transformed ones. Chlorophyll levels in transformants were elevated throughout their life span and ultimately resulted in the delay of senescence which kept transformed plants green weeks after control sibs had senesced (Fig. 4a). This is interesting since increased chlorophyll formation and delay of senescence are traits usually associated with an increase in cytokinins (Richmond and Lang, 1957; Gan and Amasino, 1995) whereas the plant architecture of Ubi::ZOG1 transformants is generally indicative of a decrease in active cytokinins (see Discussion). However, these findings are supported by the cytokinin analyses, which showed increases in active cytokinins in transformant leaves (Fig. 2). The delay of senescence was influenced by the environment since it occurred in the summer but not in the winter under artificial light (Fig. 4b), even though the daylength was adjusted to mimic summer conditions. It is most likely that the difference in light intensity and/or quality was a major factor in this seasonal effect.Fig. 4.Leaf characteristics. (a, b) Chlorophyll levels in the 10th leaf through senescence. Plants grown in the spring/summer under natural long-day conditions (a) and in the autumn under artificial lighting (b). (c) Paired chain of stomata from a non-transformed leaf. (d) Single chain of stomata from Ubi::ZOG1 leaf. (e) Number of paired stomata mm−2 (f) Total number of stomata mm−2. Values for (a) and (b) are means (±SE) of two experiments with four plants per line and six measurements per leaf. Values for (e) and (f) are means (±SD) of cross-sections of 20 random mm2 fields.Closer examination of the leaves revealed that the number of major veins was the same in non-transformed plants and hemizygous transformants but that the distance between the veins was smaller in the latter. On normal maize leaves, the stomata usually occur in rows, both in double files with stomata at alternate positions (Fig. 4c; Hernandez et al., 1999) and single files (Fig. 4d). Transformants had fewer double file stomata (Fig. 4e) but more single files. The overall density of the stomata in the leaves did not differ between the two groups (Fig. 4f).Root mass and branching is increased in Ubi:ZOG1 transformantsRoot morphology differed between non-transformed plants and hemizygous transformants. Ubi:ZOG1 roots were thicker, more branched (Fig. 5a) and longer (Fig. 5b). As a result, the total root weight was much greater in the transformants (Fig. 5c).Fig. 5.Root development. (a) Non-transformed (two at left) and hemizygous (two at right) roots 3 weeks after planting. (b) Root fresh weight. (c) Root length. Values for (b) and (c) are means (±SD) of 10 plants.Ubi:ZOG1 causes reduction of tassel size and feminization of floretsHemizygous Ubi:ZOG1 plants showed delayed tassel initiation (Fig. 6a) and, more strikingly, a drastic reduction in tassel size, branching, and spikelet production compared with the non-transformed control (Fig. 6b, c, d). The weight of transformed tassels was about 75% less than that of non-transformed tassels (Fig. 6b), due to the decreased number and smaller size of branches. Moreover, the tassels were abnormal. Normal maize tassels have many florets up to the tip of each branch (Fig. 6e), but the hemizygous Ubi:ZOG1 tassels had functional spikelets only at the lower end of the tassels while the tips of the branches were devoid of most floral structures (Fig. 6f). Sterile spikelets consisted of two external glumes with no florets, lemma or stamens. The normal florets underwent anthesis and produced functional pollen grains as demonstrated by the ability of the pollen to germinate (Fig. 6g). Pollen viability was further supported by the roughly equal numbers of transformed and untransformed kernels in the BC4 populations, which were generated using BC3 transformants as male parents.Fig. 6.Tassel development. (a) Age of plant at tassel emergence. (b) Tassel fresh weight at maturity. (c) Non-transformed tassel. (d) Hemizygous Ubi::ZOG1 tassel. (e) Control lateral rachis with fully developed fertile spikelets. (f) Hemizygous lateral rachis with empty spikelets. (g) Viable Ubi::ZOG1 pollen in liquid germination medium. (h) Homozygous tassels with various degrees of floret feminization. Values for (a) and (b) are means (±SE) of five experiments with four plants each.Homozygous F2 tassels were phenotypically even more extreme, showing a dosage effect of the ZOG1 gene. They were very small, had no or very few branches, and most interestingly, showed various degrees of floret feminization (Fig. 6h). Fertilization of these female florets resulted in formation of complete kernels (Fig. 6h). The feminization varied in the homozygous population, from a few female florets and seeds to almost complete silking of the tassel (Fig. 6h, left to right). Normal tassel floret development is initially bisexual, after which florets become unisexual through gynoecium abortion (Cheng and Pareddy, 1994). This programmed gynoecium abortion was inhibited by ZOG1 over-expression in some of the lower florets since fertile ovules and elongated silk were formed. The few male florets on the apical portion of the homozygous tassels yielded a small amount of viable pollen. The phenotype varied depending on the light source during plant growth. Plants grown in the winter under artificial light always showed some degree of feminization while plants grown in the spring and summer under ambient light conditions had only a few tasselseeds.Ubi:ZOG1 reduces seed weightEar development on hemizygous Ubi:ZOG1 plants lagged behind that of non-transformed plants but the ears looked normal (Fig. 7a). Hemizygous ears had only partial seed set, in contrast to fully filled non-transformed ears under the same greenhouse conditions; however, when the ears were artificially pollinated, the hemizygous ears were also completely filled, indicating that pollen availability (not viability) was the limiting factor. When the Ubi:ZOG1 plants were used as the male parent, plants produced kernels with a lower average weight than those pollinated with non-transformed pollen (Fig. 7b). To confirm this initial observation, hemizygous plants were used to pollinate non-transformed plants and the resulting kernels were weighed and classified as transgene positive or negative (using PCR with ZOG1-specific primers). The data indicate that kernels on the same ear resulting from pollination with Ubi:ZOG1 pollen grains were significantly smaller than those resulting from pollination with non-transformed pollen (Fig. 7c).Fig. 7.Ear and kernel development. (a) Non-transformed (left) and hemizygous (right) ears from plants of the same age before pollination. (b) Weight of kernels from non-transformed ears pollinated with pollen from non-transformed (NT) or hemizygous Ubi::ZOG1 (P+) plants. (c) Weight of kernels from non-transformed ears pollinated with pollen from hemizygous transformants. Kernels were taken from the same ear and classified as P–×P– or P–×P+ by PCR with ZOG1-specific primers. Values for (b) are means (±SE) of all kernels of four ears. Values for (c) are means (±SD) of 15±2 kernels.DiscussionRelevance of cytokinin determinationsConstitutive over-expression of ZOG1 resulted in elevated levels of O-glucosides, as expected. There was, however, no concomitant decrease in active cytokinins (bases and ribosides) in roots and an increase in overall active cytokinins in leaves. This indicates that the initial decrease in zeatin led to an increase in cytokinin biosynthesis. It should be noted though that cytokinin analyses can only determine the cytokinin composition in a large amount of tissue, whereas the distribution of cytokinins within specific tissues and cells may be most critical. As a consequence of ZOG1 over-expression, there are probably changes in the inter- and intracellular distribution of various active cytokinins and their metabolites due to differential expression of cytokinin biosynthetic genes. Thus it is possible that cytokinins are elevated at or near biosynthetic sites but not at cytokinin receptor sites.Root development and plant architecture of Ubi:ZOG1 transformants are characteristic of cytokinin deficiencyRoot development was enhanced in the Ubi:ZOG1 maize plants similar to that observed in Arabidopsis and Lotus japonicus transformed with maize cytokinin oxidase/dehydrogenase genes (Werner et al., 2001, 2003; Lohar et al., 2004; Kopečný et al., 2006). This suggests that the decrease in physiologically active cytokinins caused by increased glucosylation stimulated initiation and growth of root tips. Both the increased number of root tips and possible increased cytokinin biosynthesis may have led to the adjustment in free cytokinins to levels of non-transformed maize.While root development was stimulated by the presence of the transgene, the opposite was the case with shoot development. The hemizygous Ubi:ZOG1 transformants were shorter and less robust than the controls and this size reduction was more severe in the homozygous plants, indicating dosage effects of the transgene. The phenotypes were similar to those of maize plants having a CKX transgene (N Brugière, unpublished results). Slower growth rates and reduced plant stature were also observed in dicots having decreased cytokinin levels or reduced cytokinin sensitivity due to increased expression of oxidases/dehydrogenases (Werner et al., 2003; Kopécný et al., 2006), deficiencies in biosynthesis (Miyawaki et al., 2006), or mutations in receptor genes (Higuchi et al., 2004; Nishimura et al., 2004; Riefler et al., 2006).Observations regarding cytokinin effects on monocot leaf development are limited. Classical experiments involving exogenous cytokinin applications conducted with dicots demonstrated that cytokinins stimulated leaf expansion (Kuraishi and Okumura, 1956). However, transformants and mutants of Arabidopsis with either reduced cytokinin levels or deficient in cytokinin perception have smaller leaves, but the ratio between the length and width remains constant (Riefler et al., 2006; Werner et al., 2001, 2003). The leaf width of Ubi:ZOG1 hemizygous maize plants was significantly reduced by over-expression of ZOG1, while both the leaf width and length were decreased in Ubi:ZOG1 homozygous plants. Maize genetic mutants with smaller leaves usually have smaller meristems (Scanlon et al., 1996). This is also the case with the Ubi:ZOG1 transformants. The meristems of the homozygous Ubi:ZOG1 plants were slightly smaller than those of the hemizygous Ubi:ZOG1 plants which were smaller than those of the non-transformed control plants. The meristems of Arabidopsis plants over-expressing cytokinin oxidase/dehydrogenase genes were also smaller those of control plants (Werner et al., 2003; Kopečný et al., 2006). Another notable difference between the Ubi:ZOG1 transformants and control plants was the arrangement of leaf stomata. There were more single file stomata and fewer double file stomata in the transformants, which may be related to the reduced number of cell files between veins.Chlorophyll formation and leaf senescence reflect higher levels of active cytokininsChlorophyll levels were increased and senescence was delayed in Ubi:ZOG1 transformant leaves. These characteristics are indicative of an increase in active cytokinins since exogenous cytokinins and up-regulation of cytokinin biosynthesis produce similar phenotypic changes (Richmond and Lang, 1957; Smart et al., 1991; Li et al., 1992; Gan and Amasino, 1995; Robson et al., 2004). The cytokinin analyses confirmed that active cytokinins (free bases and ribosides) are increased in leaves. Although it is possible that accumulation of extremely high amounts of O-glucosides can eventually lead to increased free bases due to the action of β-glucosidases, higher chlorophyll and delayed senescence were also reported for tobacco over-expressing the AtCKX2 gene of Arabidopsis (Mýtnová et al., 2006). A factor common to both the transgenic maize and tobacco is the generally slower development due to cytokinin O-glucosylation and degradation, respectively, thus prolonging the growth period. However, if this slower development is delaying maturation to a significant extent, time of flowering would also be greatly delayed. This was not the case since tassels emerged only about 2 d later in transformants than in untransformed controls, whereas leaf senescence showed a much longer delay (Fig. 4a). Both the ZOG1 and AtCKX2 leaves may respond to the resulting cytokinin deficiency by increasing cytokinin biosynthesis. Since biosynthesis takes place in the plastids (Kasahara et al., 2004; Sakakibara, 2006), there may be a localized increase of cytokinins in plastids. Whether higher cytokinin levels in chloroplasts could bring about an increase in chlorophyll is an interesting question. It would assume direct action of cytokinins in chloroplasts, which has never been demonstrated.Indirect effects contributing to the higher chlorophyll in the transformed maize leaves can not be excluded. Non-transformed maize leaves expand very rapidly under natural summer conditions while those of the transformants remain much narrower, which could result in differential chlorophyll accumulation. Furthermore, the ratio between root and shoot growth is much higher in the transformants, possibly leading to higher accumulation of nutrients (including N) in the leaves. If this is the case, the much higher cytokinin levels in the mature transformant leaves may not be the cause, but rather a consequence, of the delayed senescence.Increased cytokinin conjugation leads to defective tassel development and feminization of the lower floretsThe drastic reduction in tassel size of the Ubi:ZOG1 transformants points to the importance of cytokinins in maize reproductive development. Normal tassels have 10–50 lateral branches which contain paired spikelets (Cheng and Pereddy, 1994). By contrast, the hemizygous Ubi:ZOG1 transformants had fewer and shorter lateral branches. Terminal florets were often missing. Maize plants transformed with the CKX1 gene driven by a pollen-specific promoter also had smaller tassels even though CKX1 expression was targeted to the pollen (Huang et al., 2003). A rice cultivar with reduced expression of a cytokinin oxidase/dehydrogenase gene and increased cytokinins had more reproductive organs (Ashikari et al., 2005). Although reproductive programmes of maize and rice are different in a number of aspects, these studies suggest that cytokinins have a positive effect on terminal flower development, most probably through its effects on meristem size.Most intriguing was the occurrence of female florets at the lower end of the homozygous Ubi:ZOG1 tassels which formed seed when pollinated. Maize florets begin as complete bisexual flowers containing pistil and anther initials, but later the pistils abort (Cheng and Pareddy, 1994). The process of pistil abortion must have been inhibited in the homozygous Ubi:ZOG1 plants. Tasselseeds were found on all four transformed lines examined indicating that they were not a consequence of accidental insertion in one of the tasselseed (TS) loci but rather as the result of altered cytokinin levels or composition. Tasselseed in maize is a known phenomenon and several genetic mutations causing feminization of the tassel have been described (Irish and Nelson, 1989; Dellaporta and Calderon-Urrea, 1994). The most prominent phenotypes occur in recessive mutants ts1/ts1 and ts2/ts2, which display complete reversion from male to female inflorescences, with the failure of pistil abortion and the induction of stamen abortion (Emerson, 1920; Irish et al., 1994; DeLong et al., 1993). Somewhat less extreme is the dominant Ts5 mutant, which shows positional effects, with female florets occurring at the basal portion of the tassel (Nickerson and Dale, 1955), similar to the Ubi:ZOG1 transformants. The ts4/ts4 and ts6/ts6 mutants also show partial reversions (Dellaporta and Calderon-Urea, 1994). The TS2 gene was cloned and found to have homology to short chain alcohol dehydrogenases (DeLong et al., 1993). The maize transformants described here establish the first link between tasselseed and cytokinins.Previously, cytokinins have been implicated in sex expression of a number of plants (Durand and Durand, 1994), but to our knowledge they have not been researched in connection with sex expression in maize tassels. The occurrence of tasselseed on the homozygous Ubi:ZOG1 transformants indicates that increased zeatin O-glucosylation and the associated disturbance in cytokinin homeostasis result in feminization. Whether this is due to a decrease in active cytokinins (as shown by the retarded shoot development and smaller meristem size) is difficult to assess. However, tasselseed formation was also observed on maize transformants over-expressing a cytokinin oxidase/dehydrogenase gene under the control of the Ubi promoter (N Brugière, unpublished results). In most plant species (although primarily dicots), cytokinins are feminizing, but exceptions are known (Durand and Durand, 1994). Interestingly, only the lower inflorescences showed this feminization whereas the more apical florets had the usual male characteristics. Thus there may be a gradient of active cytokinins or some polarity in signal distribution, causing female flowers to develop at the basal end of the tassel.A possible alternative explanation for the tasselseed characteristic could reside in the changed composition of cytokinins. Not much is known about the influence of particular cytokinin metabolites on sex expression. In the most extensively studied system, Mercurialis, where a number of genes control sex expression, specific cytokinins have been linked with sex differentiation and male sterility. For instance, occurrence of trans-zeatin in apices was correlated with femaleness, while its riboside and nucleotide were more abundant in males (Durand and Durand, 1994). The presence of cis-zeatin and its riboside were associated with male sterility. Thus the changes in the ratios between cis and trans isomers in the Ubi::ZOG1 homozygotes, as a result of the preference of ZOG1 for trans-zeatin, may also be contributing to the abnormal tassel phenotypes. However, this is difficult to assess since any changes in cytokinin levels and composition could be the cause, but also merely the consequence, of altered development.The influence of light intensity or quality on the tasselseed trait in Ubi:ZOG1 transformants indicates an interaction between the cytokinin and light signalling pathways. Such interactions have been previously observed. For instance, exogenous cytokinin caused de-etiolation of dark-grown Arabidopsis seedlings (Chory et al., 1994). Furthermore, Arabidopsis transformants and mutants with altered expression of the cytokinin type A response regulator ARR4 showed altered red light sensitivity (Sweere et al., 2001; To et al., 2004).Pollen grains of hemizygous Ubi:ZOG1 plants appeared normal and germinated on artificial medium. Even the few anthers on homozygous transformants contained normal pollen. Also Arabidopsis transformants with constitutive over-expression of CKX genes produced fewer but still functional pollen grains (Werner et al., 2003); however, maize plants over-expressing CKX1 via a pollen-specific promoter (pZtap) showed pollen sterility (Huang et al., 2003).Female inflorescence development is normal but seeds are smaller in Ubi:ZOG1 plantsThe female inflorescences are not altered in Ubi:ZOG1 maize plants indicating that the changes in cytokinin levels have stronger effects on male flower development. Also, no obvious abnormalities in the flowers were observed in transgenic tobacco over-expressing ZOG1 (Martin et al., 2001a) even though tobacco flowers are derived from the terminal bud. Flower abnormalities were also absent from Arabidopsis plants constitutively over-expressing CKX genes (Werner et al., 2003). However, when cytokinin levels were increased through senescence-induced expression of the Agrobacterium IPT gene, the pistil was retained in the lower floret and fused kernels were formed (Young et al., 2004).Until recently, very few studies have addressed the effects of cytokinin on seed size. Transgenic Arabidopsis with decreased cytokinin levels or sensitivity produced fewer but larger seeds (Werner et al., 2003; Kopečný et al., 2006; Riefler et al., 2006). However, the maize seeds with increased cytokinin conjugation were smaller than non-transformed seeds from the same ears when control plants were pollinated with pollen from hemizygous Ubi:ZOG1 transformants. It should be noted though that the increased seed size in Arabidopsis was associated with a decrease in the number of seeds, while seed fill of the maize cobs was complete. Moreover, maize and Arabidopsis seed development can not be compared due to the fact that maize kernels are largely endosperm while Arabidopsis seeds consist mainly of embryos.Implications of high levels of cis isomersIn the present study, the predominant cytokinin in non-transformed maize leaves was the O-glucoside of cis-zeatin. In a previous study, cis-zeatin and its derivatives were found to be the major components in maize roots and stems (Veach et al., 2003). Levels of cis isomers were also very high in kernels although lower than the trans counterparts (Veach et al., 2003). Maize has a cytokinin glucosyltransferase with a preference for cis-zeatin over trans-zeatin (Martin et al., 2001b). In addition, a maize cytokinin receptor, ZmHK1, with high affinity for cis-zeatin and trans-zeatin has been identified (Yonekura-Sakakibara et al., 2004). These findings suggest that cis-zeatin is an active cytokinin in tissues where ZmHK1 is expressed.A related issue is the possible origin of cis-zeatin. The presence of a hydroxylase for the synthesis of cis-zeatin, similar to the hydroxylase found in Arabidopsis for the formation of trans-zeatin (Takei et al., 2004), is a distinct possibility. Recent findings in Arabidopsis (Miyawaki et al., 2006) favoured tRNA degradation rather than de novo synthesis as the source of cis-zeatin. Maize may be different from Arabidopsis in that the cis-zeatin level is too high to be accounted for solely by tRNA breakdown. In addition, the presence in maize of a receptor responsive to cis-zeatin and a cis-specific glucosyltransferase indicates species-specific pathways, supporting direct synthesis of cis-zeatin in monocots.ConclusionsThe changes in plant architecture associated with ZOG1 over-expression are consistent with the reduction of active cytokinins, but chlorophyll levels and retention seem to reflect increases in cytokinins, as supported also by the cytokinin analyses. The most interesting effect of ZOG1 over-expression was the feminization of tassel floret development. This novel observation provides a link between cytokinins and sex-specific floral development in maize.Supplementary dataA supplementary figure with specifics of the construct used for transformation is available at JXB online.We thank K Cook for her assistance with microtome sectioning. We thank M Rossman and C Pereira for their help with the statistical analyses. This work was supported by the National Science Foundation under Grant No. 0514024 and by the Ministry of Education, Youth and Sports of the CR under project no. ME868.ArmstrongDJMokDWSMokMCCytokinin oxidase and the regulation of cytokinin degradationCytokinins: chemistry, activity, and function1994Boca RatonCRC Press139154AshikariMSakakibaraHLinSYamamotoTTakashiTNishimuraAAngelesERQianQKitanoHMatsuokaMCytokinin oxidase regulates rice grain productionScience2005309741745BarnabasBBajajYPSPreservation of maize pollenBiotechnology in agriculture and forestry, Vol. 25. Maize1994BerlinSpringer-Verlag607618BlockMBottermanJVandewieleMDockxJThoenCGosseleVMovvaNThompsonCMontaguMLeemansJEngineering herbicide resistance in plants by expression of a detoxifying enzymeEMBO Journal1987625132518BrayEAZeevaartJADThe compartmentation of abscisic acid and β-D-glucopyranosyl abscisate in mesophyll cellsPlant Physiology198579719722BrzobohatýBMooreIKristoffersenPBakóLCamposNSchellJPalmeKRelease of active cytokinin by a β-glucosidase localized to the maize root meristemScience199326210511054ChengPCPareddyDRFreelingMWalbotVMorphology and development of the tassel and earThe maize handbook1994New YorkSpringer-Verlag3747ChoryJReineckeDSimSWashburnTBrennerMA role for cytokinins in de-etiolation in ArabidopsisPlant Physiology1994104339347ChristensenAHSharrockRAQuailPHMaize polyubiquitin genes: structure, thermal perturbation of expression and transcript splicing, and promoter activity following transfer to protoplasts by electroporationPlant Molecular Biology199218675689DeanJVShahRPMohammedLAFormation and vacuolar localization of salicylic acid glucose conjugates in soybean cell suspension culturesPhysiologia Plantarum2003118328336DellaportaSLCalderon-UrreaAThe sex determination process in maizeScience199426615011505DeLongACalderon-UrreaADellaportaSLSex determination gene TASSELSEED2 of maize encodes a short-chain alcohol dehydrogenase required for stage-specific floral organ abortionCell199374757768DixonSCMartinRCMokMCShawGMokDWSZeatin glycosylation enzymes in Phaseolus: isolation of O-glucosyltransferase from P. lunatus and comparison to O-xylosyltransferase from P. vulgarisPlant Physiology19899013161321DurandRDurandBMokDWSMokMCCytokinins and reproductive organogenesis in MercurialisCytokinins: chemistry, activity, and function1994Boca RatonCRC Press295304EmersonRAHeritable characters in maize. II. Pistillate flowered maize plantsJournal of Heredity1920116576FalkARaskLExpression of a zeatin-O-glucoside-degrading β-glucosidase in Brassica napusPlant Physiology199510813691377FussederAZieglerPMetabolism and compartmentation of dihydrozeatin exogenously supplied to photoautotrophic suspension cultures of Chenopodium rubrumPlanta1988173104109GanSAmasinoRMInhibition of leaf senescence by autoregulated production of cytokininScience199527019861988Garcia-MartinezJLOhlroggeJBRappaportLDifferential compartmentation of gibberellin A1 and its metabolites in vacuoles of cowpea and barley leavesPlant Physiology198168865867HernandezMLPassasHJSmithLGClonal analysis of epidermal patterning during maize leaf developmentDevelopmental Biology1999216646658HiguchiMPischkeMMähönenAPIn planta functions of the Arabidopsis cytokinin receptor familyProceedings of the National Academy of Sciences, USA200410188218826HouBLimE-KHigginsGSBowlesDJN-glucosylation of cytokinins by glycosyltransferases of Arabidopsis thalianaJournal of Biological Chemistry20042794782247832HuangSCernyREQiYBhatDAydtCMHansonDDMalloyKPNessLATransgenic studies on the involvement of cytokinin and gibberellin in male developmentPlant Physiology200313112701282InskeepWPBloomPRExtinction coefficientsPlant Physiology198577483485IrishEELangdaleTMNelsonTInteractions between tassel seed genes and other sex determining genes in maizeDevelopmental Genetics199415155171IrishEENelsonTSex determination in monoecious and dioecious plantsThe Plant Cell19891737744KasaharaHTakeiKUedaNHishiyamaSYamayaTKamiyaYYamaguchiSSakakibaraHDistinct isoprenoid origins of cis- and trans-zeatin biosynthesis in ArabidopsisJournal of Biological Chemistry20042791404914054KopečnýDTarkowskiPMajiraABouchez-MahioutINoguéFLaurièreMSandbergGLaloueMHouba-HérinNProbing cytokinin homeostasis in Arabidopsis thaliana by constitutively overexpressing two forms of the maize cytokinin oxidase/dehydrogenase 1 genePlant Science2006171114122KristoffersenPBrzobohatýBHöhfeldIBakoLMelkonianMPalmeKDevelopmental regulation of the maize Zm-p60.1 gene encoding a β-glucosidase located to plastidsPlanta2000210407415KuraishiSOkumaraFSThe effect of kinetin on leaf growthBotanical Magazine195669300306LehmannHGlundKAbscisic acid metabolism: vacuolar/extravacuolar distribution of metabolitesPlanta1986168559562LethamDSRegulators of cell division in plant tissues. XII. A cytokinin bioassay using excised radish cotyledonsPhysiologia Plantarum197125391396LethamDSPalniLMSTaoG-QGollnowBIBatesCMRegulators of cell division in plant tissues. XXIX. The activities of cytokinin glucosides and alanine conjugates in cytokinin bioassaysPlant Growth Regulation19832103115LiYHagenGGuilfoyleTJAltered morphology in transgenic tobacco plants that overproduce cytokinins in specific tissues and organsDevelopmental Biology1992153386395LoharDPSchaffJELaskeyJGKieberJJBilyeuKDBirdDMcKCytokinins play opposite roles in lateral root formation, and nematode and Rhizobial symbiosisThe Plant Journal200438203214MartinRCMartinRRMokMCMokDWSA monoclonal antibody specific to zeatin O-glycosyltransferases of PhaseolusPlant Physiology19909412901294MartinRCMokDWSMokMCDevelopment of transgenic tobacco harboring a zeatin O-glucosyltransferase gene from PhaseolusIn Vitro Cellular and Developmental Biology–Plant200137354360MartinRCMokMCHabbenJEMokDWSA cytokinin gene from maize encoding an O-glucosyltransferase specific to cis-zeatinProceedings of the National Academy of Sciences, USA20019859225926MartinRCMokMCMokDWSIsolation of a cytokinin gene, ZOG1, encoding zeatin O-glucosyltransferase of Phaseolus lunatusProceedings of the National Academy of Sciences, USA199996284289MartinRCMokMCMokDWSA gene encoding the cytokinin enzyme zeatin O-xylosyltransferase of Phaseolus vulgarisPlant Physiology1999120553557McGawBAHorganRCytokinin oxidase from Zea mays kernels and Vinca rosea crown-gall tissuePlanta19831593037MiyawakiKTarkowskiPMatsumoto-KitanoMKatoTSatoSTarkowskaDTabataSSandbergGKakimotoTRoles of Arabidopsis ATP/ADP isopentenyltransferases and tRNA isopentenyltransferases in cytokinin biosynthesisProceedings of the National Academy of Sciences, USA20061031659816603MokDWSMokMCCytokinins: chemistry, activity, and function1994Boca Rotan, FLCRC PressMokMCMartinRCMokDWSShawGKaminekMMokDWSZazimalovaECytokinin activity, metabolism and function in PhaseolusPhysiology and biochemistry of cytokinins in plants1992The HagueSPB Academic Publishers4146MokMCMartinRCDobrevPIVaňkováRHoPSYonekura-SakakibaraKSakakibaraHMokDWSTopolins and hydroxylated thidiazuron derivatives are substrates of cytokinin O-glucosyltransferase with position specificity related to receptor recognitionPlant Physiology200513710571066MokMCMokDWSArmstrongDJDifferential structure–activity relationships in PhaseolusPlant Physiology1978617275MurashigeTSkoogFA revised medium for rapid growth and bioassays with tobacco tissue culturesPhysiologia Planarum196215473497MýtnováZHaiselDWilhelmováNPhotosynthesis and protective mechanisms during ageing in transgenic tobacco leaves with over-expressed cytokinin oxidase/dehydrogenase and thus lowered cytokinin contentPhotosynthetica200644599605NickersonNHDaleEETassel modifications in Zea maysAnnals of the Missouri Botanical Garden195542195212NishimuraCOhashiYSatoSKatoTTabataSUeguchiCHistidine kinase homologs that act as cytokinin receptors possess overlapping functions in the regulation of shoot and root growth in ArabidopsisThe Plant Cell20041613651377RichmondAELangAEffect of kinetin on protein content and survival of detached Xanthium leavesScience1957125650651RieflerMNovakOStrnadMSchmüllingTArabidopsis cytokinin receptor mutants reveal functions in shoot growth, leaf senescence, seed size, germination, root development, and cytokinin metabolismThe Plant Cell2006184054RobsonPRHDonnisonISWangKFrameBPeggSEThomasAThomasHLeaf senescence is delayed in maize expressing the Agrobacterium IPT gene under the control of a novel maize senescence-enhanced promoterPlant Biotechnology Journal20042101112SakakibaraHCytokinins: activity, biosynthesis, and translocationAnnual Review of Plant Physiology and Plant Molecular Biology200657431449ScanlonMJSchneebergerRGFreelingMThe maize mutant narrow sheath fails to establish leaf margin identity in a meristematic domainDevelopment199612216831691SchmittRSandermannHJrSpecific localization of β-D-glucoside conjugates of 2,4-dichlorophenoxyacetic acid in soybean vacuolesZeitschrift für Naturforschung198237772777SkoogFArmstrongDJCytokininsAnnual Review of Plant Physiology197021359384SmartCMScofieldSRBevanMWDyerTADelayed leaf senescence in tobacco plants transformed with tmr, a gene for cytokinin production in AgrobacteriumThe Plant Cell19913647656SpíchalLRakovaNYRieflerMMizunoTRomanovGAStrnadMSchmüllingTTwo cytokinin receptors of Arabidopsis thaliana, CRE1/AHK4 and AHK3, differ in their ligand specificity in a bacterial assayPlant and Cell Physiology20044512991305StrnadMThe aromatic cytokininsPhysiologia Plantarum1997101674688SweereUEichenbergKLohrmannJMira-RodadoVBaurleIKudlaJNagyFSchaferEHarterKInteraction of the response regulator ARR4 with phytochrome B in modulating red light signalingScience200129411081111TakeiKYamayaTSugiyamaTArabidopsis CYP735A1 and CYP735A2 encode cytokinin hydroxylases that catalyze the biosynthesis of trans-zeatinJournal of Biological Chemistry20042794186641872ToJPHabererGFerreiraFJDeruereJMasonMGSchallerGEAlonsoJMEckerJRKieberJJType-A Arabidopsis response regulators are partially redundant negative regulators of cytokinin signalingThe Plant Cell200416658671TurnerJEMokDWSMokMCShawGIsolation and partial purification of an enzyme catalyzing the formation of O-xylosylzeatin in Phaseolus vulgaris embryosProceedings of the National Academy of Sciences, USA19878437143717VeachYKMartinRCMokDWSMalbeckJVankovaRMokMCO-Glucosylation of cis-zeatin in maize. Characterization of genes, enzymes, and endogenous cytokininsPlant Physiology200313113741380WernerTMotykaVStrnadMSchmüllingTRegulation of plant growth by cytokininProceedings of the National Academy of Sciences, USA2001981048710492WernerTMotykaVLaucouVSmetsRvan OnckelenHSchmüllingTCytokinin-deficient transgenic Arabidopsis plants show multiple developmental alterations indicating opposite functions of cytokinins in the regulation of shoot and root meristem activityThe Plant Cell20031525322550Yonekura-SakakibaraKKojimaMYamayaTSakakibaraHMolecular characterization of cytokinin-responsive histidine kinases in maize. Differential ligand preferences and response to cis-zeatinPlant Physiology200413416541661YoungTEGiesler-LeeJGallieDRSenescence-induced expression of cytokinin reverses pistil abortion during maize flower developmentThe Plant Journal200438910922ZhaoZGuWCaiTTaglianiLAHondredDBondDKrellSRudertMLBruceWBPierceDAMolecular analysis of T0 plants transformed by Agrobacterium and comparison of Agrobacterium-mediated transformation with bombardment transformation in maizeMaize Genetics Cooperative Newsletter1998723437
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B5F8CAA2C191B84BE158AB0DAA18E6A754871DF.txt b/test/dataset/in/resources/corpus/Clean_0B5F8CAA2C191B84BE158AB0DAA18E6A754871DF.txt
deleted file mode 100644
index 2eae479..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B5F8CAA2C191B84BE158AB0DAA18E6A754871DF.txt
+++ /dev/null
@@ -1 +0,0 @@
-amjepidajeAmerican Journal of Epidemiology1476-62560002-9262Oxford University Press10.1093/aje/kwq044ORIGINAL CONTRIBUTIONSTrajectories of Neighborhood Poverty and Associations With Subclinical Atherosclerosis and Associated Risk FactorsThe Multi-Ethnic Study of AtherosclerosisMurrayEmily T.*Diez RouxAna V.CarnethonMercedesLutseyPamela L.NiHanyuO'MearaEllen S.*Correspondence to Dr. Emily T. Murray, Laboratory of Epidemiology, Demography, and Biometry, Gateway Building, 3C309, 7201 Wisconsin Avenue, Bethesda, MD 20814 (e-mail: emily.lemelin@nih.gov).15520102742010171101099110827720091522010American Journal of Epidemiology © The Author 2010. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org.2010The authors used data from the Multi-Ethnic Study of Atherosclerosis and latent trajectory class modeling to determine patterns of neighborhood poverty over 20 years (1980–2000 residential history questionnaires were geocoded and linked to US Census data). Using these patterns, the authors examined 1) whether trajectories of neighborhood poverty were associated with differences in the amount of subclinical atherosclerosis (common carotid intimal-media thickness) and 2) associated risk factors (body mass index, hypertension, diabetes, current smoking) at baseline (January 2000–August 2002). The authors found evidence of 5 stable trajectory groups with differing levels of neighborhood poverty (∼6%, 12%, 20%, 30%, and 45%) and 1 group with 29% poverty in 1980 and approximately 11% in 2000. Mostly for women, higher cumulative neighborhood poverty was generally significantly associated with worse cardiovascular outcomes. Trends generally persisted after adjustment for adulthood socioeconomic position and race/ethnicity, although they were no longer statistically significant. Among women who had moved during the 20 years, the long-term measure had stronger associations with outcomes (except smoking) than a single, contemporaneous measure. Results indicate that cumulative 20-year exposure to neighborhood poverty is associated with greater cardiovascular risk for women. In residentially mobile populations, single-point-in-time measures underestimate long-term effects.body mass indexcarotid artery, internaldiabetes mellitushypertensionmodels, statisticalresidential mobilityretrospective studiessmokingA number of studies have reported associations of neighborhood socioeconomic characteristics with cardiovascular-related outcomes that persist after statistical adjustment for individual characteristics. For example, neighborhood disadvantage has been linked to coronary heart disease prevalence, incidence, and mortality (1–6); subclinical coronary heart disease (7); and cardiovascular risk factors (3, 4, 8–11). However, almost all of these studies are based on the measurement of neighborhood characteristics at a single point in time.Very few studies have documented what trajectories of neighborhood socioeconomics people experience in adulthood and how they are related to health outcomes (including cardiovascular outcomes) later in life. Because atherosclerosis develops over long periods, long-term neighborhood exposures may be more relevant to investigate than single-point-in-time exposures. Prior work has researched cumulative effects of neighborhood exposures over the life course by summing or averaging neighborhood exposures over a specified period of time (12, 13) or has estimated “independent” effects of exposures for different life epochs.If all persons in a population stay in the same neighborhood or move between neighborhoods of a similar socioeconomic level, then an average measure and any single measure for the specified time period would show the same association with the outcome. However, if the accumulation of neighborhood effects over the life course is affecting cardiovascular disease risk, then the use of a single-point-in-time measure may result in misestimates of area effects. Additionally, socioeconomic mobility may itself have an independent effect on cardiovascular disease health (14, 15). To test whether this is the case, cardiovascular disease health outcomes need to be compared in populations of upward- or downward-moving neighborhood poverty with what would be expected under a purely cumulative model. Prior approaches do not allow for these comparisons.We used data from the Multi-Ethnic Study of Atherosclerosis (MESA), and latent trajectory class modeling (16, 17), to determine patterns of neighborhood poverty over a 20 year period during mid- to late adulthood. We then examined how these patterns are related to the amount of subclinical atherosclerosis as well as associated risk factors later in life. We also assessed effects of mobility by comparing cardiovascular outcomes for groups with decreased (upwardly mobile) neighborhood census tract poverty over time with groups with equivalent stable cumulative poverty. In addition, we investigated the added advantage of using a cumulative measure of neighborhood poverty compared with a single measure in adulthood.MATERIALS AND METHODSStudy populationMESA was initiated to investigate the prevalence, correlates, and progression of subclinical cardiovascular disease. Participants were 45–84 years of age, from 6 US communities (Baltimore, Maryland; Chicago, Illinois; Forsyth County, North Carolina; Los Angeles County, California; northern Manhattan, New York; and St. Paul, Minnesota), clinically free of cardiovascular disease at baseline, and sampled to be ethnically diverse (∼38% white, 28% African American, 23% Hispanic, and 11% Chinese). The methods used for sampling and study design have been reported elsewhere (18).Assessment of neighborhoodsDuring one study clinic visit, participants were asked by trained interviewers to complete a 20-year residential history questionnaire (January 1980–date of visit). Each address was geocoded and assigned latitude and longitude coordinates. Census tract codes from the 1980, 1990, and 2000 US Censuses were coded to each address. For intercensal years, we interpolated the value for that year based on the 2 closest censuses. The percentage of residents living below the poverty level (i.e., neighborhood census tract poverty) for each location was obtained from the Neighborhood Change Database, which enables comparison across various census years by recalculating and normalizing past census years to 2000 US Census tract boundaries (19).We selected neighborhood census tract poverty as our primary key neighborhood variable for the following reasons: it is often used in sociologic work to characterize neighborhood conditions (20), it was measured in a standardized manner for all the US Censuses relevant to the period of study, and its definition is modified over time by the US Census to account for changes in the cost of living. Using this approach, we created a database that contains a measure of census tract poverty for each month between January 1980 and the date of the MESA baseline examination for each study participant. Only participants with complete address history information for the entire 20 years were included in the analyses.Measures of long-term patterns of neighborhood povertyTrajectory classes of neighborhood poverty over the 20 years prior to the MESA examination were identified by using hierarchical latent growth curve modeling. Individuals were assigned to trajectory classes based on their pattern of repeated measurements of annual average neighborhood poverty for each year between 1980 and 1999 (16, 17). The data trajectory for each subject consisted of repeated measurements of neighborhood census tract poverty over T time periods, Yi = (Yi1, …, YiT), that were independent given the group Ci. The trajectory was modeled aswhere μijk represents the ith response of person j in group K, β0jk is the estimated mean neighborhood poverty value for group k, where Yearij = 0 (January 1980) and Yearij denotes the number of years since January 1980 at subject j’s time i. Up to a third-order polynomial in year was included to flexibly model the effects of time. K numbers of groups were assigned so that the likelihood of observing the data trajectory for subject j, given that he or she belonged to group K, was estimated by using a censored normal model.We investigated models for 2–8 trajectory groups. The number of groups selected was based on which model as a whole had the best overall Bayesian Information Criterion (BIC) value. Group membership for a given person was assigned based on which K group had the highest likelihood of observing the data trajectory observed, given that he or she belonged to group k.In addition to categories based on trajectories of neighborhood poverty, we also investigated a measure of neighborhood percent poverty at the time of the baseline examination (referred to as the contemporaneous measure) as well as an average measure for the whole 20-year period. The average measure was the area under a continuous line fitted to each participant's data, divided by the number of months from January 1980 to the date of the baseline examination (January 2000–August 2002) (mean = 0.14, range = 0.01–0.71).Outcome variablesCommon carotid intimal-media thickness (IMT) was the measure of subclinical atherosclerosis examined. IMT is a relatively simple, inexpensive, precise, reproducible, and valid measure of early atherosclerotic changes in the carotid artery. IMT has been shown to be associated with coronary heart disease risk factors, prevalent coronary heart disease, and subsequent coronary heart disease events (21, 22). Common carotid IMT was measured noninvasively with high-resolution B-mode ultrasonography (Logia 700 ultrasound machine; General Electric Medical Systems, Little Chalfont, Buckinghamshire, United Kingdom). Common carotid IMT reflects the mean of all available maximum wall thicknesses across all scans, across both left and right sides, and across the near and far walls. Central reading of IMT was performed at the Tufts-New England Medical Center (Boston, Massachusetts) (18).We studied the following cardiovascular risk factors: body mass index (BMI), hypertension, diabetes, and current smoking. Low density lipoprotein cholesterol and high density lipoprotein cholesterol were also examined, but no associations were apparent for either gender, so results are not shown in this paper. BMI was calculated as weight (kg)/height (m)2. Resting blood pressure was measured 3 times with a Dinamap PRO 100 automated oscillometric device (Critikon Inc., Tampa, Florida). The average of the final 2 blood pressure readings was used for this analysis. Hypertension was defined by the 6th report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (23) criteria of self-reported treatment of hypertension or systolic blood pressure of ≥140 mm Hg or diastolic blood pressure of ≥90 mm Hg. Diabetes mellitus was defined by the 2003 American Diabetes Association criteria of fasting glucose ≥126 mg/dL, use of insulin or oral hypoglycemic agents, or self-reported physician diagnosis. Cigarette smoking was based on self-report and was classified into 1 of 3 categories: current, former, or never (18).CovariatesAge, gender, adult socioeconomic position (SEP), and race/ethnicity were obtained from the baseline examination interview. As in prior work (13), adult SEP was created by combining information on income, education, and wealth. Participants were asked to select their total gross family income in the past 12 months from 13 categories (collapsed into 4 for this analysis: <$25,000, $25,000–$39,999, $40,000–$74,999, or ≥$75,000). Highest educational level completed was collapsed into 4 categories: ≤high school, some college but no degree/technical school certificate, associate's or bachelor's degree, or graduate/professional degree. Wealth was assessed with the following 4 items: 1) whether the participant, or his or her family, had investments such as stocks, bonds, mutual funds, retirement investments, or other investments (yes/no); 2) whether the participant owned the home (yes/no); 3) whether the participant owned a car (yes/no); and 4) whether the participant owned land or another property that was not his or her primary residence (yes/no). A summary adult SEP score was created by summing scores for income (0–3, from lowest to highest category) and education (0–3, from lowest to highest) and adding one point for each wealth indicator present. Thus, the range of values for adult SEP was 0 to 10, with higher values indicating greater adult SEP. Race/ethnicity was classified as 1 of 4 categories: white non-Hispanic, African American non-Hispanic, Chinese, or Hispanic.Statistical analysisSociodemographic characteristics and mean common carotid IMT for the chosen categories of neighborhood poverty trajectory group were compared by using analysis of variance (continuous variables) and the chi-square statistic (categorical variables). For each analysis, linear regression was used to assess mean differences in common carotid IMT and BMI, and a generalized linear model with a binomial distribution (24) was used to estimate the relative prevalence of hypertension, diabetes, and current smoking associated with the neighborhood poverty measures. These regression models were used to estimate associations of trajectory classes with the outcomes by including trajectory classes as dummy variables in the regressions. For those in stable trajectory classes, tests for trend were conducted to examine whether higher levels of neighborhood poverty were associated with worse cardiovascular outcomes.In addition, in these same stable trajectory classes, we also contrasted associations of contemporaneous (each participant's last known address at the MESA baseline examination (January 2000–August 2002)) and average poverty with the outcomes by fitting models with each of these exposures separately and estimating the adjusted mean difference or adjusted relative prevalence of the outcome for the 90th versus the 10th percentile of each measure. The nonstable trajectory group was excluded from these analyses; because of the changing nature of poverty over time in this group, comparison of the contemporaneous and average measures is not very meaningful.Because race/ethnicity and adult SEP may be partly confounding associations of neighborhood poverty with common carotid IMT, and associated risk factors, models were fit before and after adjustment for adult SEP and race/ethnicity. Analyses were stratified by gender because of potential differences in associations of SEP with atherosclerosis between men and women.RESULTSOf the total 6,814 men and women, 5,871 (86.2%) completed a residential history questionnaire at baseline (January 2000–August 2002). Of these participants, 929 (16%) were excluded because 1) one or more addresses could not be geocoded (n = 797) or 2) poverty values were not available for one or more of the tracts in which they had lived since January 1980 (n = 132); thus, data on 4,942 participants were available for analysis. Compared with persons included in the analysis, persons excluded were significantly more likely to be younger; be Chinese or Hispanic; be in the lower income or educational categories; have lower common carotid IMT, BMI, and high density lipoprotein cholesterol; and have a lower prevalence of hypertension, although differences were generally not large.The mean age of the sample was 62.4 years; 47.6% were male; 43% were white, 30% were African American, 7% were Chinese, and 20% were Hispanic; mean BMI was 28.7 (standard deviation, 5.4); 46% were hypertensive; 14% were diabetic; and 13% were current smokers. The fit of the trajectory class models, as assessed by BIC value, improved from 2 classes (BIC = 110,809) through 6 classes (BIC = 148,513), but it worsened with 7 classes (BIC = 143,758). The model with 8 classes was a slightly better fit (BIC = 151,448), but zero participants were assigned to 2 of the classes, suggesting that the 6-class model was the best fit for the data (Table 1).Table 1.Mean Neighborhood Poverty Trajectory Values for 1980, 1985, 1990, 1995, and 2000 for the 6 Neighborhood Trajectory Groups, Multi-Ethnic Study of Atherosclerosis, United StatesNo.%19801985199019952000Medium high–low2044.10.290.270.140.100.11High stable1783.60.440.470.490.450.42Medium-high stable4709.50.310.330.340.330.31Medium stable80616.20.200.210.220.240.24Low-medium stable1,11622.70.120.130.150.150.15Low stable2,16843.90.060.060.050.060.07Participants in the low stable poverty group were significantly more likely to be male, white, and US born. These participants had the most favorable risk factor profiles of all of the groups: the lowest mean BMI and the lowest prevalence of hypertension, diabetes, and current smoking. Moving from the lowest stable group to the highest stable group, the groups were younger, included more females, included fewer whites or Chinese, included more blacks and Hispanics (with a decrease in Hispanics and an increase in blacks in the highest poverty group), were less likely to be US born, and had a higher mean BMI and a higher prevalence of hypertension, diabetes, and current smoking. The percentage of persons who moved over the 20 years decreased monotonically from a high of 56% in the low stable poverty group to a low of 42% in the high stable poverty group (Table 2).Table 2.Sociodemographic Characteristics by Neighborhood Trajectory Class, Multi-Ethnic Study of Atherosclerosis, United States, 2000–2002aLow Stable (n = 2,168)Low-Medium Stable (n = 1,116)Medium Stable (n = 806)Medium-High Stable (n = 470)High Stable (n = 178)Medium High–Low (n = 204)Mean (SD)%Mean (SD)%Mean (SD)%Mean (SD)%Mean (SD)%Mean (SD)%Baseline neighborhood povertyb,c0.06 (0.04)0.12 (0.05)0.20 (0.07)0.31 (0.08)0.44 (0.10)0.29 (0.11)20-Year average neighborhood povertyb,c0.06 (0.02)0.14 (0.02)0.22 (0.03)0.33 (0.03)0.45 (0.06)0.18 (0.05)Age, yearsb62.8 (9.7)61.9 (10.2)63.5 (10.3)62.4 (10.4)60.1 (9.0)59.1 (10.3)Maleb504746454241Moved during the last 20 yearsb565246403796No. of movesb2.1 (1.7)2.1 (1.5)1.9 (1.4)1.7 (1.2)1.6 (1.0)3.1 (1.5)Raceb White non-Hispanic6344217524 Chinese8563012 African-American non-Hispanic193143476239 Hispanic101931433325US bornb,d858271555769Married at baselineb695751464851Body mass index, kg/m2b28.1 (5.2)28.5 (5.3)29.4 (5.7)29.6 (6.0)30.0 (5.5)29.2 (5.5)LDL cholesterol, mg/dL117.4 (29.9)117.1 (32.0)117.7 (33.5)116.5 (31.7)116.2 (31.0)118.7 (33.1)HDL cholesterol, mg/dL51.5 (15.1)51.6 (15.0)50.2 (14.9)51.1 (13.5)51.1 (15.3)49.4 (13.1)CIMT, μm872.5 (195.2)870.9 (189.6)893.2 (183.6)885.7 (199.5)879.2 (193.5)833.2 (166.7)Hypertensionb444451525446Diabetesb91315172411Current cigarette smokerb101415161917Abbreviations: CIMT, common carotid intimal-media thickness; HDL, high density lipoprotein; LDL, low density lipoprotein; SD, standard deviation.aNeighborhood poverty from January 1980 to 2000: ∼6% all = low stable; ∼12% all = low-medium stable; ∼20% all = medium stable; ∼31% all = medium-high stable; ∼45% all = high stable; ∼30% to ∼10% = medium-high–low.bTest for trend significant at the 0.05 level.cThe value 0.06, for example, indicates that the mean number of households in that census block living below the poverty line is 6/100.dAll participants have been in the United States since January 1980.Only one group with a changing (i.e., not stable) neighborhood poverty trajectory was identified. This group (labeled medium high–low) began the period at a medium-high neighborhood poverty level (mean, 29%) and experienced a decline in neighborhood poverty over time (mean, 11%) (Table 1). The medium high–low group was the youngest and most female. The racial/ethnic distribution was similar to that of the medium stable group except for a larger representation of Chinese. This group, compared with the other groups, also had the lowest mean common carotid IMT, a relatively low prevalence of diabetes, and a high prevalence of current smokers. Virtually all (96%) persons in this group moved at some point over the 20-year period (Table 2).Among women, there was clear evidence that, after age adjustment, groups experiencing higher stable neighborhood census tract poverty over time had a higher common carotid IMT and BMI and more diabetes and hypertension than those in the low stable poverty group. For men, this finding was true for only diabetes—with less clear patterns. For both genders, after adjustment for race and adult SEP, only the medium-high stable group for BMI and the medium-high and high stable groups for diabetes in women had mean outcomes significantly higher than those in the low stable group. However, point estimates showed that a general trend of greater common carotid IMT, greater BMI (except for the highest poverty category), and greater prevalence of diabetes and hypertension associated with higher levels of neighborhood poverty tended to persist after adjustment, although the trends were no longer statistically significant at the 0.05 level (Figures 1–4). For men and women, stable poverty trajectory groups above the low stable group had a higher prevalence of current smoking, but a dose-response relation was not apparent (Figure 5). When we adjusted for adult SEP, results were similar when the individual SEP components of the composite score were adjusted for rather than the composite score (data not shown).Figure 1.Adjusted mean differences in common carotid intimal-media thickness (CIMT) (μm), and 95% confidence intervals, for neighborhood trajectory classes compared with the low stable group for A) males and B) females. Solid vertical line: age adjusted; dotted vertical line: age–adult socioeconomic position and race adjusted. P for trend (stable groups only)—age adjusted: A) 0.2362, B) 0.0002; age–adult socioeconomic position and race adjusted: A) 0.3451, B) 0.2568.Figure 2.Adjusted mean differences in body mass index (BMI), and 95% confidence intervals, for neighborhood trajectory classes compared with the low stable group for A) males and B) females. Solid vertical line: age adjusted; dotted vertical line: age–adult socioeconomic position and race adjusted. P for trend (stable groups only)—age adjusted: A) 0.6932, B) <0.001; age–adult socioeconomic position and race adjusted: A) 0.0022, B) 0.1043.Figure 3.Adjusted relative prevalences of hypertension, and 95% confidence intervals, for neighborhood trajectory classes compared with the low stable group for A) males and B) females. Solid vertical line: age adjusted; dotted vertical line: age–adult socioeconomic position and race adjusted. P for trend (stable groups only)—age adjusted: A) 0.0962, B) <0.001; age–adult socioeconomic position and race adjusted: A) 0.1925, B) 0.5045.Figure 4.Adjusted relative prevalences, and 95% confidence intervals, of diabetes for neighborhood trajectory classes compared with the low stable group for A) males and B) females. Solid vertical line: age adjusted; dotted vertical line: age–adult socioeconomic position and race adjusted. P for trend (stable groups only)—age adjusted: A) 0.0055, B) <0.0001; age–adult socioeconomic position and race adjusted: A) 0.4063, B) 0.0012.Figure 5.Adjusted relative prevalences, and 95% confidence intervals, of current smoking for neighborhood trajectory classes compared with the low stable group for A) males and B) females. Solid vertical line: age adjusted; dotted vertical line: age–adult socioeconomic position and race adjusted. P for trend (stable groups only)—age adjusted: A) <0.0001, B) 0.0008; age–adult socioeconomic position and race adjusted: A) 0.1904, B) 0.2246.With few exceptions, cardiovascular outcomes for the medium high–low trajectory group (category shown to the right of the stable groups in each figure) were somewhere in between those for the low-medium stable and medium stable groups (which averaged 12% and 20% neighborhood poverty levels, respectively, comparable to the medium high–low group's average of 18.5%). For men, mean common carotid IMT was lower (Figure 1) and the prevalence of hypertension was nonsignificantly higher (Figure 3) in the medium high–low group than in any of the other trajectory groups. Residential mobility over the 20 years was not associated with any of the outcomes after adjustment for age, cumulative neighborhood poverty, and race/ethnicity (data not shown).For all outcomes except smoking, associations using the contemporaneous or average poverty measure were similar for women reporting 1 address. However, for women reporting 2 or more addresses, associations were consistently stronger when the average poverty measure was used. In contrast, for current smoking, among women with 2 or more addresses, associations were stronger for the contemporaneous measure than for the average measure. There was no clear pattern for men. These analyses were restricted to the stable trajectory groups (Table 3).Table 3.Mean Differences in CIMT and Body Mass Index, and Relative Prevalences of Hypertension, Diabetes, and Current Smoking, for the 90th vs. 10th Percentile of Contemporaneousa and Average Neighborhood Poverty During a 20-Year Period (1980–2000) for Women With Stable Trajectories (n = 2,592), Multi-Ethnic Study of Atherosclerosis, United StatesOutcome and AdjustmentWomen Reporting Only 1 Address (n = 1,245)Women Reporting ≥2 Addresses (n = 1,347)ContemporaneousAverageContemporaneousAverageMean Difference95% CIMean Difference95% CIMean Difference95% CIMean Difference95% CICIMT (n = 2,561) Age31.811.0, 52.528.98.2, 49.618.5−2.3, 39.335.412.5, 58.4 Age and adult SEP32.310.4, 54.229.17.3, 50.98.1−13.9, 30.125.31.0, 49.7 Age, adult SEP, and race/ethnicity28.74.5, 52.925.01.0, 48.9−14.2−37.3, 8.9−2.6−29.1, 23.9Body mass index (n = 2,592) Age2.41.7, 3.22.31.6, 3.12.51.7, 3.33.42.5, 4.2 Age and adult SEP2.11.3, 2.92.01.2, 2.81.81.0, 2.72.71.7, 3.6 Age, adult SEP, and race/ethnicity0.8−0.1, 1.60.7−0.2, 1.50.1−0.8, 0.90.4−0.6, 1.4Relative Prevalence95% CIRelative Prevalence95% CIRelative Prevalence95% CIRelative Prevalence95% CIHypertension (n = 2,592) Age1.71.5, 2.01.61.4, 1.91.51.2, 1.82.01.6, 2.3 Age and adult SEP1.71.4, 2.01.61.3, 1.91.31.0, 1.61.71.4, 2.0 Age, adult SEP, and race/ethnicity1.20.9, 1.51.10.8, 1.40.90.5, 1.21.00.7, 1.4Diabetes (n = 2,582) Age2.21.8, 2.62.42.0, 2.82.72.3, 3.13.73.3, 4.1 Age and adult SEP1.91.5, 2.32.11.7, 2.52.11.7, 2.52.92.4, 3.3 Age, adult SEP, adult and race/ethnicity1.20.7, 1.61.30.9, 1.81.51.0, 1.91.91.4, 2.3Current smoking (n = 2,584) Age2.31.7, 2.72.11.7, 2.51.81.4, 2.11.41.0, 1.8 Age and adult SEP2.11.7, 2.51.91.5, 2.31.51.1, 1.91.20.7, 1.6 Age, adult SEP, and race/ethnicity2.21.7, 2.72.01.5, 2.41.30.9, 1.70.90.4, 1.4Abbreviations: CI, confidence interval; CIMT, common carotid intimal-media thickness; SEP socioeconomic position.aContemporaneous refers to each participant's address at the baseline Multi-Ethnic Study of Atherosclerosis examination (January 2000–August 2002).DISCUSSIONWe found evidence of several different subpopulations of neighborhood trajectory groups, with most groups experiencing approximately stable neighborhood poverty over the entire time period but differing in level of neighborhood poverty. For women, higher cumulative neighborhood poverty was generally significantly associated with worse cardiovascular outcomes in age-adjusted models, and, whereas the general trends (as indicated by the point estimates) tended to persist after adjustment for adult SEP and race/ethnicity, confidence intervals for many of the estimates were wide and trends were no longer statistically significant. No consistent association of neighborhood poverty with the outcomes was observed for men. In general, the medium high–low neighborhood poverty mobility group had outcomes comparable to those for stable groups with similar cumulative exposures. In addition, for women who moved, the long-term average measure of poverty was more strongly associated with outcomes than a single, contemporaneous measure. In contrast, the opposite was true for current smoking.Few prior studies have investigated trajectories of neighborhood poverty over time. In our sample, the relative stability of residence over the study period (47% did not move during the entire 20 years), coupled with very few changes in poverty within census tracts over time, resulted in stable trajectory groups. However, almost 50% of the stable groups had moved at least once, and correlations between poverty levels in subsequent neighborhoods were positive although not very high (Pearson's correlations of neighborhood poverty between each pre- and postmove neighborhood, 0.36). In general, our sample was less mobile than the US population of the same age (percentage who moved at least once between 1995 and 2000: US Census, 33.1%; MESA, 16.7%). The peak age for mobility is 20–29 years, with mobility decreasing with age (25). Consequently, the relative residential stability of our sample may be attributable to its age (41% were older than age 65 years) or to selection effects related to participation in a long-term study such as MESA.These results are consistent with prior analyses of other cohorts showing that cumulative neighborhood SEP is associated with higher mean common carotid IMT in white women (12), with patterns still apparent but nonsignificant after adjustment for individual SEP and race/ethnicity (12, 13). The measures of neighborhood SEP used in previous work were summaries or average measures of neighborhood socioeconomic status. In contrast, we examined distinct trajectories of neighborhood poverty over time. Potential mechanisms linking long-term neighborhood conditions to cardiovascular disease include access to resources and services that promote healthy lifestyles, as well as social features such as social norms, social support, and features of neighborhoods (such as violence or disorder) that could be stressful (2). To the extent that neighborhood poverty is a reasonable proxy for these conditions, our results may reflect cumulative effects of these exposures.Few data exist on how fluctuations in neighborhood poverty over time could translate into better or worse cardiovascular disease risk factors and outcomes for residents. It is plausible that changes in neighborhood poverty are associated with changes in specific environmental features linked to cardiovascular disease. For instance, some studies have found that lower income neighborhoods have fewer supermarkets and more fast-food restaurants than wealthier neighborhoods do (26). Persons moving from higher to lower income neighborhoods or experiencing a change in the socioeconomic features of their neighborhoods over time could also be exposed to changes in food environments related to cardiovascular disease. Measuring neighborhood poverty at a single point in time would give an incomplete assessment of what neighborhood exposures an individual had experienced over that time period and could therefore lead to underestimates of neighborhood effects.Consistent with our results, previous literature has documented a stronger effect of neighborhood deprivation on the incidence and prevalence of coronary heart disease in women than men (1, 3, 27). A number of processes could explain the stronger associations among women, including greater exposure to neighborhood conditions or greater reliance on neighborhood resources for women than for men (13). Further study is needed to examine mechanisms related to heterogeneity of neighborhood effects by gender. We found that most associations of neighborhood poverty with the cardiovascular disease risk factors were attenuated after adjustment for adult SEP and race/ethnicity. General patterns tended to remain, although trend tests were no longer statistically significant. The strong association of trajectories of neighborhood poverty with adult SEP and race/ethnicity in these data makes it difficult to isolate their “independent” contributions. In addition, adult SEP may be partly influenced by a history of exposure to neighborhood poverty. Statistical controls for race/ethnicity and adult SEP may therefore result in underestimates of neighborhood effects.One group, the medium high–low group, displayed a substantial change in neighborhood poverty over the period (dropping from 29% to 11%). This change was largely attributable to residential mobility (96% had moved over the period, and Pearson's correlations between census years for all census tracts in the Neighborhood Change Database were very high (1980–2000: 0.73, 1980–1990: 0.76, 1990–2000: 0.85). Our results with respect to cardiovascular outcomes in this group support an accumulation model of neighborhood exposure to poverty with little evidence of an independent effect of social mobility (15, 28). However, our ability to examine this trajectory group was limited by small sample size (n = 204 or 4.1% of the sample). Further study is needed to examine whether residential mobility patterns in MESA are representative of patterns of moving during mid- to late adulthood.The finding that 96% of the sample was classified into a “stable” trajectory group implies that a contemporaneous single-point-in-time measure of neighborhood poverty may accurately represent past neighborhood exposure. However, we found that, even in the “stable” trajectory groups, when women had moved during the historical period of interest, use of the long-term average measure of exposure rather than a contemporaneously measured exposure resulted in stronger associations with health outcomes that develop over long periods (such as subclinical atherosclerosis). In contrast, a current behavior such as smoking was more strongly related to current conditions than to past conditions. These results suggest that, when estimating long-term exposures in the presence of residential mobility, reliance on a single-point-in-time measure could lead to underestimates of the effects of cumulative long-term exposures to neighborhood conditions.Limitations of our study include limited trajectory groups in which neighborhood poverty changed substantially over time and retrospective collection of residential history information, which could have resulted in important misclassification of poverty exposures. The exposure to neighborhood was measured with only a single indicator, neighborhood census tract poverty, and captured only 20 years of exposure in late adulthood. Census tracts may not be the most relevant geographic unit, 20 years may not fully capture the historical time frame relevant to the outcomes we studied, and the 20 years might be capturing a different period in adulthood depending on the participant's age when he or she entered the study.Our results indicate that a summary of 20-year exposure to neighborhood poverty is associated with greater cardiovascular risk for women. They also suggest that, in many cases, single-point-in-time measures may be reasonable proxies for historical exposures, although they may underestimate neighborhood effects in the presence of important residential mobility and for outcomes that develop over long periods.AbbreviationsBICBayesian Information CriterionBMIbody mass indexIMTintimal-media thicknessMESAMulti-Ethnic Study of AtherosclerosisSEPsocioeconomic positionAuthor affiliations: Department of Epidemiology, University of Michigan, Ann Arbor, Michigan (Emily T. Murray, Ana V. Diez Roux); Department of Preventive Medicine, Northwestern University, Chicago, Illinois (Mercedes Carnethon); Department of Epidemiology and Community Health, University of Minnesota, Minneapolis, Minnesota (Pamela L. Lutsey); Epidemiology Branch, Division of Prevention and Population Services, National Heart, Lung, and Blood Institute/National Institutes of Health, Bethesda, Maryland (Hanyu Ni); and Department of Biostatistics, University of Washington, Seattle, Washington (Ellen S. O’Meara).This study was supported by grant 2R01-HL071759 from the National Heart, Lung, and Blood Institute (A. D. R., Principal Investigator). The Multi-Ethnic Study of Atherosclerosis is supported by contracts N01-HC-95159 through N01-HC-95165 and N01-HC-95169 from the National Heart, Lung, and Blood Institute.The authors thank the MESA investigators and staff for their valuable contributions.A full list of participating MESA investigators and institutions can be found at the following website: http://www.mesa-nhlbi.org.Conflict of interest: none declared.1.Diez-RouxAVNietoFJMuntanerCNeighborhood environments and coronary heart disease: a multilevel analysisAm J Epidemiol1997146148632.Diez RouxAVMerkinSSArnettDNeighborhood of residence and incidence of coronary heart diseaseN Engl J Med20013452991063.SundquistKMalmströmMJohanssonSENeighbourhood deprivation and incidence of coronary heart disease: a multilevel study of 2.6 million women and men in SwedenJ Epidemiol Community Health200458171774.SundquistKWinklebyMAhlénHNeighborhood socioeconomic environment and incidence of coronary heart disease: a follow-up study of 25,319 women and men in SwedenAm J Epidemiol200415976556625.CubbinCWinklebyMAProtective and harmful effects of neighborhood-level deprivation on individual-level health knowledge, behavior changes, and risk of coronary heart diseaseAm J Epidemiol200516265595686.StjärneMKFritzellJDe LeonAPNeighborhood socioeconomic context, individual income and myocardial infarctionEpidemiology200617114237.NordstromCKDiez RouxAVJacksonSAThe association of personal and neighborhood socioeconomic indicators with subclinical cardiovascular disease in an elderly cohort. The Cardiovascular Health StudySoc Sci Med20045910213921478.HartCEcobRDavey Smith G. People, places and coronary heart disease risk factors: a multilevel analysis of the Scottish heart health study archiveSoc Sci Med19974568939029.SmithGDHartCWattGIndividual social class, area-based deprivation, cardiovascular disease risk factors, and mortality: the Renfrew and Paisley StudyJ Epidemiol Community Health199852639940510.CubbinCHaddenWCWinklebyMANeighborhood context and cardiovascular disease risk factors: the contribution of material deprivationEthn Dis200111468770011.CubbinCSundquistKAhlénHNeighborhood deprivation and cardiovascular disease risk factors: protective and harmful effectsScand J Public Health200634322823712.CarsonAPRoseKMCatellierDJCumulative socioeconomic status across the life course and subclinical atherosclerosisAnn Epidemiol200717429630313.LemelinETDiez RouxAVFranklinTGLife-course socioeconomic positions and subclinical atherosclerosis in the Multi-Ethnic Study of AtherosclerosisSoc Sci Med200968344145114.PollittRARoseKMKaufmanJSEvaluating the evidence for models of life course socioeconomic factors and cardiovascular outcomes: a systematic reviewBMC Public Health20055715.HallqvistJLynchJBartleyMCan we disentangle life course processes of accumulation, critical period and social mobility? An analysis of disadvantaged socio-economic positions and myocardial infarction in the Stockholm Heart Epidemiology ProgramSoc Sci Med20045881555156216.NaginDSAnalyzing developmental trajectories: a semi parametric group-based approachPsychol Methods19994213915717.JonesBLNaginDSRoederKA SAS procedure based on mixture models for estimating developmental trajectoriesSociol Methods Res.200129337439318.BildDEBluemkeDABurkeGLMulti-Ethnic Study of Atherosclerosis: objectives and designAm J Epidemiol2002156987188119.Normalized data—neighborhood change database [NCDB] tract data from 1970–20002006East Brunswick, NJGeolytics Inc(http://www.geolytics.com/USCensus, Neighborhood-Change-Database-1970-2000, Products.asp). (Accessed May 19, 2009)20.KriegerNWilliamsDRMossNEMeasuring social class in US public health research: concepts, methodologies, and guidelinesAnnu Rev Public Health19971834137821.BotsMLCarotid intima-media thickness as a surrogate marker for cardiovascular disease in intervention studiesCurr Med Res Opin200622112181219022.PoredosPIntima-media thickness: indicator of cardiovascular risk and measure of the extent of atherosclerosisVasc Med200491465423.The sixth report of the Joint National Committee on PreventionDetection, Evaluation, and Treatment of High Blood PressureArch Intern Med1997157212413244624.SpiegelmanDHertzmarkEEasy SAS calculations for risk or prevalence ratios and differencesAm J Epidemiol2005162319920025.PlaneDAHenrieCJPerryMJMigration up and down the urban hierarchy and across the life courseProc Natl Acad Sci U S A200510243153131531826.MooreLVDiez RouxAVAssociations of neighborhood characteristics with the location and type of food storesAm J Public Health200696232533127.WinklebyMSundquistKCubbinCInequities in CHD incidence and case fatality by neighborhood deprivationAm J Prev Med20073229710628.Davey SmithGLynchJWKuhDBen-ShlomoYSocioeconomic differentialsA Lifecourse Approach to Chronic Disease Epidemiology2003Vol 2Oxford, United KingdomOxford University Press
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B60CE69C89F751133FE358CED2307D9E9B90004.txt b/test/dataset/in/resources/corpus/Clean_0B60CE69C89F751133FE358CED2307D9E9B90004.txt
deleted file mode 100644
index b8e569d..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B60CE69C89F751133FE358CED2307D9E9B90004.txt
+++ /dev/null
@@ -1 +0,0 @@
-jmpjmpJournal of Medicine and Philosophy1744-50190360-5310Oxford University Press10.1093/jmp/jhq048ArticlesA Transhumanist Fault Line Around Disability: Morphological Freedom and the Obligation to EnhanceBradshawHeather G.*University of Bristol, Bristol, UKTer MeulenRuudUniversity of Bristol, Bristol, UK*Address correspondence to: Heather G. Bradshaw, Centre for Ethics in Medicine, University of Bristol, 3rd Floor Hampton House, Cotham Hill, Bristol BS6 6AU, UK. E-mail: heather.bradshaw@bristol.ac.uk12201012112010356Bioethics and Transhumanism670684© The Author 2010. Published by Oxford University Press, on behalf of the Journal of Medicine and Philosophy Inc. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org2010The transhumanist literature encompasses diverse nonnovel positions on questions of disability and obligation reflecting long-running political philosophical debates on freedom and value choice, complicated by the difficulty of projecting values to enhanced beings. These older questions take on a more concrete form given transhumanist uses of biotechnologies. This paper will contrast the views of Hughes and Sandberg on the obligations persons with “disabilities” have to enhance and suggest a new model. The paper will finish by introducing a distinction between the responsibility society has in respect of the presence of impairments and the responsibility society has not to abandon disadvantaged members, concluding that questions of freedom and responsibility have renewed political importance in the context of enhancement technologies.disabilityhuman enhancementmorphological freedomnegative libertytranshumanistI.INTRODUCTIONEnhancement in the context of disability draws out areas of continuing debate and development in transhumanist ethics and political philosophy. We will introduce the fault line by contrasting the words of people with disabilities with two strands of thought in the transhumanist literature. Section II explains the fault line, following Isaiah Berlin (2002/1958), via two models of freedom. In Section III, the transhumanist conception of freedom is developed and critiqued. Sections IV and V each address objections to our preferred model of freedom in transhumanist work referring to disability. The conclusion summarizes the contextual threats of the rejected model of freedom. We turn now to the fault line.James Hughes (2004, 147) writes that we have “a duty and right to provide children with the healthiest and most able bodies we can.” He calls this “a really basic idea.”But there exist persons living with differences or disabilities who do not think they do have such an obligation to enhance themselves or their children.1 For example, Louisa, a congenitally deaf academic and linguist who is also a childhood cochlear implantee (CI) said: “[I]f for example you get a deaf baby, there is enormous pressure from doctors and society etc to take CI, because ‘it is there’ . . . whilst I think it is not because it is there, that it is automatically ‘good’ or matching the particular situation/environment of the person.”“Healthiest” and “most able” can be seen as situationally dependent (Savulescu, 2006, 324) and thus less basic than Hughes’ use suggests. This partially explains why Louisa and others living in the Deaf community (Savulescu, 2002; Atkinson, 2006; Murphy, 2008) are uncomfortable with and do not see themselves as bound by what is expressed in Hughes’ “basic idea.”2 Of course, there are others living with disabilities of various types who do want the opportunity to enhance themselves, for example, a young congenitally partially sighted man interviewed in the same research project as Louisa: “It's just … ah you know … just … just be able to see in my case.”A view from the transhumanist literature that contrasts with Hughes’ acknowledges this complexity:There clearly exist many people who deeply wish to be cured from various disabilities. But there are also many people who over time have become used to them and instead integrated them into their self-image. The investment of personal growth and determination necessary to accept, circumvent or overcome a disability is enormous. Suggesting a cure to them implies a change to themselves on a far deeper level than just “fixing” a broken tool, and quite often is experienced as an attack on their human dignity. (Sandberg, 2001, S6 Why do we need morphological freedom?, Para 9)Sandberg develops the concept of “morphological freedom” to express one transhumanist attitude to morphology or body configuration:3 “The desirability to [of] many of the possibilities allowed by morphological freedom also helps support the right to not change, as people see that they are two sides of the same coin” (Sandberg, 2001, S6, Para 4). Hughes (2004, 138) similarly writes: “People should have a right to control their own genomes and have children without permission from the government”; however, he goes on:But if eugenics includes believing that individuals, free of state coercion, should have the right to change their own genes and then have children, then the advocates of human enhancement and germinal choice are indeed eugenicists. If eugenics also includes the belief that parents and society have an obligation to give our children and the next generation the healthiest bodies and brains possible, then most people are eugenicists. (Hughes, 2004, 131)It may be that most people do believe that parents and society have such an obligation (Hughes does not give evidence) but “most” leaves out perhaps significant minorities and most people can be ethically mistaken, collectively as well as individually.Strikingly, Hughes adds with reference to Deaf people having this same “right to control their own genomes and have children”: “Physicians should refuse to accede to such a request, and public and private insurance should refuse to pay for it. When the deaf child reaches maturity they should be able to sue their parents for damages . . .” (Hughes, 2004, 140).4Sandberg, in contrast to Hughes, makes an in-principle defense of the right not to change, including the right to stay disabled or choose disability (or Deafness). He is concerned with defending radical technologies against the serious charges that they might be misused “in a coercive manner, enforcing cultural norms of normality or desirability” (Sandberg, 2001, S6, Para 3). He proposes that this extension of the concept of freedom to the control of the form of one's body is required to defend against the threat of coercive normalization from radical technologies.If it is widely accepted that we have the right to control how our bodies are changed both in the positive sense (using available tools for self-transformation) and in the negative sense of being free to not change, then it becomes harder to argue for a compulsory change (Sandberg, 2001, S6, Para 4).Sandberg and Hughes represent two diverse streams of transhumanist thought that are not yet always clearly delineated. One symptom of their divergence is the different weight they place on countering the risk that enhancement technologies and the knowledge underpinning them will be used in tyrannical or exploitative ways. That is, the weight they place on protecting freedom.5 These contentious differences about how the risks should be addressed are especially prominent in transhumanist attempts to come to terms with political, social, ethical, and economic questions involving human diversity, especially, impairment and disability. As differences over freedom are at the epicenter of the fault line, we turn now to Berlin's analysis of freedom.II.ABANDONMENT AND OPPRESSION ARE THE RISKS OF FREEDOMIn our view, the fissure in transhumanist positions (and politics) is a modern continuation of older political debates. Isaiah Berlin wrote of (but does not claim to have created the distinction between) two sorts of political or social risk arising from an institution or society's attempt to meet citizens’ psychological need for control of their own actions, freedom. Two sorts of risk will be outlined in this section then related to transhumanist through a dilemma. A potential solution and objection will take us to Section III.Abandonment is the first type of risk. Abandonment occurs when freedom is produced by society withdrawing from individual lives. We shall call this the absence model of freedom. This has been experienced in recent human history as “politically and socially destructive policies which armed the strong, the brutal and the unscrupulous against the humane and the weak, the able and ruthless against the less gifted and the less fortunate. Freedom for the wolves has often meant death to the sheep” (Berlin, 2002/1969, 38). Berlin wrote early in Europe's welfare state era. He excuses his lack of critique of the absence model by alluding to the voluminous contemporaneous literature on the harms pure capitalism causes to all but a few (Berlin, 2002/1969, 38).Oppression is the second type of political or social risk. Oppression's relation to freedom is complex. With regard to the involvement of society in individual lives in the form of the welfare state or socialism, “the case for intervention” (Berlin, 2002/1969, 38) to counter abandonment is “overwhelmingly strong” (Berlin, 2002/1969, 38). Yet in his view, the case for intervention still does not justify any model of freedom that carries the risk of oppression. Models in which your control of your actions is accepted as only partial invite others to help you increase your control. Oppression follows because, in Berlin's slippery slope argument:This renders it easy for me to conceive of myself as coercing others for their own sake, in their, not my, interest. I am then claiming that I know what they truly need better than they know it themselves. What, at most, this entails, is that they would not resist me if they were rational and as wise as I and understood their interests as I do. But I may go on to claim a good deal more than this. I may declare that they are actually aiming at what in their benighted state they consciously resist, […] Once I take this view, I am in a position to ignore the actual wishes of men or societies, to bully, oppress, torture them in the name, and on behalf, of their “real” selves […]. (Berlin, 2002/1958, 180)Such models of freedom we will call partial models (Berlin's “positive” liberty; Berlin, 2002/1969, 4). We explain their relevance to transhumanist via respect for “the person-as-they-are.” By summarizing the argument so far as a dilemma, we introduce John Christman's “autonomy” resolution and Takala's liberal objection.Berlin is clear that no matter how obvious it might appear from the outside that certain people might do better, in some sense, if they were different, they are actually not other than they are. One has to start from the person as they present themselves. To do otherwise is to oppress. One cannot start from the person one hopes another will become. To do this is to fail to recognize and respect the person-as-they-are as a morally whole person, as one who has his or her own ends in mind. It is to place one's own mind in place of theirs. “This monstrous impersonation, which consists in equating what X would choose if he were something he is not, or at least not yet, with what X actually seeks and chooses, is at the heart of all theories of political self-realisation” (Berlin, 2002/1958, 180).6The dilemma abandonment and oppression pose is emotive in disability and the transhumanist human enhancement context. The absence model risks abandoning to involuntary suffering even those who are suffering from a difference, impairment, disability or not being enhanced.7 Partial models risk oppression for those who have a difference, impairment, disability, or are not enhanced. Abandoned or oppressed? A poor choice for those already perceived as at a disadvantage!John Christman (2005) attempts a solution to the version of the dilemma posed in medical care and especially relevant in mental health care (Gutridge, 2010) by choosing a partial model but attempting to provide a barrier on the slippery slope. He argues that thinking about having “idealized,” complete, control over our actions, or “effective agency” is useful (Christman, 2005, 86, 80). We have thus discovered that autonomy is partial because… Autonomy is defined in various ways, but most conceptions stress the capacity for critical self-reflection in the development of value systems and plans of action. Such capacities do not merely emerge naturally, but must be developed through various processes involving educational, social, and personal resources. (Christman, 2005, 87)But autonomy is beneficial because it is self-limiting. Unlike other models of self-realization, in autonomy the only goal that can be imposed on one by others is the goal of being able to choose one's own goals. So even coercion in the name of autonomy can never lead to oppression and Berlin's “monstrous impersonation.”But Christman's model, as he admits all defenses of self-realization are, is based on what he calls a “cognitivist” understanding of justification: an action is justified by being the right action, not by being an action that was chosen (Christman, 2005, 84). Voluntarists, in contrast, claim that justification requires only that the action was freely chosen (Christman, 2005, 84). The catch with Christman's “cognitivist” justification is that we require some way of knowing which actions are right independently of whether people choose them or not. But how can the rightness of the action be thus abstracted from the circumstances in which it is made? In particular, how can it be separated from the nature, background, and subjective judgments of the agent that makes it?Takala (2007) develops this objection to Christman's abandonment/oppression dilemma solution. She reminds us that societies like those of the United States and Britain today do not have ways of agreeing on the right action in every ethically contentious case. This is because of cultural and value divergence:If the account of autonomy is grounded on a positive notion of liberty, it annihilates the value neutrality that respect for autonomy was supposed to protect. In multicultural societies there is no justification to build the notion of autonomy on an account of liberty that presupposes a particular value system. (Takala, 2007, 228)Christman argues that any clarification of freedom must be normative, thus taking some value as prior. In contrast, Takala argues, liberally, for separation of freedom from morality. Freedom is constituted by the space to follow and develop one's own conscience, perhaps in opposition to the ways of the surrounding society. In the transhumanist future, this question of a preset guiding value versus value-free space is exceedingly difficult due to the potentially greatly increased diversity and consequent threat to social coherence from the absence model of freedom in this context. Freedom in transhumanist thought adds a further complication: variety in the sources of impediments to freedom is greater than in Berlin's work. In Section III, both Berlin's human source of impediment and transhumanist natural sources of impediment will be discussed in order to complete and criticize the transhumanist model of freedom.III.TRANSHUMANIST FREEDOM: THE HUMAN SOURCE RESTRICTION AND THE DISPOSITIONAL ACCOUNT OF VALUESHere we discuss the human source restriction's relevance to transhumanist freedom and introduce objections from resource scarcity and obligation scope. We counter by appeal to subjective, not universal, enhancement valuations, arguing that universalization requires particularly problematic empirical evidence in the enhancement context. Here the dispositional theory of value is subject to Berlin's objections from oppression. In contrast to Bostrom, we see the absence model as providing part of the solution to the evidence deficit. We conclude by returning to disability through Sandberg's work.Christman (2005) and Garnett (2007) emphasize Berlin's “human source restriction” (Christman, 2005, 82). Berlin wrote: “Helvetius made this point very clearly: ‘The free man is the man who is not in irons, not imprisoned in a goal, nor terrorised like a slave by the fear of punishment.’ It is not lack of freedom to fly like an eagle or swim like a whale” (Berlin, 2002/1958 FN, 169). So, on Berlin's view we are not unfree when subject to traditional and contemporary human limitations, for example, lacking wings. But transhumanist is “a belief that the human race can evolve beyond its current limitations, especially by the use of science and technology” (Oxford English Dictionary, 2009), and the Transhumanist FAQ version 2.1, which summarizes the movement's values, describes it thus:(1) The intellectual and cultural movement that affirms the possibility and desirability of fundamentally improving the human condition through applied reason, especially by developing and making widely available technologies to eliminate aging and to greatly enhance human intellectual, physical, and psychological capacities.(2) The study of the ramifications, promises, and potential dangers of technologies that will enable us to overcome fundamental human limitations, and the related study of the ethical matters involved in developing and using such technologies. (Bostrom, 2003a, 1, my italics)So a transhumanist may hold that if it would be possible given available or near future technology and desirable; according to her understanding of “‘enhance,” for her to fly like an eagle, then anyone preventing her from accessing, developing, or encouraging the development of such technology is interfering with her freedom to spend her money and time and her freedom to change herself—body, brain, and mind—in accordance with her ends. Sandberg writes: “As a negative right, morphological freedom implies that nobody may force us to change in a way we do not desire or prevent our change” (Sandberg, 2001, S2, Para 11, my italics). That contradicts the human source restriction.An objection is society's lack of resources to address these new limitations to freedom as well as traditional human oppression. In reply we disagree that Sandberg's position entails a universal obligation to fund, buy, or use eagle technology. Universalization would require the significant further step that others ought to come to share that particular transhumanist's value system. A universal social obligation to enhance would require the enhancement to be likely to improve anyone's well-being, not just one's own (for individual welfarist definitions of enhancement, see Savulescu, 2006, Savulescu and Kahane, 2008; Kahane and Savulescu, 2009) as well as taking one beyond one's present state. But why should others share this value system, and how could we know if they did? Transhumanists who talk of universal enhancement obligations face a normativity problem similar to Christmans’—their justification is not value neutral and so risks Berlin's oppression rather than abandonment. In effect, they are using a partial model of freedom not an absence model.To generate a universal obligation to support an enhancement, the easiest way around our reply is to start with an enhancement which it can be assumed everyone already does value, such as reducing the incidence of the diseases of old age, as demonstrated by the Transhumanist FAQ. John Harris (2007, 36) uses a version of this strategy and defines an enhancement as such a widely valued ability.We respond first that the original objection of oppression is not addressed by seeking a universal value because it leaves no room for dissent. Second, it is still an assumption in need of evidence that these technologies are valued universally by the sorts of beings who will invest in developing them. It is not enough in this context of oppression avoidance that they ought to be so valued if the arguments supporting that ought are themselves based on a partial model of freedom or are otherwise oppressive. Third, it is a different question whether beings who actually had such technologies would still value them. And this question leads to its own set of problems, which we shall call the Chimpanzee Challenge.Not even the transhumanists have conducted large-scale surveys to determine what technological changes to the human organism are actually desired by the members of all human populations currently existing, let alone what the members of all human populations think they might desire if they were different beings, that is, beings who actually had these technologies.8 As Bostrom (2003b, 3) acknowledges, this would be difficult because “Just as chimpanzees lack the brainpower to understand what it is like to be human, so too do we lack the practical ability to form a realistic understanding of what it would be like to be posthuman”. Research with differently abled people (Bradshaw, forthcoming) shows that the change does not have to be large, or in the direction of increased ability, for a being, even a human, embodied in one way to “lack the practical ability to form a realistic understanding of what it would be like to be” differently embodied, that is, to have a different morphology. Sighted teachers lack a realistic understanding of partially sighted pupils’ phenomenological experiences; even technically well educated, oral, and hearing-friendly deaf persons may lack a realistic understanding of what life with a CI will be like. Hearing, orally communicating people generally lack a realistic understanding of what it is to be a Deaf parent.9 Most of us humans, like the chimpanzee, are just not very good at projecting what different circumstances will feel like for us let alone how these will change our values. If we were, there would be little point to literature and film.So the human “chimpanzee” seeking posthuman enlightenment will, phenomenologically, tread some very foreign roads, and visit some quite fantastic towns, and may well change value direction at each of them, seeking a destination that always eludes and having forgotten after a while where he or she came from. But none of this should stop him or her from setting out on the journey if they so wish and as long as they are aware of there being risks, including the risk of losing any sense of value direction.Even the nature of the more specific, personal, risks must be largely, though not entirely, unknown. “Here be serpents, Eve.” Not every road will lead to Utopia, and if some claim they have found it and we should follow them, how are we to know they have not just been taken in by some beguiling detour? (Bostrom, 2008; Miah, 2008, Letters from and to Utopia).But this does not constitute an objection to voluntary individual enhancement or morphological variation, for instance choosing to remain deaf, as supported by morphological freedom. Indeed, it is an argument for morphological freedom. This is because none of these risks, or the chimpanzee's necessary ignorance, should detract from the value of the journey itself, for how can we learn to navigate such waters if none of the various willing volunteers are allowed to explore them?But, to return us yet again to the question of there being a social obligation to enhance, Bostrom (2003b) goes further than the case of what one may choose for oneself and writes:Additionally, we may favour future people being posthuman rather than human, if the posthumans would lead lives more worthwhile than the alternative humans would. Any reasons stemming from such considerations would not depend on the assumption that we ourselves could become posthuman beings.We agree that the ethical force of this need not depend on the posthumans being in some way continuous with the original humans, but if there is no such connection, and the posthumans are, for example, other people's children, then the poor chimpanzee's problems are multiplied.This is because, although it is clear that I can favor, or wish, that my future life, whether as the person I am now, with my present identity and values, or with a new identity and values, will be posthuman rather than human, without risk of oppressing others (assuming the means I take to this end do not themselves harm others), it is not clear that I could favor this for someone else without risking oppressing them. There are various possibilities, such as altruism, and hope in despair, but we will show that they are not adequate.Altruistically we can wish that others’ lives will be better than our own. The hope that their children's lives will be better than their own appears to have carried many parents through dark times. To support and wish for technology that will benefit others, even if it arrives too late to help oneself, is admirable and may provide something to hope for when there is no personal hope remaining (see Bradshaw, forthcoming, for an example of this).Our objection to these two options is that they depend on a certainty that what the technology has to offer will be better for that other person. But this certainty is unavailable because of the Other Chimpanzee Challenge: If an individual cannot predict his/her own future well-being, then how can we predict well-being on behalf of another? And without certainty we have to take account of the risk of loss. If it is that the technologically modified being will have known nothing else (e.g., new morphology congenitally present), then at least they will not suffer a loss. So consider if we chose to alter a child's morphology and our imagined “better” turns out worse for them than what we can see in hindsight the original morphology would have done. Then the child will have lost nothing, in the sense of having experienced loss.10 But equally they will not experience a gain from the technology either.Such a risk of loss would, however, be present for any existing being who takes our destination as his or her own, for better or, in this case, worse. Because of this risk of loss due to the difficulty of predicting future well-being, it is important that he or she chooses of his or her own accord, and not because of anyone else's preferences so as to maximize the accuracy of the well-being forecast, to clearly assign responsibility for the consequences and to preserve agency. And for the choice to be fully theirs, it must be a real option for them to refuse the technology.This sort of refusal is, we think, what Sandberg means when he says that the opportunity to enhance needs to be countered by the “right not to change” to form morphological freedom. A true enhancement, rather than just a change, will be supported by many people's uncoerced, and retrospectively valued, choices over time. But any hint of coercion will detract from the information content of such changes as well as their psychological acceptance. Furthermore, as the proportions of people choosing one way or the other varies, so will the value of either choice. Staying the same when everyone else changes is itself to explore a subspace different from staying the same when everybody else also stays the same. And this requirement of free choice is precisely what Bostrom's formulation puts in doubt, especially when interpreted with his references to Lewis’ dispositional theory of value in mind. Bostrom writes:The conjecture that there are greater values than we can currently fathom does not imply that values are not defined in terms of our current dispositions. […] According to Lewis’ theory, something is a value for you if and only if you would want to want it if you were perfectly acquainted with it and you were thinking and deliberating as clearly as possible about it. On this view, there may be values that we do not currently want, and that we do not even currently want to want …. (Bostrom, 2003b, 4)This may be quite consistent but in terms of political philosophy it is reminiscent of Berlin's passages on the risks of positive or partial models of freedom (Berlin, 2002/1958, 180). For Lewis and Bostrom are here suggesting that a person's set of present values can be split into the values they are aware of and the values they are unaware of. Berlin claims such partial consciousness entails only that if the subject were (i) rational, (ii) “as wise as I,” and (iii) “understood their interests as I do,” then they “would not resist me.” It does not imply that present resistance is not significant! Nor that such people “are actually aiming at what in their benighted state they consciously resist” (Berlin, 2002/1958, 180). Berlin sees a great risk of oppression in attributing such currently unwanted or occluded values to those who are obviously not presently in the ideal state. “Once I take this view,” he writes, “I am in a position to ignore the actual wishes of men or societies, to bully, oppress, torture them in the name, and on behalf, of their ‘real’ selves” (Berlin, 2002/1958, 180). And that, when connected with the sorts of technologies in question, is more than enough to generate, and realize, the fears Sandberg refers to that “technologies such as genetic modifications would be used in a coercive manner, enforcing cultural norms of normality or desirability” (Sandberg, 2001, S6, Para 3).Sandberg connects his idea of morphological freedom as a defense against such risks with the experience of those living with disabilities and facing present technological choices. As they are the group for whom the obligation to enhance is often thought to be strongest, but rest lightest, it is a good test case. He writes:A simple ban of coercive medical procedures would not be enough, even if it is better than nothing. The reason is that it does not imply any right to have an alternative body or protect differently bodied people. The official [a bureaucrat considering that being disabled is a very expensive lifestyle] could encourage “normal” bodies through various means, including officially pronouncing disabled people who did not change as irresponsible and wasting public resources. Without any protection of the right to have a different body, both in the legal sense to prevent discrimination and in the ethical sense as a part of public ethics guiding acceptance and tolerance, the disabled would be in a very disagreeable situation. It should be noted that the disability movement have been strong supporters of right to determine ones body just for this reason. (Sandberg, 2001, S6, Para 11)Transhumanist pushes the boundaries of the human source constraint by attributing to human actions or omissions one's bodily inability to perform posthuman actions. Thus, transhumanist reiterates in a particularly concrete form the questions of freedom and oppression wrestled with by Berlin.IV.MATERIALIZATION OF THE SELF AND TWO RESPONSIBILITIES: A NEW DISTINCTIONThis transhuman materialization precipitates the hitherto abstract “self” to be realized into a material, though malleable, body or brain. Then the suffering of the “sheep” becomes not an act of fate beyond human power (as it was considered in the early lassaiz-faire period) but a human responsibility because of our matter manipulating technology. This apparently strengthens the argument against the absence model of freedom under which not only would the disadvantaged be abandoned but also now their disadvantage as well as their abandonment could be directly attributed to others’ omissions.11But now there are two responsibilities: the disadvantage responsibility and the non-abandonment responsibility. We may reason that if we can ameliorate disadvantage, we should, to fulfill our disadvantage responsibility. But exactly this benevolent (though self-serving) intent also leads to coercion and oppression. Consider instead who is relieved of this responsibility when the person concerned voluntarily accepts it, for example, when a Deaf person freely chooses to remain Deaf even when speech reception technology is available to them. This transfer of responsibility is how morphological freedom manages disadvantage. Society has no further obligation to “cure,” enhance, or encourage self-enhancement for those disadvantaged by freely chosen morphologies.We do retain a strong obligation to continue with enhancement research for the benefit of others with the same disadvantage who, voluntarily, do not choose to retain it. Our non-abandonment responsibility is unaffected by either choice and implies support for the ends of those who choose to retain a disadvantage just as we support the ends of other members of society within the constraints of just resource allocation.In our view, there is also a further set of constraints on morphological freedom. They can be derived from consideration of a model society within which morphologically homogenous, economically independent subgroups interact in a structured way supported by a neutral state. In this model, there is no single “basic cooperative framework” (Buchanan et al., 2000, 20, 288–303). Instead, multiple cooperative frameworks, each representing different combinations of physical, social, and morphological factors, are coordinated through collectively financed state institutions.In this model, the efficiencies of morphological similarity are exploited within the subgroups, but technology and new political institutions ensure that the benefits of morphological diversity are available, collectively to the whole society and personally to individual members wishing to explore and take up other morphologies or lifestyles. Freedom to found new subgroups helps the society to explore a variety of morphology-environment fits. It should be noted that constraints on acceptable morphologies will still be present in this model, but they will be far fewer than at present in our single basic cooperative framework societies. Such multiple cooperative framework societies seem capable of supporting much greater levels of morphological freedom while minimizing the risks of both abandonment and tyranny.V.CONCLUSIONThrough the lens of transhumanist's attempts to address disability, we have glimpsed one of the schisms fundamental to the movement's history and future. The fault line lies between freedom as the absence of interference with people's existing ends—a pluralist, voluntarist, liberal conception, compatible with disability studies’ models of disability—and the less compatible, partial—monistic, rationalist, objectivist—conception of freedom. This conceptual gap has not grown with the advent of greater understanding of human psychology, neuroscience, or biotechnology; it has just become more material.Commercialization of recent technological innovations may make the political importance of this schism greater than at any time since Berlin wrote. Transhumanist literature such as Hughes (2004), Stock (2003), and Naam (2005) demonstrates that sociopolitical understanding is still stymied by humans’ inability to imagine how morphological fluidity (MacKenzie, 2008, 399) will affect their identities and value systems (Scully, 2008). Subjective experience, by analogy or experimentation, may be the only way out of the impasse. Meanwhile, reducing the risks posed by oppression, tyranny, despotism, abandonment, and the use of force against discontent requires more diligent study of existing reports, theory, and history of value change. Attention to available data may yet enable technological, political, and social progress to continue without war's destruction, fueled by perceived oppression, and conducted with the full force of our crude, present or near future, levels of technological understanding.AtkinsonRI hoped our baby would be deafThe Guardian2006BaumanH.-D. LDesigning deaf babies and the question of disabilityJournal of Deaf Studies and Deaf Education2005103115BerlinINHardyHTwo concepts of libertyIsaiah Berlin Liberty1958Oxford, UKOxford University Press166217———HardyHIntroduction to five essays on libertyIsaiah Berlin Liberty1969Oxford, UKOxford University Press355BostromNTranshumanist FAQ version 2.1 [On-line]2003World Transhumanist AssociationAvailable: http://www.transhumanist.org/resources/FAQv21.pdf (Accessed December 29, 2008)———Human genetic enhancements: A transhumanist perspectiveJournal of Value Enquiry200337493506———Letter from UtopiaStudies in Ethics, Law, and Technology200827BradshawHGDefining enhancement, disability and therapy: how technology affects identity and the ethical implications of thisPhD Dissertation, University of Bristol, forthcomingBuchananABrockDDanielsNWiklerDEFrom chance to choice: Genetics and justice2000CambridgeCambridge University PressChristmanJSaving positive freedomPolitical Theory: An International Journal of Political Philosophy2005337988GarnettMIgnorance, incompetence and the concept of libertyJournal of Political Philosophy20071542846GutridgeKSafer self-injury or assisted self-harm?Theoretical Medicine and Bioethics2010317992HarrisJEnhancing evolution: The ethical case for making better people2007Princeton, NJPrinceton University PressHughesJCitizen cyborg: Why democratic societies must respond to the redesigned human of the future2004Cambridge, MAWest View Press, Perseus Books GroupKahaneGSavulescuJBrownleeKCuretonAThe welfarist account of disabilityDisability and disadvantage2009Oxford, UKOxford University Press1452MacKenzieRSomatechnics of medico-legal taxonomies, elective amputation, transableism and functional somatic syndromesMedical Law Review200816123MiahALetter to Utopia: A reply to BostromStudies in Ethics, Law and Technology200827MurphyFDraft letter to Professor Marcus Pembrey2008 [On-line]. Available: http://www.grumpyoldeafies.com/2007/11/hfeb_bda_draft_letter_to_profe.html (Accessed October 8, 2009)NaamRMore than human: Embracing the promise of biological enhancement2005New YorkBroadway BooksOxford English Dictionary onlineOxford, UKOxford University PressAvailable: http://www.oed.com/ (Accessed October 7, 2009)ParfitDReasons and persons1984Oxford, UKOxford University PressSandbergAMorphological freedom—Why we not just want it, but need it2001[On-line]. Available: http://www.nada.kth.se/∼asa/Texts/MorphologicalFreedom.htm (Accessed February 26, 2009)SavulescuJDeaf lesbians, “designer disability” and the future of medicineBritish Medical Journal20023257713Sims BainbridgeWRoccoM———Justice, fairness and enhancement Progress in convergenceAnnals of the New York Academy of Science2006109332138SavulescuJBostromNHuman enhancement2009Oxford, UKOxford University PressSavulescuJKahaneGThe moral obligation to create children with the best chance of the best lifeBioethics20082327490ScullyJ. LDisability bioethics: Moral bodies, moral differences2008Lanhan, MDRowman and LittlefieldStockGRedesigning humans: Choosing our genes, changing our future2003Boston, MAMariner BooksTakalaTConcepts of “Person” and “Liberty,” and their implications to our fading notions of autonomyJournal of Medical Ethics20073322581For the purposes of this paper, we will consider enhancement to include any intervention that takes an individual beyond the level they have previously experienced for that ability. Therapy, in contrast, returns an individual to an ability level they had previously experienced. In the case of children, for the purposes of this paper we will take enhancement to be any intervention on the part of the parents to deliberately produce offspring who have greater abilities than either of the parents do or to increase the probability with which the offspring are likely to carry a particular trait that at least one of the parents has experience of. We would consider it unethical to deliberately cause offspring to carry traits considered to be detrimental to life success by the parents and which neither parent has personal experience of. This leaves open the risk that some traits considered by the parents to be enhancing will prove to be detrimental. It also allows parents who live with a trait considered undesirable by those without it to deliberately pass this trait on to their children.2And enshrined in UK law at present in the form of the illegality of implanting embryos known to have an “abnormality” which involves a “significant risk” of the development of a “serious” “disability,” “illness,” or “any other serious medical condition.” (Human Fertilisation and Embryology Act 2008 c.22. Clause 14 Subsection (9)). People identifying as Deaf do not consider Deafness to be a medical condition of any sort and certainly not a disability or illness.3Sandberg was the main author of Version 1 of the Transhumanist FAQ. See S7, Acknowledgments and Document History, within Bostrom (2003a).4Later the same page he acknowledges that the benefits of respect for reproductive freedom would allow us to ignore the few such cases of “bizarre” choices because the resulting harm done would be small in comparison to the overall benefits of reproductive freedom. This seems ad hoc and in need of empirical confirmation though we agree the numbers involved may be quite small.5Neither they nor we feel that these fears and risks justify preventing the development of technologies with such potential for ameliorating involuntary suffering and increasing achievement. Indeed, we would all argue strongly for more support for research in these areas.6Transhumanist is often seen as a theory of physical self-realization.7Despite common speech not everyone affected suffers from such differences.8Some such work has been done. See Savulescu and Bostrom (2009, 8).9We use Deaf where a clear allegiance to the Deaf community is known to be present and deaf where such an allegiance cannot be assumed. We also differentiate between hard of hearing but orally oriented and deaf and/or Deaf in the sense of nonorally oriented persons who may feel a greater or lesser allegiance to the Deaf community.10This is a version of Derek Parfit's nonidentity problem. Parfit (1984, 351–90).11We are using “disadvantage” here to refer to those whose competitive advantage may be considered low in their society, for example, those with body morphologies today considered “impaired” or those who are not “enhanced.” That is, those with a smaller range of abilities or a lower level of key abilities than is usual in their society. We would like to contrast this sense of “disadvantaged” with what is implied by “disabled” in the social model of disability—the negative effects of an impairment that are due not to the impairment itself but to the social circumstances in which the impaired individual finds themselves. The latter have always been within society's power to change and thus have always been the responsibility of society to ameliorate. But the impairment was until now often accepted as being beyond human power to alter. That is no longer the case.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B6423DC51059B6040DE8A5A383991C474FAB3B6.txt b/test/dataset/in/resources/corpus/Clean_0B6423DC51059B6040DE8A5A383991C474FAB3B6.txt
deleted file mode 100644
index d2d1017..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B6423DC51059B6040DE8A5A383991C474FAB3B6.txt
+++ /dev/null
@@ -1 +0,0 @@
-]>MAD1887S0047-6374(97)01887-310.1016/S0047-6374(97)01887-3Elsevier Science Ireland LtdFig. 1Proliferative responses of PBMC from 14 centenarians, 10 middle-aged and 16 young controls to soluble anti-CD3 mAb (25 ng/ml) or PMA (5 ng/ml).Fig. 2Proliferative responses of purified T cells from 14 centenarians, 10 middle-aged and 16 young controls to PMA (5 ng/ml); PMA added to immobilized anti-CD28 mAb, immobilized anti-CD3 mAb or to coimmobilized anti-CD3 plus anti-CD28 mAb.Table 1Cytofluorimetric analysis of T lymphocytes from centenarian, middle-aged and young subjectsCentenariansMiddle agedYoung adultsMean age and number of subjects101.1±0.4 (10)63.5±1.9 (17)28.8±0.9 (24)% of CD28+ among T cells50.5±3.5**77.3±3.8**90.25±1.4Number of CD28+ T cells per μl583±47**1138±70**1558±47CD28 median fluorescence channel in linear scale385±13.9358±21372±18Data are reported as mean values±S.E.M.PBMC were double stained with FITC-conjugated mAb anti-CD3 and PE-conjugated mAb anti-CD28 as described in Section 2.Analysis of CD28 expression was obtained by Lysis II® software.Statistical analysis was done by using the ANOVA Fischer test.**P<0.01 vs. either of the two other groups.T lymphocyte proliferative capability to defined stimuli and costimulatory CD28 pathway is not impaired in healthy centenariansPaoloSansonia*FrancescoFagnoniaRosannaVescoviniaMarcoMazzolaaVincenzoBriantiaGiovanniBolognaaEnricaNigroaGiampaoloLavagettoaAndreaCossarizzabDanielaMontibClaudioFranceschibcMarioPasseriaaIstituto di Clinica Medica Generale e Terapia Medica, University of Parma, via Gramsci 14, Parma 43 100, ItalybDipartimento di Scienze Biomediche, Sezione di Patologia Generale, University of Modena, via Campi 287, Modena 41 100, ItalycINRCA, via Birarelli, 8, 60 100 Ancona, Italy*Corresponding author. Tel.: +39 521 290783/4; fax: +39 521 290776.AbstractIt is generally assumed that T cell proliferation is impaired in aged individuals. We report data on the proliferative capability of peripheral blood mononuclear cells (PBMC) and T lymphocytes from 40 healthy people of different ages, (19–107 years), including 14 centenarians, to defined mitogenic stimuli. We observed no age-related proliferative impairment both in PBMC and in purified T cells stimulated by anti-CD3 mAb or phorbol myristate acetate (PMA). Furthermore, T cells stimulated by anti-CD3 mAb or PMA and costimulated by CD28 mAb did not proliferate differently among young, middle aged subjects and centenarians. Thus, short term T cell proliferation is not affected even at extreme age when well defined stimuli are used on cells deriving from carefully selected healthy subjects.KeywordsCentenariansT cellsCD28Proliferation1IntroductionAn old tenet in medicine is that aging is accompanied by a deterioration of the immune system that could contribute to increased risk of morbidity and mortality [1–4]However, the study of carefully selected healthy aged subjects showed that most immune functions are well preserved even in advanced age [5, 6]. We have shown that several immune responses are well preserved in healthy centenarians in comparison to middle-aged and young subjects[7, 8].A very important function which has been suggested to deteriorate with age and to play a major role in the aging process is the capability of cells from aged subjects to respond to mitogenic stimuli and, consequently, to undergo cell proliferation [9]. A defect in this function is very important for immune cells, whose capability to undergo clonal expansion is critical to set up an effective response to antigenic stimuli [10]. Accordingly, we thought it worthwhile to assess the proliferative capability of peripheral blood mononuclear cells (PBMC) and purified T lymphocytes from healthy centenarians in comparison to middle-aged and young subjects. Our results demonstrate that this function is well preserved when defined stimuli, such as anti-CD3 mAb or phorbol 12-myristate acetate (PMA), are used. This conclusion is further supported by experiments showing that the costimulatory pathway via the CD28 molecule is well preserved in centenarians.2Materials and methods2.1Subjects.We studied a total of 14 centenarians with a mean age of 100.8±0.3 years (range, 100–107 years), 3 males and 11 females. All subjects were in relatively good clinical condition without relevant acute or chronic disease affecting the immune system and mentally competent to give informed consent. In particular, none of the subjects had cancer, were suffering from serious cardiac, brain or kidney disease or taking drugs known to affect the immune system. Along with a complete social and medical history, we performed a physical examination. As control groups we studied 10 healthy middle-aged subjects with a mean age of 58.2±2.1 years (range, 48–68 years) and 16 healthy young donors with a mean age of 28.1±0.6 years (range, 19–36 years). All these control subjects were selected according to the SENIEUR protocol [5, 6].2.2Monoclonal antibodies.MAb to CD3 (OKT3) was generously provided by Dr E. Engleman, Stanford University (Stanford, CA); anti-CD8 conjugated with fluorescein (FITC), anti-CD28 conjugated with phycoerythrin (PE) and isotype-matched control mAb were purchased from Becton Dickinson (San José, CA).2.3Cell preparation and enrichment of T lymphocytes.PBMC were obtained by Ficoll-Hypaque gradient centrifugation from freshly drawn venous blood collected around 08:00 from a centenarian, a middle-aged and a young control on the same day. After washings with PBS, the cells were suspended in RPMI-1640 with 10% AB serum, 2mM l-glutamine, 100 μg/ml streptomycin and 100 U/ml penicillin, hereafter referred to as complete medium. PBMC were fractionated into T and non-T cells by a single step rosetting method [11]. T cells were separated from sheep red blood cells by hypotonic lysis of the latter. T cells were further depleted of monocytes by adherence (1 h) to plastic. The purity of T cells preparation was always >95% as assessed by cytofluorimetric analysis.2.4Proliferation assays.All proliferation assays were performed in round-bottomed microtitre wells in a final volume of 0.2 ml complete medium. Stimulation of 1×105 PBMC with soluble anti-CD3 mAb (25 ng/ml) or 5 ng/ml PMA was carried out for 3 days at 37°C in 6% carbon dioxide–air. Purified T cells were stimulated with immobilized mAb as follows: 100 μl anti-CD3 (0.1 μg/ml) and/or 100 μl anti-CD28 (5 μg/ml), diluted in PBS, were placed in 96 wells microtiter plates and incubated at room temperature overnight and then washed with PBS. T cells were challenged with several stimuli or a combination of them: immobilized anti-CD3; immobilized anti-CD3+immobilized anti-CD28; 5 ng/ml PMA; PMA+immobilized anti-CD28 as specified in the figure legends. Assays were performed in triplicate and 0.5 μCi [3H]thymidine ([3H]TdR) was added to each well 6 h before cell harvesting on glass fiber filter paper. Uptake of [3H]TdR was measured in a liquid scintillation counter and the results expressed as the mean counts per min±S.E.M. (cpm±S.E.M.).2.5Cytofluorimetric analysis.Cytofluorimetric analysis was performed on PBMC following standard methods [12]. Briefly, 1×106 PBMC were incubated with 1 μg anti-CD8 FITC and anti-CD28 PE for 20 min at 4°C, washed with PBS and suspended in 200 μl PBS until analysis. Two colors FACS analysis was performed on a FACScan cytofluorimeter (Becton-Dickinson, CA) as previously described [12]. The expression per cell of CD28 molecule was calculated by flow cytometry using PE-conjugated anti-CD28 mAb, considering the mean fluorescence channel of each histogram, as described [13].2.6Statistical analysis.Statistical analysis was performed by ANOVA and Fisher test by SPSS for Windows® software. P<0.05 was considered significant.3Results3.1Proliferative responsiveness of PBMC and purified T cells.When PBMC were stimulated by defined mitogenic stimuli, such as PMA or anti-CD3 mAb, no significant difference concerning the proliferative capability was observed among cells from young, middle-aged and centenarians (Fig. 1). PBMC are mainly composed of T and B lymphocytes, NK cells and accessory cells such as monocytes. It is well known that optimal T cell responsiveness requires a primary and a costimulatory signal that usually is delivered by accessory antigen presenting cells through the B7 molecule which is capable of interacting with the CD28 molecule on the T cell membrane [14]. Accordingly, purified T cells from young, middle aged and centenarians were stimulated by the above mentioned primary stimuli (PMA or anti-CD3 mAb) and costimulated by anti-CD28 mAb in order to activate the most important costimulatory pathway in T cells [15]. Fig. 2 shows that costimulation via anti-CD28 mAb significantly increased T cell responsiveness to PMA or anti-CD3 mAb in cells from all groups including centenarians and again no significant age-related difference was found.3.2CD28 expression.Since the target of anti-CD28 mAb is the CD28 molecule on the T cell membrane we performed a cytofluorimetric analysis of resting lympocytes in order to study the expression of this molecule on T lymphocytes from centenarians, middle-aged and young controls.The results indicate that the costimulatory capability of anti-CD28 mAb is not likely to be related to the percentage of CD28+ cells in culture. Indeed Table 1 shows that the percentage of CD28+ cells is significantly reduced in centenarians, whose cells, however, responded well to costimulation with anti-CD28 in the presence of PMA or anti-CD3 mAb (Fig. 2). Table 1 also shows that, notwithstanding the decrease in the percentage of CD28+ lymphocytes, the expression of CD28 molecules per cell was, however, unchanged.4DiscussionPrevious studies indicate that a profound remodelling of the immune system occurs with age [7]. In particular we have reported that, in the peripheral blood of aged people including centenarians, there is a progressive increase in serum level of IgA, IgG1, IgG2, IgG3 but not of IgM or IgG4 [16], a decrease in absolute number of T lymphocytes (CD3+) involving both CD4+ and CD8+ subsets, a marked decrease of B lymphocytes and an increase of cells with natural killer (NK) markers [8, 17]. We have also found that healthy elderly and centenarians are almost free of organ-specific autoantibodies [18, 19]and equipped with very well preserved cytotoxic activities (NK, anti-CD16 and anti-CD3 redirected killing activities) [8, 17, 20].An important and extensively studied immune function is lymphocyte capability to undergo cell proliferation. Several reports indicate that this function is impaired with age [1, 2]. The reason for this defect has been variably referred to among T cell subsets, an imbalance [21], a reduced proliferation of CD8+ T cells [22], a failure to produce IL-2 [23]or to a decreased number of cells capable of completeing the S phase of the cell cycle [24]. However, most of the previous studies refer to the capability of PBMC to proliferate when stimulated with PHA, a mitogen used to assess such a function since 1960 but whose fine mechanism is not completely understood [25]. The available data indicate that PHA induces activation of T cells in a complex way through cross-linking of several surface structures, including CD2, CD3, CD5 molecules [26, 27]and perhaps others. Thus, we thought it worthwhile to study the proliferative capability of either PBMC or purified T cells from people of different ages using more defined stimuli acting at different levels of the signal transduction cell machinery, i.e. plasma-membrane (anti-CD3) or transmembrane (PMA). When this approach was used no proliferative defect was evident with both stimuli in cells from any group, including people who are close to the maximum life span such as subjects over 100 years of age. These data are in accord with earlier reports in unseparated PBMC from subjects aged between 70 and 82 years [28, 29].We tried to further dissect the pathways known to be involved in lymphocyte proliferation and particularly the main costimulatory T-cell pathway via the CD28 molecule. To this end we used anti-CD28 mAb to act as surrogate for this most important contribution by accessory cells to T cell activation [14]. This system has the advantage of optimizing the lymphocyte proliferative conditions, indeed the ligation of CD28 molecules increases the expression of IL-2 receptors and IL-2 availability [15, 30, 31]which, in turn, are thought to be defective in lymphocyte cultures from elderly subjects [23, 32]. Under these conditions, costimulation via CD28 significantly increased T lymphocyte proliferation in middle-aged and young controls, as well as in centenarians, and this phenomenon was of the same extent in the three groups.According to the two signal theory, T lymphocyte activation derives from MHC-peptide specific engagement of TCR-CD3 and contestual crosslinking of non antigen-specific accessory molecules such as CD28 [33, 34]. Our data indicates that both CD3 and CD28 pathways are well preserved in T cells from aged people including centenarians. Our results suggest that the cell machinery responsible for the transduction of primary and costimulatory signals is functioning well even in cells from these subjects of very advanced age. The results presented are in apparent contrast with the data in the literature indicating that lymphocytes from elderly people show a proliferative defect when complex stimuli such as PHA [2]or alloantigens are used [35]. The defective response to PHA could be related to the complexity of this stimulus (proliferation of different T cell subsets? Activation-induced cell death? Induction of cell nonresponsiveness?).The results shown here suggest that, when cells from carefully selected aged subjects are used and stimulated with well defined stimuli T lymphocyte proliferation is not impaired as previously thought and that some important activation pathways, such as those involving CD3, are well preserved into the last decades of life. Furthermore, the efficient costimulation through CD28 molecules in PMA and in anti-CD3 stimulated cultures indicate that the costimulation through the CD28 pathway is not impaired in centenarians despite the reduced number of CD28 positive cells with age.About 1/3 of centenarians, despite their very advanced age, are still in good mental and physical condition. These subjects are the best example of successful aging, being free of the major age-related diseases. Therefore, healthy centenarians are a model of human longevity and constitute a very select group of exceptional individuals. The current observation of a well preserved T cell proliferative capability involved in adaptive immune responses, together with previous observations concerning well preserved innate immunity, suggests that such a well equipped immune system may contribute to longevity in human and, on the other hand, indicate that the immune system of healthy centenarians is still apparently capable of coping with infectious diseases.In summary, we suggest that a generalized age-related defect in the cell proliferative machinery can be excluded and that important activation pathways are intact in lymphocytes even in the last decade of life. The data presented here fit the hypothesis that aging is not characterized by unidirectional deterioration but by a complex remodelling of immune functions [7, 36].AcknowledgementsThis work was supported by the National Research Council (CNR) target project `Ageing', M.U.R.S.T. (40 and 60%) to PS and CF and from Fondazione Cassa di Risparmio di ParmaReferences1T.Makinodan and M.M. Kay, Age influence on the immune system. Adv. Immunol., 29 (1980) 287–330.2D.M. Murasko, B.J. Nelson, R.Silver, D.Matour and D. Kaye, Immunologic response in an elderly population with a mean age of 85. Am. J. Med., 81 (1986) 612–618.3D.M. Murasko, P. Weiner and D. Kaye, Decline in mitogen induced proliferation of lymphocytes with increasing age. Clin. Exp. Immunol., 70 (1987) 440–448.4M.J. Hicks, J.F. Jones, A.C. Thies, K.A. Weigle and L.L. Minnich, Age-related changes in mitogen-induced lymphocyte function from birth to old age. Am. J. Clin. Pathol., 80 (1983) 159–163.5G.J. Ligthart, J.X. Corberand, C. Fournier et al., Admission criteria for immunogerontological studies in man: the SENIEUR protocol. Mech. Ageing Dev., 28 (1984) 47–55.6G.J. Ligthart, J.X. Corberand, H.G. M. Geertzen, A.E. Meinders, D.L. Knook and W. Hijmans, Necessity of the assessment of health status in human immunogerontological studies: evaluation of the SENIEUR protocol. Mech. Ageing Dev., 55 (1990) 89–105.7C. Franceschi, D. Monti, P. Sansoni and A. Cossarizza, The immunology of exceptional individual: the lesson of centenarians. Immunol. Today, 16 (1995) 12–16.8P. Sansoni, A. Cossarizza, V. Brianti et al., Lymphocyte subsets and natural killer activity in healthy old people and centenarians. Blood, 82 (1993) 2767–2773.9V.J. Cristofalo and R.J. Pignolo, Replicative senescence of human fibroblast-like cells in culture. Physiol. Rev., 73 (1993) 617–638.10C.A. J. Janeway and K. Bottomly, Signals and signs for lymphocyte responses. Cell, 76 (1994) 275.11A. Saxon, J. Feldhaus and R.A. Robins, Single step separation of human T and B cells using AET treated srbc rosettes. J. Immunol. Methods, 12 (1976) 285–288.12A. Cossarizza, C. Ortolani, E. Forti et al., Age-related expansion of functionally inefficient cells with markers of natural killer activity in Down's syndrome. Blood, 77 (1991) 1263–1270.13G. Girolomoni, G. Zambruno, R. Manfredini et al., Expression of B7 costimulatory molecule in cultured human epidermal Langerhans cells is regulated at the mRNA level. J. Invest. Dermatol., 103 (1994) 54–59.14P.S. Linsley, P.M. Wallace, J. Johnson et al., Immunosuppression in vivo by a soluble form of the CTLA-4 T cell activation molecule. Science, 257 (1992) 792–795.15R.H. Schwartz, Costimulation of T lymphocytes: the role of CD28, CTLA-4, and B7/BB1 in interleukin–2 production and immunotherapy. Cell, 71 (1992) 1065–1068.16R. Paganelli, I. Quinti, U. Fagiolo et al., Changes in circulating B cells and immunoglobulin classes and subslasses in a healthy aged population. Clin. Exp. Immunol., 90 (1992) 351–354.17P. Sansoni, V. Brianti, F. Fagnoni et al., NK Cell Activity and T-lymphocyte proliferation in healthy Centenarians. Ann. N.Y. Acad. Sci., 663 (1992) 505–507.18S. Mariotti, P. Sansoni, G. Barbesino et al., Thyroid and other organ-specific autoantibodies in healthy centenarians. Lancet, 339 (1992) 1506–1508.19S. Mariotti, G. Barbesino, P. Caturegli et al., Complex alterations of thyroid function in healthy centenarians. J. Clin. Endocrinol. Metab., 77 (1993) 1130–1134.20F.F. Fagnoni, R. Vescovini, M. Mazzola et al., Expansion of cytotoxic CD8+ CD28- T cells in healthy aged people and centenarians. Immunology, 88 (1996) 501–507.21R.A. Miller, Accumulation of hyporesponsive, calcium extruding memory T cells as a key feature of age-dependent immune dysfunction. Clin. Immunol. Immunopathol., 58 (1991) 305–317.22A. Grossmann, J.A. Ledbetter and P.S. Rabinovitch, Reduced proliferation in T lymphocytes in aged human is predominantly in the CD8+ subset, and is unrelated to defects in transmembrane signaling which are predominantly in the CD4+ subset. Exp. Cell Res., 180 (1989) 367–382.23S. Gillis, R. Kozak, M. Durante and M.E. Weksler, Immunological studies of aging. Decreased production of and response to T cell growth factor by lymphocytes from aged humans. J. Clin. Invest., 67 (1981) 937–942.24L. Staiano-Coico, Z. Darzynkiewicz, M.R. Melamed and M.E. Weksler, Immunological studies of aging. IX. Impaired proliferation of T lymphocytes detected in elderly humans by flow cytometry. J. Immunol., 132 (1984) 1788–1792.25P.C. Nowell, Phytohemagglutinin: an initiator of mitosis in cultures of normal human leukocytes. Cancer Res., 20 (1960) 462–468.26K. O'Flynn, A.M. Krensky, P.C. Beverley, S.J. Burakoff and D.C. Linch, Phytohaemoagglutinin activation of T cells through the sheep red blood cell receptor. Nature, 313 (1985) 686–687.27M.A. Valentine, C.D. Tsoukas, G. Rhodes, J.H. Vaughan and D.A. Carson, Phytohemagglutinin binds to the 20-kDa molecule of the T3 complex. Eur. J. Immunol., 15 (1985) 851–854.28R.K. Chopra, N.J. Holbrook, D.C. Powers, M.T. McCoy, W.H. Adler and J.E. Nadler, Interleukin 2, interleukin 2 receptor and interferon gamma synthesis and mRNA expression in phorbol myristate acetate and calcium ionophore A23187-stimulated T cells from elderly humans. Clin. Immunol. Immunopathol., 53 (1989) 297–308.29I. Beckman, K. Dimopoulos, X.N. Xu, M. Ahern and J. Bradley, Age related changes in the activation requirements of human CD4+ T cell subsets. Cell. Immunol., 132 (1991) 17–25.30C.H. June, J.A. Ledbetter, M.M. Gillespie, T. Lindsten and C.B. Thompson, T-cell proliferation involving the CD28 pathway is associated with cyclosporine-resistant interleukin 2 gene expression. Mol. Cell. Biol., 7 (1987) 4472–4481.31M.K. Jenkins, P.S. Taylor, S.O. Norton and K.B. Urdahl, CD28 delivers a costimulatory signal involved in antigen-specific IL-2 production by human T cells. J. Immunol., 147 (1991) 2461–2466.32A. Cossarizza, D. Monti, F. Bersani et al., Extremely low frequency pulsed electromagnetic fields increase interleukin-2 (IL-2) utilization and IL-2 receptor expression in mitogen-stimulated human lymphocytes from old subjects. FEBS Lett., 248 (1989) 141–144.33P. Bretscher, The two–signal model of lymphocyte activation twenty-one years later. Immunol. Today, 13 (1992) 74–76.34C.H. June, J.A. Bluestone, L.M. Nadler and C.B. Thompson, The B7 and CD28 receptor families. Immunol. Today, 7 (1994) 321–331.35F. Licastro, P.L. Tabacchi, M. Chiricolo et al., Defective self–recognition in subjects of far advanced age. Gerontology, 29 (1983) 64–72.36C. Franceschi, D. Monti, D. Barbieri et al., Immunosenescence in humans: deterioration or remodelling? Int. Rev. Immunol., 12 (1995) 57–74.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B69B273F75DCBF65BAAB6506108A3F83D71D665.txt b/test/dataset/in/resources/corpus/Clean_0B69B273F75DCBF65BAAB6506108A3F83D71D665.txt
deleted file mode 100644
index 1304f7b..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B69B273F75DCBF65BAAB6506108A3F83D71D665.txt
+++ /dev/null
@@ -1 +0,0 @@
- geront Gerontologistgeront The Gerontologist The Gerontologist 0016-9013 1758-5341 Oxford University Press 358 10.1093/geront/48.3.358 PHYSICAL ACTIVITY Long-Term Effects of a Stage-Based Intervention for Changing Exercise Intentions and Behavior in Older Adults Greaney Mary L. PhD 1 Riebe Deborah PhD 2 Ewing Garber Carol PhD 3 Rossi Joseph S. PhD 8 4 Lees Faith D. MS 5 Burbank Patricia A. DNSc, RN 6 Nigg Claudio R. PhD 7 Ferrone Christine L. MS 8 Clark Phillip G. ScD 5 Address correspondence to Mary L. Greaney, PhD, Public Health Nutrition, Harvard School of Public Health, 677 Huntington Avenue, Boston, MA 02115. E-mail: mgreaney@hsph.harvard.edu 6 2008 48 3 358 367 12 9 2007 5 1 2007 Copyright 2008 by The Gerontological Society of America 2008 Purpose: We examined the efficacy of an intervention tailored to the individual's stage of change for exercise adoption on exercise stage of change, physical activity, and physical function in community-dwelling older adults. Design and Methods: We randomized participants to a print and telephone intervention or a contact comparison group. Through the use of longitudinal analyses we examined the intervention's effectiveness in promoting stage progression, altering decisional balance and the processes of change, increasing self-efficacy and physical activity, and improving physical function among older adults who completed the 24-month study (N = 966). We conducted similar analyses that excluded individuals who were in maintenance at baseline and 24 months. Results: At the end of the study, there were no differences in stage progression, self-efficacy, decisional balance, the processes of change, physical activity, or physical function by intervention assignment. When the analyses excluded those participants (n = 358) who were in the maintenance stage for exercise throughout the intervention, we found that, compared with the comparison group, a greater proportion of individuals who received the exercise intervention progressed in stage by 24 months. Conversely, more individuals in the comparison group remained stable or regressed in stage compared with the intervention group. Implications: Results indicate that a tailored intervention is effective in increasing motivational readiness for exercise in individuals who were in stages of change other than maintenance. Behavior change Exercise Health promotion Intervention Transtheoretical model hwp-legacy-fpage 358 hwp-legacy-dochead RESEARCH ARTICLE Regular physical activity is associated with a decreased risk of functional limitation among older adults, a delay or decrease in the incidence of chronic health problems, an improved quality of life, and a reduced risk of falling (Miller, Rejeski, Reboussin, Ten Have, & Ettinger, 2000). Despite the known physiological and psychological benefits of a physically active lifestyle, physical inactivity remains a major health problem among older adults in the United States, with a majority of older adults failing to reach the minimum recommended levels of physical activity (Kamimoto, Easton, Maurice, Husten, & Macera, 1999; Kruger, Carlson, & Buchner, 2007). Physical activity interventions that include older adults and use an assortment of behavioral and educational strategies have been somewhat effective in increasing the adoption of physical activity (Hillsdon, Foster, & Thorogood, 2005; Marcus et al., 2006). Many intervention studies with older adults have provided organized, center-based exercise classes for a small number of participants (e.g., DiBrezzo, Shadden, Raybon, & Powers, 2005), but several studies of middle-aged and older adults have shown that most of these individuals prefer to do their physical activity outside a formal setting, and home-based exercise has higher adherence rates (Ashworth, Chad, Harrison, Reeder, & Marshall, 2005; Juneau et al., 1987; King et al., 2000; King, Haskell, Taylor, Kraemer, & DeBusk, 1991). Several theories of behavior change have been used to develop interventions to promote physical activity, including the transtheoretical model of health behavior change (TTM). The TTM is an integrative theoretical model that uses individual decision-making processes to explain intentional health behavior change (Prochaska & Velicer, 1997). It has been used to tailor interventions to increase physical activity and exercise in diverse populations (Jones et al., 2001; Marcus et al., 1998), and it has been validated for use in community-dwelling older adults (Barké & Nicholas, 1990; Riebe et al., 2005). Although it has been demonstrated that interventions based on the TTM are effective in increasing exercise under some circumstances, it has not been applied to a large-scale community-based intervention in older adults. Our purpose in this study was to determine whether an exercise intervention tailored to an individual's stage of change for exercise would be effective in progressing stage of change, increasing utilization of TTM constructs (self-efficacy, decisional balance, and the processes of change), increasing physical activity behavior, and improving physical function in community-dwelling older adults. Methods Participants The study participants were volunteers for the Study of Exercise and Nutrition in Older Rhode Islanders (SENIOR) Project, a community-based health-promotion study designed to increase fruit and vegetable consumption and exercise among community-dwelling older adults (Clark et al., 2005). We recruited participants into the study over a period of 14 months by using a variety of methods, including direct mailings, advertisements, flyers, and presentations given at senior housing sites (Greaney, Lees, Nigg, Saunders, & Clark, in press). Study entry criteria included being 60 years of age or older and residing in the greater East Providence, Rhode Island area. We excluded those individuals who were not living independently in the community. Participants completed the Physical Activity Readiness Questionnaire as modified for older adults, which identifies participants who should seek out medical consultation before engaging in a new program of exercise (Thomas, Reading, & Shephard, 1992). Participants who answered “yes” to any of the questions on the Physical Activity Readiness Questionnaire were strongly urged to check with their health care provider prior to entering the study, but medical clearance was not required. The University of Rhode Island's Institutional Review Board approved the study, and all participants provided informed consent. Design The SENIOR Project consisted of a 12-month intervention followed by a 12-month nonintervention observational period (Clark et al., 2005). We randomly assigned participants to one of four interventions: Group 1 increased fruit and vegetable consumption; Group 2 increased exercise; Group 3 increased both exercise and fruit and vegetable consumption; and Group 4 was a contact-comparison group. In this article we examine the effects of the intervention designed to increase exercise, so we combined the four groups to form two: an intervention group consisting of all participants who received the intervention for increasing exercise (Groups 2 and 3) and a comparison group consisting of participants who did not receive the intervention for increasing exercise (Groups 1 and 4). Data reported here were collected at baseline, 12 months, and 24 months. Intervention The SENIOR Project's intervention components have been discussed in detail elsewhere (Clark et al., 2005). Briefly, we developed the interventions on the basis of the TTM, and they included a series of written materials and counselor calls designed to encourage movement through the stages of change to increase exercise, fruit and vegetable consumption, or both. The tailored materials included behavior-specific manuals, newsletters, expert systems reports, and coaching calls. We contacted participants monthly by means of print or telephone contact throughout the 12-month intervention period. For those participants who received the exercise intervention, we strongly encouraged them to adopt a habitual pattern of intentional physical activity (exercise) performed at home or in formal exercise settings consistent with the recommendations for apparently healthy adults by the American College of Sports Medicine (1998). This guideline recommends several sessions per week of aerobic (cardiorespiratory), flexibility, and muscle-strengthening exercises. We chose these criteria for physical activity because this level of activity has been demonstrated to improve physical fitness, physical function, and general health, which are particularly important in older adults (American College of Sports Medicine). In the intervention's print materials and during the coaching calls, we encouraged participants to engage in aerobic exercise of moderate to vigorous intensity, such as brisk walking, cycling, swimming, and exercising in aerobic exercise classes or on exercise machines, for 3 to 5 days per week for a minimum of 20 minutes. We recommended flexibility exercises, such as stretching, yoga, Pilates, and tai chi, for at least 2 days per week, and we encouraged muscle-strengthening exercises such as weight lifting, calisthenics, and resistance equipment for 2 to 3 days per week. We gave participants information on how to do these exercises at home and we encouraged them to use exercise facilities and to join formal exercise programs available within the community. Details of the intervention components follow. Manual At the start of the intervention, each participant received an intervention-specific manual that was organized by stages of change and that incorporated stage-specific behavioral strategies commonly employed to advance to the next stage. The manual also included information about the benefits of physical activity, what types of exercise are recommended, how to safely engage in exercise, and a listing of community resources for exercise, including formal exercise programs. Individuals not receiving the exercise intervention received either a manual about fruit and vegetable consumption or a fall-prevention manual; neither of these manuals included any information about physical activity or exercise. Newsletters We sent intervention-specific newsletters to participants on a monthly basis, except for Months 4, 8, and 12, when they received an expert system report (subsequently described). Newsletters were stage specific, and participants received the newsletter designed for their stage of change as determined during the evaluations conducted at baseline and at 4 and 8 months. The newsletters included stories about older adults, practical tips about exercise, suggested physical activities, interactive sections, and local resources for exercise. The newsletters addressed stage-appropriate processes of change, and they included material to effect positive changes in self-efficacy and decisional balance. Expert System Assessments and Reports During the 12-month intervention, we collected data on TTM constructs and other behavioral and health information at baseline and at 4 and 8 months. On the basis of the participants' responses, a computer-based expert system generated a 4- to 5-page report that was mailed directly to each participant's home. Reports provided normative (compared with peers in the same stage of change, or SOC) and ipsative (based on current and previous responses from the individual) feedback on the individual's physical activity behavior to encourage continued positive behavior change. Coaching Calls Participants received three 15-minute coaching calls from trained counselors during the 12-month intervention period, approximately 4 to 6 weeks after they received their expert system reports, which the counselors used to guide calls. Counselors used a standardized protocol incorporating motivational interviewing strategies in a stage-matched manner (Rollnick, Heather, & Bell, 1992). Demographics and Overall Health Participants provided demographic information, including gender, age, race or ethnicity, and educational level. Participants also completed the Medical Outcomes Study 36-item Short Form, and the answer to the first question on this form; we used the General Health subscale (general health perception) to assess perceived health status (McHorney, Ware, and Raczek, 1993). The first measure is categorical and we used it for descriptive purposes; we used the latter measure as a covariate in the statistical analyses. Outcome Measures Participants completed a baseline assessment in person at their home or in the project office. We had the assessment administered by trained project staff members who were of similar age and ethnicity as the participants and from the intervention community. The project staff members administered the assessment before randomization (baseline), at the end of the 12-month intervention period, and 12 months after the completion of the intervention (24 months), during which no intervention was delivered. The present study includes measures obtained at baseline and at 12 and 24 months; they are described in the paragraphs that follow. TTM constructs Measuring SOC for exercise is ideal for gaining a better understanding of physical activity and exercise behavior, because it focuses on intention and perception of behavior rather than actual behavior (Nigg & Riebe, 2002). We measured all TTM constructs (SOC, self-efficacy, decisional balance, and processes of change) by using instruments validated in community-dwelling older adults, and details of these instruments follow here. SOC for Exercise We determined SOC for regular aerobic exercise by using a validated questionnaire and scoring algorithm (Nigg & Riebe, 2002). The criterion used for SOC for exercise was the volume of exercise recommended by the American College of Sports Medicine (1998), which recommends aerobic exercise of moderate to vigorous intensity, lasting at least 10 minutes, for a total of 20 to 60 minutes per day, 3 to 5 days per week. This criterion is consistent with the exercise message used in the intervention and with the U.S. Surgeon General's recommendations for physical activity (U.S. Department of Health and Human Services, 1996), which emphasizes moderate to vigorous physical activity on most days of the week. On the basis of their responses to the questionnaire, we classified participants into one of five stages. Participants classified in the precontemplation stage had no intention to begin exercising in the next 6 months, whereas those in the contemplation stage intended to begin exercising within the next 6 months. Individuals in the preparation stage intended to change within the next month. Those in action had been exercising regularly over the past 6 months, and participants in the maintenance stage had been exercising regularly for 6 months or more (Nigg & Riebe, 2002). Self-Efficacy The six-item exercise self-efficacy scale measured a participant's confidence in his or her ability to exercise despite adverse or challenging situations (Benisovich, Rossi, Norman, & Nigg, 1998). Self-efficacy scores range from 1 to 5, with higher scores indicating greater self-efficacy. Decisional Balance (Pros and Cons) This two factor, ten-item Likert-format scale measures the pros (advantages) and cons (disadvantages) of adopting or maintaining exercise (Nigg, Rossi, Norman, & Benisovich, 1998). Scores range from 1 to 5, with higher scores demonstrating more exercise pros or exercise cons. Processes of Change This questionnaire includes 30 statements that participants are asked to rate in terms of frequency of occurrence over the past month on a 5-point Likert scale ranging from 1 = never to 5 = repeatedly (Nigg & Riebe, 2002). The questionnaire contains three items for each of the 10 processes of change and provides individual scores (ranging from 1 to 5) for each process. For analyses, we divided the processes into two groupings—experiential and behavioral processes. Scores range from 15 to 75, with higher scores indicating greater use. Physical Activity Behavior We measured physical activity by using the Yale Physical Activity Survey (YPAS). The YPAS is a validated survey for older adults that is designed to assess exercise, household, and recreational physical activity during a typical week in the previous month (DiPietro, 2001). We calculated five activity dimensions: vigorous activity, leisurely walking, moving on feet, standing, and sitting. We then summed the results to calculate the YPAS summary score. A higher score indicates higher levels of physical activity. Physical Function We measured physical function by using the Timed Up-and-Go (TUG) test. The TUG test is a practical, reliable measure of functional mobility in older adults (Podsiadlo & Richardson, 1991). The TUG test measures, in seconds, the time taken by an individual to stand up from a standard armchair, walk a distance of 3 m, turn, walk back to the chair, and sit down again. TUG scores have been able to distinguish between older adults who need assistance in activities of daily living, those who are independent, and those with somewhat impaired mobility (Van der Bij, Laurant, & Wensing, 2002). TUG scores of less than 12 seconds reflect normal physical function in community-dwelling adults (Bischoff et al., 2003). Analysis We conducted all analyses by using SPSS version 15.0 (SPSS, Chicago, IL). We calculated means, standard deviations, and frequencies for demographic characteristics, health status, SOC, physical activity, and physical function variables at baseline by using SPSS descriptives. We set significance a priori at p <.05. The purpose of the analyses was to evaluate the intervention's effect on progressing SOC, promoting use of TTM constructs, increasing physical activity, and improving physical function. To examine SOC progression, we conducted chi-square analyses using three stage movement groups (progress, regress, stable) to test the hypothesis that stage progression differed by exercise intervention assignment at 12 and 24 months. We defined stage progression as an increase of one or more stages from baseline; we defined stage regression as a decrease of one or more stages from baseline. We categorized participants who remained in the same stage as being in the stable stage. Because there was a large proportion of participants who remained in a maintenance mode for exercise throughout the intervention who would not be expected to progress in stage, we also evaluated SOC progression while excluding participants who remained in maintenance at baseline and 24 months. We used a series of repeated measures analysis of covariance to answer these research questions: First, do self-efficacy, decisional balance, and processes of change scores differ in participants who received the intervention? Second, did the intervention have an effect on physical activity and physical functioning? We included the following covariates in all models: gender, age at baseline, and perceived health status at baseline. Time was the within-subject factor and intervention assignment was the between-subject factor. We also evaluated these questions while excluding participants who remained in maintenance at both time points (baseline and 24 months). Results Participants The results presented in this article include study participants who completed the 24-month evaluation (N = 966). This includes 21 people who completed the 24-month assessment but did not complete the 12-month assessment. Table 1 shows the characteristics of these people at baseline. Baseline Equivalence of Intervention Groups As shown in Table 1, there were no significant differences between individuals in the intervention and comparison groups based on age, education, gender, race, perceived health status, SOC, YPAS summary score, or TUG score. Overall, 58.2% (n = 562) of the total sample was in action or maintenance stages for exercise at baseline (56.4%, n = 265, of the intervention group and 59.9%, n = 297, of the comparison group). Adherence Of the 1,274 people enrolled in the study at baseline, 80.5% (n = 1,026) completed the 12-month follow-up, and 75.8% (n = 966) completed the 24-month evaluation. Analyses published elsewhere have examined differences between participants who withdrew and those who completed the study, and there were no differences in attrition by race–ethnicity or intervention assignment (Greaney et al., in press). Compared with participants who withdrew, participants completing the study were more likely to be female (77.8% vs 70.8%, p =.01), in the action or maintenance stages for exercise at baseline (80.1% vs 70.3%, p <.01), younger (76.2, SD = 6.8, vs 78.5, SD = 6.8; p <.001), and have more years of education (12.87 years, SD = 2.2, vs 12.11 years, SD = 3.4; p <.01; see Greaney et al.). Participants who dropped out of the study also had lower self-efficacy than did participants who completed the 24-month study (3.22, SD =.95, vs 3.37, SD =.90; p =.02). Program Outcomes With the exception of participants in the precontemplation stage who progressed in stage at 24 months, participants who received the intervention were no more likely to progress in stage, regress in stage, or remain stable in stage than were participants who did not receive the intervention. In addition, there were no differences in self-efficacy, decisional balance, and the processes of change by intervention assignment. Furthermore, there were no differences in physical activity behavior and physical function, as measured by the YPAS and TUG, respectively, by intervention assignment over the course of the study. Additional details about the program's outcomes are provided in the following paragraphs. Progress in Stage of Change Table 2 shows the proportion of participants in the intervention and comparison groups who progressed, regressed, or maintained SOC from baseline to 12 months and from baseline to 24 months. At 12 months, χ2(2, N = 941) = 2.02, p =.36, and 24 months, χ2(2, N = 966) = 5.61, p =.06, there were no differences between the intervention group and comparison group with respect to the proportion of individuals who progressed, regressed, or maintained stage. When we excluded participants who were in the maintenance stage at baseline and 12 months from the analysis (n = 378), the results remained unchanged, χ2(2, n = 563) = 2.01, p =.37. However, when we excluded participants (n = 358) who were in the maintenance stage throughout the 24-month period from this analysis, we found that a greater proportion of individuals who received the exercise intervention progressed in stage by 24 months compared with individuals in the comparison group. Consistent with this finding, more people in the comparison group than people in the intervention group remained stable or regressed in stage, χ2(2, N = 608) = 6.30, p =.04. A description of stage progress according to the baseline SOC by intervention assignment is shown in Table 3. Although not significantly different, a greater percentage of participants in the intervention group who were in contemplation, preparation, and action stages at baseline had progressed in stage at 12 months, compared with the comparison group. Overall, 22.3% of the intervention group progressed in stage by 12 months compared with 19.4% of the comparison group (p =.36). We found similar results at 24 months (27.2% vs 20.7%, p =.06), and this time individuals receiving the intervention who were in the precontemplation stage at baseline were more likely to progress in stage than individuals in the comparison group who were in precontemplation stage at baseline, χ2(2, n = 177) = 4.11, p = 04. TTM Constructs The means of the TTM constructs are shown in Table 4. With the exception of exercise pros at 24 months, there was no difference in the use of these constructs by intervention assignment over the course of the study. When we excluded individuals who were in the maintenance stage at baseline and 24 months from the analysis (n = 358), the results remained unchanged (data not shown). Behavioral Outcomes The adjusted means of the YPAS summary score and the TUG score at 12 and 24 months by intervention assignment are shown in Table 5. There were no significant differences between the intervention and comparison groups with respect to physical activity (YPAS summary score) or physical function (TUG score) at 12 or 24 months. When we excluded the individuals who remained in the maintenance stage at baseline and 24 months from the analysis (n = 358), the results remained unchanged (data not shown). Discussion There were no significant differences between the intervention group and the comparison group in the percentage of individuals in each of the SOCs for exercise at 12 months, at the end of the intervention period, or a long-term effect at 24 months. There were limited differences in stage progression between the intervention and comparison groups (individuals in the precontemplation stage who received the intervention were more likely to progress in stage than were individuals in the comparison group). When we excluded the large number of participants who remained in the maintenance stage for the entire study period from the analysis, there was a long-term effect at 24 months, so that individuals who received the exercise intervention were more likely to progress in stage and individuals in the comparison group were more likely to regress in stage or remain stable in stage. Concurrently, with the exception of exercise pros at 24 months, there were no differences in use of the TTM constructs, or any changes in physical activity or physical function throughout the intervention. It is intriguing that individuals in the precontemplation stage at baseline who received the intervention had significantly more stage progression than did similar individuals in the comparison group. This suggests that the intervention was more effective in individuals who were not considering becoming physically active at the start of the study. This could have important public health implications if confirmed in subsequent studies, as it suggests that the most sedentary older adults could benefit from a low cost, primarily print-based intervention. Sedentary individuals (in the precontemplation, contemplation, or preparation stages) were more likely to drop out of the study, although the drop-out rate among sedentary individuals was similar between the intervention group and the comparison group. Participants who dropped out of the study had lower exercise self-efficacy than did those who completed the study. Some studies of exercise adherence have reported that low levels of self-efficacy and readiness to change are associated with drop out from the exercise program (Young, King, Sheehan, & Stefanick, 2002), whereas higher levels of self-efficacy predict adoption and maintenance of exercise (Oman & King, 1998). Overall, individuals completing the study were younger and had more education than those who dropped out of the study; each of these variables is associated with higher levels of physical activity and better health (Bish, et al., 2005; U.S. Department of Health and Human Services, 1996). There were no differences between the intervention and comparison groups in YPAS summary score and TUG scores at 12 or 24 months, suggesting that the intervention was not effective in improving self-reported physical activity or physical function. The lack of effect of the intervention on these variables remained when individuals in maintenance at each time point were excluded from the analyses. The lack of effect on physical activity behavior and function is not surprising, because the primary outcome of this study was SOC, which reflects intentions for and perception of exercise behavior and does not necessarily reflect actual exercise behavior. For example, an individual who moved from precontemplation to contemplation or who remained in the maintenance stage, as a large portion of our participants did, would not be expected to change physical activity behavior and consequently physical function. Advancing age is associated with declines in physical activity, physical fitness, and physical function (American College of Sports Medicine, 2004). Interestingly, neither physical activity nor physical function declined significantly in our sample, as evidenced by the stable YPAS summary scores and TUG scores over 24 months. The mean scores on the TUG test are well below the 12-second cutpoint for normal physical function in community-dwelling older adults, demonstrating that our participants were particularly vigorous. We do not know if those who dropped out of the study were more likely to experience declines in physical functions and physical activity or be institutionalized, but this is certainly a possibility. The modest results found in our study of younger and older adults are consistent with the findings of previous studies in younger and older adults that used similar methodologies (Banks-Wallace & Conn, 2002; Conn, Valentine, & Cooper, 2002; Eakin, Glasgow, & Riley, 2000; Hillsdon et al., 2005; Kahn et al., 2002; Van der Bij, et al., 2002). A recent Cochrane review concluded that physical activity interventions have moderate to little effect (Hillsdon et al.). It is believed that physical activity interventions have had equivocal results in part as a result of methodological problems, including incomplete or inconsistent application of behavioral theory (Baranowski, Anderson, & Carmack, 1998; Epstein, 1998; Prochaska & Velicer, 1997). The SENIOR Project's intervention was designed with one theoretical model, the TTM, and it maintained theoretical integrity in both the development of the intervention and the outcome measurements. Nevertheless, the results were still modest, indicating the difficulty in understanding and altering this complex behavior. The main focus of our intervention was at the intrapersonal (individual) level; the intervention did not incorporate interpersonal and environmental factors that are thought to mediate physical activity and exercise behavior (King, 2001). In a very small subsample of our participants, we have shown that a number of personal and physical environmental factors affected physical activity behavior (King et al., 2006), but we are unable to evaluate this possibility in the context of the intervention. Print interventions have generally had a small effect on physical activity behavior (Marshall, Owen, & Bauman, 2004). The use of tailoring and adding other methods such as counseling in combination with print materials, similar to the methods used in the SENIOR Project, have improved the impact of print and other types of physical activity interventions (Kahn et al., 2002; Van der Bij et al., 2002). One difficulty with print interventions, such as the SENIOR project, is that it is difficult to assess the dose of the intervention that the particpants received; we were unable to assess the amount of time spent using the materials or ensure that people actually read and understood the materials. We did not screen for literacy or cognitive function, although it is known that cognitive decline is common in aging adults (Craik & Bialystok, 2006; Raz, Rodrigue, Kennedy, & Acker, 2007). A high proportion of our sample was in action or maintenance stages at the start of the study, which may have affected our ability to assess the impact of our intervention, because the participants perceived that they were already engaging in adequate amounts of exercise. Over 70% of the individuals in maintenance for exercise at baseline remained in maintenance at 24 months. These results are not unexpected, as a recent study of long-term maintenance of regular exercise (McAuley et al., 2007) reported that regular exercise behavior was a strong predictor of maintaining exercise over the long term. The high proportion of people in the maintenance stage is consistent with other studies of adults of all ages obtained in statewide samples in Rhode Island (Garber, Allsworth, Hesser, & Marcus, 2006; Laforge et al., 1999). When we excluded participants who were in the maintenance stage at both time points (baseline and 12 months, baseline and 24 months), we found that individuals in all SOCs in the exercise intervention group were more likely to advance in stage by 24 months. This was not an immediate effect of the intervention: At 12 months there were no differences between the intervention and comparison groups with respect to the proportion who progressed, regressed, or remained stable in stage. These results are similar to results found in studies of physical activity (McAuley et al., 1999) and smoking cessation (Prochaska et al., 2001) that have found that individuals who received behaviorally based interventions continued to improve their behavior after completion of the intervention, indicating that the effects of the treatment continued long after the end of the intervention. The consistency between the SOC and self-reported physical activity and independently assessed physical function provides further validation for the SOC for exercise, extending the results of previous studies (Schumann, Estabrooks, Nigg, & Hill, 2003; Schumann et al., 2002; Spencer, Adams, Malone, Roy, & Yost, 2006). Use of TTM Constructs There were no differences in the scores of the TTM constructs by intervention assignment. We were unable to detect differences between the intervention and comparison groups with respect to exercise self-efficacy, decisional balance, experiential processes, and behavioral processes at 12 and 24 months, even when we excluded the large number of participants who remained in maintenance throughout the intervention. These results are in contrast to other studies of exercise adoption and maintenance in older and younger adults, in which self-efficacy and the use of cognitive and behavioral strategies predict exercise behavior (Bock et al., 1997; Brassington, Atienza, Perczek, DiLorenzo, & King, 2002; Cheung et al., 2007; Kosma, Cardinal, & McCubbin, 2004; McAuley et al., 2007; Oman & King, 1998; Velicer, Norman, Fava, & Prochaska, 1999). However, it is possible that the number of participants who progressed or regressed in stage was inadequate for us to detect differences in the use of these behavioral processes. Study Limitations This study had several limitations, the greatest being the potential for sampling (volunteer) bias, whereby persons in the preaction stages may have been less likely to volunteer as participants in this study, whereas persons in maintenance were more likely to volunteer. There is also the possibility of drop-out bias because individuals who were sedentary and with lower levels of self-efficacy at baseline were more likely to withdraw than participants who were regular exercisers. The drop-out pattern was similar between the exercise intervention and the comparison groups, so there is no differential group effect of the dropouts. Lastly, all of the measures, except physical function, were measured by self-report, which can reduce the sensitivity and specificity of the instruments as a result of report and response biases. However, TUG scores are an objective measure of physical function, and the TUG scores and self-report of physical activity mirrored each other, suggesting that the self-reported physical activity data were valid. Conclusions The SENIOR exercise intervention demonstrated that a community-based intervention to promote exercise that is grounded by behavioral theory can be modestly effective in promoting stage progression over the long term. Nevertheless, the intervention did not increase physical activity behavior or improve physical function. Future research evaluating the intrapersonal, interpersonal, and environmental mediator and moderators variables affecting the adoption and maintenance of exercise behavior is needed to provide important insights into factors that can be addressed in future behavior-based interventions to improve outcomes. This research was supported by Grant 1R01AG16588 from the National Institute on Aging, National Institutes of Health. Preliminary results were presented in a symposium at the 57th Annual Scientific Meeting of the Gerontological Society of America, Washington, DC, November 19–23, 2004, and at the 26th Annual Meeting of the Society of Behavioral Medicine, Boston, MA, April 13–16, 2005. We thank all participants of the Study of Exercise and Nutrition in Older Rhode Islanders (SENIOR) Project and acknowledge the contributions of all the members of the SENIOR Project Research Team, including the following individuals: Bryan Blissmer, PhD; Robert Dufresne, PhD; Catherine English, PhD, RD; Geoffrey W. Greene, PhD, RD, RN; Andrea Luisi, PharmD; Norma Owens, PharmD; Cynthia Padula, PhD, RN; James Prochaska, PhD; Susan Rossi, PhD, RN; Sandra Saunders, MS, MPH; Laurie Ruggiero, PhD; Kira Stillwell, MS; and Nancy Fey-Yensan, PhD, RD. 1 Public Health Nutrition, Harvard School of Public Health, Boston, MA. 2 Department of Kinesiology, University of Rhode Island, Kingston. 3 Department of Biobehavioral Sciences, Columbia University, New York, NY. 4 Department of Psychology, University of Rhode Island, Kingston. 5 Program in Gerontology and Rhode Island Geriatric Education Center, University of Rhode Island, Kingston. 6 School of Nursing, University of Rhode Island, Kingston. 7 Department of Public Health Sciences, University of Hawaii at Manoa, Honolulu. 8 Cancer Prevention Research Center, University of Rhode Island, Kingston. Decision Editor: William J. McAuley, PhD Table 1. Characteristics of 966 Participants Completing the SENIOR Project by Intervention Assignment. Intervention Assignment<--CO?1--> Intervention Group Comparison Group Characteristic n (M) % (SD) n (M) % (SD) p Baseline age: years (75.2) (6.7) (74.7) (6.6) .28 Education: years (12.9) (2.7) (12.9) (3.1) .59 YPAS summary score (44.9) (26.4) (45.7) (31.4) .66 TUG: seconds (8.7) (4.1) (8.8) (5.7) .74 Gender Male 128 27.2 146 29.6 Female 342 72.8 348 70.4 .43 Race or ethnicity White 362 78.4 379 77.0 Black 10 2.2 12 2.4 Portuguese–Cape Verdean 60 13.0 71 14.4 Other 30 6.5 30 6.1 .91 Perceived health status Excellent–very good 221 47.4 244 49.3 Good 188 40.3 192 38.8 Fair–poor 57 12.2 59 11.9 .65 SOC for exercise at baseline Precontemplation 89 18.9 88 17.7 Contemplation 32 6.8 29 5.8 Preparation 84 17.9 82 16.5 Action 22 4.7 30 6.0 Maintenance 243 51.7 267 53.8 .77 Note: The intervention and comparison groups are n = 470 and n = 496, respectively; some data are missing, so the sample size for all variables may not equal 966. Values given in parentheses refer to means or standard deviations, as shown. SENIOR = Study of Exercise and Nutrition in Older Rhode Islanders; YPAS = Yale Physical Activity Survey; TUG = Timed Up-and-Go test; SOC = state of change; SD = standard deviation. Table 2. SOC Movement from Baseline by Intervention Assignment. Intervention Assignment<--CO?2--> Intervention Group Comparison Group Stage Movement n % n % 12 months Progress 102 22.3 94 19.4 Regress 102 22.3 124 25.6 Stable 253 55.4 266 55.0 24 months Progress 128 27.2 103 20.8 Regress 117 24.9 138 27.8 Stable 225 47.9 255 51.4 Note: Intervention and comparison groups are n = 470 and n = 496, respectively; some data are missing, so the sample size for all variables may not equal 966. SOC = stage of change. Table 3. SOC Movement from Baseline by Stage and Intervention Assignment. Progress<--CO?3-->: %<--CO?4--> (n) Regress: % (n) Stable: % (n) Baseline Stage Intervention Comparison Intervention Comparison Intervention Comparison 12 Months Precontemplation 37.1 (33) 37.6 (32) 62.9 (56) 62.4 (53) Contemplation 59.4 (19) 48.3 (14) 37.5 (12) 44.8 (13) 3.1 (1) 6.9 (2) Preparation 45.6 (36) 38.3 (31) 41.8 (33) 40.7 (33) 12.7 (10) 21.0 (17) Action 77.8 (14) 58.6 (17) 16.7 (3) 37.9 (11) 5.6 (1) 3.4 (1) Maintenance 22.6 (54) 25.8 (67) 77.4 (185) 74.2 (193) 24 Months Precontemplation* 58.4 (52) 43.2 (38) 41.6 (37) 6.8 (50) Contemplation 62.5 (20) 51.7 (15) 31.3 (10) 31.0 (9) 6.3 (2) 17.2 (5) Preparation 51.2 (43) 46.3 (38) 33.3 (28) 40.2 (33) 15.5 (13) 13.4 (11) Action 59.1 (13) 40.0 (12) 36.4 (8) 50.0 (15) 4.5 (1) 10.0 (3) Maintenance 29.2 (71) 30.3 (81) 70.8 (172) 69.7 (186) Notes: Intervention and comparison groups are n = 470 and n = 496, respectively; some data are missing, so the sample size for all variables may not equal 966. SOC = stage of change. *Significant difference (p <.05) in stage progression by intervention assignment. Table 4. Adjusted Means of the TTM Constructs by Intervention Assignment. Intervention <--CO?5-->Group Comparison Group Construct M SE M SE p Self-efficacy Baseline 3.41 .04 3.37 .04 .57 12 months 3.50 .05 3.41 .04 .10 24 months 3.52 .05 3.41 .05 .11 Pros Baseline 3.51 .05 3.47 .05 .52 12 months 3.64 .05 3.65 .04 .98 24 months 3.81 .05 3.67 .05 .03 Cons Baseline 1.22 .02 1.24 .02 .31 12 months 1.16 .02 1.49 .02 .20 24 months 1.15 .02 1.16 .02 .80 Experiential processes Baseline 50.87 .57 50.52 .55 .63 12 months 53.32 .52 52.46 .51 .21 24 months 54.83 .58 52.83 .56 .10 Behavioral processes Baseline 45.04 .55 43.61 .53 .05 12 months 46.40 .53 45.51 .51 .20 24 months 46.13 .56 45.13 .54 .17 Note: Intervention and comparison groups are n = 470 and n = 496, respectively; some data are missing, so the sample size for all variables may not equal 966. Adjusted means indicates that the analyses controlled for gender, age, and perceived health status at baseline. TTM = transtheoretical model of behavior change; SE = standard error. Table 5. Adjusted Means of the YPAS Summary Score and TUG Scores by Intervention Assignment. Intervention Group Comparison Group Scores M SE M SE p YPAS summary score Baseline 46 1.4 46 1.3 .96 12 months 46 1.2 47 1.1 .81 24 months 47 1.3 47 1.2 .92 TUG score (seconds) Baseline 8.4 0.2 8.6 .2 .51 12 months 8.6 0.2 8.7 .2 .50 24 months 8.9 0.3 8.9 .3 .93 Note: Intervention and comparison groups are n = 470 and n = 496, respectively; some data are missing, so the sample size for all variables may not equal 966. Adjusted means indicates that the analyses controlled for gender, age at baseline, and perceived health status at baseline. YPAS = Yale Physical Activity Survey; TUG = Timed Up-and-Go test; SE = standard error. References American College of Sports Medicine. (1998). American College of Sports Medicine Position Stand. The recommended quantity and quality of exercise for developing and maintaining cardiorespiratory and muscular fitness, and flexibility in healthy adults. Medicine & Science in Sports & Exercise, 30, 975-991. American College of Sports Medicine. (2004). Physical activity programs and behavior counseling in older adult populations. Medicine & Science in Sports & Exercise, 36, 1997-2003. Ashworth, N. L., Chad, K. E., Harrison, E. L., Reeder, B. A., & Marshall, S. C., (2005). Home versus center based physical activity programs in older adults (review). Cochrane Database of Systematic Reviews, 25, (1), CD004017. Banks-Wallace, J., & Conn, V., (2002). Interventions to promote physical activity among African American women. Public Health Nursing, 19, 321-335. Baranowski, T., Anderson, C., & Carmack, C., (1998). Mediating variable framework in physical activity interventions. How are we doing? How might we do better? American Journal of Preventive Medicine, 15, 266-297. Barké, C. R., & Nicholas, P. R., (1990). Physical activity in older adults: The stages of change. Journal of Applied Gerontology, 9, 216-223. Benisovich, S. V., Rossi, J. S., Norman, G. J., & Nigg, C. R., (1998). Development of a multidimensional measure of exercise self-efficacy. Annals of Behavioral Medicine, 20, 190. Bischoff, H. A., Stähelin, H. B., Monsch, A. U., Iversen, M. D., Weyh, A., & von Dechend, M., et al (2003). Identifying a cut-off point for normal mobility: A comparison of the timed “up and go” test in community-dwelling and institutionalized elderly women. Age and Ageing, 32, 315-320. Bish, C. L., Blanck, H. M., Serdula, M. K., Marcus, M., Kohl, H. W., 3rd, & Khan, L. K., (2005). Diet and physical activity behaviors among Americans trying to lose weight: 2000 Behavioral Risk Factor Surveillance System. Obesity Research, 13, 596-607. Bock, B. C., Albrecht, A. E., Traficante, R. M., Clark, M. M., Pinto, B. M., & Tilkemeier, P., et al (1997). Predictors of exercise adherence following participation in a cardiac rehabilitation program. International Journal of Behavioral Medicine, 4, 60-75. Brassington, G. S., Atienza, A. A., Perczek, R. E., DiLorenzo, T. M., & King, A. C., (2002). Intervention-related cognitive versus social mediators of exercise adherence in the elderly. American Journal of Preventive Medicine, 23, (Suppl. 2), 80-86. Cheung, C., Wyman, J., Gross, C., Peters, J., Findorff, M., & Stock, H., (2007). Exercise behavior in older adults: A test of the transtheoretical model. Journal of Aging and Physical Activity, 15, 103-118. Clark, P. G., Rossi, J. S., Greaney, M. L., Riebe, D. A., Greene, G. W., & Saunders, S. D., et al (2005). Intervening on exercise and nutrition in older adults: The Rhode Island SENIOR Project. Journal of Aging & Health, 17, 753-778. Conn, V. S., Valentine, J. C., & Cooper, H. M., (2002). Interventions to increase physical activity among aging adults: A meta-analysis. Annals of Behavioral Medicine, 24, 190-200. Craik, F. I., & Bialystok, E., (2006). Cognition through the lifespan: Mechanisms of change. Trends in Cognitive Sciences, 10, 131-138. DiBrezzo, R., Shadden, B. B., Raybon, B. H., & Powers, M., (2005). Exercise intervention designed to improve strength and dynamic balance among community-dwelling older adults. Journal of Aging and Physical Activity, 13, 198-209. DiPietro, L., (2001). Physical activity in aging: Changes in patterns and their relationship to health and function. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences 56:13-22 Journal of Gerontology: Medical Sciences, 56A(Suppl. 2), M13–M22. Eakin, E. G., Glasgow, R. E., & Riley, K. M., (2000). Review of primary care-based physical activity intervention studies: Effectiveness and implications for practice and future research. The Journal of Family Practice, 49, 158-168. Epstein, L. H., (1998). Integrating theoretical approaches to promote physical activity. American Journal of Preventive Medicine, 15, 257-265. Garber, C. E., Allsworth, J. E., Hesser, J., & Marcus, B. H., (2006). Trends in stages of change for physical activity in a statewide surveillance sample. Medicine and Science in Sports and Exercise, 38, S378. Greaney, M. L., Lees, F. D., Nigg, C. R., Saunders, S. D., & Clark, P. G., in press. Recruiting and retaining older adults for health promotion research: The experience of the SENIOR Project. Journal of Nutrition for the Elderly.. Hillsdon, M., Foster, C., & Thorogood, M., (2005). Interventions for promoting physical activity [review]. Cochrane Database of Systematic Reviews, CD003180. Jones, N. D., DellaCorte, M. R., Nigg, C. R., Clark, P. G., Burbank, P. M., & Garber, C. E., et al (2001). Seniorcise: A print exercise intervention in older adults. Educational Gerontology, 27, 717-728. Juneau, M., Rogers, F., De Santos, V., Yee, M., Evans, A., Bohn, A., Haskell, W. L., Taylor, C. B., & DeBusk, R. F., et al (1987). Effectiveness of self-monitored, home-based, moderate-intensity exercise training in middle-aged men and women. American Journal of Cardiology, 60, 66-70. Kahn, E. B., Ramsey, L. T., Brownson, R. C., Heath, G. W., Howze, E. H., & Powell, K. E., et al (2002). The effectiveness of interventions to increase physical activity. A systematic review. American Journal of Preventive Medicine, 22, (Suppl. 4), 73-107. Kamimoto, L. A., Easton, A. N., Maurice, E., Husten, C. G., & Macera, C.A., (1999). Surveillance for five health risks among older adults—United States, 1993–1997. Morbidity & Mortality Weekly Report, 48, (SS08), 89-130. King, A. C., (2001). Interventions to promote physical activity by older adults. Journal of Gerontology: Medical Sciences, 56A, (Spec. Issue 2), M36-M46. King, A. C., Castro, C., Wilcox, S., Eyler, A. A., Sallis, J. F., & Brownson, R. C., (2000). Personal and environmental factors associated with physical inactivity among different racial-ethnic groups of U.S. middle-aged and older-aged women. Health Psychology, 19, 354-364. King, A. C., Haskell, W. L., Taylor, C. B., Kraemer, H. C., & DeBusk, R. F., (1991). Group- vs home-based exercise training in healthy older men and women. A community-based clinical trial. Journal of the American Medical Association, 266, 1535-1542. King, A. C., Toobert, D., Ahn, D., Resnicow, K., Coday, M., & Riebe, D., et al (2006). Perceived environments as physical activity correlates and moderators of intervention in five studies. American Journal of Health Promotion, 21, 24-35. Kosma, M., Cardinal, B. J., & McCubbin, J. A., (2004). Predictors of physical activity stage of change among adults with physical disabilities. American Journal of Health Promotion, 19, 114-117 Retrieved July 10, 2007, from www.cdc.gov/mmwr/preview/mmwrhtml/ss4808a5.htm. Kruger, J., Carlson, S. A., & Buchner, D., (2007). How active are older Americans? Preventing Chronic Disease, 4, Retrieved July 7, 2007, from http://www.cdc.gov/pcd/issues/2007/jul/06_0094.htm. Laforge, R. G., Rossi, J. S., Prochaska, J. O., Velicer, W. F., Levesque, D. A., & Mahoney, C. A., (1999). Stage of regular exercise and health-related quality of life. Preventive Medicine, 28, 349-360. Marcus, B. H., Emmons, K. M., Simkin-Silverman, L. R., Linnan, L. A., Taylor, E. R., & Bock, B. C., et al (1998). Evaluation of motivationally tailored vs. standard self-help physical activity interventions at the workplace. American Journal of Health Promotion, 12, 246-253. Marcus, B. H., Williams, D. M., Dubbert, P. M., Sallis, J. F., King, A. C., & Yancey, A. K., et al (2006). Physical activity intervention studies: What we know and what we need to know: A scientific statement from the American Heart Association Council on Nutrition, Physical Activity, and Metabolism (Subcommittee on Physical Activity); Council on Cardiovascular Disease in the Young; and the Interdisciplinary Working Group on Quality of Care and Outcomes Research. Circulation, 114, 2739-2752. Marshall, A. L., Owen, N., & Bauman, A. E., (2004). Mediated approaches for influencing physical activity: Update of the evidence on mass media, print, telephone and website delivery of interventions. Journal of Science and Medicine in Sport, 7, (Suppl. 1), 74-80. McAuley, E., Katula, J., Mihalko, S. L., Blissmer, B., Duncan, T. E., & Pena, M., et al (1999). Mode of physical activity and self-efficacy in older adults: A latent growth curve analysis. Journal of Gerontology: Psychological Sciences, 54B, P283-P292. McAuley, E., Morris, K. S., Motl, R. W., Hu, L., Konopack, J. F., & Elavsky, S., (2007). Long-term follow-up of physical activity behavior in older adults. Health Psychology, 26, 375-380. McHorney, C. A., Ware, J. E., & Raczek, A. E., (1993). The MOS 36-Item Short Form Health Survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Medical Care, 31, 247-263. Miller, M. E., Rejeski, W. J., Reboussin, B. A., Ten Have, T. R., & Ettinger, W. H., (2000). Physical activity, functional limitations, and disability in older adults. Journal of the American Geriatrics Society, 48, 1264-1272. Nigg, C. R., Rossi, J. S., Norman, G. J., & Benisovich, S. V., (1998). Structure of decisional balance for exercise adoption. Annals of Behavioral Medicine, 20, S211. Nigg, C. R., & Riebe, D., (2002). The transtheoretical model: Research review of exercise behavior and older adults. In P. M. Burbank & D. Riebe (Eds.), Promoting exercise and behavior change in older adults (pp. 147–180). New York: Springer. Oman, R. F., & King, A. C., (1998). Predicting the adoption and maintenance of exercise participation using self-efficacy and previous exercise participation rates. American Journal of Health Promotion, 12, 154-161. Podsiadlo, D., & Richardson, S., (1991). The timed “Up & Go”: A test of basic functional mobility for frail elderly persons. Journal of the American Geriatrics Society, 39, 142-148. Prochaska, J. O., & Velicer, W. F., (1997). The transtheoretical model of health behavior change. American Journal of Health Promotion, 12, 38-48. Prochaska, J. O., Velicer, W. F., Fava, J. L., Riebe, D., Garber, C. E., & Rossi, J. S., et al (2001). Evaluating a population-based recruitment approach and a stage-based expert system intervention for smoking cessation. Addictive Behaviors, 56, 583-602. Raz, N., Rodrigue, K. M., Kennedy, K. M., & Acker, J. D., (2007). Vascular health and longitudinal changes in brain and cognition in middle-aged and older adults. Neuropsychology, 21, 149-157. Riebe, D., Blissmer, B., Greene, G., Caldwell, M., Ruggiero, L., & Stillwell, K. M., et al (2005). Long-term maintenance of exercise and healthy eating behaviors in overweight adults. Preventive Medicine, 40, 769-778. Rollnick, S., Heather, N., & Bell, A., (1992). Negotiating behavior change in medical settings: The development of brief motivational interviewing. Journal of Mental Health, 1, 25-37. Schumann, A., Estabrooks, P. A., Nigg, C. R., & Hill, J., (2003). Validation of the stages of change with mild, moderate, and strenuous physical activity behavior, intentions, and self-efficacy. International Journal of Sports Medicine, 24, 363-365. Schumann, A., Nigg, C. R., Rossi, J. S., Jordan, P. J., Norman, G. J., & Garber, C. E., et al (2002). Construct validity of the stages of change of exercise adoption for different intensities of physical activity in four samples of differing age groups. American Journal of Health Promotion, 16, 280-287. Spencer, L., Adams, T. B., Malone, S., Roy, L., & Yost, E., (2006). Applying the transtheoretical model to exercise: A systematic and comprehensive review of the literature. Health Promotion Practice, 7, 428-443. Thomas, S., Reading, J., & Shephard, R. J., (1992). Revision of the Physical Activity Readiness Questionnaire (Par-Q). Canadian Journal of Sport Sciences, 17, 4338-4345. U.S. Department of Health and Human Services. (1996). Physical activity and health: A report of the Surgeon General. Atlanta, GA: U. S. Department of Health and Human Services, Centers for Disease Control and Prevention. Retrieved from http://www.cdc.gov/nccdphp/sgr/htm. Van der Bij, A. K., Laurant, M. G., & Wensing, M., (2002). Effectiveness of physical activity interventions for older adults: A review. American Journal of Preventive Medicine, 22, 120-133. Velicer, W. F., Norman, G. J., Fava, J. L., & Prochaska, J. O., (1999). Testing 40 predictions from the transtheoretical model. Addictive Behavior, 24, 455-469. Young, D. R., King, A. C., Sheehan, M., & Stefanick, M. L., (2002). Stage of motivational readiness: Predictive ability for exercise behavior. American Journal of Health Behavior, 26, 331-341.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B6C0887F02B47C2DBB922E056BE4C67FB660CB8.txt b/test/dataset/in/resources/corpus/Clean_0B6C0887F02B47C2DBB922E056BE4C67FB660CB8.txt
deleted file mode 100644
index 0a054bd..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B6C0887F02B47C2DBB922E056BE4C67FB660CB8.txt
+++ /dev/null
@@ -1 +0,0 @@
-cercorcercorCerebral Cortex1460-21991047-3211Oxford University Press10.1093/cercor/bhl036ArticlesExtrahippocampal Contributions to Age Differences in Human Spatial NavigationMoffatScott D.12KennedyKristen M.12RodrigueKaren M.12RazNaftali121Institute of Gerontology2Department of Psychology, Wayne State University, Detroit, MI 48202, USAAddress correspondence to email: moffat@wayne.edu.62007207200617612741282© The Author 2006. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org2007The hippocampus (HC) and associated neural structures are hypothesized to contribute to individual differences in human spatial navigation. However, functional imaging studies and theoretical models underscore the importance of extrahippocampal regions as well. The purpose of the present study was to examine age differences in virtual environment navigation and to assess possible relationships between navigation and structural integrity of hippocampal and extrahippocampal brain regions. Healthy adult volunteers completed a virtual navigation task and underwent magnetic resonance imaging to assess volumes of the caudate nucleus (CN), cerebellum, HC, prefrontal, and primary visual cortices. Results demonstrated robust age-related differences in place learning. Moreover, individual differences in regional brain volumes as well as performance on the tests of memory and executive functions contributed to age differences in human place learning. High performance in a virtual navigation task was associated with larger volume of the CN and prefrontal gray and white matter. Larger hippocampal volume was associated with improved performance in the young but not old participants. We conclude that human navigation requires both hippocampal and extrahippocampal brain systems and draws on executive resources for successful performance.agingAlzheimer's diseaseexecutive functionhippocampusplace learningwanderingIntroductionAge-related deficits in spatial maze learning are evident in spatial navigation tasks across several mammalian species (Barnes 1979; Ingram 1988; McLay and others 1999). In particular, these deficits are observed in performance on the Morris water task (MWT) (Morris and others 1982) that has been used extensively in studying cognitive aging in animals. Briefly, in the MWT, an animal is required to search for a platform hidden beneath the surface of a circular pool. Because the platform cannot be seen directly, the animal must locate the platform position with reference to external cues throughout the environment and remember its location across a series of trials. Multiple studies of MWT performance have yielded a plethora of data, and age-related deficits on the MWT have been well characterized (Gallagher and Pelleymounter 1988; Lindner 1997; Lukoyanov and others 1999; Begega and others 2001). One of the reasons for popularity of MWT in studies of cognitive aging is its sensitivity to hippocampal lesions and its functional dependence on the hippocampal formation (Morris and others 1982).The investigation of age-related differences in human spatial navigation is a topic of considerable import. Nondemented elderly experience substantial deficits in spatial navigation (Kirasic 1991; Kirasic and others 1992; Burns 1999; Wilkniss and others 1997; Newman and Kaszniak 2000), and impairment in navigational skills is often apparent in the early stages of dementia (Passini and others 1995; Klein and others 1999). Recently, virtual environment (VE) spatial learning tasks have been successfully applied to the study of age differences in spatial navigation among healthy elderly (Moffat and others 2001, 2006; Moffat and Resnick 2002; Driscoll and others 2005). In a virtual MWT, elderly individuals traveled a longer linear distance in locating the hidden goal, spent less time searching in the vicinity of the goal on a retention trial, and were impaired in constructing a cognitive map of the environment in comparison to their younger counterparts (Moffat and Resnick 2002).Functional neuroimaging and lesion studies have identified a complex network of structures that are involved in spatial navigation. The proposed network includes the hippocampus (HC), parahippocampal gyrus, cerebellum (CB), parietal cortex, posterior cingulate gyrus, prefrontal cortex, retrosplenial cortex, and other cortical and subcortical regions (Aguirre and others 1996; Barrash 1998; Maguire and others 1998; Katayama and others 1999; Gron and others 2000). In a recent functional neuroimaging study, we found that in performing a virtual navigation task, elderly participants showed reduced activation in some of the same neural systems that young subjects use to navigate successfully. The cerebral regions identified in that study included the HC, parahippocampal gyrus, and retrosplenial cortex. Thus, reduced activation in the HC and extrahippocampal regions may underlie the observed age deficits in performance (Moffat and others 2006).Most attention in research on human navigation has focused on the role of the HC and associated structures. This focus may be justified in light of the prominent place of the HC in models of human spatial and episodic memory and in animal models of spatial navigation. However, a closer examination of the executive and strategic demands of human navigation and their neural underpinnings is overdue. Successful navigation requires the selection of an appropriate search strategy and also depends on appropriate behavioral monitoring and alterations of searching behavior if the selected strategy proves unsuccessful. Such reliance on executive skills during acquisition is common in many tasks, including those that in their developed states are highly proceduralized. On other procedural tasks, shrinkage of the prefrontal cortex, neostriatum, and CB have been shown to contribute to age-related differences in these skills (Raz and others 2000; Kennedy and Raz 2005). Thus, it is plausible that age-related deficits in acquiring and maintaining navigational skills may reflect, at least in part, differential shrinkage of multiple regions rather than strictly hippocampal atrophy.The age-related deficits observed in the virtual analog of MWT cannot be fully accounted for by models of performance restricted to spatial memory. In particular, we have observed that healthy elderly perform more poorly than their younger counterparts even on the first trial, which does not depend on memory for platform location (Moffat and Resnick 2002; see Driscoll and others 2005, for similar results). According to strictly spatial memory models of performance, there should be no age difference on trial 1 performance because the search for the hidden platform on the first trial should be random. Importantly, our analyses indicated that this trial 1 deficit was not attributable to psychomotor impairment in the elderly. We hypothesize that the deficits observed among elderly participants in navigational behavior may be partially attributed to impaired executive and strategic functions that play a significant role in skill acquisition (Anderson 1983). In particular, our observations indicate that elderly individuals are reluctant to disengage from searching in locations that have been adequately explored, suggesting a perseverative component to their behavior or, more generally, the selection of inefficient spatial search strategies. This behavioral effect is illustrated in Figure 1. We therefore hypothesize that the deficits observed among elderly subjects in navigational behavior may be partially attributed to impaired executive and strategic functions and shrinkage of the neural systems that support executive control.Figure 1.Top view of virtual MWT search paths (trial 1) in 5 representative young and old subjects in a previous study (Moffat and Resnick 2002). Older subjects took longer to find the platform and traveled a longer linear distance in their initial search. Inefficient search strategies and a reluctance to disengage from nongoal locations may contribute to this age difference on trial 1.In the present study, we quantified navigational behavior in a sample of men and women whose regional brain volumes were assessed via structural magnetic resonance imaging (MRI). In addition to an extensive cognitive test battery, the individuals in the present study performed a VE place navigation task modeled after the MWT. The purpose of the study was 3-fold: 1) to investigate age differences on a virtual Morris water task (vMWT), regional brain volumes, and other cognitive tests; 2) to investigate associations between regional brain volumes and vMWT performance; and 3) to investigate associations between cognitive tasks and vMWT performance. We hypothesized age-related deficits in spatial navigation as assessed by the vMWT. We hypothesized that a significant portion of variance in vMWT performance would be explained by hippocampal, prefrontal, and striatal volume differences, whereas individual differences in the volume of a control region (primary visual cortex [VC]) would evidence no significant association with navigation performance. In addition, both hippocampal dependent measures of memory as well as frontal lobe executive measures would be positively associated with vMWT performance.Materials and MethodsParticipantsIn this study, we examined cross-sectional differences in the normal aging brain using a sample of healthy volunteers who participate in an ongoing study of brain and cognitive aging. Participants in the present study (18 years and older) were recruited from a pool of more than 140 subjects who underwent MRI scanning and cognitive testing in our laboratory. These participants, who live in the Detroit metropolitan area, were contacted and agreed to return to complete the virtual MWT. Sixty-eight participants who had undergone MRI scanning agreed to participate in the study and met all study qualifications. These participants were divided into 2 groups, hereafter referred to as young and elderly. Persons who reported history of cardiovascular (except treated essential hypertension), neurological, or psychiatric illness, diabetes, head trauma with loss of consciousness for more than 5 min, thyroid problems, treatment for drug and alcohol problems, or a habit of taking 3 or more alcoholic drinks per day were excluded from the study as were participants who use antiseizure medication, anxiolytics, or antidepressants. Persons who experience emotional distress that could interfere with cognitive performance were screened using a geriatric depression questionnaire (Radloff 1977). In addition, participants in our study were classified into those with current or history of hypertension. This was done to investigate the effects on performance of hypertension that is increasingly acknowledged as an important moderator of cognitive and brain aging (Raz and others 2003; Korf and others 2004; Wiseman and others 2004). General cognitive status was assessed by Mini Mental State Examination (Folstein and others 1975). All participants were strongly right handed as screened by the Edinburgh handedness questionnaire (Oldfield 1971). Table 1 presents demographic data for the subject sample. Mean delay between MRI and cognitive testing was 3.8 months (standard deviation [SD] = 2.7), and there were no differences in delay between the young and the old participants t = 0.01, not significant (NS). Before cognitive testing, all participants were interviewed to ensure that they had not experienced any health changes in the interim period. Because we performed a cross-sectional study comparing a group of older participants with a group of younger participants, any small cognitive or brain changes over the 3.8-month interval are unlikely to affect our cross sectional comparisons that average 44 years difference between old and young participants.Table 1Demographic characteristics of participantsNPercent femaleAge (mean ± SE)Education (mean ± SE)Percent hypertensiveYoung327524.5 (0.91)15.16 (0.32)0Old366768.5 (0.92)15.22 (0.52)47.22Note: SE, standard error.ProceduresStructural MRI and Regional VolumetryImage acquisition and processing.Participants in the existing project underwent MRI scanning at Children's Hospital of Michigan on a GE Signa 1.5-T scanner. Regional brain volumes were measured on images acquired with T1-weighted 3-dimensional spoiled gradient recalled (SPGR) sequence with 124 contiguous axial slices, echo time = 5 ms, time repetition = 24 ms, field of view = 22 cm, acquisition matrix 256 × 192, slice thickness = 1.3 mm, and flip angle = 30°. The images were reformatted to correct for undesirable effects of head tilt, pitch, and rotation with BrainImage 2.3.3 software. The procedure is described in detail elsewhere (Raz and others 2004).Volumetry: region selection, demarcation, and tracing.The regions of interest (ROIs), selected on the basis of the hypotheses to be tested in this study, included HC, lateral prefrontal cortex gray matter (LPFC), prefrontal white matter (PFW), caudate nucleus (CN), and the CB. Each of these selected structures has been implicated as contributing to spatial navigation in human neuroimaging or animal studies. Primary VC was selected as a comparison (control) region for which no correlations were expected. To control statistically for individual differences in body size, especially for sexual dimorphism, we used intracranial volume (ICV). All regional brain volumes were corrected for total ICV using a regression analysis of covariance approach. Interrater reliability of all measures exceeded an intraclass correlation of 0.90.The rules for demarcation and tracing of brain regions as well as the estimates of interrater reliability of volume estimates obtained with National Institutes of Health Image software (Version 1.62) are described in detail in previous publications (Raz and others 1997, 2004) and will be presented here in condensed form. Unless otherwise noted, we used the average of left and right hemisphere volumes in statistical analyses.Caudate nucleus.The volume of the head and the body of the CN were estimated from 15 to 20 coronal slices. The most rostral slice was the one on which the CN first appeared, usually lateral to the lateral ventricles. The CN was traced on every other slice (interslice distance 3 mm) until no longer visible.Cerebellum.The cerebellar hemispheres were measured on 32–40 coronal slices. The vermis, the cerebellar peduncles, and the forth ventricle were excluded, whereas the hemispheric gray matter, the cerebellar tonsils, the vellum, and the corpus medullare were included in the tracing of each cerebellar hemisphere. The rostral border was defined as the first slice on which cerebellar gray matter became visible and distinguishable from the cerebellar peduncles, and the caudal border ended when the hemispheres were no longer distinguishable.Hippocampus.HC volume was measured on continuous slices aligned perpendicular to the long axis of the right HC between the mammillary bodies and slice showing the fornices rising from the fimbria. The HC included sectors CA1–CA4, the dentate gyrus, and the subiculum.Intracranial volume.The ICV was estimated from the coronal sections. The operator traced ICV on every eighth slice (a total of 11–12 slices) between the first slice following the orbits and last slice on which brain tissue was visible.Lateral prefrontal cortex.Eight to 12 coronal slices located within 40% of the distance between the genu of the corpus callosum and the frontal pole are included in the prefrontal ROI.Prefrontal cortex white matter.The range is identical to LPFC above and also includes orbitofrontal cortex. The volume includes all white matter on the coronal slice, excluding ventricles and other cerebrospinal fluid spaces.Primary VC (calcarine).The volume of the VC is calculated as the volume of the cortical ribbon lining the calcarine sulcus on the anterior 50% of the coronal slices between the midvermis slice and the occipital pole.Spatial Navigation Assessments: Virtual Water MazeFor the behavioral navigation testing, all subjects completed a vMWT that has been developed and validated previously (Moffat and Resnick 2002). Before beginning the navigation task, practice trials were allowed to familiarize subjects with movement through the environment and to be certain that participants were comfortable with the computer-administered task. The task environment was a circular arena surrounded by several cues, which could be used to guide navigation. Hidden beneath the surface of the arena was a platform, and the task required the participant to locate the platform as quickly as possible. Because the platform was not visible to the participant, it must be located with reference to its position relative to the external objects and cues. When the participant passed over the platform, the platform became visible, and a tune notified the participant that he or she had located the platform. Participants were informed that the platform remained in the same location on each trial and that they should try to remember its location. On each of 6 learning trials, participants were “placed” into one of the 3 quadrants of the pool, which did not contain the platform and facing a different orientation on each trial. The primary dependent measure was the distance traveled on each trial. Following the 6 learning trials, a “probe” trial was performed in which participants explored the same VE with the exception that the platform had been removed. Participants were given 1 min of search time in the probe trial. Individuals who have properly encoded platform location will spend a greater proportion of their time and distance searching in the vicinity of the platform. Thus, the dependent measures were the percentage of distance spent in the goal quadrant and the number of platform intersections (number of times a participant's path would have passed over the platform had it been present).Control TestsPretest training and assessment of joystick visuomotor control.Prior to virtual MWT testing, extensive pretraining was provided to familiarize participants with the VE and with the use of a joystick for movement. This was accomplished with an initial period of experimenter instruction, followed by a period of free exploration of a VE using the joystick. After participants were comfortable with the joystick and had satisfactorily demonstrated their ability to guide themselves to targets designated by the experimenter, the participants underwent a joystick control speed test. During the speed test, participants were required to navigate a long winding corridor as quickly and accurately as possible until they reached a trophy at the goal point. Participants were required to demonstrate their competency with the joystick by completing the corridor in less than 120 s.Visible platform performance.As a further experimental control, a trial was performed in which the platform was visible on the surface of the pool. Subjects were instructed to move onto the platform from the starting location as quickly as possible. This served as an additional control for possible age differences in visual and motor skills. The magnitude of any age differences observed on this task should be considerably smaller than those observed in the vMWT trials and probe trials. The dependent measure was the total distance traveled to the platform.Cognitive TestsCognitive tests were selected based on theoretical and empirical considerations. In particular, we selected tests that measure cognitive skills that have been empirically demonstrated to be related to spatial navigation in previous studies (Moffat and others 1998, 2001). In addition, we selected traditional neuropsychological measures that are sensitive to frontal lobe function to assess the contribution of prefrontal cortex to spatial navigation.“Computation span” (Salthouse 1989) was obtained as a measure of verbal working memory, and the absolute span was used as the index of performance, calculated by summing the number of correct items across all trials.A modified “size-judgment span” (Cherry 1993) was administered as a measure of nonverbal working memory. In this task, participants are orally presented with a list of objects and animals and asked to repeat them in order of size from smallest to largest. Index of performance is total number of correct trials.“Letter comparison” and “pattern comparison” tests (Salthouse and Meinz 1995) were indices of perceptual speed.A computerized version of the “Wisconsin card sorting test” (WCST; Psychological Assessment Resources, Odessa, FL) was administered. Total number of perseverative errors was the index of perseveration.“Vocabulary” (V-3) is a paper-and-pencil test from Educational Testing Service Factor-Referenced Test Kit (Ekstrom 1976) and was used as a measure of word knowledge.“Spatial recall” test was modified from a task described previously (Salthouse 1994). In this task, participants were instructed to memorize the locations of a stimulus in 1 of 7 target locations within a 5 × 5 matrix and then asked to place an X in the location of the target in a blank matrix. Average number of correct items across the 25 trials was the index of performance.“Buildings memory” test is a paper-and-pencil test from the Educational Testing Service Factor-Referenced Test Kit (Ekstrom 1976). After studying a fictitious map for 4 min, participants were shown a blank map and tested on their memory for the location of buildings on the map studied in a multiple-choice format. The index of performance was the number correct minus the number incorrect.Data ReductionTo reduce the data, we analyzed cognitive performance by 4 construct domains: speed of processing, working memory, spatial memory and executive control. Each construct, except executive control, was measured by 2 tests; executive control was assessed by WCST. The working memory domain was assessed by computation span and size judgment span. The spatial memory construct was measured by spatial recall and the buildings memory test. The construct of perceptual speed was measured by letter and pattern comparison tests. Although the domains were determined on the basis of theoretical considerations, we performed principal component analysis on our data in order to verify the postulated factor structure. We conducted principal component analysis followed by varimax rotation on the young and older participants separately. For 32 young participants, the analysis yielded 3 factors that accounted for 78% of the score variance. The first factor was speed (letter comparison and pattern comparison loadings both 0.87, 29% of the variance accounted), the second factor was spatial memory (building memory 0.79, spatial recall 0.88, 25% of the variance), and the third factor was working memory (size judgment 0.87, computation span 0.86, 24% of the variance).In the older age group, a different factor structure was obtained, with the first factor (letter and pattern comparison tests, spatial recall and building memory, loadings ranging between 0.74 and 0.91) accounting for 59% of the variance and the second factor (size judgment and computation span measures, loadings 0.87 and 0.88) accounting for additional 16%. The difference illustrates an often observed tendency for dedifferentiation of cognitive abilities with age (Ghisletta and Lindenberger 2004). For statistical analyses, we converted all raw scores from cognitive tests to standardized (Z) scores and computed composites consisting of the average of test performance in common cognitive domains as described above.Statistical Approach and AnalysisAs noted earlier, there were 3 overarching aims of the study: 1) to investigate basic age differences in vMWT performance, cognitive performance, and regional brain volumes; 2) to investigate correlations between vMWT performance and regional brain volumes; and 3) to investigate correlations between cognitive measures and vMWT performance. For all 3 aims, our hypotheses were tested within the framework of the general linear model with age and sex as independent variables and appropriate dependent variables. The primary dependent variable on the vMWT was the total distance traveled on each trial. Because the distribution of that index was positively skewed, a logarithmic transformation was applied. Following the evaluation of hypothesis 1, which established basic age effects on vMWT, cognitive tests and ROI's, analyses were performed incorporating the regional brain volumes into the identical models, separately for each brain region with distance traveled on the 6 trials of vMWT as dependent variable. This allowed us to evaluate the contribution of each brain region to vMWT performance (hypothesis 2). Lastly, we deleted brain volumes from our statistical models and incorporated cognitive test scores allowing us to assess the correlation between performance on vMWT and other cognitive domains (hypothesis 3). To assess the specificity of possible associations between vMWT performance and specific cognitive domains, we also evaluated correlations with vocabulary scores as a control measure for which we expected no significant positive correlations.ResultsAge and Sex Effects on Brain, vMWT, and Cognitive MeasuresRegional Brain VolumesThe effects of age and sex on regional brain volumes were evaluated in the framework of a general linear model with age group and sex as independent variables and regional volumes as a 6-level repeated measure. Before being entered into the equations, the volumes of all brain regions were adjusted for individual differences in ICV. Overall, there was a significant main effect of age (F1,65 = 34.06, P < 0.001; η2 = 0.34) and no significant sex differences (all main effects and interactions F values < 1). The significant age × ROI interaction, F5,325 = 10.09, P < 0.001, η2 = 0.13, indicated that the magnitude of the age differences varied across ROIs. There was a significant effect of hemisphere (F1,65 = 6.19, P < 0.05, η2 = 0.09), with right hemisphere volumes, on average, larger than left but no interactions involving hemisphere were significant (but see below for the analysis of simple effects).Analysis of simple effects with univariate F-tests indicated that the differences between the age groups were significant in LPFC (F1,65 = 46.38, P < 0.001), the HC (F1,65 = 25.34, P < 0.001), the caudate (F1,65 = 21.07, P < 0.001), the CB (F = 11.37, P < 0.001), and PFW (F1,65=6.21, P < 0.05) but not in the primary VC (F1,65 = 2.82, P = 0.10). Comparison of effect sizes revealed that the magnitude of age differences in the prefrontal volume was significantly greater than that for the pericalcarine volume. Other regional differences fell between the 2 values and did not differ significantly from each other (see Table 2).No lateral differences were found in LPFC (F < 1), VC (F < 1), the caudate (F1,65 = 1.71, NS), or the CB (F1,65 = 2.57, NS). Right PFW was significantly larger than left, F1,65 = 13.87, P < 0.001. Left HC was larger than the right but only in the young (F1,30 = 25.74, P < 0.001) not in the older participants (F < 1).We examined the possible influence of hypertension in the older age group on the observed volume differences. Only the older age group could be considered in these analyses as there were no young subjects with a diagnosis of hypertension. Separate analyses of the older participants revealed a trend for the diagnosis of hypertension to be associated with overall smaller regional brain volumes (F1,32 = 3.53, P < 0.07, η2 = 0.09).Virtual Water Maze: LearningThere was a significant main effect of age, F1,65 = 84.34, P < 0.001, η2 = 0.56, with younger subjects outperforming older subjects and a main effect of trial, F5,325 = 9.11, P < 0.001, η2 = 0.12, in which performance improved across trials. There was a nonsignificant trend for an age × trial interaction: F5,325 =1.97, P < 0.09, η2 = 0.03. The linear component of the trial effect was significant, F1,65 = 23.29, P < 0.001, η2 = 0.26, as was the quadratic component for an age × trial interaction: F1,65 = 5.99, P < 0.05 η2 = 0.08. Changes in performance across trials for both age groups are depicted in Figure 2. As seen in Figure 2, there was an unexpected increase in distance traveled on trial 4 in both the younger and older groups. This effect was an artifact of the random placement procedure across the 6 learning trials. On trial 4, participants were placed at the maximum possible distance and facing a direction opposite from the platform resulting in the increase in distance traveled. We performed all analyses again deleting trial 4 from our repeated measure and obtained virtually identical results. In particular, the task resulted in the same robust age and trial effects. In all remaining analyses, all 6 trials are used as dependent measure.Figure 2.Total distance traveled in the vMWT as a function of age and trial. Performance improved across trials and older participants traversed a longer path in locating the hidden platform on all 6 trials.The effects of hypertension on vMWT performance were investigated in the older participants. Within the older group, there was a trend for normotensive participants to perform better than hypertensive participants: F1,32 = 3.94, P = 0.056, η2 = 0.11.Virtual Water Maze: Probe TrialIn this analysis, the new dependent variable was the distance traveled in the quadrant of the pool that formerly contained the platform. The results indicated that younger (Mean = 52.50, SD = 22.16) individuals spent more of their distance searching in the goal quadrant than the older (Mean = 28.17, SD = 20.06) participants: main effect of age, F1,64 = 6.49, P = 0.013. Similarly, there was a trend for younger subjects (Mean = 2.94, SD = 1.56) to have more frequent platform intersections than their older (Mean = 1.97, SD = 1.34) counterparts: F1,64 = 3.69, P = 0.059. We observed neither a main effect of sex nor significant interactions.Cognitive TestsThe descriptive statistics and the effect size indices (Cohen's d) for all cognitive tests by age group are presented in Table 3. The distribution of perseverative errors was skewed and was entered into the analysis after a logarithmic transformation.Table 2Descriptive statistics (mean ± standard error of mean) and effect sizes for regional brain volumes (cm3) in the 2 age groupsLPFCPFWHCCBCNPVCYoung20.94 (0.40)43.51 (1.07)7.39 (0.13)133.06 (1.98)10.15 (0.17)5.36 (0.14)Old16.30 (0.38)37.59 (1.01)6.36 (0.12)123.12 (1.87)8.77 (0.16)5.23 (0.13)Effect size (d)2.010.971.440.881.410.37P value<0.0010.001<0.0010.001<0.0010.12Note: Lateral prefontal cortex white matter (PFW); Primary visual cortex (PVC).Table 3Descriptive statistics (mean ± standard error of mean) and effect sizes (d) for cognitive test scores in the 2 age groupsWCSTaSpatialVocabBuildSizeCSPANPatcompLetcompYoung6.75 (0.64)5.33 (0.10)18.25 (1.14)7.86 (0.52)9.84 (0.31)26.94 (3.07)41.21 (1.26)24.87 (0.72)Old19.23 (2.08)4.22 (0.13)26.52 (1.25)4.26 (0.56)8.50 (0.32)18.06 (1.90)32.19 (0.97)18.26 (0.86)Effect size (d)1.311.41−1.161.120.720.611.381.40P value<0.001<0.001<0.001<0.0010.0040.014<0.001<0.001Note: Wisconsin card sorting test (WCST) perseverative errors; spatial recall (Spatial); vocabulary (Vocab); buildings memory (Build.); size judgment (Size); computation span (CSPAN); pattern completion (Patcomp); letter completion (Letcomp). One older subject had missing data on Letcomp, making for that test the total N = 67.aScoring scale is reversed. Higher scores reflect poorer performance.The results of the analysis revealed a significant main effect of age (F1,64 = 10.87, P < 0.01, η2 = 0.15) and no sex differences (F1,64 = 1.56, NS, η2 = 0.02). However, the magnitude of age differences varied across the tested domains, as indicated by a significant age × test interaction, F7,448 = 15.35, P < 0.001, with Hyunh-Feldt correction, η2 = 0.19. Univariate analyses showed that the age differences (favoring the younger participants) were the smallest in the working memory domain (F1,64 = 6.05, P < 0.05 for computation span and F1,64 = 9.49, P < 0.01 for size judgment span) in comparison to the other domains: speed (F1,64 = 33.64, P < 0.001 for pattern comparison and F1,64 =32.92, P < 0.001 for letter comparison); spatial memory (F1,64 =33.95, P < 0.001 for spatial recall and F1,64 = 20.64, P < 0.001 for building memory); and perseveration: F1,64 =38.37, P < 0.001. The effect of age on vocabulary was significant (F1,64 = 23.70, P < 0.001) but in the opposite direction (favoring the older participants).Regional Brain Volumetry and Virtual MWT PerformanceBecause high correlations among the regional volumes would result in multicollinearity if all entered in the same model, we performed a separate analysis for each ROI. Before being entered into the equations, the volumes of all brain regions were adjusted for individual differences in ICV. In the interest of brevity, main effects of age and trial on vMWT performance that were reported above are not reported in these analyses. All zero-order correlations between ROI's and vMWT performances on each trial are presented in Table 4.Table 4Correlations between regional brain volumes and distance traveled on each trial of the vMWTTrial 1aTrial 2aTrial 3aTrial 4aTrial 5aTrial 6aLPFC−0.28**−0.47**−0.43**−0.40**−0.40**−0.37**PFW−0.25*−0.40**−0.41**−0.31**−0.23−0.27*HC−0.17−0.33**−0.38**−0.38**−0.30**−0.36**CB−0.10−0.19−0.34**−0.34**−0.31*−0.27*CN−0.16−0.43**−0.39**−0.50**−0.47**−0.46**PVC−0.11−0.39**−0.15−0.100.03−0.10Note: Lateral prefrontal cortex white matter (PFW); Hippocampus (HC); Primary visual cortex (PVC). Correlations were essentially the same for right and left hemisphere volumes of each structure (data not shown).aScoring scale is reversed. Higher scores reflect poorer performance.*P < 0.05; ** P < 0.01.The inclusion of hippocampal volume as an independent variable revealed a complex pattern of differences in performance reflected in a significant interaction between age, trial, and HC volume: F5,300 = 2.37, P = 0.04, η2 = 0.04, although the main effect of HC on vMWT performance was NS (F1,60 = 1.18, NS, η2 = 0.02). Analysis of simple effects showed that the association between HC volume and maze performance was significant only among the younger participants and only on the first trial with young adults with larger HC accumulating shorter distances in their initial search for the virtual platform (r = −−0.38, P < 0.05).The analyses of the influence of other regional volumes revealed a less complex pattern of results. Larger volumes of prefrontal gray matter and PFW were linked to better vMWT performance: F1,60 = 3.89, P = 0.05, η2 = 0.06 and F1,60 = 7.28, P = 0.009, η2 = 0.11, respectively. A similar association between better vMWT performance and larger volumes was observed for the CN: F1,60 = 11.07, P < 0.001, η2 = 0.16. A trend in the same direction was observed for the CB volume: F1,60 = 2.89, P = 0.09, η2 = 0.05. Consistent with our hypothesis, VC, designated as a control region, showed no significant association with vMWT performance: main effect F < 1. In none of the models were there significant interactions between the regional volumes and age, sex, or trial.In the models that included brain regional volumes, the effect of hypertension was significant only in conjunction with the caudate and HC volumes: F1,31 = 7.43, P = 0.01, η2 = 0.19 and F1,31 = 4.48, P < 0.05, η2 = 0.13, respectively. For all other ROI models, the trends for hypertension effect on performance were in the same direction but failed to reach the 0.05 level of significance. The probability levels of the trends ranged from P = 0.056 to P = 0.20, all NS.Cognitive Test Performance and Virtual MWT PerformanceAll zero-order correlations between cognitive composites and vMWT performance on each trial are presented in Table 5. The analyses revealed that performance on executive function tests predicted vMWT performance: participants with better working memory (F1,64 = 3.97, P = 0.05) or fewer perseverative errors on WCST (F1,64 = 7.98, P = 0.006) attained higher vMWT scores. There were no significant interactions involving age, sex, or trial.Table 5Correlations between cognitive domains and distance traveled on each trial of the vMWTTrial 1aTrial 2aTrial 3aTrial 4aTrial 5aTrial 6aWorking memory0.00−0.28*−0.39**−0.28*−0.24*−0.33**Spatial memory−0.11−0.41**−0.47**−0.62**−0.38**−0.57**Perseverationa0.170.38**0.42**0.50**0.34**0.55**Perceptual speed−0.13−0.46**−0.48**−0.52**−0.25*−0.53**Vocabulary0.090.160.120.33**0.22−0.09aScoring scale is reversed. Higher scores reflect poorer performance.*P < 0.05, **P < 0.01.Better spatial memory was associated with better performance on vMWT as indicated by the significant main effect of spatial memory: F1,64 = 17.68, P < 0.001, η2 = 0.22. A trend for trial by spatial memory interaction (F5,320 = 2.17, P = 0.057, η2 = 0.03) suggested that spatial memory may be linked to performance on each trial except the first (see Table 5). A similar pattern of results was observed for perceptual speed: main effect F1,63 = 7.15, P = 0.01, η2= 0.10. Individuals who were faster on the perceptual speed tests performed better on the vMWT. Moreover, a significant interaction between perceptual speed and trial (F5,315 = 2.52, P = 0.03, η2 = 0.04) indicated that perceptual speed was positively associated with performance on each trial but the first (see Table 5).When vocabulary scores, intended as a measure of general crystallized abilities, were introduced into the model, we observed an unexpected significant main effect on vMWT performance: F1,64 = 5.77, P = 0.02, η2 = 0.08. Individuals with higher vocabulary scores performed more poorly on the vMWT. There were no significant interactions.DiscussionThe results of the present study replicate extant findings of age differences in spatial navigation (Moffat and others 2001; Moffat and Resnick 2002; Driscoll and others 2005) and contribute new evidence toward understanding of their neural substrates. We found that age effects on VE place learning are indeed substantial; older individuals traversed a longer linear distance in solving the learning trials of the vMWT. On a probe trial, a measure of retention of spatial location, younger individuals traveled a greater distance in the vicinity of the platform and had more frequent platform intersections relative to older subjects. These findings indicate that younger individuals learned and retained more accurate knowledge of platform location than their older counterparts. In search of neural correlates of the observed behavioral differences, we found that larger LPFC, PFW, and CN volumes were positively associated with navigational skill. However, individual differences in HC volume did not explain the age-related deficit in navigation performance. In addition, we found that success in spatial navigation was associated with superior executive skills, spatial memory, and speed of processing. Notably, the observed relationships between regional brain volumes, cognitive skills, and navigation performance were specific and dissociable from the control brain region (primary VC) and control cognitive variables (vocabulary).In addition, our study replicated our earlier findings demonstrating that the age difference in vMWT performance emerges on the very first trial (see Figs 1 and 2) arguing against a strictly hippocampal/episodic memory explanation of the age difference. This behavioral difference may reflect planning, strategy selection or perseveration (i.e., persistence in searching of the same region despite ample opportunity to determine that a particular pool location has been searched adequately), cognitive skills that are dependent on prefrontal executive resources. As well, it is important to note that within a trial, participants must remember preceding moves and locations and continually update this information to avoid returning to the same (incorrect) locations. Although clearly memory related, such within-trial immediate recall might best be characterized as requiring working memory, an age-sensitive ability (Salthouse and Meinz 1995) that has also been shown to be dependent on the prefrontal regions (Courtney and others 1998; Smith and Jonides 1999; D'Esposito and others 2000).The most important evidence supporting our view that vMWT performance is influenced by frontal systems is the fact that both the gray and white matter volumes of the prefrontal cortex were positively associated with vMWT performance. In addition, 2 cognitive measures of executive/frontal function (the number of perseverative errors and working memory span) were both associated with performance on the vMWT. This strongly suggests that successful navigation in humans require substantial contribution from prefrontal circuits and associated cognitive systems. It should be noted that frontal lobe contributions to vMWT performance may not be restricted to the first trial but may extend across several trials, as evidenced by the significant correlations between frontal lobe volumes and vMWT performance across all 6 trials (see Table 4). However, it should also be acknowledged that our primary measure of frontal lobe executive resources (perseveration on the WCST) predicted performance on trials 2–6 but not trial 1 (see Table 5). Although this may be attributable to chance variation in a single study, it is also possible that other frontal lobe measures (or other cognitive domains) may better capture individual differences in trial 1 performance.Studies in nonhuman species confirm important contributions from prefrontal systems in solving the MWT. The respective roles of the frontal cortex and hippocampal system have been delineated by adopting both the place and response versions of the MWT. In the former, the platform remains in the same position while the release site varies, thus requiring a place strategy for successful solution. In the latter, both the platform and release site vary from trial to trial with the spatial relationship between the release site and platform location held constant (e.g., a fixed distance to the right). Several studies have reported double dissociations with hippocampal and or fimbria/fornix lesions impairing place learning in the place version of the MWT and frontal cortex lesions impairing response learning in the MWT (de Bruin and others 1997, 2001). It is noteworthy that we did not specifically assess response learning in our vMWT. However, as we outlined above, even in the place MWT, there are likely substantial executive demands required and at least one study observed impairments in the place version of the MWT following temporary inactivation of the orbital frontal cortex by tetrodotoxin (Vafaei and Rashidy-Pour 2004). Interestingly, consistent with the age differences observed in our study, this impairment was evident on the very first block of trials, suggesting that frontal dysfunction may impair MWT place learning early in the acquisition phase before there are significant mnemonic demands. Although separate tasks may be designed to investigate place and response strategies, respectively, animals (Packard and McGaugh 1996) and humans likely employ both strategies to some extent concurrently or successively across trials.In addition to positive associations with frontal cortex measures, we also observed a positive correlation between CN volume and navigation performance. These findings are compatible with previous studies in humans and other mammals. The CN plays an important role in learning and spatial memory and is often activated in young subjects during virtual navigation tasks in functional imaging studies (Maguire and others 1998; Moffat and others 2006). Because of its extensive connections with the LPFC and HC (Alexander and others 1986), the CN is a part of distributed frontostriatal and striato-hippocampal systems, which are vulnerable to aging (Raz 2000; Raz 2003). Studies investigating contributions of the CN to human spatial navigation suggest that the CN may work in concert with hippocampal systems (Voermans and others 2004) and may play a role in nonspatial or procedural response components of spatial behavior (Hartley and others 2003; Iaria and others 2003). As the MRI methods used in this study were not geared to assessment of the integrity of corticocortical and corticosubcortical connections, the intriguing question of the joint effects of multiple systems on navigation remains open.Somewhat unexpectedly, we observed that hippocampal volume was associated with vMWT performance in the young but not in the elderly participants. The extant literature on HC memory relations in human aging consists primarily of studies that relied on verbal and visual memory tasks with positive associations between HC volume and episodic memory reported in some samples or specific groups of participants (Soininen and others 1994; Raz and others 1998; Jack and others 2000; Petersen and others 2000; but c.f. Van Petten 2004). A notable exception to that focus is a recent study that assessed the effects of age and HC volume and navigation performance in the framework of a virtual MWT similar to the one used in the current study (Driscoll and others 2003). In that investigation, HC volume was positively correlated with vMWT performance across all subjects. However, when these researchers incorporated age into their model as we did in this study, the relationship between HC volume and vMWT performance was reduced to nonsignificant. Thus, to our knowledge, the only studies to investigate the contributions of HC volume to age differences in navigation have produced similar results.Although, the lack of a relationship between HC volume and vMWT performance in the elderly group should be interpreted with caution, a possible explanation for this finding may be found in the animal literature. It is known that animals have an initial preference for the use of a place strategy in the MWT (de Bruin and others 1997). Thus, when confronted with the response version of the MWT, animals first adopt a place strategy, which when unsuccessful is replaced by a response strategy. Importantly, Barnes has demonstrated that the preference for initial selection of a place strategy is age dependent with younger animals adopting a place (hippocampal) strategy, whereas older animals adopt a response (extrahippocampal) strategy. If a similar phenomenon is at work in humans, then it could be argued that our younger and elderly participants were adopting different solution strategies that were differentially dependent on the HC with younger participants adopting a place (hippocampal) strategy and older participants adopting a reference (extrahippocampal strategy), resulting in positive and null associations with hippocampal structure, respectively. Conceivably, after a reference strategy proved unsuccessful, elderly participants may switch to a place strategy. However, the number of trials employed in the present study was not sufficient to test this hypothesis. In the future, the design of distinct place and response versions of the vMWT incorporating more learning trials may help to address this question.It is also important to note that we assessed only the volume of the HC and not other important neurochemical and/or functional measures of HC integrity. An alternative conceptualization is that a variety of “functional” changes in the HC that were not reflected in our volume measurement may play a more important role in cognitive aging (Driscoll and others 2003; Grady and others 2003; Moffat and others 2006).In the framework of a cross-sectional study, it is impossible to discern the influence of longstanding individual differences in brain and cognitive variables from true longitudinal declines. It is possible that smaller frontal and striatal volumes that predict poorer performance in the young and the elderly alike do so for age-specific reasons. For the younger participants, the frontal executive links may reflect developmental or even inborn differences in the size of the prefrontal cortex and white matter. In the older subgroup, better performance may be due to preservation of these age-vulnerable regions.One of the factors that could have affected performance and the relationships between brain variables and navigation is hypertension, a commonly acknowledged negative modifier of age trajectories (Breteler and others 1994; Raz and Rodrigue 2006). An interesting trend was observed for the influence of hypertension on brain and cognitive performance. The diagnosis of hypertension is usually acquired in older age, and the current sample of older adults was relatively small to allow a powerful test of that hypothesis. However, our results indicated that hypertensive participants tended to show poorer navigation skills than their normotensive counterparts. Thus, hypertension and possibly other vascular risk factors need to be taken into account in future evaluation of age differences in cognitive performance.In summary, the present study confirms robust age-related deficits in navigational skill. It indicates that successful navigation does not depend solely on the HC but is also associated with larger regional volumes of multiple cortical and subcortical structures and draws on age-sensitive executive abilities as well as task-specific spatial memory resources.Conflict of Interest: None declared.AguirreGKDetreJAAlsopDCD'EspositoMThe parahippocampus subserves topographical learning in manCereb Cortex19966823829AlexanderGEDeLongMRStrickPLParallel organization of functionally segregated circuits linking basal ganglia and cortexAnnu Rev Neurosci19869357381AndersonJRThe architecture of cognition1983NJ: Lawrence Erlbaum AssociatesHillsdaleBarnesCAMemory deficits associated with senescence: a neurophysiological and behavioral study in the ratJ Comp Physiol Psychol19799374104BarrashJA historical review of topographical disorientation and its neuroanatomical correlatesJ Clin Exp Neuropsychol199820807827BegegaACienfuegosSRubioSSantinJLMirandaRAriasJLEffects of ageing on allocentric and egocentric spatial strategies in the Wistar ratBehav Processes2001537585BretelerMMvan SwietenJCBotsMLGrobbeeDEClausJJvan den HoutJHvan HarskampFTangheHLde JongPTvan GijnJCerebral white matter lesions, vascular risk factors, and cognitive function in a population-based study: the Rotterdam studyNeurology19944412461252BurnsPCNavigation and mobility of older driversJ Gerontol Soc Sci199954BS49S55CherryKEParkDCIndividual difference and contextual variables influence spatial memory in younger and older adultsPsychol Aging199384517526CourtneySMPetitLHaxbyJVUngerleiderLGThe role of prefrontal cortex in working memory: examining the contents of consciousnessPhilos Trans R Soc Lond B Biol Sci199835318191828de BruinJPMoitaMPde BrabanderHMJoostenRNPlace and response learning of rats in a Morris water maze: differential effects of fimbria fornix and medial prefrontal cortex lesionsNeurobiol Learn Mem200175164178de BruinJPSwinkelsWAde BrabanderJMResponse learning of rats in a Morris water maze: involvement of the medical prefrontal cortexBehav Brain Res1997854755D'EspositoMPostleBRRypmaBPrefrontal cortical contributions to working memory: evidence from event-related fMRI studiesExp Brain Res2000133311DriscollIHamiltonDAPetropoulosHYeoRABrooksWMBaumgartnerRNSutherlandRJThe aging hippocampus: cognitive, biochemical and structural findingsCereb Cortex20031313441351DriscollIHamiltonDAYeoRABrooksWMSutherlandRJVirtual navigation in humans: the impact of age, sex, and hormones on place learningHorm Behav200547326335EkstromRFrenchJHarmanHManual for kit of factor-referenced cognitive tests1976NJ: Educational Testing ServicePrincetonFolsteinMFFolsteinSEMcHughPRMini-mental state. A practical method for grading the cognitive state of patients for the clinicianJ Psychiatr Res197512189198GallagherMPelleymounterMASpatial learning deficits in old rats: a model for memory decline in the agedNeurobiol Aging19889549556GhislettaPLindenbergerUStatic and dynamic longitudinal structural analyses of cognitive changes in old ageGerontology2004501216GradyCLMcIntoshARCraikFIAge-related differences in the functional connectivity of the hippocampus during memory encodingHippocampus200313572586GronGWunderlichAPSpitzerMTomczakRRiepeMWBrain activation during human navigation: gender-different neural networks as substrate of performanceNat Neurosci20003404408HartleyTMaguireEASpiersHJBurgessNThe well-worn route and the path less traveled: distinct neural bases of route following and wayfinding in humansNeuron200337877888IariaGPetridesMDagherAPikeBBohbotVDCognitive strategies dependent on the hippocampus and caudate nucleus in human navigation: variability and change with practiceJ Neurosci20032359455952IngramDKComplex maze learning in rodents as a model of age-related memory impairmentNeurobiol Aging19889475485JackCRJrPetersenRCXuYO'BrienPCSmithGEIvnikRJBoeveBFTangalosEGKokmenERates of hippocampal atrophy correlate with change in clinical status in aging and ADNeurology200055484489KatayamaKTakahashiNOgawaraKHattoriTPure topographical disorientation due to right posterior cingulate lesionCortex199935279282KennedyKMRazNAge, sex and regional brain volumes predict perceptual-motor skill acquisitionCortex200541560569KirasicKCSpatial cognition and behavior in young and elderly adults: implications for learning new environmentsPsychol Aging199161018KirasicKCAllenGLHaggertyDAge-related differences in adults' macrospatial cognitive processesExp Aging Res1992183339KleinDASteinbergMGalikESteeleCSheppardJMWarrenARosenblattALyketsosCGWandering behaviour in community-residing persons with dementiaInt J Geriatr Psychiatry199914272279KorfESWhiteLRScheltensPLaunerLJMidlife blood pressure and the risk of hippocampal atrophy: the Honolulu Asia aging studyHypertension2004442934LindnerMDReliability, distribution, and validity of age-related cognitive deficits in the Morris water mazeNeurobiol Learn Mem199768203220LukoyanovNVAndradeJPDulce MadeiraMPaula-BarbosaMMEffects of age and sex on the water maze performance and hippocampal cholinergic fibers in ratsNeurosci Lett1999269141144MaguireEABurgessNDonnettJGFrackowiakRSFrithCDO'KeefeJKnowing where and getting there: a human navigation networkScience1998280921924McLayRNFreemanSMHarlanREKastinAJZadinaJETests used to assess the cognitive abilities of aged rats: their relation to each other and to hippocampal morphology and neurotrophin expressionGerontology199945143155MoffatSDElkinsWResnickSMAge differences in the neural systems supporting human allocentric spatial navigationNeurobiol Aging200627796572MoffatSDHampsonEHatzipantelisMNavigation in a virtual maze: sex differences and correlation with psychometric measures of spatial ability in humansEvol Hum Behav1998197387MoffatSDResnickSMEffects of age on virtual environment place navigation and allocentric cognitive mappingBehav Neurosci2002116851859MoffatSDZondermanABResnickSMAge differences in spatial memory in a virtual environment navigation taskNeurobiol Aging200122787796MorrisRGGarrudPRawlinsJNO'KeefeJPlace navigation impaired in rats with hippocampal lesionsNature1982297681683NewmanMKaszniakASpatial memory and aging: performance on a human analog of the Morris water mazeAging Neuropsychol Cogn200078693OldfieldRCThe assessment and analysis of handedness: the Edinburgh inventoryNeuropsychologia1971997113PackardMGMcGaughJLInactivation of hippocampus or caudate nucleus with lidocaine differentially affects expression of place and response learningNeurobiol Learn Mem1996656572PassiniRRainvilleCMarchandNJoanetteYWayfinding in dementia of the Alzheimer type: planning abilitiesJ Clin Exp Neuropsychol199517820832PetersenRCJackCRJrXuYCWaringSCO'BrienPCSmithGEIvnikRJTangalosEGBoeveBFKokmenEMemory and MRI-based hippocampal volumes in aging and ADNeurology200054581587RadloffLSThe CES-D scale: a self-report depressive scale for research in the general populationJ App Psychol Meas19771385401RazNCraikFIMSalthouseTAAging of the brain and its impact on cognitive performance: integration of structural and functional findingsHandbook of aging and cognition—II2000Mahwah, NJErlbaumRazNGunningFMHeadDDupuisJHMcQuainJBriggsSDLokenWJThorntonAEAckerJDSelective aging of the human cerebral cortex observed in vivo: differential vulnerability of the prefrontal gray matterCereb Cortex19977268282RazNGunning-DixonFMHeadDDupuisJHAckerJDNeuroanatomical correlates of cognitive aging: evidence from structural magnetic resonance imagingNeuropsychology19981295114RazNGunning-DixonFHeadDRodrigueKMWilliamsonAAckerJDAging, sexual dimorphism, and hemispheric asymmetry of the cerebral cortex: replicability of regional differences in volumeNeurobiol Aging200425377396RazNRodrigueKMDifferential aging of the brain: patterns, cognitive correlates and modifiersNeurosci Biobehav Rev2006ForthcomingRazNRodrigueKMAckerJDHypertension and the brain: vulnerability of the prefrontal regions and executive functionsBehav Neurosci200311711691180RazNRodrigueKMKennedyKMHeadDGunning-DixonFAckerJDDifferential aging of the human striatum: longitudinal evidenceAm J Neuroradiol2003249184956RazNWilliamsonAGunning-DixonFHeadDAckerJDNeuroanatomical and cognitive correlates of adult age differences in acquisition of a perceptual-motor skillMicrosc Res Tech2000518593SalthouseTAThe aging of working memoryNeuropsychology19948535543SalthouseTAMeinzEJAging, inhibition, working memory, and speedJ Gerontol B Psychol Sci Soc Sci199550P297P306SalthouseTAMitchellDRSkovronekEBabcockRLEffects of adult age and working memory on reasoning and spatial abilitiesJ EXP Psychol Learn Mern Cogn198915350716SmithEEJonidesJStorage and executive processes in the frontal lobesScience199928316571661SoininenHSPartanenKPitkanenAVainioPHanninenTHallikainenMKoivistoKRiekkinenPJSr.Volumetric MRI analysis of the amygdala and the hippocampus in subjects with age-associated memory impairment: correlation to visual and verbal memoryNeurology19944416601668VafaeiAARashidy-PourAReversible lesion of the rat's orbitofrontal cortex interferes with hippocampus-dependent spatial memoryBehav Brain Res20041496168Van PettenCRelationship between hippocampal volume and memory ability in healthy individuals across the lifespan: review and meta-analysisNeuropsychologia20044213941413VoermansNCPeterssonKMDaudeyLWeberBVan SpaendonckKPKremerHPFernandezGInteraction between the human hippocampus and the caudate nucleus during route recognitionNeuron200443427435WilknissSMJonesMGKorolDLGoldPEManningCAAge-related differences in an ecologically based study of route learningPsychol Aging199712372375WisemanRMSaxbyBKBurtonEJBarberRFordGAO'BrienJTHippocampal atrophy, whole brain volume, and white matter lesions in older hypertensive subjectsNeurology20046318921897
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B716D778D468DEFC32EE4B62F70AA93A8601B17.txt b/test/dataset/in/resources/corpus/Clean_0B716D778D468DEFC32EE4B62F70AA93A8601B17.txt
deleted file mode 100644
index 2f3c18a..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B716D778D468DEFC32EE4B62F70AA93A8601B17.txt
+++ /dev/null
@@ -1 +0,0 @@
- geront Gerontologistgeront The Gerontologist The Gerontologist 0016-9013 1758-5341 Oxford University Press 66710.1093/geront/45.5.667 EXERCISE Characteristics of Physical Activity Programs for Older Adults: Results of a Multisite Survey Hughes Susan L. DSW 1 Williams Barbara PhD 2 Molina Lourdes C. MPH 1 Bayles Constance PhD 3 Bryant Lucinda L. PhD, MHSA 4 Harris Jeffrey R. MD, MPH 2 Hunter Rebecca MEd 5 Ivey Susan MD, MHSA 6 Watkins Ken PhD 7 Address correspondence to Susan L. Hughes, DSW, Center for Research on Health and Aging, University of Illinois at Chicago, 1747 W. Roosevelt Road, Room 558, M/C 275, Chicago, IL 60608. E-mail: shughes@uic.edu 10 2005 45 5 667 675 8 2 2005 23 8 2004 The Gerontological Society of America 2005 Purpose: Although increased participation in physical activity by older adults is a major public health goal, little is known about the supply and use of physical activity programs in the United States. Design and Methods: Seven academic centers in diverse geographic areas surveyed physical activity programs for older adults. Five sites conducted surveys by mail with telephone follow-up, and two administered surveys primarily by telephone. Reported program attendance rates were compared with local census data to assess unmet needs. Results: Of the 2,110 targeted facilities, 77% responded. Aerobic programs were offered by 73%, flexibility by 47%, and strength training by 26%. Commercial gyms or YMCAs, senior centers, park or recreation centers, and senior-housing facilities offered 90% of available programs. The 2000 Census enumerated 1,123,401 total older adults across the seven sites. Facilities reported 69,634 individuals as current weekly program participants, equaling 6% of the sites' total older-adult population. This percentage varied from 3% in Pittsburgh to 28% in Colorado. Implications: Based on conservative estimates of demand, the number of physical activity programs would have to increase substantially (by 78%) to meet the needs of older adults. The data also indicate the need to develop more strength-training programs and to engage a higher percentage of older adults in these programs. There is a clear need to stimulate demand for programs through health promotion. Exercise Needs and demand Providers Facilities hwp-legacy-fpage 667 hwp-legacy-dochead RESEARCH ARTICLE Current estimates predict that the number of adults over the age of 65 years (older adults) will increase from 13% of the U.S. population in 2000 to 20% by 2030, with the most rapid expansion occurring among those aged 85 and older (Federal Interagency Forum on Aging-Related Statistics, 2000). Although 17% of younger adults aged 18 to 64 have disabilities, disability prevalence increases to 50% in older adults (Centers for Disease Control and Prevention [CDC], 2001). Physical inactivity, a known modifiable risk factor for future disability, also increases with age (Huang et al., 1998; Hubert, Bloch, & Fries, 1993; LaCroix, Guralnik, Berkman, Wallace, & Satterfield, 1993; Seeman et al., 1995; Strawbridge, Cohen, Shema, & Kaplan, 1996). Physical activity interventions have demonstrated multiple benefits among older adults, including improved functioning (Fiatarone et al., 1994; Singh, 2002), improved health-related quality of life, and decreased levels of depression (CDC, 1996). Physical activity also has been shown to benefit older adults with specific chronic conditions, including arthritis, heart disease, and diabetes (Stahle, Nordlander, & Bergfeldt, 1999). Specifically, studies show that regular physical activity reduces the risk of dying prematurely and of developing diabetes, high blood pressure, and colon cancer; reduces feelings of depression and anxiety; helps control weight and maintains bone mineral density; and promotes psychological well-being (Binder, Birge, & Kohrt, 1996; Blumenthal et al., 1991; Fried et al., 1998; Preisinger et al., 1996; Singh et al., 1999). Despite these documented benefits, estimates suggest that 33% of men and 50% of women over the age of 75 engage in no physical activity (CDC, 2004). The prevalence of inactivity varies by racial and ethnic group and by gender, from 47% in White women aged 75 and older to 59% in older Black men and 61% in older Black women (Rejeski, Brawley, McAuley, & Rapp, 2000). Healthy People 2010 national health objectives recommend an increase in the proportion of adults who engage in regular, moderate physical activity for 30 min or more per day or vigorous physical activity 3 or more days per week for 20 min or more per occasion (U.S. Department of Health and Human Services, 2000). Current estimates suggest that, of older adults who engage in any physical activity, only 25% aged 65 to 74 and 15% aged 75 and older meet these recommendations for vigorous or moderate physical activity (CDC, 2002). A Robert Wood Johnson Foundation (2001) report identified the removal of barriers to increased physical activity among older adults as a major current national public health need. However, a potential key barrier to engaging in physical activity—the available supply of affordable physical activity programs for older adults—has not been studied to date. In this article we address this important issue by presenting findings from a multisite survey of physical activity programs for older adults in seven diverse geographic locations across the United States. Members of the Healthy Aging Research Network (HAN) conducted the survey between January and July 2003 with support from a CDC program—the Prevention Research Centers (PRCs). The goal of the PRC program is to support the development of academic and community partnerships to conduct prevention research, and each PRC conducts research and demonstration projects to address the most pertinent public health problems (Doll & Kreuter, 2001). The HAN survey built on the methods and findings of an earlier survey conducted by the Senior Health Alliance Promoting Exercise (SHAPE) in Cook County, IL, which compared the existing supply of physical activity programs to potential demand among older adults. The survey found that approximately 4% of the potential demand, defined as persons older than 65 in Cook County, based on the 2000 Census, could be met by the current supply (Hughes & Molina, 2002). The purposes of the current survey were to (a) obtain new information regarding the types of facilities that provide programs for older adults and the types of activities provided; (b) undertake an initial effort to estimate the supply of these programs in a variety of geographic areas across the nation; and (c) compare capacity data for the specific geographic areas studied to estimate the potential demand for programming by using U.S. census data. To our knowledge, this is the first study of the extent of physical activity programming for older adults at multiple sites across the United States. Methods HAN Sites The seven selected participating HAN academic centers target areas for study that are diverse with respect to geographic and demographic characteristics. Table 1 describes the population of older adults in each target area, including the size of each target area, its population density, the percentage of ethnic minorities, and the percentage of the population living below poverty level. The target-area size ranged from 26 square miles (67.34 km2) in southeast Seattle, WA, to 8,194 square miles (21,222.46 km2) in San Luis Valley, CO. The population of persons aged 65 or older ranged from 5,921 in San Luis Valley to 630,265 in Cook County, IL. In addition, the population density of persons aged 65 or older varied across target areas, from fewer than 1 older adult per square mile in San Luis Valley to almost 700 older adults per square mile in Cook County. Differences also existed with respect to the percentage of persons aged 65 or older who are minorities or non-Hispanic Whites, ranging from 9.1% in Allegheny County, PA, to 57.6% in southeast Seattle. The percentage of older adults living below poverty level ranged across target areas from 8.1% in Alameda County to 14.6% in the San Luis Valley, with an average of 10.7%. Survey Instrument The HAN survey evolved from a previous survey by the SHAPE, administered in 2001 to more than 1,000 facilities in Cook County, IL (Hughes & Molina, 2002). Revisions to the SHAPE survey included more detailed questions on program capacity, accessibility, features, participation barriers, and reasons why facilities did not provide programs for older adults. The survey asked about programming designed for older adults (those over the age of 65) and also sought information on programs older adults used that were not necessarily designed for them. The eight-page survey included an activity grid (e.g., type of activity offered, frequency), yes–no questions, open-ended questions, and checklists. Copies of the instrument are available from the corresponding author. In addition to collecting data on existing physical activity programs for older adults, the survey collected information about reasons that some organizations do not provide programs. The first survey questions asked respondents if they provided programming specifically for older adults or if older adults participated in any of their programs (e.g., 1. Do you provide physical activity programs designed specifically for older adults? 2. Do you have physical activity programs that younger as well as older adults attend?). If respondents answered “no” to both questions, they were asked why they did not provide programming for seniors. If respondents answered “yes” to either question, the instructions asked respondents to complete the entire questionnaire. Several survey items addressed program capacity. They included questions on the maximum capacity of older adults (aged 65+) per week, the actual number of older-adult participants per week, whether there were waiting lists for activities, and the estimated unduplicated number of older-adult participants in the past year. Organizations that provide physical activity programs for older adults pilot tested the survey twice. The National Council on the Aging conducted the first pilot test in 14 senior centers across the country. HAN members at the University of North Carolina conducted the second pilot test at 10 community sites. Survey items were revised and clarified based on the pilot findings. Sampling Frame All seven participating HAN academic centers agreed to participate in the study, but each site chose its own geographic target area to study. The Seattle site, for example, wanted to do an in-depth assessment of programming availability within a precisely defined, underserved minority community. In contrast, the Chicago site had previously surveyed Cook County, had already developed an initial sampling frame, and wanted to update its findings on the same population of respondents. As previously noted, the size of the geographic target area included in the surveys varied across sites from 26 square miles (67.34 km2) in southwest Seattle to 8,194 square miles (21,222.46 km2) in San Luis Valley. Despite this variation in size, each participating site attempted to assemble as exhaustive a list of potential physical activity providers as possible within its geographic area, given available resources. To err on the side of inclusiveness, each site initially included a broad spectrum of community organizations thought to provide physical activity programs for older adults. Thus, the initial sampling frames included a wide array of potential providers, including senior centers, community centers, YMCAs or YWCAs, commercial gyms, county and city parks and recreation facilities, churches, schools, hospitals, private or public housing for seniors, and residential facilities for able-bodied older adults (e.g., independent-living facilities). These sampling frames included most of the organizations that might have had contracts with Area Agencies on Aging and all known senior centers. Information gathered during the survey process led to refinement of the sampling frames and elimination of initially suggested facilities or programs that no longer existed, had moved from the area, or did not provide physical activity programming for persons of any age. Survey Administration Each HAN site acquired approval for the study protocol from its human subject research institutional review board. All HAN sites began survey administration at the same time and followed similar protocols. Five sites conducted surveys initially by mail and followed up by telephone or in person. The remaining two sites administered the survey primarily by telephone. One of the seven sites also made the survey accessible on the Internet. In most cases, sites used a combination of methods to maximize response. Initial contact by an introductory letter or phone call gave recipients information about the HAN, the survey's history and purpose, benefits of the survey to the organization (specifically, future publication of directories of available programs and facilities), HAN site-specific contact information, and expected survey timing. This introduction also requested confirmation of the appropriate contact person in the facility or organization to receive the survey. Most HAN sites waited 2 weeks after the introduction letter or call to mail the survey. For the HAN site that conducted the survey by telephone, the initial call also served as the first attempt to complete the survey. If the first mailing or phone calls did not generate responses, all sites made follow-up calls or on-site visits to nonrespondents. The follow-up process spanned several months. During follow-up, sites offered organizations the choice of responding to the survey by telephone or in person, and having the survey re-sent or faxed. Some sites with a comparatively small number of providers in their target areas were able to achieve very high response rates by calling repeatedly until they achieved a response. Others with larger numbers of providers in their target areas had lower response rates, but still used multiple call backs to all potential respondents. The use of a combination of administration methods yielded higher response rates than any single method and offered the opportunity to clarify respondents' questions about the survey. Each HAN site recorded survey activities on an Excel tracking sheet that was separate from data received from programs and facilities. This tracking sheet listed the organization name or identification number, contact information, date of initial mailing, follow-up activity, response dates, and any specific data-collection issues that arose. The University of Washington HAN site's Health Promotion Research Center (UWHPRC) served as the central data-collection site. UWHPRC created a universal Microsoft Access database and provided each HAN site with detailed instructions for data entry. Submitted data excluded all personal identifiers (institutional or individual). UWHPRC conducted a database reliability check 1 month after each HAN site began using the database. HAN sites sent all survey data to UWHPRC for analysis when data collection and entry were completed. Table 2 provides sampling frame sizes for each HAN site (adjusted for initially incorrect or incomplete information as described earlier), response rates, and the number of facilities in each sample that offered programs. Respondents included organizations that offered physical activity programs for older adults and completed the entire survey, organizations that did not offer physical activity programs for older adults and completed the abbreviated version of the survey, and organizations that responded orally that they did not offer physical activity programs for older adults and did not complete any version of the survey. Results Although the number of facilities surveyed by each site varied more than 20-fold, ranging from 29 organizations in Colorado to 737 in Chicago, response rates were good for all seven sites (see Table 2). Response rates ranged from 67% in Alameda County, CA, to 100% in San Luis Valley, CO, with an average response rate of 77% across sites. Among the 1,168 responding facilities, 675 (58%) reported that they offered programs for older adults. Of 326 facilities that stated they did not offer programs and provided a reason for lack of programs, 161 (50%) identified the most common reason as a perceived lack of interest from older adults, followed by lack of funding (46%), lack of staff interest (44%), lack of staff knowledge regarding frail adults (34%), staff shortage (34%), lack of staff training regarding older adults (24%), and concerns about liability (23%). Respondents could give multiple reasons for not offering programs. Of 675 facilities that provided programming designed for or used by older adults, 652 (97%) provided information regarding the specific types of programs offered (Table 3). Overall, aerobic programs were offered most frequently (73%), in contrast to flexibility (47%) and strength training (26%), and 31% of facilities offered multicomponent programs. Among facility types, senior centers most frequently offered aerobic programs, and hospitals and clinics most frequently offered strength-training and flexibility programs. Most facilities surveyed provided more than one program. Specifically, the 652 facilities surveyed provided 2,546 programs to 69,634 older adults weekly (Table 4). This amounts to an average of four programs per facility. The most commonly offered and best-attended programs were aerobics (47% of programs, 53% of attendance), followed by flexibility (24% of programs, 19% of attendance). Although strength training was offered by 26% of facilities, it represented only 10% of total programs and 11% of attendance. The most popular aerobic programs were aerobic exercise (unspecified), stationary equipment, chair-based activities, walking, and dance. Four types of facilities accounted for 90% of programs offered (Table 5). Although we originally tracked commercial gyms and YMCAs separately, we combined them into one type of respondent in Table 5 because calls to both types of facilities showed that their fee structures were similar in terms of both initial membership and monthly fees. An analysis of program offerings by facility type showed that commercial gyms and YMCAs offered 27% (685 programs), senior centers offered 24% (604), park and recreation centers or community centers offered 23% (582), and senior housing facilities offered 16% (415). These same facility types have the largest number of older adults participating in programs per week, constituting 88% of attendance. Across facility types, the percentage of programs offered closely reflects the percentage of older-adult participants per week, with the exception of park and recreation centers or community centers, which have a greater percent of programs offered than attended. It is important to note, however, that facility types with the greatest number of programs might also have the largest capacity to serve older adults. Several items on the survey addressed issues of access. First, with respect to populations served, facilities reported that they served the following specific, nonexclusive subpopulations of older adults: sedentary (52% of all older adults served), low income (47%), frail (43%), and non-English speaking (25%). Second, regarding physical access to programs, of 675 respondents, 88% reported having parking available on site, 66% were within one fourth of a mile (0.40 km) of public transportation, and 33% had senior transportation or shuttles that conveyed participants to the facility. The survey also inquired about program fees with respect to financial access, but it found too much variation across respondents in terms of monthly memberships versus daily fees versus class fees to be able to report meaningful information in a consistent way. We also report findings regarding two measures of demand. The first measure estimates demand as a function of the existence of waiting lists to gauge demand for physical activity programming in each geographic area (Table 6). Only 4% of programs reported that they had waiting lists—a consistent finding across all sites. The second measure compares the U.S. census population at each site with reported participation in programs. Of 1,123,401 total older adults that the 2000 Census enumerated across the seven sites combined, the facilities identified 69,634 individuals as current weekly program participants. This number of participants equals 6% of the total older-adult population across the sites. The participation percentage varied from 3% in Pittsburgh to 28% in Colorado and was generally higher in areas with the fewest numbers of programs. It is important to note that these percentages, although low, may overestimate participation because they likely include individuals who participated in more than one activity. Facilities also estimated their maximum capacity to serve older adults seeking programs. The total estimated maximum capacity across all sites, 207,328, would meet the needs of only 18% of older adults residing in the survey sites. Finally, responses to several survey items indicate the presence of program-management issues. First, nonresponses and follow-up calls to a number of facilities regarding the aforementioned capacity items revealed that several respondents had difficulty documenting the number of persons served per year or the number of persons attending programs during a given year and had a particularly difficult time estimating the number of persons who could be served by equipment as opposed to classes. Other responses concerning program-management issues show that, for the 675 facilities offering programs, 49% conducted program evaluations, 74% tracked attendance, 47% tracked participant progress, and 56% trained instructors. Senior centers and hospitals, more than other facilities, performed program evaluations (62% and 70%, respectively) and tracked attendance (83% and 89%, respectively), whereas churches were least likely to conduct program evaluations (38%) or track attendance (64%). Hospitals also were most likely to train instructors (81%). Forty-one percent of organizations indicated an interest in obtaining assistance with programming for older adults. Discussion Evidence regarding the benefits of physical activity to older adults is strong and compelling, especially when coupled with the prevention imperative posed by the rapidly growing aging population and costs known to be associated with inactivity. Equipped with this knowledge, the public health, personal health, and aging communities increasingly call for older adults to take to the trails, dance floor, or pool. Although these calls may actively stimulate demand, we know remarkably little about the available supply of organizations that provide physical activity programming for older adults. Is supply adequate to meet current demand? Is there room for growth? How can we build increased capacity among active organizations or potential program providers? These questions become increasingly urgent as the older population grows, and we continue to encourage physical activity as sound prevention. We know that many older adults seek physical activity opportunities independently, but others need structured programs (King et al., 2000). Structured programs may be particularly helpful for sedentary older adults who need instruction and support in getting started and integrating behavioral change into their lifestyles (King, Haskell, Taylor, Kraemer, & DeBusk, 1991). The HAN physical activity program capacity survey provides new information regarding the types of organizations that provide programs to older adults and the types of activities provided. The data also yield an estimate of the supply of these programs in a variety of settings across the nation, and they provide crude estimates of the existing supply compared with the potential demand for programming based on U.S. census data for the specific geographic areas studied. With respect to those facilities that do not provide programs, the most commonly reported reason cited was a perceived lack of interest from older adults (50%), followed by lack of funding, lack of staff interest, and lack of staff knowledge regarding frail older adults. This finding indicates that substantial effort has to be invested in educating older adults about the benefits of physical activity, in increasing funding for programs, and conducting staff training regarding the exercise needs and capacities of older adults. The overall low percentage of total population served (6% across study sites) is another noteworthy finding of the survey. The fact that study sites with the highest response rates reported greater attendance on average tempers our confidence in this finding. Improved response rates could result in an upward adjustment of attendance. In contrast, the data may overestimate participation because some individuals participate in more than one physical activity program in a given week. For example, they may attend an aerobics class and also participate in a senior golf league. Despite these cautionary notes, this indicator of program participation causes concern regarding the engagement of older adults in physical activity. Even if we take into account individuals in institutions and the portion of the older adult population that prefers individual physical activity to structured programs, a significant gap remains to be filled. For example, consider Cook County, which, according to data from the 2000 Census, has 630,265 persons aged 65 and older (U.S. Census Bureau, 2000). CDC (2003) data indicate that nationally about 4% of the elderly population resides permanently in nursing homes, reducing the size of the relevant Cook County population to roughly 600,000. If we further estimate, using national averages on participation in leisure-time physical activity, that 33% of the population is currently sedentary, that reduces the number of persons in Cook County needing programs to 198,000 (CDC, 2001). If we further assume, on the basis of prior reports in the literature, that 28% of the remaining population prefer group as opposed to individual or home-based exercise, then we estimate that 55,440 persons not currently involved in facility-based programming in Cook County might participate in and benefit from programming if it were available (Mills, Stewart, Sepsis, & King, 1997). Because the findings presented in this article indicate that 31,171 older adults already participate in programs, program capacity in Cook County would have to increase by 78% to meet the needs of this group, assuming that we can motivate them to become involved. Findings regarding attendance per facility type are also of interest. A majority of participants (57%) attend programs that are not geared specifically to older adults, and 43% attend programs tailored for older adults. Across all programs offered, 55% are not geared to older adults, and 45% of programs target only older adults. These data underscore the important role of general physical activity programs in addressing the needs of both older and younger adults. Not all older adults need specialized or senior-only programs. These data demonstrate the key contribution that private-sector organizations make to physical activity programming. Accordingly, as we move to build capacity, we must look both to the private sector and to traditional senior-service organizations for growth in size and program scope. Additional need for programming documented by survey results is also of interest. First, findings indicate that only 31% of 163 facilities that serve non-English-speaking populations tailor programs for non-English-speaking participants. This finding demonstrates a need for additional focus on non-English-speaking populations. Second, given that attendance largely parallels program offerings and that 73% of facilities offer some form of aerobic activity, it is not surprising that aerobic programs are best attended. However, this finding also raises the question of whether facilities provide what consumers request or whether consumers use what is available. Another consideration relates to consumer and organization perceptions about what constitutes appropriate and safe activity for older adults. For example, flexibility programs are widely believed to be safe and low risk, whereas strength training is less familiar to older adults, and both older adults and organizations that serve them may perceive it as a more risky undertaking. Given the strong evidence concerning the benefits of strength-training activity for older adults (CDC, 2004), these data point clearly to the need to develop more strength-training programs and to engage a higher percentage of older adults in these programs. A need to create demand for such programs through public education also may be indicated. It is likely that strength training requires technical assistance for providers to support sound program development. This study has limitations that merit discussion. The communities surveyed have diverse ethnic and geographic characteristics but constitute a convenience sample and do not represent the nation as a whole. Although the actual data collected were consistent across sites, administration methods varied (e.g., telephone, mail, and in-person interviews). For example, South Carolina used a telephone survey as the primary administration method because the site believed it would be better received, particularly at churches in its sampling frame. This variability can affect response rates. Moreover, some sampling-frame differences existed across sites. For example, in some areas, churches generally did not offer programs; in other communities, faith-based organizations represented a key provider type. In those cases, sites chose to include or exclude specific types of organizations. Other types of organizations like Federally Qualified Health Centers also were not included. In some instances, responses indicated possible problems among responding organizations relative to the estimation of capacity. For example, some organizations could not provide data pertaining to the number of persons served over the course of a year or the number of persons in attendance during a week. Organizations had a particularly difficult time estimating the number of persons who could be served by using equipment as opposed to attending classes. Similarly, organizations serving both older and younger adults had to estimate participants' ages (older or younger than age 65), and they found estimating use by older adults to be challenging. Other program-management issues noted by survey respondents included limited attention to evaluating programs, tracking participant progress, and training instructors. These management issues also may be appropriate foci for technical assistance, especially because 41% of the surveyed organizations indicated an interest in obtaining assistance with their programming for older adults. Despite these limitations, findings from this survey suggest that the current supply of physical activity programs designed for older adults in the target communities does not adequately meet potential demand for programs by older adults. More energy should be focused on increasing demand by raising awareness of the importance of physical activity among older adults and reducing barriers to exercise. The barriers perceived by providers indicate a substantial need for health-promotion campaigns for users and providers, as well as increased funding for programs. Future research should examine ways to refine the survey sampling methodology, streamline the survey instrument, and replicate this survey in other communities. These efforts are vital if we are to obtain valid data on the existing supply of physical activity programs and be empowered to make valid assessments of the fit between supply and demand in the future. This research was supported by grants from the Centers for Disease Control and Prevention (CDC) Prevention Research Centers Program (Grant U48/CCU0009654) and the Health Care and Aging Studies Branch, and from the National Council on the Aging and the Robert Wood Johnson Foundation. We acknowledge the following people for their contributions to survey development, data collection, or manuscript review: Melissa Kealey, University of California, Berkeley; Jennifer I. McLean, University of Colorado; Thomas Prohaska, Ella Fermin, Megan Renehan, University of Illinois at Chicago; members of the Cook County Senior Health Alliance Promoting Exercise; Christen Sible, Cindy Schrauder, Michael Randall, Carol Giuliani, Franzi Zabolitizki, Victor Marshall, Mary Altpeter, and Tiffany Small, University of North Carolina; Jane Schall, University of Pittsburgh; Harriet Williams, Sara Wilcox, Bridget Kane, Larissa Oberrecht, Jill Maxwell, and Joey Vrazel, University of South Carolina; Gwen Moni and James LoGerfo, University of Washington; and Nancy Whitelaw at the National Council on the Aging. 1 Center for Research on Health and Aging, University of Illinois at Chicago. 2 Health Promotion Research Center, University of Washington, Seattle. 3 Center for Healthy Aging, University of Pittsburgh, PA. 4 Division of Health Care Policy and Research, University of Colorado Health Sciences Center, Denver, CO. 5 Chapel Hill School of Medicine Program on Aging, University of North Carolina. 6 Berkeley School of Public Health, University of California. 7 Department of Health Promotion, Education, and Behavior, University of South Carolina, Columbia. Decision Editor: Linda S. Noelker, PhD Table 1. Description of Target Areas. City or County State Area (square miles) 65+b Densityc Minority (%)d Below Poverty (%) Alameda CA 737 147,591 200.2 43.1 8.1 San Luis Valleya CO 8,194 5,921 0.7 39.7 14.6 Cook IL 945 630,265 666.9 30.3 10.3 Durham, Henderson, and Northampton NC 1,200 44,755 37.3 20.9 11.4 Allegheny PA 730 228,416 312.9 9.1 9.0 Richland and Lexington SC 1,457 53,459 36.7 22.4 10.9 Southeast Seattlea WA 26 12,994 500.0 57.6 10.8 aSan Luis Valley is Alamosa, Conejos, Costilla, Mineral, Rio Grande, and Saguache counties; Southeast Seattle is King County census tracts 93–95, 99–104, 107–114, and 117–119. bPopulation aged 65 and older (U.S. Census 2000 table P12I; available at http://www.census.gov). cDensity = population aged 65 and older per square mile. dPopulation aged 65 and older who reported other than non-Hispanic White race or ethnicity (http://www.census.gov). Table 2. Response Rates by Site. City or County State No. of Facilities in Initial Sampling Framea No. of Eligible Facilitiesb No. of Facilities That Responded (%) No. of Facilities That Offer Programsc Alameda CA 289 251 168 (67) 88 (52) San Luis Valley CO 36 29 29 (100) 29 (100) Cook IL 804 737 529 (72) 273 (52) Durham, Henderson, and Northampton NC 617 179 143 (80) 58 (41) Allegheny PA 70 70 60 (86) 57 (95) Richland and Lexington SC 237 197 191 (97) 138 (72) Southeast Seattle WA 57 49 48 (98) 32 (67) aNumber of facilities believed to have the potential to offer physical activity programs in the target area and were mailed a questionnaire. bNumber of eligible facilities included facilities that did not offer physical activity programs but completed the survey. Nonrespondents, except churches, were assumed to have physical activity programs and were included. Nonrespondent churches were assumed to not have physical activity programs and were excluded (see text). Facilities that indicated they did not offer physical activity programs and did not fill out the survey were removed. cNumber of facilities that responded positively to at least one of the two screening questions. Table 3. Percent of Facilities Offering Various Types of Physical Activity Programs, for Facilities With Any Programs. Type of Facility Total No. of Facilities Types of Programs (% of Facilities) Aerobic Strength Training Flexibility Multicomponent Recreational Other Commercial gym or YMCA 184 70 22 45 35 25 15 Senior center 150 81 21 55 29 15 13 Park or recreation and community center 146 73 32 40 18 32 10 Housing 103 71 30 53 41 11 6 Church 39 51 13 28 38 18 8 Hospital or clinic 26 77 46 62 35 15 23 School 4 100 50 75 0 50 0 Total 652 73 26 47 31 22 12 Table 4. Program Attendance by Physical Activity Program Subtype. Program Type Programs Offered Attendance per Week N % of Total N a % of Total Aerobic 1,184 47 37,041 53 Aerobic exercise 290 25 8,952 24 Stationary equipment 168 14 8,057 22 Walk 178 15 4,660 13 Chair based 181 15 4,591 12 Dance 154 13 4,436 12 Water aerobics 106 9 3,253 9 Swimming 58 5 1,988 5 Other 49 4 1,104 3 Flexibility 607 24 13,281 19 Strength training 266 10 7,356 11 Multicomponentb 205 8 4,949 7 Recreational 188 7 4,649 7 Otherc 96 4 2,358 3 Total 2,546 69,634 100 aAttendance is the sum of actual number of older adults reported as participants in all programs or activities (except those that were educational only). If an actual number is missing, then attendance is computed as the median value for that activity or as the mean proportion of maximum capacity (whichever is less). Because adults may participate in more than one activity per week, the sum may not be an unduplicated number. bMulticomponent program example: aerobic and free weights. Multicomponent programs are counted as one program, assuming that same older adults attend both components. If two different attendance numbers are listed, then take maximum of the two (and assume that some older adults left before other component). cOther does not include programs that were only educational. Table 5. Program Attendance by Facility Type. Facility Type Programs Offered Attendance per Weeka N % N % Commercial gym or YMCA 685 27 20,917 30 Senior centerb 604 24 16,830 24 Park or recreation and community center 582 23 13,379 19 Housing 415 16 10,813 15 Hospital or clinic 146 6 5,970 9 Church 78 3 1,165 2 School 36 1 560 1 Total 2,546 100 69,634 100 aSum of actual number of older adults for all programs or activities (Q7–Q34, but not Q32). If missing actual number, we used the median value for that activity or the mean proportion of maximum capacity (whichever is less). Because adults may participate in more than one activity per week, sum is an not unduplicated number. bSenior center includes any center or day program geared to older adults. Table 6. Waiting Lists and Weekly Attendance by Site as Percentage of Total Population Aged 65 and Older. City or County State Total No. of Programsa Programs With Waiting List Attendance per Week N % of Programs N b % of Total Population Aged 65+c Alameda CA 305 12 4 11,338 8 San Luis Valley CO 78 0 0 1,643 28 Cook IL 1079 43 4 31,171 5 Durham, Henderson, and Northampton NC 236 12 5 9,710 22 Allegheny PA 286 10 3 7,601 3 Richland and Lexington SC 305 4 1 5,384 10 Southeast Seattle WA 97 3 3 2,787 21 Total 2,386 84 4 69,634 6 aTotal number of programs is not the same as those in Table 4 because some programs in Table 5 are missing waiting list information and are not included. bSum of actual number of older adults (65+) per week for all activities reported by responder, aerobic exercise to other, not including educational materials (Q32). If missing actual number, we used the median value for that activity or the mean proportion of maximum capacity (whichever is less). The number of older adults per week is not exactly the same as in Table 4 because attendance per week is an estimate (when missing actual number) and rounding errors occur when calculating by site (Table 5) or by program (Table 4). cPercent of older adults per week is the actual number of older adults per week divided by total population 65+ from Table 1. References Binder, E. F., Birge, S. J., & Kohrt, W. M., (1996). Effects of endurance exercise and hormone replacement therapy on serum lipids in older women. Journal of the American Geriatrics Society, 44, 231-236. Blumenthal, J. A., Emery, C. F., Madden, D. J., Coleman, R. E., Riddle, M. W., & Schniebolk, S., et al (1991). Effects of exercise training on cardiorespiratory function in men and women older than 60 years of age. American Journal of Cardiology, 67, 633-639. Centers for Disease Control and Prevention. (1996). Physical activity and health: A report of the surgeon general. Atlanta, GA: CDC National Center for Chronic Disease Prevention and Health Promotion. Centers for Disease Control and Prevention. (2001). Prevalence of disabilities and associated health conditions among adults—United States, 1999. Morbidity and Mortality Weekly Report, 50, 120-125. Centers for Disease Control and Prevention. (2001). U.S. physical activity statistics 2001 state demographic data comparison. Atlanta, GA: Author. Centers for Disease Control and Prevention. (2003). Health United States, 1999, with health and aging chartbook. Atlanta, GA: Author. Centers for Disease Control and Prevention. (2004). Strength training among adults aged ≥65 years—United States, 2001. Morbidity and Mortality Weekly Report, 53, 25-26. Doll, L., & Kreuter, M., (2001). The Prevention Research Centers program: Linking science and practice through community collaborations. Journal of Public Health Management and Practice, 7, x-xii. Federal Interagency Forum on Aging-Related Statistics. (2000). Older Americans 2000: Key indicators of well-being. Washington, DC: U.S. Government Printing Office. Fiatarone, M., O'Neill, E. F., Ryan, N. D., Clements, K. M., Solares, G. R., & Nelson, M. E., et al (1994). Exercise training and nutritional supplementation for the physical frailty in very elderly people. New England Journal of Medicine, 330, 1769-1775. Fried, L. P., Kronmal, R. A., Newman, A. B., Bild, D. E., Mittelmark, M. B., & Polak, J. F., et al (1998). Risk factors for 5-year mortality in older adults. Journal of the American Medical Association, 279, 585-592. Huang, Y., Macera, C. A., Blair, S. N., Brill, P. A., Kohl, H. W., III, & Kronenfeld, J. J., (1998). Physical fitness, physical activity, and functional limitation in adults aged 40 and older. Medicine and Science in Sports and Exercise, 30, 1430-1435. Hubert, H. B., Bloch, D. A., & Fries, J. F., (1993). Risk factors for physical disability in an aging cohort: The NHANES I epidemiologic follow-up study. Journal of Rheumatology, 20, 480-488. Hughes, S. L., & Molina, L. C., (2002). Overcoming barriers to physical activity: Public health studies on the adequacy of community-based programs and resources. Symposium conducted at the 55th annual scientific meeting of the Gerontological Society of America, Boston, MA. King A. C., Castro, C., Wilcox, S., Eyler, A. A., Sallis, J. F., & Brownson, R. C., (2000). Personal and environmental factors associated with physical inactivity among different racial–ethnic groups of U.S. middle-aged and older-aged women. Health Psychology, 19, 354-364. King A. C., Haskell, W. L., Taylor, C. B., Kraemer, H. C., & DeBusk, R. F., (1991). Group- vs home-based exercise training in healthy older men and women. A community-based clinical trial. Journal of the American Medical Association, 266, 1535-1542. LaCroix, A. Z., Guralnik, J. M., Berkman, L. F., Wallace, R. B., & Satterfield, S., (1993). Maintaining mobility in later life II. Smoking, alcohol consumption, physical activity, and body mass index. American Journal of Epidemiology, 137, 858-869. Mills, K. M., Stewart, A. L., Sepsis, P. G., & King, A. C., (1997). Consideration of older adults' preferences for format of physical activity. Journal of Aging and Physical Activity, 5, 50-58. Preisinger, E., Alacamlioglu, Y., Pils, K., Bosina, E., Metka, M., & Schneider, B., et al (1996). Exercise therapy for osteoporosis: Results of a randomised controlled trial. British Journal of Sports Medicine, 30, 209-212. Rejeski, W., Brawley, L., McAuley, E., & Rapp, S., (2000). An examination of theory and behavior change in randomized clinical trials. Controlled Clinical Trials, 21, 164S-170S. Robert Wood Johnson Foundation. (2001). National blueprint: Increasing physical activity among adults aged 50 or older. Princeton, NJ: Author. Seeman, T., Berkman, L. F., Charpentier, P. A., Blazer, D. G., Albert, M. S., & Tinetti, M. E., (1995). Behavioral and psychosocial predictors of physical performance. MacArthur studies of successful aging. Journal of Gerontology: Medical Sciences, 50A, M177-M183. Singh, M. A., (2002). Exercise comes of age: Rationale and recommendations for a geriatric exercise prescription. Journal of Gerontology: Medical Sciences, 57A, M262-M282. Singh, M. A., Ding, W., Manfredi, T. J., Solares, G. S., O'Neill, E. F., & Clements, K. M., et al (1999). Insulin-like growth factor I in skeletal muscle after weight-lifting exercise in frail elders. American Journal of Physiology, 277, E135-E143. Stahle, A., Nordlander, R., & Bergfeldt, L., (1999). Aerobic group training improves exercise capacity and heart rate variability in elderly patients with a recent coronary event. A randomized controlled study. European Heart Journal, 20, 1638-1646. Strawbridge, W. J., Cohen, R. D., Shema, S. J., & Kaplan, G. A., (1996). Successful aging: Predictors and associated activities. American Journal of Epidemiology, 144, 135-141. U.S. Census Bureau. (2000). Summary File 1 and Summary File 3—Census 2000 demographic profile highlights for Cook County, IL. Washington, DC: U.S. Government Printing Office. U.S. Department of Health and Human Services. (2000). Healthy people 2010: Understanding and improving health and objectives for improving health (2nd ed.). Washington, DC: U.S. Government Printing Office.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B7707F0F4E4D9DD633419401936BBA794C59402.txt b/test/dataset/in/resources/corpus/Clean_0B7707F0F4E4D9DD633419401936BBA794C59402.txt
deleted file mode 100644
index af8c840..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B7707F0F4E4D9DD633419401936BBA794C59402.txt
+++ /dev/null
@@ -1 +0,0 @@
-JAGP61041S1064-7481(12)61041-010.1097/00019442-199800640-00003American Association for Geriatric PsychiatryTABLE 1Prevalence of psychiatric symptoms in patients with and without hoarding behavior, n (%)CharacteristicsHoarding(n = 30)Nonhoarding (n = 103)χ2[1]PDelusions21 (70.0)67 (65.0)0.2540.614 Theft16 (53.3)56 (54.4)0.0100.920 Persecutory5 (16.7)30 (291)1.8600.173 Jealousy3 (10.0)18 (17.5)0.9770.405Hallucinations9 (30.0)37 (35.9)0.3600.548 Auditory5 (16.7)19 (18.4)0.0500.823 Visual7 (23.3)27 (26.2)0.1010.750Misidentifications14 (46.7)30 (291)3.2290.072 Phantom boarders symptoma8 (26.7)15 (14.6)2.3800.123 Residence is not home7 (23.3)15 (14.6)1.2940.271 Mirror signb3 (10.0)5 (4.9)1.0800.379 Caregiver is impostor2 (6.7)11 (10.7)0.4240.732Other disturbances Repetitive behaviors24 (80.0)55 (53.4)6.8170.009 Day/night disturbance8 (26.7)28 (27.2)0.0030.955 Aggressiveness18 (60.0)48 (46.6)1.6680.196 Agitation18 (60.0)49 (47.6)1.4350.231 Wandering15 (50.0)44 (42.7)0.4990.480 Hyperphagia18 (60.0)24 (23.3)14.482<0.001 Pilfering15 (50.0)26 (25.2)6.6780.010athe belief that other (imaginary) people are in the house.bthe inability to recognize one's own mirror reflection (or statements such as “Someone else is in the mirror.”)Received March 24, 1997; revised August 10, 1997, December 24, 1997; accepted January 15, 1998.The authors are indebted to the geropsychiatric staff at Veterans General Hospital–Taipei for their assistance in this study.This study was supported by Veterans General Hospital–Taipei, Grant VGH-85-150.Regular ArticlesHoarding Behavior in Dementia: A Preliminary ReportJen-PingHwangM.D.Shih-JenTsaiM.D.*Chen-HongYangM.D.King-MingLiuM.D.Jiing-FengLirngM.D.Department of Psychiatry, Veterans General Hospital-Taipei, Republic of China*Address correspondence to Dr. Tsai, Department of Psychiatry, Veterans General Hospital-Taipei, No. 201, Shih-Pai Road Sec. 2, Taipei, Taiwan, Republic of ChinaHoarding behavior has been reported in several mental disorders and is occasionally reported by the caregivers of dementia patients. Such behavior may have adverse effects on the patients and increase the burden of the caregivers. This study was conducted to investigate the prevalence of hoarding behavior in patients with dementia and identify the characteristics and psychiatric symptoms associated with it. The sample was 133 dementia patients admitted to a geropsychiatric ward. Of the 133 dementia patients, 30 (22.6%) showed hoarding. Hoarding was found in various types of dementia. Patients with hoarding had a higher prevalence of repetitive behaviors, hyperphagia, and pilfering. Results suggested that hoarding behavior is a common symptom in dementia patients and a complex phenomenon. Better understanding of the underlying pathogenesis may highlight specific pharmacological or behavioral methods for treatment of the behavior.Hoarding—collecting a large number of unneeded objects—is commonly found in the general population1 and in a variety of mental disorders, including schizophrenia,2–4 dementia,2 obsessive-compulsive disorder,2 and eating disorders.5 Greenberg et al.2 proposed that when hoarding is the main symptom, it is the final common pathway for a spectrum of different processes. At one end of the spectrum is obsessive-compulsive disorder; at the center, paranoid disorders, and at the other end, organic mental disorders.Hoarding behavior has long been recognized as a feature of dementia2 and is frequently noted by the caregivers of dementia patients. It can interfere with the hygienic management and health of patients, and patients may become extremely agitated and even violent when family members threaten to discard their possessions.2,6 Also, many nurses view hoarding as negative and assume that the patient is a bit “strange” or psychotic.7 A search of the MEDLINE database revealed only one case report concerning hoarding in dementia patients.2 Our study was designed to investigate the prevalence of hoarding behavior in dementia patients admitted to the geropsychiatric ward of a general hospital. We also made an attempt to identify the characteristics and psychiatric symptoms that are associated with hoarding behavior.METHODSThe sample population consisted of 133 consecutive dementia patients admitted to the geropsychiatric ward of Veterans General Hospital–Taipei between August 1989 and February 1996. Most of the patients were admitted because of violence, inappropriate behaviors, or emotional problems. All patients met the criteria for dementia set by the Diagnostic and Statistical Manual of Mental Disorders (DSM-III-R).8 Those with Dementia of the Alzheimer's type (DAT) met DSM-III-R criteria for primary degenerative dementia as well as the diagnostic criteria developed by the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association (NINCDS-ADRDA).9 All patients with multi-infarct dementia met the DSM-III-R criteria for multi-infarct dementia. The other nonspecific dementias were defined on the basis of their etiologies, according to the DSM-III-R criteria. Diagnosis was based on a complete medical and neuropsychiatric examination, including history, physical examination, ECG, EEG, blood count, biochemistry (electrolytes, liver and renal function, vitamin B12, folic acid), thyroid function tests, and a serological test for syphilis. Computed tomographic scans of the brain were available for all patients.Patients were divided into two groups, based on the presence or absence of hoarding. Hoarding was defined as repeatedly collecting mostly useless or unneeded objects during some time from the onset of the illness. Such behaviors were indiscriminate and developed after the onset of dementia.All demographic data were supplied by caregivers. Psychotic symptoms and behavioral problems were assessed by a checklist adapted from the Behavioral Pathology in Alzheimer's Disease Rating Scale (BEHAVE-AD) by Reisberg et al.,10 with some modifications. In each case, the patient, as well as at least one caregiver familiar with the behaviors of the patient, was interviewed by a geriatric psychiatrist. Symptoms were evaluated and classified as dichotomous variables. Dementia severity was assessed on admission by use of the Folstein Mini-Mental State Exam (MMSE); scores ranged from 0 to 30.11Data were analyzed by use of SPSS for Windows. Two-tailed t tests were used to compare the differences between the means of continuous variables. The chi-square test was used to compare differences in dichotomous variables. A P value less than 0.05 was considered to be statistically significant.RESULTSThe mean age of these patients was 74.1±6.3 years (range: 65–91), and the mean MMSE score was 11.6±6.2 (median: 11). There were 98 men and 35 women; the large male predominance is characteristic of a veterans' hospital. Of the 133 dementia patients, 30 (22.6%) showed hoarding. Patients with hoarding were found in various types of dementia classifications: Alzheimer's dementia (24 of 75), multi-infarct dementia (4 of 40), and dementia not otherwise specified (NOS; 2 of 18). The items patients hoarded included daily necessities (10), food (7), garbage (6), newspapers or magazines (6), broken umbrellas or electrical items (6), plastic bags (4), old clothes (4), and cigarette butts (1). Patients in the hoarding group stored their objects around their apartment, in closets, in drawers, under pillows or bedclothes, under trash cans, under the bed, or in boxes, or carried the items with them. Reasons given by patients for hoarding included “It will come in handy or can be sold.”, “This object belongs to me.”, “It will be stolen.”, “It is too good to throw away.”, “I don't want to be caught without a needed item.”, “Even though the food is outdated, it is still edible.”, and “I usually feel hungry.”There were no significant differences between the two groups in regard to age, age at onset, gender, educational level, or MMSE score. Table 1 summarizes the frequency of psychiatric symptoms in the two groups. The prevalence of repetitive behaviors was significantly higher in the hoarding group. Also, patients with hoarding were more likely to have hyperphagia and to pilfer other people's possessions.DISCUSSIONThe prevalence of hoarding (22.6%) in dementia in this study was high. In a study of behavioral complications of dementia, the prevalence of hoarding was 1.9%.12 The discrepancy in these two studies may stem from the differences in the definition of hoarding and the sample source. In the previous study, the patients were in the community, and their disease severity varied widely. Behavior was assessed for only one week. The sample in our study was patients with moderate-to-severe dementia admitted to a geropsychiatric ward. The duration being assessed was from the onset of illness. Our study found that hoarding occurred in Alzheimer's dementia, multi-infarct dementia, and dementia NOS; this finding suggests that hoarding is a universal problem in various kinds of dementia.It was believed that older women hoard more than older men,7 but our study showed no sex differences. We also did not find a difference in the MMSE score between groups. This finding may be because our study group was more homogenous, with most patients having moderate-to-severe dementia. Homma et al.12 reported that hoarding was found in 5.3% of patients with moderate dementia but was not found in those with mild dementia. Persons suffering from cognitive impairment may be unable to determine the relative importance of articles, so they tend to save everything, resulting in “conditions of hoarding and filth.” Some patients in this study tended to save everything they considered valuable even before disease onset. Initially, they stored their collection well, but it eventually became disorganized as their dementia progressed.Analysis of the psychiatric symptoms showed that patients who hoarded had a higher prevalence of repetitive behaviors, hyperphagia, and pilfering. Hoarding has been classified with repetitive behaviors and has been postulated to involve hippocampus dysfunction.3,4 In a recent report, magnetic resonance imaging (MRI) showed some atrophy of the hippocampus in both Alzheimer's dementia and multi-infarct dementia patients.13 Alteration of eating habits has been found in dementia patients, and one study demonstrated that 26% of patients had eaten significantly more at some stage since the onset of dementia.14 One patient in this study stored spoiled food, which caused gastrointestinal problems. He insisted that he was hungry and the food was still edible. Some patients with hyperorality stored food around their bedside during hospitalization. During hospitalization, some hoarders took objects from the ward or from other patients. Often these patients failed to discriminate and claimed that these objects belonged to them. Newspapers, magazines, food, soap, and toilet paper were stored under the bed, in the bedclothes, or in the closet. Greenberg et al.2 also reported this phenomenon.One study reported that some patients carried or stored their hoarded items out of fear of theft.2 We also found this to be true in our patients, but in this study, the delusion of theft was not significantly different in the two groups. It is possible that the delusion of theft is common in dementia and explains hoarding behavior in only some patients.Hoarding sometimes is a safety hazard and can damage health. Patients may collect dangerous objects, newspapers may cause a fire, and outdated food spoils. Intervention is needed in these circumstances. However, the causes of hoarding vary from individual to individual. Before there is intervention, the underlying etiology must be explored. For example, dementia patients with hyperphagia who store food may benefit from drugs such as fluvoxamine.15 Phenylpropanolamine, which has been reported to inhibit feeding and hoarding in rats, may also help these patients.16 Patients who exhibit hoarding because of delusion of theft may respond to antipsychotic treatment. Also, Hogstel7 has proposed some nonpharmacological interventions for hoarding in elderly patients, such as distraction, limiting opportunities, behavioral therapy, group discussion, and removing potentially dangerous objects.7This is a preliminary report of hoarding behavior in dementia patients. There were two limitations in this study. First, because of the protean expression of hoarding behavior, it is difficult to clearly define such behavior. Second, the population in this study consisted of inpatients with moderate-to-severe dementia. Further study is needed to identify hoarding behavior in a community-based population. Also, it would be of interest to compare transnational and cross-cultural differences in hoarding behavior among dementia patients.CONCLUSIONHoarding is common in patients with dementia of various etiologies, particularly in those who exhibit repetitive behaviors, hyperphagia, and pilfering. The pathogenesis of hoarding is complex. The underlying etiologies must be clarified before there is intervention.References1LWWarrenJCOstromPack rats: world-class saversPsychology Today22198858622DGreenbergEWitztumALevyHoarding as a psychiatric symptomJ Clin Psychiatry5119904174213DJLuchinsA possible role of hippocampal dysfunction in schizophrenic symptomatologyBiol Psychiatry28199087914DJLuchinsMBGoldmanMLiebRepetitive behaviors in chronically institutionalized schizophrenic patientsSchizophr Res819921191235FRFrankenburgHoarding in anorexia nervosaBr J Med Psychol57198457606DGreenbergCompulsive hoardingAm J Psychother4119874094167MOHogstelUnderstanding hoarding behaviors in the elderlyAm J Nurs93199342458American Psychiatric AssociationDiagnostic and Statistical Manual of Mental Disorders3rd Edition1987American Psychiatric AssociationWashington, DCRevised9GMcKhannDDrachmanMFolsteinClinical diagnosis of Alzheimer's disease: reports of the NINCDS-ADRDA Work Group under the auspices of the Department of Health and Human Services Task Force on Alzheimer's DiseaseNeurology34198493994410BReisbergJBorensteinSPSalobBehavioral symptoms in Alzheimer's disease: phenomenology and treatmentJ Clin Psychiatry48suppl198791511MFFolsteinSEFolsteinPRMcHughMini-Mental State: a practical method for grading the cognitive state of patients for the clinicianJ Psychiatr Res12197518919812AHommaTIshiiRNiinaRelationship of behavioral complications and severity of dementia in Japanese elderly personsAlzheimer Dis Assoc Disord81994465313TMatsuzawaMTHishinumaHMatsuiSevere atrophy of amygdala and hippocampus in both Alzheimer's disease and multi-infarct dementiaScientific Reports of the Research Institute, Tohoku Univ. (Med.)371990232514CHMorrisRAHopeCGFaiburnEating habits in dementia: a descriptive studyBr J Psychiatry154198980180615RHopePAllmanHyperphagia in dementia: fluvoxamine takes the biscuitJ Neurol Neurosurg Psychiatry5419918816PJWellmanALevyInhibition of feeding and hoarding behaviors by phenylpropanolamine in the adult ratPharmacol Biochem Behav2919887981
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B78A3E78B5029ADA30D3B9CB9FF382C6880564F.txt b/test/dataset/in/resources/corpus/Clean_0B78A3E78B5029ADA30D3B9CB9FF382C6880564F.txt
deleted file mode 100644
index bca53dc..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B78A3E78B5029ADA30D3B9CB9FF382C6880564F.txt
+++ /dev/null
@@ -1 +0,0 @@
-gerontgerontThe Gerontologist1758-53410016-9013Oxford University Press10.1093/geront/gnp170FUNCTIONAssistive Device Use as a Dynamic Acquisition Process in Later LifePresslerKaris A.MA12FerraroKenneth F.PhD22Center on Aging and the Life Course and Department of Sociology, Purdue University, West Lafayette, Indiana1Address correspondence to Karis A. Pressler, MA, Center on Aging and the Life Course, Purdue University, Ernest C. Young Hall, 155 South Grant Street, West Lafayette, IN 47907-2114. E-mail: kpressler@purdue.eduDecision Editor: William J. McAuley, PhD620102712010503371381281020098122009© The Author 2009. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org.2010Purpose: This study identifies risk factors, including incident disability, for the use of assistive devices (ADs) among older people. Design and Methods: Three waves of data from the National Long-Term Care Survey (NLTCS) are used to examine whether upper and lower body disability lead to use of ADs (both number of devices used and number of activities of daily living domains for which ADs are used). Predictors of AD use include demographic variables, body mass index, and disability (both initial and incident). Relationships are estimated with negative binomial regression models. Results: Lower body disability, advanced age, and obesity are consistent predictors of the number of ADs used. An interaction between age and incident disability revealed the highest rate of adoption among the younger respondents who experienced increases in disability. Implications: Many older adults use ADs in response to the disablement process. In addition to need driven by rising disability, obese older adults use more ADs. Results from this study clarify who and why ADs are adopted by older persons and should facilitate effective intervention by health care personnel and caregivers.Assistive devicesDisabilityDisablement processAssistive technologyObesityAssistive devices (ADs), also referred to as assistive technology, are “items frequently used by people with functional deficits as alternative ways of performing actions, tasks, and activities” (LaPlante, Hendershot, & Moss, 1992, p. 2). Examples of ADs include walkers, wheelchairs, raised toilet seats, and grab bars. It is now widely recognized that ADs aid functional independence, enabling older adults to remain active when facing disability in activities of daily living (ADL; Agree & Freedman, 2003; Cornman, Freedman, & Agree, 2005). The use of ADs has risen substantially in the past 20 years, and current estimates reveal that approximately one third of all adults aged 65 years or older use at least one device (Freedman, Agree, Martin, & Cornman, 2006; Schoeni, Freedman, & Martin, 2008; Spillman, 2005). From another viewpoint, more than half of all AD users are older adults, and demand is expected to increase due to population aging and growing sales and acceptance of such aids (Centers for Disease Control and Prevention and The Merck Company Foundation, 2007).ADs can be an important resource for older adults, but timely adoption can mean the difference between maintaining and losing one’s functional independence. We conceptualize the adoption of ADs as a response to the disablement process, but find that many previous studies of the topic rely on interpretations from cross-sectional analyses or point estimates of disability. The present study makes use of 10 years of data from a longitudinal panel study to identify incident disability and whether it, in turn, increases the likelihood of using one or more ADs.Disability Dynamics and Use of ADsThe disablement process can be an abrupt transition for some people, such as for stroke victims (Gitlin, Luborsky, & Schemm, 1998), but it is a long and somewhat circuitous process for others. Research during the past decade has revealed that many functional limitations emerge during late middle age and that many older people experience spurts, plateaus, and declines in disability (Manton, 2008; Manton, Gu, & Lamb, 2006; Schoeni et al., 2008). Despite predictions of a compression of morbidity (and disability), most people spend years adjusting to growing disability, seeking help through environmental modifications, health care, and activity accommodations to manage their conditions (Verbrugge & Jette, 1994). ADs are one such resource that many people use effectively in the struggle to maintain functional independence (Verbrugge, Rennert, & Madans, 1997).Broadly speaking, previous research focused on ADs has had one of two main aims. First, several studies identify trends over time in the use of ADs, and most such studies use data from cross-sectional studies, including repeated cross sections (Freedman et al., 2006; Spillman, 2005). The overall trend is straightforward—increasing use of ADs—but especially for devices to aid walking, bathing, and toileting (Freedman et al., 2006; Russell, Hendershot, LeClere, Howie, & Adler, 1997).Second, a number of studies have been conducted to identify which older adults are most likely to adopt ADs to help maintain their functional independence. Findings from these studies generally concur that advanced age and disability are the most consistent predictors of use of ADs by older adults (Agree, Freedman, Cornman, Wolf, & Marcotte, 2005; Agree, Freedman, & Sengupta, 2004; Cornman et al., 2005; Hartke, Prohaska, & Furner, 1998; Tomita, Mann, Fraas, & Stanton, 2004). At the same time, there is some evidence of a nonlinear relationship between disability and AD use: a rise in AD use across low-to-moderate levels of disability but less so thereafter suggesting a threshold effect (Mathieson, Kronenfeld, & Keith, 2002). Research on predictors of AD use has also addressed the related question of whether such devices supplement or substitute for personal care. Most studies of these processes provide support for a process of supplementation (Agree, 1999; Hoenig, Taylor, & Sloan, 2003; Manton, Corder, & Stallard, 1993), although ADs may substitute for informal care on specific tasks (Agree & Freedman, 2000; Agree et al., 2005).Previous research has contributed much to our understanding of the magnitude of recent increases in AD use and who is likely to adopt them, but we identify four limitations of prior research in order to advance our knowledge and improve interventions. First, most studies examine overall disability, but Verbrugge and colleagues (1997) showed that lower body disability is more consequential to AD use. Accordingly, we differentiate upper and lower body disability as predictors of ADs.Second, beyond age and disability, there is little agreement in the literature as to other factors that predict AD use. For instance, some studies show that persons with less education are more likely to use ADs (e.g., Kaye, Kang, & LaPlante, 2000), whereas other studies show the opposite (e.g., Cornman et al., 2005). Similarly, there is inconsistency in reports of racial differences in AD use (cf. Cornman & Freedman, 2008; Kaye, Yeager, & Reed, 2008). Moreover, some variables that are theoretically relevant have received scant attention. One example is obesity, which many studies show is related to disability (Ferraro & Kelley-Moore, 2003; Guralnik, Fried, & Salive, 1996), but few studies actually examine whether obesity increases AD use (Cornman & Freedman; Verbrugge & Sevak, 2002). The current study considers these and other covariates while prospectively modeling use of ADs over time.Third, there is some disagreement as to what actually constitutes AD use. Some studies (including the present one) identify ADs used specifically for physical disability (Hoenig et al., 2003; Jette, 1994; Manton et al., 1993), but other studies use a more expansive definition that includes visual, cognitive, and hearing devices (Mann, Ottenbacher, Hurren, & Tomita, 1995; Mann et al., 2008; Tomita, Mann, Fraas, & Burns, 1997). Of course, with the more inclusive definition, nearly all older adults would be AD users. In addition, Cornman and colleagues (2005) show quite convincingly that some surveys underestimate AD use by posing the question to only those respondents who report difficulty with daily activities. This decision may contribute to inconsistency in findings on the topic and understate the importance of AD use in the daily lives of older people.Fourth, if AD use is presumed to be a resource that is activated in response to growing disability, longitudinal panel studies are needed to uncover how people respond to the disability episodes. Cross-sectional analyses might even give the appearance of less disability if AD use is part of the question tapping performance of activities. The extant literature is replete with cross-sectional analyses as well as calls for panel studies (e.g., Agree et al., 2004). Nevertheless, we were able to identify only one study that used a prospective two-wave panel design on a national sample (Cornman & Freedman, 2008) and a few others that are experiments, testing the effect of AD use on a select sample over time (Gosman-Hedström, Claesson, & Blomstrand, 2002; Mann, Ottenbacher, Fraas, Tomita, & Granger, 1999; Mann et al., 2008).There are two primary reasons why longitudinal panel designs make sense for research on ADs. First, illness episodes are the likely turning points in AD adoption. During the early stages of a disease, a person may be able to manage their condition without an AD, but the advanced stages of a disease may mean that the person needs to do more to maintain the same functional level. Tracking this over time is important. Indeed, Agree and colleagues (2004, p. 267) called for incorporating change in disability into the study of the “dynamic acquisition process” whereby older people adopt an AD—and this is precisely the aim of the present analysis.Second, as disability increases, it may diffuse to other domains of functioning (Ferraro & Shippee, 2009). Do most people use multiple devices within one domain of functioning? Or might the disablement process lead to AD use in multiple domains of functioning? To answer these questions, it would be helpful if research could not only identify who uses ADs—and how many—but also the number of functional domains affected. As the number of domains rises for which a person uses one or more ADs, this is likely an indicator of overall vulnerability.Three main research questions guide the present research. First, does incident disability lead to the use of one or more ADs? Second, if yes, do both upper and lower body disability increase the likelihood of adopting ADs? Third, does rising disability also lead to domain diffusion in the adoption of ADs? Answers to these questions are vital to effective intervention by health care personnel and caregivers who assist older adults with ADLs. Our aim is to capture important transitions in the disability process as it unfolds and test the effect that this change has on subsequent use of ADs.MethodsSampleData for this study come from the National Long Term Care Survey (NLTCS), a nationally representative sample of those aged 65 years and older in the United States, drawn from the Medicare beneficiary list, accounting for 97% of U.S. older adults (Manton, Corder, & Stallard, 1997). Even though the NLTCS began in 1984, the current study focuses on the three most recent waves of data: 1994, 1999, and 2004. The 1994 survey will be referred to as W1 (Wave 1), 1999 as W2, and 2004 as W3. Although the NLTCS tracks respondents who enter institutions, measures needed for this study vary across the institutional and community surveys. Therefore, only community-dwelling older adults were included here.A detailed questionnaire was administered during 1994 to 4,126 chronically disabled persons (i.e., difficulty performing an ADL or independent activity of daily living for at least 3 months) and a sample of 963 nondisabled older persons, resulting in 5,089 respondents. The inclusion of initially nondisabled persons is important given our aim to capture incident disability as it occurs naturally.Among the 5,089 W1 respondents, 1,706 (33.5%) died by W2; 320 (6.3%) were institutionalized; and 600 (11.8%) were nonrespondents. Among the 2,463 interviewed at both W1 and W2, 957 (38.9%) died by W3, 131 (5.3%) were institutionalized, and 328 (13.3%) were nonrespondents. Analyses proceeded with 2,436 respondents between W1 and W2 and 1,047 respondents across three waves.To account for sample attrition, multivariate analyses are adjusted with selection bias modeling (Heckman, 1979). A probit model predicting death was estimated with the following predictors: age, gender, lives alone, body mass index (BMI), personal care, upper body disability, lower body disability, incident lower body disability, self-rated health, smoker, and whether a proxy provided responses to the cognitive questions. A selection factor based on the inverse Mills ratio of the probit results was used to adjust for attrition in the final equations.MeasuresTwo dependent variables were created to assess use of ADs. (Of note, we follow Agree & Freedman’s, 2000, use of the term “assistive device” as opposed to “assistive technology” because of our focus on ADL disability; see also Cornman et al., 2005). The first is an overall count of ADs that respondents reported currently using, referred to as number of ADs used. When respondents indicated they used devices for both getting into or out of bed and getting around indoors, each device was counted only once. Respondents could report using as many as 29 ADs, but 88% reported using 5 or fewer ADs in 2004. A complete list of ADs used in the analyses is found in Table 1 (note a).Table 1.Descriptive Statistics of NLTCS Respondents Present at W1, W2, and W3 Community SurveysSurvey yearW1W2W3VariableRange199419992004Number of ADs useda0–161.26b,c (2.01)d1.64 (2.47)2.05 (2.83)Number of ADLs with ADse0–60.94 (1.31)1.04 (1.37)1.23 (1.44)Age65–98f78.67 (7.66)Male0–134.09%White0–187.74%Education1–73.35 (1.60)Lives alone0–137.91%42.27%Underweight (BMI <18.5)0–17.28%4.90%Normal weight (BMI 18.5–24.9)0–144.20%43.07%Overweight (BMI 25–29.9)0–131.49%34.02%Obese (BMI >30)0–117.03%18.01%Personal care0–124.37%21.84%Upper body disability (Nagi)0–121.93 (2.86)1.78 (2.87)Lower body disability (Nagi)0–123.76 (3.65)3.39 (3.60)Incident upper body disability, W1–W20–60.55 (1.32)Incident lower body disability, W1–W20–60.74 (1.37)Number of cases5,0892,4631,047Note: AD = assistive device; ADL = activity of daily living; BMI = body mass index (kg/m2); NLTCS = National Long Term Care Survey.aNumber of ADs used included the following, which are ranked in order from most to least frequently used at W3 among respondents who participated at all three waves: Shower seat/tub stool, grab bars/handle bars for bathing, walker, cane, raised toilet seat, rubber mat for bathing, hand-held shower, rail/grab bar for toileting, wheelchair, portable toilet/bedside commode, special underwear/diapers, indoor railing, furniture/walls to get around indoors, bedside railing, bedpan or urinal, lift to get in/out of bed, oxygen/respirator, orthopedic shoes, elevator/escalator, brace (leg or back), crutch, prosthesis, and chairlift on stairs. Respondents also indicated whether they used “any” AD for eating and dressing and were asked whether they used any “other” ADs for each of the ADL domains—both of these items were included in count of ADs at W2 and W3.bMean or percentage.cFor respondents available at all three waves, the mean values for number of ADs used were 0.47 at W1 and 0.89 at W2.dStandard deviation.eADLs include eating, dressing, getting in/out of bed, getting around indoors, bathing, and toileting.fAge at W1 (1994).The second dependent variable is the number of ADL domains for which respondents reported using one or more ADs. This measure, number of ADL domains with AD, taps AD use across domains of ADL (i.e., eating, dressing, getting in/out of bed, getting around indoors, bathing, and toileting). Use of these two outcomes enables us to compare the robustness of predictors by differentiating between people who use many devices within a single ADL domain and those who use few devices spread across several ADL domains.Disability was measured with the Nagi items and differentiated as upper and lower body (Nagi, 1976). Lower body disability questions asked respondents the degree of difficulty experienced climbing one flight of stairs, walking to the end of the room and back, bending to put on socks or stockings, and lifting a 10-pound package and holding it for a few minutes. Upper body disability captured the degree of difficulty respondents experienced reaching above their head, combing and brushing their hair, washing their hair, and using fingers to grasp and handle small objects (no Nagi items were used for both disability variables). Responses for each disability variable were captured on a 4-point scale (0 = not difficult, 1 = somewhat difficult, 2 = very difficult, and 3 = cannot do the task at all). Each disability index has an alpha reliability coefficient greater than .80 at each wave. To capture incident disability, persons who experienced an increase in disability were assigned the difference between waves (top-coded so that 6 = 6 or more). Persons who were followed at all three waves and had no change between W1 and W2 (n = 411) or experienced a disability decline (n = 266) were assigned a value of 0.Personal care was measured with a binary variable if the respondent reported someone helping them to perform at least one ADL during the previous week. To assess the potential influence of BMI, we used a categorical form of kilograms per meters squared. Binary variables were created for the categories outlined by the National Institutes of Health, National Heart, Lung, and Blood Institute (1998): underweight, BMI less than 18.5; normal weight, BMI of 18.5–24.9; overweight, BMI of 25.0–29.9; obese, BMI greater than or equal to 30. Control variables included age, gender (male), race (White or non-White), education, and lives alone. Several additional variables were included in preliminary multivariate analyses but omitted from the final analyses because they were nonsignificant in various specifications. These variables included marital status, comorbidity, and social isolation.Analytic PlanGiven the overdispersion of the outcome variables, negative binomial regression was selected as the statistical estimator (Long, 1997). All analyses were conducted with Stata 10 (StataCorp, 2007).The analysis was divided into two main stages. First, equations predicting number of ADs used at W2 and the number of ADL domains with ADs at W2 were estimated; two models were estimated for each outcome, where the W1 variable of the outcome was excluded and then included. The first model provides point estimates at W2 and the second reveals change in the outcome. Second, negative binomial regression models were estimated to examine change in the outcomes by W3. We capitalized on the three waves of data to isolate the influence of both initial disability (W1) and incident disability (between W1 and W2, using lagged models). In both stages, we tested for potential statistical interactions (e.g., age and disability) and polynomial terms for independent variables (e.g., disability).ResultsDescriptive statistics for the sample are presented in Table 1. AD use increased substantially over the decade, from 1.26 at W1 to 2.05 at W3 (p < .001). The number of ADL domains with AD also increased over time (p < .001). At W1, the majority of respondents were either normal weight (44%) or overweight (31%); about 17% were obese. Approximately one quarter of respondents reported receiving personal care at W1. Lower body disability was greater than upper body disability, with W2 respondents reporting a value of about 2 for upper body disability and about 4 for lower body disability.Table 2 displays the results of the negative binomial regression analyses for the two W2 dependent variables, and two models are presented for each outcome. Model 1 provides estimates of overall AD use at W2; Model 2 adds AD use at W1 in order to examine the adoption of ADs since W1.Table 2.Negative Binomial Regression Predicting Number of ADs Used at W2 (1999)Number of ADs usedNumber of ADLs with ADsW1 independent variablesModel 1Model 2Model 1Model 2Age0.065a,*** (0.010)b0.057*** (0.010)0.059*** (0.008)0.053*** (0.008)Male−0.073 (0.104)−0.128 (0.101)−0.033 (0.083)−0.054 (0.082)White−0.004 (0.108)−0.021 (0.105)−0.063 (0.085)−0.080 (0.083)Education0.087*** (0.022)0.061** (0.022)0.059** (0.018)0.041* (0.018)Lives alone0.064 (0.074)0.000 (0.073)0.078 (0.059)0.047 (0.059)Underweight (BMI <18.5)c0.133 (0.256)0.076 (0.246)0.066 (0.198)0.058 (0.195)Overweight (BMI 25–29.9)c−0.003 (0.088)0.032 (0.087)0.004 (0.072)0.030 (0.071)Obese (BMI >30)c0.343** (0.106)0.382*** (0.104)0.259** (0.084)0.267** (0.083)Personal care0.407** (0.130)0.145 (0.130)0.328** (0.099)0.138 (0.101)Upper body disability0.019 (0.019)0.022 (0.019)0.008 (0.014)0.009 (0.014)Lower body disability0.181*** (0.019)0.122*** (0.020)0.155*** (0.015)0.116*** (0.016)Number of ADs used, W10.190*** (0.023)0.127*** (0.017)Mortality λ−0.346 (0.300)−0.179 (0.290)−0.305 (0.221)−0.193 (0.218)Observations2,1612,1612,1612,161Pseudo R20.0600.0710.0870.096Likelihood ratio χ2437.45512.42529.60587.62Note: AD = assistive device; ADL = activity of daily living; BMI = body mass index (kg/m2).aUnstandardized coefficient.bStandard error.cCompared with those with a normal BMI (18.5–24.9).*p < .05. **p < .01. ***p < .001.Findings from Model 1 of Table 2 are quite similar across the W2 two outcomes. The number of ADs used and the number of ADL domains with ADs were higher for persons who were older, had more education, obese, received personal care, and had more lower body disability. When AD use at W1 was added in Model 2—predicting change in AD use from W1 to W2—most results are consistent with Model 1, except that personal care is no longer significant. Across all four equations, lower body disability predicts AD use, but upper body disability is not independently related to the outcomes.Supplementary analyses reveal that 42% of respondents present at W1 reported using ADs for bathing and/or indoor mobility difficulty at W2, which are devices that are more useful to those with compromised lower body function and require a certain degree of upper body strength and dexterity to operate. The percentage of respondents using bathing and/or mobility devices increased to 50% by W3.The results displayed in Table 3 carry the analyses forward to W3 and take advantage of data from the prior two waves. Both Models 1 and 2 include the W2 measure of AD use in order to focus on change in use of ADs. Model 1 for number of ADs reveals that age, obesity, and lower body disability at W2 are associated with an increase in AD use by W3. Again, upper body disability was not a significant predictor of AD use. In addition to initial level of lower body disability (W1), Model 1 also includes a variable for incident lower body disability (between W1 and W2), which is associated with the adoption of additional ADs (p < .01). (Incident upper body disability was not included in the equation due to the W2 level of upper body disability having no significant influence on AD use in W3.)Table 3.Negative Binomial Regression Predicting ADs Used at W3 (2004), NLTCSNumber of ADs usedNumber of ADLs with ADsIndependent variablesModel 1Model 2Model 1Model 2Number of ADs used at W20.154a,*** (0.031)b0.133*** (0.032)0.099*** (0.020)0.078*** (0.021)Age0.059*** (0.016)0.068*** (0.016)0.054*** (0.012)0.058*** (0.012)Male−0.219 (0.133)−0.261* (0.123)−0.138 (0.104)−0.179 (0.097)White0.147 (0.153)0.155 (0.152)0.082 (0.119)0.061 (0.119)Education0.026 (0.032)0.036 (0.032)0.034 (0.025)0.041 (0.025)Lives alone0.096 (0.097)0.103 (0.097)0.126 (0.075)0.138 (0.075)Underweight (BMI <18.5)c−0.433 (0.353)−0.506 (0.345)−0.437 (0.275)−0.534* (0.272)Overweight (BMI 25–29.9)c0.184 (0.131)0.261 * (0.126)0.082 (0.100)0.149 (0.096)Obese (BMI >30)c0.539*** (0.145)0.594*** (0.140)0.344** (0.109)0.394*** (0.105)Personal care0.011 (0.206)−0.098 (0.213)0.083 (0.143)−0.015 (0.148)Upper body disability at W20.004 (0.029)0.012 (0.029)−0.005 (0.020)−0.003 (0.020)Lower body disability at W10.067* (0.030)0.060* (0.026)0.079** (0.023)0.070*** (0.020)Incident lower body disability, W1–W20.135** (0.048)1.630** (0.488)0.111** (0.035)1.150*** (0.329)Interaction between age and incident lower body disability between W1–W2−0.019** (0.006)−0.013** (0.004)Mortality λ0.422 (0.387)0.659 * (0.326)0.263 (0.272)0.518 * (0.230)Observations933933933933Pseudo R20.0680.0710.1020.106Likelihood ratio χ2232.74244.07287.94300.84Note: AD = assistive device; ADL = activity of daily living; BMI = body mass index (kg/m2); NLTCS = National Long Term Care Survey.aUnstandardized coefficient.bStandard error.cCompared with those with a normal BMI (18.5–24.9).*p < .05. **p < .01. ***p < .001.Several interactions between incident lower body disability and predictors were tested in supplementary analyses, and the product term of incident disability and age was significant as shown in Model 2 of Table 3. The effect of incident lower body disability on AD use varies across age groups. Incident lower body disability increased AD use, but this relationship was moderated somewhat for the oldest respondents. Note also that with the interaction specified in Model 2, overweight emerged as a significant predictor of using more ADs by W3, and the mortality selection factor (λ) was also significant.The last two equations (columns) in Table 3 display similar models for the second outcome. Findings for the number of number of ADL domains with ADs are quite similar. The pattern of significant relationships for Model 1 is similar to the equation for number of ADs used, but the effect due to initial lower body disability is slightly stronger for number of ADLs with ADs.The findings from Model 2, which include the interaction between age and incident lower body disability, are similar to those for the first outcome, albeit overweight is not significant. The product term in Model 2 reveals the same pattern: Incident lower body disability increases AD use but this relationship is stronger for the younger members of the sample.To better understand the interaction between age and incident lower body disability that was observed for both outcomes, we provide a graphical representation for two different levels of AD use at W3 while controlling for the demographic, BMI, and disability variables. Using a logistic regression model, Figure 1 displays the predicted probability of using one or more ADs at W3 across four age groups while accounting for incident lower body disability between W1 and W2. As shown in Figure 1, the probability of using one or more ADs by W3 varies considerably by age even if there was no new disability between W1 and W2. Incident lower body disability, however, raised the likelihood of using ADs for all age groups, and the slope for persons aged 65–69 years was fairly linear. By contrast, for the oldest respondents, the slope rose faster across incident disability and then tapered off at the highest levels approaching 100% saturation.Figure 1.Predicted probability of using 1+ assistive devices (ADs) by W3 (2004) by incident lower body disability and age (N = 933).Figure 2 predicts the probability of a less frequent scenario—when a person used three or more ADs at W3. Again, there are age differences even with no incident disability. The slopes for the middling age categories are fairly linear. By contrast, the probability of using three or more ADs rose more slowly across incident disability for the youngest age group and rose faster for the oldest age group before tapering off at about 80%.Figure 2.Predicted probability of using 3+ assistive devices (ADs) by W3 (2004) by incident lower body disability and age (N = 933).DiscussionSeveral years ago, Agree and colleagues (2004, p. 267) called for research to examine use of ADs as a “dynamic acquisition process, with attention to age and disability severity.” We sought to answer that call by using three waves of panel data from a nationally representative sample of older adults to examine how ADs are adopted by persons facing ADL disability. The current study examines not only who is more likely to use ADs but also how incident disability and prior AD use influence number of ADs used and the number of ADL domains for which ADs are used.Three research questions were examined in the present study. First, we sought to determine whether incident disability leads to the use of one or more ADs. Although many gerontologists no doubt conceptualize AD use as a consequence of the disablement process, we are unaware of any nationally representative studies that used three or more waves of data to isolate whether incident disability accelerates the adoption of ADs. The present study found consistent evidence across two different outcome variables that incident disability is the precursor of AD adoption. This is not to say that baseline levels are unimportant; they are and result in AD adoption. Nevertheless, incident disability leads to additional use of ADs. For practice, therefore, it may be helpful to identify persons in the early stages of incident disability to optimize the effectiveness of ADs.Second, we sought to determine if both upper and lower body disability increase the likelihood of adopting ADs. Although some studies do not differentiate types of disability, findings from the current study are consistent with other investigations showing that lower body disability is the engine of AD adoption (Cornman & Freedman, 2008; Spillman, 2005; Verbrugge et al., 1997). In addition, the present research builds upon this finding by demonstrating that AD use is not only driven by baseline levels of lower body disability but also by incident lower body disability as well. We believe that there is now ample evidence to call for distinguishing between these two forms of disability in order to identify persons who are likely adopters of ADs.Third, we asked whether increasing disability also leads to domain diffusion in the adoption of ADs. This is an important question for intervention. Are older people in need of additional devices for one domain of ADL functioning or will the need spill over into other domains? Our results point to the latter. When incident lower body disability occurs, it leads not only to additional AD use within one domain, but across multiple domains. Underlying physical capacity is important for the performance of many ADLs. Persons who have difficulty with transfers to the toilet are probably also at risk for transfers involving getting in/out of bed. Future research could confirm this by investigating the sequence of AD adoption among older adults. These results from the NLTCS clarify that incident lower body disability diffuses to AD use across ADL domains.While answering these three research questions, two additional conclusions emerged as noteworthy. First, as suggested by Agree and colleagues (2004), age merits special attention. Based on this national sample of older adults, AD use varied by age in a nonlinear way. Although there are age differences in the likelihood of AD use—such that persons of advanced age (80+ years at W1) used more devices—we also observed a tapering of the rising probability of AD adoption for the oldest age groups (perhaps reflecting a saturation effect for the oldest respondents). Small changes in incident lower body disability were associated with rising AD use among the youngest respondents, an important finding for intervention efforts.Second, we were struck by the consistent predictive ability of obesity as a risk factor for AD use. Across all models, regardless of the outcome considered, obesity was a consistent, significant, and positive predictor. Obesity is clearly a risk factor for disability, but few previous studies have included it as a predictor of AD use. Of those that did, the results revealed either no influence on AD use (Verbrugge & Sevak, 2002) or few findings that merited discussion (Cornman & Freedman, 2008). The current study reveals the lasting impact that obesity has on the use of ADs: Those who were obese at either W1 or W2 were more likely to use a greater number of ADs at W2 and W3, respectively. Given the increased prevalence of adult obesity during recent decades, it should be recognized as a critical risk factor for both disability and AD use.This study’s contributions must be weighed against three limitations. First, although we devised an analytic plan to take advantage of the sequence of events observed in these data, the NLTCS 5-year time frame is fairly coarse when attempting to measure both incident disability and AD use (see also Freedman, Martin, & Schoeni, 2002; Spillman, 2005). Whereas the precise date of AD adoption was not known, we know only that it occurred during the 5-year period between surveys. Accordingly, we used a lagged analytic strategy to guard against conclusions that could be due to reverse causality (because both incident disability and AD adoption can occur within either of the 5-year periods). This strategy enabled us to isolate the influence of incident disability on adoption of ADs, but the 5-year periods are fairly long for studying these dynamic processes. There is clearly a need for research with shorter time frames to detect early stages of incident disability.Second, although we theorized that incident disability is a risk factor for AD use, there are other factors that merit attention, including illness episodes and hospitalizations (Wolff, Agree, & Kasper, 2005). Hospitalizations linked to specific illness episodes such as a stroke serve as a gateway to both incident disability and AD use (Gitlin et al., 1998). Now that we have documented that AD adoption is linked to incident disability, there is a need for future research to identify these “upstream” factors that may lead to both rising disability and AD use.Third, although we know that AD adoption does not occur in a social vacuum, the NLTCS does not provide information on which social network members facilitate the process. We envision caregivers as pivotal agents in the decision to use ADs, but this proposition merits systematic examination in future research.Implications for Policy and PracticeOur findings have implications for caregivers, clinicians, and policy makers, and we articulate two that are especially important. First, according to several studies, poor lower body function is a strong indicator of additional functional decline among older adults (e.g., Guralnik et al., 2000). We found that incident lower body disability triggers a response of increased reliance on ADs—and across multiple ADL domains—to counteract the slippery slope associated with functional decline. Moreover, those who begin experiencing this decline after age 70 are especially sensitive to respond to this change by using more ADs. Therefore, clinicians and caregivers should take note of increases in lower body disability among those benefitting from their care and be prepared to respond by suggesting the use of ADs to effectively bridge the gap between impairment and function. Medicare’s durable medical equipment benefit has helped millions of Americans obtain ADs, but many remain unaware of this benefit (Wolff et al., 2005).Second, obesity is more prevalent among the baby boom generation (born between 1946 and 1965) compared with those born in the prior decades (Leveille, Wee, & Iezzoni, 2005). Based on our finding that obese individuals are more likely to use more ADs—and for a wider number of ADLs—it should be anticipated that AD use will climb as the baby boom population ages. Beyond their sheer numbers, the higher prevalence of obesity among baby boomers and their higher levels of education will make ADs even more widely used in the future.We appreciate the comments on an earlier version of the manuscript by Kathy Abrahamson, Neale R. Chumbler, Ann Howell, Markus H. Schafer, Tetyana P. Shippee, and two anonymous reviewers.AgreeEMThe influence of personal care and assistive devices on the measurement of disabilitySocial Science & Medicine199948427443AgreeEMFreedmanVAIncorporating assistive devices into community-based long-term care: An analysis of the potential for substitution and supplementationJournal of Aging and Health200012426450AgreeEMFreedmanVAA comparison of assistive technology and personal care in alleviating disability and unmet needThe Gerontologist200343335344AgreeEMFreedmanVACornmanJCWolfDAMarcotteJEReconsidering substitution in long-term care: When does assistive technology take the place of personal care?Journal of Gerontology: Social Sciences200560BS272S280AgreeEMFreedmanVASenguptaMAFactors influencing the use of mobility technology in community-based long-term careJournal of Aging and Health200416267307Centers for Disease Control and Prevention and The Merck Company FoundationThe state of aging and health in America 20072007Whitehouse Station, NJMerck Company FoundationRetrieved October 13, 2009, from http://www.cdc.gov/aging/pdf/saha_2007.pdfCornmanJCFreedmanVARacial and ethnic disparities in mobility device use in late lifeJournal of Gerontology: Social Sciences200863S34S41CornmanJCFreedmanVAAgreeEMMeasurement of assistive device use: Implications for estimates of device use and disability in late lifeThe Gerontologist200545347358FerraroKFKelley-MooreJACumulative disadvantage and health: Long-term consequences of obesity?American Sociological Review200368707729FerraroKFShippeeTPAging and cumulative inequality: How does inequality get under the skin?The Gerontologist200949333343FreedmanVAAgreeEMMartinLGCornmanJCTrends in the use of assistive technology and personal care for late-life disability, 1992-2001The Gerontologist200646124127FreedmanVAMartinLGSchoeniRFRecent trends in disability and functioning among older adults in the United StatesJournal of the American Medical Association200228831373146GitlinLNLuborskyMRSchemmRLEmerging concerns of older stroke patients about assistive device useThe Gerontologist199838169180Gosman-HedströmGGClaessonLBlomstrandCAssistive devices in elderly people after stroke: A longitudinal, randomized study—The Göteborg 70+ stroke studyScandinavian Journal of Occupational Therapy20029109118GuralnikJMFerrucciLPieperCFLeveilleSGMarkidesKSOstirGVLower extremity function and subsequent disability: Consistency across studies, predictive models, and value of gait speed alone compared with the short physical performance batteryJournal of Gerontology: Medical Sciences200055AM221M231GuralnikJMFriedLPSaliveMEDisability as a public health outcome in the aging populationAnnual Review of Public Health1996172546HartkeRJProhaskaTRFurnerSEOlder adults and assistive devices: Use, multiple-device use, and needJournal of Aging and Health19981099116HeckmanJJSample selection bias as a specification errorEconometrica197947153161HoenigHTaylorDHSloanFADoes assistive technology substitute for personal assistance among the disabled elderly?American Journal of Public Health200393330337JetteAMHow measurement techniques influence estimates of disability in older populationsSocial Science & Medicine199438937942KayeHSKangTLaPlanteMPMobility device use in the United States. Disability Statistics Report, (14)2000Washington, DCU.S. Department of Education, National Institute on Disability and Rehabilitation ResearchKayeHSYeagerPReedMDisparities in usage of assistive technology among people with disabilitiesAssistive Technology200820194203LaPlanteMPHendershotGEMossAJAssistive technology devices and home accessibility features: Prevalence, payment, need, and trends (No. 217)1992Hyattsville, MDNational Center for Health StatisticsLeveilleSGWeeCCIezzoniLITrends in obesity and arthritis among baby boomers and their predecessors, 1971–2002American Journal of Public Health20059516071613LongJSRegression models for categorical and limited dependent variables1997Thousand Oaks, CASageMannWCJohnsonJLLynchLGJustissMDTomitaMWuSSChanges in impairment level, functional status, and use of assistive devices by older people with depressive symptomsAmerican Journal of Occupational Therapy200862917MannWCOttenbacherKJFraasLTomitaMGrangerCVEffectiveness of assistive technology and environmental interventions in maintaining independence and reducing home care costs for the frail elderly: A randomized controlled trialArchives of Family Medicine19998210217MannWCOttenbacherKJHurrenDTomitaMRelationship of severity of physical disability to pain, functional status, and assistive device use of home-based elderly clientsHome Health Care Management & Practice199587584MantonKGRecent declines in chronic disability in the elderly U.S. population: Risk factors and future dynamicsAnnual Review of Public Health20082991113MantonKGCorderLStallardEChanges in the use of personal assistance and special equipment from 1982 to 1989: Results from the 1982 and 1989 NLTCSThe Gerontologist199333168178MantonKGNational Long-Term Care Survey: 1994, 1999, and 2004 [Computer file]. ICPSR09681-v1Ann Arbor, MIInter-university Consortium for Political and Social Research [distributor]2007-02-07. doi:10.3886/ICPSR09681MantonKGCorderLStallardEChronic disability trends in elderly United States population: 1982-1994Proceedings of the National Academy of Sciences of the United States of America19979425932598MantonKGGuXLambVLChange in chronic disability from 1982 to 2004/2005 as measured by long-term changes in function and health in the U.S. elderly populationProceedings of the National Academy of Sciences of the United States of America20061031837418379MathiesonKMKronenfeldJJKeithVMMaintaining functional independence in elderly adults: The roles of health status and financial resources in predicting home modifications and use of mobility equipmentThe Gerontologist2002422431NagiSZAn epidemiology of disability among adults in the United StatesMillbank Memorial Fund Quarterly197654439467National Institutes of Health, National Heart, Lung, and Blood InstituteClinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults: The evidence report1998Hyattsville, MDNational Center for Health StatisticsRetrieved October 13, 2009, from http://www.nhlbi.nih.gov/guidelines/obesity/ob_gdlns.pdfRussellJNHendershotGELeClereFHowieLJAdlerMTrends and differential use of assistive technology devices: United States, 1994 (No. 292)1997Hyattsville, MDNational Center for Health StatisticsSchoeniRFFreedmanVAMartinLGWhy is late-life disability declining?Milbank Quarterly2008864789SpillmanBAssistive device use among the elderly: Trends, characteristics of users, and implications for modeling2005Report to the Department of Health and Human Services, Assistant Secretary for Planning and Evaluation, Office of Aging and Long-Term Care Policy. Washington, DC: The Urban Institute. Retrieved October 13, 2009, from http://www.urban.org/url.cfm?ID=1001277StataCorpStata Statistical Software: Release 102007College Station, TXStata CorporationTomitaMRMannWCFraasLFBurnsLLRacial differences of frail elders in assistive technologyAssistive Technology19979140151TomitaMRMannWCFraasLFStantonKMPredictors of the use of assistive devices that address physical impairments among community-based frail eldersJournal of Applied Gerontology200423141155VerbruggeLMJetteAMThe disablement processSocial Science & Medicine199438114VerbruggeLMRennertCMadansJHThe great efficacy of personal and equipment assistance in reducing disabilityAmerican Journal of Public Health199787384392VerbruggeLMSevakPUse, type, and efficacy of assistance for disabilityJournal of Gerontology: Social Sciences200257BS366S379WolffJLAgreeEMKasperJDWheelchairs, walkers, and canes: What does Medicare pay for, and who benefits?Health Affairs20052411401149
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B84673E2A77F8873ADE3BF869C67FECB1786C32.txt b/test/dataset/in/resources/corpus/Clean_0B84673E2A77F8873ADE3BF869C67FECB1786C32.txt
deleted file mode 100644
index 5a8c676..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B84673E2A77F8873ADE3BF869C67FECB1786C32.txt
+++ /dev/null
@@ -1 +0,0 @@
-cercorcercorCerebral Cortex1460-21991047-3211Oxford University Press10.1093/cercor/bhp250ArticlesPrefrontal Cortical Contribution to Risk-Based Decision MakingSt. OngeJennifer R.FlorescoStan B.Department of Psychology and Brain Research Center, University of British Columbia, Vancouver, BC, Canada V6T 1Z4Address correspondence to Stan B. Floresco, PhD, Department of Psychology and Brain Research Center, University of British Columbia, 2136 West Mall, Vancouver BC V6T 1Z4, Canada. Email: floresco@psych.ubc.ca.82010511200920818161828© The Author 2009. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org2010Damage to various regions of the prefrontal cortex (PFC) impairs decision making involving evaluations about risks and rewards. However, the specific contributions that different PFC subregions make to risk-based decision making are unclear. We investigated the effects of reversible inactivation of 4 subregions of the rat PFC (prelimbic medial PFC, orbitofrontal cortex [OFC], anterior cingulate, and insular cortex) on probabilistic (or risk) discounting. Rats were well trained to choose between either a “Small/Certain” lever that always delivered 1 food pellet, or another, “Large/Risky” lever, which delivered 4 pellets, but the probability of receiving reward decreased across 4 trial blocks (100%, 50%, 25%, and 12.5%). Infusions of gama-aminobutyric acid agonists muscimol/baclofen into the medial PFC increased risky choice. However, similar medial PFC inactivations decreased risky choice when the Large/Risky reward probability increased over a session. OFC inactivation increased response latencies in the latter trial blocks without affecting choice. Anterior cingulate or insular inactivations were without effect. The effects of prelimbic inactivations were not attributable to disruptions in response flexibility or judgments about the relative value of probabilistic rewards. Thus, the prelimbic, but not other PFC regions, plays a critical role in risk discounting, integrating information about changing reward probabilities to update value representations that facilitate efficient decision making.anterior cingulateorbitofrontal cortexprelimbic cortexreversible inactivationratIntroductionDecision making entailing cost–benefit evaluations about risks and rewards recruits numerous regions of the prefrontal cortex (PFC). Initial work by Damasio, Bechara, and colleagues, employing what is now referred to as the Iowa Gambling task, revealed that patients with damage to the ventromedial and orbital regions of the PFC make risky, disadvantageous choices compared with control subjects (Bechara et al. 1994, 1999; Damasio 1994). Subsequent research, using a variety of different tasks, patient populations, as well as neuroimaging in healthy subjects, has implicated multiple interconnected regions of the PFC in mediating these forms of decisions. These include the dorsolateral PFC (Ernst et al. 2002; Manes et al. 2002; Brand et al. 2004; Fellows and Farah 2005; Labudda et al. 2008), the medial and lateral regions of the orbitofrontal cortex (OFC; Rogers, Everitt, et al. 1999; Rogers, Owen, et al. 1999; Manes et al. 2002; Clark et al. 2003, 2008; Ernst et al. 2004; Fukui et al. 2005), the dorsal (Brodmann's Area 24) and ventral (Area 32) regions of the anterior cingulate (Ernst et al. 2002; Labudda et al. 2008; Lawrence et al. 2009), and the insular cortex (Ernst et al. 2002; Bar-On et al. 2003; Clark et al. 2008; Smith et al. 2009).Although multiple frontal lobe regions have been implicated in facilitating risk/reward decisions, the specific contribution that different PFC regions make to risk-based decision making remains unclear. For example, impaired decision making on the Iowa Gambling task displayed by patients with ventromedial PFC damage may reflect a failure to learn the win–lose contingencies associated with the different choice options, a preference for high-risk options regardless of potential loss or deficits in strategy acquisition and maintenance (Clark et al. 2004). Likewise, damage to either the OFC or dorsolateral PFC has been associated with impaired decision making in some studies (Bechara et al. 1994, 1999; Rogers, Owen, et al. 1999; Brand et al. 2004; Fellows and Farah 2005) but not in others (Bechara et al. 1998; Manes et al. 2002; Clark et al. 2003). Different regions of the anterior cingulate have been implicated in multiple aspects of decision making, depending on the type of task used. These include monitoring increasing potential or received gain (Rogers et al. 2004; Tom et al. 2007), loss or negative feedback (Blair et al. 2006; Marsh et al. 2007; Cohen et al. 2008), or changes in reward contingencies that require modifications in behavior (Bush et al. 2002; Cohen et al. 2008). These discrepant findings may be due in part to differences in the specific loci and laterality of damage that may vary considerably between subjects and may also encompass multiple PFC regions. Thus, studies that afford greater experimental control, as may be achieved with animal models, may help clarify how different PFC regions contribute to aspects of risk-based decision making.One component of decision making that can be assessed in rodents is the evaluation of certain costs associated with different actions relative to the potential reward that may be obtained by those actions. In these studies, rats typically choose between smaller rewards associated with a nominal cost, and larger, yet more costly rewards (Floresco, St Onge, et al. 2008). Dissociable regions of the rat PFC appear to mediate different forms of these cost–benefit judgments. The OFC appears to be a critical component of the circuitry underlying delay discounting, where animals choose between small, immediate rewards and larger, delayed rewards (Mobini et al. 2002; Winstanley et al. 2004; Rudebeck et al. 2006). On the other hand, the dorsal anterior cingulate has been implicated in mediating decisions related to effort costs associated with larger rewards (Walton et al. 2003; Rudebeck et al. 2006; Floresco and Ghods-Sharifi 2007).Interestingly, there have been relatively few studies on the involvement of different regions of the rat PFC in risk-based decision making. Of these, the OFC has received the most attention. Mobini et al. (2002) reported that large lesions of the OFC (encompassing portions of the medial PFC) increased preference for a small, but certain, food reward relative to a larger reward delivered in a probabilistic manner. Another study used a procedure modeled after the Iowa Gambling task, where rats learned contingencies of decision alternatives in 1 session and adjusted choice based on evaluations of previous outcomes (Pais-Vieira et al. 2007). Rats with OFC lesions developed a preference for the high-risk/large reward (LR) over the low-risk/small reward (SR) option, even though the expected value of the 2 options was equal. In both of these studies, lesions were induced prior to training on the task, indicating that the OFC may play a role in the initial learning of risk/reward contingencies. In contrast, the contribution of the rat OFC to risk/reward judgments after animals have learned the relative value of different response options is unknown. Moreover, to our knowledge, there have been no studies on the contribution of the rodent medial PFC or insular cortex to risk-based decision making.To address this issue, the present study investigated the effects of reversible inactivation of different subregions of the rat PFC on risk-based decision making, using a probabilistic discounting task conducted in an operant chamber. Rats chose between responses that yielded a small/certain food reward and a larger reward delivered in a probabilistic manner, where the odds of receiving the larger reward decreased over a session. Previous studies using similar procedures have implicated the nucleus accumbens (Cardinal and Howes 2005), the basolateral amygdala (Ghods-Sharifi et al. 2009), and the dopamine system (St Onge and Floresco 2009) in biasing choice toward larger, probabilistic rewards. Each of these systems shares connections with different regions of the PFC (Beckstead et al. 1979; McDonald 1991). In the present study, rats were well trained on this task prior to receiving reversible inactivations of 1 of 4 different PFC subregions. Each of these regions sends projections to the nucleus accumbens core (Brog et al. 1993), which has been shown previously to mediate to risk discounting (Cardinal and Howes 2005). Risk discounting was altered selectively by inactivation of the prelimbic region of the medial PFC, a region that is anatomically homologous to Area 32 of the human anterior cingulate (Uylings and van Eden 1990) but also shares functional homology to the dorsolateral PFC (Brown and Bowman 2002; Uylings et al. 2003). Accordingly, subsequent experiments were conducted to clarify the specific nature of the disruption in risk-based decision making induced by medial PFC inactivation.Experiment 1: The Effects of Inactivations of the Medial PFC, OFC, Anterior Cingulate, and Insular Cortex on Risk DiscountingMaterials and MethodsAnimalsMale Long Evans rats (Charles River Laboratories, Montreal, Canada) weighing 275–300 g at the beginning of behavioral training were used for the experiment. On arrival, rats were given 1 week to acclimatize to the colony and food restricted to 85–90% of their free-feeding weight for an additional 1 week before behavioral training. Rats were given ad libitum access to water for the duration of the experiment. Feeding occurred in the rats’ home cages at the end of the experimental day, and body weights were monitored daily to ensure a steady weight loss during food restriction and maintenance or weight gain for the rest of the experiment. All testing was in accordance with the Canadian Council of Animal Care and the Animal Care Committee of the University of British Columbia.ApparatusBehavioral testing for all experiments described here was conducted in 12 operant chambers (30.5 × 24 × 21 cm; Med-Associates, St Albans, VT) enclosed in sound-attenuating boxes. The boxes were equipped with a fan to provide ventilation and to mask extraneous noise. Each chamber was fitted with 2 retractable levers, one located on each side of a central food receptacle where food reinforcement (45 mg; Bioserv, Frenchtown, NJ) was delivered via a pellet dispenser. The chambers were illuminated by a single 100-mA houselight located in the top center of the wall opposite the levers. Four infrared photobeams were mounted on the sides of each chamber. Locomotor activity was indexed by the number of photobeam breaks that occurred during a session. All experimental data were recorded by an IBM personal computer connected to the chambers via an interface.Lever-Pressing TrainingOur initial training protocols have been described previously (Floresco, Tse, et al. 2008; St Onge and Floresco 2009). On the day prior to their first exposure to the chambers, rats were given approximately 25 sugar reward pellets in their home cage. On the first day of training, 2–3 pellets were delivered into the food cup, and crushed pellets were placed on a lever before the animal was placed in the chamber. Rats were first trained under a fixed-ratio 1 schedule to a criterion of 60 presses in 30 min, first for one lever, and then repeated for the other lever (counterbalanced left/right between subjects). Rats were then trained on a simplified version of the full task. These 90 trial sessions began with the levers retracted and the operant chamber in darkness. Every 40 s, a trial was initiated with the illumination of the houselight and the insertion of 1 of the 2 levers into the chamber. If the rat failed to respond on the lever within 10 s, the lever was retracted, the chamber darkened, and the trial was scored as an omission. If the rat responded within 10 s, the lever retracted, and a single pellet was delivered with 50% probability. This procedure was used to familiarize the rats with the probabilistic nature of the full task. In every pair of trials, the left or right lever was presented once, and the order within the pair of trials was random. Rats were trained for approximately 5–6 days to a criterion of 80 or more successful trials (i.e., ≤10 omissions).Risk-Discounting TaskThe task was modified from Cardinal and Howes’ (2005) procedure, which we have previously used to assess the role of dopamine and the basolateral amygdala in risk-based decision making (Ghods-Sharifi et al. 2009; St Onge and Floresco 2009). Rats received daily sessions consisting of 72 trials, separated into 4 blocks of 18 trials. The entire session took 48 min to complete, and animals were trained 6–7 days per week. A session began in darkness with both levers retracted (the intertrial state). A trial began every 40 s with the illumination of the houselight and, 3 s later, insertion of one or both levers into the chamber (the format of a single trial is shown in Fig. 1). One lever was designated the Large/Risky lever, the other the Small/Certain lever, which remained consistent throughout training (counterbalanced left/right). If the rat did not respond by pressing a lever within 10 s of lever presentation, the chamber was reset to the intertrial state until the next trial (omission). When a lever was chosen, both levers retracted. Choice of the Small/Certain lever always delivered one pellet with 100% probability; choice of the Large/Risky lever delivered 4 pellets but with a particular probability (see below). When food was delivered, the houselight remained on for another 4 s after a response was made, after which the chamber reverted to the intertrial state. Multiple pellets were delivered 0.5 s apart. The 4 blocks were comprised of 8 forced-choice trials where only one lever was presented (4 trials for each lever, randomized in pairs) permitting animals to learn the amount of food associated with each lever press and the respective probability of receiving reinforcement over each block. This was followed by 10 free-choice trials, where both levers were presented and the animal chose either the Small/Certain or the Large/Risky lever. The probability of obtaining 4 pellets after pressing the Large/Risky lever varied across blocks: It was initially 100%, then 50%, 25%, and 12.5%, respectively, for each successive block. For each session and trial block, the probability of receiving the LR was drawn from a set probability distribution. Therefore, on any given day, the probabilities in each block may vary, but on average across many training days, the actual probability experienced by the rat will approximate the set value. Using these probabilities, selection of the Large/Risky lever would be advantageous in the first 2 blocks, and disadvantageous in the last block, whereas rats could obtain an equivalent number of food pellets after responding on either lever during the 25% block. Therefore, in the last 3 trial blocks of this task, selection of the larger reward option carried with it an inherent “risk” of not obtaining any reward on a given trial. Latencies to initiate a choice and overall locomotor activity (photobeam breaks) were also recorded.Figure 1.Task design. Cost–benefit contingencies associated with responding on either lever (A) and format of a single free-choice trial (B) of the risk-discounting task.A priori we determined that if inactivation of a particular PFC region resulted in an increased preference for the Large/Risky lever, a separate group of animals would be trained on a similar discounting task, in which the probabilities associated with the Large/Risky lever increased, rather than decreased across the 4 blocks (i.e., 12.5%, 25%, 50%, and 100%). The purpose of this was to determine if an increased preference for the Large/Risky lever following PFC inactivation was due specifically to a general increase in risky choice or to alterations in other cognitive processes that facilitate risk discounting. All other aspects of the training procedures were identical to those described above.Training Procedure, Surgery, and Microinfusion ProtocolRats were trained on the task until as a group they 1) chose the Large/Risky lever during the first trial block (100% probability) on at least 80% of successful trials and 2) demonstrated stable baseline levels of choice. Brain inactivations were conducted once a group of rats displayed stable patterns of choice for 3 consecutive days, assessed using a procedure similar to that described by Winstanley et al. (2005), Floresco, Tse, et al. (2008), and St Onge and Floresco (2009). In brief, data from 3 consecutive sessions were analyzed with a repeated-measures analysis of variance (ANOVA) with 2 within-subjects factors (Day and Trial Block). If the effect of Block was significant at the P < 0.05 level but there was no main effect of Day or Day × Trial Bock interaction (at P > 0.1 level), animals were judged to have achieved stable baseline levels of choice behavior. After the stability criterion was achieved, rats were provided food ad libitum and 2 days later were subjected to surgery.Rats were anesthetized with 100 mg/kg ketamine hydrochloride and 7 mg/kg xylazine and implanted with bilateral 23-gauge stainless steel guide cannulae into one of the following regions: the medial PFC (prelimbic cortex; anteroposterior [AP] = +3.4 mm; medial–lateral [ML] = ± 0.7 mm from bregma; and dorsoventral [DV] = −2.8 mm from dura); lateral OFC ([AP] = +3.9 mm; [ML] = ± 2.6 mm from bregma; [DV] = −2.9 mm from dura); anterior cingulate ([AP] = +2.0 mm; [ML] = ± 0.7 mm from bregma; and [DV] = −1.0 mm from dura); and agranular insular cortex ([AP] = +2.7 mm; [ML] = ± 3.8 mm from bregma; and [DV] = −3.8 mm from dura). For all surgical preparations, the mouthbar was set to −3.3 mm (flat skull). Thirty-gauge obdurators flush with the end of guide cannulae remained in place until the infusions were made. Rats were given at least 7 days to recover from surgery before testing. During this recovery period, animals were handled at least 5 min each day and food restricted to 85% of their free-feeding weight. Body weights were continued to be monitored daily to ensure a steady weight loss during this recovery period. Rats were subsequently retrained on the task for at least 5 days and until, as a group, they displayed stable levels of choice behavior. For 3 days before the first microinfusion test day, obdurators were removed, and a mock infusion procedure was conducted. Stainless steel injectors were placed in the guide cannulae for 2 min, but no infusion was administered. This procedure habituated rats to the routine of infusions in order to reduce stress on subsequent test days. The day after displaying stable discounting, the group received its first microinfusion test day.A within-subjects design was used for all experiments. Inactivation of PFC regions was achieved by infusion of a drug solution containing the GABAA agonist muscimol (Sigma–Aldrich Canada, Oakville, Ontario, Canada) and the GABAB agonist baclofen (Sigma–Aldrich). Both drugs were dissolved separately in physiological saline at a concentration of 500 ng/μL and then combined in equal volumes, with the final concentration of each compound in solution being 250 ng/μL. Intracranial microinfusions used a volume of 0.5 μL so that the final dose of both baclofen and muscimol was 125 ng per side. These doses were chosen because we have found them to be effective at altering risk discounting when infused in other brain regions (Ghods-Sharifi et al. 2009). In addition, infusions of comparable doses of baclofen and muscimol in the OFC (Takahashi et al. 2009) or muscimol alone into the anterior cingulate (Ragozzino and Rozman 2007) have been reported to disrupt cognition. Note that these doses are substantially higher than doses used in previous experiments that have also reported behavioral effects (e.g., 10 ng; Corrigall et al. 2001). Infusions of gama-aminobutyric acid (GABA) agonists or saline were administered bilaterally into one of the PFC regions via 30-gauge injection cannulae that protruded 0.8 mm past the end of the guide cannulae, at a rate of 0.5 μL/75 s by a microsyringe pump. Injection cannulae were left in place for an additional 1 min to allow for diffusion. Each rat remained in its home cage for another 10-min period before behavioral testing. On the first infusion test day, half of the rats in each group received saline infusions and the other half received muscimol/baclofen. The next day, they received another training day with no infusion. If, for any individual rat, choice of the Large/Risky lever deviated by more than 15% from its preinfusion baseline, it received an additional day of training prior to the second infusion test. On the following day, rats received a second counterbalanced infusion of either saline or muscimol/baclofen.HistologyAfter completion of all behavioral testing, rats were sacrificed in a carbon dioxide chamber. Brains were removed and fixed in a 4% formalin solution. The brains were frozen and sliced in 50-μm sections before being mounted and stained with Cresyl Violet. Placements were verified with reference to the neuroanatomical atlas of Paxinos and Watson (1998). The locations of acceptable infusions for rats used in Experiment 1 are presented in the right panels of Figure 2.Figure 2.Effects of inactivation of different PFC regions on risk discounting. The panels of each figure plots percentage choice of the Large/Risky lever (left) and response latencies (middle) as a function of the trial block (x-axis) following infusions of saline (white circles) or muscimol/baclofen (inactivation, black squares) into each of the 4 PFC subregions: (A) Medial PFC; (B) OFC; (C) anterior cingulate; and (D) insular cortex. Symbols represent mean + standard error of the mean. Black stars denote a significant (P < 0.05) main effect of treatment. White stars denote significant (P < 0.05) differences between treatments during a specific block. Right panels are schematics of coronal sections of the rat brain showing the location of acceptable infusion placements through the rostral–caudal extent of each of the 4 PFC regions.Data AnalysisThe primary dependent measure of interest was the percentage of choices directed toward the Large/Risky lever for each block of free-choice trials factoring in trial omissions. For each block, this was calculated by dividing the number of choices of the Large/Risky lever by the total number of successful trials. The choice data were analyzed using 2-way within-subjects ANOVAs, with Treatment and Trial Block as the within-subjects factors. The main effect of block for the choice data was significant in all experiments (P < 0.05) indicating that rats discounted choice of the Large/Risky lever as the probability of the LR changed across the 4 blocks. This effect will not be mentioned further. Locomotor activity (photobeam breaks) and the number of trial omissions were analyzed with 1-way within-subjects ANOVAs. Response latencies were analyzed using either 2-way or 1-way within-subjects ANOVAs.ResultsMedial PFC (Prelimbic)Initially, 16 rats were trained for this experiment. Two animals died during surgery, and the data from another rat was eliminated due to inaccurate placement. The remaining rats (n = 13) required an average of 35 days of training on the risk-discounting task prior to receiving counterbalanced infusions of saline or muscimol/baclofen into the medial PFC. Analysis of choice behavior following bilateral infusions of muscimol/baclofen or saline into the medial PFC revealed a significant main effect of treatment (F(1,12) = 6.20, P < 0.05; Fig. 2A, left panel). Medial PFC inactivation caused a significant increase in the proportion of choices directed toward the Large/Risky lever relative to saline infusions. Interestingly, inactivation of the medial PFC also significantly increased response latencies but only in the 25% and 12.5% probability blocks, where the relative long-term value of the Large/Risky option was comparable or less than that of the Small/Certain option (treatment × block interaction: F(3,36) = 3.85, P < 0.05, Dunnett's, P < 0.05; Fig. 2A, middle panel). However, this effect was only apparent during free-choice trials, as analysis of the response latency data collected during forced-choice trials revealed no differences between treatment conditions (F(1,12) = 1.89, n.s.). There were no differences in the number of trial omissions or locomotor activity between saline and muscimol/baclofen treatments (all F's < 0.89, n.s.). Thus, under these conditions, inactivation of the medial PFC reduced risk discounting, leading to an apparent increase in risky choice.OFCA total of 16 rats were initially trained for this experiment. Five of these animals had placements that extended ventral to the OFC, and their data were excluded from the analyses. The remaining rats (n = 11) required an average of 28 days of training on the risk-discounting task prior to receiving counterbalanced infusions of saline or muscimol/baclofen into the OFC. Analysis of the choice data revealed no main effect of treatment (F(1,10) = 0.38, n.s) or treatment x block interaction (F(3,30) = 0.43, n.s.; Fig. 2B, left panel). Although bilateral inactivation of the OFC had no effect on choice behavior, this manipulation did alter response latencies, as indicated by a significant treatment × block interaction (F(3,30) = 4.26, P < 0.05; Dunnett's, P < 0.05; Fig. 2B, middle panel). Simple main effects analyses revealed that OFC inactivation did not affect response latencies during the 100% and 50% probability blocks, but significantly increased deliberation times in the latter 2 blocks, relative to saline infusions. As was observed from the medial PFC group, OFC inactivation did not affect response latencies during forced-choice trials (F(1,10) = 3.48, n.s.). There was also a slight decrease in locomotion following OFC inactivation (1580 ± 229) relative to saline treatment (1712 ± 253), but this difference failed to reach significance (F(1,10) = 4.27, P = 0.07). There was no effect of OFC inactivation on trial omissions (F(1,10) = 1.79, n.s.). Thus, inactivation of the OFC did not affect risky choice under conditions where rats were familiar with the relative risk/reward contingencies associated with different response options and changes in these contingencies that occurred over the course of a session.Anterior CingulateThirteen rats were trained for this experiment. The data from 2 rats were excluded due to inaccurate placements and another rat died during surgery. The remaining 10 rats took an average of 33 days of training prior to receiving counterbalanced infusions of saline or muscimol/baclofen into the anterior cingulate. Inactivation of this region caused no discernable change in choice of the Large/Risky lever relative to saline treatment. This observation was confirmed by the lack of a significant main effect of treatment (F(1,9) = 0.00, n.s.) and treatment × block interaction (F(3,27) = 0.31, n.s; Fig. 2C, left panel). Likewise, response latencies during free-choice trials did not differ between treatment conditions (all F's < 1.0, n.s.; Fig. 2C, middle panel). However, during forced-choice trials, rats displayed slightly longer response latencies following anterior cingulate inactivation (1.3 ± 0.1 s), compared with saline infusions (1.0 ± 0.1 s; F(1,9) = 11.26, P < 0.01). Locomotion and trial omissions were not altered significantly by anterior cingulate inactivations (all F's < 1.0, n.s.).Insular CortexThirteen rats were trained for this experiment, but data for 6 rats were excluded due to inaccurate placements. The remaining 7 rats took an average of 28 days of training on the risk-discounting task prior to surgery and receiving counterbalanced infusions of saline or muscimol/baclofen into the agranular insular cortex. Again, we observed no change in choice behavior following muscimol/baclofen infusions, as indicated by the lack of a significant main effect of treatment (F(1,12) = 0.16, n.s), or treatment × block interaction (F(3,18) = 1.67, n.s; Fig. 2D, left panel). Insular inactivation did not significantly affect response latencies during either free-choice (Fig. 2D, middle panel) or forced-choice trials nor did it alter locomotion or trial omissions (all F's < 2.51, n.s.).Medial PFC Inactivation (Ascending Probabilities)Medial PFC inactivation increased selection of the Large/Risky lever when the probabilities of obtaining the larger reward decreased over the course of a risk discounting session. In a subsequent experiment, a separate group of 12 rats were trained for 25 days on a variant of this task, where the probability of obtaining the LR “increased” over the session, prior to receiving counterbalanced infusions of saline or muscimol/baclofen into the medial PFC. In contrast to the effects on choice behavior reported above, inactivation of the medial PFC induced the opposite effect on risk discounting under these conditions, with rats displaying a decreased preference for the Large/Risky lever over the 4 trial blocks. Analysis of these data revealed a significant main effect of treatment (F(1,11) = 12.28, P < 0.05; Fig. 3A). Rats chose the Large/Risky lever less often following medial PFC inactivations relative to saline infusions. There was no significant effect of inactivation on trial omissions or locomotion (F's < 2.5, n.s.). However, there was a significant increase in average response latencies during free-choice trials across all the blocks after medial PFC inactivation (F(1,11) = 8.82, P < 0.05; Fig. 3B). Again, this effect was selective to trials where rats were required to choose between the 2 levers, as response latencies during forced-choice trials did not differ between treatments (F(1,11) = 2.87, n.s.).Figure 3.Effects of medial PFC inactivation on risk discounting with ascending probabilities. (A) Percentage of choice of the Large/Risky lever during free-choice trials following infusions of saline or muscimol/baclofen into the medial PFC for all 12 rats used in this experiment. Inactivation of the medial PFC decreased risky choice. Inset displays data from a subset of rats (n = 9) that demonstrated comparable levels of choice of the Large/Risky lever after both saline and inactivation treatments during the initial, 12.5% probability block (i.e., matching for performance). (B) Response latencies displayed as a function of trial block. Black stars denote a significant (P < 0.05) main effect of treatment. (C) Location of acceptable infusion placements for rats used in this experiment.Inspection of Figure 3A reveals that the decreased preference for the Large/Risky lever after inactivation of the medial PFC was apparent during the first trial block (12.5%) and persisted over the remainder of the session. However, the decreased choice of the Large/Risky lever during the first block was primarily attributable to 3 rats. We conducted a supplemental analysis on the data from the 9 remaining rats that displayed a comparable preference for the Large/Risky lever during the first trial block (i.e., matching for performance) after inactivation and saline treatments. In this instance, the analysis of these data again revealed a significant overall decrease in choice of the Large/Risky lever (F(1,8) = 7.39, P < 0.05; Fig. 3A, inset). Thus, in this subset of rats, even though medial PFC inactivation did not reduce preference for the Large/Risky lever during the first block, these animals were still slower to adjust their choice behavior as the probability of obtaining the larger reward increased over subsequent blocks. This suggests that the medial PFC may facilitate adjustments in preference for the Large/Risky option in response to increases in the probability of obtaining the larger reward.Experiment 2: Effects of Medial PFC Inactivation on Performance of a Within-Session ReversalInactivation of the medial PFC (but not other frontal lobe regions) induced differential effects on risk discounting depending on the manner in which the probabilities of obtaining the larger reward changed over a session. When the odds of the Large/Risky reward were initially 100% and then decreased, medial PFC inactivation increased risky choice, whereas when the probabilities were initially low and subsequently increased, similar inactivations reduced preference for the Large/Risky lever. Both variants of this task require a shift in the direction of responding, whereby rats choose the Large/Risky lever less or more often as the probabilities decrease or increase over the session, respectively. Thus, one potential interpretation of the effects of medial PFC inactivation on risk discounting may be that these manipulations impaired behavioral flexibility, such that rats were perseverating on the lever that they preferred during the first block (Large/Risky or Small/Certain). Notably, similar inactivations of the medial PFC do not impair performance of standard or probabilistic reversal tasks, although they do impair shifting between discrimination strategies (Ragozzino et al. 1999; Block et al. 2007; Ragozzino 2007; Floresco, Block, and Tse 2008). Nevertheless, we conducted a subsequent experiment to determine whether alterations in risk-based decision making induced by medial PFC inactivations were due to perturbations in response flexibility. Rats were trained on a modified within-session reversal task that was similar to the risk-discounting procedure in a number of respects.Materials and MethodsWithin-Session Reversal TaskRats underwent initial lever-pressing training in a manner identical to that described in Experiment 1. They then received daily training sessions consisting of 56 trials, separated into 2 blocks of 28 trials. The 2 blocks were each comprised of 8 forced-choice trials where only one lever was presented (4 trials for each lever, randomized in pairs) permitting animals to learn the amount of food associated with each lever. This was followed by 20 free-choice trials, where both levers were presented (i.e., 40 free-choice trials in total, as in the risk-discounting task). Each 38-min session began in darkness with both levers retracted, and trials began every 40 s with the illumination of the houselight and insertion of one or both levers into the chamber 3 s later. One lever was designated the LR lever, the other the SR lever (counterbalanced left/right). When a lever was chosen, both levers retracted. During the initial block, a press on the LR lever delivered 4 pellets, whereas the SR lever delivered 1 pellet, both with 100% probability, similar to the first block of the risk discounting task. However, during the second, reversal block, selection of the LR resulted in the immediate retraction of both levers, extinguishing of the houselight, and no pellets were delivered. During this block, responses on the SR lever continued to deliver 1 pellet. All other aspects of the task were identical to those described in Experiment 1. Rats were trained with these same contingencies until as a group, they chose the LR lever on at least 80% of successful trials during the first block, less than 20% of successful trials during the reversal block, and the group displayed stable patterns of choice for 3 consecutive days. Subsequently, they were subjected to surgery, and then retrained on the task for at least 5 days prior to receiving intracranial microinfusions.Surgery, Microinfusions, and HistologyRats were implanted with bilateral guide cannulae into the medial PFC as described in Experiment 1. All other procedures were identical to those described in Experiment 1.ResultsEight animals were trained for 18 days prior to receiving counterbalanced infusions of muscimol/baclofen into the medial PFC. One rat had 2 infusions located in 1 hemisphere of the PFC, and its data were excluded from the analyses. Figure 4A displays choice data for the remaining rats (n = 7). After saline infusions, rats showed a strong preference for the LR lever during the initial block when it delivered 4 pellets, compared with the SR lever associated with 1 pellet. During the reversal block, rats shifted their choice away from the LR lever that now delivered 0 pellets, instead choosing the SR lever on the vast majority of free-choice trials, which continued to deliver 1 pellet. This pattern of choice was not altered by inactivation of the medial PFC. Analysis of the choice data revealed a significant main effect of block (F(1,6) = 103.16, P < 0.001) but no significant main effect of treatment (F(1,6) = 0.84, n.s.) or treatment x block interaction (F(1,6) = 2.75, n.s.; Fig. 4A). Interestingly, in this experiment, there were no effects of medial PFC inactivations on response latencies (F(1,6) = 0.01, n.s.). Likewise, locomotion and trial omissions did not differ between treatment conditions (all F's < 0.9, n.s.). The location of acceptable medial PFC infusion placements in this experiment is displayed in Figure 4B. Thus, medial PFC inactivation did not impair the ability to perform a within-session reversal in response to changes in reinforcement contingencies. This suggests that alterations in risk discounting induced by medial PFC inactivation cannot be easily attributed to a general impairment in response flexibility.Figure 4.Medial PFC inactivation does not affect performance of a within-session reversal. Data are plotted as percentage choice of the LR lever over 2 blocks of 20 free-choice trials (initial and reversal). The LR lever was associated with a 4-pellet reward during the initial block, but did not deliver any reward (0 pellets) during the reversal block. Responding on the SR lever delivered 1 pellet during both blocks. Rats were able to shift responding away from the LR lever during the reversal block after both saline and muscimol/baclofen treatment. (B) Location of acceptable infusion placements for rats used in this experiment.Experiment 3: Effects of Inactivation of the Medial PFC on Risk Discounting with Fixed ProbabilitiesAnother potential explanation for the alterations in risk discounting observed in Experiment 1 may be that medial PFC inactivation disrupted monitoring of “changes” in the probability of obtaining the larger reward that occurred within a session. Alternatively, this manipulation may have induced a more fundamental disruption in calculating the relative value of the Large/Risky option within each trial block. To assess this latter hypothesis, a separate group of rats were trained on a simplified version of the task, in which the probability of obtaining the larger reward remained constant throughout a session. Initially, the probability was fixed at 40%, making the relative long-term value of the Large/Risky option (4 pellets @ 40%) higher than the Small/Certain option (1 pellet @ 100%). Rats were then retrained and tested under conditions where the probability of obtaining 4 pellets was set at 10%. In this instance, selection of the Small/Certain option would yield a greater amount of reward in the long term.Materials and MethodsRisk Discounting with Fixed ProbabilitiesEight rats underwent initial lever-pressing training in a manner identical to that described in Experiment 1. They were then trained on the modified risk discounting procedure, consisting of 40 trials, broken down into 20 forced-choice, followed by 20 free-choice trials. Each session took 26 min to complete. All other aspects of the task were identical to the risk-discounting task used in Experiment 1. On each trial, selection of the Large/Risky lever delivered 4 pellets with a 40% probability, whereas pressing the Small/Certain lever delivered 1 pellet with 100% probability. The 40% probability was chosen to maximize the possibility of detecting either increases or decreases in the preference for the Large/Risky lever, given our observation in Experiment 1 that rats tend to select the Large/Risky lever on ∼90% of trials when the probability of obtaining 4 pellets was higher (i.e., 50%; e.g., see Fig. 2A). Rats were trained 6–7 days a week until, as a group, they chose the Large/Risky lever on at least 60% of successful trials and demonstrated stable performance for 3 consecutive days. They were then subjected to surgery, retraining, and their first sequence of intracranial microinfusions.Following the first set of microinfusion test days, rats were retrained on the same task for another 11 days but with the probability of obtaining the larger reward now set to 10%. This probability was chosen because it biases rats to use a risk-aversive strategy but was large enough to ensure that, over 20 forced-choice trials, rats will experience at least some reward so that the probability would not be inferred to be 0%. By the end of this training period, rats as a group chose the Large/Risky lever on less than 40% of the free-choice trials and demonstrated stable performance for 3 consecutive days. They then received a second series of counterbalanced infusions into the medial PFC.Surgery, Microinfusions, and HistologyRats were implanted with bilateral guide cannulae into the medial PFC as described in Experiment 1. All other procedures were identical to those described in Experiment 1.ResultsRats were trained for 24 days prior to receiving the first sequence of counterbalanced infusions of muscimol/baclofen or saline into the medial PFC. Of the initial 8 rats, one died after surgery, and the infusion placement for another was outside the boundary of the prelimbic cortex. Its data were excluded from the analyses. Following saline infusions, the remaining 6 rats displayed a slight preference for the Large/Risky option when the probability of obtaining 4 pellets was 40%, selecting this lever on ∼60% of free-choice trials (Fig. 5A, left panel). Importantly, inactivation of the medial PFC did not alter this preference (F(1,5) = 0.40, n.s.; Fig. 5A left panel). There were also no differences between treatment conditions in terms of locomotion, omissions, or response latencies (all F's < 2.0, n.s.).Figure 5.Inactivation of the medial PFC does not affect risky choice using fixed probabilities associated with larger reward. Bars represent percentage choice of the Large/Risky lever following saline (white) or muscimol/baclofen (black) infusions into the medial PFC. The left panel displays data from infusion tests days when the Large/Risky reward probability was fixed at 40%. The right panel displays data from infusions test days following 11 days of retraining with the Large/Risky probability at 10%. The inset diagrams the significant reduction in preference for the Large/Risky lever averaged across the last 3 days of training prior to the first (40%) and second infusion tests (10%). The star denotes a significant (P < 0.05) difference between reward probability conditions. (B) Location of acceptable infusion placements for rats used in this experiment.Rats were subsequently retrained for another 11 days with the probability of obtaining the Large/Risky reward set at 10%. By the end of this retraining period, rats were selecting the Large/Risky lever less often, compared with their performance when the probability was set at 40% (Fig. 5, inset). Analysis of the choice data obtained from the last 3 baseline days prior to the first (40% chance of the Large/Risky reward) and second (10% chance) infusion tests days confirmed that rats significantly discounted the Large/Risky lever as the probability of reward was reduced (F(1,5) = 7.82, P < 0.05). Again, inactivation of the medial PFC failed to alter choice behavior (F(1,5) = 3.91, n.s.; Fig. 5A right panel), locomotion, omissions, or response latencies relative to saline infusions (all F's<1.9, n.s.). The location of acceptable medial PFC infusion placements in this experiment is displayed in Figure 5B. Thus, inactivation of the medial PFC does not alter choice between small/certain and larger/uncertain rewards when the relative value of larger, probabilistic rewards remains constant during a session.DiscussionThe major finding of the present study is that inactivation of the prelimbic region of the medial PFC induces dramatic effects on risky choice assessed with a probabilistic discounting task. However, the specific effect of medial PFC inactivation was dependent on the manner in which the reward probabilities changed over a session. When the probabilities associated with the Large/Risky option were initially high (100%) and subsequently decreased, medial PFC inactivations increased risky choice. Yet, when the 4-pellet option was initially disadvantageous (12.5%) and then increased in value, similar inactivations had the opposite effect. Subsequent experiments indicated that these effects were not easily attributable to general disruptions in response flexibility or to evaluating the relative value of larger/uncertain versus smaller/certain rewards. In contrast, inactivation of the OFC did not affect risk discounting but did increase response latencies during the latter portion of the session. Similar inactivations of the anterior cingulate or insular cortex did not reliably alter choice behavior. Collectively, these findings indicate that the medial prelimbic PFC plays a critical role in guiding choice between certain and uncertain rewards of different magnitudes in response to changes in reward probability.Involvement of the Rat Medial PFC in Risk DiscountingAt first glance, the increase in risky choice induced by medial PFC inactivation in Experiment 1 may have indicated that this region normally biases choice toward small, certain rewards. However, similar inactivations had the opposite effect when the probabilities of obtaining the Large/Risky reward were initially low and then increased, suggesting that this relatively simple explanation is insufficient to account for our findings. Alternatively, it is possible that these treatments may have induced a general impairment in response flexibility, given that in both experiments, rats persisted in selecting the lever they displayed a bias for during the first probability block. It is notable that following medial PFC inactivations, rats were able to adjust choice of the Large/Risky lever in response to changes in reward probability, albeit at a slower rate than that displayed after saline treatments. Nevertheless, we conducted a subsequent experiment where rats were required to perform a within-session reversal, patterned after the risk-discounting task. Responding on 1 of 2 levers initially delivered a large 4-pellet (LR) or smaller 1-pellet (SR) reward. In the latter part of the session, the LR lever no longer delivered food, but the SR lever continued to deliver 1 pellet. If medial PFC inactivations induced a general tendency to perseverate toward the lever they initially displayed a bias for at the beginning of a session, we would have expected to observe an impaired ability to shift choice between levers after the reversal. However, this effect was not observed, arguing against the notion that the results of Experiment 1 were attributable to impairments in response flexibility. These results are in keeping with numerous other studies showing that inactivation of the prelimbic region of the medial PFC does not affect reversal learning (Ragozzino et al. 1999; Birrell and Brown 2000; Floresco, Block, and Tse 2008).The prelimbic PFC has been implicated in guiding goal-directed behavior in response to changes in the value of rewards linked to particular actions (Balleine and Dickinson 1998; Killcross and Coutureau 2003). Experiment 3 examined whether medial PFC inactivations interfered with choice behavior when the relative long-term value of larger, probabilistic rewards remained constant throughout a session. Here, the probability of obtaining 4 pellets was fixed so that over 20 free-choice trials, this option either had a greater (40% = ∼32 pellets) or lesser (10% = ∼8 pellets) long-term value than the certain, 1 pellet option (20 pellets). Following saline infusions, rats displayed an appropriate bias toward the Large/Risky or Small/Certain lever when the odds of obtaining the larger reward were 40% or 10%, respectively. Importantly, inactivation of medial PFC did not alter choice behavior under either condition. This suggests that the medial PFC is not critical for basic estimations of reward probabilities when animals must evaluate the relative long-term value associated with smaller, certain versus larger, uncertain rewards. Moreover, this lack of effect on choice indicates that the decreased choice of the Large/Risky option following medial PFC inactivations observed in Experiment 1 (i.e., risk discounting with ascending probabilities) cannot be attributed to a general reduction in preference for larger, probabilistic rewards. Rather, it appears that this region plays a more specialized role in biasing choice in response to changes in the probabilities of obtaining larger, uncertain rewards.Further insight into the contribution of the medial PFC to risk-based decision making comes from a discussion of how rats may identify changes in reward probabilities to facilitate risk discounting. Over repeated training, animals learn that changes in the probabilities associated with the Large/Risky reward are signaled by intermittent blocks of forced-choice trials that precede each set of 10 free-choice trials. Thus, rats must remember the outcomes of previous trials and use this information to update their representation of the relative value associated with the Large/Risky lever as a session proceeds. Rats may also use internal temporal cues to estimate the relative value of the Large/Risky option, as they are trained that the likelihood of obtaining 4 pellets early in a session is substantially different from that later in a session (Catania 1970). There is some suggestion that damage to the medial PFC impairs aspects of time perception (Dietrich and Allen 1998; Thorpe et al. 2002) and induces insensitivity to within-session shifts in delays in a delay-discounting task (Cardinal et al. 2001). Thus, in this context, the medial PFC may subserve an “updating function,” whereby mnemonic and temporal information used to identify changes in reward probabilities is integrated to adjust the direction of behavior to maximize reward. Following inactivation of the medial PFC, rats were impaired at using within-session cues to update their choice according to the reward contingencies of the current trial block. Therefore, dysfunction in this region may limit the ability to use internally generated information to update representations of expected reward probabilities based on recently experienced events.The rat medial PFC displays anatomical connectivity homologous to the ventral portion of the anterior cingulate (Area 32) in primates (Uylings and van Eden 1990) and recent studies suggest that this region plays an important role in mediating risk-based decisions in humans. General increases in activity in this region are observed when healthy subjects perform common cost–benefit decision-making tasks (Ernst et al. 2002; Labudda et al. 2008; Lawrence et al. 2009), particularly when they make risky relative to safe choices on the Iowa gambling task (Lawrence et al. 2009). Similarly, using a task where the risk associated with the different choices was “not winning” a given reward (as in the task used in the present study), Smith et al. (2009) found greater activation in the ventral anterior cingulate (Area 32) during risky versus safe choices, as well as during low-probability versus high-probability choices. Moreover, patients with ventromedial PFC lesions that include damage to Area 32 show impaired decision making on the Cambridge Gamble Task (Clark et al. 2008). Yet, the rat medial PFC has also been proposed to share “functional” homology to the dorsolateral PFC of primates (Brown and Bowman 2002; Uylings et al. 2003). For example, the medial and lateral PFCs have been implicated in working memory in rats (Seamans et al. 1995; Floresco et al. 1999), and humans (Owen et al. 1990; Stuss and Alexander 2000), respectively. In humans, dorsolateral PFC activity has been linked to changes in contexts associated with different gains and losses of reward, particularly when outcomes are the opposite of expectations (Akitsuki et al. 2003). Activity in this region is also increased during performance of the Iowa gambling task (Ernst et al. 2002), particularly in the decision phase (Labudda et al. 2008). Likewise, damage to the dorsolateral PFC has been associated with impairments in different forms of risk-based decision making (Manes et al. 2002; Brand et al. 2004; Fellows and Farah 2005). These findings, viewed in light of the present data, suggest that with respect to risk-based decision making, the medial PFC of the rat may share functions that are similar to those mediated by both the ventral anterior cingulate cortex and dorsolateral PFC in humans.The Role of Other PFC Regions in Cost–Benefit Decision MakingInactivation of the OFC did not affect risk discounting, although these manipulations dramatically increased response latencies in a manner similar to that induced by medial PFC inactivation, whereby deliberation times were increased only in the last 2 trial blocks. This effect on response latencies was only apparent during free-choice trials when rats had to choose 1 of the 2 response options. Notably, it is during these blocks that the relative long-term value of the Large/Risky option switched from having a greater value (100% and 50%) to roughly equal (25%) or lesser (12.5%) value than the Small/Certain option. This suggests that although the OFC may not be necessary for making advantageous choices involving risks and rewards under these conditions, neural activity in this region may aid speed of processing about the cost–benefit contingencies associated with each decision following shifts in the relative long-term values of probabilistic rewards. Consistent with these findings, patients with OFC damage also display increased deliberation times when performing certain risk-based decision making tasks (Rogers, Everitt, et al. 1999; Manes et al. 2002; Clark et al. 2008).Our finding that inactivation of the OFC did not alter risk discounting was somewhat surprising, considering previous studies with both rodents and humans that have implicated this region in cost–benefit decision making tasks about risks and rewards (Rogers, Everitt, et al. 1999; Rogers, Owen, et al. 1999; Manes et al. 2002; Clark et al. 2003, 2008; Ernst et al. 2004; Fukui et al. 2005). Lesions of the rat OFC induced prior to training led to a reduced preference for larger probabilistic rewards (2 pellets) relative to a smaller certain reward (1 pellet; Mobini et al. 2002). In that study, the larger reward probability remained constant over a daily session and was decreased systematically every 20–25 days. The “risk-averse” tendencies displayed by OFC lesioned rats were most prominent when the relative long-term value of the larger, probabilistic option (32% or 20%) was lesser than that of the small certain option. This suggests that OFC lesions may impair learning about the relative value of larger, probabilistic rewards. This notion is supported by findings from human neuroimaging studies, where lateral OFC activity is associated with learning risk/reward contingencies of the Iowa Gambling task over a session (Lawrence et al. 2009). In contrast, rats in the present study were trained for 3–4 weeks until they demonstrated stable patterns of risk discounting, implying that they had learned to adjust their choice behavior in response to within-session changes in reward probabilities. Thus, the lateral aspects of the OFC may facilitate the initial learning of risk/reward contingencies, but neural activity in this region may not be required to guide decision making once these contingencies have been acquired.It is notable that patients with OFC damage also display impaired decision making on tasks where subjects are explicitly informed of reward probabilities prior to making a choice (Clark et al. 2008), similar to the task used in the present study, where rats were familiarized with changes in reward probability over many training days. However, in that study, the patient group that showed impairments included those with damage to the medial portions of the OFC. In fact, a lesion control group that included individuals with lateral OFC damage performed similar to healthy controls. Thus, it may be that the medial, rather than the lateral OFC, plays a more prominent role in facilitating risk-based decisions in situations where subjects have prior knowledge about reward probabilities.Damage to the insular cortex in humans has been associated with alterations in risk-related judgments (Bar-On et al. 2003; Clark et al. 2008). These patients are able to make advantageous decisions (i.e., they correctly choose options yielding the most reward), but do not adjust their betting according to gain–loss probability, suggesting that they are less “risk sensitive” than healthy controls. In a similar vein, imaging studies in healthy controls suggest that the insular cortex may have a specific role in using negative outcomes (i.e., punishment) to bias choice behavior to maximize potential reward (O'Doherty et al. 2003; Cohen et al. 2008; Preuschoff et al. 2008). In contrast, we did not observe significant alterations in risk discounting following insular cortex inactivation. Note that the task used in the present study measured whether rats were able to choose the most advantageous option based on changing reward probabilities. However, our procedures did not employ any explicit punishments per se; the risk was a lost opportunity to obtain a reward, which may explain the lack of effect on risk discounting following inactivation of this region. Thus, it would be of particular interest to investigate the role of the rat insular cortex in guiding decision making when animals evaluate potential aversive consequences associated with different options.Inactivation of the dorsal anterior cingulate cortex also did not affect risk discounting, even though infusions of GABA agonists into the adjacent prelimbic cortex did affect choice and response latencies. These 2 regions of the medial PFC receive dissociable patterns of innervation from the mediodorsal thalamus (Conde et al. 1990) and also exhibit different afferent and efferent connectivity with other cortical and subcortical structures (Sesack et al. 1989; McDonald 1991; Heidbreder and Groenewegen 2003). The present findings suggest that despite their close proximity, these 2 cortical regions have dissociable functions. This notion is in keeping with numerous studies demonstrating that lesions or inactivation of the prelimbic versus anterior cingulate cortex induce differential effects on cognitive, executive, and decision-making functions (Seamans et al. 1995; Bussey et al. 1997; Walton et al. 2003).It is interesting to compare the present findings with previous studies on the dissociable contribution of rat PFC subregions to other forms of cost–benefit decision making, such as delay or effort discounting. The OFC appears to play a role in delay discounting, guiding choice between larger, delayed versus smaller, immediate rewards, whereas lesions of the anterior cingulate do not interfere with this form of decision making (Cardinal et al. 2001; Mobini et al. 2002; Winstanley et al. 2004; Rudebeck et al. 2006). Conversely, the anterior cingulate, but not the OFC, is a component of the circuitry underlying effort-related decisions, but not delay discounting (Walton et al. 2003; Rudebeck et al. 2006; Floresco and Ghods-Sharifi 2007; but see Schweimer and Hauber 2005). Notably, medial PFC lesions do not significantly alter cost–benefit judgments involving either delay- or effort-related costs (Cardinal et al. 2001; Walton et al. 2003). Our findings that the prelimbic PFC (but not the OFC or anterior cingulate) contributes to guiding choice between certain and probabilistic rewards further supports the notion that different response costs (e.g., delays, effort, and uncertainty) are processed by separate frontal lobe regions (Rudebeck et al. 2006; Floresco, St Onge, et al. 2008). It is important to highlight that with delay or effort discounting procedures, animals typically obtain “some” reward after each choice, receiving immediate feedback about the costs associated with larger rewards. Imposition of these costs may reduce the perceived value of larger rewards, biasing choice toward smaller but more easily obtainable rewards. On the other hand, risk discounting requires monitoring of rewarded and nonrewarded outcomes, as well as integration of this information over many trials to estimate the long-term value of small/certain versus large/uncertain rewards. In turn, this information may be incorporated in computations about the probability of obtaining the larger reward under conditions where reward probabilities change. Thus, the seemingly selective involvement of the medial PFC in this form of discounting may be linked to its role in the temporal organization of behavior (Fuster 2000), whereby internally generated information is used to keep track of changes in reward probabilities and update value representations that, in turn, may facilitate efficient decision making.FundingCanadian Institutes of Health Research (MOP-89861 to S.B.F., a Michael Smith Foundation for Health Research Senior Scholar); Natural Sciences and Engineering Research Council of Canada and the Michael Smith Foundation for Health Research (to J.R.S.O.).We are grateful to Ying Zhang for her assistance with behavioral testing and surgical procedures and Colin Stopper for his helpful comments on the manuscript. Conflict of Interest: None declared.AkitsukiYSugiuraMWatanabeJYamashitaKYAwataSMatsuokaHMaedaYMatsueYFukudaHKawashimaRContext-dependent cortical activation in response to financial reward and penalty: an event-related fMRI studyNeuroImage20031916741685BalleineBWDickinsonAGoal-directed instrumental action: contingency and incentive learning and their cortical substratesNeuropharmacology199837407419Bar-OnRTranelDDenburgNLBecharaAExploring the neurological substrate of emotional and social intelligenceBrain200312617901800BecharaADamasioARDamasioHAndersonSWInsensitivity to future consequences following damage to human prefrontal cortexCognition199450715BecharaADamasioHDamasioARLeeGPDifferent contributions of the human amygdala and ventromedial prefrontal cortex to decision-makingJ Neurosci19991954735481BecharaADamasioHTranelDAndersonSWDissociation of working memory from decision making within the human prefrontal cortexJ Neurosci199818428437BecksteadRMDomesickVBNautaWJEfferent connections of the substantia nigra and ventral tegmental area in the ratBrain Res1979175191217BirrellJMBrownVJMedial frontal cortex mediates perceptual attentional set shifting in the ratJ Neurosci20002043204324BlairKMarshAAMortonJVythilingamMJonesMMondilloKPineDCDrevetsWCBlairJRChoosing the lesser of two evils, the better of two goods: specifying the roles of ventromedial prefrontal cortex and dorsal anterior cingulate in object choiceJ Neurosci2006261137911386BlockAEDhanjiHThompson-TardifSFFlorescoSBThalamic-prefrontal cortical–ventral striatal circuitry mediates dissociable components of strategy set shiftingCereb Cortex20071716251636BrandMKalbeEKrachtLWRiebelUMunchJKesslerJMarkowitschHJOrganic and psychogenic factors leading to executive dysfunctions in a patient suffering from surgery of a colloid cyst of the foramen of monroNeurocase200410420425BrogJSSalyapongseADeutchAZahmDSThe patterns of afferent innervation of the core and shell in the “accumbens” part of the rat ventral striatum: immunohistochemical detection of retrogradely transported fluoro-goldJ Comp Neurol1993338255278BrownVJBowmanEMRodent models of prefrontal cortical functionTrends Neurosci200225340343BushGVogtBAHolmesJDaleAMGreveDJenikeMARosenBRDorsal anterior cingulate cortex: a role in reward-based decision makingProc Natl Acad Sci U S A200299523528Bussey TJ EverittBJRobbinsTWDissociable effects of cingulate and medial frontal cortex lesions on stimulus-reward learning using a novel Pavlovian autoshaping procedure for the rat: implications for the neurobiology of emotionBehav Neurosci1997111908919CardinalRNHowesNJEffects of lesions of the nucleus accumbens core on choice between small certain rewards and large uncertain rewards in ratsBMC Neurosci200569CardinalRNPennicottDRSugathapalaCLRobbinsTWEverittBJImpulsive choice induced in rats by lesions of the nucleus accumbens coreScience200129224992501CataniaACSchoenfeldWNReinforcement schedules and psychophysical judgment: a study of some temporal properties of behaviorThe theory of reinforcement schedules1970New YorkAppleton (Century/Crofts)142ClarkLBecharaADamasioHAitkenMRSahakianBJRobbinsTWDifferential effects of insular and ventromedial prefrontal cortex lesions on risky decision-makingBrain200813113111322ClarkLCoolsRRobbinsTWThe neuropsychology of ventral prefrontal cortex: decision-making and reversal learningBrain Cogn2004554153ClarkLManesFAntounNSahakianBJRobbinsTWThe contributions of lesion laterality and lesion volume to decision-making impairment following frontal lobe damageNeuropsychologia20034114741483CohenMXElgerCEWeberBAmygdala tractography predicts functional connectivity and learning during feedback-guided decision-makingNeuroimage20083913961407CondeFAudinatEMaire-LepoivreECrepelFAfferent connections of the medial frontal cortex of the rat. A study using retrograde transport of fluorescent dyes. I. thalamic afferentsBrain Res Bull199024341354CorrigallWACoenKMZhangJAdamsonKLGABA mechanisms in the pedunculopontine tegmental nucleus influence particular aspects of nicotine self-administration selectively in the ratPsychopharmacology2001158190197DamasioARDescartes’ error: emotion, reason and the human brain1994New YorkPutnamp. 312DietrichAAllenJDFunctional dissociation of the prefrontal cortex and the hippocampus in timing behaviorBehav Neurosci199811210431047ErnstMBollaKMouratidisMContoreggiCMatochikJAKurianVCadetJLKimesASLondonEDDecision-making in a risk-taking task: a PET studyNeuropsychopharmacology200226682691ErnstMNelsonEEMcClureEBMonkCSMunsonSEshelNZarahnELeibenluftEZametkinATowbinKChoice selection and reward anticipation: an fMRI studyNeuropsychologia20044215851597FellowsLKFarahMJDifferent underlying impairments in decision-making following ventromedial and dorsolateral frontal lobe damage in humansCereb Cortex2005155863FlorescoSBBlockAETseMTInactivation of the medial prefrontal cortex of the rat impairs strategy set-shifting, but not reversal learning, using a novel, automated procedureBehav Brain Res20081908596FlorescoSBBraaksmaDNPhillipsAGThalamic-cortical-striatal circuitry subserves working memory during delayed responding on a radial arm mazeJ Neurosci1999191106111071FlorescoSBGhods-SharifiSAmygdala-prefrontal cortical circuitry regulates effort-based decision makingCereb Cortex200717251260FlorescoSBSt OngeJRGhods-SharifiSWinstanleyCACortico-limbic-striatal circuits subserving different forms of cost-benefit decision makingCogn Affect Behav Neurosci20088375389FlorescoSBTseMTGhods-SharifiSDopaminergic and glutamatergic regulation of effort- and delay-based decision makingNeuropsychopharmacology20083319661979FukuiHMuraiTFukuyamaHHayashiTHanakawaTFunctional activity related to risk anticipation during performance of the Iowa gambling taskNeuroimage200524253259FusterJMExecutive frontal functionsExp Brain Res20001336670Ghods-SharifiSSt. OngeJRFlorescoSBFundamental contribution by the basolateral amygdala to different forms of decision makingJ Neurosci20092952515259HeidbrederCAGroenewegenHJThe medial prefrontal cortex in the rat: evidence for a dorso-ventral distinction based upon functional and anatomical characteristicsNeurosci Biobehav Rev200327555579KillcrossSCoutureauECoordination of actions and habits in the medial prefrontal cortex of ratsCereb Cortex200313400408LabuddaKWoermannFGMertensMPohlmann-EdenBMarkowitschHJBrandMNeural correlates of decision making with explicit information about probabilities and incentives in elderly healthy subjectsExp Brain Res2008187641650LawrenceNSJollantFO'DalyOZelayaFPhillipsMLDistinct roles of prefrontal cortical subregions in the Iowa Gambling TaskCerebral Cortex20091911341143ManesFSahakianBClarkLRogersRAntounNAitkenMRobbinsTDecision-making processes following damage to the prefrontal cortexBrain2002125624639MarshAABlairKSVythilingamMBusisSBlairRJRResponse options and expectations of reward in decision-making: the differential roles of dorsal and rostral anterior cingulate cortexNeuroImage200735979988McDonaldAJOrganization of amygdaloid projections to the prefrontal cortex and associated striatum in the ratNeuroscience199144114MobiniSBodySHoMYBradshawCMSzabadiEDeakinJFAndersonIMEffects of lesions of the orbitofrontal cortex on sensitivity to delayed and probabilistic reinforcementPsychopharmacology (Berl)2002160290298O'DohertyJCritchleyHDeichmannRDolanRJDissociating valence of outcome from behavioral control in human orbital and ventral prefrontal corticesJ Neurosci20032379317939OwenAMDownesJJSahakianBJPolkeyCERobbinsTWPlanning and spatial working memory following frontal lobe lesions in manNeuropsychologia19902810211034Pais-VieiraMLimaDGalhardoVOrbitofrontal cortex lesions disrupt risk assessment in a novel serial decision-making task for ratsNeuroscience2007145225231PaxinosGWatsonCThe rat brain in stereotaxic coordinates19984th ednSan Diego (CA)Academic PressPreuschoffKQuartzSRBossaertsPHuman insula activation reflects risk prediction errors as well as riskJ Neurosci20082827452752RagozzinoMEThe contribution of the medial prefrontal cortex, orbitofrontal cortex, and dorsomedial striatum to behavioral flexibilityAnn N Y Acad Sci20071121355375RagozzinoMEDetrickSKesnerRPInvolvement of the prelimbic–infralimbic areas of the rodent prefrontal cortex in behavioral flexibility for place and response learningJ Neurosci19991945854594RagozzinoMERozmanSThe effect of rat anterior cingulate inactivation on cognitive flexibilityBehav Neurosci2007121698706RogersRDEverittBJBaldacchinoABlackshawAJSwainsonRWynneKBakerNBHunterJCarthyTBookerEDissociable deficits in the decision-making cognition of chronic amphetamine abusers, opiate abusers, patients with focal damage to prefrontal cortex, and tryptophan-depleted normal volunteers: evidence for monoaminergic mechanismsNeuropsychopharmacology199920322339RogersRDOwenAMMiddletonHCWilliamsEJPickardJDSahakianBJRobbinsTWChoosing between small, likely rewards and large, unlikely rewards activates inferior and orbital prefrontal cortexJ Neurosci19991990299038RogersRDRamnaniNMackayCWilsonJLJezzardPCarterCSSmithSMDistinct portions of anterior cingulate cortex and medial prefrontal cortex are activated by reward processing in separable phases of decision-making cognitionBiol Psychiatry200455594602RudebeckPHWaltonMESmythANBannermanDMRushworthMFSeparate neural pathways process different decision costsNat Neurosci2006911611168SchweimerJHauberWInvolvement of the rat anterior cingulate cortex in control of instrumental responses guided by reward expectancyLearn Mem200512334342SeamansJKFlorescoSBPhillipsAGFunctional differences between the prelimbic and anterior cingulate regions of the prefrontal cortexBehav Neurosci199510910631073SesackSRDeutchAYRothRHBunneyBSTopographical organization of the efferent projections of the medial prefrontal cortex in the rat: an anterograde tract-tracing study with Phaseolus vulgaris leucoagglutininJ Comp Neurol1989290213242SmithBWMitchellDGHardinMGJazbecSFridbergDBlairRJErnstMNeural substrates of reward magnitude, probability, and risk during a wheel of fortune decision-making taskNeuroimage200944600609St OngeJRFlorescoSBDopaminergic modulation of risk-based decision makingNeuropsychopharmacology200934681697StussDTAlexanderMPExecutive functions and the frontal lobes: a conceptual viewPsychol Res200063289298TakahashiYKRoeschMRStalnakerTAHaneyRZCaluDJTaylorARBurkeKASchoenbaumGThe orbitofrontal cortex and ventral tegmental area are necessary for learning from unexpected outcomesNeuron200962269280ThorpeCFlorescoSCarrJWilkieDAlterations in time-place learning induced by lesions to the rat medial prefrontal cortexBehav Proc20025987100TomSMFoxCRTrepelCPoldrackRAThe neural basis of loss aversion in decision-making under riskScience2007315515518UylingsHBGroenewegenHJKolbBDo rats have a prefrontal cortex?Behav Brain Res2003146317UylingsHBvan EdenCGQualitative and quantitative comparison of the prefrontal cortex in rat and in primates, including humansProg Brain Res1990853162WaltonMEBannermanDMAlterescuKRushworthMFFunctional specialization within medial frontal cortex of the anterior cingulate for evaluating effort-related decisionsJ Neurosci20032364756479WinstanleyCATheobaldDECardinalRNRobbinsTWContrasting roles of basolateral amygdala and orbitofrontal cortex in impulsive choiceJ Neurosci20042447184722WinstanleyCATheobaldDEDalleyJWRobbinsTWInteractions between serotonin and dopamine in the control of impulsive choice in rats: therapeutic implications for impulse control disordersNeuropsychopharmacology200530669682
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B853D5B870F3D88CB012A22899484E2ECE632AD.txt b/test/dataset/in/resources/corpus/Clean_0B853D5B870F3D88CB012A22899484E2ECE632AD.txt
deleted file mode 100644
index 73c998a..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B853D5B870F3D88CB012A22899484E2ECE632AD.txt
+++ /dev/null
@@ -1 +0,0 @@
-]>NBA5261S0197-4580(99)00016-010.1016/S0197-4580(99)00016-0Elsevier Science Inc.Fig. 1Mean (SEM) body weight values. ∗Significant age effect for data combined across group (ANOVA, p < 0.0001). Tukey pairwise notations for combined data: 25-month and 17-month significantly differ; 317-month and 27-month significantly differ. Sample sizes are listed in parentheses.Fig. 2Morris swim maze performance. The first group of mice received 20 trials (5 days) of hidden platform training (A). The second group received 28 trials (7 days) of hidden platform training (B). ∗Significant age effect (ANOVA, p < 0.05). The 90-s probe trial was subdivided into six 15-s epochs for the first group (C), and for the second group (D). +Significantly different from chance (25%) for data combined across age (one-sample t, p < 0.05). Mean (SEM) cued platform distance for the second group collapsed across 5 trials (1 day) of training (E), and for each training trial (F). Tukey pairwise notations: 15-month and 27-month significantly differ; 25-month and 17-month significantly differ; 317-month and 27-month significantly differ. Sample sizes are listed in parentheses.Fig. 3Psychomotor battery performance. Mean (SEM) tightrope suspension time (A), crossings (B), and transformed score (C) collapsed across three trials. ∗Significant age effect for data combined across group (Brown–Forsythe, p < 0.05). Mean (SEM) square open field activity during a 10-min exposure (D). ∗Significant age effect for data combined across group (ANOVA, p < 0.05). Tukey pairwise notations for combined data: 15-month and 27-month significantly differ; 25-month and 17-month significantly differ; 317-month and 27-month significantly differ. Sample sizes are listed in parentheses.Fig. 4Eye pathology correlation with Morris swim maze performance of 17- and 27-month 129/SvJ mice. Spearman’s rank correlation was not significant for mean distance to find the hidden platform (A) nor mean probe trial performance (B). However, this correlation was slightly significant (p < 0.05) for mean distance traveled to the cued platform (C). Gross eye pathology was ranked on a four-point scale ranging from 0 (no eye pathology) to 3 (severe eye pathology; see text).Table 1Two groups, each comprising three counter-balanced age cohorts of 129/SvJ mice, were subjected to both a psychomotor battery and Morris swim maze. The schedule of behavioral tests is listed along with the day of the test for each groupBehavioral TestGroup 1 (Weeks 1–2)Group 2 (Weeks 8–9)Swim MazePretrainingN/ADay 1 (5 trials)Hidden PlatformDays 1–5 (20 trials)Days 2–8 (28 trials)Probe TrialDay 5 (1 trial 6 epochs)Day 8 (1 trial 6 epochs)Cued PlatformN/ADay 9 (5 trials)Psychomotor BatteryTightropeDay 10 (3 trials)Day 10 (3 trials)Open FieldDay 16 (10-min exposure)Day 11 (10-min exposure)ArticlesAge-related psychomotor and spatial learning deficits in 129/SvJ miceJohn MHengemihleaJeffrey MLongaJenniferBetkeybMathiasJuckercDonald KIngrama*doni@vax.grc.nia.nih.govaLaboratory of Cellular and Molecular Biology, Molecular Physiology and Genetics Section, National Institute on Aging, Intramural Research Program, Gerontology Research Center, Baltimore, MD 21224, USAbDenison University, Granville, OH, USAcDepartment of Neuropathology, Institute of Pathology, University of Basel, Schonbeinstr. 40, CH-4003 Basel, Switzerland*Corresponding author. Tel.: +1-410-558-8180; fax: +1-410-558-8323AbstractThe 129 mouse strain has been widely used to construct mutations that model behavioral aging in humans. The current study found significant age-related declines in both psychomotor and swim maze performance of 5-, 17-, and 27-month-old 129/SvJ mice. However, the age differences in swim maze acquisition were inconsistent with poor performance in the probe trial which assesses spatial memory. This inconsistency may result from the high degree of genetic polymorphisms and age-related visual pathology which afflicts this mouse strain. Therefore, we concluded that 129/SvJ mice present a problematic model of mammalian cognitive aging and involve a risk for behavioral contamination in studies involving mutant mice derived from this strain.KeywordsAgingAnimal behaviorEye pathologyGeneticsMaze learningTransgenic mice1IntroductionThe use of transgenic and knockout murine models has expanded opportunities for examining complex behavioral processes and may be clinically relevant in establishing models of neurodegenerative disease states. Homologous recombination in embryonic stem (ES) cells has allowed the development of chimeric mice with stable, germ line transmission of heritable targeted mutations [28]. Homozygous F2 offspring carrying the mutation can be backcrossed for several generations to create a congenic line on a fixed genetic background suitable for behavioral testing. The behavioral phenotype of these genetically engineered mutant mice is influenced by the mutation but also reflects the interaction of background genes [4], compensatory developmental mechanisms [18], and epistatic interactions between genes [15]. Therefore, before conclusions can be reached about the effect of a targeted mutation on polygenic learning and memory processes, it is important to understand the endogenous behavioral phenotype of the host strain. Because complex behaviors are polygenic, the genetic background of the host strain may affect the behavioral expression of a single gene mutation [28].ES cells isolated from the 129 mouse strain show the highest success rate for germ line integration [22,23]. Consequently this inbred mouse strain has been widely used in the construction of mutant mice, including murine models of Alzheimer’s disease [19,31]. However, relatively little has been published about age-related behavioral changes in this strain that would be useful for assessing the impact of mutant genes. It is possible that the behavioral abnormalities seen in mutant mice derived from 129 ES cells arise from polymorphisms in the mutant 129 genetic background rather than from any exogenous genetic manipulation [4]. Because of potential behavioral contamination from these background or hitchhiker genes [3,5,7,14], there is clearly a need for improved behavioral characterization of wild type 129 mice.To the best of our knowledge, this is the first study of aging in wild type 129/SvJ mice. We attempt to determine the suitability of this inbred strain to studies on behavioral aging. This is important because most previous studies on behavioral aging in mice have used the C57BL/6 strain [9,10,13]. In the current study, 129/SvJ mice of three ages underwent a motor and cognitive assessment, specifically locomotor activity in an open field, strength and coordination in a tightrope test, and spatial learning ability in a Morris water maze. Because previous studies had indicated relatively poor performance of 129/SvJ mice in the water maze [20], we examined whether extended acquisition training (28 versus 20 trials) could enhance performance in this task. We also attempted to ascertain the influence of age-related gross eye pathology on water maze performance of this albino mouse strain [24,25]. Previous studies have also indicated that 129 mice may be especially vulnerable to handling stress [17]. Therefore, we attempted to habituate our mice to handling and thereby overcome some of the known stress related activity confounds of the swim maze, such as passive floating [30] and jumping.2Methods2.1AnimalsExperimentally naive male 129/SvJ mice were obtained from the Jackson Laboratory, Bar Harbor, ME at 5-weeks of age (JAX® Stock Number 000691; now designated as 129XI/SvJ). Over a period of 22 months, three cohorts of mice were acquired and maintained at the Gerontology Research Center (GRC). At the time of behavioral testing, mice were either 5, 17 or 27 months old. Mice were housed in groups of five in a 30 × 19 × 13 cm plastic cage with corncob bedding and had access to food (NIH formula 07) and filtered water ad libitum. Mice were maintained on a 12:12-h light:dark cycle. Lights were turned on at 0700. The room temperature of the vivarium was about 22°C and the relative humidity was about 48%.2.2Behavior testing2.2.1HandlingFor logistical reasons, mice were divided into two counterbalanced groups and behaviorally tested 8 weeks apart. Before testing, mice were systematically handled in an attempt to minimize the previously observed passive floating behavior of this strain in the Morris swim maze (unpublished observation). Each mouse received 10-min of daily handling by the same experimenter. This consisted of picking up each mouse and continuously grooming its fur with brushes of various textures. This was done each day for a period of 2 weeks before behavior testing. To facilitate identification, a sterile AVID® microchip (AVID Inc., Norco, CA USA) was aseptically placed subcutaneously (s.c.) into each mouse before behavior testing.2.2.2Apparatus and procedureThe testing area consisted of a black rectangular table (112 cm × 104 cm) elevated 64 cm from the floor. This table was surrounded by white curtains (183 cm long) and was surveyed by a video camera mounted 122 cm overhead. The area was illuminated with overhead fluorescent lights that provided adequate contrast for the camera. Locomotor activity was monitored using a Videomex® tracking system (Columbus Instruments, Columbus, OH USA) which measured the amount of distance traveled by the mouse during the test period. Before testing, mice were transferred from the vivarium to the testing room 30 min before testing to allow for acclimation. All testing was conducted in the same testing room by the same experimenter between 0730 and 1500 h in the order described in Table 1. 2.2.3Maze trainingThe Morris swim maze is a spatial learning task in which mice must learn to escape by swimming to a concealed, submersed platform within a tank. The maze was a black, circular plastic pool, 80 cm in diameter and 50 cm in height, filled with clear (i.e. non-opaque) tap water to a depth of 25 cm. Water was maintained at 21–22°C. A transparent circular platform, 7 cm in diameter, was fixed in the center of one quadrant of the maze, 1 cm below the water surface. The swim maze was always maintained in the same position in the room as were all visible extra-maze cues. A white cloth curtain was draped around all sides of the maze to allow for adequate video camera tracking of mice within the maze. Previous studies conducted in our laboratory determined that C57BL/6J mice learn the Morris swim maze significantly faster in the presence of multiple three-dimensional cues in close proximity to the maze (unpublished data). It was assumed that these same cues would also benefit 129/SvJ mice. Therefore, seven three-dimensional objects (soda can, toy mouse, floppy disk box, nondescript foam rubber cutout, origami crane, light bulb, rubber stopper) were anchored equidistantly around the top inside wall of the maze and served as cues in this study. The cues varied in shape, size, color, and were no larger than 15 × 15 cm. Because a curtain surrounded the tank, the mice were unable to view the experimenter at any time. Distance traveled (cm) to reach the platform was the dependent variable.As shown in Table 1, the second group of mice (but not the first) were administered five pretraining trials to familiarize them with the presence of a submersed goal and to minimize any potentially confounding effect of nonspatial kinesthetic strategies that the mice may employ to solve the task. During pretraining, all cues were removed, and a Plexiglas® insert was placed into the tank forming a 7-cm wide alley leading to the submerged platform. Each mouse in the second group was given a total of five 90-s trials to locate the platform starting at the opposite end of the alley.For the hidden platform trials, the cues were replaced, and a trial was begun by placing the mouse into a counterbalanced starting quadrant (either N, S, E, or W) facing the wall of the pool. The computer then began recording the distance traveled. Each mouse had a maximum of 90-s to swim to the platform, at which point activity tracking was terminated. When the mouse reached the platform, it was permitted to remain there for 30 s before being removed from the maze. If the mouse failed to find the platform within 90 s, it was placed onto the platform by the experimenter and was removed after 30 s. Following a trial, the mouse was returned to its home cage with free access to food and water between trials. Following completion of a trial for all mice, a 70-min inter-trial interval (ITI) was provided before the start of a new trial. All mice dried themselves off during this rest interval and did not seem to be fatigued. Each mouse received four trials per day; 1 trial per day from each of the four counterbalanced starting quadrants. The platform’s location remained constant throughout training for each mouse. As shown in Table 1, the first group of mice received 20 trials (5 days) of hidden platform training, whereas the second group received 28 trials (7 days). Acquisition was measured as mean distance traveled each day, i.e. each block of four trials, for each group of mice. Thus, a decrease in distance traveled over hidden platform trials implies an improvement in reference memory spatial learning processes.A probe trial was given to all mice following their final trial on the last day of training (see Table 1). The probe trial started from a position directly in the center of the tank and was identical to previously described trials, except that the goal platform was removed. The percent of total distance traveled in the training quadrant (the quadrant where the platform had been located throughout the experiment) was calculated for the entire 90-s trial. The probe trial was divided into six contiguous 15-s epochs, and the percent training quadrant distance was calculated for each epoch in an attempt to access extinction. The probe trial was used to determine the amount of spatial bias or habit strength that mice developed for the location of the platform. An increased spatial bias in the probe trial indicated that the mice had developed a memory for the spatial location of the platform and was an additional indicator of the degree of spatial learning improvement with training.As shown in Table 1, on the day following the probe trial, the second group of mice received five cued platform trials with a 70-min ITI. For this block of trials, the peripheral cues described above were removed. The submerged platform was restored to a quadrant 180° diagonal to its original location and was prominently cued by the insertion of a dowel stick with a green Styrofoam® ball (10-cm diameter) attached 15 cm above the water. The 129/SvJ strain is albino, and therefore subject to visual impairments that may confound a pure spatial learning interpretation of the swim maze task [27]. The cued platform trials can address this problem by serving to identify gross impairments in visual discrimination ability in addition to motoric and motivational impairments all of which are potential confounds of the swim maze.2.2.4TightropeStrength and motor coordination of each mouse was tested by its ability to grasp and to remain suspended from a rope (2 mm in diameter and 50 cm in length). The rope was stretched taut inside a white polyethylene tank (50 cm diameter × 30 cm high) which was filled with water (at 21–22°C) to a depth of 10 cm. During testing, each mouse was kept in a dry holding cage that did not contain any bedding material. Before the test began, the mouse was introduced to the water by holding it by the tail in the water for 5 s. For each of three subsequent trials, the mouse was raised by its tail above the rope and then lowered slowly until it grasped the center of the rope with both front paws. The body of the mouse was then slowly lowered below the rope and released so that the animal had to support its body weight by its grip or fall into the water 20 cm below. The suspension time until a fall into the water was recorded (60-s maximum); the clock was stopped when the mouse entered the water (not when it lost its grip). In addition, the rope had pen markings at 5 cm intervals along its length to measure the amount of horizontal movement (crossings) which occurred while the mouse was suspended (a crossing was counted when the animals leading front paw touched the marking). When the mouse fell into the water, it was immediately removed to the holding cage and allowed to rest for 30-s before the next trial. The suspension time and number of crossings were averaged across three trials. A transformed score was calculated for each mouse: Transformed Score = (mean suspension time) + 10 (mean number of crossings). Thus, an elevated transformed score indicates superior performance either in terms of grip strength or motor coordination or both.2.2.5Open fieldSpontaneous locomotor activity in a novel environment was assessed in this test. A black Plexiglas® square (40 cm wide × 17 cm high) with wire mesh (1.5-cm square openings) attached to one side was used to monitor total distance traveled during a 10-min exposure. The device was elevated 14 cm from the surface of a black table to allow for passage of waste. The apparatus was illuminated from above with fluorescent lights such that the center of the square received approximately 9 foot-candles whereas the corners received 5 foot-candles of light. Mice were allowed to acclimate to the test room for 30 min before testing. Ten seconds after the mouse was placed at the center of the square, distance tracking began and was measured for 10 min. The apparatus was cleaned with 100% ethanol after each mouse was tested to remove both odors and boli. All alcohol vapors were thoroughly removed before the next mouse was tested. Two mice (5 and 17 months old) from the first group died following tightrope testing; open field data were not collected for these subjects.2.3Eye pathologyAt the conclusion of behavior testing, the second group of mice was examined for blepharoconjunctivitis. This ophthalmic condition can affect humans and is observed in several strains of mice, A/HeJ, BALB/cJ, CBA/J, and 129/J [24,25]. It is characterized either by a flaking hyperkeratosis of the eyelids or by excessive production of sebum-like material from the meibomian glands, and the incidence increases with advancing age. To assess the degree of pathology, we examined the eyeball presentation of each mouse and ranked the observations on a four-point scale as follows: 0 = no detectable eye involvement; 1 = mildly swollen eyelid; 2 = moderate or bulging eye; 3 = severe periorbital abscess. No attempt was made to determine the etiology of the ocular pathology nor the relationship between this pathology and visual acuity.2.4Statistical analysisBody weight, tightrope and open field data were first analyzed using a 2(Group) by 3(Age) analysis of variance (ANOVA; BMDP 7D) to determine the appropriateness of combining the two separately tested groups of mice for statistical analysis. If the main effect of group was not significant, data were collapsed across group, and a 3(Age) ANOVA was calculated. If the main effect of age was significant, the Tukey Studentized Range method (with harmonic mean adjustment for unequal sample size) was used to determine individual differences between age groups. If Levene’s test revealed heterogeneity of variance between the groups, the Brown–Forsythe equality of means test [1] was used in lieu of ANOVA because it does not assume that the group variances are equal. Critical values for the BrownForsythe test were obtained from the F distribution with a loss of degrees of freedom to allow for unequal group variances. To determine the degree to which body weight affected tightrope performance, the Pearson correlation coefficient (r) was calculated.Swim maze data for each group were analyzed separately. Hidden platform data for the first group of mice was analyzed using a 3(Age) by 5(Block) ANOVA with repeated measures (ANOVA-RM; BMDP 2V) on the last factor—mean swim distance for each block of four trials. Likewise, hidden platform data for the second group of mice was analyzed using a 3(Age) by 7(Block) ANOVA-RM. Probe trial data for each group of mice were analyzed using a 3(Age) by 6(Epoch) ANOVA-RM on the last factor—percent total distance in the training quadrant for each 15-s epoch. Probe trial data for each group were also analyzed using a one-sample t-test to determine if the observed spatial bias was significantly above chance level (25%). In addition, probe trial data for each quadrant were separately analyzed using a 3(Age) ANOVA collapsed across group. This analysis was conducted to verify selective searching in the training quadrant. Cued platform data for the second group was analyzed using a 3(Age) by 5(Trial) ANOVA-RM on the last factor—swim distance for each cued platform trial. In all cases, significant F tests were followed by pairwise comparisons using the Tukey method. If Levene’s test was significant for any repeated measure, a variance stabilizing square root transformation was conducted.The relation between eye pathology and swim maze performance was determined using the Spearman rank-order correlation (rs; BMDP 3D). This coefficient was selected because exact numerical values could not be assigned to gross eye pathology rankings.3Results3.1Body weightThe body weight of each mouse was taken before the start of behavior testing. Data were collapsed across the two separately tested groups of mice and a 3(Age) ANOVA was computed. As shown in Fig. 1, this analysis revealed a significant main effect of age (F(2,70) = 13.41, p < 0.0001). Pairwise mean comparisons using the Tukey method showed that 17-month-old mice weighed significantly more than either 5- or 27-month-old mice (p < 0.01).3.2Morris swim mazeThe hidden platform data for each group of mice was analyzed separately to ascertain the influence of extended training (see Table 1). Because Levene’s test was significant, a variance stabilizing square root transformation was conducted on both groups of hidden platform data. Fig. 2A presents 20 trials of data for the first group of mice. A 3(Age) by 5(Block) ANOVA-RM showed a significant main effect of age (F(2,33) = 6.56, p < 0.01). Likewise, Fig. 2B presents 28 trials of data for the second group of mice. A 3(Age) by 7(Block) ANOVA-RM revealed a significant main effect of age (F(2,34) = 7.35, p < 0.01). Tukey pairwise mean comparisons were used to access age differences at different time points (see Fig. 2). Both groups of mice demonstrated appreciable learning across trials as evidenced by the significant main effect of block (Group 1: F(4,132) = 15.26, p < 0.0001; Group 2: F(6,204) = 6.99, p < 0.0001). The significant interaction between block and age in each group reflects impaired learning among the 27-month-old mice compared to the 5- and 17-month-old mice in each group (Group 1: F(8,132) = 2.87, p < 0.01; Group 2: F(12,204) = 2.70, p < 0.01). When examining the data collected during extended training provided to the second group of mice, two salient points emerged. First, no age group improved their performance when training was extended from 4 to 7 days. Second, the 27-month-old mice showed significant deterioration in performance during extended training.On their last day of hidden platform training, both groups of mice were administered a 90-s probe trial consisting of six 15-s epochs. Fig. 2C presents probe trial data (spatial bias for training quadrant) for the first group of mice which received 20 trials of hidden platform training; Fig. 2D presents probe trial data for the second group of mice which received 28 trials of hidden platform training. A 3(Age) by 6(Epoch) ANOVA-RM did not reveal a significant main effect of age, epoch nor interaction for neither group. One-sample t-tests revealed that when data were collapsed across age, the first group of mice performed at or slightly below chance level (25%) for each 15-s epoch (see Fig. 2C). However, the same analysis revealed that the second group of mice, which received extended hidden platform training, performed significantly above chance level in all but the first epoch of the probe trial (see Fig. 2D). Thus, despite not showing improved acquisition learning with extended training trials as measured by swim distance, probe trial memory for the location of the platform was improved with extended training, although no significant age differences were evident using percent quadrant distance as the measure of memory. Moreover, the relative probe trial performance was not great considering that levels of 35% to 40% would be clear evidence of a spatial bias. Based on an analysis of distance swam in each quadrant in the combined groups, no significant spatial bias was observed in any maze quadrant. In addition, there were no significant age differences observed in any of the other quadrants. The percent distance scores for other (i.e. non-platform) quadrants were as follows: 5-month group (23.5%, 25.3%, 22.7%), 17-month group (24.2%, 24.4%, 24.8%), 27-month group (25.2%, 23.4%, 27.1%).The second group of mice was given five cued platform trials on the day following the probe trial (see Table 1). As shown in Fig. 2E, when the data were collapsed across all five cued trials, a 3(Age) ANOVA was not significant. As shown in Fig. 2F, when the data were analyzed as a function of training trials, a 3(Age) by 5(Trial) ANOVA-RM did not reveal any significant main effects nor interaction. There was an apparent age-related increase in distance to find the cued platform; however, this trend was not significant.3.3Psychomotor batteryBoth groups of mice were tested on the tightrope task following maze training (see Table 1). Data from the two separately tested groups were combined. Because Levene’s test revealed heterogeneity of variance between groups, the Brown–Forsythe statistic (F∗) was used. As shown in Fig. 3A, a 3(Age) Brown–Forsythe test on the tightrope suspension time did not show a significant main effect of age. However, as shown in Fig. 3B, a 3(Age) Brown–Forsythe test on tightrope crossings revealed a significant main effect of age (F∗ (2,55) = 15.51, p < 0.0001). Pairwise mean comparisons using the Tukey method showed that 5-month-old mice made significantly more crossings than either 17- or 27-month-old mice (p < 0.01). Fig. 3C presents transformed data from the tightrope task for which a 3(Age) Brown–Forsythe test revealed a significant main effect of age (F∗ (2,55) = 4.43, p < 0.05). Tukey pairwise comparisons showed that 5-month-old mice performed significantly better than 17-month-old mice (p < 0.05).Interpretation of the age differences in the tightrope task were complicated to some degree by differences in body weight. Previous studies in our laboratory have shown that heavier mice tend to perform somewhat more poorly on the tightrope task [8,12]. As evidence of possible extraneous variation in the present study, the correlation between body weight and the transformed score for all age groups was significant, r (71) = −0.27, p < 0.05. However, this correlation was significant only within the 27-month-old group, r (16) = −0.48, p < 0.05. The significant decline in the tightrope transformed score between 5- and 17-month-old groups could be related to the heavier body weight of the 17-month-old group. However, as shown in Fig. 1, there was no significant difference in body weight between the 5- and 27-month groups to account for the decline in performance of this task.Both groups of mice were tested in the square open field (see Table 1). Data from the two groups were combined. As shown in Fig. 3D, a 3(Age) ANOVA revealed a significant main effect of age (F(2,68) = 5.50, p < 0.01). Pairwise Tukey comparisons showed that 27-month-old mice were significantly less active than either 17- (p < 0.05) or 5-month-old mice (p < 0.01).3.4Eye pathologyThe second group of mice was examined for gross eye pathology at the completion of behavior testing. Only 6% (1:18) of the 5-month-old mice were observed to have moderate to severe eye pathology suggesting that this cohort was relatively healthy. In contrast, 46% (6:13) of the 17- and 67% (4:6) of the 27-month-old mice were observed to have an eye pathology ranking of 2 or 3. Because the 5-month-old cohort was relatively healthy, their data were culled from the eye pathology correlation; only data from the 17- and 27-month-old cohorts were included because these mice presented most of the overt eye pathology. As shown in Fig. 4A and B, the Spearman rank correlation with eye pathology was not significant for hidden platform distance (mean of first 20 trials) nor probe trial performance (mean of 6 epochs). However, as shown in Fig. 4C, a significant Spearman correlation was found between eye pathology and distance traveled to the cued platform (mean of 5 trials; rs (17) = 0.46, p < 0.05).4DiscussionEvidence of behavioral aging in male 129/SvJ mice was observed in tests of psychomotor performance and spatial learning in this cross-sectional study. Specifically, age was associated with diminished strength and motor coordination as measured by tightrope performance and decreased spontaneous locomotor activity measured in the open field. Significant age-related deficits in the water maze were evident in this strain of mice during the acquisition trials measured as distance to platform; however, there were no significant age differences in probe trial performance. The later observation would confound any conclusions about age differences in spatial memory abilities, despite the age differences in learning in this task.The behavioral phenotype of 129 mice is continuing to emerge in the literature. Previous studies have reported hypoactivity of 129/SvJ mice in the open field [15,17,18]. However, this robust finding was not reproduced in the current study. When compared to unpublished observations in 6-month-old male C57BL/6J mice from our own laboratory, the 5-month-old male 129/SvJ mice of the current study were found to be significantly (18%) more active in the open field. One previous study [18] using the tightrope task found no difference between C57BL/6J and 129/Sv mice. In contrast, the 5-month-old male 129/SvJ mice from the current study were found to demonstrate a significantly (36%) lower transformed score in the tightrope task when compared to 6-month-old male C57BL/6J mice (unpublished data). Methodological differences, such as the handling, may account for these discrepancies. However, it is more likely that the mixed genetic background of the 129/SvJ strain [26] accounts for many of these discrepancies with the previous psychomotor literature on 129/SvJ mice.The Morris swim maze is heavily dependent upon visual, motor, and cognitive abilities. As such, interpretation of a cognitive deficit in the current study is problematic from several perspectives. First, the observed age-related decline in psychomotor performance of this strain may confound a pure spatial deficit interpretation of the swim maze results. Second, although learning was evident in the two youngest groups, the aged group showed no significant improvement across sessions. Third, the level of performance achieved over 20 trials in 129/SvJ mice was substantially less than we have observed in young C57BL/6NIA [11] and C57BL/6J [2] mice as measured by distance traveled. Fourth, performance in the probe trial was poor for all three age groups and did not differ significantly. There was no evidence of spatial bias to indicate use of a spatial strategy based on utilization of visual cues. Although no significant age effect was found in the cued platform test, it was clear that many older mice seemed impaired as shown by the great variability. This variability was correlated with the degree of eye pathology. Over 52% (10:19) of the 17- and 27-month-old mice presented with overt eye symptoms, suggestive of blepharoconjunctivitis [24,25]. It is possible that the high incidence of age-related visual pathology in these mice may cause a visual acuity deficit resulting in a potentially confounding measurement error [9]. As the severity of the pathology increased, performance in the cued trials diminished (Fig. 4C). However, no such correlation was observed in the hidden platform task (Fig. 4A) or in the probe trial (Fig. 4B). These latter observations would indicate that the mice were not utilizing available visual cues to direct their performance in the swim maze. In summary, interpreting the observed age-related differences in spatial learning ability is very problematic to the extent that these learning deficits may not be a function of normal aging in 129/SvJ mice [9].As noted previously, the 5-month-old 129/SvJ mice in the current study presented a very low incidence of overt visual pathology. Therefore, their swim maze behavior was not likely to be influenced by the measurement errors discussed above. As such, the current study supports previous findings which showed young 129/SvJ mice achieved an inferior level of asymptotic performance in the swim maze relative to pigmented strains of mice [20]. Although 5-month-old male 129/SvJ mice from the current study were able to consistently find the hidden platform, this asymptote was only achieved with a highly significant (98%) increase in distance traveled compared to 7-month-old male C57BL/6NIA mice from our laboratory [11]. The current study demonstrated that without extended hidden platform training, 129/SvJ mice lack development of appreciable spatial bias as measured in the probe trial (see Fig. 2C); this finding replicates negative probe trial results of a previous study in 129/SvJ mice [20]. It is possible that the percent training quadrant distance measure frequently used in the probe trial is not a sensitive measure of spatial bias in 129/SvJ mice. Annulus crossings seems to be a better choice because another study [18] was able to demonstrate significant probe trial spatial bias in 129/Sv mice using that measure. The current study reports two manipulations which taken together were partially successful in improving probe trial performance. First, extended hidden platform training was able to improve spatial bias above chance performance (Fig. 2D). Second, our handling manipulation was apparently successful in reducing stress in that neither passive floating nor jumping behaviors were observed in any of our mice in the swim maze. This observation is significant because previous studies have shown that handling reduces stress and thereby improves spatial learning in mice [6]. Despite these relative improvements in spatial bias, however, probe trail performance did not vary significantly across age.Previous studies using aged rats have shown that a shortened ITI can impair water maze performance due to fatigue [21]. Therefore, we were careful to space the trials out in an attempt to avoid fatigue in our aged subjects. Absence of fatigue was evidenced by the fact that all the mice completely groomed themselves to dryness between trials and did not shiver. Moreover, the distance measure used in the current study to assess acquisition is more robust to fatigue-related activity confounds than the latency measure used in previous studies [18,20]. These previous studies have relied on distal room cues for navigation, however, in the current study proximal cues were placed inside the rim of tank. Because it has been suggested that mice are less adapted to swimming tasks than other rodent species [29], it is possible that this proximal location of cues favors the development of a spatial rather than kinesthetic strategy to solve the task.In conclusion, we feel that the high degree of documented genetic polymorphisms [26] and possible ancestral contamination [22] makes the 129/SvJ mouse strain generally unsuitable to studies of behavioral aging. Although we feel that the 129/SvJ strain should be avoided in experiments assessing age-related effects on behavior, we appreciate the value of this strain in the production of mutant mice. The strategy of maintaining the targeted mutation heterozygous (by backcrossing chimeras into a common inbred strain such as C57BL/6J) and then testing the homozygous mutant in an F1 cross is sound. Because F1s are used, specific deficits in the 129 strain influenced by deleterious homozygous genes will be attenuated. In addition, the current study reports visual pathology in aged 129/SvJ mice severe enough to potentially confound swim maze studies of aged mutant mice derived from 129/SvJ ES cells. Future studies using conditional mutants may avoid potential behavioral problems resulting from genetic variation and contamination [16].AcknowledgementsThe Gerontology Research Center (NIH-NIA-IRP) is fully accredited by the American Association for the Accreditation of Laboratory Animal Care. The authors wish to thank: Dr. Robert Meyer and Edward Spangler for critical review of the manuscript; Keith Staton and Richard Zichos for construction of equipment; Carol Lindsay and Kim Miller for editorial assistance.References[1]M.B.BrownA.B.ForsytheThe small sample behavior of some statistics which test the equality of several meansTechnometrics161974129132[2]M.E.CalhounD.KurthA.L.PhinneyJ.M.LongJ.HengemihleP.R.MoutonD.K.IngramM.JuckerHippocampal neuron and synaptophysin-positive bouton number in aging C57BL/6 miceNeurobiol Aging191998599606[3]J.N.CrawleyUnusual behavioral phenotypes of inbred mouse strainsTrends Neurosci1951996181182[4]J.N.CrawleyJ.K.BelknapA.CollinsBehavioral phenotypes of inbred mouse strainsimplications and recommendations for molecular studiesPsychopharmacology (Berl)13221997107124[5]W.E.CrusioGene-targeting studiesnew methods, old problemsTrends Neurosci1951996186187[6]D.E.FordyceJ.M.WehnerPhysical activity enhances spatial learning performance with an associated alteration in hippocampal protein kinase C activity in C57BL/6 and DBA/2 miceBrain Res6191–21993111119[7]R.GerlaiGene-targeting studies of mammalian behavioris it the mutation or the background genotype?Trends Neurosci1951996177181[8]D.K.IngramToward the behavioral assessment of biological aging in the laboratory mouseconcepts, terminology, and objectivesExp Aging Res941983225238[9]D.K.IngramMotor performance variability during aging in rodents. Assessment of reliability and validity of individual differencesAnn N Y Acad Sci51519887096[10]D.K.IngramPerspectives on genetic variability in behavioral aging of miceD.E.HarrisonGenetic effects on aging II1990The Telford Press IncCaldwell, NJ205231[11]Ingram DK, Jucker M. Developing mouse models of aging: a consideration of strain differences in age-related behavioral and neural parameters. Neurobiol. Aging, in press.[12]D.K.IngramM.A.ReynoldsAssessing the predictive validity of psychomotor tests as measures of biological age in miceExp Aging Res1231986155162[13]D.K.IngramR.WeindruchE.L.SpanglerJ.R.FreemanR.L.WalfordDietary restriction benefits learning and motor performance of aged miceJ Gerontol42119877881[14]R.LatheMice, gene targeting and behaviormore than just genetic backgroundTrends Neurosci1951996183186[15]S.F.LogueE.H.OwenD.L.RasmussenJ.M.WehnerAssessment of locomotor activity, acoustic and tactile startle, and prepulse inhibition of startle in inbred mouse strains and F1 hybridsimplications of genetic background for single gene and quantitative trait loci analysesNeuroscience804199710751086[16]M.MayfordI.M.MansuyR.U.MullerE.R.KandelMemory and behaviora second generation of genetically modified miceCurr Biol791997R580R589[17]L.L.MinerCocaine reward and locomotor activity in C57BL/6J and 129/SvJ inbred mice and their F1 crossPharmacol Biochem Behav58119972530[18]A.MontkowskiM.PoettigA.MedererF.HolsboerBehavioral per-formance in three substrains of mouse strain 129Brain Res7621–219971218[19]U.MullerN.CristianaZ.W.LiBehavioral and anatomical deficits in mice homozygous for a modified β-amyloid precursor protein geneCell7951994755765[20]E.H.OwenS.F.LogueD.L.RasmussenJ.M.WehnerAssessment of learning by the Morris water task and fear conditioning in inbred mouse strains and F1 hybridsimplications of genetic background for single gene mutations and quantitative trait loci analysesNeuroscience804199710871099[21]J.T.RickM.P.MurphyG.O.IvyN.W.MilgramShort intertrial intervals impair water maze performance in old Fischer 344 ratsJ Gerontol Biol Sci51A41996B253B260[22]E.M.SimpsonC.C.LinderE.E.SargentM.T.DavissonL.E.MobraatenJ.J.SharpGenetic variation among 129 substrains and its importance for targeted mutagenesis in miceNature Genet16119971927[23]A.G.SmithMouse embryo stem cellstheir identification, propagation and manipulationCell Biol361992385399[24]R.S.SmithX.MontagutelliJ.P.SundbergUlcerative blepharitis in aging inbred miceU.MohrD.L.DungworthC.C.CapenW.W.CarltonJ.P.SundbergJ.M.WardPathobiology of the aging mouse1996International Life Sciences Institute PressWashington, DC131138[25]J.P.SundbergK.S.BrownR.BatesT.L.Cunliffe–BeamerH.BedigianSuppurative conjunctivitis and ulcerative blepharitis in 129/J miceLab Anim Sci4151991516518[26]D.W.ThreadgillD.YeeA.MartinJ.H.NadeauT.MagnusonGenealogy of the 129 inbred strains129/SvJ is a contaminated inbred strainMamm Genome861997390393[27]M.UpchurchJ.M.WehnerDifferences between inbred strains of mice in Morris water maze performanceBehav Genet18119885568[28]J.M.WehnerB.J.BowersR.PaylorThe use of null mutant mice to study complex learning and memory processesBehav Genet2631996301312[29]I.Q.WhishawA comparison of rats and mice in a swimming pool place task and matching to place tasksome surprising differencesPhysiol Behav5841995687693[30]D.P.WolferM.Bozicevic–StagliarH–P.LippThe role of genetic background in assessing the behavior of transgenic miceSoc Neurosci Abstr21480.1519951227[31]D.P.WolferU.MullerM.StagliarH–P.LippAssessing the effects of the 129/Sv genetic background on swimming navigation learning in transgenic mutantsa study using mice with a modified β-amyloid precursor protein geneBrain Res77111997113
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B87791589602536E912B35C9D930D15FE291C4D.txt b/test/dataset/in/resources/corpus/Clean_0B87791589602536E912B35C9D930D15FE291C4D.txt
deleted file mode 100644
index ff92245..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B87791589602536E912B35C9D930D15FE291C4D.txt
+++ /dev/null
@@ -1 +0,0 @@
-jexbotexbotjJournal of Experimental Botany1460-24310022-0957Oxford University Press10.1093/jxb/ern330Research PapersDifferent ways to die: cell death modes of the unicellular chlorophyte Dunaliella viridis exposed to various environmental stresses are mediated by the caspase-like activity DEVDaseJiménezCarlos1CapassoJuan M.2EdelsteinCharles L.2RivardChristopher J.2LuciaScott3BreusegemSophia2BerlTomás2SegoviaMaría1*1Department of Ecology, Faculty of Sciences, University of Málaga, Bvd. Louis Pasteur s/n, E-29071 Málaga, Spain2Department of Renal Diseases and Hypertension, School of Medicine, University of Colorado Health Sciences Center, 4200 E. 9th Av. Denver, CO 80262, USA3Department of Pathology, School of Medicine, University of Colorado Health Sciences Center, 4200 E. 9th Av. Denver, CO 80262, USA*To whom correspondence should be addressed: E-mail: segovia@uma.es320096038158284720082011200824112008© 2009 The Author(s).2009This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.This paper is available online free of all access charges (see http://jxb.oxfordjournals.org/open_access.html for further details)Programmed cell death is necessary for homeostasis in multicellular organisms and it is also widely recognized to occur in unicellular organisms. However, the mechanisms through which it occurs in unicells, and the enzymes involved within the final response is still the subject of heated debate. It is shown here that exposure of the unicellular microalga Dunaliella viridis to several environmental stresses, induced different cell death morphotypes, depending on the stimulus received. Senescent cells demonstrated classical and unambiguous apoptotic-like characteristics such as chromatin condensation, DNA fragmentation, intact organelles, and blebbing of the cell membrane. Acute heat shock caused general swelling and altered plasma membrane, but the presence of chromatin clusters and DNA strand breaks suggested a necrotic-like event. UV irradiated cells presented changes typical for necrosis, together with apoptotic characteristics resembling an intermediate cell-death phenotype termed aponecrosis-like. Cells subjected to hyperosmotic shock revealed chromatin spotting without DNA fragmentation, and extensive cytoplasmic swelling and vacuolization, comparable to a paraptotic-like cell death phenotype. Nitrogen-starved cells showed pyknosis, blebbing, and cytoplasmic consumption, indicating a similarity to autophagic/vacuolar-like cell death. The caspase-like activity DEVDase was measured by using the fluorescent substrate Ac-DEVD-AMC and antibodies against the human caspase-3 active enzyme cross-reacted with bands, the intensity of which paralleled the activity. All the environmental stresses tested produced a substantial increase in both DEVDase activity and protein levels. The irreversible caspase-3 inhibitor Z-DEVD-FMK completely inhibited the enzymatic activity whereas serine and aspartyl proteases inhibitors did not. These results show that cell death in D. viridis does not conform to a single pattern and that environmental stimuli may produce different types of cell death depending on the type and intensity of the stimulus, all of which help to understand the cell death-dependent and cell death-independent functions of caspase-like proteins. Hence, these data support the theory that alternative, non-apoptotic programmed cell death (PCDs), exist either in parallel or in an independent manner with apoptosis and were already present in single-celled organisms that evolved some 1.2-1.6 billion years ago.Aponecrosis-likeapoptosis-likeautophagic/vacuolar cell deathcaspase-likecell deathDEVDase activityDunaliellaenvironmental stressmicroalgaenecrosis-likeparaptosis-likephytoplanktonTUNELIntroductionWhether cell death is beneficial or detrimental for the organisms depends on the biological circumstances in which this happens and on the cell type or species. For both, the question arises concerning the significance of programmed cell death (PCD) in the biology of multi- and unicellular organisms. The origin of the capacity for self-destruction may be as ancient as the origin of the very first cell (Ameisen, 1998). Thus, if effectors of the cell survival machinery can also be effectors of the self-destruction of the cell in which they operate, then the requirement for coupling cell survival to the prevention of self-destruction may be as old as the origin of the first cell, and cell suicide is therefore an unavoidable consequence of self-organization. PCD is ubiquitous in multicellular systems and is essential for normal growth and development (Leist and Nicotera, 1997). The PCD type known as apoptosis, is characterized by very specific morphological and biochemical requirements such as chromatin condensation and margination, ordered DNA cleavage while the cytoplasm and organelles remain unchanged, and the participation of a family of cysteine proteases named caspases, as central regulators (Kerr et al., 1972). Apoptosis is clearly different from necrosis which is characterized by a loss of membrane integrity, cell swelling, and lyses. However, the classical necrotic nature described as passive, unprogrammed cell death, can be questioned. There are examples of cells presenting necrotic morphologies that are subjected to an active cell death programme called ‘necrotic-like’-PCD that can be either caspase-dependent or -independent (Edelstein et al., 1999; Kitanaka and Kuchino, 1999). Consistently, it is considered a type of PCD due to the presence of underlying regulatory mechanisms. However, other types of PCDs, which do not completely fulfil either the typical apoptotic features or the necrotic ones, have also been reported (Clarke, 1990; Sperandio et al., 2000; Leist and Jäätelä, 2001; Papucci et al., 2004; Elmore, 2007). In this regard, a cell death type named ‘paraptosis’, which is fundamentally different from apoptosis, was discovered. It involves cytoplasmic vacuolization, mitochondrial swelling, and the absence of caspase activation or typical nuclear changes, including pyknosis and DNA fragmentation, and is mediated by mitogen-activated protein kinases (MAPKs) (Sperandio et al., 2000, 2004). Other cells have been shown to exhibit features of both apoptosis and necrosis and the term ‘aponecrosis’ was coined (Formigli et al., 2000; Papucci et al., 2004). These authors suggested that apoptosis and necrosis represent two extremes of a wide range of aponecrotic responses that operate under caspase activation. One distinct model is ‘autophagic/vacuolar’-like cell death. The typical hallmark of this form of cell death is the consumption of the cytoplasm in the absence of leakage of the intracellular content and intact cell membranes. This type of cell death may display nuclear degradation and pyknosis, depending on the nature of the cell (Jones, 2000) and it has been shown to be a lysosomal degradation pathway, since final evidence for lysosome-like organelles in plants has been reported (Swanson et al., 1998).Among all the cell death classes, apoptosis has been mainly studied in metazoans, but vascular plants, as well as some unicellular eukaryotic organisms, show characteristic apoptotic-like features. Apoptosis-like phenomena occur in vascular plants (Greenberg, 1996; Pennell and Lamb, 1997; Lam and del Pozo, 2000; Lam et al., 2001) and they have also been portrayed in unicellular organisms, including chlorophytes (Berges and Falkowski, 1998; Segovia et al., 2003; Segovia and Berges, 2005), dinoflagellates (Vardi et al., 1999; Dunn et al., 2004; Franklin and Berges, 2004; Segovia, 2007), diatoms (Casotti et al., 2005), yeast (Frohlich and Madeo, 2000), kinetoplastids and slime moulds (Cornillon et al., 1994), and bacteria (Lewis, 2000), including cyanobacteria (Berman-Frank et al., 2004). Berges and Falkowski (1998) described a form of autocatalysed cell death in the single cell algae D. tertiolecta. When this chlorophyte was deprived of light, it underwent apoptotic cell death (Segovia et al., 2003). During the process of cell dismantling in D. tertiolecta, the DNA suffered fragmentation and the nucleus disintegrated while the cytoplasm and organelles remained intact. In parallel, both caspase-like activity and expression increased. Despite the identification of metacaspases in vascular plants and protists their roles are not clear yet (Vercammen et al., 2007; Deponte, 2008) and neither is the nature of caspase-like activities in vascular plants (Bonneau et al., 2008). In the majority of cases, measurements of these activities in unicellular species have been carried out by using the classical aspartate-containing caspase substrates. Consequently, the activities measured must be ‘caspase-like’ activities. Although data are scarce, modes of cell death different from apoptosis are adopted by other phytoplanktonic species under stress conditions (Dunn et al., 2002, 2004; Franklin and Berges, 2004). The factors that cause cell death in unicells, their roles, and the details of the apoptotic process remain unclear and, nowadays, the existence and evolution of PCD in unicellular organisms is still controversial (Deponte, 2008). Hence evidence that these alternative, non-apoptotic, PCDs exist has important implications for understanding the fate of unicellular organisms.Microalgae from the genus Dunaliella are among the most ubiquitous eukaryotic organisms in hypersaline environments, and often the major primary producers in salt lakes and in the evaporation ponds of salt works (Borowitzka, 1981). As an adaptation to the strong environmental seasonal changes operating in these systems, they show a remarkable degree of acclimation to salinity, temperature, nitrogen, and irradiance (Ginzburg, 1987). These features make this species a perfect candidate as a model organism.Exposure of D. viridis to environmental stresses that impair cell division such as hyperosmotic shock, UV radiation, heat shock, and nutrient starvation, causes a marked decrease in the phosphorylated form of an extracellular signal-regulated kinase (ERK), known to be involved in cell proliferation and differentiation in mammalian cells, through protein kinase cascades (Jimenez et al., 2007). The authors had formerly demonstrated that ERK phosphorylation was critical for cell division in D. viridis (Jiménez et al., 2004). Cell numbers and viability of D. viridis cultures under these conditions were tested and suffered no changes when exposed to sub-lethal stress conditions.In the present study, evidence is presented that D. viridis has the capacity to undergo different modes of cell death depending on the stress factor and on its intensity. The existence of several biochemical and morphological features typical of each cell death-like morphotype in this microalga is further demonstrated. In all the cases, cell death in D. viridis was linked to an increase in the caspase-like activity DEVDase that matched their accumulation.Materials and methodsCulture conditionsDunaliella viridis Teodoresco was grown in Johnson et al. (1968) medium, at 2 M NaCl (Jiménez et al., 2004) and 25 °C under continuous photosynthetic active radiation (PAR) (400–700 nm) at 150 μmol quanta m−2 s−1, while maintaining gentle stirring and bubbling with filtered air. When cultures reached mid log-phase they were submitted to the following environmental stress treatments: (i) osmotic shock, (ii) UV radiation, (iii) heat shock, (iv) nitrogen starvation, and (v) cells were left to reach late stationary phase (senescence). All experiments were performed in triplicate and each sample was measured by triplicate.Stress treatmentsOsmotic shock:Osmotic stress consisted of an increase in the osmotic pressure of the medium by the addition of NaCl. In acute hypertonic stress experiments, a final concentration of 4 M was considered non-lethal while 5.5 M NaCl was lethal. Cells were sampled at 0, 0.5, 1, and 2 h after the shock.UV radiation (UVR):Cultures were placed in 14 cm diameter Petri dishes transparent to UVR and exposed to either 40 (non-lethal) or 70 mJ cm−2 (lethal) of UVR in the range 200–400 nm using a GS Gene Linker UV chamber (Bio-Rad). The UVR spectral band is divided into three sub-bands corresponding to UV-C (<280 nm), UV-B (280–315 nm), and UV-A (315–400 nm). After UV exposure, cultures were maintained with continuous orbital shaking and 150 μmol m−2 s−1 PAR. Cultures were sampled at 0, 2, 4, and 6 h after UV radiation.Heat shock:For thermal stress, cultures were transferred to 50 ml conical tubes and submerged in a water bath at either 35 °C (non-lethal) or 40 °C (lethal). Continuous light was provided at the same irradiance as the PAR used during culture growth. Cultures were sampled at 0, 1, 2, 3, and 4 h after the heat shock.Nitrogen starvation:Cells were harvested by centrifuging the cultures at 1500 g for 10 min and resuspending them in nitrogen-free growth medium for 7 d. Control cultures were maintained with normal nitrogen-containing medium under the same conditions. Cultures were sampled at 0, 2, 5, and 7 d.Senescence:Cells were left to grow under continuous PAR, at the same irradiance as above, for 12 d until they reached late stationary phase. Sampling took place at 0, 2, 5, 7, 9, and 12 d after inoculation.Transmission electron microscopy (TEM)Cell pellets of the last point of the time-course for each treatment were used for morphological analysis. For that, such pellets were fixed in 2% glutaraldehyde for 1 h, washed and resuspended in 1 ml of 0.01 M phosphate buffer (pH 8). The pellets were post-fixed in 1% buffered osmium tetroxide followed by dehydration in a graded series of ethanol and embedded in epoxy resin. Ultra-thin sections were viewed and photographed on a Philips EM 201 electron microscope. TEM images of all the treatments were examined at ×3750, ×11 750, and ×25 000. Representative pictures (×25 000) under the different stress conditions are presented. Quantification of the cells under the TEM is always difficult, therefore counting of cells showing each different morphotype was carried out by analysing three fields of view (FOV) for each treatment under the smallest magnification.Dead cells stainingEvans Blue is an acidic dye which has the inverse staining properties of a vital stain when determining the survival of plant and planktonic cells. Cells with intact semi-permeable membranes exclude the dye, whereas the dye penetrates and stains dead cells. Therefore, Evans Blue is referred as a mortal stain rather than a vital stain (Crippen and Perrier, 1974). Since this method does not give an indication of the mode of cell death, it should only be used in conjunction with other techniques. Accordingly, the method described by Crippen and Perrier (1974) was applied for marine plankton. Samples were measured in triplicate independent cultures for each treatment. Two aliquots of 1 ml of each culture were taken. One of them was preserved in Lugols iodine for counting the total number of cells. The other, was used for counting the number of Evans Blue-stained cells, by using a final concentration of 1:2000 (w/v) of Evans Blue stock 1% (w/v). Samples were counted under the microscope using a haemocytometer.Confocal laser microscopy (CLM)DAPI is a popular nuclear counterstain for use in multicolour fluorescent techniques. Its blue fluorescence stands out in vivid contrast to green, yellow, or red fluorescent probes of other structures and stains nuclei specifically, with little or no cytoplasmic labelling. DAPI (Invitrogen) was added to a concentration of 1–10 μM and incubated for 5 min at room temperature. A small droplet of algae suspension was placed on a pre-cleaned glass slide (VWR, Aurora, CO) and trapped under a coverslip. Algae were imaged on an inverted laser scanning confocal microscope (LSM 510, Carl Zeiss Inc., Thornwood, NY) through a ×40 1.2 N.A. water immersion objective. DAPI and endogenous chlorophyll fluorescence were simultaneously excited using two-photon excitation at 780 nm from a Ti: sapphire laser (Coherent Inc., Santa Clara, CA). A dichroic filter (HFT KP 700/543) was used to split the fluorescence emission into two channels. DAPI fluorescence was observed through a 435–485 nm band pass filter and chlorophyll fluorescence through a 650–710 nm bandpass filter. More than 100 cells from each condition were examined under a ×40 objective and representative pictures were taken.TUNEL stainingNuclear DNA fragmentation was identified in situ by TUNEL labelling (Gavrieli et al., 1992). Basically, cells were fixed with 0.1% glutaraldehyde, centrifuged for 5 min at 14 000 g at 4 °C. The pellet containing cells was permeabilized by the addition of 0.1% Triton-X100 for 15 min, washed with PBS, and labelled following the manufacturers’ instructions (Apoptag Direct Kit, Chemicon). Samples were then resuspended in PBS, and green fluorescence was observed using an epifluorescence microscope (Nikon, Eclipse E 800, Japan) (excitation 490 nm, emission of 525 nm). Samples were also analysed by using a DAKO cytomation flow cytometer (MoFlo, Beckman Coulter, Fullerton, CA, USA). Counts were triggered using forward scatter (FSC) signals. Positive controls consisted of cells pretreated with 10 μg ml−1 of DNAse I (nickase); for negative controls, distilled water was substituted for the terminal deoxynucleotidyl transferase.DEVDase activityFor estimating the protein concentration of samples, 50 ml of culture were centrifuged at 1500 g for 10 min. The pellets were resuspended in 1 ml of lysis buffer and kept at 4 °C for 1 h. The lysis buffer contained 25 mM Na+ HEPES, 2 mM dithiothreitol (DTT), 1 mM EDTA, 0.1% 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulphonate (CHAPS), 10% sucrose, 1 mM phenylmethylsulphonyl fluoride (PMSF), and 1 μM pepstatin A, pH 7.2. Then the homogenate was sonicated and centrifuged at 4 °C at 100 000 g in a Beckman ultracentrifuge using a Ti70 rotor for 1 h. The resultant supernatants were immediately frozen in liquid N2 and stored at –80 °C until use. Lysate protein concentration was measured by the bicinchoninic acid method (Pierce, Rockford, IL), with bovine serum albumin as standard.The DEVDase activity was determined by use of fluorescent substrates as previously described (Edelstein et al., 1999). The DEVDase assay was performed using 20–50 μl of the supernatant obtained as described above containing 50 μg total protein. A caspase assay buffer was added to the supernatant to achieve a total sample volume of 190 μl. The assay buffer consisted of 250 mM K+ HEPES, 50 mM KCl, 1 mM dithiothreitol, 1 mM EDTA, and 0.1% CHAPS, pH 7.4. The solution was pre-incubated for 10 min at 30 °C before the addition of the caspase substrate (Ac-Asp-Glu-Val-Asp-7-amido-4-methyl coumarin (Ac-DEVD-AMC) in 10% DMSO) (Thornberry et al., 1997). Ten μl of the substrate (25 μM final concentration) were added to make a final assay volume of 200 μl. Peptide cleavage was measured over 1 h at 25 °C using a Cytofluor 4000 series fluorescent plate reader (Perseptive Biosystems) at an excitation wavelength of 380 nm and an emission wavelength of 460 nm. An AMC standard curve was determined for each experiment. Caspase activity was expressed in nmol AMC released min−1 of incubation time mg−1 of lysate protein. To check whether or not these activities were real DEVDase, increasing concentrations of the irreversible caspase-3 inhibitor Z-DEVD-FMK (Calbiochem) were added to the homogenates according to Segovia et al. (2003) and samples were pre-incubated for 60 min before running reactions.ImmunodetectionSDS-PAGE electrophoresis in 12% gels run in Tricine buffer, as well as immunedetection and band analysis were carried out according to Capasso et al. (2001). Antibodies against the full-length precursor form of caspase-3 of human origin were purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). This antibody is highly specific and reacts with the active form and the full-length precursor of caspase-3. Jurkat (human T-cell leukemia) whole cell lysates were used (Santa Cruz Biotechnology Inc., Santa Cruz, CA) as positive controls.Statistical analysisMultiple group comparisons were performed using a one-way analysis of variance (ANOVA). Where significant differences were detected, post-hoc multiple comparisons were made by using the Tukey tests. A P value of <0.05 was considered statistically significant. To quantify the relationship between the variables, the Pearson product moment correlations were performed (considering P <0.05 as significant). Values are expressed as means ±standard deviation. The statistical analyses were carried out by using the STATISTICA 7 statistical package (StatSoft Inc. Tulsa, Oklahoma, USA).ResultsCellular morphologyTEM micrographs of control cells actively growing in light (Fig. 1A) showed that D. viridis has one big cup-shaped chloroplast that occupies more than 50% of the cell volume with the pyrenoid in the centre containing starch granules. The nucleus is located in the apical part of the cell surrounded by a well-defined nuclear membrane. Mitochondria and Golgi apparatus are easily recognized. The flagellar insertion point can also be seen in the cellular apex. Ninety-eight per cent of the cells presented this morphology (Table 1). When cells were injured with sub-lethal doses of the environmental stressors, they did not die and were able to cope with the damage (essentially identical to that described by Jimenez et al., 2007). However, exposure of cells to sudden hyperosmotic shock (5.5 M NaCl) (Fig. 1B) revealed that 55% of the cells had swollen and showed chromatin condensation but also extensive cytoplasmic vacuolation (arrow 1) in the absence of nuclear fragmentation and cellular blebbing (arrow 2). This morphology seems to resemble the paraptotic cell death phenotype. Seventy-six per cent of cells exposed to lethal UV radiation (Fig. 1C) presented symptoms of morphological changes typical of necrosis such as cell swelling, disruption of organelle membranes, condensation of mitochondria, and the formation of cytoplasmic blebs (arrow 1). Nevertheless, other features clearly indicated some apoptotic characteristics such as an intact cell membrane and neat membrane blebbing (arrow 2). This appearance gives the impression of an intermediate cell-death phenotype combining both apoptotic and necrotic features, or aponecrosis. However, chromatin condensation was not clear. Fig. 1D corresponds to cells exposed to acute heat shock (40 °C). Under these conditions, 96% of the cells displayed swelling and nuclear oedema. Mitochondrial rupture and disrupted organelle membranes were also observed (arrow 1). Eventually, altered plasma membranes typical of necrosis were also apparent. However, since some chromatin clusters can be identified as spots within the nucleus, it is suggested that the morphology might be necrotic-like (arrow 2). In nutrient-starvation experiments, cells were deprived of nitrogen for 7 d (Fig. 1E) and 77% of the cells demonstrated margination of the nucleus in the cell, pyknosis, clumping and condensation in the nucleus (arrow 1), and intact cell membrane with blebbing (arrow 2) and the absence of leakage of the intracellular content. However, cytoplasmic consumption was observed as indicated by the disappearance of chloroplast as well as other organelles (arrow 3). This morphotype is very difficult to interpret as it seems to be coincident with autophagic/vacuolar cell death. When control cultures of Dunaliella are allowed to reach a late stationary phase of growth or senescence (Fig. 1F), cell death started to happen and cell density declined sharply. In this case, chromatin aggregation and a certain level of karyorrhexis can be seen (arrow 1) as well as membrane blebs (arrow 2). Cytoplasmic disassembling suggests secondary necrosis after apoptosis. Eighty-three per cent of the cells were in this state.Table 1.Percentage of cells showing each different morphotype (Fig. 1) counted for each environmental stress under the TEM with the smallest magnification used (×3750)Cell morphotype% of cells(A) (normal)98 (8.76)B (paraptotic-like)65 (11.23)C (aponecrotic-like)76 (10.57)D (necrotic-like)96 (2.23)E (autophagic-like)77 (6.16)F (apoptotic-like)83 (5.44)Standard deviation in brackets. (A) Control cells growing in PAR. (B) Cells exposed to sudden hyperosmotic shock (5.5 M NaCl). (C) Cells exposed to lethal UV radiation. (D) Cells submitted to acute heat shock. (E) Cells under nitrogen deprivation. (F) Cells left to reach the late stationary phase of growth, i.e. senescence. n=200–250 (cells treatment−1).Fig. 1.Representative transmission electron micrographs showing the morphological changes in Dunaliella viridis subjected to different lethal stress treatments. C, chloroplast; Fi, flagellar insertion; G, Golgi, M, mitochondria; Mb, plasmalemma; N, nucleus; P, pyrenoid; S, starch. Arrows indicate the alterations indicated in the text. (A) Normal vegetative cells grown in PAR. (B) Cells After 3 h of lethal hyperosmotic shock (5.5 M NaCl). Note chromatin condensation together with extensive cytoplasmic swelling and vacuolation (arrow 1) in the absence of nuclear fragmentation and cellular blebbing (arrow 2). (C) Cells after 4 h of UV radiation. Characteristic cell swelling, organelle membranes are disrupted, mitochondria are condensed and cytoplasmic blebs appear (arrow 1). Cell membrane is intact and shows blebbing (Arrow 2). (D) Cells after 2 h of heat shock. Cells experience nuclear edema and swelling, mitochondrial rupture, disrupted organelle membranes and eventually altered plasma membrane (arrow 1), some chromatin clusters are observed (arrow 2). (E) Cells after 7 d of nitrogen starvation. Note pyknosis (arrow 1) but not margination of chromatin, cell membrane with blebbing (arrow 2), and cytoplasmic consumption (arrow 3). (F) Senescent cells after 12 d. Nucleus suffers margination in the cell, chromatin is condensed (arrow 1), clumped and marginated in the nucleus, while cell membrane is intact with blebbing (arrow 2) and undamaged organelles. Picture augmentation was ×25 000.Dead cellsCounting of Evans Blue-stained cells revealed that under the non-lethal stress the cells did not incorporate the dye. After the first 30 min following osmotic shock around 35% of the cells were already dead. One and 2 h later, more than half of the population was dead (Fig. 2A). The response was different after UV radiation and thermal shock. After the UV stress, the cells also died during the first 4 h of treatment. However, the amount of dead cells increased 2-fold between 4 h and 6 h, reaching almost 80% of the total number of cells (Fig. 2B). The number of dead cells rose to about 3-fold, 2 h after the heat shock treatment and remained constant during the whole period of time (Fig. 2C). Sixty per cent of the cells died suddenly after day 7 under nitrogen deprivation whilst during the previous days only 5% or 10% (days 2 and 5, respectively) of the cells showed blue staining (Fig. 2D). Senescent cells started to die during day 7 and by the end of the sampling period (day 12), around 90% of the population was already dead (Fig. 2E).Fig. 2.Cell death judged by Evans Blue mortal staining of cells under light microscopy with lethal stress treatments. The dashed horizontal line represents live control cells in PAR at t=0. (A) Cells after osmotic shock. (B) Cells after UV radiation. (C) Cells after heat shock. (D) Cells under nitrogen starvation. (E) Culture senescence.DNA condensationD. viridis cells were also analysed by means of CLM using DAPI nuclear staining. Normal cells growing in PAR (Fig. 3A) showed homogeneous DAPI staining well confined to the nuclear area. Two hours after the onset of a lethal osmotic shock (5.5 M NaCl) (Fig. 3B) cells presented irregular DAPI staining and a slight degree of chromatin clumping could be observed. Fig. 3C represents cells exposed for 2 h to lethal UV irradiation. In this case, chromatin aggregation was not only restricted to the nuclear area, blue-stained granules were also observed in the centre of the cell, therefore suggesting some degree of karyolysis. After 2 h of acute heat shock (Fig. 3D) cells demonstrated chromatin aggregation but again, not strictly restricted to the nuclear area. Figure 3E corresponds to cells subjected to 7 d of nitrogen starvation in which well-defined DNA condensation is observed, appearing only in the nuclear zone. Senescent cells also showed clear chromatin condensation in the nucleus (Fig. 3F).Fig. 3.DNA condensation in Dunaliella viridis revealed by DAPI staining and confocal laser microscopy. (A) Control cells in PAR. (B) Cells after osmotic shock (5.5 M NaCl). (C) Cells after 4 h of UV radiation. (D) Cells after 2 h of heat shock. (E) Cells after 7 d of nitrogen starvation. (F) Culture senescence. Horizontal bar is 1 μm. (This figure is available in colour at JXB online.)DNA strand breaksWhile the morphological changes mentioned above occurred, nuclear DNA was concurrently degraded in some of the stress treatments. Free 3′ OH ends of DNA, generated by activation of endonuclease activity in dying cells, were fluorescently labelled with a conventional TUNEL assay (Gavrieli et al., 1992). No labelling was observed either in cells growing in PAR (Fig. 4A) or in the the non-lethal conditions. In the FACS chart, quadrant R1 contain cells with positive TUNEL labelling, and quadrant R3 corresponds to the absence of TUNEL green fluorescence. Red chlorophyll fluorescence comprised both quadrants. After osmotic shock (Fig. 4B) only about 5% of the cells presented green fluorescence and such labelling was very faint. Around 68% of cells showed labelling after UV radiation (Fig. 4C) and a similar pattern, was observed after 3 h of heat shock (Fig. 4D). Results similar to those obtained under osmotic shock were also found under the nitrogen-starvation treatment (Fig. 4E), in which the number of green fluorescent cells was about 28%. A greater number of cells (93%) presented green fluorescence under the senescence treatment (Fig. 4F) after 12 d. Some of the counts in R1 correspond to cellular debris (for instance the dots at 100 both for FITC and red chlorophyll). Positive controls, consisting of cells treated with DNAse showed strong staining, indicative of DNA degradation. Negative controls were analysed by using 9 d senescent cells and by substituting MilliQ water for the TdT enzyme. These cells did not stain (controls data not shown).Fig. 4.DNA fragmentation in Dunaliella tertiolecta after different environmental stresses revealed by TUNEL staining. (A) Control cells in PAR. (B) Cells after 3 h of osmotic shock. (C) Cells after 4 h of UV radiation. (D) Cells after 3 h of heat shock. (E) Cells after 7 d of nitrogen starvation. (F) Senescence after 12 d. Horizontal bar is 20 μm. (This figure is available in colour at JXB online.)DEVDase activityDEVDase activity was detected in D. viridis cell extracts, in response to the different stress conditions tested in this work (Fig. 5). Exposure of cultures of D. viridis to either lethal or sub-lethal hyperosmotic stress by the addition of NaCl (5.5 M and 4 M final concentrations, respectively) produced different results (Fig. 5A). While non-lethal hyperosmotic stress did not result in significant increases in activity, cells exposed to 5.5 M NaCl exhibited a rapid increase in DEVDase activity within 0.5 h after the shock, and resulted in values of 337±67 nmol min−1 mg−1 protein after 2 h (about a 17-fold increase). An increase in activity was also determined for cultures subjected to either lethal doses of UV radiation (70 mJ cm−2; Fig. 5B) or temperature (40 °C; Fig. 5C). In these cases a similar increase in DEVDase activity (100–120 nmol min−1 mg−1 protein) was noted under both treatments in the first few hours with a significant drop in activity at later time points most likely representing cell death. The activity of the enzymes was around half of that obtained under osmotic shock. Unlike hyperosmotic stress, sub-lethal UV and temperature shock did not induce a significant increase of DEVDase activity.Fig. 5.DEVDase activity in Dunaliella viridis following exposure to various stresses. Changes in enzymatic activity were measured as hydrolysis of 7-amino-4-fluoromethyl coumarin-labelled specific substrate DEVD, in all stress treatments. (A) Osmotic shock. (B) UV radiation. (C) Heat shock. Lethal conditions of stress factors are represented by black bars and sub-lethal conditions by white bars. (D) Nitrogen starvation. (E) Senescent cultures. (F) Inhibition of the enzymatic activity by using the irreversible caspase-3-inhibitor Z-DEVD-FMK after 4 h of UV lethal stress. The highest point of DEVDase activity for the rest of the treatments was inhibited with the highest concentration of the inhibitor (200 μM) and any of them showed activity. Statistical differences (P <0.05) during the time-course for the lethal stress are marked with letters. Same letter means no differences.Nitrogen starvation (Fig. 5D) is a much slower process leading to cell death, and a peak of maximal DEVDase activity (17-fold increase compared with control cells) was found 5 d after nitrogen removal from the medium. The maximum DEVDase activity in nitrogen-starved cells at 5 d was more than double compared to treatment with UV or temperature and similar to that caused by osmotic stress (Fig. 5B, C, respectively), reaching average values of 250 nmol min−1 mg−1 protein.Having demonstrated that DEVDase activity was induced in cells exposed to different forms of lethal stress, the influence of natural cell deterioration due to culture ageing was studied. Fig. 5E depicts a dramatic increase in DEVDase activity in senescent cultures of D. viridis at 12 d following initial inoculation (late stationary phase of growth). The activity reached values of more than 350 nmol min−1 mg−1 protein, indicating that activation of this particular caspase-like activity naturally occurs in ageing cultures of D. viridis. DEVDase activity dropped 50% after the addition of 100 μM of the irreversible caspase 3 inhibitor Z-DEVD-FMK (Fig. 5F), and was not inhibited at all by 1 mM PMSF or 1 μM pepstatin A, well-known inhibitors of serine proteases and acid proteases (aspartyl peptidases), respectively.Immunoblot analysisA mammalian anti-caspase 3 antibody was used to detect the caspase-like protein in D. viridis in response to lethal environmental stress (Fig. 6). Such protein was essentially absent in normal steady-state cells; however, a rapid increase in the intensity of the active 32 kDa caspase-like band was detected after the shock treatments. According to the apparent molecular weight and to the migration of the caspase-3 positive control, the band is similar to the caspase-3 of human origin. The presence of the caspase-like band was only slightly detected after the lethal hyperosmotic shock (Fig. 6A) and no bands appeared under the sub-lethal condition. Under UV radiation (Fig. 6B), maximal band intensity occurred 4–6 h after the treatment, while 3–4 h was needed following heat shock (Fig. 6C). It started to accumulate 4 d after the transfer of cells to nitrogen-free medium, reaching maximal band intensity after 7 d (Fig. 6D). Senescence of cultures of D. viridis also induced the appearance of the caspase-like enzyme (Fig. 6E) 9 d after inoculation, showing maximal band intensity after 12 d (late stationary phase of growth). Thus, there was a clear induction of caspase-like enzyme synthesis in response to lethal stress, irrespective of the cell death morphology presented.Fig. 6.Western blot showing cross-reactions of protein extracts from Dunaliella viridis with antibodies raised against human caspase 3. Lanes correspond to the hours or days exposed to each particular stress treatment. (A) Osmotic shock. (B) UV radiation. (C) Heat shock. (D) Nitrogen starvation. (E) Culture senescence.DiscussionApoptosis is an active form of cell death by which individual cells commit suicide. It is a highly controlled and organized process characterized by well-defined morphological changes. The possible role of programmed cell death (PCD) in unicellular organisms has received much attention recently (Cornillon et al., 1994; Ameisen, 1996; Vardi et al., 1999; Frolich and Madeo, 2000; Lewis, 2000; Ning et al., 2002; Segovia et al., 2003). The presence of key components of cell death pathways in some of the earliest-evolved organisms (Berman-Frank et al., 2004) suggests that their origins are truly ancient, and it has been speculated that they may be the result of viral-eukaryote genomic mixing during ancient evolutionary history (Berges and Falkowski, 1998; Segovia et al., 2003; Bidle and Falkowski, 2004). However, the existence and evolution of PCD in unicellular organisms is controversial (Deponte, 2008) and obviously confusing because unlike multicellular organisms, it results in complete loss of the organism. However, apoptosis is not the only way by which cells may die. Evidence shows that other alternative forms of non-apoptotic PCD exist in parallel or in an independent manner with apoptosis (Golstein and Kroemer, 2005; Bredesen, 2007), with important implications for understanding the type of PCD that occurs in unicellular microalgae and how they operate. It is now commonly thought that subtle or dramatic changes in the cell-death phenotype are a direct result of the relative degree of the injury to which the cells are exposed. Supporting this contention, there are data that environmental stimuli can produce different types of cell death depending on the intensity of the stimulus and on ATP availability within the cell, and that classic apoptosis and necrosis may represent only two extremes of a continuum of intermediate forms of cell death (Papucci et al., 2004).In the present work, both morphological and biochemical approaches have been used to study the forms of cell death in D. viridis. Only senescent cultures showed unambiguous features of apoptosis, followed by secondary necrosis; in turn, the other stress factors provoked different cell-death phenotypes under lethal conditions. In the case of heat shock, necrosis-like was the single cell-death process, whilst UV radiation produced an intermediate cell-death morphotype combining both apoptotic and necrotic features, with aponecrotic-like characteristics. A different pattern was observed under hyperosmotic shock, where cells highly resembled the paraptotic-like phenomenon. Finally, algae exposed to nitrogen starvation showed a different death morphotype, that, acording to the results obtained, suggest that it is a similar mechanism to that of autophagic/vacuolar-like cell death. The ends of the death chain would then correspond to necrosis and apoptosis (Aigner, 2002; Papucci et al., 2004), fitting within the cell-death phenotypes found in D. viridis under heat shock and senescence, respectively. The necrotic-like morphology observed after heat stress implies active cell death (Kitanaka and Kuchino, 1999). Chromatin condensation was not observed, but the cells presented chromatin clusters as well as disorganized spots accompanied by cytoplasmic swelling, also described by Leist and Jäätelä (2001), and TUNEL labelling was positive. This is opposed to the necrotic event per se, which is passive and unprogrammed. Cell disintegration under heat stress seems to be programmed and organized in Chlorella pyrenoidosa (Leu and Hsu, 2004) and shows apoptotic morphology as well as intermediate morphotypes in Chlorella saccharopila (Zuppini et al., 2007). Interestingly, the temperature used in both species was higher than the temperature used for D. viridis, which showed a necrotic-like pattern. In the same context, Symbiodinium sp., the symbiotic dinoflagellate of the sea anemone Aiptasia sp., presented PCD characteristics under experimental bleaching, that shifted between both necrosis and apoptosis, depending on the extent of the stress (Dunn et al., 2002, 2004). When D. viridis cells were left to reach late stationary phase, they died apoptotically-like. The cell-death phenotype shown by these cells was totally coincident with that described for D. tertiolecta under light deprivation (Segovia et al., 2003). This is not surprising as ageing has been widely reported to be one of the conditions by which apoptosis takes place in vascular plants (Fukuda, 1994; Buchanan-Wollaston et al., 2003) and by which phytoplankton blooms disappear (Berges and Falkowski, 1998; Vardi et al., 1999; Ross and Sharples, 2007). In fact, dead cells in natural phytoplankton populations closely resemble senescent cultured cells (Veldhuis et al., 2001).Intermediate features of both apoptosis and necrosis morphotypes or aponecrosis (Formigli et al., 2000) were exhibited when D. viridis was irradiated with UV, including a high degree of TUNEL labelling. UV has been reported to cause apoptotic-like cell death in the unicellular alga Chlamydomonas reinhardtii subjected to high doses of UV (100 J m−2) (Moharikar et al., 2006). However, the difference between D. viridis showing an aponecrotic-like appearance and C. reinhardtii with apoptotic-like morphology might reside in the UV doses received, i.e. in the level of the injury caused to the cells. So, depending on the damage infringed, the cells seem to ‘choose’ how to die.Completely distinct responses took place under nitrogen starvation and hyperosmotic shock. Under nitrogen starvation, cells may die by means of a mechanism similar to autophagic/vacuolar-like cell death as indicated by the disappearance of clear cytoplasm. Nevertheless, the nucleus was not degraded and chromatin was slightly clumped and condensed. Strikingly, a small percentage of cells showed TUNEL positive labelling. Several reports support that autophagic cell death goes through without the generation of DNA strand breaks (Kissova et al., 2006; Bassham, 2007). Nuclear degradation and pyknosis seem not to be universal in all cells undergoing autophagic death (Jones 2000), and this would rather depend on the cell type. Although the positive TUNEL assay was not statistically significant, the reason lying beyond the chromatin cluttering observed with DAPI and TEM, deserves further studies. Yet, the most important fact is that autophagic/vacuolar cell death occurs in nutrient-starved cells, serving as a cellular protective mechanism up-regulated by nutrient starvation (Yue et al., 2003; Shimizu et al., 2004; Levine and Yuan, 2005). In vascular plants, autophagy has been known for some time to be important for nutrient remobilization during sugar and nitrogen starvation and leaf senescence, and recent reports focus on its role in housekeeping functions related to oxidative stress (Bassham, 2007).Finally, when cells were subjected to hyperosmotic shock they showed an analogous appearance to paraptotic-like morphology, i.e. chromatin condensation and also cytoplasmic swelling and vacuolation. Like apoptosis, it seems to be an active process requiring transcription and de novo protein synthesis (Aigner, 2002). Paraptosis has been reported to occur in the unicellular dinoflagellate A. carterae when cultured in darkness and during culture senescence (Franklin and Berges, 2004) concurring with rapid vacuolization, loss of internal structure, intact membranes, and lack of DNA fragmentation. Paraptosis is mediated by mitogen-activated protein kinases (MAPKs) in human cells (Sperandio et al., 2004). Coincidently, authors have demonstrated the presence of mitogen-activated protein (MAP) kinase signalling pathways in D. viridis, and that operation of the p38 and the c-Jun N-terminal kinase (JNK) cascades are crucial for adaptation and survival of this microalga upon hyperosmotic stress (Jiménez et al., 2004), the very stress treatment described in this work under which the cells seem to undertake a paraptotic-like demise. Moreover, hyperosmotic shock, nitrogen starvation, and UV irradiation, impaired cell division and caused a marked decrease in the phospho-ERK levels in D. viridis (Jiménez et al., 2007). These data suggest that, indeed, D. viridis might undergo a paraptotic-like event.Hence, it is now generally accepted that multiple forms of programmed cell death exist and that some of them do not require the activation of caspases (Leist and Jäätelä, 2001; Clarke, 2002). Thus, the term apoptosis in most cases, but not always, is exclusively used for caspase-dependent cell death (Blagosklonny, 2000; Leist and Jäätelä, 2001). On the other hand, while necrosis does not require caspase activation, necrosis-like, being active cell death might or might not require caspase activation (Kitanaka and Kuchino, 1999), and the same is true for aponecrosis (Papucci et al., 2004). However, paraptosis and autophagic/vacuolar cell death traditionally do not call for the participation of caspases (Sperandio, 2000; Jones, 2000; Wyllie and Golstein, 2001; Leist and Jäätelä, 2001). A highly significant increase of DEVDase activity and activation in cell extracts was found after the induction of cell death by means of the environmentally relevant stresses. Caspase-like enzymatic activities and immunodetection were reported for the first time in the unicellular cholorophyte D. tertiolecta (Segovia et al., 2003), other reports regarding metacaspase activity in several microalgal species have also been described (Berman-Frank et al., 2004; Moharikar et al., 2006; Zuppini et al., 2007) and analyses of completed genome sequences of prokaryotic and eukaryotic phytoplankton have revealed the widespread presence of metacaspases in some cyanobacteria, and in unicellular eukaryotic microalgae (Bidle and Falkowski, 2004). However, despite the fact that metacaspases have been reported to operate in an analogous manner to caspases, they are distinct in terms of target site specificity from caspases, at least in vascular plants [their target substrate contains either lysine or arginine at the P1 position, whilst caspase-like enzymes seem to have specificity for aspartate (Vercammen et al., 2004., 2007; Watanabe and Lam, 2005)], and their roles in protists are not obvious yet (Deponte, 2008). DEVDase was not apparent in steady-state D. viridis cells; however, a rapid increase in the intensity of the 32 kDa band was detected after shock treatments. Our data suggest that these activities must be DEVDases because PMSF (an inhibitor of serine proteases) and pepstatin A (an inhibitor of aspartyl peptidases) did not inhibit the activity, whilst the irreversible caspase-3 inhibitor Z-DEVD-FMK inhibited them completely. In this context, it seems clear that environmental treatments induced the activation and the increase in DEVDase activity in unicellular organisms. Yet, the nature of caspase-like proteins in plants is diverse (Bonneau et al., 2008) and a more thorough inhibitor analyses is necessary. Similar results were obtained in other unicellular chlorophytes exposed to UV (Moharikar et al., 2006) and to heat shock (Zuppini et al., 2007). In both cases, antibodies against a mammalian caspase-3 from human origin cross-reacted with a protein of 28–32 kDa and its pattern of expression correlated with the onset of cell death. DEVDase activity and accumulation of the 32 kDa band was observed in cells undergoing all the different cell-death types described in this work, despite the fact that some of those cell-death events, as for example autophagic-like or paraptotic-like deaths, do not necessarily concur with caspase activation. A non-apoptotic role for proapoptotic caspases was previously proposed (Zeuner et al., 1999), indicating that high caspase-3 activity is not exclusive of apoptotic cell death, and in phytoplankton they have been reported to be constitutive and having housekeeping functions (Segovia and Berges, 2005). Thus, growing evidence suggests the participation of caspases in other cellular processes such as development, cell cycle, cell proliferation, and receptor internalization (Algeciras-Schimnich et al., 2002), in addition to their well-characterized role in apoptosis, helping to understand apoptosis-dependent and apoptosis-independent functions of caspases.PCD is difficult to explain as it is a mechanism that offers negative selective pressure. In spite of this, cells have managed to use PCD for several ecologically relevant purposes. One theory suggests that PCD in unicellular organisms would provide for evolutionary advantage for their genome to survive harsh conditions, optimize adaptation of cell numbers to environmental conditions (i.e. nutrient availability), maintain tight regulation of the cell cycle and differentiation (i.e. formation of resistant forms: cysts), enhance defence against pathogens (i.e. viral infection), and to promote and maintain clonality within the population (Welburn et al., 1997; Ameisen, 2000). In this sense, PCD in unicellular organisms may be considered as a safety mechanism for the population. The evidence that alternative, non-apoptotic PCD exists in unicellular organisms has important implications for understanding cell dynamics. That environmental stimuli can produce different types of cell death depending on the intensity of the stimulus, and that classic apoptosis and necrosis may represent only two extremes of a continuum of intermediate forms of cell death, is also applicable to unicellular organisms. If PCD in phytoplankton is truly the result of a programme activated by environmental factors, then it is important to understand how activation happens and what occurs within the cell in response. The existence of genetically driven cell-death phenomena in phytoplankton indicates that the regulation mechanisms responsible for the dichotomy cell survival/cell death were already present in single-celled organisms 1.2–1.6 billion years ago and preceeded multicellularity.It has been demonstrated in this work, by using morphological and biochemical approaches, that the presence of different cell-death programmes, mediated by DEVDase activity, can occur in a single-celled organism such as the unicellular chlorophyte D. viridis, and that cell death may vary from apoptosis to necrosis, going through different intermediate morphotypes such as aponecrosis, or adopt autophagic/vacuolar or paraptotic-like appearance, depending on the stimulus received. New advances in the elucidation of the molecular pathways leading to cell death in phytoplankton will decipher the genes involved in the responses to different stress factors.This research was supported by grants from the Ministry of Science and Innovation (MICINN) (Spain) to C Jimenez (CGL05-01071 partially funded by FEDER) and to M Segovia (CTM06-09710), and grants from the National Institutes of Health (NIH) to CL Edelstein (DK-56851). We thank the two anonymous reviewers for their exhaustive criticisms.AignerTApoptosis, necrosis, or whatever: how to find out what really happens?Journal of Pathology200219814Algeciras-SchimnichABarnhartBCPeterMEApoptosis: independent functions of killer caspasesCurrent Opinion in Cell Biology200214721726AmeisenJCThe origin of programmed cell deathScience199627212781279AmeisenJCLockshinRZakeriZTillyJThe evolutionary origin and role of programmed cell death in single celled organisms: a new view of executioners, mitochondria, host–pathogen interactions, and the the role of death in the process of natural selectionWhen cells die1998New YorkWiley-Liss356AmeisenJCOn the origin, evolution, and nature of programmed cell death: a timeline of four billion yearsCell Death and Differentiation20029367393BasshamDCPlant autophagy: more than a starvation responseCurrent Opinion in Plant Biology200710587593BergesJAFalkowskiPGPhysiological stress and cell death in marine phytoplankton: induction of proteases in response to nitrogen or light limitationLimnology Oceanography199843129135Berman-FrankIBidleKDHaramatyLFalkowskiPGThe demise of the marine cyanobacterium, Trichodesmium spp., via an autocatalyzed cell death pathwayLimnology and Oceanography2004499971005BidleKDFalkowskiPGCell death in planktonic, photosynthetic microorganismsNature Reviews in Microbiology20042643655BlagosklonnynMVCell death beyond apoptosisLeukemia20001415021508BonneauLGeYDruryGEGalloisPWhat happened to plant caspases?Journal of Experimental Botany200859491499BorowitzkaLJThe microflora: adaptations to life in extremely saline lakesHydrobiologia1981813346BredesenDEToward a mechanistic taxonomy for cell death programsStroke200738652660Buchanan-WollastonVEarlSHarrisonEMathasENavabpourSPageTPinkDThe molecular analysis of leaf senescence: a genomics approachPlant Biotechnology Journal20031322CapassoJMRivardCBerlTThe expression of the gamma subunit of Na-K-ATPase is regulated by osmolality via C-terminal Jun kinase and phosphatidylinositol 3-kinase-dependent mechanismsProceedings of The National Academy of Sciences, USA2001981341413419CasottiRMazzaSBrunetCVantrepotteVIanovaAMiraltoAGrowth inhibition and toxicity of the diatom aldehyde 2-trans, 4-trans-decadienal on Thalassiosira weissflogii (Bacillariophyceae)Journal of Phycology200541720ClarkePGHDevelopmental cell death: morphological diversity and multiple mechanismsAnatomy and Embryoology1990181195213ClarkePGHApoptosis: from morphological types of cell death to interacting pathways: comment from ClarkeTrends in Pharmacological Science200223308309CornillonSFoaCDavoustJBuonavistaNGrossJDGolsteinPProgrammed cell-death in DictyosteliumJournal of Cell Sciences199410726912704CrippenRWPerrierJLThe use of neutral red and Evans blue for live–dead determinations of marine planktonStaining Technology19744997104DeponteMProgrammed cell death in protistsBiochemica et Biophysica Acta2008178313961405DunnSRBythellJCLe TissierMDABurnettWJThomasonJCProgrammed cell death and cell necrosis activity during hyperthermic stress-induced bleaching of the symbiotic sea anemone Aiptasia spJournal of Experimental Marine Biology and Ecology20022722953DunnSRThomasonJCThisslerMLBythellJCHeat stress induces different forms of cell death in sea anemones and their endosymbiotic algae depending on temperature and durationCell Death and Differentiation2004112131222EdelsteinCLShiYXSchrierRWRole of caspases in hypoxia-induced necrosis of rat renal proximal tubulesJournal of the American Society of Nephrology19991019401949ElmoreSApoptosis: a review of programmed cell deathToxicologic Pathology200735495516FormigliLPapucciLTaniASchiavoneNTempestiniAOrlandiniGECapaccioliSOrlandiniSZAponecrosis: morphological and biochemical exploration of a syncretic process of cell death sharing apoptosis and necrosisJournal of Cellular Physiology20001824149FranklinDJBergesJAMortality in cultures of the dinoflagellate Amphidinium carterae during culture senescence and darknessProceedings of the Royal Society London, Biology200427120992107FrohlichKUMadeoFApoptosis in yeast: a monocellular organism exhibits altruistic behaviourFEBS Letters200047369FukudaHProgrammed cell death of tracheary elements as a paradigm in plantsPlant Molecular Biology199444245253GavrieliYShermanYBen-SassonSAIdentification of programmed cell-death in situ via specific labelling of nuclear-DNA fragmentationJournal of Cell Biology1992119493501GinzburgMDunaliella: a green alga adapted to saltAdvances in Botanical Research19871493183GolsteinPKroemerGRedundant cell death mechanisms as relics and backupsCell Death and Differentiation20051214901496GreenbergJTProgrammed cell death: a way of life for plantsProceedings of the Natural Academy of Sciences, USA1996931209412097JimenezCCossioBRRivardCJBerlTCapassoJMCell division in the unicellular microalga Dunaliella viridis depends on phosphorylation of extracellular signal-regulated kinases (ERKs)Journal Experimental Botany20075810011011JimenezCBerlTRivardCJEdelsteinCCapassoJMPhosphorylation of MAP kinase-like proteins mediates the response of the halotolerant alga Dunaliella viridis to hypertonic shockBiochimica et Biophysica Acta200416446169JohnsonMKJohnsonEJMacElroyRDSpeer Hl, BruffBSEffects of salts on halophilic alga Dunaliella viridisJournal of Bacteriology19689514611468JonesADoes the plant mitochondrion integrate cellular stress and regulate programmed cell death?Trends in Plant Science20005225230KerrJFRWyllieAHCurrieARApoptosis: a basic biological phenomenon with wide-ranging implications in tissue kineticsBritish Journal of Cancer197226239257KissovaIPlamondonLTBrissonLPriaultMRenoufVSchaefferJCamougrandNManonSEvaluation of the roles of apoptosis, autophagy, and mitophagy in the loss of plating efficiency induced by bax expression in yeastJournal of Biological Chemistry20062813618736197KitanakaCKuchinoYCaspase-independent programmed cell death with necrotic morphologyCell Death and Differentiation19996508515LamEDel PozoOCaspase-like protease involvement in the control of plant cell deathPlant Molecular Biology200044417428LamEKatoNLawtonMProgrammed cell death, mitochondria and the plant hypersensitive responseNature2001411848853LeistMNicoteraPThe shape of cell deathBiochemical and Biophysical Research Communications199723619LeistMJäätteläMFour deaths and a funeral: from caspases to alternative mechanismsNature Reviews in Molecular and Cell Biology20012589598LeuKLHsuBDA programmed cell disintegration of Chlorella after heat stressPlant Science2005168145152LevineBYuanJAutophagy in cell death: an innocent convict?Journal of Clinical Investestigation200511526792688LewisKProgrammed cell death in bacteriaMicrobiology and Molecular Biology Reviews200064503514MoharikarSD'SouzaJSKulkarniABRaoBJApoptotic-like cell death pathway is induced in unicellular chlorophyte Chlamydomonas reinhardtii (Chlorophyceae) cells following UV irradiation: detection and functional analysesJournal of Phycology200642423433NingSBGuoHLWangLSongYCSalt stress induces programmed cell death in prokaryotic organism AnabaenaJournal of Applied Microbiology2002931528PapucciLFormigliLSchiavoneNApoptosis shifts to necrosis via intermediate types of cell death by a mechanism depending on c-myc and bcl-2 expressionCell and Tissue Research2004316197209PennellRILambCProgrammed plant cell in plantsThe Plant Cell1997911571168RossONSharplesJPhytoplankton motility and the competition for nutrients in the thermoclineMarine Ecology-Progress Series20073472138SegoviaMPerez MartinJMProgrammed cell death in dinoflagellatesProgrammed cell death in Protozoa2007Church Street Georgetown USALandes Bioscience-Springer Wiley126142SegoviaMBergesJAEffect of inhibitors of protein synthesis and DNA replication on the induction of proteolytic activities, caspase-like activities and cell death in the unicellular chlorophyte Dunaliella tertiolectaEuropean Journal of Phycology2005402130SegoviaMHaramatyLBergesJAFalkowskiPGCell death in the unicellular chlorophyte Dunaliella tertiolecta: an hypothesis on the evolution of apoptosis in higher plants and metazoansPlant Physiology200313299105ShimizuSKanasekiTMizushimaNMizutaTArakawa-KobayashiSThompsonCbTsujimotoYRole of bcl-2 family proteins in a non-apoptotic programmed cell death dependent on autophagy genesNature Cell Biology2004612211228SperandioSde BelleIBredesenDEAn alternative, nonapoptotic form of programmed cell deathProceedings of the National Academy of Sciences, USA2000971437614381SperandioSPoksayKde BelleILafuenteMJLiuBNasirJBredesenDEParaptosis: mediation by MAP kinases and inhibition by AIP-1/AlixCell Death and Differentiation20041110661075SwansonSJBethkePCJonesRLBarley aleurone cells contain two types of vacuoles: characterization of lytic organelles by use of fluorescent probesThe Plant Cell199810685698ThornberryNARanonTAPietersonEPA combinatorial approach defines specificities of members of the caspase family and granzyme B: functional, relationships established for key mediators of apoptosisJournal of Biological Chemistry19972721790717911VardiABerman-FrankIRozenbergTHadasOKaplanALevineAProgrammed cell death of the dinoflagellate Peridinium gatunense is mediated by CO2 limitation and oxidative stressCurrent Biology1999910611064VeldhuisMJWKraayGWTimmermansKRCell death in phytoplankton: correlation between changes in membrane permeability, photosynthetic activity, pigmentation and growthEuropean Journal of Phycology200136167177VercammenDDeclercqWVandenabeelePVan BreusegemFAre metacaspases caspases?Journal of Cell Biology2007179375380VercammenDVan de CotteBDe JaegerGEeckhoutDCasteelsPVandepoeleKVandenbergheIVan BeeumenJInzéDVan BreusegemFType II metacaspases Atmc4 and Atmc9 of Arabidopsis thaliana cleave substrates after arginine and lysineJournal of Biological Chemistry2004444532945336WatanabeNLamETwo Arabidopsis metacaspases AtMCP1b and AtMCP2b are 18 arginine/lysine-specific cysteine proteases and activate apoptosis-like cell death in 19 yeastJournal of Biological Chemistry20052801469114699WelburnSCBarcinskiMAWilliamsGTProgrammed cell death in trypanosomatidsParasitotology Today1997132226WyllieAHGolsteinPMore than oneway to goProceedings of the National Academy of Sciences, USA2001981113YueZJinSYangCLevineAJHeintzNBeclin 1, an autophagy gene essential for early embryonic development, is a haploin sufficient tumor suppressorProceedings of the National Academy of Sciences, USA20031001507715082ZeunerAEramoAPeschleCDe MariaRCaspase activation without deathCell Death and Differerentiation1999610751080ZuppiniAAndreoliCBaldanBHeat stress: an inducer of programmed cell death in Chlorella saccharophilaPlant and Cell Physiology20074810001009
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B8AA39E3E4530D655DC168C153C6510FEA35523.txt b/test/dataset/in/resources/corpus/Clean_0B8AA39E3E4530D655DC168C153C6510FEA35523.txt
deleted file mode 100644
index b47662c..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B8AA39E3E4530D655DC168C153C6510FEA35523.txt
+++ /dev/null
@@ -1 +0,0 @@
- geront Gerontologistgeront The Gerontologist The Gerontologist 0016-9013 1758-5341 Oxford University Press 90810.1093/geront/43.6.908 PRACTICE CONCEPTS The Savvy Caregiver Program: Developing and Testing a Transportable Dementia Family Caregiver Training Program Hepburn Kenneth W. PhD 1 Lewis Marsha PhD, RN 1 Sherman Carey Wexler MA 2 Tornatore Jane PhD 3 Address correspondence to Kenneth W. Hepburn, PhD, School of Nursing, 6-169 Weaver-Densford Hall, University of Minnesota, 308 Harvard St. SE, Minneapolis, MN 55455. E-mail: hepbu001@umn.edu 12 2003 43 6 908 915 10 10 2002 23 7 2002 The Gerontological Society of America 2003 Purpose: This article reports on the development and field testing of the Savvy Caregiver Program, the transformation of a successful, academic-based caregiver psychoeducational program into a self-contained program that can be adopted in other locations.Design and Methods: Program development began with a prototype of a 12-hr course with the aims of introducing family caregivers to the caregiving role, providing them with the knowledge, skills, and attitudes needed to carry out that role, and alerting them to self-care issues. Results from initial field trials dictated a substantial revision of the workshop materials. The next version was field tested in multiple sites in southern rural Minnesota, Colorado, and Alaska. In this expanded testing, participants evaluated the program, and cross-group comparisons were conducted by use of well-established caregiver well-being scales.Results: Virtually all respondents reported increased skill, knowledge, and confidence, and all would recommend the program to others. A preintervention versus postintervention analysis indicates that caregivers' reaction to the overall behavior of the persons for whom they provide care (i.e., “total reaction”), their self-reported burden, and their beliefs about caregiving (emotional enmeshment) changed significantly in directions indicating better caregiver well-being. Implications:Results suggest that it is feasible to translate a research-based caregiver intervention into a packaged program that can be adopted in other settings without the direct involvement of the program initiators. Alzheimer's disease Dissemination Community intervention Translational research hwp-legacy-fpage 908 hwp-legacy-dochead RESEARCH ARTICLE Family members provide the bulk of community-based care for persons with dementia and often do so for long periods of time (National Alliance for Caregiving, 1997; Stone, Cafferata, & Sangl, 1987). This practice provides a substantial cost benefit to society (Arno, Levine, & Memmott, 1999; Max, Webber, & Fox, 1995), but it can exact a heavy toll on caregivers in the form of depression, social isolation, family stress, and other adverse outcomes (Alspaugh, Zarit, Stephens, Townsend, & Greene, 1999; Schulz & Beach, 1999; Schulz, O'Brien, Bookwala, & Fleissner, 1995). The most frequently identified sources of caregiver burden and distress are issues associated with care receiver behavior and efforts to manage that behavior (Gaugler, Davey, Pearlin, & Zarit, 2000; George & Gwyther, 1986; Teri, 1997). Effectively dealing with these issues while maintaining one's own well-being is particularly challenging for family caregivers, given their close emotional ties to the care receiver, the changing nature of their relationship with the ill family member, and their need to take on new caregiving-related tasks for which they are typically unprepared. Recent research suggests that caregivers can be trained to succeed in their new role with less detriment to themselves. In a National Institute for Nursing Research (NINR)-supported randomized controlled trial, we demonstrated that the Minnesota Family Workshop, a 14-hr psychoeducation program that drew on stress mediation theory (Lazarus & Folkman, 1984), was effective in reducing caregiver burden, depression, and reaction to behavior and in helping caregivers achieve a less emotionally enmeshed attitude (Hepburn, Tornatore, Center, & Ostwald, 2001; Ostwald, Hepburn, Caron, Burns, & Mantell, 1999). A number of other psychoeducational and counseling interventions have demonstrated benefits for caregiver well-being, delay of institutionalization of the care recipient, and even care recipient mortality (Brodaty, McGilchrist, Harris, & Peters, 1993; Buckwalter et al., 1999; Gallagher-Thompson & DeVries, 1994; Gendron, Poitras, Dastoor, & Perodeau, 1996; Mittelman, Ferris, Shulman, Steinberg, & Levin, 1996). For large numbers of family caregivers to benefit from successful research-based training programs, the programs must be translatable to other locations and auspices. One might expect to see widespread translation efforts, given the anticipated increase in the number of persons with dementing disorders (Brookmeyer, Gray, & Kawas, 1998). However, reports of such efforts are notably absent from the literature. In this article, we report on the development and testing of an innovative “practice-translation” project, the Savvy Caregiver Program (SCP). Modeled on the Minnesota Family Workshop (MFW), the SCP was designed to offer organizations a turn-key means of implementing our successful, academic-based caregiver psychoeducational program—without the direct involvement of the program initiators. We describe developing the SCP and field testing it with 140 caregivers in rural Minnesota; Denver, Colorado; and Anchorage, Alaska. Challenges to Creating a Transportable Intervention The original MFW program provided 7 weekly 2-hr training sessions to dementia caregivers and other family members. The sessions were designed and conducted by university faculty with disciplinary backgrounds in nursing, family therapy, education, and occupational therapy. A hallmark of the program was its emphasis on training caregivers for the unfamiliar work role into which they had been thrust. The program's curriculum was designed to help caregivers objectively appraise their situation (e.g., assess the progression of the disease in the care receiver and identify feasible caregiving goals) while providing them with the knowledge, skills, and attitudes needed to mediate their stress situation and more effectively carry out the caregiving role they have assumed. We recognized four major challenges in developing a transportable version of the MFW. The first was translating expertise. Teaching activities in the MFW curriculum drew on individual faculty knowledge and skills in the provision of tasks and exercises. The problem we faced was how to codify the teaching activities in a way that did not render them sterile and canned and yet did not require trainers and workshop leaders in other sites to have the same knowledge and skill set as the expert faculty. Moreover, because the program touches on sensitive issues, caregivers often sought additional information, causing segments to run longer than expected. In such circumstances, our faculty were able to draw on their well-developed sense of the program's content and pace to bring closure to the segment and to abridge future segments so as to fit all critical content into the session. Inherent in this problem was the recognition that providing individualized training to SCP workshop leaders would not be feasible. The second challenge was in transporting content. The MFW information base, presented in talks and exercises by four faculty members, had not formally been written. Although these instructors used overheads, they did not have verbatim texts of their talks. MFW participants were given manuals that contained some of the materials covered in the talks, but these also were not comprehensive. The third challenge was in maintaining program integrity. The MFW curriculum and order of presentation remained constant in the randomized trial. The final version had been refined over time in a series of previous field tests and beta versions. In developing the transportable SCP, we sought to preserve the program intact and particularly to ensure that the stress-mediation theoretical framework (Lazarus & Folkman, 1984) would be maintained. We also wanted to ensure that the workshop did not serve as a platform for any idiosyncratic permutations a local trainer might wish to use. The fourth challenge was in obtaining care recipient assessment. During the MFW, a parallel activity program for care recipients offered the opportunity for an occupational therapist to assess the cognitive performance status of the care recipients, using a modified Allen scale (Allen, 1988; Burns, Mortimer, & Merchak, 1994). A key part of the learning activity in the MFW involved participants' viewing a videotape of their care recipient being assessed and then learning the results of the assessment from the occupational therapist. For the SCP, we needed to find a way to teach caregivers to estimate their care recipient's cognitive performance abilities, short of having potential SCP sites hire an occupational therapist to conduct such assessments. Development and Piloting of SCP Curriculum and Materials In 1998, working with a staff member from the Metro Lakes chapter of the Alzheimer's Association of Minnesota, we developed a prototype version of the SCP, designed to overcome the challenges to transportability. The prototype was a 12-hr, self-contained dementia caregiver training program that could be offered in the community by governmental, educational, medical, or social service provider organizations and led by volunteer professionals from the sponsors. Table 1 outlines the key learning objectives of this SCP curriculum. The prototype version of the SCP included the following materials. Trainer's Manual This manual orients the workshop facilitator to the SCP and helps him or her to conduct the program. The manual introduces the facilitator to the intent of the program, emphasizing the tone he or she should strive to maintain (training, in contrast to a support group). The manual specifies the order of talks and exercises in each of six 2-hr sessions. It provides specific learning objectives for each segment of each session, describes how the segment fits into the session and the overall program, and gives detailed directions for conducting each segment. The manual provides the slides and handouts the instructor needs for each segment of the program (also given to the participants). Finally, the manual specifies the “homework” the instructor should encourage participants to engage in after each session. Caregiver's Manual The caregiver's manual provides a written version of the material conveyed in the workshop. Initially, the manual was designed to follow along with the instructional sequence of the program, so the sections were divided by where the material fell in the curriculum. As we discuss in what follows, the structure of this manual changed in the field trial. CD-ROM In a separate project, a CD-ROM, Alzheimer's Caregiving Strategies, was developed at the Minneapolis Veterans Affairs Medical Center (Adelson, Burns, Hepburn, Maddux, & Smith, 1999). This CD-ROM involved faculty from the MFW and drew on its curriculum for content. Equally, it drew on material that was developed at the Minneapolis Geriatric Research, Education, and Clinical Center program. Although the content of the CD-ROM parallels that of the MFW and the SCP, of particular interest is the cognitive performance estimation procedure that is embedded in the program. Using video clips of persons with dementia who are being tested in the performance of an everyday task (making toast), the CD-ROM introduces caregivers to the performance staging system we use and encourages them to estimate where their care recipient falls along the performance stages. Once the prototype material had been developed, two of the authors (K. Hepburn and C. Sherman) led a workshop for five family caregivers, using the materials and following the scripts. The participating caregivers provided detailed and extensive commentary on the program itself and the materials used in it. This pilot resulted in a revision of the curriculum (primarily a reshuffling, but in some cases a revising or a deletion, of materials and segments). It also resulted in an expansion of both the caregiver's and the trainer's manuals. Field Testing Local Field Testing The Metro Lakes chapter of the Alzheimer's Association recruited volunteers to serve as trainers for SCP workshops in their home organizations. These persons were from a variety of backgrounds such as public health nursing, social work, and therapeutic recreation. All had direct service experience with dementia patients and their caregivers. They received approximately 6 hr of training or orientation for the role to familiarize them with the pace and structure of the SCP, to show how certain exercises worked, and to illustrate the linkages between the trainer's and the caregiver's manuals. From the spring through the fall of 1999, three of these volunteer trainers conducted a total of five SCP workshops throughout the metropolitan area of Minneapolis–St. Paul, Minnesota. The volunteers recruited participants for the workshops and found community locations in which to offer the workshops. We observed the trainers conducting three complete workshops and portions of two others, providing us with firsthand experience in seeing trainers who were not invested in the original program attempt to teach from the materials we provided. By agreement with the volunteers, we did not take part in these workshops but observed in silence. We kept notes on what worked and what did not work in the program, particularly assessing the extent to which (a) the trainers appropriately conveyed the intent and meaning of each segment; (b) which segments of the program were effective; (c) the segment achieved our intended or expected result; and (d) the content was (could be) presented in the time available. At the end of each session and then again at the end of each of these programs, we asked program participants to complete evaluation forms. These forms sought the family caregivers' reactions to the individual segments within each session, to the teaching materials (slides and handouts), to the reading and homework assigned for the session, and, where appropriate, to the segments of the CD-ROM assigned for the session. We also sought more general responses and judgments about the trainers' style and delivery and about the value of the segments, sessions, and the program as a whole. The results of this initial field trial informed a substantial revision of the SCP materials. Overall, we tightened and removed elements from the curriculum program to sharpen the program's focus on caregiver role mastery. Acquiring the skills, knowledge, and attitudes important for success in the caregiver role became the central design feature of the program. Because the trainers, typically, did not get through all of the material in the sessions, we eliminated content we determined to be less than central to the purpose of the program. For example, we eliminated an exercise that sought to have participants work through ways of improving family interaction related to caregiving. We retained this content in the manual and the homework assignments, but it was no longer an in-class exercise. We also eliminated or simplified some of the handout materials and slides. Although the caregiver's manual remained essentially intact, we added sections on communication and difficult behaviors and carried out minor revisions to simplify the language and remove jargon from the text. We added text to the trainer's manual to better convey the intent of the segments and to provide fuller direction for their conduct. As a means of ensuring the inclusion of key material, we added directions for how much time training should spend on each segment in each session. Our observations reinforced the importance of using a specified amount of unstructured “debriefing time” at the beginning of each session to address issues and questions raised by the previous session or the homework and to review the results of experiential assignments. Expanded Field Test Next we initiated a field test of the SCP in settings outside the Twin Cities. One site, in southern Minnesota, provided linkages to a number of projects and agencies (including the regional Area Agency on Aging) with whose missions the SCP aligned. The second site, in Denver, Colorado, involved the Centura Health System Senior Clinic and the Rocky Mountain chapter of the Alzheimer's Association, both participants in a National Chronic Care Consortium/Alzheimer's Disease project seeking to implement a caregiver training program as part of the demonstration. The authors provided brief on-site training to the leadership of each site and have since provided problem-solving consultation by phone. The first implementation of this round of field testing of the SCP began in the fall of 2000. In the spring of 2001 we added a third field test site, the Alzheimer's Resource Center of Alaska in Anchorage. This site initiated programs without the use of any face-to-face training; the program leader read through the materials and began programs after only a brief phone consultation with one of the authors (K. Hepburn). This larger, three-site field trial was conducted in the framework of a quasi-experimental design, following review by the University of Minnesota Institutional Review Board. Each participating site recruited participants for SCPs in their area. Participants were informed that the organizations were taking part in a formative study of the program and were asked if they would consent to be part of the study. The organizations sent a list of these participants (and their signed consent forms) to our study center. We then sent participants a brief questionnaire, which we asked them to complete and return (using a stamped envelope we provided) prior to the beginning of the SCP at their site. In addition to basic background questions about the caregiver and care recipient (e.g., age, gender, relationship to care recipient, living situation, education, and income), the questionnaire included five brief, well-established scales. First, the Revised Memory and Behavior Problem Checklist (Teri et al.,1992) includes a caregiver rating of the extent and severity of care recipient behaviors in four domains (memory, behavior, paranoia, and depression); it also includes a caregiver self-assessment of the extent to which she or he is bothered by each of these behaviors. Second, the Caregiver Burden Scale (Zarit, Orr, & Zarit, 1985) seeks to assess the extent to which various facets of caregiving are problematic for caregivers. Third, we used the Center for Epidemiologic Studies-Depression scale (Radloff, 1977) to measure one element of the emotional aspect of caregiving. Fourth, we used a Mastery scale (Pearlin, Mullan, Semple, & Skaff, 1990) to record caregiver self-assessment of caregiving competence. Fifth and finally, we used the Beliefs About Caregiving Scale (Phillips, Rempusheski, & Morrison, 1989) to assess the emotional enmeshment of the caregiver. Approximately 3–4 months after the beginning of each program, we again mailed the same questionnaire to caregivers and asked them to complete and return it. At the end of each SCP workshop, the site leaders distributed evaluation questionnaires that sought caregivers' ratings of the benefits of the program (e.g., improvements in knowledge, confidence, or attitude), their ratings of the effectiveness of the program and its materials, and their suggestions for ways to improve the program. Participants were asked to complete the evaluations and mail them back to the study center (using stamped envelopes we provided). We did no phone or mail follow-up to increase our survey response rate, so we report on considerably fewer participants than total entrants into the program. All data were coded and managed at our study center. Expanded Field Test Results Participant Demographics In the year and a half of field trials held in Alaska, Colorado, and rural Minnesota, 22 separate SCP programs were offered at these sites by a wide variety of professionals—educators, nurses, social workers, recreational therapists, and even geriatricians. A total of 140 family caregivers took part; 40% were spouses of the care recipient, and 47% were adult children; 55% lived with the care recipient. Several paid caregivers (e.g., nursing assistants) also took part, but their data were not included in the analyses reported here, nor were those of participants in the earlier Twin Cities prototype development groups (local field test participants). Participants in the larger field trial were relatively well educated (approximately 70% had more than a high school education) and affluent (modal household income was $30,000–39,000). Almost all (95.7%) were non-Hispanic White. Evaluation Findings As the data in Table 2 indicate, participant response to the program, its materials, and the quality of the workshop trainers was very favorable. Virtually all respondents reported increased skill, knowledge, and confidence as a result of participation, and all would recommend the program to others in similar situations. The caregiver manual was also judged favorably. We note that, although they were still quite positive, the relevance items in Table 2 were scored lower than the other items, particularly in the “strongly agree” category. A closer examination of the data revealed that caregiving daughters tended to rate the program as slightly less relevant to them than did caregiving spouses. Of particular importance, the participants reported a high degree of linkage and concordance between the text in the caregiver manual and the material presented by the trainers. This gives us reassurance that the trainers were faithful to the curriculum of the program, as specified in the trainer's manual. Table 2 also provides caregivers' reports on the effectiveness of the trainers and the training process. These results suggest that the program can be taught effectively by trainers from a wide variety of professional backgrounds. Preintervention Versus Postintervention Findings Table 3 reports on the results of a preintervention versus postintervention t-test analysis comparison of caregivers on the small set of measures we used (taken prior to and 3–4 months after participation in the SCP). It is of note that on three important dimensions—caregivers' reaction to the overall behavior of the persons for whom they provide care (i.e., “total reaction”), their self-reported burden, and their beliefs about caregiving (emotional enmeshment)—significant changes occurred in directions that indicate better caregiver well-being. Discussion The principal finding of this effort is that it is feasible to “translate” a research-based caregiver intervention into a packaged program that can be adopted in other settings. The experience of dissemination suggests that the SCP is flexible enough to be facilitated by a variety of leaders with different areas of expertise. Results indicate that we met the four primary challenges we faced. Participant assessment of strong concordance between classroom activities and material in the caregiver manual indicates that the program is kept intact when other leaders implement it. The other evaluation data indicate that participation yields benefits that are similar to those found in the parent program. This encourages us to conclude that expertise was transported and program integrity was maintained. The participants' evaluation of the caregiver manual leads us to conclude that the transportation of content was successful. Finally, when caregivers used the CD-ROM-based assessment of their care recipient's performance stage, they judged it to be helpful and put the information to use. However, caregivers' access to computers limited the utility of the CD-ROM, a situation we have sought to resolve (see the paragraphs that follow). Thus far, we have learned three main lessons regarding implementation of the SCP. First, leaders have to be committed to the program and willing to lead within the parameters of a training paradigm. Given the SCP's training emphasis, workshop leaders have to be willing to refer caregivers to other resources to address their other needs (e.g., for information or support) and curtail impulses to transform the SCP into a counseling program. Second, leaders have to prepare to lead the program. It takes time to read through both manuals and to understand the flow and timing of the curriculum. It does not work well to try to teach directly from the manual during the sessions. The manual can cue the leaders about the exercises and talks during the sessions, but leaders have to be in possession of the material beforehand. Most facilitators report that the first time through can be challenging but that they gain mastery through this experience. Third, the SCP benefits from strong organizational support. Those leaders working in organizations in which aiding caregivers fits well and centrally with the mission found it easier to recruit caregivers and to offer the SCP as an integral part of their work. Thus, in Denver, where a natural linkage existed with a large clinical operation, the leaders found that caregivers were readily identified by clinicians and referred into the SCP. Since our field testing, SCP has continued to evolve. Using feedback from the field trial trainers, we completely revised the caregiver's manual. The book was restructured so it could be read more as a stand-alone text. The language, reading level, and printing format were carefully examined; the resulting text is less complex, easier to read, visually less dense, and more reader friendly. Although the CD-ROM was well received, it was not used by everyone. Lack of equipment (a CD-ROM-equipped computer) or computer skills made it inaccessible to some participants. At the same time, we heard from trainers that a component of the training related to decision-making skills was difficult to teach. To remedy both problems, we produced a video that incorporates key elements of the CD-ROM (particularly the segment on care recipient assessment), provides a guided lecture on decision making by one of the authors (M. Lewis), and includes a segment illustrating the care recipient involvement strategies that SCP teaches. The new training video has helped to make the program even more portable. Lastly, we found that a greater proportion of caregiving daughters took part in the SCP than in either of our NINR-supported trials. The evaluation responses suggest that we need to attend better to their situation, both in the curriculum and in the program materials. Thus, we are in the process of developing additional material for this caregiver group. Several limitations of this report should be acknowledged. Approximately 37% of those taking part in the SCP trials did not provide follow-up data. Our examination of the nonrespondent demographic data does not suggest any differences between the two groups. The positive findings we are able to report stem from a within-group, preintervention versus postintervention analysis. A randomized trial of the SCP is currently underway in sites in Alaska, Colorado, and Mississippi; this trial will provide a more rigorous test of SCP effectiveness. Despite these limitations, the SCP has demonstrated that it can offer a wide range of organizations the opportunity to provide dementia caregivers with training from which the caregivers themselves claim to experience the same kinds of benefits as those demonstrated in the randomized trial of the MFW. The SCP offers organizations a turn-key means to add to the repertoire of assistance they can provide to this important and growing group of family caregivers. This research was supported by grants from the UCare Minnesota Foundation (1009-98), the National Alzheimer's Association (IIRG-99-1591), and the National Institute for Nursing Research (NR02981). 1 School of Nursing, University of Minnesota, Minneapolis. 2 Department of Family Practice and Community Health, University of Minnesota, Minneapolis. 3 Screen, Inc., 8017½ 8th Avenue S., Seattle, WA. Decision Editor: Eleanor S. McConnell, RN, PhD Table 1. Caregiving Learning Objectives in the Savvy Caregiver Curriculum. Objective Description Acknowledge the disease Understand and come to grips with the disease that is affecting the person. Recognize it is not personal—it is the disease and not the person that is causing difficulties. Make the cognitive shift Develop a strategic sense of what cognitive losses are occurring and how caregiver behavior has to adapt to these as they take place and progress. Develop emotional tolerance Recognize the central role of confusion in dementia and how it contributes to troubling behaviors (e.g., shadowing or repetitive questioning). Appreciate the care recipient's need for emotional stability in the face of confusion. Accept the caregiver's role of providing calm and stability. Take control Understand that dementia gradually erodes autonomy. Appreciate the social and emotional difficulties involved in taking control of another adult (thinking of the other as somehow not equal). Recognize the need to take control and be willing and able to do so. Establish a realistic care goal Accept that striving to ensure as good a quality of life as possible is the most realistic goal for caregiving. Recognize that rehabilitation, restoration, or retardation of dis-ease progress are not feasible goals and let go of them, when applicable. Gauge the care recipient's capabilities Be able, using an occupational-therapy-based staging system, to estimate the care recipient's capacity for involvement in tasks and activities (kind, complexity, duration, etc.). Design opportunities for satisfying occupation Be able, based on the estimate of capability, to design tasks or activities in which the person can become engaged and which she or he will enjoy. Be able to communicate effectively with the person to promote and maintain involvement. Become a sleuth Adopt a problem-solving approach to caregiving, one that involves being able to stand back from the situation (emotionally), create hypotheses about causes of behavior, formulate and try out responses, observe and learn from results, and repeat the sequence as needed. Table 2. Results of Participant Evaluation of the Savvy Caregiver Program. Evaluation Response Category Strongly Agree (5) Agree (4) Total (4 + 5) Overall response to SCP (%) More knowledgeable as a caregiver 86.4 11.4 97.8 Learned useful caregiving strategies 85.2 11.4 95.6 Feel more confident as a caregiver 68.2 27.3 95.5 Have more caregiving skills 75.0 19.3 94.3 Program content relevant to caregiving 72.7 18.2 90.9 Would recommend to other caregivers 95.5 4.5 100.0 Evaluation of SCP caregiver manual (%) Clear and easy to understand 75.0 19.3 94.3 Information relevant to caregiving 68.2 22.7 90.9 Information helps in caregiving 75.0 18.2 93.2 Information ties in with sessions 84.1 12.5 96.6 Expect to use manual in the future 68.2 22.7 90.9 Evaluation of workshop trainers (%) Information (talks or exercises) clear 86.4 9.1 95.5 Trainers presented clearly and well 85.2 11.4 96.6 Trainers encouraged discussion 92.0 6.8 98.8 Class time used effectively 78.4 15.9 94.3 Trainers made information relevant 88.5 10.3 98.8 Note: For all evaluation response categories, N = 88. Table 3. t-Test Comparison of Caregiver Measures Taken Before and After Participation in the SCP. Measurea Score p value n Before After Reaction to care recipient behavior (−) 8.56 7.29 .075 56 Reaction to care recipient paranoia (−) 4.62 3.65 .104 56 Reaction to care recipient depression (−) 9.71 7.47 .023 57 Total reaction (−) 31.02 26.12 .020 56 Zarit burden (−) 40.40 38.13 .047 63 CES–Depression scale (−) 16.33 15.10 .186 63 Master scale—caregiving competence (+) 2.48 2.47 .941 62 Beliefs about caregiving scale (+) 47.75 49.85 .003 60 Notes: Time after participation in the program was 3–4 months. SCP = Savvy Caregiver Program; CES = Center for Epidemiologic Studies. aThe plus or minus sign following each measure indicates the direction of change sought through the intervention, thus a decrease in the reaction to behavior score and an increase in the mastery score would both be good outcomes. References Adelson, R., Burns, T., Hepburn, K., Maddux, M., Smith, S, 1999;. Alzheimer's caregiving strategies. Minneapolis, MN: Veterans Affairs Education Service. Allen, C. K., 1988;. Occupational therapy: Functional assessment of the severity of mental disorders. Hospital and Community Psychiatry,. 39:140-142. Alspaugh, M. E. L., Zarit, S. H., Stephens, M. A. P., Townsend, A. L., Greene, R., 1999;. Longitudinal patterns of risk for depression in dementia caregivers: Objective and subjective primary stress as predictors. Psychology and Aging,. 14:34-43. Arno, P. S., Levine, C., Memmott, M. M., 1999;. The economic value of informal caregiving. Health Affairs,. 18:182-188. Brodaty, H., McGilchrist, C., Harris, L., Peters, K. E., 1993;. Time until institutionalization and death in patients with dementia: Role of caregiver training and risk factors. Archives of Neurology,. 50:643-650. Brookmeyer, R., Gray, S., Kawas, C., ;. Projections of Alzheimer's disease in the United States and the public health impact of delaying disease onset. American Journal of Public Health,. 88:1337-1342. Buckwalter, K. C., Gerdner, L., Kohout, F., Hall, G. R., Kelly, A., Richards, B., et al 1999;. Nursing intervention to decrease depression in family caregivers of persons with dementia. Archives of Psychiatric Nursing,. 13:80-88. Burns, T., Mortimer, J. A., Merchak, P., 1994;. Cognitive performance test: A new approach to functional assessment in Alzheimer's disease. Journal of Geriatric Psychiatry and Neurology,. 7:46-54. Gallagher-Thompson, D., DeVries, H. M., 1994;. “Coping with frustration” classes: Development and preliminary outcomes with women who care for relatives with dementia. The Gerontologist,. 34:548-552. Gaugler, J. E., Davey, A., Pearlin, L. I., Zarit, S. H., 2000;. Modeling caregiver adaptation over time: The longitudinal impact of behavior problems. Psychology and Aging,. 15:437-450. Gendron, C., Poitras, L., Dastoor, D. P., Perodeau, G., 1996;. Cognitive-behavior group intervention for spousal caregivers: Findings and clinical considerations. Clinical Gerontologist,. 17:3-19. George, L. K., Gwyther, L. P., 1986;. Caregiver well-being: A multidimensional examination of family caregivers of demented adults. The Gerontologist,. 26:253-259. Hepburn, K., Tornatore, J., Center, B., Ostwald, S. W., 2001;. Dementia family caregiver training: Affecting beliefs about caregiving and caregiver outcomes. Journal of the American Geriatrics Society,. 49:450-457. Lazarus, R. S., Folkman, S., 1984;. Coping and adaptation. In W. D. Gentry (Ed.), The handbook of behavioral medicine (pp. 282–325). New York: Guilford. Max, W., Webber, P., Fox, P., 1995;. Alzheimer's disease: The unpaid burden of caring. Journal of Aging and Health,. 7:179-199. Mittelman, M. S., Ferris, S. H., Shulman, E., Steinberg, G., Levin, B., 1996;. A family intervention to delay nursing home placement of patients with Alzheimer disease: A randomized controlled trial. Journal of the American Medical Association,. 276:1725-1731. National Alliance for Caregiving, American Association of Retired Persons., 1997;. Family caregiving in the U.S.: Findings from a national survey (Final Report). Bethesda, MD: Author. Ostwald, S. K., Hepburn, K. W., Caron, W., Burns, T., Mantell, R., 1999;. Reducing caregiver burden: A randomized psychoeducational intervention for caregivers of persons with dementia. The Gerontologist,. 39:299-309. Pearlin, L. I., Mullan, J. T, Semple, S. J., Skaff, M. M., 1990;. Caregiving and the stress process: An overview of concepts and their measures. The Gerontologist,. 30:583-594. Phillips, L. R., Rempusheski, V. F., Morrison, E., 1989;. Developing and testing the Beliefs about Caregiving Scale. Research in Nursing and Health,. 12:207-220. Radloff, L. S., 1977;. The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement,. 1:385-401. Schulz, R., Beach, S. R., 1999;. Caregiving as a risk factor for mortality: The caregiver health effects study. Journal of the American Medical Association,. 282:2215-2219. Schulz, R., O'Brien, A. T., Bookwala, J., Fleissner, K., 1995;. Psychiatric and physical morbidity effects of dementia caregiving: Prevalence, correlates, and causes. The Gerontologist,. 35:771-791. Stone, R., Cafferata, G. L., Sangl, J., 1987;. Caregivers of the frail elderly: A national profile. The Gerontologist,. 27:616-626. Teri, L., 1997;. Behavior and caregiver burden: Behavioral problems in patients with Alzheimer disease and its association with caregiver distress. Alzheimer Disease and Associated Disorders,. 11:S35-S38. Teri, L., Truax, P., Logsdon, R., Uomoto, J., Zarit, S., Vitaliano, P. O., 1992;. Assessment of behavioral problems in dementia: The revised memory and behavior problems checklist. Psychology and Aging,. 7:622-631. Zarit, S., Orr, N., Zarit, J., 1985;. The hidden victims of Alzheimer's disease: Families under stress. New York: New York University Press.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B8CE3F28C7ECF202C365FB46B0D7DF0825FD78C.txt b/test/dataset/in/resources/corpus/Clean_0B8CE3F28C7ECF202C365FB46B0D7DF0825FD78C.txt
deleted file mode 100644
index 94fa14f..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B8CE3F28C7ECF202C365FB46B0D7DF0825FD78C.txt
+++ /dev/null
@@ -1 +0,0 @@
-]>AGG997S0167-4943(99)00048-510.1016/S0167-4943(99)00048-5Elsevier Science Ireland LtdFig. 1(a) The structure of the hip protector. (b) The appearance of the patient with hip protector.Fig. 2Profile of fall (orientation and time).Table 1Summary table of falls and fractures dataControlTreatmentCombinedFracture groupFalls10119129232Residents3140718aAverage falls per resident3.264.784.114Fractures639aOne resident had two fractures (one on each side).Effectiveness and acceptability of a newly designed hip protector: a pilot studyDaniel KChana*d.chan@unsw.edu.auGaryHillierbMichelleCoorebRosemaryCookebRebeccaMonkbJanetteMillscWai THungdaDepartment of Geriatric Medicine, Prince of Wales Hospital, High Street, Randwick 2031, NSW, AustraliabAged Care Assessment Team, Community Health Services, Orange, NSW, AustraliacDepartment of Orthopaedics, Prince of Wales Hospital, High St., Randwick 2031, NSW, AustraliadFAR, School of Mathematical Sciences, University of Technology, Broadway 2007, NSW, Australia*Corresponding author. Tel.: +61-2-93824242; fax: +61-2-93824241AbstractHip fracture has a significant economic and personal cost, involving hospital admission and functional impairment for elderly people. To assess the benefit of using a newly designed hip protector (new material and new design) to prevent fracture in a realistic setting, a randomised intervention-control design was used to trial the effectiveness of pads worn by high falls risk residents (n=71) in nursing home for 9 months. 40 residents were in the intervention group and 31 were in the control group. A profile of falls, including time of day, and orientation was obtained to demonstrate the potential effectiveness of the protectors for injury prevention. Acceptance of the hip protector was also surveyed amongst nursing home staff and residents. One hundred and one falls and six fractures occurred in the control group. In contrast, one hundred and ninety one falls and three fractures occurred in the hip protector (pads) group. The three fractures in the protector wearing group occurred when pads were not in place. This was extrapolated as 1 in every 16.8 falls and 1 in every 63.7 falls resulting in fracture in the two groups, respectively. The relative risk of fracture was 0.264 (95% CI=0.073–0.959) when the fracture incidence rate in the intervention group (three fractures per 191 falls) was compared to the control group (six fractures per 101 falls). This is a statistically significant result and implies that this newly designed hip protector is effective in preventing hip fracture. The majority of falls occurred during the day, which was when protectors were worn in this study, but the data on orientation was incomplete, with direction unknown in 74% of falls. Compliance was an issue, which was interpreted as only 50.3% of falls recorded with protectors in place. Dementia was identified as the explanation for this as the pads were often removed by these residents who comprised the majority of participants. Perception of low risk was the primary barrier to residents accepting the intervention. Comfort of protectors was not a significant concern for staff or residents, and only staff described appearance as an issue. In conclusion, the newly designed hip protector is protective against fractures in a realistic setting. Compliance and acceptance of the protectors will ultimately determine the viability of this prophylaxis.KeywordsNovel hip protectorHip fractureDementia1IntroductionFracture of the hip has been recognised as a significant problem in the elderly population from both an economic perspective, and in terms of the personal impact on the individual experiencing the injury. The impairment in function associated with the fracture, and surgical intervention alters the lives of individuals, increasing the likelihood of permanent institutionalisation five times for previously community dwelling elderly (Cumming et al., 1996). Pre-fracture levels of physical functioning are not regained by more than half the injured patients, and fractured neck of femur increases mortality in the elderly significantly (Butler et al., 1996; Fox et al., 1996).In the majority (60–90%) of fractures direct trauma is the aetiology (Cummings and Nevitt, 1989), which indicates the importance of addressing falls in the elderly. However, this is a complex issue as most falls are multifactorial in cause, including, lower limb weakness, gait instability, vision, cognitive and functional impairments as well as the effects of medication (Rubenstein et al., 1994). External factors have been found to be unreliable determinants of risk, as multiple hazards have been implicated in fractures (Clemson et al., 1996). Therefore, the value of a secondary prophylaxis is suggested as a fracture prevention, to mediate the impact of the fall, when it inevitably occurs.Hip fracture risk is known to be reduced by a larger body mass index as increased soft tissue appears to absorb the impact of a fall and protect the femur (Hayes et al., 1991; Robinovitch et al., 1995). Even in a direct fall onto the hip in the elderly, research has shown that natural padding is insufficient to attenuate the force, and other mechanisms such as reflex contraction of the quadriceps protect the hip fracture (Robinovitch et al., 1995).As a solution to the prevention of fracture, several clinical trials have demonstrated that external hip protector is effective in protecting the hip, as no fractures occurred in subjects while wearing the appliances (Lauritzen et al., 1993; Tracey et al., 1998). However, the compliance rate is only about 50% (Lauritzen et al., 1993; Tracey et al., 1998) and it drops to about 30% (Tracey et al., 1998) in the long term. The design of these hip protectors are such that they have to be worn with underpants to cover a small area over the greater trochanter. Any movement of the protector may mean that the covering position may be shifted. Therefore, the hip protectors are quite tightly fitted and this may explain a significant proportion of people feeling the discomfort (Tracey et al., 1998). Furthermore, the polypropylene material used is quite hard which may add to the discomfort.We therefore designed a new hip protector using softer material, which is fitted to the inner surfaces of trousers or tracksuit pants. The hip protectors also cover the greater trochanters but have the propensity to allow for some movement, and hence may cause less discomfort. Our aims are to find out both the effectiveness and acceptability of this newly designed hip protector.2Patient and method2.1ParticipantsSubjects were drawn from nine nursing homes and the criteria for inclusion was that the nursing home staff identified the residents as high falls risk. This was not on the basis of particular diagnosis, or any formal evaluation of falls risk, but the perception of the staff themselves. The subjects or the person responsible was then asked to sign consent to participate. Therefore, the purposes of the study were clear to participants and/or the guardians, and involvement was voluntary. Random assignment of subjects was achieved in most nursing homes with some participants designated as control and some to wear the protectors. The current results are based on 71 participants (control=31, treatment=40) on whom data has been collected after 9 months.2.2Hip protectorThe hip protectors were designed to absorb the impact in the upper femur, hip region, in particular to cover the greater trochanter in a fall. They are made of pads as shown in Fig. 1. The pads are worn in pockets sewn into the inner surfaces of tracksuit pants or trousers. Pads are 2×3 rows of cube shapes with dimensions 6 (width)×7 (length)×2.5 (depth) cm in each cube. The material is made from EVA foam. This material is waterproof and the shock absorbency is demonstrated through the successful use in Tai Kwan Do matting. The mould used is also jointed allowing flexibility and comfort. Therefore, this new hip protector is different from that by Lauritzen (Lauritzen et al., 1993) in both material and design.2.3Measurement2.3.1Effectiveness of protectorsA simple form was designed for inclusion in residents files which staff were asked to complete for every fall. Important information, which would indicate the effectiveness of the protector, was injuries, orientation of the fall, and the time of day, as protectors were not worn at night in bed in this current study.2.3.2Acceptance of protectorsCompliance was extrapolated from the percentage of falls recorded for which protectors were worn.Separate surveys were designed for staff and residents, with closed responses to enhance acceptability to nursing staff in terms of time constraints. However, the survey was administered verbally and response options were not designed to quantify responses, but facilitate those actually using the protectors to articulate their opinion of their acceptability.2.4Data analysisThe calculation of the fracture incidence rate ratio (or relative risk) was as follows:three fractures per 191 falls (hip pad group)/six fractures per 101 falls (control group).This is based on the test-based method by Sahai and Khurshid (Sahai and Khurshid, 1996).3Results3.1Effectiveness of protectorsFifty of the 71 subjects had fallen with a total 292 falls recorded. Nine fractures occurred with none of the eight residents wearing protectors at the time of injury (one resident fractured both hips 1 month apart). 101 of the falls and six of the fractures were recorded in the control group. This may be interpreted as a risk of 1 in every 16.8 falls resulting in a hip fracture. The protector group in comparison accounted for three of the fractures and 191 of the falls with an extrapolated fracture risk of one in every 63.7 falls. The relative risk of fractures in the hip protector group as compared with the control group was 0.264 (95% CI=0.073–0.959). The average falls for control, protector, total and fracture groups is shown in Table 1.The orientation of falls for those resulting in fractures was that three were unknown, one was backwards, and five were sideways. All occurred during the day.In the total study group 81.8% of falls occurred during the day (6:00–22:00 h) and 9.6% were sideways. As 74% of falls were of unknown orientation, it is possible that the protectors may have been more useful in mediating impact than indicated. A profile of falls (orientation and time) is shown in Fig. 2.3.2Acceptance of protectors3.2.1ComplianceIn this study, compliance was defined as the percentage of falls recorded for which hip protectors were worn in the treatment group. The compliance of the participants was 50.3%.In the staff survey dementia was indicated as the reason for non-compliance. 64.8% of the total 71 participants have been identified as having dementia.3.2.2Perceptions of protectorsThe surveys were administered to seven staff and four residents. As most participants were diagnosed with dementia this limits the reliable resident sample. Three of the residents surveyed had been nominated as suitable for the study, but declined to participate. Comments from staff and residents noted informally have also been documented.3.2.3ComfortOnly one staff member noted concern about the comfort suggesting the seams on adapted underwear had rubbed. Also only one resident had serious concerns regarding the protectors highlighting the bulkiness and awkwardness, especially when continence pads are already worn. The tracksuit pants were also an issue for hot weather, so alternative clothing fitted with the protectors was preferred.3.2.4CostSince a pair of protectors cost only $10.00 AUD ($30 with tracksuit pants), no one was unable to afford it.3.2.5AppearanceFifty seven percent of staff described this as concern. However, no residents indicated it was an issue. The obvious bulkiness of the protectors was noted, and one staff member commented public dignity was compromised, such as at a doctors appointment.3.2.6Barriers to acceptanceTwo of the staff indicated they would have serious concerns regarding the use of protectors in resident management. The reasons suggested included, hygiene and the inconvenience of putting them on, especially when resident non-compliance was an issue. Enthusiasm for the protectors appeared to have a direct relationship with perceived effectiveness, as one staff member was sufficiently convinced by their own observation of the protectors, that they would like to continue their use beyond the study. Another respondent indicated that implementation of the protectors would ultimately depend on resident feelings.Residents explanations for not using the protectors centred on a perceived lack of personal risk. This included the belief in the two respondents already post fracture, simply that as they had experienced one fractured hip, they were now safe, and an indication from the doctor that the first fracture was pathological, therefore the resident felt external protection would be of no benefit in preventing fracture. Another resident noted that she had ridden horses for years, and would have needed them more then, as would other people in higher risk occupations, than simply being in danger of falling. Independence to manage their own risk was also an issue as one resident felt she was now too old to care, and another asserted the right to refuse intervention and experience the consequences of their own decision. The one resident surveyed who was wearing the protectors queried the position of the protectors for effectiveness, but was otherwise satisfied with the intervention. His daughter actually commented on the peace of mind provided by knowing he was wearing the protectors.4DiscussionUltimately the aim of this study was to implement protectors in a realistic setting and record data as to their effectiveness. This data has supported the value of pads worn to prevent fracture of the hip as no fractures occurred while they were in place. Furthermore, the reduction of relative risk of hip fracture in this new hip protector (RR=0.264) is comparable or marginally better than the old design (RR=0.44) by Lauritzen (Lauritzen et al., 1993). As most falls occurred during the day, this suggested that current wearing regime should be sufficient, it is the actual compliance with wearing the protectors that is problemsome. 50.3% is a low compliance, even allowing for the minority of falls occurring at night when the protectors were not worn (i.e. compliance rate is actually higher if night time falls are excluded from the calculation). However, this compliance rate is consistent with other designs of hip protectors (Lauritzen et al., 1993; Tracey et al., 1998). Dementia is the most significant factor in this study regarding compliance, which is an area that perhaps should be addressed in the design of the pads so they are not able to be removed easily and are less obvious to the person as their presence appears to irritate these residents.Unfortunately the low compliance could also reduce the strength of the fracture risk findings comparing the treatment and control groups. Although compliance was related to dementia in this research, the surveys regarding acceptance have gauged the importance of attitudes to the protectors in actually implementing them as an intervention tool.As for the perception and acceptance survey, the sample used was small, especially of residents. However, the diversity of responses even in this group suggests the difficulty of assessing attitudes and marketing the protectors to the consumer. In terms of staff it is interesting that they accorded more concern regarding the appearance of the protectors than the residents. It is the actual perception and understanding of fracture risk that appears to be the central barrier to acceptance amongst residents. This supports the value of education to the elderly who are high risk, such as frequent fallers or post fracture clients, in understanding the risk not simply of falling, but of fracture (Cameron and Quine, 1994). Therefore, the preferred management of the problem of fractured neck of femur would include not just falls prevention but fracture protection.The findings of this study were limited to a nursing home population where protectors have had been implemented and monitored by staff. Therefore, issues such as ease of use for the client have not been explored which could be a significant difficulty for community living elderly with mobility impairments. A similar study using community living individuals would be appropriate, not to validate the actual effectiveness of protectors but their practicality.In terms of the participant group selected as suitable for the research the variables were not monitored by the researcher, it was based on the subjective opinion of staff. However, the comparison in the average falls per year per bed of 1.5 in institutions to the 4.11 recorded for this group in a 9-month period, suggests that staff are accurate in determining high falls risk residents. Another point of note is the higher rate of fall (4.78 falls per bed per 9 months) in treatment group compared to controls (3.26) may be due to willingness of staff to allow patients to ambulate after they are put on hip protectors (rather than restricting them).The hip fracture per fall rate in the control group is 5.94% (six fractures in 101 falls). This is considerably higher than fracture rates of elderly in the community but is in accord with findings from Butler et al. (Butler et al., 1996) (the risk of hip fracture in institutions is 101.5 times those living in private homes).Since a lot of the fallers are demented people whose rehabilitation and functional outcome after a fracture hip is poor, this raises the ethical question as to their rights to choose or deny wearing the hip protectors. As our study actually includes proxy consent from carers or guardians, this raises the important question as to whether the important information about hip protectors should be focused on carers or guardians, not just elderly fallers.The price of a pair of protectors is only $10.00 AUD, making it possible to be affordable to nearly all pensioners. Cost was a concern in some other protectors (Cameron and Quine, 1994).5ConclusionThe descriptive data about falls and fractures in a realistic trial in nursing homes indicates the viability of protectors as a valid instrument in the prevention of fractured hips as none occurred while the device was in place. However, further information regarding the profile of falls, especially direction of fall, would increase the evidence for the potential effectiveness of this device. Potential barriers to implementation include compliance, and attitudes to perceived risk.AcknowledgementsThe authors express their gratitude and thanks to Calare Nursing Home, Orange; Cudal Memorial Hospital; Eugowra Memorial Hospital; Mater Misericordiae Nursing Home, Forbes; Moyne Eventide Home, Canowindra; Niola Nursing Home, Parkes; Wontama Nursing Home, Orange; Sir Joseph Banks Nursing Home, Botany; Camelot Nursing Home, Maroubra for their co-operation and help.ReferencesButler et al., 1996M.ButlerR.NortonL.TrevorA.ChengA.J.CampbellThe risk of hip fracture in older people from nursing homes and institutionsAge Ageing251996381385Cameron and Quine, 1994I.CameronS.QuineExternal hip protectors: likely non-compliance among high risk elderly people living in the communityArch. Gerentol. Geriatr.191994273281Clemson et al., 1996L.ClemsonR.G.CummingM.RolandCase-control study of hazards in the home and risk of falls and hip fracturesAge Ageing25199697101Cumming et al., 1996R.G.CummingR.KlinebergA.KatelarisCohort study of risk of institutionalisation after hip fractureAust. NZ J. Public Health201996579582Cummings and Nevitt, 1989S.R.CummingsM.C.NevittA hypothesis: the causes of hip fracturesJ. Gerontol.441989M107M111Fox et al., 1996K.M.FoxG.FelsenthalJ.R.HebelS.I.ZimmermanJ.MagazinerA portable neuromuscular function assessment for studying recovery from hip fractureArch. Phys. Med. Rehabil.771996171175Hayes et al., 1991W.C.HayesE.R.MyersL.A.MaitlandN.M.ResnickL.A.LipsitzS.L.GreenspanRelative risk for fall severity, body habits and bone density in hip fracture among the elderlyTrans. Orthop. Res. Soc.161991139Lauritzen et al., 1993J.B.LauritzenM.M.PetersonB.LundEffect of external hip protectors on hip fracturesLancet34119931113Robinovitch et al., 1995S.N.RobinovitchT.A.McMahonW.C.HayesForce attenuation in trochanteric soft tissues during impact from a fallJ. Orthop. Res.1319955662Rubenstein et al., 1994L.Z.RubensteinK.R.JosephsonA.S.RobbinsFalls in the nursing homeAnn. Intern. Med.1211994442451Sahai and Khurshid, 1996H.SahaiA.KhurshidStatistics in Epidemiology: Methods, Techniques and Applications1996CRC PressBoca Raton, FL171174Tracey et al., 1998M.TraceyA.VillarP.HillH.InskipP.ThompsonC.CooperWill elderly rest home residents wear hip protectorsAge Ageing271998195198
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B937F9461CCAEEA575B6A276E7894E53C3759CC.txt b/test/dataset/in/resources/corpus/Clean_0B937F9461CCAEEA575B6A276E7894E53C3759CC.txt
deleted file mode 100644
index 255757c..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B937F9461CCAEEA575B6A276E7894E53C3759CC.txt
+++ /dev/null
@@ -1 +0,0 @@
-gerontgerontThe Gerontologist1758-53410016-9013Oxford University Press10.1093/geront/gnp031INEQUALITY AND AGEISMManaging Age Discrimination: An Examination of the Techniques Used When Seeking EmploymentBergerEllie D.PhD122Department of Sociology, Nipissing University, North Bay, Ontario, Canada1Address correspondence to Ellie D. Berger, PhD, Department of Sociology, Nipissing University, 100 College Drive, North Bay, Ontario, Canada P1B 8L7. E-mail: ellieb@nipissingu.caDecision Editor: Nancy Schoenberg, PhD62009174200949331733221120071382008© The Author 2009. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org.2009Purpose: This article examines the age-related management techniques used by older workers in their search for employment. Design and Methods: Data are drawn from interviews with individuals aged 45–65 years (N = 30). Results: Findings indicate that participants develop “counteractions” and “concealments” to manage perceived age discrimination. Individuals counteract employers’ ageist stereotypes by maintaining their skills and changing their work-related expectations and conceal age by altering their résumés, physical appearance, and language used. Implications: This research suggests that there is a need to reexamine the hiring practices of employers and to improve legislation in relation to their accountability.AgeismHiring practicesManagement techniquesOlder workersUnemploymentIt is widely recognized that the Canadian population is aging, and as a result, there are a wide number of social and economic changes expected to occur in the coming years. In particular, changes in the domain of work will be considerable as individuals aged 45–65 years are estimated to represent close to half of the working-age population by the year 2015 (Forum of Labour Market Ministers, 2002). Mandatory retirement was eliminated in Ontario in 2006, and thus, increasing attention has been paid to the aging of the workforce by both lay and academic audiences. In fact, the Canadian government has been trying to encourage older workers to remain in the labor force longer through various policy changes (Policy Research Initiative, 2005). Despite these changes, the average age of retirement in Canada has been decreasing from 64.9 years in 1976 to 61.4 years in 2005 (Conference Board of Canada, 2006).Further, industries in Canada, such as manufacturing, have had large layoffs in recent years and switched their focus to more contract and part-time work, which leaves older workers with little protection under employment standards law and employment insurance (McMullin, Cooke, & Downie, 2004). In fact, it has been recognized that recent changes in the Canadian economy have led to the need for workers to be geographically, occupationally, and industrially mobile as well as invest more heavily in job searches, training, and education (Secretariat for the Expert Panel on Older Workers, 2007). Thus, displaced older workers face the most challenges due to their lack of mobility, education, and training and due to the existence of age discrimination. These factors cause older Canadian workers to have a reduced chance of finding reemployment once unemployed. Further, older workers take considerably longer than younger ones to find work once unemployed due in part to the existence of age discrimination (Ontario Human Rights Commission, 2000; Walker, 2002). In fact, as age increases, so does the average duration of unemployment for working-age Canadians. Individuals aged 15–24 years spend an average of 7.9 weeks unemployed, those aged 25–54 years spend 17.9 weeks unemployed, and 55- to 64-year-olds spend an average of 22.8 weeks unemployed (Statistics Canada, 2007). This finding is similar to that in the United States, where individuals aged 16–19 years spend an average of 11.2 weeks unemployed and those aged 20–24 years spend 14.4 weeks unemployed on average (U.S. Bureau of Labor Statistics: Division of Labor Force Statistics, 2007). This pattern continues, where number of weeks spent unemployed increases as age increases, and 45- to 54-year-olds spend an average of 21.2 weeks unemployed and 55- to 64-year-olds spend an average of 21.9 weeks unemployed.Therefore, despite the growing number of older individuals and policy changes that encourage them to remain in the workforce longer, labor shortages will still occur due to their tendency to retire earlier (both voluntarily and involuntarily due to discrimination) and to their poor retention and recruitment. Although an increasing amount of research has been focused on policy changes related to the aging workforce and quantitative studies examining employers’ attitudes toward older workers (reviewed subsequently), there is a lack of research on how various policies and practices (both positive and negative) affect older workers. Specifically, little is known about the hiring practices of employers from the older worker's perspective and how these practices influence older workers when searching for employment. Therefore, the purpose of this article is to examine older workers’ perceptions of employers’ attitudes toward them and their management of these (mostly) negative attitudes.Literature ReviewEmployers' Attitudes Toward Older WorkersEmployers’ first impressions of older workers are often based on stereotypes—“mental pictures of people based on their membership in a group” (American Association for Retired Persons [AARP], 1994, p. 8). Although research suggests positive attitudes toward older workers on certain dimensions (knowledge, experience, honesty, dependability, work ethic, mentoring, and stability; Berger, 1999; Marshall, 1996; Walker & Taylor, 1998), they are perceived negatively in many other areas. For example, studies indicate that employers believe older workers are less flexible, in poorer health, less creative, less interested in technological change, and less trainable than are younger workers (Chiu, Chan, Snape, & Redman, 2001; Guillemard & Walker, 1994). In addition, older workers are seen as being more prone to accidents and staying with the company for a shorter period of time as compared with younger workers. Older workers are also viewed by employers as less productive and less able to engage in physically demanding work than younger workers (Henkens, 2005).Although most of these stereotypes have been disproved in the literature, employers continue to believe many of them and often make their hiring decisions accordingly (Berger, 1999, 2004). The key reasons that employers cite for not hiring older workers relate to beliefs concerning their lack of computer literacy, followed by lack of other appropriate skills, lack of appropriate qualifications, and difficulty adapting to varying work settings (Marshall, 1996). Employers’ stereotypical attitudes have also been linked to poor opportunities for training, promotion, and retention of older workers (Chiu et al., 2001). Another study that specifically examined the relationship between employers’ attitudes and their behaviors toward older workers also found a direct association between the two in relation to hiring, training, and advancement opportunities (Taylor & Walker, 1998). Further, Taylor and Urwin (2001) found that older workers participate less than younger workers in employer training programs due in large part to an age bias in employers’ decision making. In addition, stereotypical attitudes of employers have been tied to negative views concerning the retirement or retention of their older employees (Henkens, 2005). In fact, employers who have less personal contact with older workers have more negative attitudes toward them than employers who have more personal contact with older workers. This could suggest that although employers may be positive toward older workers already employed by their organization, they are negative toward the recruitment of new older employees. Employers’ attitudes toward older workers also vary significantly according to company size, employers’ age, and employers’ gender, with older female employers from smaller companies displaying the most positive attitudes (Berger, 1999).Although there is a growing body of literature on employers’ attitudes toward older workers, there is a lack of research that examines older workers’ perceptions regarding ageism. However, according to a national American work and career survey of more than 2,500 individuals aged 45–74 years, 80% of individuals who are searching for employment feel that age discrimination exists in the workplace (AARP, 2002, p. 66). Also, 15% of older individuals feel that they were not hired for a particular job because of age (AARP, 2002, p. 13). Findings from the survey speculate that older workers’ perceptions about age-related barriers to reemployment are a reality due to the increasing number of complaints about age discrimination made to the Equal Employment Opportunity Commission. In fact, between 2000 and 2002, complaints increased by 25% (AARP, 2002, p. 14).Managing Attitudes and BehaviorsConcepts that are taken from the symbolic interactionist framework are used in this article to gain a meaningful understanding of individuals’ motives, behaviors, and interactions (Blumer, 1969). In this perspective, individuals are seen as having agency and as engaging in mindful interaction and self-reflexive behavior. Therefore, individuals act, they do not simply respond, and they make choices within the constraints of the broader social context. Under the symbolic interactionist umbrella, Goffman (1959, 1963) has demonstrated the importance of stigma, identity, self-presentation, and impression management through his dramaturgic perspective. The concept of stigma refers to “an attribute that is deeply discrediting” (Goffman, 1963, p. 3) and results in discriminatory attitudes and behaviors being directed toward the stigmatized individual. More recently, the conceptualization of stigma has been redefined as existing “when elements of labeling, stereotyping, separating, status loss, and discrimination co-occur in a power situation that allows these processes to unfold” (Link & Phelan, 2001, p. 382).In relation to identity, Goffman (1963) described three types of identity that are used in the developmental process: “felt identity,” the meaning attributed to one's own identity; “presented identity,” the view of oneself that is projected to others; and “social identity,” the meaning that others attribute to the individual. He argued that one's presented identity can be altered to meet the needs of any given situation. By using certain techniques, individuals actively control the image they are projecting to others (presented identity), to obtain a desirable social identity. Goffman's (1959, 1963) use of the term “impression management” explains how individuals present themselves to others by conveying certain images that they consider appropriate in a given situation. In this way, individuals can manipulate the presentation of themselves displayed to others and create a distinct “definition of the situation” (Thomas, 1972) to sway others’ perceptions of them in a positive direction. For example, in their sociological study of the process of becoming doctors, Haas and Shaffir (1991) discussed medical students’ “constant need to create and manage the image of a competent self through the process of impression management” (p. 74). Thus, the importance of managing self-presentations is a well-developed concept in many realms of the sociological literature. With respect to aging, research on older widows has shown that to avoid being stigmatized and maintain a positive identity, the women “suppressed” evidence of age (Matthews, 1979, pp. 74–75). “People have time allotted to tend to physical and mental needs . . . to cover potential age stigma signs, such as wrinkles or gray hair, by surgical intervention or hair dyes” (Luken, 1987, p. 185). Thus, research suggests that individuals use certain management strategies to avoid being classified as “old” and being stigmatized by others.Studies indicate that impression management tactics are often used in the context of a job interview (Delery & Kacmar, 1998; Rosenfeld, 1997). More specifically, Delery and Kacmar showed three types of impression management techniques that are useful in an interview setting: (a) entitlements (taking credit for a prior work-related success), (b) enhancements (making statements reflecting one's own positive attributes), and (c) self-promotion (highlighting strengths in relation to those required by the ideal applicant). Their research considered the importance of applicants’ age in an interview setting but only included participants with a maximum age of 38 years and considered age to be an advantage in relation to having more experience with the interview process itself. Other research indicates that individuals use impression management strategies when they perceive that discrepancies exist between the feedback received in an organizational setting and their desired social identity (Bozeman & Kacmar, 1997). When discrepancies do occur, individuals acquire alternative techniques to use and then wait for additional feedback, which determines the future interaction between the participants. Accounts (Scott & Lyman, 1968) and disclaimers (Hewitt & Stokes, 1975) are also used in the employment context (Bozeman & Kacmar), to avoid negative stigmatization and negotiate identity in relation to past and future behaviors. Thus, although previous research has not focused on older workers seeking employment, it does suggest that certain strategies are particularly useful for enabling individuals to focus the interview on their strengths and positive attributes.Despite these findings, there is a lack of research that seeks to understand older workers’ experiences during the job search process. This article is part of a larger research project, which examined the meaning and import of age in the job search process (Berger, 2004). The majority of participants in this project experienced ageism when looking for work. The present article examines how these participants manage this ageism when searching for employment. To do this, I draw on data from 30 semistructured interviews with older unemployed workers. More specifically, this qualitative study addresses the following research questions: Once older job hunters perceive they have been discriminated against by potential employers, do they develop specific management techniques that they feel help them in the job search process? Are these techniques similar to the ones used by younger workers or are there specific age-related techniques that develop during the job search process?MethodsThis study used a qualitative methodological approach that allowed for a direct examination of older workers’ experiences in the job search process. The use of qualitative methodologies allowed for the subjective meanings and interpretations of my participants’ experiences to be revealed in a way that is often not possible through quantitative methodologies (George, 1990). This study also used the grounded theory approach (Glaser & Strauss, 1967) in an attempt to let the data generate the theory, and thus, initial research questions were reformulated as the study progressed to better reflect the emerging data.The data for this study were drawn from semistructured interviews with 30 unemployed individuals aged 45–65 years. Individuals became unemployed for a variety of reasons, including corporate downsizing, layoffs, or desire to leave their previous place of employment. I also conducted 35 hrs of participant observation in three older worker programs in the greater Toronto area to help provide a context for older workers’ experiences. Human Resources Development Canada (1999) created these programs to assist older workers in their job search process by providing them with specific workshops to meet their needs. These workshops covered topics such as interviewing skills, personality assessments, computer training, résumé writing, and specific sessions on understanding the myths and realities of being an older worker. I recruited most of my respondents by attending these workshops (n = 20) and posting recruitment advertisements in older worker programs (n = 4) and in employment agencies geared to all age groups (n = 0). To obtain participants who may not have sought out specific employment assistance (anticipating that this may lead to distinct results), recruitment advertisements were also placed in community centers (n = 1) and libraries (n = 0) in a major Canadian city, in a monthly newspaper aimed at individuals aged 55 years and older (n = 3), on a university Web site (n = 0), and through personal referral techniques (n = 2). To be included in my study, participants had to be between the ages of 45 and 65 years (in congruence with the definition of older workers given in the majority of literature), actively searching for work, and unemployed for 3 months or longer. (The stipulation that individuals must be unemployed for 3 months or longer was used because individuals unemployed for less than this may still be highly distressed from losing their jobs and may not have had the opportunity to develop age-related management techniques for their job search.)The interviews conducted in this study lasted from 45 min to 2 hrs. The first 8 interviews were conducted in 1999. An additional 22 interviews were done in 2002 to expand the initial research project. (As the unemployment rates were similar in 1999 and 2002 [7.6% and 7.7%, respectively], I do not believe that the economic conditions differentially affected participants’ ability to secure employment in the two time periods [Statistics Canada, 2004a, 2004b]. Further, due to the qualitative nature of this project and the related focus on individuals’ meanings that were attributed to their experiences, the fact that data were collected at two different time points is not perceived to have altered the findings of my study.) Prior to conducting the interviews, the study was approved by the McMaster University Research Ethics Board. Interviews were held at a variety of locations (including older worker programs, community centers, libraries, and coffee shops), depending on individuals’ personal preferences. A semistructured interview format was used to guide the interview along several key areas, as well as to provide enough flexibility for the participants to discuss issues freely that arose during the course of the interview. The interview guide included questions on demographics, employment background, perceptions of age discrimination, feelings related to these perceptions, and strategies used to gain employment. With consent from the participants, interviews were tape-recorded and later transcribed verbatim.The sample for this study included individuals who varied by age, gender, marital status, ethnic or religious background, country of origin, education, income, and length of time unemployed. Just over half of the participants were aged 45–54 years (n = 16) and the remaining ones were 55–65 years (n = 14). There was an equal number of men and women in this study and the majority of participants were married (n = 16). The remaining participants were divorced (n = 6), never married (n = 6), or widowed (n = 2). In terms of ethnic or religious background, the majority of participants indicated that they were atheist or preferred not to answer this question (n = 16). The remaining respondents varied in their ethnic or religious background. The sample was quite well educated, as the majority of the participants had a college or university degree (n = 19). In terms of participants’ personal income before they became unemployed, individuals who were comfortable disclosing their income were quite dispersed among the income categories provided to them in the interview, aside from the lowest income category ($19,999 or less). Participants were also asked their personal and family income level since being unemployed; however, only a small number of participants were willing to provide this information. The length of time that individuals were unemployed also varied substantially, ranging from 3 months to 6 years. The most common length of time unemployed for the participants was between 6 and 8 months. The occupational backgrounds of the participants interviewed were quite varied. Occupational categories represented in this study included administrative, construction, consulting, education, engineering, executive, finance, food and service, health care, human resources, law, management, sales, marketing, and television and film.Data AnalysisThe data were coded according to several key themes, which were broken down further into various subthemes and subcategories (Taylor & Bogdan, 1998, p. 142). I used three stages to analyze my data based on the grounded theory approach (Strauss & Corbin, 1988). First, I used open coding and explored preliminary themes that emerged from the data. I then used axial coding, which allowed me to create a framework involving these themes and helped me to identify and link more specific subthemes and subcategories. Finally, I used selective coding, which involved confirming and refining my initial set of codes.This article is part of a larger research study containing six main research themes. “Management techniques,” the focus of this article, was one of my six key themes, which was broken down further into three subthemes and nine subcategories. After reading through the data and recognizing similar statements from various participants in terms of a pattern beginning to present itself, I created a theme, subtheme, or subcategory. For example, looking at the theme “management techniques” as a whole, I began to notice some common strategies that participants were using in an effort to find employment. I divided these strategies into three subthemes based on “preparation for the interview,” “impression management during the interview,” and “alteration of employment.” I then more closely examined these subthemes and found that within the subtheme “preparation for the interview,” various patterns began emerging, such as “physical preparations” (related to appearance), “training,” “résumé modification,” and “mental preparations.” These became the subcategories. Two subcategories also presented themselves with respect to “impression management during the interview,” which I grouped into “clothing for the interview” and “youth-oriented language during the interview.” Finally, the third subtheme, “alteration of employment,” was broken down into the subcategories “alteration of employment goals,” “alteration of employment type,” and “alteration of geographic location for employment.” I developed codes (e.g., 1.1, 1.2) and recorded them on the transcripts by hand and then I loaded the data into the qualitative analytical software program QSR NUD*IST (Non-Numerical Unstructured Data Indexing Searching and Theorizing). This program allows researchers to code large amounts of text, which can then be stored according to themes and accessed easily in the future. Thus, once the transcripts were loaded into QSR NUD*IST, I entered the codes I had created by hand so that when I was ready to write about a particular subtheme or subcategory, this software would highlight these previously coded quotations.To ensure trustworthiness of the data (Lincoln & Guba, 1985) or accuracy of its interpretation, several techniques were used. First, interviews were transcribed verbatim by a research assistant, and I then checked them independently for errors. Second, although I coded the data myself, themes that emerged were shared with my peers and only finalized once this consultation took place. At that point, a consensus was reached that these themes accurately represented the data they were describing. Transcripts were also reexamined on several occasions to confirm accuracy of the themes that initially presented themselves. Finally, I continued to recruit participants until I felt that no new themes were emerging and data saturation was achieved (Patton, 2002). It is anticipated that by using the methods previously described, the transferability of my findings (similar to external validity in quantitative studies) to other research contexts will be heightened (Lincoln & Guba). In addition, member checks, a technique that involves the corroboration of preliminary findings with research participants, were used early on in my study to ensure the credibility (similar to internal validity in quantitative studies) of my findings. These member checks involved follow-up telephone conversations with several participants. During these conversations, I suggested specific themes that were emerging from the data and asked respondents if these themes accurately reflected their job search experiences.Results: Managing Age DiscriminationIn the job-seeking process, participants felt that prospective employers used various mechanisms to discriminate against them that reflected negative stereotypes with respect to skills, training, adaptability or flexibility, and financial costs. Further, participants felt that employers examined their résumés in a discriminatory fashion (i.e., choosing candidates to interview based on the year their degree was received or the number of years of experience they had), that the job interview was a mechanism for employers to assess the age of job candidates, and that ageist language was used during the hiring process. (For more detail on participants’ perceptions concerning age discrimination in the hiring process, see Berger [2004].) As a result of these negative experiences, the vast majority of individuals in this study developed specific strategies that they felt helped them avoid being stigmatized as old when searching for employment. Although it is recognized that most individuals, regardless of age, use various management techniques during their job searches (e.g., wearing appropriate attire, preparing mentally for an interview), the strategies that I refer to in this section are management techniques that are related to age. I have grouped these techniques into two categories. The first type of technique used by respondents I refer to as “counteractions”—tactics used to offset negative stereotypes. These techniques involve participants attempting to counteract employers’ ageist stereotypes by maintaining their skills and changing their work-related expectations. The second type of technique used by participants was “concealments”—strategies used to hide specific information (i.e., age). These techniques, developed to conceal age, included participants altering their résumés, physical appearance, and language used.CounteractionsSkill Maintenance.—The first age-related management technique used by respondents to counteract employers’ negative attitudes toward them involved keeping up-to-date with training. This technique was used by the majority of participants, slightly more often by women as compared with men. Overall, they felt that employers did not want to invest the money and time in training older workers:They look at a person and—this has been told to me point-blank by human resources people—they look at people as a monetary investment. They invest a certain amount of time and effort in training a new recruit—the higher level of the recruit, the more expensive the time in training. And they expect to get back ten times their investment, or else it isn't worth it to them to hire them. So if they spend a year acclimatizing me, training me, getting me integrated into the system, then they would expect to get ten years of profitable time out of me. Now, if I am 50 years old, they look at me as a poor prospect. Even at 46, they were looking at me as a very poor prospect. (man aged 52)Respondents kept their training current to avoid being classified as “out of touch” by potential employers:Any training going on, I signed up for it, and I did it. You know, when it came to computer training, I was one of the older ones. Everyone realized it, but I'm computer literate, so that's not an issue. … I have also gone for all the training on how to job search. They have my résumé, a follow-up letter, and the whole bit—doing it professionally. (woman aged 60)I went to the University of Toronto to take some courses to update my knowledge. I also go when different organizations offer specific workshops that I am interested in. (woman aged 56)Despite the recognized importance of keeping training and skills up-to-date, many of the participants were very frustrated with their current inability to pay for the training needed to obtain employment and described a type of “catch-22” situation during our interview:I need the qualifications, but I need money to get the qualifications, but I need the qualifications in the first place to get the job to have the money. It's kind of like a catch-22. It's very similar to the 17-year-old fresh out of high school who can't get a job because he doesn't have any experience and doesn't have any experience because he can't get a job. It is the same kind of catch-22, only 35 years later. (woman aged 53b)If you have an education here [Canada] you can reach whatever you want. But I have no money to take courses to upgrade my education from [my previous country of residence]. It is not a nice feeling. It is unpleasant and very difficult. … I am planning to learn more and get a Canadian license but the problem for me is money. If I can get a job and start to work, I will be able to save money and get a license. (woman aged 45b)There were three groups of women and four groups of men who shared the same age in my study. To distinguish between participants, I have used the letters “a,” “b,” and “c” after their ages. A few of the participants mentioned that they were doing volunteer work, in the hopes that this would help them gain reentry into their desired occupation:We both [she and her husband] are volunteering and always keeping busy. This allows us to improve our knowledge. … And I am reading every day on the computer to improve my professional skills. (woman aged 45b)I have worked for two years as a volunteer. … I am probably too chicken to get a [paid] job because I haven't worked in a while. (woman aged 56)Volunteering was viewed as a temporary (and undesirable) solution to prolonged unemployment; yet, participants felt that this would help them update their skills or keep them current and hoped that it would give them the chance to get a foot back in the door.Another way that participants were able to update their skills was by attending workshops held by older worker programs:I never really used the Internet very much where I worked ‘cause you know, you were busy with your other things. … They [older worker program] spent a lot of time on computer skills, like using Word and getting on the Internet. (man aged 62)I was advised to go for that [older worker] program. … I needed some updating in my computer skills because the courses I had in computers was from more than 20 years ago. (man aged 57b)I have changed majorly since I took the training [provided through the older worker program]. … There were a lot of books available there. There was a lot of advice given. … I went to the centre and every day I learned something new. (woman aged 50a)Interestingly, what was previously considered an advantage by participants—having job security by remaining in one job for a long period of time—was now something that they believed was hindering their job search success. To overcome this newfound limitation, participants felt that they not only needed to update their skills but they also needed to mentally alter their expectations to find work. Participants’ changing expectations will now be discussed.Changing Expectations.—Another way that participants attempted to counteract employers’ ageist stereotypes was in relation to their work-related expectations. Most respondents (men and women equally) changed their employment expectations with respect to altering their original employment goals, the type of employment being sought, or the geographical location of the place of employment. First, a third of participants indicated that they altered their original employment goals. Second, a majority of participants changed the type of employment they were seeking (i.e., monetary remuneration, career change, employment status). Finally, one third of respondents also considered geographical relocation to secure employment. Although these mental strategies may be applicable to all individuals regardless of age, older workers have had to use these management techniques to heightened degrees due to the experience and higher salaries that are often associated with advancing age. Thus, once older workers perceived that their age was a barrier to reemployment, they soon realized that they could no longer have the same expectations they once had if their main goal was to secure employment.In terms of employment goals, although monetary rewards were viewed as important to a degree, satisfaction and fulfillment with work were seen as the primary motivators in seeking employment by some of the participants in higher-income brackets. For example, according to one respondent:You know, I'm not an entry-level person, and I don't expect an entry-level kind of remuneration. Not that money is the most important thing to me—satisfaction to me is as important or more important at this stage of my life than money. (man aged 52)This tendency may assist older individuals in their job-seeking process, as employers often cite higher financial costs (associated with employee turnover, salaries, and benefits) as a reason for not hiring older employees. Respondents described how they came to this realization:The end result for me is to find a position. I'm not finding a position. The end result of a company is to make money. They think they are going to make money off younger people better than they are going to make money off an older person because they think that the younger person will be with the company longer and is going to be able to produce or benefit revenue for a longer period of time. It's not true. An older person doesn't make the mistakes of the less experienced. So, the learning curve is a lot shorter. But they don't look at that. They don't look at the result's line. (man aged 52)I think salary is a barrier because employers think we expect more money. (woman aged 45b)It follows that many of the respondents realized that receiving the same level of financial compensation as they had prior to becoming unemployed was not necessarily feasible and thus lowered their monetary expectations in relation to the type of employment being sought. For example:I know that I will take a lower [pay] rate to come in [to the labor force]. … You just have to keep going, you know. Lower your expectations, unfortunately, and just keep going. … It seems the longer you're unemployed, the more you can rationalize it. You don't feel any better, but you just sort of, like I said, you lower your expectations just basically to get back into the job market. (woman aged 58)Thus, job advertisements stating lower salaries, a mechanism that participants believed employers used to discriminate against them, would no longer be considered a barrier to these respondents.Furthermore, after long periods of unemployment, many individuals begin to consider a change in career or industry to gain reentry into the workforce:The frustration is immense because you keep sending your résumés back out to the same people. … I've been about 18 months out of work, and a couple of months ago I decided I've got to switch this trade. That's when I considered bartending. … Then I decided on this building maintenance course. . . . But [it is hard] knowing that no matter what industry I go into now, I'll never make that kind of money again. (man aged 50c)I am trying now to shift my focus. … I am trying to diversify. … It has been very difficult to get interviews, so now I am trying to look into other areas. (woman aged 50b)Although a change in career was not their preference, many respondents realized that this sacrifice had to be made at some point in their job search process. Some individuals even deliberated between returning to their previous careers with lowered salary expectations and changing careers in the hopes of achieving any level of financial stability. This decision was not an easy one, as many participants soon realized that once they left their original career, they would have great difficulty reentering it at a later date. For example, one, woman who was a dental assistant explained her current predicament:I am looking into—I mean I'm not really sure that I would be just as good doing other things and I'm thinking why should I waste 25 years of training. … I have set up a deadline for myself that at least by the end of January if I do not get a job as a dental assistant—I never thought I could think along the lines and I'm still not happy doing that—but I just might consider other jobs. … But then that's where the problem lies, that if you go into other things, it…s that much harder to go back. If you go for an interview and the doctor…s asking you what you are doing and you say, “I’m in a bookstore” or something like that—“So why are you applying for the job? I can have 20 assistants that are actually assistants.” So that’s the dilemma there. (woman aged 50a)Expectations may also change in relation to the status of employment that is being sought. Many older individuals decide (voluntarily or not) to work on a contract or part-time basis with various organizations. For example:I would accept part-time work, and at the moment I'm doing some consulting work for a boutique; it gets me out of the house. (woman aged 60)There are some jobs that admittedly the salary is not what I want, but right now, would I take a part-time job? Yes. … Right now I'm willing to listen to any offers that come around. (man aged 62)Although the security of full-time employment was not present for these individuals, participants felt that they were able to display their strengths to employers, which may ultimately lead to more stable employment for them in the future. Many participants felt that changing their own “definition of the situation” (Thomas, 1972) from unemployed to semiretired helped them accept a possible change in the type of employment that was being sought. For example:I sort of call myself semi-retired ‘cause that means that I could tolerate, say, a lesser job than I had before. Like part-time work or something. … When I first looked [for a job], that's what I told people—that I wasn't anxiously looking. Like when I say people, I mean the neighbor I bump into on the street, or something, wondering why I was home during the day. (man aged 60)You know, I almost made up my mind that this is semi-retirement, but I'm not ready for it, I'll tell you. (woman aged 60)Individuals reconceptualized their status from unemployed to semiretired to provide a new and more acceptable social meaning to their current situation, similar to findings reported in the literature on displaced workers (McMullin & Marshall, 1999).Despite making drastic changes in monetary, career, and employment status expectations, participants were still unable to secure employment:I'd like to have full-time but if like I could only get part-time, well, I would do that, you know, in the interim. … I'd take another job. Like I even applied at different places like IGA and Blockbuster, you know, I would take something like that, you know, just to be working and doing something. (woman aged 53c)I'm willing to try anything and I'm not asking 30 dollars an hour, what I used to make before, I'm not asking that. I'm going to go for—how much is minimum wage—six eighty-five? I’m willing to start with that if the job is normal and I’m capable of doing it. I’m not picky. I’m not asking this kind of wages, minimum wage is going to do it. I’m going to be there. I’m going to try. (man aged 56)I’ve junk-mailed every employer out there with my résumé—even survival jobs where I went to a car dealer and wanted to be a lot person—you know, park cars, take them in and get them washed, vacuum them out, and that kind of stuff—even to get minimum wage. And I can’t even land that ‘cause they look at you and go—“What is your problem?” You know, it’s just tough. (man aged 46b)Making these extreme changes in their expectations without success (i.e., securing employment) was obviously quite disturbing to many of the respondents. In fact, many of the comments made by the participants conveyed both a sense of desperation and a sense of adaptableness. For example:I’m willing to do practically anything to work rather than not work. (woman aged 64)Many of the individuals in this study were also quite willing to move to another geographical location to find work:Don't just depend on Toronto, Mississauga, Brampton, or York, because there is nothing out there. And you gotta have your mind made up to want to leave your home and to be traveling wherever the job is. …You have to try all different angles. (woman aged 59)When I started my job search.… I placed that restriction—my job search parameters—and then realized it. I used to say, “I want only full-time and I want it downtown.” Then I said, “I want full-time but I’m willing to stay anywhere in greater Toronto or contract in greater Toronto.” So now it's full-time, part-time, or contract anywhere in Ontario.… And this month I've also begun sending résumés off to the [United] States.… So now I've widened my search.… So we’ll see. (man aged 62)However, several participants noted that although they would personally be willing to relocate to secure employment, they felt that the needs of their families would best be met by remaining in their current geographical location. For example, despite the hardships that they had encountered since immigrating to Canada, some of the participants felt that there would be better long-term opportunities for their families if they remained in Canada due to the educational system or a general feeling of safety and security:I would have, you know, left the country [if] I didn't have the problem of my daughter studying here. It's only for her that I continue to stay here. (man aged 57b)We came to Canada because of the dangers of living in [my previous country of residence]. …My husband was a big company director and I was a medical doctor and now we are both unemployed and on welfare. … We can't find work here. We have been unemployed for five years—since we have been in Canada. (woman aged 45b)The changes in mental expectations described in this section were deemed necessary by the participants to try to overcome some of the stereotypes associated with being an older worker, such as higher salary expectations. Unfortunately, making these mental adjustments did not help participants secure employment, so many of them felt that more drastic measures were needed to conceal their ages to potential employers. The analysis now turns to these types of management techniques.ConcealmentsRésumé Modification.—It should be recognized that stereotypes can also be positive, and some of the participants felt that they possessed certain positive characteristics that are associated with being older. For example, the experience that comes with age was recognized by one respondent when describing her previous work situation:I mean, there is no substitute for experience. … It does make the work that much more efficient; that much more professional. Honestly, I think a lot of people—say, my age—when they go to an office and see a person my age, they would say “Okay, this person knows what she or he is doing. It seems that they’ve been doing this for a long enough time.” (woman aged 50a)While some participants tried to emphasize positive stereotypes associated with older workers, such as loyalty and experience, many were not able to do this due to a variety of reasons. For example, assuming that age and experience go hand in hand reflects a gendered bias in that most of the women in my sample had less paid work experience than did the men, due mainly to raising their children. Thus, they were not able to highlight to employers one of the few positive stereotypes associated with older workers in the same way that men could do so. See McMullin and Berger (2006) for more details on gender differences in the route to reemployment for older workers.However, having experience can also be a negative thing:To just look at my résumé, they know how many years I’ve been in the business and they can sort of deduct that I’m not 35 or 40. (woman aged 60)You don’t know if they're looking at your résumé and looking at … like, figuring you got all that experience so obviously you got to be a certain age, right? (woman aged 53c)In the aforementioned examples, respondents described how their extensive experience was being used by employers as a way to determine age from their résumés and subsequently identify and select younger candidates to interview for job openings. Therefore, most respondents felt that they had to conceal age on their résumés by eliminating some of their work experience, number of jobs held, or the year a degree was received:I try to hide it [my experience]. Sometimes I don’t mention my degree and that sort of thing, and I only show the last ten years of work experience. (man aged 60)I changed my résumé somewhat to exclude the number of years I’ve worked and said, “I have extensive [experience].” … Even though—I guess now everybody who uses “extensive,” they [employers] say, “Oh. This guy's got 20 plus years.” (man aged 62)Participants also indicated that they switched the style of their résumés from chronological ones to “functional” ones, where the importance of skills rather than specific jobs or years in the labor force is highlighted:I changed my résumé around. I was using a chronological résumé at first, but for the last four months I have been using a functional one. (man aged 50a)I use a particular style of résumé that emphasizes skills more than the number of jobs or length of time that I have been in certain jobs. (man aged 50b)Responses from the participants made it increasingly clear that “de-emphasizing age” was a strategy taught by personnel at the older worker programs:At the [older worker program] they helped me modify my résumé—to de-emphasize age. … They would tell us don't give them [employers] any more information, basically, than is necessary. Like sort of just give them your last 10 years of work, let them see you versus, you know, putting down everywhere you've worked for the last 30 years or 40 years. (woman aged 58)I was taught [at the older worker program] that when you send a résumé, you don't put the date you graduated from school … or stuff like that. So they really don’t know how old I am when I send them the résumé. (woman aged 55)Thus, although some individuals may have eventually decided to alter their résumés on their own using some of the techniques previously described, it was clear that the overt encouragement at the older worker programs was the reason that the majority of adjustments were made.“Improving” Appearances.—Another way respondents felt that employers discriminated against them related to the interview process. These individuals believed that after employers had the opportunity to see them during an initial interview, they could immediately estimate how old they were and only request second interviews from the younger applicants. Several participants provided justification for this belief:What is quite interesting is that a friend of mine who is 45 years old, about the same age as me and who just got her degree, is getting more interviews than me. She is only getting the first interview though. I am not sure if it is because the employers assume that she is younger because of her recent degree, but once she gets in there and has an interview, she is never asked back for a second interview. I am not sure if it is age per se, but that is just my feeling. (woman aged 45a)We [she and other older workers] get interviews but not jobs. … When they [employers] look at you, they decide. (woman aged 53a)Thus, another type of concealment used by the majority of participants—by women far more often than men—was to portray a “youthful” appearance to potential employers. Several of the participants referred to physical preparations for the job interview. For example, many men and women acknowledged dying their hair prior to an important interview:When I actually get an interview, … I dye my hair. (man aged 50b)When you look around you and you see the faces getting younger and younger, you know you've got to keep it up. Even if you can't afford it, that's the one thing you need is your bottle of dye. (woman aged 58)Several men also considered using a toupee if they were balding and shaving their beards to appear younger, despite the fact that they did not necessarily agree with this or acknowledge that this was their idea in the first place:They want to see me have a full head of dark brown hair. Maybe I should go to Honest Ed's [discount department store] and buy a nylon toupee. That would improve my appearance, don't you think? (man aged 52)People say to me, because I'm bald, “Why don’t you get one of these wigs?” or whatever you call them, or “Why don’t you shave your beard?” I have had my beard all of my life—since I was 20. It was never a problem, so why is it a problem now? (man aged 50a)Similarly, several respondents mentioned that after going to job interviews, they recognized the importance of “fitting in” to an employer’s organizational culture, which essentially meant giving the impression of being younger:I think that the organizational culture in an organization is very important. You need to fit into the company’s culture. Many companies have youth-oriented cultures and they may only hire someone who fits into this culture. (man aged 50b)Therefore, he explained how he would dress a certain way for an interview in an attempt to “fit in” with what he believed was a young organizational culture:There are certain techniques that I use [in an interview] to seem younger. … I always wear young-looking clothes. … I dress in certain clothes that fit in with their company's culture. (man aged 50b)Another strategy that respondents used to maintain youthful appearances was to make a concerted effort to maintain their health. Thus, although many of the participants encountered changes in their mental health following experiences with age discrimination and prolonged unemployment (see Berger [2006] for more detail on the psychosocial consequences of perceived age discrimination), the majority of participants were in good physical health prior to becoming unemployed and remained this way following unemployment. In fact, several of the respondents felt that their physical health had improved since they became unemployed. For example, when asked to describe his health since he had become unemployed, one individual explained:My health has improved quite a bit because I've got back into a running program. … My previous job, you know, they wanted the 60-hour weeks. … So physically I'm doing a lot better and my eating habits have improved. (man aged 46a)Other respondents expressed how they felt fortunate that they did not physically look “old” due to their good health and thus felt they did not have to alter their physical appearance to avoid experiencing age discrimination from employers:I guess [in relation to] how I look—I really haven’t had that kind of age discrimination. I think that when I go to a [work-related] event I always look good, so I think that also helps a lot. (woman aged 50b)Appearance-wise I find I’m okay … to be able to say, “yes, I'm still capable of working; I don’t have a cane yet.” (woman aged 58)Finally, it should be noted that not all the participants agreed with modifying their physical appearance to obtain employment. Some individuals actually were quite against this active age concealment behavior. For example:I have grey hair and I refuse on general principle to dye my hair. … I don't want to cater to this. … I can see the reaction of people to somebody who has dyed hair—you know they dye their hair blond or whatever—they tend to think of them as younger and prettier and so on and I refuse to do it. I’m stubborn that way. It’s costing me I know but I don’t believe in catering to that. I don’t believe the color of your hair should be an influence on how people treat you. (woman aged 53b)It is clear that this woman felt very strongly about not wanting to conform to society’s expectations regarding her physical appearance; yet, she also recognized that this conviction may be harming her job search success.The “Right” Age Talk.—The last way that many participants attempted to conceal their age was by managing their outward portrayal during job interviews through language. This was done by either using certain discourse to reframe otherwise undesirable conversations or using “youth-oriented” language. Turning first to reframing undesirable age-related conversations, one woman said:I was told that [I was too qualified for a position] but I turned it around because I had so much experience or knowledge in finance that—I was applying for a job in mutual funds and I said “You are getting kids out of school who don’t even own them so how do you expect them to sell them.” I turned it around. (woman aged 53a)Many of the respondents explained how they managed to avoid or deflect discussions in which their age could become known to employers and explained the importance of anticipating age-related questions during job interviews to mentally rehearse desirable answers:I’m not saying falsify an image [during an interview]. Be yourself and be honest. But, again, depending on the questions, you know—“How old are you?”—rather than respond—“That’s none of your business”—I would have said something like—“Well, I’m old enough to have 25 years of experience in this business,” which gives them an age bracket and it doesn’t affect me. “As you can see by my résumé, I have 25 years [experience], so you know I’m not 35.” (man aged 50c)I try to verbalize what I’m going to say to them [employers] because they seem to do this constantly …. I guess I just sort of prepare myself mentally for these, I call them, stupid questions—“What are your goals, let’s say ten years down the road?” Well, I don’t have ten years left, you know. So you have to verbalize what you are going to say. (woman aged 58)Several individuals also engaged in what they considered to be youthful language. For example, one participant explained the lengths he went through to avoid being defined as “out of touch” more generally in terms of current societal trends:What I talk about is really youth-oriented. I make sure I discuss very physically active kinds of sports and I mention various social groups that I have joined. (man aged 50b)Participants also realized the importance of using certain “buzzwords” (e.g., in the sales field, what was once referred to as a “forecast” is now called a “pipeline”) during a job interview to illustrate to employers that they were up to date with the language being used in their field. For example:The high-tech industry has so many different skills and languages. Like, you have to be an expert. … You’ve got to really move quickly and learn new stuff. (man aged 46a)I think the lingo has changed considerably. … There are a whole slew of buzzwords that I am sure are being used today by corporations that perhaps a person that was trained 20, 25 years ago understands but is not comfortable using because it’s not in their everyday vocabulary. They were trained with a different sets of words. … I think it's a barrier. … As soon as you go into an interview and you use your own terminology, you date yourself—automatically, subconsciously. (man aged 57a)This participant went on to describe a whole list of buzzwords that he felt were being used by individuals in his field and in human resources more generally:We were just discussing a whole list of words yesterday [at the older worker program]. And they were words that I personally had not seen, let’s say 20 years ago. Such as “telecommuting” and “motor-skilling” and “life-long learning” and “career planners” and “dejobbing.” This is just on the human resources side. … The same thing occurred in the marketing industry in a different way. For example, you know, having “new enhancements.” That’s one way people would refer to things. In the old days you would just call it, you know, “increased sales.” … I do believe that it would be useful to create, I don’t know, a whole set of words, buzzwords, that are currently in use in an industry that show involvement and progression—at least of terminologies. (man aged 57a)Another individual used some other buzzwords to explain his difficulties in an interview situation:Let's say they [employers] describe a particular position you have to do. … You have to meet “revenue targets” and you have to “integrate pricing and portfolio managements” and “database marketing” and you have to work “retention strategies.”. … Unless you are able to repeat these words that they put in their job description, you may sound outdated. You may sound like you will not be able to fit because you don’t understand their terminology, which is not true. You are [just] more comfortable working with established business terms. (man aged 45b)Thus, some of the respondents felt they were at a disadvantage in relation to the language being used in their field after realizing that the lingo they were trained with in the past was already outdated; being unemployed for a lengthy period of time complicated the matter. They found this quite frustrating, similar to the catch-22 situation that they found themselves in in relation to their skills and training. Thus, individuals felt that the only way to maintain a current lingo in their field was to be employed in it, and this was not happening for them at the time of these interviews.Discussion and ConclusionsThis study explored how older individuals’ perceptions of age discrimination in the job search process led to the development of age-related management techniques. Participants’ feedback on seeking employment illustrated how age became a discrediting attribute or stigma (Goffman, 1963), which resulted in discriminatory attitudes and behaviors by potential employers. Using Goffman’s three notions of identity, individuals began to feel old (felt identity) when the image they were projecting to employers (presented identity) was perceived to be that of an “older” worker (social identity). Because presented identity can be modified depending on social context, by using the management techniques described in this article, participants were actively controlling the image they were projecting to others to present a more advantageous social identity.Similar to findings discussed previously in relation to impression management techniques in general (Goffman, 1959, 1963) and their use in the organizational behavior literature (Bozeman & Kacmar, 1997), once individuals perceived that a discrepancy existed between the image they wanted to convey to potential employers and the feedback they received (perceived age discrimination), the vast majority of them developed alternative management techniques that were age related to project a positive presented identity (Goffman, 1959) and avoid being stigmatized by potential employers. For example, most respondents altered their physical appearance to portray a more youthful image in a job interview setting to manipulate their self-presentations and sway potential employers to hire them.Overall, counteractions were used by participants more often than concealments. Looking at all five types of management techniques, “changing expectations” was used most often, followed by “skill maintenance,” “the ‘right’ age talk,” “résumé modification,” and “‘improving’ appearances.” With respect to gender differences, in descending order of use, women were most likely to “‘improve’ appearances,” equally used “changing expectations” and “skill maintenance,” and then used “the ‘right’ age talk” and “résumé modification.” For men, the pattern of usage was slightly different. They predominately used “changing expectations,” followed equally by “résumé modification” and the “right age talk,” “skill maintenance,” and “improving appearances.” I was surprised that there were not larger gender differences in terms of the management techniques used overall, but it is interesting to note that the biggest gender difference appeared with respect to the subcategory “improving” appearances. Aged bodies are more commonly perceived to be a barrier to finding employment for women than for men because youth is more often equated with attractiveness for women (McMullin & Berger, 2006). As many of the women in my sample had come to this realization over the course of their job search, most felt the need to alter their appearance before a job interview.Management techniques were developed by the participants to combat employers’ negative attitudes in relation to their skills, training, adaptability or flexibility, and higher perceived monetary costs by staying up-to-date with their skills and training and changing their work-related expectations. In addition, participants felt that employers examined their résumés and chose candidates to interview in a discriminatory fashion. After coming to this realization, individuals began to use concealment tactics on their résumés that involved eliminating the year that they received their degree or reducing the amount of experience (i.e., number of jobs or number of years in a job) displayed on their résumés to gain a more competitive advantage in the job market. Similarly, respondents felt that the job interview and language used by employers were key mechanisms for discrimination. By altering their appearance or using specific language to mentally prepare for age-related concerns, they felt a more youthful image was being conveyed to potential employers.This study contributes theoretically and substantively to the literature by applying concepts related to management techniques and the aging process to older individuals who feel stigmatized in their job search process. More specifically, in relation to management techniques, previous researchers have argued that the following types of impression management techniques have been used in an interview setting: (a) entitlements (claiming responsibility for a positive event that occurred in the past), (b) enhancements (making statements that reflect one’s positive attributes), and (c) self-promotion (highlighting strengths in relation to those required by the ideal applicant (Delery & Kacmar, 1998). (Although this was not the focus of my current investigation, I found evidence that participants in my study were using all three of the impression management techniques described by Delery and Kacmar. Many of them made statements to me that highlighted their past accomplishments, emphasized their positive attributes, and stressed how their work-related experience would benefit them in a job situation. Further, they explained that they would make similar comments to employers in an interview setting.) Findings from my study expand on the impression management techniques described by Delery and Kacmar. My research demonstrated that older workers use these three techniques in a similar fashion to younger workers. However, they also use two additional techniques in an attempt to combat age discrimination. The first technique was counteractions—tactics used to offset negative stereotypes. The second technique used by participants was concealments—strategies used to hide specific information (i.e., age). The notion of concealment has been examined in the literature on aging in relation to older widows who suppressed evidence of age (Matthews, 1979, pp. 74–75). However, the use of these management techniques has not been applied specifically to age discrimination in the hiring process prior to this study. I anticipate that this contribution to knowledge can also be applied to a broader area relating to impression management and aging more generally. Both counteractions and concealments can relate to behaviors that attempt to offset any type of stereotypes or hide any type of stigma, although further research is needed to demonstrate the utility of these concepts beyond their applicability to older workers and the hiring process.The findings from my research suggest that more needs to be done to assist older individuals in their search for reemployment; thus, I believe that additional interviews with older workers, employment counselors, and employers themselves will greatly enrich our understanding of the job search process. These findings would help expand those presented in this article, which could then be investigated on a larger scale. A quantitative study that tests the frequency that these management techniques are used would expand our knowledge regarding structural barriers in the job search process. Examining these strategies on a larger scale in future research would help quantify the extent that counteractions and concealments are being used by older workers. This would highlight the magnitude of this issue, which is becoming of growing importance due to the aging of the workforce. This suggestion highlights the utility of combined research methods or triangulation (Creswell, 1994). I suspect that the insights into the lived experiences of the participants could only be obtained from qualitative interviews; yet, to illustrate the applicability of the themes I created beyond the participants in my study, a larger-scale investigation is warranted. Although gender was examined to some extent in this article, the ways in which age combines with other social structural factors such as race, ethnicity or country of origin, and occupation in the job search process need to be examined.Another area worthy of future research relates to conducting interviews with individuals who found work immediately following unemployment or those older than 65 years who are working past the “typical” retirement age. This research would provide insights into whether these individuals have used the specific age-related techniques described in this article. This could lead to an exploration of lessons learned from “success stories” of older workers who have succeeded in finding and retaining employment, which could help unemployed older individuals avoid or manage encounters with ageist employers. Further, it would be interesting to discover if these individuals are located in certain industries or occupations (e.g., those with less physically demanding jobs).Next, longitudinal research is needed to follow older individuals who experience prolonged periods of unemployment. Interviewing individuals several times over an extended period would provide greater insight into the job-seeking process and advance knowledge in relation to the amount of time spent in each stage of the job search process. Further, both physical and mental health could be assessed at various points in time to observe specific changes and relate them to their coinciding experiences with their search for employment. Thus, both long- and short-term health of older workers could be assessed in relation to experiences with age discrimination and unemployment.Further research is also needed in relation to the newfound contradiction that has surfaced during my research. On the one hand, prior quantitative analyses have demonstrated that employers’ attitudes toward older workers have become more positive in recent years; yet, on the other hand, my interviews with older workers suggest that age discrimination currently exists in the hiring process. Although quantitative studies that document employers’ attitudes toward older workers are growing, there is little research that examines employers’ accounts of the hiring process. Therefore, qualitative interviews with employers would allow for greater insight into the meanings that they give to various experiences with older job candidates and provide a greater understanding of their attitudes toward older workers. It would also provide clarification in relation to whether employers’ attitudes lead to the development of age-related hiring and retention practices and policies.In summary, research indicates that as a result of prolonged periods of unemployment, older workers are often considered to be “peripheral and marginal to the workforce” (Henkens, Sprengers, & Tazelaar, 1996, p. 575). This is due in large part to the existence of age discrimination in employers’ hiring processes. Thus, the current societal norms that favor youth have created structural barriers for older workers seeking reemployment. However, despite the existence of these structural barriers, findings from this study suggest that participants actively negotiate age-related strategies and thus clearly have agency in the job search process. Although the findings from this study cannot be generalized to all older workers due to the small nonrepresentative sample, the sociological insights gained contribute to knowledge of older workers’ job search process and provide a foundation for further research into this area on a broader scale. This research suggests that there is a need to reexamine the hiring practices of employers and to improve legislation in relation to their accountability. The Conference Board of Canada (2006) has suggested that policy makers work with employers to make age part of their diversity strategies; yet, this is still not largely occurring. It is only by combating age discrimination on a broader societal level that the age-related management strategies used by the participants will no longer be necessary.FundingThe research reported in this article was funded by a Social and Economic Dimensions of an Aging Population fellowship, supported by the Social Sciences and Humanities Research Council of Canada, Byron Spencer, principal investigator.I thank Dr. Carolyn Rosenthal, Dr. Margaret Denton, and Dr. William Shaffir for comments on various drafts of this article as well as the three anonymous reviewers for their helpful feedback. In addition, many thanks go to the participants in this study for sharing their experiences.American Association for Retired PersonsHow to manage older workers1994Washington, DCAuthorAmerican Association for Retired PersonsStaying ahead of the curve: The AARP Work and Career Study2002Washington, DCAuthorBergerEDOrganizational and personal characteristics influencing Canadian employers’ attitudes toward older workers. Unpublished MSc thesis1999Toronto, Ontario, CanadaUniversity of TorontoBergerED‘Older’ workers: The negotiation of age discrimination and identity in the job search process. Unpublished doctoral dissertation2004Hamilton, Ontario, CanadaMcMaster UniversityBergerED‘Aging’ identities: Degradation and negotiation in the search for employmentJournal of Aging Studies200620303316BlumerHSymbolic interactionism: Perspective and method1969Rev. ed., 1986 edBerkeleyUniversity of California PressBozemanDPKacmarKMA cybernetic model of impression management processes in organizationsOrganizational Behavior and Human Decision Processes199769930ChiuWCKChanAWSnapeERedmanTAge stereotypes and discriminatory attitudes towards older workers: An East-West comparisonHuman Relations200154629661Conference Board of CanadaCanada's demographic revolution: Adjusting to an aging population2006Ottawa, Ontario, CanadaAuthorCreswellJWResearch design: Qualitative and quantitative approaches1994LondonSageDeleryJEKacmarKMThe influence of applicant and interviewer characteristics on the use of impression managementJournal of Applied Social Psychology19982816491669Forum of Labour Market MinistersOlder workers in the labour market: Employment challenges, programs and policy implications2002Winnipeg, Manitoba, CanadaSecretariat, Forum of Labour Market MinistersGeorgeLKBinstockRGeorgeLSocial structure, social processes, and social-psychological statesHandbook of aging and the social sciences19903rd edSan Diego, CAAcademic Press186203GlaserBGStraussALThe discovery of grounded theory1967Rev. ed., 1999 edChicagoAldineGoffmanEThe presentation of self in everyday life1959New YorkDoubledayGoffmanEStigma: Notes on the management of spoiled identity1963New YorkSimon & SchusterGuillemardA-MWalkerAEmployers’ responses to workforce ageing—A comparative Franco-British exploration1994(For presentation at the World Congress of Sociology, Bielefeld). Paris: University of ParisHaasJShaffirWBecoming doctors: The adoption of a cloak of competence1991Greenwich, CTJai PressHenkensKStereotyping older workers and retirement: The managers’ point of viewCanadian Journal on Aging200524353366HenkensKSprengersMTazelaarFUnemployment and the older worker in The Netherlands: Re-entry into the labour force or resignationAgeing and Society199616561578HewittJPStokesRDisclaimersAmerican Sociological Review197540111Human Resources Development CanadaOlder worker adjustment programs: Lessons learned: Final report1999Ottawa, Ontario, CanadaEvaluation and Data Development, Strategic Policy, Human Resources Development CanadaLincolnYSGubaEGNaturalistic inquiry1985Beverly Hills, CASageLinkBGPhelanJCConceptualizing stigmaAnnual Review of Sociology200127363385LukenPCSocial identity in later life: A situational approach to understanding old age stigmaInternational Journal of Aging and Human Development198725177193MarshallVWIssues of an aging workforce in a changing society: Cases and comparisons (Monograph Series: Issues of an Aging Workforce)1996Toronto, Ontario, CanadaUniversity of Toronto, Centre for Studies of Aging, and CARNET, the Canadian Aging Research NetworkMatthewsSHThe social world of old women: Management of self-identity1979Beverly Hills, CASageMcMullinJCookeMDownieRLabour force ageing and skill shortages in Canada and Ontario2004Ottawa, Ontario, CanadaCanadian Policy Research NetworksMcMullinJABergerEDCalasantiTMSlevinKFGendered ageism/age(ed) sexism: The case of unemployed older workersAge matters: Re-aligning feminist thinking2006New YorkRoutledge201223McMullinJAMarshallVWRyffCDMarshallVWStructure and agency in the retirement process: A case study of Montreal garment workersThe self and society in aging processes1999New YorkSpringer305338Ontario Human Rights CommissionDiscrimination and age: Human rights issues facing older persons in Ontario2000Toronto, Ontario, CanadaPolicy and Education BranchPattonMQQualitative research and evaluation methods20023rd edThousand Oaks, CASagePolicy Research InitiativeEncouraging choice in work and retirement: Project report2005Ottawa, Ontario, CanadaProject Research InitiativeRosenfeldPImpression management, fairness, and the employment interviewJournal of Business Ethics199716801808ScottMBLymanSMAccountsAmerican Sociological Review1968334662Secretariat for the Expert Panel on Older WorkersOlder workers: Challenges and policy issues2007Ottawa, Ontario, CanadaHuman Resources and Social Development CanadaStatistics CanadaParticipation rates and unemployment rates by age and sex, selected countries2004Ottawa, Ontario, CanadaMinister of IndustryStatistics CanadaUnemployment rate, labour force aged 15 and over, Canada, provinces, health regions and peer groups, 2002-20032004Ottawa, Ontario, CanadaMinister of IndustryStatistics CanadaLabour force historical review 20072007Ottawa, Ontario, CanadaAuthorStraussACorbinJBasics of qualitative research: Grounded theory procedures and techniques19882nd edNewbury Park, CASageTaylorPUrwinPAge and participation in vocational education and trainingWork, Employment & Society200115763779TaylorPWalkerAEmployers and older workers: Attitudes and employment practicesAgeing and Society199818641658TaylorSJBogdanRIntroduction to qualitative research methods: A guidebook and resource19983rd edNew YorkJohn Wiley & SonsThomasWIManisJGMeltzerBNThe definition of the situationSymbolic interaction: A reader in social psychology19722nd edBostonAllyn and Bacon331336U.S. Bureau of Labor Statistics: Division of Labor Force StatisticsUnemployed persons by age, sex, race, Hispanic or Latino ethnicity, marital status, and duration of unemployment2007Washington, DCAuthorWalkerAA strategy for active ageingInternational Social Security Review2002551121139WalkerATaylorPCombating age barriers in employment: A European portfolio of good practice1998LuxembourgOffice for the Official Publications of the European Communities
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B94909EDE7981E184D1F4DEEA400564BC0B9249.txt b/test/dataset/in/resources/corpus/Clean_0B94909EDE7981E184D1F4DEEA400564BC0B9249.txt
deleted file mode 100644
index 2efd3d5..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B94909EDE7981E184D1F4DEEA400564BC0B9249.txt
+++ /dev/null
@@ -1 +0,0 @@
-]>HEARES3641S0378-5955(01)00217-910.1016/S0378-5955(01)00217-9Elsevier Science B.V.Fig. 1Schematic drawing of the lateral (a) and medial (b) side of the right ear of the anabantoid T. vittata. (a) Anterior is to the right. (b) Anterior is to the left. The saccular and lagenar maculae and nerve branchelets are indicated for orientation purposes. AC, anterior semicircular canal; HC, horizontal semicircular canal; L, lagena; LM, lagenar macula; LO, lagenar otolith (sagitta); N, vestibular and auditory branches of the eighth nerve; PC, posterior semicircular canal; S, saccule; SM, saccular macula; SO, saccular otolith; U, utricle; UM, utricular macula; UO, utricular otolith.Fig. 2Saccular maculae from (a) B. splendens (1.6 g, 39 mm), (b) M. opercularis (1.87 g, 39 mm) and (c) T. vittata (2.00 g, 44 mm). In each case, the long dashed lines separate groups of ciliary bundles oriented in different directions while short dashed lines in the saccular epithelia separate the anterior portion of the macula from the ribbon-shaped portion. The arrows point in the direction of the kinocilia on the cells in a particular group.Fig. 3Scanning electron micrograph showing the posterior saccular epithelium of M. opercularis. Cells shown are divided into two groups that differ in their orientation (separated by a dashed line). All cells in these two groups are oriented in the same direction and have their kinocilia pointed in the same way (arrows). Scale bar=3 μm.Fig. 4Lagenar maculae from (a) B. splendens, (b) M. opercularis and (c) T. vittata. See Fig. 2 for further explanations.Fig. 5Utricular maculae from (a) B. splendens, (b) M. opercularis and (c) T. vittata. See Fig. 2 for further explanations. Lc, lacinia; St, striola.Fig. 6Scanning electron micrographs of the saccular sensory epithelia of (a) B. splendens, (b) M. opercularis and (c) T. vittata. These SEMs show F1 and F2 ciliary bundles at the ventral margin on the posterior part of the maculae. The bundles are oriented ventrally. The arrows point in the direction of the kinocilia. Note that the kinocilia (K) of the F2 bundles are several times longer than the kinocilia of F1 bundles compared to stereocilia. Scale bar=5 μm.Fig. 7Scanning electron micrograph of the ventral saccular sensory epithelium of T. vittata. Arrowheads point to the separated row of F2 ciliary bundles at the ventral margin. Scale bar=10 μm.Fig. 8Scanning electron micrographs of the lagenar and utricular maculae of T. vittata showing ciliary bundle types. (a) Caudal margin of the lagenar macula. (b) Lateral margin of the striola of the utricle near the lacinia. Scale bar=5 μm.Table 1Mean (±S.E.M.) body weights, standard lengths (SL) and length of the saccular maculae in the rostral caudal direction in the three species examinedB. splendensM. opercularisT. vittataANOVASignificanceWeight (g)1.67±0.12.06±0.21.03±0.3F2,15=1.55P=0.01SL (mm)39.0±1.340.4±1.634.2±3.3F2,14=1.25P=0.14Macula (mm)1.05±0.011.04±0.010.98±0.09F2,14=0.63P=0.55Significances of the differences between species were tested using one-way ANOVA.Table 2Mean (±S.E.M.) ciliary bundle densities per 1000 μm2 in the rostral, middle and caudal area of the saccular maculae in the three species examinedB. splendensM. opercularisT. vittataANOVASignificanceRostral53.1±7.450.9±2.148.4±4.7F2,13=0.25P=0.78Middle49.5±3.754.0±1.451.6±5.8F2,15=0.42P=0.67Caudal51.1±2.156.3±2.360.5±2.4F2,14=3.70P=0.051Significances of the differences between species were tested using one-way ANOVA.Comparison of the inner ear ultrastructure between teleost fishes using different channels for communicationFriedrichLadicha*friedrich.ladich@univie.ac.atArthur NPopperbaInstitute of Zoology, University of Vienna, Althanstrasse 14, A-1090 Vienna, AustriabDepartment of Biology, University of Maryland, College Park, MD 20742, USA*Corresponding author. Tel.: +43 (1) 4277 54227; Fax: +43 (1) 4277 9544AbstractThe anatomy and ultrastructure of the inner ear of three species of gouramis which differ widely in acoustic behavior were studied using scanning electron microscopy. Of the three species, Trichopsis possess a pectoral sound-producing mechanism while Macropodus and Betta lack sonic organs. The general structure of the inner ear and the shapes of the sensory epithelia are very similar, although they do differ on the posterior part of the saccular macula which is more S-shaped in Trichopsis and Macropodus than in Betta. The maculae on the three species do not differ either in ciliary bundle type (cells with long kinocilia on the periphery of the maculae and cells with short kinocilia in the central region) or in hair cell orientation pattern. Quantitative measurements of hair cell densities and the size of the sensory epithelia of the saccule did not show significant differences between species. Data presented correlate with physiological results from other investigators showing similar auditory sensitivity in Trichopsis and Macropodus. The similarity in structure and function of the inner ears of gouramis on one hand, and the occurrence of sound-generating organs in just one genus, suggests that hearing evolved prior to vocalization and thus acoustic communication in this taxon.KeywordsSound productionAcoustic communicationHearing specialistHair cellAnabantoidEvolutionAbbreviationsAC, anterior semicircular canalF1, F2, ciliary bundle typesHC, horizontal semicircular canalK, kinociliaL, lagenaLc, laciniaLM, lagenar maculaLO, lagenar otolithN, vestibular and auditory branches of the eighth nervePC, posterior semicircular canalS, sacculusSBC, suprabranchial chamberSL, standard lengthSM, saccular maculaSO, saccular otolithSt, striolaU, utriculusUM, utricular maculaUO, utricular otolith1IntroductionTeleost fishes have evolved a diverse array of sound-generating organs, and many species regularly vocalize during social interactions (Myrberg, 1981; Ladich, 1997; Ladich and Bass, 1998; Zelick et al., 1999). Interestingly, only a few groups of fishes are known in which sonic organs are taxonomic characteristics. One such a group is the ‘arioids’, a suborder of siluriforms which includes five families such as the African mochokids (upside down catfishes) and the neotropical doradids (thorny catfishes). This group is characterized by having an elastic spring mechanism that is used in sound production (Lundberg, 1993). In many other fish groups, such as gadids (cods and relatives) and pomacentrids (damselfishes), some genera are known to be vocal while others are not (Hawkins and Rasmussen, 1978; Hawkins and Myrberg, 1983). Furthermore, within mormyrids (elephant-nose fishes), sound generation is only described in the genus Pollimyrus (Crawford et al., 1997; Crawford and Huang, 1999), and within anabantoids (gouramis) in the genera Trichopsis and Colisa (Kratochvil, 1985; Schuster, 1986; Ladich et al., 1992; Bischof, 1996). Within the family Cyprinidae (minnows and carps), a group which comprises about 2000 species, only representatives of a few genera such as Cyprinella, Gobio and Pimephales are known sound producers (Winn and Stout, 1960; Ladich, 1988; Johnston and Johnson, 2000).This limitation of sound production to certain genera of particular families has several behavioral and physiological implications which are poorly understood. Non-vocal species may primarily utilize other channels for intraspecific communication such as vision or chemoreception. Nelissen (1978) analyzed acoustical behavior of cichlid species and found that the number of sounds produced is inversely correlated to the number of color patterns shown. Fish which do not display visually or acoustically during social interactions might communicate chemically via pheromones (goldfish, Sörensen et al., 1989), via the lateral line (salmon, Satou et al., 1994), or through use of electric signals (knifefishes, Hagedorn and Heiligenberg, 1985).Does the occurrence of sound-generating mechanisms and acoustical communication correlate with the structure and function of the inner ear? Correlations between the region of the greatest hearing sensitivity and sound energy spectra in the characid Serrasalmus nattereri, the toadfish Porichthys notatus, the pomacentrids Eupomacentrus spp., and the gourami Trichopsis vittata support this idea (Cohen and Winn, 1967; Myrberg and Spires, 1980; Stabentheiner, 1988; Ladich and Yan, 1998; McKibben and Bass, 1999). However, comparative studies which also included closely related non-vocal species failed to find vocalization-specific differences in auditory sensitivity among anabantoids and otophysans (Ladich and Yan, 1998; Ladich, 1999).In order to determine if structural differences exist in the inner ear between teleosts that use acoustical communication and those which exploit other communication channels, we examined the ear and hearing ability of several anabantoid species (gouramis or labyrinth fishes). These are a group of small perciforms from Southeast Asia and Africa which differ widely in their methods of intraspecific communication. Only representatives of the genus Trichopsis (croaking gouramis) possess an elaborate sound-producing mechanism based on modifications of the pectoral fins (Kratochvil, 1978). Both males and females of these species emit loud croaking sounds during agonistic behavior. Closely related genera, such as the Siamese fighting fish Betta splendens and the paradise fish Macropodus opercularis, lack sonic organs and do not vocalize except with pharyngeal teeth as an incidental by-product of feeding. Betta and Macropodus, however, produce intensive visual displays by erecting fins, opercula, and gill membranes and by pronounced color changes (Simpson, 1968; Bischof, 1996).Contrary to this variation in the presence of sonic organs, all of the labyrinth fishes are characterized by having a ‘labyrinth’ which is an air-filled cavity located dorsal to the gills (=suprabranchial chamber). These chambers are supplementary respiratory organs utilized for air breathing since these fishes live in poorly oxygenated waters. At the same time, the chambers are located lateral to the inner ears and thereby potentially serve to enhance the hearing ability of these fish (Schneider, 1941; Ladich and Yan, 1998; Yan, 1998). Prior comparison of the saccular sensory epithelia in the blue gourami, Trichogaster trichopterus, and the kissing gourami, Helostoma temmincki, representatives of two different anabantoid families (Belontiidae vs. Helostomatidae), revealed major differences in the size and number of sensory cells of the saccule, despite the two species having similar hearing capabilities (Saidel and Popper, 1987).The goal of the present study was to investigate the ultrastructure of the otic end organs of T. vittata, M. opercularis, and B. splendens. We have analyzed the shape of the sensory epithelia (maculae), distribution of the hair cell types, and the hair cell orientation patterns on the saccular, lagenar and utricular sensory epithelia. In addition, we compared hair cell densities and the size of the saccular maculae in all three species since the saccule is generally considered to be the main hearing end organ in most fishes (Popper and Fay, 1999).2Materials and methodsSix specimen of B. splendens (1.27–1.9 g; 34–42 mm in standard length), eight of M. opercularis (1.41–2.65 g; 34–45 mm SL), and nine of T. vittata (0.29–2.0 g; 25–44 mm SL) obtained from aquarium stores were used for investigations. Fish were kept at 28°C in aerated and filtered aquaria and fed daily until used in the study. All work was done with the approval of the Institutional Animal Care and Use Committee of the University of Maryland.2.1Scanning electron microscopyFish were deeply anesthetized with tricaine methanosulfonate (MS 222, Sigma) and the trunk of the fish, opercula, and gills were dissected away. Cold fixative (2% glutaraldehyde in 0.1 M phosphate or cacodylate buffer) was put into the opened ear capsules. The head was then placed in fixative for at least 24 h. The tissue was then rinsed twice in buffer before the ears were dissected out and otoliths removed. The ears were post-fixed in 1–2% osmium tetroxide for 1 h, washed three times in double distilled water, and dehydrated through a graded series of alcohols to 100% ethanol. Sensory epithelia were then critical point dried with liquid carbon dioxide as the intermediate fluid. The samples were mounted flat on aluminum stubs, coated with gold-palladium alloy and viewed with an Amray 1820 or a Philips XL 20 SEM operating at 20 kV.2.2Density measurements and statisticsIn order to compare the density of ciliary bundles, low power scanning electron micrographs were taken perpendicular to the surface of the maculae of six B. splendens, seven M. opercularis and five T. vittata. Three areas were sampled from each saccule, one in the rostral region of the epithelium in which the ciliary bundles on the hair cells are oriented along the fish’s horizontal axis (see below), and one each in the middle and caudal (where the macula curves ventrally) regions in which ciliary bundles are oriented along the fish’s vertical axis. Although this procedure may not give precise density values because of the shrinkage of the tissue during fixation, it does not effect a general comparison since shrinkage would be about the same in all specimens. The same regions of each end organ were sampled in all specimens. In all cases, peripheral margins of the sensory epithelia were avoided because density increases on the periphery and due to the presence of otolithic membrane which, in these species, were almost impossible to remove from the very edges of the maculae with any consistency.The density (number of hair cells/1000 μm2) was calculated from each of the three sample areas in pictures taken at 3000×. Each picture was 31×24 μm. Approximately 2000 ciliary bundles were counted. Statistical analysis between species were calculated for each area as well as for the mean using one-way ANOVA. In addition, the total length of the saccular macula in the anterior–posterior axis was determined using the scanning electron micrograph (SEM) measuring device.The hair cell orientation patterns were mapped by examining the orientation of ciliary bundles along transects of the maculae. Orientation of transects generally followed the orientation of ciliary bundles. Dorsal–ventral transects were taken from the caudal part of the saccular epithelium while anterior–posterior transects were taken from the anterior part of the saccular macula. Radially oriented transects were taken from the utricular macula, and anterior–posterior transects from the lagena. At least 10 such transects were taken from each epithelium. The orientations of ciliary bundles in each region were marked on a low power picture of the whole epithelium. Orientation patterns were determined for at least two specimens for each species.3Results3.1Gross morphologyThe gross morphology of the ear of all three species is basically similar to that found in most other teleosts. There are three semicircular canals and associated cristae (for detection of angular acceleration) and three otolithic end organs (saccule, lagena, utricle), each containing a sensory epithelium and an overlying single otolith (Fig. 1a,b). No macula neglecta was found. The saccule is the largest otic end organ and the lagena the smallest. The utricle is anterior and dorsal to the saccule and the two regions are connected by a narrow opening. The lagena is located dorso-posteriorly to the saccule. The utricular macula lies on the animal’s horizontal plane, while the saccular and lagenar maculae lie in the vertical plane.Otoliths of the end organs differ considerably in size. The sagitta and the small asterisci fill almost all of the volume of the large saccular and the lagenar chambers respectively, whereas the lapilli do not fill the whole utricular chamber (Fig. 1). Sensory epithelia are smaller than the overlying otoliths especially within the saccule where the ovoid sagittae only contact the sensory macula at the sulcus, an L-shaped shallow indentation. However, despite the otolith’s size, the dorsal parts of the rostral region of the saccular maculae are not covered by the sagittae. In this region, the otolith membrane, which lies between the sensory epithelium and otolith, is the only contact for the ciliary bundles.The otolithic end organs in all three species are innervated by four branches of the eighth cranial nerve, an anterior utricular, two saccular, and one posterior lagenar branch. The first saccular branch leaves the brain along with the utricular nerve and innervates the oval rostral region of the saccular maculae. This branch also goes to the cristae of the anterior and lateral semicircular canals. The posterior part of the saccular macula is innervated by a massive separate branch of the eighth nerve (Fig. 1b). The posterior branch innervates the lagena and the crista of the posterior semicircular canal.3.2Sensory epitheliaThe shapes of the sensory epithelia are very similar in all three species (Figs. 2, 4, 5). The saccular macula consists of two regions, the oval anterior region and a more ribbon-shaped posterior region. The posterior region differs slightly between species. It was rather stretched in B. splendens and more S-shaped in M. opercularis and T. vittata (Fig. 2). The lagenar macula is moon-shaped, while the utricular macula is almost round with a lateral finger-like projection pointing laterally (=lacinia) (Figs. 4, 5).Morphological comparisons revealed that species differed in weight but not in standard length or length of the saccular maculae in the rostro-caudal dimensions (Table 1). Saccular epithelia ranged from 0.66 mm to 1.15 mm long and the size was correlated significantly with body measures (standard length: r=0.68, n=17, P=0.003; weight: r=0.58, P=0.016).Hair cells on each macula are divided into groups, with all of the ciliary bundles in each group oriented in approximately the same direction as defined by having their kinocilia on the same side of the ciliary bundle (see Fig. 6). Saccular hair cells are divided into four orientation groups. The rostral region of the saccular macula has posteriorly oriented hair cells in front of, and directed towards, a group of cells that are oriented anteriorly. The posterior region of the saccular macula has a dorsally oriented hair cell group on the dorsal half and a ventrally oriented group on the ventral half. These cells generally change orientation to follow macula curvatures (Fig. 2). The transitions between groups of hair cells, such as that from dorsal to ventral groups, occur over one or two rows of hair cells (Fig. 3).Lagenar hair cells are oriented dorsally on the anterior side of the epithelium while the cells on the posterior side are oriented ventrally (Fig. 4). Both cell groups tend to curve as the maculae curves. The regions at which the two orientation groups overlap by a small amount is usually called the transition zone. Utricular maculae differed from the other epithelia in the size of the two orientation groups. The cells along the narrow anterior border are oriented medially while the cells of the large central groups are oriented peripherally. This pattern is also found within the lacinia (Fig. 5), a lateral extension of the anterior margin of the utricular epithelium. The actual transition occurs in a narrow region that has been called the striola (e.g. Werner, 1933). In this region, as discussed below, the lengths of the ciliary bundles are somewhat smaller than in surrounding regions (called the extrastriola region).3.3Hair cell types and densitiesSensory epithelia in all three species consist of different types of hair cell ciliary bundles. The most common ciliary bundle type is F1 (definitions according to Popper, 1977) and consists of a series of short graded stereocilia and a kinocilium that is no more than two times longer than the longest stereocilia. Type F2 bundles, on the other hand, are characterized by short graded stereocilia, and kinocilia that are approximately three times longer than the largest stereocilia. Ciliary bundle types are fairly discrete, but there are intermediary forms making it difficult to discriminate between various bundle types.F1 bundles are found in the central region of the saccular macula and F2 ciliary bundles at the periphery (Fig. 6) of all three species. A common feature of all three species is the presence of one row of isolated F2 cells at the ventral saccular epithelium. This row of ciliary bundles is only found at the ribbon-shaped posterior part of these epithelia (Fig. 7).Within the utricular macula, type F1 bundles are found in the central region of the striola and in the extrastriolar region. F2 bundles are found at the margins of the striola and within the lacinia. Similarly, the central part of the moon-shaped lagenar maculae is built up of type F1 bundles whereas the periphery consists of type F2 bundles (Fig. 8). Densities of ciliary bundles varied but are always smaller in the central region of the saccular, lagenar and striolar maculae. Particularly low densities were observed within the extrastriolar regions of the utricle.Densities of hair cells ranged from 48.5 to 60.4 bundles per 1000 μm2 (Table 2) on all three maculae in the three species. No significant differences were found between species in the mean densities or in the densities of the anterior, middle or posterior areas. Differences were close to significance in the caudal area (Table 2). In addition, densities did not differ between the rostral, middle and caudal areas within species (one-way ANOVA, F2,48=1.97, P=0.15).4Discussion4.1Gross and fine structure of anabantoid earsThe ears of B. splendens, M. opercularis and T. vittata resemble each other and the typical pattern of most teleosts other than for the otophysan fishes (goldfish and relatives) (Retzius, 1881; Popper, 1977, 1981; Popper and Coombs, 1982). The saccule is by far the largest end organ and the lagena the smallest.The acoustical coupling of the inner ears of labyrinth fishes to the lateral air-filled suprabranchial chambers (SBCs) is based on the reduction of otic bones surrounding the saccule and separating it from the SBC. According to Schneider (1941), the thin bony sheet may be replaced by a membraneous window which markedly improves hearing in some individuals of Macropodus. In essence, Schneider (1941) first proposed that this air-filled chamber could serve as an accessory hearing organ, a role that is similar to the swim bladder in fishes that have connections between this organ and the inner ear, or air bubbles that are closely associated with the ear, as in the mormyrid fishes (Stipetić, 1939).In contrast to the lagenar and utricular maculae, the saccular maculae do not follow the shape of the otolith. The ovoid sagittae are much larger than the corresponding J-shaped sensory epithelia.4.2Hair cell orientation patternsHair cell orientation patterns within the lagenar and utricular maculae are quite similar to those described in other teleosts (Popper, 1977, 1980, 1981). However, significant differences from many other teleost groups are found in the saccule. Five hair cell orientation patterns have been identified on saccular maculae in different teleosts (Popper, 1981; Popper and Coombs, 1982). Except for the otophysans (goldfish and relatives) and a few non-otophysan species, saccular maculae generally have two horizontally and two vertically oriented groups of ciliary bundles (Popper and Coombs, 1982). The anabantoids described here represent the ‘opposing’ pattern described by Popper and Coombs (1982) in that the fish have two opposing horizontally oriented ciliary bundles on the anterior saccule. Vertically oriented bundles are found on the posterior ribbon-shaped part. This pattern is encountered in other anabantoids (e.g. Popper and Hoxter, 1981) and in the unrelated deep-sea myctophids (lantern fishes) (Popper, 1977).Popper and Coombs (1982) argued that there is a close correlation between the presence of an air bubble near the ear or a connection between the swim bladder and the ear with specializations in the structure of the anterior end of the saccular macula. They also argued for a correlation between such saccular epithelial specializations and enhanced hearing capabilities as compared to fishes that have no connections to an air bubble and the ‘general’ saccular hair cell orientation pattern. The results from the three species of anabantoids reported here tend to support this hypothesis.Thus, based upon the morphological results we not only show a strong correlation between the saccular structure of these three species and the presence of the suprabranchial organ as a hearing structure, but we would predict that were hearing studies to be done on the Siamese fighting fish Betta, it could be shown that it too has a wide hearing bandwidth (ranging from 100 Hz to 5 kHz).4.3Sensory hair cellsThe ciliary bundle types of gouramis investigated are basically similar to those found in other fishes. In the central populations the ratio of kinocilium to stereocilia length is between 1:1 and 2:1 (F1 type). At the margins of the sensory epithelia cells have bundles with ratios of about 3:1 (F2 type). These cells have very short stereocilia. Platt and Popper (1981, 1984) described similar patterns in the blue gourami T. trichopterus.4.4Interspecific differences between labyrinth fishesOur results show that all three species studied have specialized saccular sensory epithelia, although the function of this type of specialization is still not understood (Popper and Fay, 1999). These results support the data in the literature (Ladich and Yan, 1998) showing that two of these species, Trichopsis and Macropodus, have an extended hearing bandwidth and increased sensitivity when compared to fishes without such specializations. Fishes without hearing specializations tend to have saccular maculae where the rostral region is not very much deeper (from dorsal to ventral margin) than the caudal region. While we have not done statistical comparisons, examination of the literature shows that the rostral end of the saccular epithelium in non-specialists is generally less than twice the depth of the caudal region (e.g. Popper, 1977, 1981). In these anabantoids, however, the rostral end of the saccular macula is three times as deep as the caudal macula and the caudally oriented cells are in front of the anteriorly oriented cells. Further supporting the hypothesis that hearing specializations are paralleled by changes in the shape and/or the hair cell orientation pattern on the rostral region of the saccular macula (Popper and Coombs, 1982) is the finding of Saidel and Popper (1987) which showed that the blue gourami T. trichopterus, a species with a saccular specialization similar to that reported here, had lower hearing thresholds than the kissing gourami H. temmincki that has a ‘general’ saccular pattern.Comparison of shape and size of sensory epithelia as well as hair cell orientation patterns did not reveal any significant differences between the species investigated. The J-shaped saccular epithelium is also found in the blue gourami (Popper and Hoxter, 1981; Saidel and Popper, 1987) and seems to be a characteristic of the family Belontiidae. Differences exist in the shape of the saccular maculae between belontiids and the kissing gourami H. temmincki, the only representative of the family Helostomatidae. The rostral portion of the saccular macula of this species is widened but lacks the characteristic ventral extension of belontiids. In addition, the total area of the saccular macula in Helostoma is approximately 40% larger as compared to similar-sized specimens of T. trichopterus.Differences in hair cell densities are minor among representatives of the belontiid family. (While we did not do a statistical analysis on cells at the very margins of the epithelia, examination of tissue from all of our animals leads to the impression that the density was similar between species.) No significant differences were found between B. splendens, M. opercularis and T. vittata and densities of ciliary bundles appear to be similar to those found in T. trichopterus. However, the densities of hair cells in T. trichopterus were slightly higher than in H. temmincki in each sample area, again indicating an interfamiliar difference between anabantoids. With the larger area and slightly higher hair cell densities, Helostoma had about 90% more hair cells than T. trichopterus (44 000 vs. 23 000, respectively – Saidel and Popper, 1987).Do these differences between families account for any difference in hearing abilities? Saidel and Popper (1987) observed that the number of hair cells in the saccules of these two anabantoid fishes inversely correlates with the saccular microphonic thresholds. The somewhat lower microphonic thresholds were almost always obtained from T. trichopterus which had fewer hair cells than Helostoma. In general, the auditory sensitivities of the fish resembled one another in that both species have very similar thresholds and bandwidth (Saidel and Popper, 1987; Yan, 1998). So far it is unknown if differences in hair cell densities result in any physiological differences such as sound localization abilities, frequency discrimination or detection of masked sound.4.5Behavioral and evolutionary considerationsOur findings, and the similarity in gross and fine structure of the inner ears of B. splendens, M. opercularis, T. vittata and T. trichopterus, suggest that no major differences in hearing sensitivities are to be expected among the belontiid family. Ladich and Yan (1998) measured auditory sensitivities by utilizing auditory brainstem response (ABR) recording techniques and observed the lowest thresholds in T. trichopterus (76 dB re 1 μPa at 800 Hz). The paradise fish and the croaking gourami were about 8 dB less sensitive and possessed higher most sensitive frequencies (MSFs) than the blue gourami but otherwise did not differ from one other. This difference in hearing capacities among belontiids might rather be attributed to the size differences of the peripheral suprabranchial chambers than to fine structural differences of the inner ears such as number of hair cells. One possibility is that smaller pharyngeal air bubbles in smaller species such as Macropodus and Trichopsis might have higher resonant frequencies than larger bubbles, but resonance properties of constrained air bubbles, such as these (and the swim bladder) are hard to predict based upon bubble size alone (Popper, 1971, 1974; Sand and Enger, 1973a,b; Clarke et al., 1975).Are differences in auditory sensitivities between the anabantoids investigated (Ladich and Yan, 1998) related to their acoustical behavior? Dominant frequencies of sounds correspond with best hearing bandwidth in T. vittata (1–2 kHz). However, B. splendens, M. opercularis and T. trichopterus which communicate primarily using visual signals, and only incidentally with sound, have similar or even better hearing sensitivities than T. vittata and did not differ in their inner ear morphology. These findings strongly suggest that enhanced sound-detecting abilities in anabantoids evolved prior to, or independently of, the evolution of pectoral sound-generating mechanisms in the genus Trichopsis. Most likely the SBC is an adaptation to waters poor in oxygen and the improvement of hearing a by-product of this process. A similar development obviously took place in otophysans. The evolution of the Weberian apparatus resulted in a pronounced improvement of hearing capacities in cypriniforms, catfishes, and characiforms. However, auditory sensitivities in vocalizing otophysans resemble those of non-vocalizing species and most sensitive frequencies lack a clear relationship to the main energies of sounds (Ladich, 1999). Interestingly, only a few representatives of cypriniforms, the most primitive otophysan group, are known to be vocal and in no case was there any vocal organ described.This again points to the fact that constraints other than optimization of acoustical communication is likely to have caused the ancestors of otophysans, and perhaps fishes in general, to develop and enhance their sound-detecting abilities. A major selective pressure might have been the analysis of the auditory scene which means separating sounds of different origins due to differences in their frequency content or temporal patterns (Bregman, 1990). This could have been a major advantage in avoiding predators and detecting prey (Popper and Fay, 1993; Ladich, 1999).If one accepts the idea that detection of the general auditory scene is of importance for fishes, and particularly for fishes that may live in dark or murky waters, the evolution of mechanisms to enhance detection of higher frequencies than found in most hearing non-specialists may have been a response to particular environments in which species ancestral to hearing specialists lived. Rogers and Cox (1988) pointed out that the transmission characteristics of sound in shallow water (up to several meters in depth) are quite different than in deeper water. In shallow waters, there is a very high attenuation rate for low frequencies that is correlated with depth. While this would not have any impact on fishes living in deeper waters, fishes living in shallow waters would only be able to detect sounds from very near-by sources, unless they have evolved hearing to higher frequencies (Schellart and Popper, 1992). We speculate that anabantoids, like otophysans, clupeids (herrings and relatives), and a number of other diverse teleost taxa may have arisen in shallow waters and evolved high frequency hearing. The fact that the specificities of the specializations are so different in various species argues for the evolution of high frequency hearing several times. For example, as the anabantoids evolved the ability to use the air bubble in the pharyngeal chamber for high frequency hearing, otophysans evolved Weberian ossicles while clupeids evolved a highly specialized utricle (e.g. Mann et al., 1997, 1998).In contrast, there are many other fishes currently living in relatively shallow waters that do not have hearing specializations. We suggest that living in shallow water is secondary to these species, and that ancestral species either evolved in deeper waters, or there were no other selective pressures on the ancestors that could have lead to higher frequency hearing.AcknowledgementsWe thank Marcy Souza, Tim Maugel, Christian Beisser and Anton Losert for their technical advice and assistance and Heidemarie Grillitsch for the line drawings. This research was supported by the Austrian Science Fund (FWF Grant No. 12411 to F.L. and grant DC03936-01 from the National Institute on Deafness and Other Communication Disorders of the NIH to A.N.P.).ReferencesBischof, 1996C.BischofDiversity in agonistic behavior of croaking gouramis (Trichopsis vittata, T. schalleri, and T. pumila; Anabantoidei) and the paradise fish (Macropodus opercularis; Anabantoidei)Aggr. Behav.221996447455Bregman, 1990Bregman, A., 1990. Auditory Scene Analysis: The Perceptual Organization of Sound. MIT Press, Cambridge, MA.Clarke et al., 1975N.L.ClarkeA.N.PopperJ.A.MannJr.Laser light scattering investigations of the teleost swimbladder response to acoustic stimuliBiophys. J.151975307318Cohen and Winn, 1967M.J.CohenH.F.WinnElectrophysiological observations on hearing and sound production in the fish Porichthys notatusJ. Exp. Zool.1651967355370Crawford et al., 1997J.D.CrawfordA.P.CookA.S.HeberleinBioacoustic behaviour of African fishes (Mormyridae): potential cues for species and individual recognition in PollimyrusJ. Acoust. Soc. Am.1021997113Crawford and Huang, 1999J.D.CrawfordX.HuangCommunication signals and sound production mechanism of mormyrid electric fishJ. Exp. Biol.202199914171426Hagedorn and Heiligenberg, 1985M.HagedornW.HeiligenbergCourt and spark: electric signals in the courtship and mating of gymnotoid fishAnim. Behav.331985254265Hawkins and Myrberg, 1983Hawkins, A.D., Myrberg, A.A., 1983. Hearing and sound communication underwater. In: Lewis, B. (Ed.), Bioacoustics, a Comparative Approach. Academic Press, London, pp. 347–405.Hawkins and Rasmussen, 1978A.D.HawkinsK.J.RasmussenThe calls of gadoid fishJ. Mar. Biol. Assoc. U.K.581978891911Johnston and Johnson, 2000C.E.JohnstonD.L.JohnsonSound production in Pimephales notatus (Rafinesque) (Cyprinidae)Copeia20002000567571Kratochvil, 1978H.KratochvilDer Bau des Lautapparates vom Knurrenden Gurami (Trichopsis vittatus Cuvier and Valenciennes) (Anabantidae, Belontiidae)Zoomorphologie9119789199Kratochvil, 1985H.KratochvilBeiträge zur Lautbiologie der Anabantoidei-Bau, Funktion und Entwicklung von lauterzeugenden SystemenZool. Jahrb. Physiol.891985203255Ladich, 1988F.LadichSound production by the gudgeon, Gobio gobio L., a common European freshwater fish (Cyprinidae, Teleostei)J. Fish. Biol.321988707715Ladich, 1997F.LadichAgonistic behaviour and significance of sounds in vocalizing fishMar. Freshw. Behav. Physiol.29199787108Ladich, 1999F.LadichDid auditory sensitivity and vocalization evolve independently in otophysan fishes?Brain Behav. Evol.531999288304Ladich and Bass, 1998F.LadichA.H.BassSonic/vocal motor pathways in catfishes: comparison with other teleostsBrain Behav. Evol.511998315330Ladich et al., 1992F.LadichC.BischofG.SchleinzerA.FuchsIntra- and interspecific differences in agonistic vocalization in croaking gouramis (Genus: Trichopsis, Anabantoidei, Teleostei)Bioacoustics41992131141Ladich and Yan, 1998F.LadichH.Y.YanCorrelation between auditory sensitivity and vocalization in anabantoid fishesJ. Comp. Physiol. A1821998737746Lundberg, 1993Lundberg, J.G., 1993. African-South American freshwater fish clades and continental drift: problem with a paradigma. In: Goldblatt, P. (Ed.), Biological Relationship between Africa and South America. Yale University Press, New Haven, CT, pp. 156–199.Mann et al., 1997D.A.MannZ.LuA.N.PopperA clupeid fish can detect ultrasoundNature3891997341Mann et al., 1998D.A.MannZ.LuM.HastingsA.N.PopperDetection of ultrasonic tones and simulated dolphin echolocation clicks by a teleost fish, the American shad (Alosa sapidissima)J. Acoust. Soc. Am.1041998562568McKibben and Bass, 1999J.R.McKibbenA.H.BassPeripheral encoding of behaviorally relevant acoustic signals in a vocal fish: single tonesJ. Comp. Physiol. A1841999563576Myrberg, 1981Myrberg, A.A. Jr., 1981. Sound communication and interception in fishes. In: Tavolga, W.N., Popper, A.N., Fay, R.R. (Eds.), Hearing and Sound Communication in Fishes. Springer, New York, pp. 395–425.Myrberg and Spires, 1980A.A.MyrbergJr.J.Y.SpiresHearing in damselfishes: an analysis of signal detection among closely related speciesJ. Comp. Physiol. A1401980135144Nelissen, 1978M.H.J.NelissenSound production by some tanganyikan cichlid fishes and a hypothesis for the evolution of the R communication mechanismsBehaviour641978137147Platt and Popper, 1981Platt, C., Popper, A.N., 1981. Fine structure and function of the teleost ear. In: Tavolga, W.N., Popper, A.N., Fay, R.R. (Eds.), Hearing and Sound Communication in Fishes. Springer, New York, pp. 3–36.Platt and Popper, 1984C.PlattA.N.PopperVariation in lengths of ciliary bundles on hair cells along the macula of the sacculus in two species of teleost fishesScann. Electr. Microsc.4198419151924Popper, 1971A.N.PopperThe effects of size on the auditory capacities of the goldfishJ. Audit. Res.111971239247Popper, 1974A.N.PopperThe response of the swimbladder of the goldfish (Carassius auratus) to acoustic stimuliJ. Exp. Biol.601974295304Popper, 1977A.N.PopperA scanning electron microscopic study of the sacculus and lagena in the ears of fifteen species of teleost fishesJ. Morphol.1531977397417Popper, 1980A.N.PopperScanning electron microscopic studies of the sacculus and lagena in several deep-sea fishesAm. J. Anat.1571980115136Popper, 1981A.N.PopperComparative scanning electron microscopic investigations of the sensory epithelia in the teleost sacculus and lagenaJ. Comp. Neurol.2001981357374Popper and Coombs, 1982A.N.PopperS.CoombsThe morphology and evolution of the ear in actinopterygian fishAm. Zool.221982311328Popper and Fay, 1993A.N.PopperR.R.FaySound detection and processing by fish: critical review and major research questionsBrain Behav. Evol.4119931438Popper and Fay, 1999Popper, A.N., Fay, R.R., 1999. The auditory periphery in fishes. In: Fay, R.R., Popper, A.N. (Eds.), Comparative Hearing: Fish and Amphibians. Springer, New York, pp. 43–100.Popper and Hoxter, 1981A.N.PopperB.HoxterThe fine structure of the sensory epithelia of the sacculus and lagena of a teleost fishHear. Res.51981245263Retzius, 1881Retzius, G., 1881. Über das Gehörorgan der Wirbelthiere: Morphologisch-histologische Studien, Vol. 1. Das Gehörorgan der Fische und Amphibien. Samson and Wallin, Stockholm.Rogers and Cox, 1988Rogers, P., Cox, M., 1988. Underwater sound as a biological stimulus. In: Atema, J., Fay, R.R., Popper, A.N., Tavolga, W.N. (Eds.), Sensory Biology of Aquatic Animals. Springer, New York, pp. 131–149.Saidel and Popper, 1987W.M.SaidelA.N.PopperSound reception in two anabantid fishesComp. Biochem. Physiol.8819873744Sand and Enger, 1973aO.SandP.S.EngerEvidence for an auditory function of the swimbladder in the codJ. Exp. Biol.591973405414Sand and Enger, 1973bSand, O., Enger, P.S., 1973b. Function of the swimbladder in fish hearing. In: Moller, A. (Ed.), Basic Mechanisms of Hearing. Academic Press, New York, pp. 893–910.Satou et al., 1994M.SatouH.A.TakeuchiJ.NishiiM.TanabeS.KitamuraN.OkumotoM.IwataBehavioral and electrophysiological evidences that the lateral line is involved in the inter-sexual vibrational communication of the hime salmon (Landlocked red salmon, Oncorhynchus nerka)J. Comp. Physiol. A1741994539549Schellart and Popper, 1992Schellart, N.A.M., Popper, A.N., 1992. Functional aspects of the evolution of the auditory system of actinopterygian fish. In: Webster, D.B., Fay, R.R., Popper, A.N. (Eds.), The Evolutionary Biology of Hearing. Springer, New York, pp. 295–322.Schneider, 1941H.SchneiderDie Bedeutung der Atemhöhle der Labyrinthfische für ihr HörvermögenZ. Vergl. Physiol.291941172194Schuster, 1986S.SchusterUntersuchungen zur Bioakustik des Zwergfisches Colisa lalia (Perciformes, Belontiidae)Zool. Beitr.291986295306Simpson, 1968M.J.ASimpsonThe display of the Siamese fighting fish Betta splendensAnim. Behav. Monogr.11968173Sörensen et al., 1989P.W.SörensenN.E.StaceyK.J.ChamberlainDiffering behavioral and endocrinological effects of two female sex pheromones on male goldfishHorm. Behav.231989317332Stabentheiner, 1988A.StabentheinerCorrelations between hearing and sound production in piranhasJ. Comp. Physiol. A16219886776Stipetić, 1939E.StipetićÜber das Gehörorgan der MormyridenZ. Vergl. Physiol.261939740752Werner, 1933C.F.WernerDie Differenzierung der Maculae im Labyrinth, insbesondere bei SäugetierenZ. Anat. Entwicklungsgesch.991933696709Winn and Stout, 1960H.E.WinnJ.F.StoutSound production by the satinfin shiner, Notropis analostanus, and related fishesScience1321960222223Yan, 1998H.Y.YanAuditory role of the suprabranchial chamber in gourami fishJ. Comp. Physiol. A1831998325333Zelick et al., 1999Zelick, R., Mann, D., Popper, A.N., 1999. Acoustic communication in fishes and frogs. In: Fay, R.R., Popper, A.N. (Eds.), Comparative Hearing: Fish and Amphibians. Springer, New York, pp. 363–411.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B980D08A04604E69176319ABDFB77351E8A67B5.txt b/test/dataset/in/resources/corpus/Clean_0B980D08A04604E69176319ABDFB77351E8A67B5.txt
deleted file mode 100644
index feba1ad..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B980D08A04604E69176319ABDFB77351E8A67B5.txt
+++ /dev/null
@@ -1 +0,0 @@
-geronbgeronbJournals of Gerontology: Series B1758-53681079-5014Oxford University Press10.1093/geronb/gbp003Journal of Gerontology: Psychological SciencesArticlesEffects of Optic Flow Speed and Lateral Flow Asymmetry on Locomotion in Younger and Older Adults: A Virtual Reality StudyChouYing-hui1WagenaarRobert C.1SaltzmanElliot1GiphartJ. Erik1YoungDaniel1DavidsdottirRosa2Cronin-GolombAlice21Department of Physical Therapy and Athletic Training, Boston University, Massachusetts2Department of Psychology, Boston University, MassachusettsAddress correspondence to Ying-hui Chou PhD, ScD, Department of Occupational Therapy, Medical College, Fu-Jen Catholic University, No. 510, Zhongzheng Road, Xinzhuang City, Taipei County 242, Taiwan. Email: yinghuichou@gmail.com32009103200964B222223182200820122008© The Author 2009. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org.2009The purpose of the study is to investigate whether there are age-related differences in locomotion due to changes in presence of vision, optic flow speed, and lateral flow asymmetry using virtual reality technology. Gait kinematics and heading direction were measured using a three-dimensional motion analysis system. Although older and younger adults were affected differentially by the availability of vision, a greater dependence on optic flow information in older adults during walking was not found. Linear relations were observed between walking performance and flow speed as well as heading direction and flow asymmetry. The findings suggest that the ability to integrate optic flow information into the multimodal system for assessment of walking speed and heading direction is comparable in older and younger adults.Optic flow speedVirtual realityLocomotionInterlimb coordinationHeading directionVisual dependenceTHE control of locomotion depends on the integration of information from visual, vestibular, and somatosensory systems. When information from one or more sensory systems is withdrawn or becomes less reliable, the central nervous system changes its weighting of sensory inputs to maintain appropriate postural responses (van der Kooij, Jacobs, Koopman, & van der Helm, 2001). A number of studies have reported that older adults are more dependent on the availability of vision for posture and gait than are younger adults. For example, older adults show greater sway with eyes closed during quiet standing (Choy, Brauer, & Nitz, 2003; Hay, Bard, Fleury, & Teasdale, 1996; Manchester, Woollacott, Zederbauer-Hylton, & Marin, 1989) and display diminished head stability with eyes closed during walking (Cromwell, Newton, & Forrest, 2001, 2002) compared with younger adults. A greater reliance on vision in older adults is considered a central sensory reweighting deficit (Teasdale, Stelmach, & Breunig, 1991) and/or age-related peripheral sensory loss in the vestibular and somatosensory systems. However, some studies (Anderson, Mulder, Nienhuis, & Hulstijn, 1998; Konczak, 1994; Schubert, Prokop, Brocke, & Berger, 2005) did not find significant differences in locomotion between younger and older adults due to changes in visual information, especially when optic flow parameters were involved. Additionally, research findings on age-related changes in perception of optic flow are equivocal. Billino, Bremmer, and Gegenfurtner (2008) and Atchley and Andersen (1998) found that perception of radial optic flow was not affected by age, whereas Warren, Blackwell, and Morris (1989) and Gilmore, Wenk, Naylor, and Stuve (1992) reported increased thresholds for heading detection and motion coherence, respectively, with age. Whether older adults are more dependent on optic flow information compared with younger adults during locomotion is still open to further investigation.Optic flow patterns and their changes are used to assess speed and direction of self-motion (Gibson, 1958). Previous studies demonstrate that gait parameters are inadvertently modulated by manipulations of optic flow speed. When optic flow speed decreases, walking speed (Prokop, Schubert, & Berger, 1997; Schubert et al., 2005; Varraine, Bonnard, & Pailhous, 2002) and stride length (Prokop et al., 1997; Schubert et al., 2005) increase; decreasing optic flow speed shows opposite effects. The manipulation of lateral asymmetry of optic flow speed influences heading direction in honey bees (Srinivasan, Lehrer, Kirchner, & Zhang, 1991) and young adults (Duchon & Warren, 2002). Bees and young adults steered away from the faster moving wall with larger differences of optic flow speeds, resulting in more drift. The purpose of the present study is to examine whether there are age-related changes in gait patterns and heading direction as a function of the presence of vision (eyes open vs. blindfolded vs. virtual environment), optic flow speed, and lateral flow asymmetry.In addition to walking speed and stride parameters, how limbs dynamically coordinate in response to changes in optic flow speed during locomotion remains to be determined. Wagenaar and colleagues have shown that shifts between patterns of interlimb coordination would occur when an appropriate control parameter is scaled (Donker, Beek, Wagenaar, & Mulder, 2001; Wagenaar & van Emmerik, 2000). For example, when walking speed (i.e., a control parameter) is systematically increased from 0.2 to 1.6 m/s, the arm-to-leg movement frequency ratio gradually changes from 2:1 (a less stable pattern) to 1:1 (a more stable pattern). The transition between coordination patterns is accompanied by an increased standard deviation of relative phase between arm and leg movements (Wagenaar & van Emmerik, 2000). Clinically, the degree of flexibility in interlimb coordination has been found to correlate with bradykinesia and rigidity in Parkinson's disease (Winogrodzka, Wagenaar, Booij, & Wolters, 2005) and with asymmetry in stroke (Ford, Wagenaar, & Newell, 2007). Older adults with risk for falls showed a reduced flexibility in their coordination of walking at speeds higher than 0.8 m/s compared with the walking patterns of healthy older adults (Wagenaar, Holt, Kubo, & Ho, 2003). The present study will investigate whether optic flow speed is an appropriate control parameter for changes in the coordination dynamics during walking and whether the coordination dynamics in response to changes in optic flow speed is age-related.The first hypothesis is that older adults are more dependent on the presence of visual information during walking compared with younger adults, as shown by a reduced walking speed, stride length, and stride frequency during the blindfolded condition compared with the conditions of normal vision and virtual reality. Second, we expect that older adults will modulate their walking performance and heading direction to a greater extent in response to changes in optic flow speed and lateral flow asymmetry than younger adults. Third, for both younger and older adults, we hypothesize that walking speed, stride length, and stride frequency will be modulated linearly with manipulations of optic flow speed during overground walking. For example, increasing optic flow speed will result in a reduced walking speed, stride frequency, and stride length. It is also expected that the interlimb coordination will demonstrate a change from a 1:1 to a 2:1 arm-and-leg frequency ratio with increasing optic flow speed. Finally, with regard to heading direction, it is hypothesized that participants will drift away from the wall that is moving faster, and the degree of lateral drift will be proportional to the level of lateral flow asymmetry.METHODSParticipantsThe participants consisted of two groups: 16 younger adults (8 men, 8 women; 15 right-handers, 1 left-hander; mean age 22.2 ± 3.4 years; age range 18−30 years) and 17 older adults (8 men, 9 women; 16 right-handers, 1 left-hander; mean age 60.5 ± 8.7 years, age range 46−73 years). Older adults were recruited through local newspaper advertisements and personal contacts, and younger adults were undergraduates at Boston University. Exclusion criteria included walking disability or history of leg, knee, hip impairments, serious cardiac disease, other serious chronic medical illness, history of traumatic brain injury, psychiatric or neurological diagnoses, history of alcoholism or other drug abuse, or history of eye disease or other abnormalities as noted on neuro-ophthalmological examination. Individuals whose corrected binocular acuity was poorer than 20/40 were excluded. Informed consent approved by the Institutional Review Board of Boston University and conforming to the 1964 Declaration of Helsinki was obtained prior to participation. All participants were paid for participation.MaterialThe VR system.—A virtual hallway was created using World ToolKit Release 9 (Sense8, San Francisco, CA) on an Onyx2 Reality graphics work station (Silicon Graphics, Inc., Mountain View, CA). The hallway was composed of two sidewalls of white random dots on a black background (Figure 1). The width of the hallway was 2 m, the height of the hallway 2.55 m, and the depth of the hallway 8.88 m. The middle of the virtual hallway along the anteroposterior axis was always aligned with a reference axis in the real environment. The orientations of the virtual hallway in the frontal and sagittal planes were calibrated with respect to each participant's head orientation, whereas the participant was instructed to stand upright and face forward. The visual scene was displayed on a ProView 60 head-mounted display (HMD; Kaiser Opto-Electronics, Inc., Mountain View, CA) weighting 1.75 lbs. This HMD contained two active LCD panels (640 × 480 resolution, true color, 60 Hz) and had a 60° field-of-view (diagonal) with 100% overlap to allow for true stereo viewing. Participant's field of view was restricted to the VR environment by a mask that occluded vision outside the LCD panels. Head coordinates were tracked in real time, using an IS 900 LAT system (InterSense, Burlington, MA), and the information was used to update the visual scene with a delay of four frames (67.7−83.3 ms). The movement tracking system covered the overground walking area with an approximate error of less than 4 mm root mean square for position data and 0.1° for orientation data. The optic texture of the virtual hallway's walls in the anteroposterior direction could be manipulated by a computer. The validity of this VR simulation has been demonstrated by Giphart, Chou, Kim, Bortnyk, and Wagenaar (2007).Figure 1.Display of virtual reality environment.Three-dimensional kinematics.—Three-dimensional kinematic data were collected using an Optotrak 3020 System (Northern Digital Inc., Waterloo, ON, Canada), with a spatial resolution of 0.1 mm. One position sensor was placed on each side of the walkway, and a third position sensor was located at the end of the walkway in order to provide an environmental reference plane for capturing bilateral locomotor movements for at least eight strides. Calibrations among the three position sensors were accepted when the mean error was 1.5 mm or less. Infrared light-emitting diodes (IREDs) were placed on participant's chin (lower mandible) and bilaterally on the ankle (lateral calcaneus), hip (anterior superior iliac spine), wrist (radiocarpal joint), and shoulder (humeral head). The instantaneous position of each IRED was sampled at a rate of 100 Hz and stored to disk for further analysis.Experimental ProcedureAfter intake, three experimental conditions were carried out in a fixed order, that is, Trial 1 (manipulation of vision), Trial 2 (manipulation of optic flow speed), and Trial 3 (manipulation of lateral flow asymmetry). The experimental procedure was described as follows.Intake.—Visual dependence was assessed by using a rod and frame test (Azulay, Mesure, Amblard, & Pouget, 2002). Participants were presented with a 1.5-m horizontal line that was slightly tilted at the outset of each run (initial tilt ranging from 9° to 12°) in a large screen. The angle of the line was gradually altered to become more horizontal by 1° increments. Participants were asked to indicate when the rod was horizontal. Mean deviation (degree) averaged from 10 runs was taken as the measure of the participants’ level of visual dependence. All stimuli sequences were created using Adobe Photoshop and PowerPoint for presentation.Trial 1.—Participants were trained to walk 8 m at 0.8 ± 0.05 m/s because the state of interlimb coordination at this walking speed is relatively unstable (Wagenaar & van Emmerik, 2000). Therefore, if there were any changes in coordination pattern as a result of changes in optic flow speed, it would be observed more sensitively. Three vision conditions were presented in a fixed sequence: (1) eyes opened (EO) condition—participants were instructed to walk down the middle of a walkway in the laboratory with eyes open; (2) blindfolded (BF) condition—participants were instructed to walk down the middle of the walkway in the lab blindfolded; (3) virtual reality (VR) condition—participants were instructed to walk down the middle of a virtual hallway created by the virtual reality system. In the VR condition, the optic flow speed was set by feedback control to equal −1 times the participant's walking speed in the anteroposterior direction. The minus sign indicated that the optic flow was programmed to move in a direction opposite to the participant's walking direction. Participants had the opportunity to practice walking in the virtual environment for at least 10 min, until they demonstrated that they were able to walk comfortably at 0.8 ± 0.05 m/s. For each condition, at least five runs were performed. Each participant received feedback on walking speed after each run. The average walking speed was measured in real time with two sets of infrared switches (Safe House, Fort Worth, TX) placed 6 m apart along the walkway.Trial 2.—Five different optic flow speed conditions (0, −0.4, −0.8, −1.2, and −1.6 m/s) were presented in random order across participants. Five consecutive runs were performed for each optic flow speed condition, yielding 25 runs in total for each participant. Participants were instructed to walk at 0.8 m/s throughout the experimental session. During Trials 2 and 3, the speed of optic flow was held constant within a condition and not a function of the participant's walking speed. For example, when the optic flow speed was −0.8 m/s, participants had the impression that they were walking at the trained speed through a stationary hallway; when the optic flow speed was faster/slower than −0.8 m/s, participants had the impression that they were walking more quickly/slowly than the trained speed. A special case was that when the optic flow speed was equal to 0 m/s, participants had the impression that they were not moving forward relative to the hallway, even though they were walking (i.e., a situation similar to that experienced visually during walking on a treadmill). No feedback on walking speed was provided during Trials 2 and 3.Trial 3.—Participants were trained to walk at 0.8 ± 0.05 m/s again in the VR environment for at least five runs prior to Trial 3. In Trial 3, the optic flow speed in one visual field was always equal to −0.8 m/s, whereas the flow speed in the other visual field was 0, −0.4, −0.8, −1.2, or −1.6 m/s. The 10 different optic flow speed conditions were presented in random order across participants. Five runs were performed for each condition, yielding 50 (2 × 5 × 5) runs in total for each participant.Data AnalysisFor kinematic analysis, the position data were filtered using a zero-lag, fourth-order Butterworth low-pass filter with a cutoff frequency of 5 Hz. Shoulder and wrist time series were used to compute the angular displacement of arm movements and hip and ankle time series to calculate the angular displacement of leg movements relative to vertical. Forward wrist or ankle movement resulted in a positive angle. Each stride cycle was identified by two consecutive maxima from the angular position data of leg movements. All the variables were computed for only the middle six strides in order to avoid acceleration and deceleration variations at the beginning and at the end of the distance walked. The data were processed using MatLab (The MathWorks, Inc., Natick, MA).Walking speed and stride parameters.—The average walking speed was estimated by the displacement of the chin time series in the anteroposterior axis divided by the time it took the participant to travel for six stride cycles. Stride length was calculated from both the left and right ankle time series, separately, by dividing the displacement in the anteroposterior axis divided by 6 (i.e., the number of strides). Stride frequency was calculated in strides per second (as the inverse of the time it took the participant to travel six stride cycles divided by the number of strides).Relative power index.—The angular displacement data of the arm and leg movements were used for the analysis of interlimb coordination. Figure 2a shows an example of the time series data of arm and leg movements during walking at lower speeds. The power spectral density function was used to transform the time series data into movement frequencies and corresponding spectral power for leg (Figure 2b) and arm movements (Figure 2c). The spectral power was normalized by dividing each frequency power by the mean power calculated over the 0.2- to 2.5-Hz frequency range for each run.Figure 2.Fourier transform of time series of arm and leg movements.A step is heel contact of one leg to heel contact of the other leg, a stride from heel contact to heel contact of the same leg. As in Figure 2b, the frequency with the largest power was regarded as the stride frequency (Wagenaar & van Emmerik, 2000). The step frequency was the peak at twice the stride frequency. The corresponding power at the stride and step frequencies of arm movements was identified using the stride and step frequencies from the leg movements, respectively (see Figure 2c). To quantify the frequency ratio of arm and leg movements, a relative power index (RPI) was calculated using the following Equation:(1)where P1 is the power at the stride frequency observed in the arm movements and P2 is the power at the step frequency observed in the arm movements (Figure 2c). RPI ranges from −1 to 1. RPI equal to 1 indicates a 1:1 frequency ratio between arm and leg movements. RPI equal to −1 points out a 2:1 coordination pattern in which the arms cycle twice during every stride cycle of the legs.Drift.—The average of left and right hip position data in the mediolateral axis was used as the body reference point. The drift is defined as the difference in mediolateral coordinates between the last and the first stride. A positive value indicates rightward drift and a negative value indicates leftward drift. Each participant's drift data in Trials 2 and 3 would be normalized by his or her baseline measures of drift in the VR condition during Trial 1.Statistical AnalysisVisual dependence.—The difference of visual dependency between the older and younger adults was examined using a one-way analysis of variance.Trial 1: Manipulation of visioen.—Data across the five runs were averaged for each condition. Whether the average walking speed in the older adults and the younger adults was significantly different from 0.8 m/s was evaluated by means of Wilcoxon one-sample signed-rank tests (with a two-tailed level of significance alpha = .05). In order to test whether there were significant effects of group and vision conditions, a general linear model was performed with a within-subject factor for Vision (EO, BF, and VR) and a between-subject factor for Group (younger and older adults).Trial 2: Manipulation of optic flow speed.—Data across the five runs were averaged for each condition. A general linear model was performed with a linear contrast for Optic Flow Speed (within-subject factor, five levels) and a between-subject factor for Group.Trial 3: Manipulation of lateral flow asymmetry.—Data across the five runs were averaged for each condition. A general linear model was performed with a within-group factor for Visual Field (i.e., optic flow speed changed in the left or the right visual field), a linear contrast for Optic Flow Speed (within-subject factor, five levels), and a between-subject factor for Group.Post hoc analyses were employed in all three trials in the case of a significant interaction effect or a significant main effect of a variable with more than two levels. Multiple comparisons were conducted with Bonferroni corrections. All analyses were performed using SPSS (SPSS, Inc., Chicago, IL).RESULTSThere was no statistically significant difference between younger and older adults in visual dependency as measured by the rod and frame task, F(1, 31) = 0.68, p = .42.Trial 1: Manipulation of VisionWalking speed and stride parameters.—The walking speeds in the younger adults were not significantly different from 0.8 m/s (EO: p = .86; BF: p = .15; VR: p = .80). The mean walking speed in the older adults was significantly lower in BF than 0.8 m/s (p = .03). However, no significant differences were found for EO (p = .36) and VR (p = .75). At all conditions, the 95% confidence intervals of the mean walking speed were within 0.8 ± 0.05 m/s for both groups.Analysis of walking speed yielded a significant main effect for Vision, F(2, 62) = 18.76, p = .0001, and a significant interaction effect between Vision and Group, F(2, 62) = 6.33, p = .003 (Figure 3a). Post hoc analyses revealed that a significant effect of Vision was observed in the older adults, F(2, 32) = 17.97, p = .0001, but not in the younger adults, F(2, 30) = 2.66, p = .09. Post hoc analyses of the vision effect in the older adults showed that walking speed in BF was significantly lower than that in EO (p = .0001) and VR (p = .006).Figure 3.Influence of vision condition on (a) walking speed and (b) stride length in younger and older adults. Error bars show 95% confidence intervals around the means. EO = eyes opened; BF = blindfolded; VR = virtual reality.Significant main effects for Vision, F(2, 62) = 12.38, p = .0001, and Group, F(1, 31) = 4.42, p = .05, were observed for stride frequency. Older adults had a significantly higher stride frequency (1.10 Hz) than the younger adults (1.03 Hz). Post hoc analyses revealed that stride frequency in BF (1.14 Hz) was significantly higher than that in EO (1.01 Hz, p = .001) and VR (1.05 Hz, p = .001).A significant main effect for Vision, F(2, 62) = 32.12, p = .0001, and a significant Vision × Group interaction effect, F(2, 62) = 5.51, p = .006 (Figure 3b) were found for stride length. Post hoc analyses revealed that (1) there were significant differences among vision conditions in the older adults, F(2, 32) = 22.25, p = .0001, as well as the younger adults, F(2, 30) = 11.24, p = .0001, and (2) the group effect was significant only in BF, F(1, 31) = 8.24, p = .007, indicating that the older adults had significantly shorter stride length compared with the younger adults in BF. In addition, both groups showed that the greatest stride length was observed in EO, followed by VR, and then BF.The RPI.—Analysis of RPI yielded a significant main effect for Vision, F(2, 62) = 3.88, p = .03. Post hoc analyses showed that the RPI in VR (0.74 ± 0.34) was significantly greater than that in EO (0.65 ± 0.33, p = .03).Drift.—A significant main effect for Vision, F(2, 62) = 4.51, p = .02, was found. The average drifts are 32.30 (SD = 50.44), −119.82 (SD = 395.66), and 32.37 (SD = 62.20) mm for the EO, BF, and VR conditions, respectively. Post hoc analyses showed no significant effect among the three vision conditions.Trial 2: Manipulation of Optic Flow SpeedWalking speed and stride parameters.—The linear main effects for Optic Flow Speed were significant for walking speed, F(1, 31) = 16.42, p = .0001, and stride frequency, F(1, 31) = 22.87, p = .0001, indicating that both groups decreased walking speed (see Figure 4a) and stride frequency (see Figure 5a) when the optic flow speed was increased. The main Group effects were significant for walking speed, F(1, 31) = 6.38, p = .02, and stride frequency, F(1, 31) = 9.52, p = .004, showing that older adults walked at a higher speed (0.86 ± 0.08 m/s) and stride frequency (1.09 ± 0.11 Hz) than the younger adults (0.81 ± 0.08 m/s; 1.00 ± 0.09 Hz, respectively). No significant interaction effects between Group and Optic Flow Speed were found for walking speed, F(1, 31) = 0.42, p = .52, and stride frequency, F(1, 31) = 1.43, p = .24. No significant main or interaction effects were found for stride length.Figure 4.Influence of optic flow speed on walking speed in (a) Trial 2 and (b) Trial 3. Error bars show 95% confidence intervals around the means.Figure 5.Influence of optic flow speed on stride frequency in (a) Trial 2 and (b) Trial 3. Error bars show 95% confidence intervals around the means.The RPI.—There was a significant linear effect for Optic Flow Speed, F(1, 31) = 5.73, p = .02 (see Figure 6a). The RPI in both groups decreased with the increment of optic flow speed, and the impacts of Optic Flow Speed on RPI were not significantly different between the older and the younger adults, F(1, 31) = 3.61, p = .07. As shown in Figure 6a, the mean RPIs were positive, indicating that the overall distribution of RPI was confined primarily to the 1:1 frequency ratio.Figure 6.Influence of optic flow speed on relative power index in (a) Trial 2 and (b) Trial 3. Error bars show 95% confidence intervals around the means.Drift.—No significant main or interaction effects were observed (Figure 7a).Figure 7.Influence of optic flow speed on drift in (a) Trial 2 and (b) Trial 3. Error bars show 95% confidence intervals around the means.Trial 3: Manipulation of Lateral Flow AsymmetryWalking speed and stride parameters.—The linear effects for Optic Flow Speed were significant for walking speed, F(1, 31) = 16.69, p = .0001, and stride frequency, F(1, 31) = 11.99, p = .0001, indicating that both groups decreased walking speed (see Figure 4b) and stride frequency (see Figure 5b) when the optic flow speed in unilateral visual field was increased. The Group effects were significant for walking speed, F(1, 31) = 16.96, p = .0001, and stride frequency, F(1, 31) = 14.78, p = .001, showing that older adults walked at a higher speed (0.89 ± 0.08 m/s) and stride frequency (1.11 ± 0.11 Hz) than the younger adults (0.81 ± 0.07 m/s; 0.99 ± 0.10 Hz, respectively). No significant interaction effects between Group and Optic Flow Speed were found for walking speed, F(1, 31) = 0.09, p = .76, and stride frequency, F(1, 31) = 0.35, p = .56. No significant main or interaction effects were found for stride length.The RPI.—A significant linear effect for Optic Flow Speed, F(1, 31) = 7.23, p = .01, was found. The RPI in both groups decreased with the increment of optic flow speed in left or right visual field, and the effects of Optic Flow Speed on RPI were not significantly different between the older and the younger adults, F(1, 31) = 0.99, p = .33. As shown in Figure 6b, the mean RPIs were positive, indicating that the overall distribution of RPI was confined primarily to the 1:1 coordination pattern.Drift.—A significant linear effect for Optic Flow Speed was observed, F(1, 31) = 26.37, p = .0001, indicating that both groups drifted away from the wall that was moving faster, and the degree of drift was positively related to the difference in optic flow speeds between the two walls (see Figure 7b). The effects of Optic Flow Speed manipulations were not significantly different between the older and the younger adults, F(1, 31) = 0.04, p = .85. No other significant main or interaction effects were observed.DISCUSSIONThe purpose of the study was to investigate whether older adults show different modulations in locomotion compared with younger adults in response to availability of vision, optic flow speed, and lateral flow asymmetry. The result in Trial 1 corroborates previous findings showing that, in terms of walking speed and stride length, older and younger adults were affected differentially by the availability of vision (Choy et al., 2003; Cromwell et al., 2001, 2002; Hay et al., 1996; Manchester et al., 1989). However, in Trials 2 and 3, the effects of flow speed and lateral flow asymmetry were not influenced by age. Our findings, complementing previous studies (Anderson et al., 1998; Konczak, 1994; Schubert et al., 2005), did not support a greater dependence on optic flow information in older adults during walking. Additionally, no age-related difference was found for the rod and frame test. Taken together, the findings suggest that the ability to integrate optic flow information into the multimodal system for assessment of walking speed and heading direction was comparable in older and younger adults. When vision was deprived, adapted walking performance in older adults emerged possibly due to changes in vestibular or somatosensory function or previous experiences, making them walk more cautiously. We are aware that some participants in the older group were younger than 65. Could the insignificant interaction effect between age and optic flow speed be attributed to the existence of subgroups in the older group? We, thus, divided the older group into two by a cutoff age of 65 years and ran the statistical analyses again. The findings with three groups were not different from those with two groups, indicating that this possibility can be ruled out.Confirming previous results (Prokop et al., 1997; Schubert et al., 2005; Varraine et al., 2002), changes in absolute value of optic flow speed were linearly and negatively correlated with variations in walking speed, stride frequency, and RPI. The finding can be interpreted as that optic flow speed presented in Trials 2 and 3 was compared with the flow speed perceived during training (i.e., −0.8 m/s). When the flow speed was faster/slower than −0.8 m/s, participants felt that they were walking more quickly/slowly than the trained speed, and they modulated locomotion inadvertently or intentionally to keep their walking speed as instructed. Interestingly, even when changes in flow speed occurred in the unilateral wall only, the effects of optic flow speed could still be observed. One possible explanation is that the perceived flow speed might be the average of the flow speeds from the two walls. More experiments are needed to understand how optic flow speed is perceived when flow speeds are different in two visual fields.With respect to the effects of lateral flow asymmetry, participants drifted away from the wall that was moving faster, and the degree of drift was positively related to the difference of optic flow speeds between the two walls. This agrees with data from previous studies (Duchon & Warren, 2002; Srinivasan et al., 1991), suggesting that when optic flow speeds are different in two visual fields, participants consider their trajectory as curved and would drift to a balance point at which the flow speed in both sides appears equal.In all, 30 of 33 participants retained a pattern of 1:1 arm-to-leg frequency ratio across all conditions and did not show any transitions with the manipulations of optic flow speed. This finding suggests that optic flow speed may not be an appropriate control parameter for changes in interlimb coordination during walking. However, it may also be the case that walking overground at 0.8 m/s reduced the likelihood of finding effects of optic flow speed manipulations on interlimb coordination. It appears that more stable patterns (i.e., 1:1 arm-to-leg frequency ratio) were observed during overground walking at 0.8 m/s than during treadmill walking at the same speed, and there is a possibility that the transition point during overground walking is at a lower walking speed compared with treadmill walking. Future work on interlimb coordination during overground walking with a systematic manipulation of walking speed will provide further insight into coordination dynamics.Older adults walked significantly faster than younger adults in Trials 2 and 3, but not in Trial 1. There are two possible explanations. First, the reported preferred walking speed was between 0.82 and 1.38 m/s for older adults and around 1.20 m/s for younger adults (Malatesta et al., 2004; McGibbon & Krebs, 2001; Patel et al., 2006). Because the trained walking speed might not be the preferred walking speed for all the participants, older adults who could not maintain the muscle strength needed for smooth walking at 0.8 m/s, or found the exposure to the VR environment challenging (Giphart et al., 2007), might have unintentionally walked at their preferred speed to reduce gait variability (i.e., to improve stability). Second, it has been reported that the perceived speed reduced after prolonged exposure to optic flow (Krekelberg, van Wezel, & Albright, 2006; Smith, 1985). It is possible that the perceived speed in older adults was lower than that in younger adults after adaptation, thereby inducing higher walking speeds in older adults. Further studies are needed to test these hypotheses.In the current study, we provide evidence that older adults are able to integrate optic flow information into the multimodal system to monitor their walking speed and heading direction in much the same manner as younger adults. Future research on the fall-prone elderly is needed to understand whether perception of optic flow is degraded in this population and how they use the optic flow information to guide locomotion.FUNDINGThis research was supported by the National Institute of Neurological Disease and Stroke (grant 1R21 NS043730-01 awarded to A.C.-G.) and the Dudley Allen Sargent Research Fund (awarded to Y.-H.C.).We thank Dr. Terry Ellis and Ms. Julie Starr for help with participant recruitment.AndersonGAMulderTNienhuisBHulstijnWAre older adults more dependent on visual information in regulating self-motion than younger adults?Journal of Motor Behavior199830104113AtchleyPAndersenGJThe effect of age, retinal eccentricity, and speed on the detection of optic flow componentsPsychology and Aging199813297308AzulayJPMesureSAmblardBPougetJIncreased visual dependence in Parkinson's diseasePerceptual and Motor Skills20029511061114BillinoJBremmerFGegenfurtnerKRDifferential aging of motion processing mechanisms: Evidence against general perceptual declineVision Research20084812541261ChoyNLBrauerSNitzJChanges in postural stability in women aged 20 to 80 yearsJournals of Gerontology: Series A, Biological Sciences and Medical Sciences200358525530CromwellRLNewtonRAForrestGHead stability in older adults during walking with and without visual inputJournal of Vestibular Research200111105114CromwellRLNewtonRAForrestGInfluence of vision on head stabilization strategies in older adults during walkingJournals of Gerontology: Series A, Biological Sciences and Medical Sciences200257M442M448DonkerSFBeekPJWagenaarRCMulderTCoordination between arm and leg movements during locomotionJournal of Motor Behavior20013386102DuchonAPWarrenWHJr.A visual equalization strategy for locomotor control: Of honeybees, robots, and humansPsychological Science200213272278FordMPWagenaarRCNewellKMThe effects of auditory rhythms and instruction on walking patterns in individuals post strokeGait and Posture200726150155GibsonJJVisually controlled locomotion and visual orientation in animalsBritish Journal of Psychology195849182194GilmoreGCWenkHENaylorLAStuveTAMotion perception and agingPsychology and Aging19927654660GiphartJEChouY.-H.KimDHBortnykCTWagenaarRCEffects of virtual reality immersion and walking speed on coordination of arm and leg movementsPresence: Teleoperators and Virtual Environments200716399413HayLBardCFleuryMTeasdaleNAvailability of visual and proprioceptive afferent messages and postural control in elderly adultsExperimental Brain Research1996108129139KonczakJEffects of optic flow on the kinematics of human gait: A comparison of young and older adultsJournal of Motor Behavior199426225236KrekelbergBvan WezelRJAlbrightTDAdaptation in macaque MT reduces perceived speed and improves speed discriminationJournal of Neurophysiology200695255270MalatestaDSimarDDauvilliersYCandauRBen SaadHPréfautCCaillaudCAerobic determinants of the decline in preferred walking speed in healthy, active 65- and 80-year-oldsPflugers Archiv: European Journal of Physiology2004447915921ManchesterDWoollacottMZederbauer-HyltonNMarinOVisual, vestibular and somatosensory contributions to balance control in the older adultJournal of Gerontology198944M118M127McGibbonCAKrebsDEAge-related changes in lower trunk coordination and energy transfer during gaitJournal of Neurophysiology20018519231931PatelITuranoKABromanATBandeen-RocheKMuñozBWestSKMeasures of visual function and percentage of preferred walking speed in older adults: The Salisbury Eye Evaluation ProjectInvestigative Ophthalmology and Visual Science2006476571ProkopTSchubertMBergerWVisual influence on human locomotion. Modulation to changes in optic flowExperimental Brain Research19971146370SchubertMProkopTBrockeFBergerWVisual kinesthesia and locomotion in Parkinson's diseaseMovement Disorders200520141150SmithATVelocity coding: Evidence from perceived velocity shiftsVision Research19852519691976SrinivasanMVLehrerMKirchnerWHZhangSWRange perception through apparent image speed in freely flying honeybeesVisual Neuroscience19916519535TeasdaleNStelmachGEBreunigAPostural sway characteristics of the elderly under normal and altered visual and support surface conditionsJournal of Gerontology199146B238B244van der KooijHJacobsRKoopmanBvan der HelmFAn adaptive model of sensory integration in a dynamic environment applied to human stance controlBiological Cybernetics200184103115VarraineEBonnardMPailhousJInteraction between different sensory cues in the control of human gaitExperimental Brain Research2002142374384WagenaarRCHoltKGKuboMHoCLGait risk factors for falls in older adults: A dynamic perspectiveGenerations2003262832WagenaarRCvan EmmerikREResonant frequencies of arms and legs identify different walking patternsJournal of Biomechanics200033853861WarrenWHJrBlackwellAWMorrisMWAge differences in perceiving the direction of self-motion from optical flowJournal of Gerontology198944P147P153WinogrodzkaAWagenaarRCBooijJWoltersECRigidity and bradykinesia reduce interlimb coordination in Parkinsonian gaitArchives of Physical Medicine and Rehabilitation200586183189Decision Editor: Rosemary Blieszner, PhD
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B98637223569DFC105B062206637AA93134062C.txt b/test/dataset/in/resources/corpus/Clean_0B98637223569DFC105B062206637AA93134062C.txt
deleted file mode 100644
index fe3a04d..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B98637223569DFC105B062206637AA93134062C.txt
+++ /dev/null
@@ -1 +0,0 @@
-EXG5174S0531-5565(98)00078-310.1016/S0531-5565(98)00078-3Elsevier Science Inc.Table 1Means (± standard errors) for longevity and female fecundity indices of the short-lived and long-lived bean weevils lines legendTraitShort-lived lineLong-lived linet-testnmean ± S.E.nmean ± S.E.LongevityFemales9511.58 ± 0.279719.19 ± 0.5412.59aMales9211.61 ± 0.259315.42 ± 0.437.67aFecundityEarly (1–2 days)10123.83 ± 1.1510015.60 ± 1.214.93aLate (≥9 days)1000.36 ± 0.201003.59 ± 0.754.14aTotal10049.57 ± 1.8210048.02 ± 2.341.19legendn = The number of analyzed individuals.ap < 0.001.Table 2Means (± standard errors) for activities (in units/mg proteins) of SOD and catalase in the young (one-day-old) and old (10-day-old) virgin and mated females and males of the short-lived and long-lived lineslegendEnzymeShort-lived lineLong-lived lineYoungOldYoungOldSODFemalesVirgin31.69 ± 3.4822.57 ± 1.3929.59 ± 1.5427.40 ± 2.67Mated26.86 ± 1.2427.59 ± 0.55MalesVirgin31.87 ± 1.6419.14 ± 1.1832.49 ± 2.4223.79 ± 1.65Mated19.52 ± 1.5227.99 ± 3.28Grand Mean31.78 ± 1.9122.02 ± 0.9631.04 ± 1.5126.69 ± 1.09CatalaseFemalesVirgin29.62 ± 0.4229.89 ± 2.7227.43 ± 0.8234.77 ± 3.38Mated35.33 ± 5.9343.91 ± 6.18MalesVirgin8.84 ± 0.298.37 ± 0.998.03 ± 0.507.36 ± 0.15Mated10.99 ± 0.4612.79 ± 2.33Grand Mean19.23 ± 3.6621.15 ± 3.1617.73 ± 3.4434.71 ± 3.96legendSample sizes for each mean are 5. Grand means are calculated on the pooled data of the corresponding groups.Table 3Two-way ANOVA for SOD and catalase activities in the young (one-day-old) females and males originated from the short-lived and long-lived linesSource of variationdfSOD M.S. (×10−3)FCAT M.S. (×10−3)FLine10.320.077.585.24aSex12.750.581409.48974.27bLine × sex11.140.230.130.09Error164.771.47ap < 0.05;bp < 0.001.Table 4Two-way ANOVA for SOD and catalase activities for virgin weevils of different ages (one day old and 10-days old) from short- and long-lived linesSource of variationdfSOD M.S. (×10−3)FCAT M.S. (×10−3)FFemalesLine13.960.631.150.27Age140.866.51a10.862.58Line × age113.332.1212.012.85Error166.284.21MalesLine112.583.279.922.35Age1159.4641.48b5.991.42Line × age19.762.540.000.00Error163.844.22ap < 0.05;bp < 0.001.Table 5Three-way ANOVA for SOD and catalase activities in the old (10-day-old) virgin and mated females and males from the short- and long-lived linesSource of variationdfSOD M.S. (×10−3)FCAT M.S. (×10−3)FLine (A)172.4615.59b13.590.79Sex (B)146.289.96b3070.31179.69bMating status (C)116.193.49126.507.40aA × B114.913.2115.950.93A × C10.030.007.480.43B × C10.140.0333.691.97A × B × C19.842.122.160.13Error324.6517.09ap < 0.05;bp < 0.001.Original PapersActivity of superoxide dismutase and catalase in the bean weevil (acanthoscelides obtectus) selected for postponed senescenceSeslijaDarkaŠešlijaaBlagojevicDuškoBlagojevićaSpasicMihajloSpasićaTucicNikolaTucića*tucic@opennet.orgaDepartment of Evolutionary Biology, Institute for Biological Research, 29 Novembra 142, 11000 Belgrade, Serbia, Yugoslavia*Corresponding authorAbstractRelationship of superoxide dismutase and catalase activities and aging were tested using bean weevil lines selected for postponed senescence. The beetles of different age (young and old) and mating status (virgin and mated) from the extended longevity lines were compared with their counterparts derived from the short-lived lines for activities of SOD and catalase. The old beetles from the long-lived lines had statistically significant higher activity of SOD than their controls. Although we did not find a significant effect of catalase on longevity, beetles originating from both types of lines exhibited an increased catalase activity during mating processes. In addition, we did observe an increased activity of catalase in one-day-old beetles of the short-lived lines relative to the same-aged individuals of the long-lived lines.KeywordsAcanthoscelides obtectusSuperoxide dismutaseCatalaseLongevitySelectionThe identification of underlying mechanisms of aging is regarded to be an elusive task, largely due to the complexity of aging processes. In addition, most of experimental methods approaching this problem have proven to be inadequate for delineation of the physiological mechanisms that determine rates and patterns of aging. For example, the usual methods of pursuing this goal, such as the studies of the temporal correlates of aging in defined cohorts, have not contributed much to the understanding of genetic and physiological mechanisms controlling aging (see review in Finch, 1990; Rose, 1991). The most important drawback to chronological measures of aging is that many age-related physiological changes are not the same as time-dependent changes (see, e.g., Baker, 1976). Also, experimental studies attempting to test causal mechanism of the aging processes by using a single gene mutant or different environmental factors that “accelerate the rate of aging” seem to be inappropriate because under these circumstances organisms could be dying because of novel pathologies rather than an acceleration of normal aging processes (Maynard Smith, 1966; Hutchinson & Rose, 1987). For these reasons, it has been argued (Johnson, 1987; Rose, 1991; Arking, 1995) that organisms undergoing genetically postponed aging represent the best experimental system that would allow study of the processes important in the normal aging situation.Such a model system for the analysis of underlying genetic and physiological mechanisms of aging constitutes lines with genetically postponed aging obtained from selection for survival to, and reproduction at, later ages. The important advantage to the use of these lines is that they have been obtained by the action of selection on allelic variation affecting aging at great many loci within outbred populations of Drosophila melanogaster (Luckinbill et al., 1984; Rose, 1984, Partridge & Flower, 1992) and Acanthoscelides obtectus (Tucić et al., 1996, 1997). In experiments with these holometabolous insect species in which selection on quantitative genetic variation proceeded by using either young or old adults as parents in each generation, it was found that the “old” lines had higher longevity (and delayed fertility) than the “young” ones. Besides, these experiments have successfully corroborated the basic assumption of the general evolutionary theory of aging (Medawar, 1952; Rose, 1991). As pointed out above, the obtained lines are an excellent experimental system for the study of the nonpathological processes of aging.Because short-lived and long-lived lines are derived by imposing selection on natural variants of many genes affecting aging, it is very likely a priori, that different species (e.g., Drosophila vs. Acanthoscelides) or lines of the same species originating from different laboratories have attained specific longevities by various underlying mechanisms. This expectation seems to be fulfilled with lines of D. melanogaster, with extented longevity stemmed out from different laboratories (see Arking, 1995). Accordingly, when dealing with a topic as complex as aging, it is essential to determine empirically mechanism(s) operative in the short-lived and long-lived lines of A. obtectus obtained in our laboratory.As a part of our effort to elucidate actual mechanisms affecting the aging processes in the bean weevils, in the present study we investigated the activities of two enzymes of great interest to gerontologists—superoxid dismutase (SOD), and catalase. These free radical scavenging enzymes are of interest because of SOD’s role in catalyzing the conversion of superoxide radicals to hydrogen peroxide (McCord & Fridovich, 1969), the latter then undergoing conversion to water due to the action of catalase (Deisseroth & Dounce, 1970). Because free radicals act damaging macromolecules within a cell, it has been proposed that they are important factors in aging (Harman, 1956), and that SOD and catalase could play the antiaging role. The most convincing experimental evidence that supports the free radical theory of aging was provided by Orr and Sohal (1994). In their experiments, D. melanogaster transgenic flies carrying three copies of SOD and catalase genes exhibited about a one-third extension of longevity, a lower rate of senescence, and a lower amount of protein oxidative damage compared to diploid controls. However, analyses of these enzymes in D. melanogaster lines with markedly different adult longevities led to contradictory results. For example, Bartosz et al. (1979) and Fleming et al. (1992) detected a positive correlation between longevity and antioxidant capacity, whereas Massie et al. (1975), Niedzwiecki et al. (1992), and Durusov et al. (1995) did not observe such a correlation. Moreover, a comparison of antioxidant defense in several mammalian species did not suggest a clear association between this defense and longevity (Sohal & Orr, 1992).The long-lived D. melanogaster lines selected by Luckinbill et al. (1984) and those by Rose (1984) differ substantially in their physiology, including the antioxidant defense system. The long-lived flies obtained in the former laboratory show a significant increase in the level and activity of both SOD and catalase (and some other enzymatic and nonenzymatic components of the antioxidant defense system) relative to the control flies, but their resistance to starvation and desiccation is low (Arking et al., 1993; Arking, 1995; Dudas & Arking, 1995). However, although more active allele of the SOD gene is associated with postponed aging lines of Rose (1984), it seems that this allele does not directly increase life span or later fecundity (Tyler et al., 1993). Here, however, the long-lived flies display enhanced resistance to starvation, desiccation, ethanol vapor, and heat (Service et al., 1985; Service, 1987; Graves et al., 1992). Thus, in D. melanogaster there is more than one route to extended longevity. The present work is an attempt to obtain further information on the activity of SOD and catalase in aging of Acanthoscelides obtectus, another insect species for which long-lived and short-lived lines produced by the manipulation with reproductive schedules exist.The fact that sexual activity reduces longevity complicates the interpretation of the long-term studies that use artificial manipulation of reproductive schedules. Although the hypothesis that selection for increased longevity acts on genes that regulate short-term mortality risk through the regulation of sexual behavior (Partridge, 1987) has been refuted (Service, 1989; Partridge & Flower, 1992; Tucić et al., 1996), it is possible that elevated sexual activity in the later ages of the long-lived lines is in a positive correlation with the efficient system of antioxidant defense. This could be expected because, in accordance with “rate of living theory” (Pearl, 1928), a higher sexual activity of females and males means that they are more metabolically active, which implies that these individuals produce a higher level of free radicals. To test this hypothesis, activities of the SOD and catalase were analyzed in the virgin and mated females and males originating from the short- and long-lived lines of the bean weevil.1Materials and methods1.1The short- and long-lived bean weevil linesThe lines used in the present study were derived from a population of bean weevil (Acanthoscelides obtectus) maintained in the laboratory from 1986 (Tucić et al., 1996). This population (the “base”) was synthesised by mass mating equal numbers of adults from three local populations of A. obtectus captured in the vicinity of Belgrade (Yug.). The base population was maintained at large size (about 5,000 individuals each generation) on Phaseolus vulgaris, c.v. “gradistanac” seeds for about a 40-day interval. All cultures were maintained in a dark incubator at 30°C and about 70% humidity.The lines used in this study were obtained from later generations of lines on which selection had been imposed for either early (“young” lines, Y) or late-life (“old” lines, O) reproductive success (Tucić et al., 1996, 1997). The four replicate lines per each selection regime were established. The following summarizes the selection procedures used to maintain the Y-selected and O-selected lines (details are given in Tucić et al., 1996).Beetles in the Y-selected lines were allowed to lay eggs at the ages of one to two days after emergence. In each generation about 300–500 individuals were used as the next generation breeders. This treatment should have given rise to beetles with enhanced fitness during early adult life. In the present study we have used “young” lines selected for the 69 generations.In the O-selected lines (55 generations), the beetles were reproduced from 10th day after emergence until death. Prior to that period experimental adults were kept in vials (females and males together) without bean seeds (in the absence of seeds the egg production is low).All the assays described below were done using four-way crosses within each selection regime. These crosses were obtained by crossing the F1s of different pairs of replicate lines within each selection regime, that is, (Y1 × Y2) × (Y3 × Y4), and (O1 × O2) × (O3 × O4), where the subscript numbers refer to the specific replicate lines (thereafter denoted simply as Y and O lines). The outcrossing of replicate lines should have removed any effect of inbreeding depression, and it diminished any epistatic interactions among genes originating during the long-term selection due to Wahlund’s effect or mutation pressures. Also, protocols used to construct four-way crosses (see Tucić et al., 1997 for details) ensured that selected lines passed through two generations of common conditions.1.2Enzyme assaysEnzyme assays were carried out for females and males of different age (1 and 10 days old) and mating status (virgin and mated) from each experimental group. Twenty individuals were homogenized in 0.25 M sucrose, 50 mM Tris, and 0.1 mM EDTA solution, pH 7.4, using a Janke-Kunkel Ka-Werk Ultra-Turrax homogenizer at 0–4°C Homogenates were sonicated according to Takada et al. (1982). After centrifugation (90 min, 10,500g, 4°C), protein content was determined in the supernates (Lowry et al., 1951) used also in enzyme activity assays. Measurements of the five different samples of each enzyme were made for each sex/age/mating status group.1.3Superoxide dismutaseSuperoxide dismutase (SOD, EC 1.15.1.1) activity was measured as described by Misra and Fridovich (1972). The rate of inhibition of epinephrine autoxidation at alkaline pH in the presence of SOD was spectrophotometrically estimated. One unit of SOD activity was defined as the amount of enzyme inhibiting oxidation of epinephrine by 50% under appropriate reaction conditions. The SOD specific activity was expressed as units per mg protein.1.4CatalaseCatalase (EC 1.11.1.6) activity was determined after Beutler (1982). Decomposition of H2O2 was monitored spectrophotometrically, and a unit of catalase activity was defined as 1 μM H2O2 decomposed per minute, assuming a molar extinction coefficient of 62.4 at 230 nm.2ResultsData shown in Table 1 are the mean longevities and three important fecundity indices of females that characterized lines selected for early and late reproduction for 69 and 55 generations, respectively. As may be seen, there are significant direct responses to both selection regimes; females from the Y lines exhibited significantly higher early fecundity than those from the O lines, whereas for the late fecundity, the opposite trend was observed. In addition, we have corroborated our earlier findings (Tucić et al., 1996, 1997) that selection for early and late reproductive effort does not change the total number of eggs laid by females during the whole lifespan. It is important to note that here, as in the earlier generations of selection, the effect of imposed age-specific selection on longevity, cohorts sampled from the O lines have significantly greater mean longevities relative to Y lines.Records of SOD and catalase activities in 1- and 10-day-old virgin and mated females and males originating from the short-lived and long-lived lines are listed in Table 2. To test differences in the activities of the analyzed enzymes among these groups we performed several ANOVAs (data are log transformed).A two-way ANOVA, with lines and sex of one-day-old beetles as the factors (Table 3), revealed significant differences in the catalase activities (but at the 0.05 level), both between the lines (at the 0.05 level; with the activity of the catalase significantly higher in the short-lived than in the long-lived lines) and sex (in both lines catalase activities in females exceeded more than three times those in males; Table 2).The differences observed above in catalase activities between the lines disappeared when virgins of both 1- and 10-day-old beetles were compared (Table 4). Because the data in Table 1 indicate the opposite pattern of age-dependent catalase activity between the sexes, a two-way ANOVA (with line as one factor and age of beetles as the other factor) were done separately for each sex (Table 4). Increased catalase activity in old females relative to the one-day-old females and its opposite trend in males was, however, statistically insignificant. Although the “line effect” was also absent with regard to SOD activity, virgin 10-day-old females and males had significantly lower activity of this enzyme than their one-day-old counterparts (Table 4).The results of three-way ANOVAs on the SOD and catalase activities over line, sex, and mating status of 10-day-old beetles as factors are listed in Table 5. Here we observed interesting differences between the SOD and catalase activities with respect to line and mating status effects. The old beetles sampled from the long-lived lines exhibited significantly higher SOD activity (at the 0.01 level) than the same aged beetles from the short-lived lines. At the same time, significant “line effect” was absent for catalase activity. Although both the SOD and catalase activities were somewhat higher in females and males kept together their whole life (Table 1), a significant difference between virgin and mated beetles has been observed only for catalase (but at the 0.05 level; Table 5). Also here, in contrast to the data for one-day-old beetles, both enzyme activities were significantly lower in males than in females.3DiscussionThe results presented in this study suggest that SOD might play a role in the postponing, and thus controlling, aging in Acanthoscelides obtectus. Although there are no available data on allelic differences at the SOD structural gene between the lines, pattern of differentiation in the SOD activities between short- and long-lived lines (i.e., statistically detectable difference between 10-day-old beetles and the lack of differentiation between young individuals sampled from different lines; Tables 3 and 5) could be attributed to differences in the regulation of structural genes common to both lines. In other words, in both lines we observed a significant age-specific decline in SOD activity (the “age effect” in Table 4), but this decrease was less pronounced in the long-lived lines. In this respect, our lines seem quite different from those obtained by Tyler et al. (1993), who found that more active SOD alleles were associated with postponed aging of Drosophila melanogaster laboratory lines. However, our interpretation is in accordance with the findings of Dudas and Arking (1995) and Burde and Arking (1998). Their data, obtained by gel electrophoresis, did not provide any evidence on the allelic differences at the SOD gene between the short- and long lived Drosophila lines, which, however, exhibited differences at the level of SOD activity.Orr and Sohal (1994) showed that transgenic Drosophila flies carrying extra copies of SOD and catalase exhibited lifespan extension as a consequence of a slower rate of aging (evidenced by the Gompertz parametric analysis), whereas selected long-lived Drosophila lines seemed to live long as a consequence of a decreased initial mortality rate with no change in the aging rate (Arking, 1995; Dudas & Arking, 1995). Our data, based on the Gompertz parametric analyses (Tucić et al., 1996), suggest that the “young” and “old” selection regimes in the bean weevil produce a change both in the timing of mortality (or in the initial mortality rate) and in rate of change of age-specific mortality. Which of these two important aging events in the bean weevil is possibly connected with the observed level of SOD activity remains to be shown.If, as our data suggest, the long-lived lines exhibit elevated level of the SOD activities, then a concordant increase could be expected in the activity of the catalase, which, acting together with SOD provides the primary enzymatic antioxidant defense. Elevated levels of SOD in the absence of a compensatory upregulation of catalase are thought to lead to H2O2 accumulation, which may give rise to cytotoxic effects (Fleming et al., 1992, and references therein). However, we did not detect any deterioration in the life history traits of the long-lived beetles (on the contrary; egg-to-adult viability, e.g., of the long-lived weevils, was higher than that of the short-lived individuals; Tucić et al., 1996), despite the fact that we failed to demonstrate the increase of catalase activity in these lines. However, this failure arose from the lack of a statistically detectable (in our sample sizes) increase of catalase activity in the long-lived lines but not from the entire absence of any increment in the activity of this free radical scavening enzyme. Careful inspection of the data in Table 2 indicates that grand mean (over both mating status and sexes) of 10-day-old beetles for the catalase is about 14% higher in the long-lived than in the short-lived lines. At the same time, the increase in SOD activity of the same age group in the long-lived relative to the short-lived lines was, on average, some 18%. There are two conclusions that could be drawn from these data. First, the observed increase in the SOD activity appears to provide additional protection against oxidative stress and results in an increased lifespan. Fleming et al. (1992) estimated the increase of SOD expression to be up to 35% (without any change of catalase activity), which could result in an increased longevity, and that levels of SOD above those appear to be toxic for Drosophila. Unfortunately, in the absence of transgenic individuals such an estimation is not possible for the bean weevil. Second, and more important from the standpoint of the evolutionary biologists, it seems that imposed selection regimes generate genotypes in which functionally coupled activities of the free radical scavening enzymes are elevated in a balanced way.Our data also raise the possibility that changes in catalase activity may be an important mechanism operating in the mating activities of the bean weevil. Although both the SOD and catalase activities increased in the mated females and males relative to the virgin beetles (Table 2), only in the latter was a statistically detectable increment of activity recorded (Table 5). It seems, therefore, that different aspects of reproductive activity in the bean weevils, such as courtships and other mating activities of females and males, oogenesis, and spermatogenesis, are protected by antioxidant enzymes. This is not unexpected, because neuropeptides involved in the expression of ecdysone during ovarian development (Lagaueux et al., 1977; Goltzene et al., 1978; Zhu et al., 1983) and spermatogenesis (Koolman et al., 1979) are important regulators of the mating activities, fertilization, and production of gametes in insects. Physiological effects of the ecdysone are realized through 20-hydroxyecdysone, which is regulated by the activity of the cytochrome P-450–linked enzymes which in turn, generate free radicals (Hoffman & Hetru, 1983).However, how could one explain increased level of catalase activity in the one-day-old beetles in the short-lived lines relative to the same aged individuals of the long-lived lines? The reason for such a result could be found in our experimental procedures involved in production of the short- and long-lived lines. Namely, our “young” lines (which bring about to the short-lived beetles) are leaving their progenies only during the first two days after emergence. They are, therefore, under strong selection for shortening the period between emergence and peak of reproductive activity in this insect species (about three to four days after emergence; Tucić et al., 1996). As a by-product of such a selection regime, we have genotypes characterized by elevated catalase activity coincident with the period of the “permitted” reproduction. Presently, we do not know what these genotypes are. They could be more active catalase alleles expressed in the early adult life, or the alleles of the other genes involved in the regulation of catalase activity.ReferencesArking 1995R.ArkingAntioxidant genes and other mechanisms involved in the extended longevity of DrosophilaR.G.CutlerL.PackerJ.BertramA.MoriOxidative Stress and Aging1995Birkhauser VerlagBasel123138Arking et al 1993R.ArkingS.P.DudasG.T.BakerGenetic and environmental factors regulating the expression of an extended longevity phenotype in a long lived strain of DrosophilaGenetica911993127142Baker 1976G.T.BakerInsect flight muscleMaturation and senescenceGerontology221976334361Bartosz et al 1979G.BartoszW.LeykoR.FriedSuperoxidase dismutase and life-span of Drosophila melanogasterExperientia35197911931194Beutler 1982E.BeutlerCatalaseE.BeutlerRed Cell Metabolism, A Manual of Biochemical Methods1982Grune and Straton, IncNew York105106Burde and Arking 1998H.R.BurdeR.ArkingImmunological confirmation of elevated levels of CuZn superoxide dismutase protein in an artificially selected long-lived strain of Drosophila melanogasterExp Geront331998227237Deisseroth and Dounce 1970A.DeisserothA.L.DounceCatalasePhysical and chemical properties, mechanism and catalysis and physiological rolePhysiol Rev501970319375Dudas and Arking 1995S.P.DudasR.ArkingA coordinate up-regulation of the antioxidant gene activities is associated with the delayed onset of senescence in a long-lived strain of DrosophilaJ Gerontol Biol Sci501995B117B127Durusov et al 1995M.DurusovN.DirihN.BozcukAge-related activity of catalase in different genotypes of Drosophila melanogasterExp Gerontol3019957786Finch 1990C.E.FinchLongevity, Senescence, and the Genome1990University of Chicago PressChicagoFleming et al 1992J.E.FlemingI.ReveillaudA.NiedzwieckiRole of oxidative stress in Drosophila agingMutat Res2751992267279Goltzene et al 1978F.GoltzeneM.LaugueuxM.CharletJ.A.HoffmannThe follicle cell epithelium of maturing ovaries of Locusta migratoriaA new biosynthetic tissue for ecdysoneHoppe-Seyler’s Z Physiol Chem359197814271434Graves et al 1992J.L.GravesE.C.ToolsonC.JeongL.N.VuM.R.RoseDesiccation, flight, glycogen, and postponed senescence in Drosophila melanogasterPhysiol Zool651992268286Harman 1956D.HarmanAging—A theory based on free radical and radiation chemistryJ Gerontol111956298300Hoffmann and Hetru 1983J.A.HoffmannC.HetruEcdysoneR.G.H.DownerH.LauferEndocrinology of Insects1983Alan R. Liss, IncNew York6588Hutchinson and Rose 1987E.W.HutchinsonM.R.RoseGenetics of aging in insectsRev Biol Res Aging319876270Johnson 1987T.E.JohnsonAging can be genetically dissected into component processes using long-lived lines of Caenorhabditis elegansProc Natl Acad Sci USA84198737773781Koolman et al 1979J.KoolmanK.SchellerB.BodensteinEcdysteroids in the adult male blowfly Calliphora vicinaExperientia351979134135Lagueux et al 1997M.LagueuxM.HirnJ.A.HoffmanEcdysone during ovarian development in Locusta mugratiraInsect Biochem231997109120Lowry et al 1951D.H.LowryN.J.RosebroughA.L.FarrR.J.RandalProtein measurements with folin-phenol reagentJ Biol Chem1931951265275Luckinbill et al 1984L.S.LuckinbillR.ArkingM.J.ClareW.C.CiroccoS.A.BuckSelection for delayed senescence in Drosophila melanogasterEvolution3819849961003Massie et al 1975H.R.MassieM.B.BairdM.M.McMahonLoss of mitochondrial DNA with agingGerontology211975231237Maynard Smith 1966J.Maynard SmithTheories of agingP.L.KrohnTopics in the Biology of Aging1966InterscienceNew York135McCord and Fridovich 1969J.McCordI.FridovichSuperoxidase dismutase. An enzymic function for erythrocupein (hemocupein)J Biol Chem244196960496055Medawar 1952P.B.MedawarAn Unsolved Problem of Biology1952H. K. LewisLondonMisra and Fridovich 1972H.P.MisraI.FridovichThe role of superoxide anion in the autooxidation of epinephrine and a simple assay for superoxide dismutaseJ Biol Chem247197231703175Niedzwiecki et al 1992A.NiedzwieckiL.ReveillaudJ.F.FlemingChanges in superoxyde dismutase and catalase in aging heat shocked DrosophilaFree Radic Res Commun171992355367Orr and Sohal 1994W.C.OrrR.S.SohalExtension of life-span by overexpression of superoxide dismutase and catalase in Drosophila melanogasterScience263199411281130Partridge 1987L.PartridgeIs accelerated senescence a cost of reproduction?Funct Ecol11987317320Partridge and Flower 1992L.K.PartridgeK.FlowerDirect and correlated response to selection on age at reproduction Drosophila melanogasterEvolution4619927691Pearl 1928R.PearlThe Rate of Living1928KnopfNew YorkRose 1991M.R.RoseEvolutionary Biology of Aging1991Oxford University PressNew YorkRose 1984M.R.RoseLaboratory evolution of postponed senescence in Drosophila melanogasterEvolution38198410041010Service 1987P.M.ServicePhysiological mechanisms of increased stress resistance in Drosophila melanogaster selected for postponed senescencePhysiol Zool601987321326Service 1989P.M.ServiceThe effect of mating status on lifespan, egg laying and starvation resistance in Drosophila melanogaster in relation to selection on longevityJ Insect Physiol351989447452Service et al 1985P.M.ServiceE.W.HutchinsonM.D.MackinleyM.R.RoseResistance to environmental stress in Drosophila melanogaster selected for postponed senescencePhysiol Zool581985380389Sohal and Orr 1992R.S.SohalW.C.OrrRelationship between antioxidants, prooxidants, and the aging processAnn NY Acad Sci66319927484Takada et al 1982Y.TakadaT.NogushiM.KayiyamaSuperoxide dismutase in various tissues from rabbits bearing the Vx-2 carcinoma in the maxillary sinusCancer Res42198242334235Tucic et al 1996N.TucićI.GliksmanD.ŠešlijaD.MilanovićS.MikuljanacO.StojkovićLaboratory evolution of longevity in the bean weevil (Acanthoscelides obtectus)J Evol Biol91996485503Tucic et al 1997N.TucićO.StojkovićI.GliksmanD.MilanovićD.ŠešlijaLaboratory evolution of life history traits in the bean weevil (Acanthoscelides obtectus)The effects of density-dependent and age-specific selectionEvolution51199718961909Tyler et al 1993R.H.TylerH.BrarM.SinghA.LatorreJ.L.GravesL.D.MuellerM.R.RoseF.J.AyalaThe effects of superoxidase dismutase alleles on aging in DrosophilaGenetica911993143150Zhu et al 1983X.X.ZhuH.GfellerB.LanzreinEcdysteroids during oogenesis in the ovoviviparous cockroach Nauphoeta cinereaJ Insect Physiol291983225236
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B9B5C3FB69DC6DD7CFB2B73AC782DA4F5CF81EC.txt b/test/dataset/in/resources/corpus/Clean_0B9B5C3FB69DC6DD7CFB2B73AC782DA4F5CF81EC.txt
deleted file mode 100644
index 9d465fc..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B9B5C3FB69DC6DD7CFB2B73AC782DA4F5CF81EC.txt
+++ /dev/null
@@ -1 +0,0 @@
- geronb J Gerontol B Psychol Sci Soc Scigeronb The Journals of Gerontology Series B: Psychological Sciences and Social Sciences J Gerontol B Psychol Sci Soc Sci 1079-5014 1758-5368 Oxford University Press 0002010.1093/geronb/56.5.P257 Journal of Gerontology: Psychological Sciences Attachment Bonds Between Adult Daughters and Their Older Mothers Associations With Contemporary Caregiving Carpenter Brian D. a aCase Western Reserve University, Cleveland, Ohio 1 9 2001 56 5 P257 P266 28 3 2001 6 3 2000 The Gerontological Society of America 2001 This study examined associations between attachment bonds and the care that daughters were providing to their community-dwelling mothers. Adult daughters (40 African American, 40 European American) completed assessments of adult attachment, instrumental and emotional caregiving, and caregiver burden. The author performed hierarchical linear regressions to examine the relationships between attachment dimensions (Security and Anxiety) and the provision of instrumental and emotional care. Both attachment dimensions were unrelated to the provision of instrumental care. In contrast, high scores on the Security dimension and low scores on the Anxiety dimension were associated with the provision of more emotional care to mothers. A final analysis revealed that high scores on the Security dimension were associated with less caregiver burden. These results suggest that practical care that daughters provide to their mothers may be independent of attachment patterns within the child–parent relationship, whereas affective, discretionary care may be promoted or hindered by attachment patterns. Moreover, the stress of caregiving may be mediated by a more secure attachment bond. The potential impact of adult attachment patterns has implications for both family intervention and public policy, given that more families are expected to participate in caregiving in coming decades. hwp-legacy-fpage P257 hwp-legacy-dochead RESEARCH ARTICLE Decision Editor: Margie E. Lachman, PhD Anumber of factors influence the degree to which adult children assist their older parents, such as obligation, affection, reciprocity, and other cultural norms (Blieszner and Hamon 1992). More concrete factors are important as well and include geographical distance, size of the family, gender of caregiver and care recipient, and socioeconomic resources and constraints (Dwyer and Coward 1991; Litvin, Albert, Brody, and Hoffman 1995; Wolf, Freedman, and Soldo 1997). In the current study I examined an additional factor, adult attachment patterns, to determine whether attachment-based emotional bonds between daughters and their older mothers were associated with (a) the amount and nature of care that daughters provide and (b) caregiving burden experienced by daughters. Basic Concepts in Attachment Theory Originating with the work of John Bowlby 1982, attachment theory describes a socioemotional behavioral system that guides how individuals manage their need for emotional security. This system is first evident early in life as children interact with their primary caregiver. When they are physically or psychologically threatened, children turn to their caregiver for comfort, and ideally their caregiver responds with immediate, positive, and consistent support. In reality, of course, caregivers do not always respond in ways that children expect. On the basis of their accumulated experiences with caregivers, children develop mental representations, or internal working models (Bowlby 1988), that reflect their beliefs about the responsiveness of caregivers and the environment more generally. Seminal work by Mary Ainsworth (Ainsworth, Blehar, Waters, and Wall 1978) identified behavioral manifestations of internal working models in the form of attachment styles, secure versus insecure attachment being the most broad differentiation. Children with a secure attachment were likely to seek and savor contact with their caregiver, to use that person as a secure base for exploration. Meanwhile, children with an insecure attachment were likely to avoid their caregiver or demonstrate anxiety in contact with him or her. A broad array of research has suggested that a child's initial attachment bond has an impact well beyond their first critical relationship and influences not only subsequent relationships but also a wide range of social and emotional outcomes later in life (Feeney and Noller 1996; Rothbard and Shaver 1994). Despite a burgeoning empirical literature on attachment dynamics, significant conceptual issues remain. For example, it is unclear whether attachment patterns represent an aspect of individuals (i.e., attachment as a trait), a facet of specific relationships (e.g., types of attachment differing within an individual's social network), or some combination of the two (Bartholomew and Shaver 1998). Likewise, there is debate about whether attachments are categorical phenomena or are best thought of as graded entities (Feeney, Noller, and Hanrahan 1994). Diversity of theory is also apparent in the variety of attachment taxonomies that exist, although a parsimonious nomenclature for attachment patterns is beginning to emerge (see Feeney et al. 1994; Griffin and Bartholomew 1994). Another issue is whether a construct originally validated with research on children is relevant to individuals at other ages. Although Bowlby first focused on the attachment dynamic between infants and their caregivers, he asserted that "attachment behaviour is held to characterize human beings from the cradle to the grave" (Bowlby 1979, p. 129), a life-span perspective shared by Ainsworth 1989. And indeed, research on attachment in childhood has been complemented by expanding attention to attachment in adults. Attachment in Adulthood Attachment patterns are hypothesized to persist across the life span through the reinforcing properties of internal working models (Bowlby 1973; Main, Kaplan, and Cassidy 1985). The first attachment relationship provides a template, a self-perpetuating schema that influences subsequent relationships. Mental representations from early attachment bonds thereby influence how individuals seek, anticipate, and interpret future interpersonal interactions (West and Sheldon-Keller 1994). Reflecting its roots, attachment theory also has emerged as a framework for understanding the relationship between adult children and their parents. Some theorists have suggested that adult children relinquish their parents as attachment figures (Weiss 1982), whereas others have amassed secondary evidence that attachments to parents are sustained (Krause and Haverkamp 1996). In a study of adult children whose parents were institutionalized, Crispi, Schiaffino, and Berman 1997 found that children's attachment style predicted aspects of their well-being. Specifically, a secure attachment style was associated with less caregiving difficulty and less psychiatric symptomatology. Another study of children whose parents were in nursing homes found that children's attachment was related to parent mood (Pruchno, Peters, Kleban, and Burant 1994). Attachments were less intense when parents were depressed and no longer able to provide the emotional support that children expected from the relationship. It may be that attachment dynamics, forged in childhood, continue to influence child–parent relationships later in the life span. But even if attachment patterns in adult children are discontinuous from their earliest manifestation, contemporary attachment patterns may still be relevant to the way in which adult children interact with their parents via children's capacity for self-reflectiveness, empathy, and their own needs for security (Crose 1994). Cicirelli 1991 has suggested that adult children provide care to their parents to forestall the dissolution of their attachment relationship. As parents age and weaken, their impermanence becomes more apparent to children. Anxious about the threatened loss of their attachment figure, children may provide support to bolster their parent and preserve the important attachment object. Cicirelli 1993 study of caregiving daughters revealed that attachment had a direct and positive relationship with the care that daughters were providing to their mothers: Stronger attachment bonds were associated with greater amounts of care, independent of mothers' level of functional dependency. Stronger attachment bonds also were associated with lower caregiver burden. One limitation of Cicirelli 1993 work is that it employed a global index of attachment rather than an assessment of specific attachment styles or dimensions. Yet just as attachment styles are associated with different behavior patterns and outcomes in other realms (e.g., Crispi et al. 1997; Hazan and Shaver 1990), they also may be associated with differences in parent care. Securely attached adult children, whose parents have been responsive and supportive, may be highly motivated to care for their parents to protect the valued attachment figure. They may continue to experience felt security (Sroufe and Waters 1977) in the attachment relationships with their parents, and they may desire to preserve that secure base as long as possible. In contrast, insecurely attached children, whose parents have been unresponsive and unsupportive, may be less eager to care for their parents because the psychological rewards for sustaining that relationship are unreliable. Their anxiety, distrust, or unease may translate into a lack of contact compared with that of peers with more secure attachments. Thus, one question I pursued in the current study was whether specific attachment patterns are related to the amount of care children provide to older parents. Another question was whether attachment patterns also are associated with the nature of parent care provided. Securely attached children may be comfortable providing a range of both emotional and practical care. Their stable, internal working models enable them to get close to their parents and offer a balance of pragmatic support as well as emotional availability. Also, because of their security and emotional flexibility they are likely to perceive caregiving as less burdensome. In contrast, insecurely attached children may be less willing to provide emotional support because of the psychological risks, although they may still provide practical assistance because of its relative emotional safety. Moreover, any care that insecurely attached children do provide is likely to be stressful because of underlying uneasiness in the dyad. In summary, my purpose in this study was to examine the relationship between attachment dimensions and the nature of care that adult daughters were providing to their older mothers. In this initial study I chose to focus on this one type of family dyad given its commonality in caregiving, but future research will need to include the full range of family relationships. I explored whether attachment dimensions contribute to the explanation of instrumental and emotional caregiving above and beyond the influence of demographic and contextual factors that may be related to parent care. The following hypotheses were tested:Securely and insecurely attached daughters provide comparable amounts of instrumental care to their older mothers.Securely attached daughters provide more emotional care to their mothers than insecurely attached daughters.Securely attached daughters experience lower caregiver burden than insecurely attached daughters. Methods Participants and Procedure Eighty adult daughters completed one semistructured interview. Participants were recruited from the greater Cleveland community via announcements in employee newsletters (29% of the sample) and local newspapers (19%), from agencies that provide social services (21%), and from recommendations made by other participants (31%). Inclusion criteria for the study were that daughters provided at least 5 hr of care each week to their mother and had been doing so for at least 2 months. In addition, mothers were required to be aged 60 and older and living alone or with their daughters. The sample size reflected a desired power of .80, an alpha level set at .05, and a medium effect size of .61 based on Cicirelli 1993 findings. Interviews were conducted by an advanced clinical psychology graduate student, lasted approximately 1–2 hr, and took place without the mother being present. After completing the interview daughters were provided with a description of the study, debriefed, and paid $10. Referrals were made to social services agencies (e.g., the Alzheimer's Association, caregiver support groups) when appropriate. The current sample expanded on Cicirelli 1993 European American sample and included 40 African Americans and 40 European Americans. Statistical comparisons between ethnic groups, however, revealed no significant differences on study variables. Daughters ranged in age from 28 to 77 (M = 50.19, SD = 10.77). Many were married (41%), a large majority had children (75%), and of them 63% had children currently living at home. As a group daughters were well educated, with 95% having completed at least high school and 69% having at least a college degree. Two thirds of the daughters were working, with most in sales, technical, or administrative support positions. Fifty-one percent of the daughters were residing with their mother. Daughters had been providing care for an average of 4.65 years (SD = 4.09). Mothers ranged in age from 60 to 99 (M = 78.60, SD = 9.30) and represented a broad range of functional dependence (activities of daily living [ADL] score M = 17.16, SD = 8.43, range = 0–28). Measures Care provided to mother. The assessment of caregiving focused on two content areas, instrumental and emotional care. An inventory of 31 instrumental activities and 22 emotional activities was presented to daughters. Examples of instrumental activities included preparing meals, performing housework, assisting with grooming, and obtaining medical equipment and supplies. Examples of emotional activities included providing gifts, hugging and kissing, voicing one's love, and offering comfort and sympathy. Daughters rated how often they had engaged in each activity with their mother in the past month on a 5-point scale: 1 ("not at all"), 2 ("once or twice"), 3 ("about once a week"), 4 ("several times a week"), and 5 ("about every day"). I summed ratings to provide two caregiving scores for each daughter: total frequency of instrumental care (α = .87) and total frequency of emotional care (α = .86), with higher scores representing more frequent caregiving. Adult attachment. I used two instruments to assess adult attachment. The Adult Attachment Scale (AAS; Cicirelli 1995) contains 16 items about the relationship between daughter and mother. Items reflect the theoretical underpinnings of attachment behavior, such as distress upon separation, joy upon reunion, and felt security provided by the relationship. Daughters rate items on a 7-point Likert-type scale ranging from 1 ("not at all true, disagree completely") to 7 ("definitely true, agree completely"). Items are summed to yield a total score that can range from 16 to 112, with higher scores indicating a stronger attachment to mother according to Cicirelli. Internal consistency reliability of the AAS has been reported as .95, with a 1-year test-retest reliability of .73 (Cicirelli 1995). In the current sample, the coefficient alpha was .95. The Relationship Questionnaire (RQ; Bartholomew and Horowitz 1991) is based on Bartholomew's theory of adult attachment (Bartholomew 1990; Griffin and Bartholomew 1994). Drawing on Bowlby 1988 conception of internal working models, Bartholomew has suggested there are two dimensions, or models, at work in an individual's attachment style. The other model represents the degree to which other people are expected to be available when needed and the disposition to seek or avoid close relationships. The self model embodies the degree to which individuals possess an internalized sense of self-worth and experience anxiety in relationships. The other and self models interact to yield four attachment styles: Secure, Preoccupied, Fearful, and Dismissive. To complete the RQ, participants rate the extent to which descriptions of the four styles correspond to the relationship they have with a particular individual. In this study the text was modified so that descriptions pertained to mothers. Ratings are made on a 7-point Likert-type scale, ranging from 1 ("not at all like me") to 7 ("very much like me"; see Appendix, Note 1). Scharfe and Bartholomew 1994 reported moderate stability of self-reported attachment patterns. In the current study internal consistency reliability among the four attachment style ratings was .81. Mother's level of functional dependency. Functional dependence was measured with a 14-item scale from the Duke Older Americans Resources and Services Multidimensional Functional Assessment Questionnaire (OARS; Fillenbaum 1988). Each daughter was asked to rate her mother's abilities on physical and instrumental ADLs using a 3-point scale: 2 ("able to perform the activity without help"), 1 ("able to perform the activity with some help"), or 0 ("completely unable to perform the activity"). Total ADL score was the sum of the 14 items, with higher scores representing greater functional independence. Internal consistency reliability for the scale was .95. Caregiver burden. The seven-item scale used by Cicirelli 1993 to measure caregiver burden was used in this study. Daughters rate the extent to which caring for mother interferes with their social life, relationship with partner, physical health, and mental health. Ratings are made on a 5-point Likert-type scale ranging from 1 ("not at all") to 5 ("very much"). Ratings are summed, and higher scores indicate greater burden. Cicirelli reported an internal consistency for the scale of .83; in the current study coefficient alpha was .80. Daughters also were asked how many months they had been providing the level of care they were currently providing. Duration of caregiving was included because some studies have documented a relationship between length of caregiving and negative caregiver outcomes (Hannappel, Calsyn, and Allen 1993). Demographic and family structure. Additional information that was collected included daughter's age, ethnicity, marital status, number of years of education, current or most recent occupation, number of children currently living with daughter, mother's age, and whether mother was currently living with daughter. A rating of socioeconomic status (SES) was based on Duncan 1961 socioeconomic index (SEI), which was updated by G. Stevens and Cho 1985 to reflect the 1980 census. I calculated an overall SES score for each daughter by summing standard scores for SEI and education, with higher scores representing higher SES. Data Analysis I calculated descriptive statistics to examine the characteristics of the sample. Preliminary exploration of the relationships among demographic variables, functional dependence, attachment variables, caregiving, and burden were performed with bivariate Pearson product-moment correlations and analyses of variance (ANOVAs). I used post-hoc comparisons to examine mean differences in caregiving across attachment styles, with the Scheffe method used to control for Type I errors. Multivariate analyses included a series of hierarchical least squares linear regressions to examine the variance in caregiving accounted for by attachment. I performed two hierarchical regressions using the same demographic, contextual, and attachment independent variables. Instrumental caregiving was the dependent variable in the first regression; emotional caregiving was the dependent variable in the second. Independent variables were entered in four blocks. Demographic and contextual variables were entered in the first step, on the rationale that they are the least modifiable. I entered mother's ADL score in the next step to account for caregiving driven by actual functional need. In the next step the amount of caregiving of the type complementary to the dependent variable was entered (emotional care when instrumental care was the dependent variable, and vice versa) in recognition of the probable overlap between types of care. In the last step I entered two attachment dimensions (discussed later) to examine their impact on caregiving independent of the preceding variables. I performed a final hierarchical regression using a similar logic to predict caregiver burden based on demographic and contextual variables, mother's level of functional dependence, amount of instrumental and emotional caregiving, and attachment dimensions. Results Care Provided to Mothers The possible range of scores for the instrumental caregiving scale was 31–115, and the mean in the current sample was 62.29 (SD = 16.81, range = 33–106). The possible range of scores for the emotional caregiving scale was 22–110, and the mean in the current sample was 64.51 (SD = 14.60, range = 27–98). Overall, daughters in this sample were providing a range of care in both domains. The bivariate correlation between the two types of caregiving was not significant (r = .22, p > .05). Adult Attachment AAS scores ranged from 29 to 112 and had a distribution (M = 81.13; SD = 24.01) similar to that reported by Cicirelli 1993(M = 81.23, SD = 20.86). As a group, daughters in this sample appeared to experience moderate to strong feelings of attachment toward their mother, although there was wide variability in the attachment scores. On the RQ, using the highest dimensional rating as an indicator of dominant attachment style, I found that the distribution of attachment styles was as follows: Secure, 66%, n = 53; Fearful, 9%, n = 7; Preoccupied, 6%, n = 5; Dismissing, 19%, n = 15. These figures are consistent with the distribution of attachment styles found in other studies with adults (e.g., Collins and Read 1990; Feeney and Noller 1990). Correlations among the AAS and RQ ratings appear in Table 1 . Scores on the AAS were significantly positively correlated with self-ratings on the Secure style and significantly negatively correlated with self-ratings on the Fearful and Dismissive styles. Correlations within the RQ were also significant and in expected directions. To maximize variance accounted for by the two highly correlated attachment measures and to maintain a concise number of variables, I combined the AAS and RQ ratings into factor scores from a principal components analysis. Factor scores were calculated with unit weighting. The first factor was composed of the following scores (rotated factor loadings are in parentheses): AAS score (.90), RQ Secure rating (.83), and RQ Dismissive rating (−.82). The eigenvalue for the first factor was 2.91, which accounted for 58% of the variance. On the basis of its components, this factor seemed to represent the security of attachment to mother and a daughter's tendency to seek or avoid a close relationship with her. This factor was named the Security dimension, and it corresponds to Bartholomew's other model of internal representations (Bartholomew and Horowitz 1991). Although the eigenvalue for the second factor (.97, 20% of the variance) was below the standard 1.0 cutoff, the factor accounted for a significant portion of variance and was therefore included. In addition, it mapped coherently onto Bartholomew 1990 two-dimension theory of internal attachment representations. The second factor was composed of RQ Fearful (factor loading = .67) and RQ Preoccupied (factor loading = .92) ratings and appeared to symbolize anxiety and self-doubt experienced by daughter in her relationship with her mother. This factor was named the Anxiety dimension, and it corresponds to Bartholomew's self model of internal representations. Univariate Analyses of Attachment and Caregiving Estimated marginal mean caregiving levels and standard errors for each attachment style appear in Table 2 . A one-way ANOVA revealed no significant differences between groups on instrumental care, F(3,76) = 1.84, p = .15, although the low sample size in the Fearful group limited the power to detect differences. Significant group differences were found among the groups in terms of emotional care, F(3,76) = 13.64, p < .001. Post-hoc comparisons revealed that daughters with a secure attachment style provided more emotional care than daughters with any of the insecure attachment styles. No significant differences were found among the insecure attachment styles. A similar pattern of relationships among attachment dimensions and caregiving is apparent in the bivariate correlations in Table 1 . The only attachment variable significantly associated with instrumental care was the RQ Fearful index (r = −.26, p < .05), with higher scores associated with less instrumental care. Meanwhile, every attachment variable was significantly correlated with emotional care. Multivariate Analyses of Attachment and Caregiving In the next set of analyses I used hierarchical linear regression to explore the relationship between caregiving and demographic, family structure, contextual, and attachment variables. Prior to the regressions two cases with univariate outliers were identified: one daughter with a high instrumental caregiving score and one daughter with a high score on the Anxiety dimension. These cases were retained with their original data because they were not statistically influential cases, as indicated by a Cook's distance less than 1.0 (Cook and Weisberg 1982), and they did not produce a substantial change in the regression coefficients when they were dropped (J. Stevens 1996). As a set, the independent variables were found to be neither singular nor adversely multicollinear. Residual scatterplots suggested adequate fulfillment of the assumptions of linearity, homoscedasticity, and independence of residuals. Bivariate correlations for the variables in these analyses appear in Table 3 . The large number of correlations precluded direct interpretation because of the possibility of a Type I error, but relationships throughout the correlation matrix were in the directions expected. Table 4 contains results from the three hierarchical regressions, the first with instrumental care as the dependent variable, the second with emotional care as the dependent variable, and the third with caregiver burden as the dependent variable. For each independent variable the table contains the unstandardized regression coefficient (b); the standard error of the regression coefficient (SE b); the standardized regression coefficient (β); and the squared semipartial correlation (sr2), which represents the unique proportion of variance contributed by the variable. In the prediction of instrumental care, number of coresident children emerged as a significant predictor; daughters living with more children provided more instrumental care. In the second step, mother's level of functional dependence was a significant predictor; mothers with more impairment received more instrumental care. In the third step, amount of emotional care was a significant predictor, indicating the overlap between both types of care. In the final step, when the Security and Anxiety attachment dimensions were added, neither added significantly to the explanation of instrumental caregiving, ΔR2 = .00, F(10,69) = .08, p = .94. The variance contributed by the independent variables that were significant at this final step was as follows: number of children in the household (3% of the variance), mother's functional dependence (32%), and amount of emotional care (6%). To summarize the interpretation of this first regression, demographic and family structure variables appeared unrelated to the amount of instrumental care daughters were providing to their mothers with the exception that daughters with more children in their household were providing more instrumental care. Mothers who were more functionally dependent also received more instrumental care. And mothers who were receiving more emotional care from their daughters tended to be receiving more instrumental care as well. Attachment dimensions were unrelated to the amount of instrumental care daughters were providing. In the first step in the prediction of emotional care, work status was significant; working daughters provided less care. In subsequent steps mother's functional dependence emerged as a significant predictor (more impaired mothers received more emotional care), as did instrumental care, again indicating overlap between types of care. In the final step, when the Security and Anxiety attachment dimensions were added to the regression equation, there was a significant change in the multiple correlation, ΔR2 = .21, F(10,69) = 14.69, p < .001. Both attachment dimensions contributed unique variance: Higher scores on the Security dimension were associated with more emotional care (15% of the variance), and higher scores on the Anxiety dimension were associated with less emotional care (6% of the variance). Three other independent variables made significant contributions: work status (3% of the variance), mother's functional dependence (6%), and instrumental care (11%). To summarize the results of this second regression, the only demographic variable that added significantly to the explanation of emotional care was work status, where working daughters tended to provide less emotional care. Mothers who were more functionally dependent were receiving more emotional care, which also accompanied larger amounts of instrumental care. Finally, daughters high on the Security dimension and low on the Anxiety dimension provided more emotional care (see Appendix, Note 2). In a final regression to predict caregiver burden, only two independent variables emerged as significant at the end of the sequence. The number of children currently living with daughters accounted for 6% of the variance in caregiver burden; daughters with more children at home reported more burden. The Security attachment dimension accounted for 6% of the variance in caregiver burden; daughters with a higher degree of security reported less burden. Discussion In the current study I used the life-span developmental framework of attachment theory to examine associations between socioemotional bonds and the caregiving that daughters were providing to their older mothers. As predicted, attachment patterns are unrelated to the amount of instrumental care daughters provide. In contrast, attachment patterns are significantly associated with the amount of emotional care daughters provide, with securely attached daughters providing more emotional care than insecurely attached daughters. Daughters with a more secure attachment bond to their mothers also report less caregiver burden. Attachment patterns appear to play a role in the amount and nature of care provided to parents and in the psychological outcomes for caregiving children, although there are important theoretical and methodological issues that remain to be addressed in future research. Attachment and Caregiving In the current study I extend work on attachment and parent care by exploring the relationship between specific attachment dimensions, Security and Anxiety, and two types of care, instrumental support and emotional support. First, adult daughters' attachment patterns are unrelated to the provision of instrumental care to their older mothers. Assistance with shopping, housework, medication management, and social service arrangement are offered to mothers regardless of the quality of the attachment bond between daughter and mother. This finding is consistent with research that has shown a continued provision of instrumental care despite its sometimes overwhelming physical and psychological burden (Cantor 1983). One explanation for this finding may be that genuine functional needs supercede attachment dynamics. A daughter in the current study admitted, "I know what I have to do, and I don't mind doing it, but sometimes it stresses me out … She and I were never really close, but you have to do what you have to do." Social and familial norms may propel instrumental caregiving regardless of the quality of the child–parent relationship (Ikkink, Van Tilburg, and Knipscheer 1999). In fact, practical support may be just the kind of care that enables some daughters to be involved yet emotionally safe. Concrete, task-focused activities such as picking up prescriptions and doing laundry require relatively little direct contact with a parent and can be managed so that they involve minimal emotional exchange and, consequently, minimal risk of friction or disappointment. In contrast, in the current study the provision of emotional care is associated with two attachment dimensions, Security and Anxiety. Higher scores on the Security dimension correspond to a more secure attachment to mother, an internalized sense of her supportiveness, and a willingness to seek intimacy with her. Daughters who score high on this dimension provide more emotional care. As Cicirelli 1991 has suggested, secure daughters may provide emotional support as a way to protect their mother as old age compromises health and hints at the attachment relationship's inevitable end. Insecure daughters, on the other hand, may be less motivated to invest emotional energy in a relationship that has been unsupportive or painful and whose end may be unremarkable or even a relief. A converse relationship exists between attachment Anxiety and emotional care. Higher scores on this dimension represent feelings of poor self-worth and apprehension in relationship with mother, and daughters who score high on this dimension provide less emotional care. They may be fearful of providing emotional care because they doubt whether they can do so effectively or whether they deserve to be in a supportive relationship. One daughter commented, "It never seems like I can do enough. And if I don't go over there for a couple days I end up feeling guilty … It's like I'm still trying to please her." With their precarious self-esteem, insecurely attached daughters may withdraw from their mothers rather than risk conflict or rejection. Overall these results are consistent with Cicirelli 1991, Cicirelli 1993 contention that adult children may provide care to preserve the attachment relationship. Yet the current study also provides evidence that attachment patterns may have more complex associations with caregiving than previously considered. Namely, the provision of care appears to depend on the nature of the attachment bond with parents (i.e., whether it is characterized by security or insecurity). Moreover, the current study clarifies that different types of care (instrumental vs emotional) have different associations with attachment patterns. Instrumental care is provided regardless of attachment patterns, yet the amount of emotional care depends on attachment, with more care provided by securely attached children who want to prolong the nurturance they receive back from their parent (Kramer 1997). Indeed, this study demonstrates that daughters with more secure attachment bonds report lower levels of caregiver burden, as have other studies (Crispi et al. 1997). The correlational nature of these data do not allow causal conclusions, but the data do suggest that secure attachment bonds may be associated with positive psychological outcomes for adult children, adding to a broader literature of similar findings (Main 1996). Altogether, adult attachment patterns appear to have a dual impact on child–parent caregiving interactions: an impact on the nature of care that older adults receive and an influence on how successfully children cope with parent care responsibilities. To conclude the findings from this study, two other factors also are related to parent care. First, there is a significant association between caregiving and daughter's work status, although that relationship differs depending on the domain of care: Working daughters provide less emotional care, but work status is unrelated to instrumental care. This result suggests that when daughters experience multiple demands on their time emotional care for a parent may seem secondary to pragmatic needs. Sitting with mother, reminiscing with her, listening to her talk about things that are important to her—these activities may seem less essential than more conspicuous needs such as dinner that must be provided or laundry that must be done. One indirect effect of the multiple burdens on caregiving children may be, then, that parents experience a qualitative difference in how their children interact with them. Second, the only family structure characteristic that is significantly associated with caregiving is the number of children living with daughters: Daughters with more children at home provide more instrumental care (no systematic relationship emerged for emotional care). With children at home daughters may have the liberty to devote more time to caring for their mother as they delegate responsibilities in their own household to their children. More children at home, however, was also associated with greater caregiver burden, suggesting that the stress of caring for a parent may exist regardless of the supports a woman has in her own home. It is important to return for a moment to review the effect sizes of factors that are associated with caregiving (Rosenthal and Rosnow 1991). Demographic factors have small effect sizes (number of children at home, sr2 = .03 for caregiving, sr2 = .06 for burden; work status, sr2 = .03 for caregiving). Mother's functional dependence has an effect size that is large in regard to instrumental care (sr2 = .32) but small to medium in regard to emotional care (sr2 = .06). Finally, the Security and Anxiety attachment dimensions have small to medium effect sizes for emotional caregiving (sr2 = .15 and .06, respectively) and a small effect size in relation to burden (sr2 = .06). In terms of clinical significance, then, changes in functional dependence can be associated with quite substantial shifts in the amount of instrumental care that is provided to mothers. When an older parent develops incontinence, for instance, a substantial increase in caregiving time and energy may occur. In addition, even small variations in attachment patterns are associated with significant differences in emotional care and burden. The effect sizes for attachment dimensions are particularly noteworthy because those factors represent possible points of intervention. Implications and Limitations Results from this study suggest that attachment patterns may provide some indication of how likely daughters are to provide emotional care to their mothers, thereby making it possible to identify older adults who may be at risk for receiving less emotional support from their families. What to do with that information is, of course, another question. The extent to which adult children's attachment patterns can be changed is unclear (Slade 1999). Yet Bowlby's term, internal working model, was carefully chosen to reflect its nature as a "dynamic representation" (Bretherton 1993, p. 239) that could be modified by direct intervention or in response to reparative attachment relationships. Other authors have described earned attachments and the positive modification of attachment representations suggesting the possibility of creating adaptive relationships between adult children and their parents even when before they were contentious (Byng-Hall 1999; Krause and Haverkamp 1996). Even if attachment modification is not a primary goal, family education programs that examine interpersonal dynamics before caregiving becomes a necessity might offer families an opportunity to discuss caregiving expectations and address discrepancies across generations before they evolve into disappointment or conflict. Results of the current study also have implications for policy efforts that address the division of responsibility for caregiving between families and other sources. Cicirelli 1991 has suggested that interaction between adult children and older parents is an important, ongoing developmental process with benefits for both sides, and policy "should be formulated to assist adult children to maintain attachment and protective behavior toward the parent" (p. 38). How might this proposition be met by children who do not get along with their parents, children with less secure and more anxious attachments? To encourage extensive caregiving in those circumstances, and to expect that care to have a positive emotional tone, may be unrealistic. Strategies that address family involvement in caregiving will need to recognize that not all families want or can tolerate close, emotionally focused interactions (Magai and Cohen 1998). In their recent work on long-term care, Kane, Kane, and Ladd 1998 stated that some family members should not be caregivers and that public policies should recognize the contribution of informal caregivers but not rely on them exclusively. The current study also brings into relief questions that require further empirical attention. First, the theoretical and measurement questions in adult attachment remain numerous and complex. The psychometric properties of the AAS and the RQ are not firmly established, and what they measure remains open to a certain degree of speculation. For instance, the AAS and RQ were worded to measure daughters' attachments to their mothers, but it is possible that an overarching attachment representation rather than a relationship-specific one might be more directly responsible for daughters' behavior (Cicirelli 1998). Future applications of attachment theory to the study of adult child–parent relationships also will require consideration of how those relationships differ in adulthood from their earlier incarnation: changes in the power dynamic, differences in resources and the direction in which they are shared, the newly real potential end of the relationship, and the host of intervening events and relationships both inside and outside the family that can influence contemporary interactions. Another limitation of the current study is that it did not include the perspective of mothers receiving care. Attachment has always been conceived of as a behavioral system, with one party influencing the other in mutual feedback. The reciprocal nature of the caregiving dynamic demands a transactional approach in future research (Magai and Cohen 1998). In fact, it is probably important for investigators to examine the entire system of attachments within a family to understand how attachments influence relationships across and within generations. Some additional caveats deserve mention. The conceptual overlap between the questions I used to measure mother's level of functional dependence and those that assessed instrumental caregiving might explain their relationship in the regressions. In addition, researchers have pointed out the potential error inherent in self-reports of caregiving (Brody, Litvin, Albert, and Hoffman 1994). Regarding the sample itself, this was a group of homogeneous daughters who volunteered to discuss their caregiving and family relationships and therefore may represent a particular kind of family, one with relatively positive relationships. Missing, of course, were daughters who refused to participate and might have more troubled relationships and different patterns of attachment and caregiving. Future research will need to include more diverse caregiving dyads and will need to use creative strategies to recruit families in which caregiving takes place in the context of more overt discord. Final considerations are the cross-sectional nature of these data and the reality that hierarchical regression is not a substitute for experimental control. Variables in the hierarchical regression are entered in a logical sequence, but the technique does not preclude the possibility of reciprocal causality. Attachment may influence caregiving, but caregiving also may influence attachment. At the beginning of caregiving a daughter may feel relatively comfortable with the situation, but her attitude may change if the demands of caregiving become unmanageable (Walker, Acock, Bowman, and Li 1996). Conversely, a daughter who spends more and more time helping her mother may experience burgeoning affection for her, a stronger sense of security, and a more positive attachment bond (Davila, Burge, and Hammen 1997). The possible bidirectional influence of attachment patterns points to the importance of longitudinal designs in clarifying the role of attachment across the caregiving history. Despite these limitations, the current study does suggest that attachment theory is a useful framework for understanding caregiving in adult family relationships. In approaching decades families are likely to experience expanded responsibility caring for older adults, and research that explores connections between emotional and behavioral family patterns will continue to be vital. Notes In a complementary method of administration, respondents can indicate which one description best represents the relationship they have with their mother. In this study the categorical ratings were uniformly consistent with the highest dimensional rating. Because the dimensional ratings provide more detail about the interplay among attachment styles, they were used in analyses. In follow-up analyses potential interaction effects between the attachment dimensions and mother's functional dependence were explored, on the basis of Cicirelli 1993 findings. No significant interactions were found. Table 1. Zero-Order Correlations Among Attachment Variables and Caregiving 1 2 3 4 5 6 7 8 9 1. AAS — .80** −.49** −.19 −.60** .76** .40** .16 .51** 2. RQ–Secure — −.60** −.35** −.59** .83** .64** .21 .49** 3. RQ–Fearful — .43** .34** −.65** −.77** −.26* −.40** 4. RQ–Preoccupied — .19 −.02 −.73** −.04 −.34** 5. RQ–Dismissive — −.80** −.01 .02 −.51** 6. Security dimension — .31** .15 .51** 7. Anxiety dimension — −.13 −.30** 8. Instrumental care — .22 9. Emotional care — Note: AAS = Adult Attachment Scale; RQ = Relationship Questionnaire. * p < .05; **p < .01. Table 2. Estimated Marginal Means and Standard Errors for Instrumental and Emotional Caregiving Instrumental Care Emotional Care Attachment Style M SE M SE Secure .10 .14 .45 .12 Fearful −.77 .35 −.68 .29 Dismissive .12 .40 −.89 .34 Preoccupied .01 .25 −.75 .21 Note: Scores are standardized because of different scales of measurement for instrumental and emotional caregiving. Table 3. Correlations Among Demographics, Attachment Dimensions, Caregiving, and Burden 2 3 4 5 6 7 8 9 10 11 12 13 1. Ethnicity .24* .16 .28* .14 −.13 .17 .03 −.19 .12 −.07 .04 .00 2. SES — .13 .07 −.02 −.08 .21 −.08 −.16 −.06 −.20 −.08 −.03 3. Work status — .20 .05 −.25* .32** .08 −.18 .13 −.21 −.21 −.03 4. Marital status — .34** −.10 .06 −.07 −.19 −.06 .00 .00 .04 5. No. of coresident children — −.06 .06 −.04 −.06 −.10 .12 −.02 .26* 6. Mother's coresidence — −.66** −.05 .10 −.07 .48** −.09 .34* 7. ADL score — .19 .02 .06 −.76** .11 −.52** 8. Duration of caregiving — −.10 .04 −.21 .06 −.11 9. Security dimension — −.01 .15 .51** .26* 10. Anxiety dimension — −.13 −.30** .00 11. Instrumental care — .22* .49** 12. Emotional care — −.15 13. Caregiver burden — Notes: Ethnicity, 1 = African American, 2 = European American; work status, 0 = not working, 1 = working; marital status, 0 = unmarried, 1 = married; coresidence, 0 = no, 1 = yes. ADL = activities of daily living; SES = socieconomic status. * p < .05; **p < .01. Table 4. Summary of Hierarchical Regression Analyses of Demographic, Caregiving, and Attachment Dimensions on Instrumental Care, Emotional Care, and Caregiver Burden Instrumental Carea Emotional Careb Caregiver Burdenc Predictor Variables b SE b β sr 2 b SE b β sr 2 b SE b β sr 2 Step 1 Ethnicity 0.73 2.42 0.02 .00 3.92 2.61 0.14 .01 0.42 1.16 −0.04 .00 SES −0.25 0.66 −0.03 .00 −0.18 0.72 −0.02 .00 0.20 0.32 0.06 .00 Work status 4.82 2.57 0.14 .01 −6.48 2.78 −0.21 .03* 1.20 1.26 0.10 .01 Marital status −1.69 2.47 −0.05 .00 2.96 2.70 0.10 .01 −1.12 1.19 −0.10 .01 No. of coresident children 3.25 1.21 0.19 .03* −2.54 1.36 0.17 .02 1.71 0.61 0.29 .06** Mother's coresidence 0.48 2.95 0.01 .00 −4.05 3.20 −.014 .01 0.92 1.41 0.08 .00 Duration of caregiving −0.02 0.01 −0.09 .01 0.03 0.02 0.15 .02 0.00 0.01 −0.03 .00 Step 2 ADL score −1.64 0.19 −0.82 .32*** 0.82 0.28 0.48 .06** −0.24 0.13 −0.35 .03 Step 3 Instrumental care 0.49 0.12 0.57 .11*** 0.07 0.06 0.22 .01 Emotional care 0.41 0.10 0.35 .07*** 0.03 0.05 0.07 .00 Step 4 Security dimension −0.06 1.40 −0.00 .00 6.37 1.32 0.44 .15*** −1.82 0.67 −0.32 .06** Anxiety dimension 0.43 1.24 0.03 .00 −4.02 1.27 −0.27 .06** 0.58 0.59 0.10 .01 Notes: Coefficients are from the final Step 4 model. Only two significant differences in coefficients were noted from the final results: In the regression on instrumental care, coresidence was significant in Step 2 but in no subsequent steps, and in the regression on caregiver burden, coresidence was significant in Step 1 and ADL score was significant in Step 2, but neither was significant in subsequent steps. ADL = activities of daily living; SES = socioeconomic status; sr2 = squared semi-partial correlation. a For the regression on instrumental care, R2 = .33 for Step 1 (p < .001); ΔR2 = .29 for Step 2 (p < .001); ΔR2 = .11 for Step 3 (p < .001); ΔR2 = .00 for Step 4 (p = .94). b For the regression on emotional care, R2 = .08 for Step 1 (p = .48); ΔR2 = .01 for Step 2 (p = .30); ΔR2 = .25 for Step 3 (p < .001); ΔR2 = .21 for Step 4 (p < .001). c For the regression on caregiver burden, R2 = .21 for Step 1 (p < .05); ΔR2 = .17 for Step 2 (p < .001); ΔR2 = .02 for Step 3 (p = .41); ΔR2 = .06 for Step 4 (p < .05). * p < .05; **p < .01; ***p < .001. Brian D. Carpenter is now at the Department of Psychology, Washington University in St. Louis. Support for this research was provided by a grant from the Retirement Research Foundation and Division 20 of the American Psychological Association. Generous cooperation was provided by Linda Noelker and MaryAnn Caston at the Benjamin Rose Institute, Cleveland, OH. This study was the dissertation research of the author, conducted at Case Western Reserve University under the guidance of Milton E. Strauss. Portions of these data were presented at the 1997 Annual Meeting of the Gerontological Society of America. Address correspondence to Brian D. Carpenter, Department of Psychology, Washington University, Box 1125, St. Louis, MO 63130. Ainsworth M. D. S., 1989. Attachments beyond infancy. American Psychologist 44:709-716. Ainsworth M. D. S., Blehar M. C., Waters E., Wall S., 1978. Patterns of attachment: A psychological study of the strange situation Erlbaum, Hillsdale, NJ. Bartholomew K., 1990. Avoidance of intimacy: An attachment perspective. Journal of Social and Personal Relationships 7:147-178. Bartholomew K., Horowitz L. M., 1991. Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology 61:226-244. Bartholomew K., Shaver P. R., 1998. Methods of assessing adult attachment: Do they converge?. Simpson J. A., Rholes W. S., , ed.Attachment theory and close relationships 25-45. Guilford, New York. Blieszner R., Hamon R. R., 1992. Filial responsibility: Attitudes, motivators and behavior. Dwyer J. W., Coward R. T., , ed.Gender, families, and elder care 105-119. Sage, Newbury Park, CA. Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation: Anxiety and anger. New York: Basic Books. Bowlby J., 1979. The making and breaking of affectional bonds Tavistock, London. Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment (2nd ed.). New York: Basic Books. Bowlby J., 1988. A secure base: Clinical applications of attachment theory Routledge, London. Bretherton, I. (1993). From dialogue to internal working models: The co-construction of self in relationships. In C. A. Nelson (Ed.), Memory and affect in development: The Minnesota Symposium on Child Psychology (pp. 237–263). Hillsdale, NJ: Erlbaum. Brody E. M., Litvin S. J., Albert S. M., Hoffman C. J., 1994. Marital status of daughters and patterns of parent care. Journal of Gerontology: Social Sciences 49:S95-S103. Byng-Hall J., 1999. Family and couple therapy. Cassidy J., Shaver P. R., , ed.Handbook of attachment: Theory, research, and clinical applications 625-645. Guilford, New York. Cantor M. H., 1983. Strain among caregivers: A study of experience in the United States. The Gerontologist 23:597-604. Cicirelli V. G., 1991. Attachment theory in old age: Protection of the attached figure. Pillemer K., McCartney K., , ed.Parent–child relations across the life span 25-42. Erlbaum, Hillsdale, NJ. Cicirelli V. G., 1993. Attachment and obligation as daughters' motives for caregiving behavior and subsequent effect on subjective burden. Psychology and Aging 8:144-155. Cicirelli V. G., 1995. A measure of caregiving daughters' attachment to elderly mothers. Journal of Family Psychology 9:89-94. Cicirelli V. G., 1998. A frame of reference for guiding research regarding the relationship between adult attachment and mental health in aging families. Lomranz J., , ed.Handbook of aging and mental health: An integrative approach 341-353. Plenum, New York. Collins N. L., Read S. J., 1990. Adult attachment, working models, and relationship quality in dating couples. Journal of Personality and Social Psychology 58:644-663. Cook R. D., Weisberg S., 1982. Residuals and influence in regression Chapman & Hall, New York. Crispi E. L., Schiaffino K., Berman W. H., 1997. The contribution of attachment to burden in adult children of institutionalized parents with dementia. The Gerontologist 37:52-60. Crose R., 1994. Family bonding and attachment patterns in late life. Family Therapy 21:17-21. Davila J., Burge D., Hammen C., 1997. Why does attachment style change?. Journal of Personality and Social Psychology 73:826-838. Duncan O. D., 1961. A socioeconomic index for all occupations. Reiss A. J., Jr. , ed.Occupations and social status 109-138. Free Press, New York. Dwyer J. W., Coward R. T., 1991. A multivariate comparison of the involvement of adult sons versus daughters in the care of impaired parents. Journal of Gerontology: Social Sciences 46:S259-S269. Feeney J. A., Noller P., 1990. Attachment style as a predictor of adult romantic relationships. Journal of Personality and Social Psychology 58:281-291. Feeney J. A., Noller P., 1996. Adult attachment Sage, Thousand Oaks, CA. Feeney J. A., Noller P., Hanrahan M., 1994. Assessing adult attachment. Sperling M. B., Berman W. H., , ed.Attachment in adults: Clinical and developmental perspectives 128-152. Guilford, New York. Fillenbaum G. G., 1988. Multidimensional assessment of older adults: The Duke Older Americans Resources and Services Procedures Erlbaum, Hillsdale, NJ. Griffin D. W., Bartholomew K., 1994. Models of the self and other: Fundamental dimensions underlying measures of adult attachment. Journal of Personality and Social Psychology 67:430-445. Hannappel M., Calsyn R. J., Allen G., 1993. Does social support alleviate the depression of caregivers of dementia patients?. Journal of Gerontological Social Work 20:35-51. Hazan C., Shaver P. R., 1990. Love and work: An attachment-theoretical perspective. Journal of Personality and Social Psychology 59:270-280. Ikkink K. K., Van Tilburg T., Knipscheer K. C. P. M., 1999. Perceived instrumental support exchanges in relationships between elderly parents and their adult children: Normative and structural explanations. Journal of Marriage and the Family 61:831-844. Kane R. A., Kane R. L., Ladd R. C., 1998. The heart of long-term care Oxford University Press, New York. Kramer B. J., 1997. Gain in the caregiving experience: Where are we? What next?. The Gerontologist 37:218-232. Krause A. M., Haverkamp B. E., 1996. Attachment in adult child–older parent relationships: Research, theory, and practice. Journal of Counseling and Development 75:83-92. Litvin S. J., Albert S. M., Brody E. M., Hoffman C., 1995. Marital status, competing demands, and role priorities of parent-caring daughters. Journal of Applied Gerontology 14:372-390. Magai C., Cohen C. I., 1998. Attachment style and emotion regulation in dementia patients and their relation to caregiver burden. Journal of Gerontology: Psychological Sciences 53B:P147-P154. Main M., 1996. Introduction to the special section in attachment and psychopathology: 2. Overview of the field of attachment. Journal of Consulting and Clinical Psychology 64:237-243. Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood and adulthood: A move to the level of representation. In I. Bretherton & E. Waters (Eds.), Growing points in attachment theory and research. Monographs of the Society for Research in Child Development, 50 (1-2, Serial No. 209), 66–104. Pruchno R. A., Peters N. D., Kleban M. H., Burant C. J., 1994. Attachment among adult children and their institutionalized parents. Journal of Gerontology: Social Sciences 49:S209-S218. Rosenthal R., Rosnow R. L., 1991. Essentials of behavioral research: Methods and data analysis McGraw-Hill, New York. Rothbard J. C., Shaver P. R., 1994. Continuity of attachment across the life span. Sperling M. B., Berman W. H., , ed.Attachment in adults: Clinical and developmental perspectives 31-71. Guilford, New York. Scharfe E., Bartholomew K., 1994. Reliability and stability of adult attachment patterns. Personal Relationships 1:23-43. Slade A., 1999. Attachment theory and research: Implications for the theory and practice of individual psychotherapy with adults. Cassidy J., Shaver P. R., , ed.Handbook of attachment: Theory, research, and clinical applications 575-594. Guilford, New York. Sroufe L. A., Waters E., 1977. Attachment as an organizing construct. Child Development 48:1184-1199. Stevens G., Cho J. H., 1985. Socioeconomic indexes and the new 1980 census occupational classification scheme. Social Science Research 14:142-168. Stevens J., 1996. Applied multivariate statistics for the social sciences Erlbaum, Mahwah, NJ. Walker A. J., Acock A. C., Bowman S. R., Li F., 1996. Amount of care given and caregiving satisfaction: A latent growth curve analysis. Journal of Gerontology: Psychological Sciences 51B:P130-P142. Weiss R. S., 1982. Attachment in adult life. Parkes C. M., Stevenson-Hinde J., , ed.The place of attachment in human behavior 171-184. Basic Books, New York. West M. L., Sheldon-Keller A. E., 1994. Patterns of relating: An adult attachment perspective Guilford, New York. Wolf D. A., Freedman V., Soldo B., 1997. The division of family labor: Care for elderly parents. Journal of Gerontology 52B: (Special Issue) 102-109.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0B9BFCC313049B368C3A2E4BDE041AEFF8CA4CD9.txt b/test/dataset/in/resources/corpus/Clean_0B9BFCC313049B368C3A2E4BDE041AEFF8CA4CD9.txt
deleted file mode 100644
index 89127c0..0000000
--- a/test/dataset/in/resources/corpus/Clean_0B9BFCC313049B368C3A2E4BDE041AEFF8CA4CD9.txt
+++ /dev/null
@@ -1 +0,0 @@
-]>Nature45071692007112249750260028-08361476-46872007Nature Publishing GroupSupplementary InformationThe file contains Supplementary Figures 1-9 with Legends, Supplementary Tables 1-5 and the MIAME Checklist. Supplementary VideoThe file contains Supplementary Video 1 showing contracting cardiomyocytes derived from differentiated CRES cell lines. 10.1038/nature06357Producing primate embryonic stem cells by somatic cell nuclear transferJ. A.ByrneJ AD. A.PedersenD AL. L.ClepperL LM.NelsonMW. G.SangerW GS.GokhaleSD. P.WolfD PS. M.MitalipovS MOregon National Primate Research Center and,Oregon Stem Cell Center, Oregon Health & Science University, 505 N.W. 185th Avenue, Beaverton, Oregon 97006, USAMunroe-Meyer Institute, 985450 Nebraska Medical Center, Omaha, Nebraska 68198, USAWhitehead Institute for Biomedical Research, 9 Cambridge Center, Cambridge, Massachusetts 02142, USAPresent address: Stanford Institute for Stem Cell Biology and Regenerative Medicine, Stanford University, Palo Alto, California 94304, USA.Correspondence and requests for materials should be addressed to S.M.M. (mitalipo@ohsu.edu).Derivation of embryonic stem (ES) cells genetically identical to a patient by somatic cell nuclear transfer (SCNT) holds the potential to cure or alleviate the symptoms of many degenerative diseases while circumventing concerns regarding rejection by the host immune system. However, the concept has only been achieved in the mouse, whereas inefficient reprogramming and poor embryonic development characterizes the results obtained in primates. Here, we used a modified SCNT approach to produce rhesus macaque blastocysts from adult skin fibroblasts, and successfully isolated two ES cell lines from these embryos. DNA analysis confirmed that nuclear DNA was identical to donor somatic cells and that mitochondrial DNA originated from oocytes. Both cell lines exhibited normal ES cell morphology, expressed key stem-cell markers, were transcriptionally similar to control ES cells and differentiated into multiple cell types in vitro and in vivo. Our results represent successful nuclear reprogramming of adult somatic cells into pluripotent ES cells and demonstrate proof-of-concept for therapeutic cloning in primates.&e071122-1; &nature06357-s1;ES cells can differentiate into multiple cell types, representatives of which could be used in replacement therapy for ageing or diseased cells and tissues. However, ES cells derived from in vitro fertilized (IVF) embryos are genetically divergent from the patient (allogenic) and thus any resultant transplanted cell would be rejected without the continual application of immunosuppressive drugs. One way to resolve completely the transplant rejection issue would be to generate ES cells that are genetically identical to the patient. Recent reports suggest that somatic cells can be directly reprogrammed into ES-cell-like cells in mice, after transfection with key stem-cell-specific (stemness) genes. However, the ability to extrapolate this approach to other mammals, including primates, remains in question, and the propensity of such cells to develop into tumours owing to c-myc transgene reactivation remains a concern. Alternatively, ES cells can be derived by epigenetic reprogramming of somatic cells via SCNT in spindle-free oocytes, a process commonly referred to as therapeutic cloning (Supplementary Fig. 1). However, despite encouraging results in the murine model, the feasibility of therapeutic cloning in primates remains in question. Although many mammals have been successfully produced by means of SCNT no successful primate reproductive cloning has been achieved. Moreover, the efficiency of blastocyst formation after human and non-human primate SCNT has typically been very low, suggesting lack or incomplete nuclear reprogramming with existing SCNT protocols. We recently reported incomplete nuclear remodelling, including nuclear envelope breakdown and premature chromosome condensation, after standard SCNT in rhesus macaques (Macaca mulatta) and correlated this observation with a decline in maturation promoting factor activity. SCNT protocols designed to prevent premature cytoplast activation and maturation promoting factor decline resulted in robust nuclear envelope breakdown and premature chromosome condensation and in significantly increased blastocyst development, suggesting that maturation promoting factor activity is instrumental for efficient nuclear reprogramming. Here we used these SCNT protocols to demonstrate therapeutic cloning in the rhesus macaque and provide a translational model for testing the feasibility, therapeutic effectiveness and long-term safety of therapeutic cloning as a concept.Embryo development and ES cell derivationThe primary culture of skin fibroblasts, used as the source of nuclear donor cells for SCNT, was established from a nine-year-old adult rhesus macaque male (male 1) housed at the Oregon National Primate Research Center (Supplementary Fig. 2). Mature metaphase II rhesus macaque oocytes were rendered spindle-free with the Oosight spindle imaging system that uses polarized light to visualize the oocyte meiotic spindle (Supplementary Fig. 3). Analysis of the removed karyoplasts, for the presence of the meiotic spindles, consistently confirmed a 100% efficiency of spindle removal using this approach. The donor fibroblast nuclei were introduced into cytoplasts by electrofusion, incubated for 2 h to allow nuclear remodelling to occur, and subsequently activated and cultured to the blastocyst stage as described previously (see also Methods). We observed a 16% (35 out of 213) blastocyst formation rate with this nuclear donor cell line (Supplementary Table 1). SCNT blastocysts demonstrated a similar morphology to low-grade IVF blastocysts (Supplementary Fig. 4). Twenty expanded or hatching SCNT blastocysts were used for ES cell derivation via mechanical inner cell mass (ICM) isolation (n = 2), immunosurgery (n = 15) or direct culture of intact blastocysts (n = 3) on mouse embryonic fibroblast (MEF) feeder layers. Two ES cell lines (cloned rhesus embryonic stem: CRES-1 and CRES-2) were derived, both after immunosurgery (10% derivation efficiency from blastocysts). Overall, 304 oocytes collected from 14 rhesus macaque females were used to generate two ES cell lines, a 0.7% derivation efficiency from oocytes.Genetic analysis and pluripotencyAs the number of mitochondria, each with 16.6 kilobases (kb) of mitochondrial DNA (mtDNA), in the cytoplast dwarfs any mitochondrial contribution from the donor somatic cells, embryos derived by SCNT should predominantly, if not exclusively, possess mitochondria inherited from the oocyte. Therefore, ES cells derived from SCNT embryos should contain mtDNA identical to the female providing the recipient cytoplasts and nuclear DNA genetically identical to the male providing the nuclear donor cells. To investigate whether the CRES-1 and CRES-2 cell lines contained the same nuclear DNA as the donor (male 1) fibroblasts, we performed microsatellite typing using 39 short tandem repeat (STR) loci and analysis of 56 single nucleotide polymorphisms (SNPs), 30 of which were informative for inheritance. Both the STR analysis—which included 25 common STR loci (Table 1) and 14 major histocompatibility complex (MHC)-linked STRs (Supplementary Table 2)—and the SNP analysis (Supplementary Table 3) demonstrated a complete match of both CRES lines to each other and to the nuclear DNA isolated from skin fibroblasts and peripheral blood leucocytes of male 1. In contrast, DNA obtained from the oocyte donor females for CRES-1 (female 1) and CRES-2 (female 2) demonstrated no significant similarity to CRES-1 or CRES-2 (Table 1 and Supplementary Tables 2 and 3). The genomic constitution of an IVF-derived rhesus macaque ES cell line (ORMES-22; ref. 17), and the ORMES-22 oocyte donor female (female 3) and sperm donor male (male 2), were also included to demonstrate STR allele inheritance (Table 1 and Supplementary Table 2).To investigate whether the CRES-1 and CRES-2 cell lines contained the same mtDNA as their respective oocyte donor females, we performed mtDNA sequence analysis investigating an informative domain 1 (ID1) in the rhesus macaque mitochondrial D-loop hypervariable region 2 (RhDHV2). This RhDHV2 sequence contained multiple informative SNPs including at ID1 nucleotide positions 4, 22 and 28 (Fig. 1). Analysis of SNP22 (an A-to-G polymorphism) demonstrated that CRES-1 mtDNA was derived from the oocyte donor female 1 and not from the nuclear donor for CRES-1. Similarly, analysis of SNP4 (a C-to-T polymorphism) and SNP28 (an A-to-G polymorphism) confirmed that the CRES-2 mtDNA was derived from the CRES-2 oocyte donor female 2 and not from the nuclear donor. Thus, microsatellite, SNP and mtDNA analyses verified that CRES-1 and CRES-2 contained nuclear DNA genetically identical to the nuclear donor fibroblasts and mtDNA inherited from oocytes, a hallmark of SCNT-produced ES cells and offspring.Both CRES lines demonstrated typical ES cell morphology (Fig. 2a–d), maintained an undifferentiated morphology after repeated manual passaging (>20 passages per line so far) and expressed key primate stemness markers including OCT4 (also called POU5F1), SSEA-4, TRA1-60 and TRA1-81 (Fig. 2). Moreover, transcripts of other stemness genes including NANOG, SOX2, LEFTYA (also called LEFTY2), TDGF and TERT were detected by polymerase chain reaction with reverse transcription (RT–PCR) analysis in both IVF-derived ES cell controls (ORMES-10 and ORMES-22) and CRES cell lines (Supplementary Fig. 5). Conventional cytogenetic G-banding analysis of the nuclear donor fibroblasts used for SCNT (Supplementary Fig. 6) and the CRES-2 cell line (Fig. 3a) demonstrated a normal male rhesus macaque chromosome (42, XY) complement in all cells analysed. However, analysis of the CRES-1 cell line indicated the presence of three metaphase cells representing a hypodiploid clone characterized by loss of the Y chromosome (Supplementary Fig. 7), and seventeen cells representing a diploid clone characterized by an isochromosome comprised of two copies of the long arm of the Y chromosome (41,X[3]/42,X,i(Y)q10)[17]) (Fig. 3b). Subsequent fluorescent in situ hybridization (FISH) analysis confirmed the G-banding findings; metaphase cells revealed the presence of a signal for bacterial artificial chromosome (BAC) CH250-283K14 on both arms of the Y chromosome (Fig. 3d), indicating the presence of the i(Y)(q10) observed in the G-banding study. Additional studies were positive for loss of the Y chromosome in 12% of the CRES-1 cells analysed.Transcriptional profilingGlobal transcription profiles of three biological replicates each of male 1 skin fibroblasts (nuclear donor for both CRES lines), both CRES-1 and CRES-2 cell lines and two control ES cell lines derived from fertilized embryos (ORMES-10 and ORMES-22) were examined by Affymetrix microarray analysis. For the primary microarray comparison (Supplementary Data 3), three types of analyses were performed: (1) replicates of each cell line were compared against each other, (2) each cell line was compared against the somatic donor cell line; and (3) each cell line was compared to a control IVF-derived ES cell line (see Methods). For each comparison, the detected signal for each present ‘P’ probe set (P < 0.05) was plotted in a scatter graph, the number of present probe sets (PP) used was recorded and the correlation value was calculated. All comparisons of control ORMES biological replicates with each other demonstrated a correlation value of greater than 60% and all unrelated sample comparisons (that is, between ES cell and somatic cell biological replicates) demonstrated a correlation value of significantly less than 60%; therefore a correlation value of 60% or greater was considered indicative of a significant transcriptional correlation. When the replicates of the somatic donor cells were compared, 99% transcriptional correlation was observed (Supplementary Fig. 8a), suggesting that minimal artificial variation was introduced via the protocols used. Although it was not possible to determine with certainty the degree of technical versus biological variation between replicates, it should be noted that all samples were processed identically and at the same time, and the level of technical variation between the donor somatic cell samples was 1% or less, suggesting that most of the 20–30% transcriptional variation observed between ES cell replicates was biological in origin. If so, ES cells show significant transcriptional plasticity not observed in somatic cells (Supplementary Fig. 8a). However, further transcriptional profiling will be required to confirm this observation. Comparisons of the CRES cell lines to the somatic donor cells and control IVF-derived ES cells demonstrated that both CRES lines had fully reprogrammed into an ES cell transcriptional state, with no significant transcriptional correlation between the CRES lines and the donor somatic cells (Supplementary Fig. 8b) but a significant correlation between CRES cells and the control ES cells (Supplementary Fig. 8c).To identify ES-cell-specific genes, comparison analysis was performed between each of the three control ORMES-10 replicates and each of the three somatic donor cell replicates, to give a total of nine ES cell–somatic comparisons (Supplementary Data 4). This set of ES cell comparisons identified 4,998 somatic-cell-specific probe sets/genes (Supplementary Data 5) and 6,178 ES-cell-specific probe sets/genes (Supplementary Data 6 and Methods). Over 90% of the somatic-cell-specific genes were significantly downregulated in the CRES cell line replicates (Supplementary Table 4), and over 85% of the ES-cell-specific genes demonstrated significantly greater expression in the CRES cell line replicates (Table 2) when compared with the somatic donor cells. Transcriptional analysis of the control ORMES-22 replicates also demonstrated that over 90% of the somatic-specific genes had significantly less expression (Supplementary Table 4) and over 85% of the ES-cell-specific genes had significantly greater expression (Table 2) when compared with the somatic donor cells.The final microarray analysis involved examining the level of expression of 12 putative rhesus macaque stemness genes identified in previous transcriptional profiling. These putative stemness genes had the highest average fold change in gene expression when undifferentiated ES cell biological replicates were compared to their in vitro differentiated counterparts, and all 12 were significantly upregulated in the five different ES cell lines examined. All 12 stemness genes were significantly upregulated in all of the ORMES-10, ORMES-22, CRES-1 and CRES-2 replicates (Supplementary Data 7), and the average fold change in gene expression for both CRES-1 and CRES-2 was comparable to that for ORMES-10 and ORMES-22 when compared to somatic donor cell replicates (Table 3). As a control, the relative fold change in expression for these putative stemness genes between the donor somatic cell replicates was insignificant (Table 3).Our overall conclusion after analysis of global transcriptional profiles is that both CRES-1 and CRES-2 cells are transcriptionally similar to control ES cell lines derived from IVF-produced blastocysts.Differentiation potentialTo define pluripotency further, both CRES lines were exposed to conditions for cardiomyocyte differentiation in vitro. CRES-1 and CRES-2 efficiently produced contracting aggregates (Supplementary Video) expressing markers of cardiac muscle tissue (Supplementary Fig. 9). Directed neural differentiation resulted in efficient formation of various neural phenotypes, demonstrating elongated cellular morphology and expression of neural markers, including microtubule-associated protein 2 (MAP2; Fig. 2e), &bgr;-III-tubulin (Fig. 2f) and tyrosine hydroxylase (Fig. 2g). When injected into severe combined immune deficiency (SCID) mice, both CRES lines formed teratomas and subsequent histological analysis identified representatives of all three germ layers (Fig. 2h–k), confirming their pluripotent status.DiscussionOur results demonstrate successful reprogramming of primate somatic cells into pluripotent ES cells, provided that an efficient SCNT method is available. We achieved a significant increase in the blastocyst formation rate (from 1% to 16%) when the use of Hoechst 33342 was avoided during the oocyte spindle removal step. The detrimental effect of fluorochrome bisbenzimide (Hoechst 33342) and ultraviolet light on cytoplast quality has been previously documented, and we speculate that the impaired blastocyst formation rate after conventional SCNT in primates may result from one or more of the following factors: Hoechst 33342 and/or ultraviolet damage to the relatively transparent primate oocyte; Hoechst 33342/UV-induced oocyte activation and/or maturation promoting factor degradation; reaction of the residual Hoechst 33342 in cytoplasts with the introduced donor cell DNA, thereby impairing reprogramming; and/or Hoechst 33342 contact with mitochondrial DNA, thus reducing cytoplast mitochondrial function.Recognizing the importance of high-quality cytoplasts for successful reprogramming, we have sought non-invasive approaches for spindle detection and removal. ‘Blind’ enucleation techniques involving ‘squish’ or ‘one-step manipulation’ were inefficient, at least in our hands, because they failed to enucleate all oocytes. In fact, our initial effort to derive ES cells from SCNT blastocysts produced with these protocols resulted in the isolation of parthenogenetic ES cells owing to failed spindle removal. Fortunately, recent developments in high-performance imaging resulted in an Oosight spindle imaging system supporting rapid and highly efficient real-time enucleation of primate oocytes. Introduction of the donor nucleus can be accomplished by either direct injection or electrofusion—the latter was dictated here by the relatively large donor cells used. The SCNT blastocyst development rate was significantly lower than fertilized controls but within the 7–29% range described by us when fetal fibroblasts, adult ear fibroblasts, cumulus and oviductal epithelial cells were used as nuclear donor cells in the monkey. Blastocysts produced by SCNT usually demonstrated poor morphology and have, so far, failed to support a term pregnancy after embryo transfer despite considerable efforts (D.P.W., unpublished data). This result may not be predictive for ES cell isolation, however, as the requirements for reproductive and therapeutic cloning are different in that a normal trophectoderm is not required for the latter.With an adequate supply of SCNT blastocysts, the final challenge in therapeutic cloning is ES cell isolation, and although several methods were examined, the conventional method involving immunosurgical dispersal of the trophectoderm seemed to be the best. The derivation efficiency in the present study is within the range reported in the mouse, where the ES cell derivation efficiency from SCNT embryos was 0.2–3.4% per oocyte and 4–10% per blastocyst.Regarding the origin of the CRES lines, in addition to the 100% spindle removal efficiency, karyotype, microsatellite and SNP analyses confirmed that both CRES lines originated from SCNT embryos and not from parthenotes. CRES lines demonstrated typical ES cell morphology, self-renewal capacity and expression of stemness markers. These cell lines were also transcriptionally similar to ES cells derived from fertilized blastocysts, and pluripotent, as demonstrated by the generation of representatives of all three germ layers after in vivo teratoma formation. Our results parallel findings in the mouse and confirm the possibility of complete nuclear reprogramming of somatic cells into pluripotent ES cells via SCNT. The CRES-1 cell line revealed a translocation of unknown origin characterized by an isochromosome comprised of two copies of the long arm of chromosome Y, whereas CRES-2 exhibited a normal euploid male karyotype. In our previous study, five out of seven analysed SCNT blastocysts were karyotypically normal with two embryos demonstrating aneuploidy. Karyotyping individual SCNT embryos based on either biopsied blastomere or trophectodermal cell analysis followed by ES cell isolation could address whether chromosomal abnormalities in SCNT-derived ES cells originate from aneuploid embryos or occur during ES cell isolation and culture.The primary rationale for therapeutic cloning is transplantation of histocompatible ES-cell-derived phenotypes back into the patient. The MHC profile of both CRES lines perfectly matched the donor male in all MHC loci examined via microsatellite analysis, suggesting that transplantation of differentiated derivatives back into the donor animal would not lead to rejection. It is possible that immune response will occur in the donor animal resulting from the epitopes derived from the allogenic recipient oocyte mitochondria contribution; although preliminary research with cloned bovine and porcine cells and tissues with allogenic mtDNA suggests that grafting back into the nuclear donor organism can occur without destruction by the immune system. The use of biological materials from other species, such as fetal bovine serum, mouse embryonic fibroblasts and guinea-pig complement, may also contaminate ES cells with xeno-epitopes and pathogens that would make such cells unsuitable for clinical applications. Other possibly minor factors that should nevertheless also be investigated include the problems that may arise from the lack of sperm-derived centrosomes, the possible effects of non-random X-inactivation and the possibility that cloned ES cells may possess the shortened telomeres of their donor somatic cell source. The possibility has also been suggested that SCNT embryos are epigenetically abnormal based on DNA methylation patterns. Therefore, the epigenetic status of CRES cell lines including imprinted gene expression would also be an interesting area for future research.Assuming that our modified SCNT protocols are applicable to humans, the low efficiency of ES cell derivation (0.7%) along with restricted availability and high cost of human oocytes suggest that a significant increase in the overall SCNT embryo generation and ES cell derivation rates would be required before a clinical implementation of human therapeutic cloning could occur. Recently, mouse ES cells were relatively efficiently cloned using aged, failed-to-fertilize oocytes, and substantial numbers of such human oocytes augmented by immature oocytes are routinely discarded, representing an untested source for human SCNT. In addition to therapeutic cloning, human SCNT-derived ES cells would also be extremely valuable in studies devoted to understanding disease mechanisms and developing possible treatments through in vitro experimentation.Methods SummaryA primary culture of fibroblasts was established from a skin biopsy of an adult rhesus macaque male (male 1) and prepared for SCNT as previously described. Mature metaphase II oocytes were rendered spindle-free using the Oosight imaging system (CRI, Inc.) and a donor somatic cell nucleus was introduced into a cytoplast through electrofusion. Reconstructed embryos were activated 2 h after fusion by exposure to 5 &mgr;M ionomycin (CalBiochem) for 5 min followed by a 5-h incubation in 2 mM 6-dimethylaminopurine (DMAP), placed in HECM-9 medium and cultured at 37 °C in 6% CO2, 5% O2 and 89% N2 until the expanded blastocyst stage. The ICMs of selected SCNT blastocysts were plated onto MEF feeder layers and cultured in ES cell culture medium for 5–7 days. ICMs that attached to the feeder layer and initiated outgrowth were manually dissociated into small clumps with a microscalpel and replated onto fresh MEFs. After the first passage, colonies with ES-cell-like morphology were selected for further propagation, characterization, low-temperature storage and in vitro and in vivo differentiation, as previously described.AnimalsAdult rhesus macaques housed in individual cages were used in this study. All animal procedures were approved by the Institutional Animal Care and Use Committee at the ONPRC/OHSUOvarian stimulation of rhesus macaquesControlled ovarian stimulation and oocyte recovery was performed as has previously been described. Briefly, cycling females were subjected to follicular stimulation using twice-daily intramuscular injections of recombinant human FSH as well as concurrent treatment with Antide, a GnRH antagonist, for days 8–9. Unless indicated otherwise, all reagents were from Sigma-Aldrich and all hormones were from Ares Advanced Technologies Inc. Females received recombinant human luteinizing hormone on days 7-9 and recombinant human chorionic gonadotrophin (hCG) on day 10.Somatic cell nuclear transferA primary culture of fibroblasts was established from an adult rhesus macaque male (male 1) as previously described. Briefly, a small skin biopsy was surgically derived, washed in Ca2+- and Mg2+-free Dulbecco PBS (Invitrogen) and minced into pieces before incubation in Dulbecco Modified Eagle’s Medium (DMEM, Invitrogen) containing 1 mg ml-1 collagenase IV (Invitrogen) at 37 °C in 5% CO2 for 40 min. Tissue pieces were then vortexed, washed, seeded into 75 cm3 cell culture flasks (Corning) containing DMEM supplemented with 100 IU ml-1 penicillin, 100 &mgr;g ml-1 streptomycin (Invitrogen), 10% FBS (DMEM/FBS culture media) and cultured at 37 °C in 5% CO2 until reaching confluency. Fibroblasts were then disaggregated with trypsin treatment and frozen down in aliquots of 1 × 106 cells in medium containing 10% dimethyl sulphoxide (DMSO).Fibroblasts were subsequently thawed, plated onto 4-well dishes (Nunc) and cultured under standard conditions until reaching 50–90% confluency. Cells were then synchronized in the G0/G1 phase of the cell cycle by culturing in DMEM medium with 0.5% FBS for 4 days before SCNT. Cumulus–oocyte complexes were collected from anaesthetized animals by laparoscopic follicular aspiration (28–29 h after hCG) and placed in TALP/HEPES medium (modified Tyrode solution with albumin, lactate and pyruvate) containing 0.3% BSA (TH3) at 37 °C. Oocytes were stripped of cumulus cells by mechanical pipetting after brief exposure (<1 min) to hyaluronidase (0.5 mg ml-1) and placed in chemically defined, protein-free HECM-9 medium (hamster embryo culture medium) at 37 °C in 6% CO2, 5% O2 and 89% N2 until further use.Recipient MII oocytes were transferred to 30 &mgr;l manipulation droplets of TH3 with 5 &mgr;g ml-1 cytochalasin B on a glass bottom manipulation dish (http://www.willcowells.com) covered with paraffin oil (Zander IVF) and incubated at 37 °C for 10–15 min before spindle removal. The chamber was then mounted on an inverted microscope (Olympus) equipped with the Oosight imaging system (CRI, Inc.), glass stage warmer (Tokai Hit, http://www.tokaihit.com) and Narishige micromanipulators. The spindle was located and extracted by aspiration into an enucleation pipette (20–25 &mgr;m outer diameter). Metaphase spindle removal was confirmed by its presence in the enucleation pipette. The Oosight imaging system allows non-invasive, polarized light imaging and detection of the spindle based on birefringence. Using this innovative approach, we were able to locate and quickly remove the oocyte spindle real-time with 100% efficiency. After oocyte spindle removal a disaggregated donor somatic cell was aspirated into a micropipette and placed into the perivitelline space of the cytoplast on the side opposite the first polar body. Cell fusion was induced by two 50 &mgr;s DC pulses of 2.7 kV cm-1 (Electro Square Porator T-820, BTX, Inc.) in 0.25 M d-sorbitol buffer containing 0.1 mM calcium acetate, 0.5 mM magnesium acetate, 0.5 mM HEPES and 1 mg ml-1 fatty-acid-free BSA. Successful fusion was confirmed visually 30 min after electroporation by the disappearance of the donor cell in the perivitelline space. All nuclear transfer micromanipulation and fusion procedures were conducted on microscope stage warmers (Tokai Hit) maintaining 37 °C. Reconstructed embryos were activated by exposure to 5 &mgr;M ionomycin for 5 min in TALP/HEPES medium supplemented with 1 mg ml-1 fatty-acid-free BSA and then transferred for 5 min in TALP/HEPES medium supplemented with 30 mg ml-1 fatty-acid-free BSA and 2 mM 6-dimethylaminopurine (DMAP) followed by a 5 h incubation in HECM-9 medium containing 2 mM DMAP at 37 °C in 6% CO2. Activated oocytes were placed in 4-well dishes containing HECM-9 medium supplemented with 10% FBS and 12 &mgr;M &bgr;-mercaptoethanol (BME) and cultured at 37 °C in 6% CO2, 5% O2 and 89% N2 for a maximum of 10 days with medium change every other day.ES cell derivation and cultureZonae pellucidae of selected expanded SCNT blastocysts were removed by brief exposure (45–60 s) to 0.5% pronase in TH3 medium. A small proportion of embryos were either transferred directly to MEFs as whole blastocysts or after mechanical dissection of the ICM. Remaining blastocysts were subjected to immunosurgical isolation of the ICMs as previously described. Briefly, zona-free blastocysts were exposed to rabbit anti-rhesus spleen serum (a gift from J. A. Thomson) for 30 min at 37 °C. After extensive washing in TH3, embryos were incubated in guinea-pig complement reconstituted with HECM-9 (1:2, v/v) for an additional 30 min at 37 °C. Partially lysed trophectodermal cells were mechanically dispersed by gentle pipetting with a small bore pipette (125 &mgr;m inner diameter; Stripper pipette, Midatlantic Diagnostics Inc.) followed by the rinsing of ICMs three times with TH3 medium. Isolated ICMs were plated onto Nunc 4-well dishes containing mitotically inactivated feeder layers consisting of MEFs and cultured in DMEM/F12 medium with glucose and without sodium pyruvate supplemented with 1% nonessential amino acids, 2 mM l-glutamine, 0.1 mM &bgr;-mercaptoethanol and 15% FBS at 37 °C, 3% CO2, 5% O2 and 92% N2. ICMs that attached to the feeder layer and initiated outgrowth were manually dissociated into small cell clumps with a microscalpel and replated onto new MEFs. After the first passage, colonies with ES-cell-like morphology were selected for further propagation, characterization and low-temperature storage as previously described. Medium was changed daily and ES cell colonies were split every 5–7 days by manual dissociation and replating collected clumps onto dishes with fresh MEFs.In vitro and in vivo differentiation of ES cellsFor embryoid body formation, entire ES cell colonies were loosely detached from feeder cells and transferred into feeder-free, 6-well, ultra-low adhesion plates (Corning Costar) and cultured in suspension in ES cell medium for 5–7 days. To induce further differentiation, embryoid bodies were transferred into collagen--well culture dishes (Becton Dickinson) to allow attachment. For neuronal differentiation, medium was replaced with serum-free DMEM/F12 containing ITS supplement (insulin, transferrin and sodium selenite, Invitrogen) and fibronectin (5 &mgr;g ml-1; Invitrogen). Cultures were maintained for 7 days, with medium replenishment every 2 days. The resulting cultures were disaggregated with collagenase or trypsin treatment and replated onto polyornithine- and laminin-coated plates or glass coverslips in N2 medium consisting of DMEM/F12 supplemented with laminin (1 &mgr;g ml-1; Invitrogen), bFGF (10 ng ml-1; R&D Systems), and N2 supplement (Invitrogen). Cultures were maintained for an additional 7 days with a daily medium change. After 7 days, bFGF was omitted from the medium and cultures were maintained for an additional 7–12 days to induce differentiation into mature neuronal phenotype. Differentiation into cardiac cells was initiated by embryoid body formation in suspension as described above. Embryoid bodies were then plated into collagen-coated dishes and cultures were maintained in ES cell medium for 2–4 weeks. In order to obtain teratomas, 3–5 million undifferentiated ES cells from each cell line were harvested and injected into the hind-leg muscle or subcutaneously into 4-week-old, SCID, beige male mice using an 18 gauge needle. Six to seven weeks after injection, mice were killed and teratomas were dissected, sectioned and histologically characterized for the presence of representative tissues of all three germ layers.Immunocytochemical proceduresUndifferentiated and differentiated ES cells were fixed in 4% paraformaldehyde for 20 min. After permeabilization with 0.2% Triton X-100 and 0.1% Tween-20, nonspecific reactions were blocked with 10% normal serum (Jackson ImmunoResearch Laboratories, Inc.). Cells were then incubated for 40 min in primary antibodies, washed three times and exposed to secondary antibodies conjugated with fluorochromes (Jackson ImmunoResearch) before co-staining with 2 &mgr;g ml-1 4',6-diamidino-2-phenylindole (DAPI) for 10 min, whole-mounting onto slides and examination under epifluorescence microscopy. Primary antibodies for OCT4, SSEA-4, TRA1-60 and TRA1-81 were from Santa Cruz Biotechnology Inc. Neural-specific antibodies including microtubule-associated protein (MAP2C), &bgr;-III-tubulin and tyrosine hydroxylase were from Chemicon International Inc.Cytogenetic analysisMitotically active ES cells in log phase were incubated with 120 ng ml-1 ethidium bromide for 40 min at 37 °C, 5% CO2, followed by 120 ng ml-1 colcemid (Invitrogen) treatment for 20–40 min. Cells were then dislodged with 0.25% trypsin, and centrifuged at 200g for 8 min. The cell pellet was gently re-suspended in 0.075 M KCl solution and incubated for 20 min at 37 °C followed by fixation with methanol:glacial acetic acid (3:1) solution. Cytogenetic analysis was performed on 20 metaphase cells from each ES cell line following standard GTW-banding procedures. Images were acquired using the Cytovision Image Analysis System (Applied Imaging) and karyotypes were arranged as previously described. FISH analysis was performed on metaphase cells from the donor somatic cell line and CRES-1 using BAC CH250-283K14, specific for the rhesus macaque Yq11.21 region. The clone was obtained from the CHORI-250 rhesus macaque BAC library (Children’s Hospital Oakland Research Institute, BACPAC Resources), grown in LB media supplemented with 25 µg ml-1 chloramphenicol and isolated using a mini-prep protocol adapted from the standard Qiagen plasmid purification method (Qiagen Inc.). Probes were nick-translated with rhodamine-5-dUTP (Enzo Life Sciences) using a modification of the manufacturer’s protocol (Vysis). Approximately 200 ng of the probe was precipitated along with 2 &mgr;g human Cot-I DNA (Invitrogen). Before hybridization, slides were pre-treated at 73 °C in 2× SSC for 2 min, followed by digestion at 37 °C in a protease solution for 5 min, post-fixation in 10% buffered formaldehyde with 2 M MgCl, and dehydration. The slide and probe mixture were co-denaturated at 74 °C for 2 min and incubated overnight at 37 °C using the HYBrite system (Vysis). Images were acquired using the Cytovision image analysis system.RT–PCRTotal RNA was extracted from cells using an RNA purification kit (Invitrogen) according to the manufacturer’s instructions. Total RNA was treated with DNase I before cDNA preparation using the SuperScript III first-strand synthesis system for RT–PCR (Invitrogen) according to the manufacturer’s instructions. The first strand cDNA was further amplified by PCR using individual primer pairs for specific genes. The primer sequence, annealing temperature and amplicon size for each pair of primers are listed in Supplementary Table 5. All PCR samples were analysed by electrophoresis on 2% agarose gel containing 0.5 &mgr;g ml-1 ethidium bromide and visualized on a transilluminator.Microsatellite analysisFor STR genotyping, DNA was extracted from blood or cultured cells using commercial kits (Gentra). Six multiplexed PCR reactions were set up for the amplification of 39 markers representing 25 autosomal loci and 14 autosomal, MHC-linked loci. On the basis of the published rhesus macaque linkage map, these markers are distributed in 19 chromosomes. Two of the markers included in the panel, MFGT21 and MFGT22, were developed from Macaca fuscata and do not have a chromosome assignment. PCRs were set up in 25 &mgr;l reactions containing 30–60 ng DNA, 2.5 mM MgCl2, 200 &mgr;M dNTPs, 1× PCR buffer II, 0.5 U Amplitaq (Applied Biosystems) and fluorescence-labelled primers in concentrations ranging from 0.06 to 0.9 &mgr;M, as required for each multiplex PCR. Cycling conditions consisted of four cycles of 1 min at 94 °C, 30 s at 58 °C, 30 s at 72 °C, followed by 25 cycles of 45 s at 94 °C, 30 s at 58 °C, 30 s at 72 °C and a final extension at 72 °C for 30 min. PCR products were separated by capillary electrophoresis on the ABI 3730 DNA analyser (Applied Biosystems) according to the manufacturer’s instructions. Fragment size analysis and genotyping was done with the computer software STRand (http://www.vgl.ucdavis.edu/informatics/STRand/). Primer sequences for MHC-linked STRs 9P06, 246K06, 162B17(A and B), 151L13, 268P23 and 222I18 were designed from the corresponding rhesus macaque BAC clone sequences deposited in GenBank (accession numbers AC148662, AC148696, AC148683, AC148682, AC148698 and AC148689, respectively). Loci identified by the letter ‘D’ prefix were amplified using heterologous human primers.SNP analysisSNP analysis was performed as previously described. Briefly, SNP genotyping was performed using iPLEX reagents and protocols for multiplex PCR, single base primer extension and generation of mass spectra, as per the manufacturer’s instructions (for complete details see iPLEX Application Note, Sequenom). Four multiplexed assays containing 56 SNPs were included, 30 which were informative for CRES inheritance (see Supplementary Table 3). All 56 SNPs demonstrated a 100% match between male 1 donor cells and the CRES-1 and CRES-2 cells lines. Initial multiplexed PCRs were performed in 5 &mgr;l reactions on 384-well plates containing 5 ng of genomic DNA. Reactions contained 0.5 U HotStar Taq polymerase (QIAGEN), 100 nM primers, 1.25× HotStar Taq buffer, 1.625 mM MgCl2 and 500 &mgr;M dNTPs. After enzyme activation at 94 °C for 15 min, DNA was amplified with 45 cycles of 94 °C × 20 s, 56 °C × 30 s, 72 °C × 1 min, followed by a 3-min extension at 72 °C. Unincorporated dNTPs were removed using shrimp alkaline phosphatase (0.3 U, Sequenom). Single-base extension was carried out by addition of single base primer extension primers at concentrations from 0.625 &mgr;M to 1.25 &mgr;M using iPLEX enzyme and buffers (Sequenom). Single base primer extension products were measured using the MassARRAY Compact system, and mass spectra were analysed using TYPER software (Sequenom).Mitochondrial DNA analysisDNA was extracted from cell lines using commercial kits (Gentra). The rhesus macaque mitochondrial D-loop hypervariable region 2 (RhDHV2) sequence was amplified for each sample using primers RhDF2 (5′-TAACATATCCGATCAGAGCC-3′) and RhDR (5′-TTAAACACCCTCTACGCCG-3′). PCR for each sample was performed using Platinum PCR SuperMix (Invitrogen) containing 0.5 &mgr;M of each primer (final volume 50 &mgr;l). Reaction conditions were initial denaturation at 94 °C for 2 min; 35 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, extension at 72 °C for 90 s and a final extension at 72 °C for 3 min, generating 544 bp of sequence covering the RhDHV2 region. Products from these reactions were then sequenced to determine polymorphic variation in the RhDHV2 region by direct sequencing. The informative domain 1 (ID1) sequence encompassing Macaca mulatta mtDNA nucleotide positions 451–480 (GenBank NC_005943) was identified as containing SNPs informative for the mitochondrial inheritance of both CRES-1 and CRES-2. Each ID1 sequence was confirmed by three other sequencing reactions and all of the RhDHV2 chromatograms used in this project were obtained with Sequencher v. 4.7 (GeneCodes).Microarray analysisThe MIAME (Minimum Information About a Microarray Experiment) guidelines for microarray research were incorporated into the design and implementation of these studies. All the microarray comparison analysis data files cited in the text are available online at the Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc = GSE7748). All the microarray comparison analysis data cited in the text are available at ftp://ftp.ncbi.nih.gov/pub/geo/DATA/supplementary/series/GSE7748. Total RNA was isolated from cell colonies selected for the appropriate ES cell morphology (flat monolayer colony with distinctive cobbled stem cell morphology and a high nucleo-cytoplasmic ratio) using the Invitrogen TRIZOL reagent, followed by further purification with the Qiagen RNeasy MinElute Cleanup Kit. For each cell line examined three biological replicates were used and each replicate represented either a different passage or different sub-line (which had been cultured independently and without mixing for several passages). The RNA samples were quantified using the NanoDrop ND-1000 UV-Vis spectrophotometer (NanoDrop Technologies) and the quality of the RNA was assessed using Lab-on-a-Chip RNA Pico Chips and the 2100 Bioanalyser (Agilent Technologies). Samples with electropherograms that showed a size distribution pattern predictive of acceptable microarray assay performance were considered to be of good quality. Two micrograms of total RNA from each sample were amplified and labelled using a single-cycle cDNA synthesis and an in vitro transcription cRNA-RNA labelling system (GeneChip one-cycle target labelling and control reagents; Affymetrix). Following successful cRNA amplification, 10 &mgr;g of labelled target cRNA was hybridized to rhesus macaque Genome Arrays (Affymetrix) using standard protocols, as described in the GeneChip Expression Analysis manual (http://www.affymetrix.com/support/technical/manual/expression_manual.affx). The rhesus macaque array contains 52,865 probe sets, representing over 20,000 genes. The arrays were scanned using the GeneChip laser scanner (Affymetrix) and image processing, normalization and expression analysis were performed with the Affymetrix GCOS version 1.4 software. MAS-5 statistical analysis was performed to calculate the signal log ratio (SLR) for each probe set comparison, and the gene expression fold changes between two samples were calculated from the SLR using the following formula: fold change = (2SLR). The GCOS 1.4 MAS 5.0 software was used to calculate statistically significant differences in gene expression (P < 0.002) between samples.For the primary microarray comparison analysis (Supplementary Data 3) the log10 of the absolute detected signal for each present ‘P’ probe set (P < 0.05) was plotted in a scatter graph (Supplementary Fig. 8) using Affymetrix GeneChip Operating Software (GCOS) version 1.4. For the correlation value calculations, the microarray data for each individual cell line comparison was filtered to only include probe sets present (P < 0.05) in both cell lines. The present probe sets (PP) value details the number of probe sets, post filtering, with a present (P < 0.05) signal in both compared cell lines. The correlation value for each cell line comparison was calculated using MAS-5 (Affymetrix microarray suite 5) analysis to calculate the proportion of compared probe sets which demonstrated no significant change ‘NC’ in gene expression (Supplementary Data 3). Each cell line had three biological replicates and the letter after the cell line name details which replicate was used in the primary microarray comparison analysis.For the secondary gene-specific analysis, comparison analysis was performed between each of the three control ORMES-10 biological replicates and each of the three somatic donor cell replicates, to give a total of nine somatic–ES cell comparisons (Supplementary Data 4). The following selection criteria were used to identify rhesus somatic-specific genes: (1) genes that were considered to be present (P < 0.05) in all three somatic donor cell replicates; and (2) genes that demonstrated statistically significant decrease ‘D’ in gene expression in the ORMES-10 replicates in all nine comparisons with the somatic donor cell replicates after GCOS comparisons with MAS-5 statistical analysis. A total of 4,998 somatic-specific probe sets were identified in this way (Supplementary Data 5). The following selection criteria were used to identify rhesus ES-cell-specific genes: (1) genes that were considered to be present (P < 0.05) in all three ORMES-10 replicates; and (2) genes that demonstrated a statistically significant increase ‘I’ in gene expression in the ORMES-10 ES cell replicates in all nine comparisons with the somatic donor cell replicates after GCOS comparisons with MAS-5 statistical analysis. A total of 6,178 ES-cell-specific probe sets were identified in this way (Supplementary Data 6). It should be noted that the general approximation when working with large numbers of probe sets is to assume that each probe set represents hybridization to a single gene. However, multiple probe sets can exist for certain genes, so the actual number of genes included in the analysis is significantly lower than the number of probe sets analysed. The somatic-specific and ES-cell-specific genes identified from this comparison analysis were then used to investigate whether the CRES cell lines had successfully downregulated somatic-specific genes (Supplementary Table 4) and successfully upregulated ES-cell-specific genes (Table 2) after comparison analysis with the three somatic donor cell replicates.For the tertiary stemness gene analysis, stemness genes were derived from our previous transcriptional profiling work with rhesus macaque embryonic stem cell lines. These 12 putative stemness genes had the highest average fold change in gene expression when three undifferentiated biological replicates of ORMES-6 were compared to their in vitro differentiated counterparts, and all 12 were significantly upregulated in five different rhesus macaque embryonic stem cell lines examined. Comparison analysis was performed between the ORMES and CRES cell line replicates and the donor somatic cell replicates (Supplementary Data 7), and the average fold change increase in gene expression of the 12 stemness genes in the ORMES and CRES cell lines was calculated (Table 3). All of the normalized microarray data sets generated from these studies can be found on the GEO website, as noted above.Statistical analysisNon-microarray results were analysed by chi-squared and unpaired t-test using Statview Software (SAS Institute, Inc.) with statistical significance set at 0.05. Microarray results were analysed by MAS-5 statistical analysis in Affymetrix GeneChip Operating Software (GCOS) version 1.4.Chromatograms of the rhesus macaque mitochondrial D-loop hypervariable region 2 informative domain 1 (RhDHV2 ID1).The ID1 sequence encompassed Macaca mulatta mtDNA nucleotide positions 451–480 (GenBank accession number NC_005943). Sequence analysis revealed that SNP22 (A-to-G) was informative for the mitochondrial inheritance for CRES-1 whereas SNP4 (C-to-T) and SNP28 (A-to-G) were informative for the mitochondrial inheritance for CRES-2. Results were confirmed by three independent sequences.Morphological and immunocytochemical analysis of CRES cells and their differentiated derivatives.a–d, Morphology of CRES colonies and expression of primate ES-cell-specific markers: OCT4, SSEA-4, TRA1-60 and TRA1-81. For each stemness marker analysed, a positive signal was only observed in CRES colonies, not in mouse embryonic fibroblast feeder cells, demonstrating antibody specificity. e–g, Cellular morphology and expression of neural markers in differentiated CRES cells, including microtubule-associated protein 2 (MAP2), &bgr;-III-tubulin and tyrosine hydroxylase. The left hand column of a–g represents phase contrast images and the right hand column demonstrates marker expression as detected via immunocytochemistry. The scale bars in a–d, g represent 100 &mgr;m and the scale bars in e, f represent 50 &mgr;m. h, j, Presence of ectoderm (Ec)-derived neuroepithelium (×400, haematoxylin and eosin (H&E)) in teratomas produced by injection of CRES-1 and CRES-2 into SCID mice. i, k, Presence of mesoderm (Me)-derived cartilage and endoderm (En)-derived glandular epithelium (×100, H&E) in CRES-1 and CRES-2 teratomas.Cytogenetic analysis of CRES cells.a, G-banding of CRES-2 cells demonstrating a normal euploid rhesus karyotype (42, XY). b, CRES-1 cell line with aneuploid karyotype, characterized by an isochromosome comprised of two copies of the long arm of the Y chromosome (41,X[3]/42,X,i(Y)q10)[17]). c, FISH analysis of the nuclear donor fibroblasts demonstrating a normal karyotype with two red fluorescent signals on the long (q) arm of the Y chromosome. d, FISH analysis of CRES-1 cells indicating the presence of four signals for the Y chromosome long (q) arm, confirming the presence of the i(Y)(q10) aneuploidy.Short tandem repeat (STR) analysis of CRES cell linesSTR lociFemale 1oocyte donor for CRES-1Female 2oocyte donor for CRES-2Male 1 somatic nuclear donor for CRES-1 and CRES-2CRES-1CRES-2Female 3oocyte donor for ORMES-22Male 2 sperm donor for ORMES-22ORMES-22 CRES-1 is an ES cell line isolated from a SCNT blastocyst produced by fusion of a cytoplast donated by female 1 and a skin fibroblast originated from male 1. Similarly, CRES-2 was derived from a cytoplast donated by female 2 and a donor somatic nucleus from male 1. ORMES-22 is an ES cell line derived from an embryo produced by IVF of a female 3 oocyte with a male 2 sperm. AME, amelogenin. Sex (AME)XXXXXYXYXYXXXYXXD1S548190/206190/198190/190190/190190/190190/190190/190190/190D2S1333301/301293/301289/301289/301289/301273/293285/289273/285D3S1768221/221205/213193/217193/217193/217205/213205/205205/205D4S2365283/283275/287283/283283/283283/283283/283283/283283/283D4S413131/131133/145131/139131/139131/139131/145125/141131/141D5S1457136/136132/136132/136132/136132/136132/136132/140136/140D6S501176/180176/180176/180176/180176/180188/192180/180180/188D7S513191/205205/209189/191189/191189/191189/217193/199199/217D7S794108/124124/128128/128128/128128/128108/108108/128108/128D8S1106144/144148/160144/148144/148144/148148/168160/168168/168D9S921183/195183/191179/179179/179179/179183/195175/195183/195D10S1412157/166160/160157/157157/157157/157157/157160/160157/160D11S2002256/256256/256260/264260/264260/264252/252256/260252/256D11S925308/338310/316308/310308/310308/310308/308338/338308/338D12S364282/290282/288281/290281/290281/290282/290268/296268/290D12S67121/129192/204117/125117/125117/125117/133109/117109/133D13S765228/240212/220228/256228/256228/256216/236228/228228/236D15S823333/349329/353329/361329/361329/361357/385345/353345/385D16S403164/168156/158158/164158/164158/164152/164152/152152/164D17S1300232/280244/280272/276272/276272/276248/252228/284252/284D18S537178/178178/178174/178174/178174/178174/178162/174162/178D18S72306/308306/322306/308306/308306/308308/308306/308308/308D22S685315/319291/303315/327315/327315/327311/311327/327311/327MFGT21113/115117/119115/115115/115115/115111/113115/125113/125MFGT22104/104104/104100/104100/104100/104100/104104/110104/104Analysis of ES-cell-specific gene expression in rhesus macaque stem cell linesCell lineBiological replicateNumber of ES cell genes* upregulated compared to donor line replicate A†Number of ES cell genes* upregulated compared to donor line replicate B†Number of ES cell genes* upregulated compared to donor line replicate C† *The general approximation when working with large numbers of probe sets is to assume that each probe set represents hybridization to a single gene. However, multiple probe sets can exist for certain genes, so the actual number of genes included in the analysis is significantly lower than the number of probe sets analysed. †The percentage value represents the proportion of probe sets, out of the 6,178 ES-cell-specific probe sets, significantly upregulated after comparison analysis. Nuclear donor cellsAN/A21 (0.3%)47 (0.8%)Nuclear donor cellsB30 (0.5%)N/A77 (1.2%)Nuclear donor cellsC18 (0.3%)13 (0.2%)N/AORMES-22A5,482 (89%)5,388 (87%)5,389 (87%)ORMES-22B5,558 (90%)5,607 (91%)5,644 (91%)ORMES-22C5,766 (93%)5,672 (92%)5,723 (93%)CRES-1A5,974 (97%)6,001 (97%)5,984 (97%)CRES-1B5,896 (95%)5,919 (96%)5,926 (96%)CRES-1C5,748 (93%)5,845 (95%)5,784 (94%)CRES-2A5,931 (96%)5,843 (95%)5,850 (95%)CRES-2B5,658 (92%)5,552 (90%)5,483 (89%)CRES-2C5,863 (95%)5,933 (96%)5,889 (95%)Expression analysis of putative rhesus macaque stemness genesAffymetrix probe set IDStemness gene*Nuclear donor cellsfold change†ORMES-10fold change†ORMES-22fold change†CRES-1fold change†CRES-2fold change† *The stemness genes were identified in previous transcriptional profiling of rhesus macaque ES cell lines. These 12 genes had the highest average fold change in gene expression of five different rhesus macaque ES cell lines. †Microarray suite-5 (MAS-5) statistical analysis was performed to calculate the signal log ratio (SLR) for each cell line in comparison to the somatic donor cell line biological replicates, and the averaged gene expression fold change between compared samples was calculated from the SLR using the formula: fold change = (2SLR). All 12 stemness genes were significantly upregulated (P < 0.002) in CRES cell lines. MmugDNA.26523.1.S1_s_atNFE2L31429389532378MmuSTS.2285.1.S1_atPOU5F1 (OCT4)0315288320281MmugDNA.9427.1.S1_atNR5A21282310278325MmugDNA.32128.1.S1_atNANOG1246180256190MmuSTS.1436.1.S1_atLCK117920621894MmugDNA.11728.1.S1_atVTCN11245139153125MmugDNA.42677.1.S1_atDPPA4415411717878MmugDNA.28461.1.S1_atSLC12A1171128185169MmugDNA.6836.1.S1_atC14orf115 (LOS699513)181878564MmuSTS.3122.1.S1_atMYRIP071517553MmugDNA.15193.1.S1_atADH4068135258MmuSTS.2310.1.S1_atPRDM14128444640The authors acknowledge the Division of Animal Resources and the Endocrine Services Cores at the Oregon National Primate Research Center for assistance and technical services. We thank M. Sparman, C. Ramsey and V. Dighe of the Assisted Reproductive Technology Core for their embryological and logistical assistance; J. Fanton and D. Jacobs for laparoscopic oocyte retrievals; B. Ferguson for performing the SNP analysis; C. Penedo for microsatellite analysis; and R. Stouffer, M. Grompe and R. Reijo Pera for reviewing this manuscript. Microarray assays were performed in the Affymetrix Microarray Core of the OHSU Gene Microarray Shared Resource. This study was supported by funds from ONPRC and NIH grants to S. Mitalipov, R. Stouffer and D. Dorsa.Author Contributions S.M.M. and J.A.B. designed experiments, conducted SCNT and ES cell derivation. L.L.C. performed DNA/RNA isolations and stemness gene expression. J.A.B. analysed the microarray data and performed the mitochondrial DNA analysis. D.A.P. assisted with ES cell derivation and performed ES cell culture, characterization and differentiation. W.G.S. and M.N. performed the cytogenetic analysis. S.G. analysed teratomas. S.M.M., J.A.B. and D.P.W. analysed the data and wrote the paper.Microarray data, including CEL and CHP files, and Supplementary Data files containing microarray analyses (Supplementary Data 3–7) have been deposited in the Gene Expression Omnibus (GEO) database with accession number GSE7748 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE7748).McKay, R. Stem cells–hype and hope. Nature 406, 361–364 (2000)Drukker, M. & Benvenisty, N. The immunogenicity of human embryonic stem-derived cells. Trends Biotechnol. 22, 136–141 (2004)Takahashi, K. & Yamanaka, S. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126, 663–676 (2006); published online 10 August 2006Okita, K., Ichisaka, T. & Yamanaka, S. Generation of germline-competent induced pluripotent stem cells. Nature 448, 313–317 (2007); published online 6 June 2007Wernig, M. et al. In vitro reprogramming of fibroblasts into a pluripotent ES-cell-like state. Nature 448, 318–324 (2007); published online 6 June 2007Rideout, W. M., Hochedlinger, K., Kyba, M., Daley, G. Q. & Jaenisch, R. Correction of a genetic defect by nuclear transplantation and combined cell and gene therapy. Cell 109, 17–27 (2002)Barberi, T. et al. Neural subtype specification of fertilization and nuclear transfer embryonic stem cells and application in parkinsonian mice. Nature Biotechnol. 21, 1200–1207 (2003)Wilmut, I. et al. Somatic cell nuclear transfer. Nature 419, 583–586 (2002)Mitalipov, S. M., Yeoman, R. R., Nusser, K. D. & Wolf, D. P. Rhesus monkey embryos produced by nuclear transfer from embryonic blastomeres or somatic cells. Biol. Reprod. 66, 1367–1373 (2002)Simerly, C. et al. Molecular correlates of primate nuclear transfer failures. Science 300, 297 (2003)Hall, V. J. et al. Developmental competence of human in vitro aged oocytes as host cells for nuclear transfer. Hum. Reprod. 22, 52–62 (2007)Mitalipov, S. M. et al. Reprogramming following somatic cell nuclear transfer in primates is dependent upon nuclear remodeling. Hum. Reprod. 22, 2232–2242 (2007)Birky, C. W. Uniparental inheritance of mitochondrial and chloroplast genes: mechanisms and evolution. Proc. Natl Acad. Sci. USA 92, 11331–11338 (1995)Bowles, E. J., Campbell, K. H. & St John, J. C. Nuclear transfer: preservation of a nuclear genome at the expense of its associated mtDNA genome(s). Curr. Top. Dev. Biol. 77, 251–290 (2007)Penedo, M. C. et al. Microsatellite typing of the rhesus macaque MHC region. Immunogenetics 57, 198–209 (2005)Ferguson, B. et al. Single nucleotide polymorphisms (SNPs) distinguish Indian-origin and Chinese-origin rhesus macaques (Macaca mulatta). BMC Genomics 8, 43 (2007)Mitalipov, S. et al. Isolation and characterization of novel rhesus monkey embryonic stem cell lines. Stem Cells 24, 2177–2186 (2006)Byrne, J. A., Mitalipov, S. M., Clepper, L. & Wolf, D. P. Transcriptional profiling of rhesus monkey embryonic stem cells. Biol. Reprod. 75, 908–915 (2006)Byrne, J. A., Clepper, L., Wolf, D. P. & Mitalipov, S. M. in International Society for Stem Cell Research, 5th Annual meeting 52 (Cairns, Queensland, Australia, 2007)Simerly, C. et al. Embryogenesis and blastocyst development after somatic cell nuclear transfer in nonhuman primates: overcoming defects caused by meiotic spindle extraction. Dev. Biol. 276, 237–252 (2004)Zhou, Q. et al. A comparative approach to somatic cell nuclear transfer in the rhesus monkey. Hum. Reprod. 21, 2564–2571 (2006)Ng, S. C. et al. The first cell cycle after transfer of somatic cell nuclei in a non-human primate. Development 131, 2475–2484 (2004)Yang, X. et al. Nuclear reprogramming of cloned embryos and its implications for therapeutic cloning. Nature Genet. 39, 295–302 (2007)Munsie, M. J. et al. Isolation of pluripotent embryonic stem cells from reprogrammed adult mouse somatic cell nuclei. Curr. Biol. 10, 989–992 (2000)Wakayama, T. et al. Differentiation of embryonic stem cell lines generated from adult somatic cells by nuclear transfer. Science 292, 740–743 (2001)Hochedlinger, K. & Jaenisch, R. Monoclonal mice generated by nuclear transfer from mature B and T donor cells. Nature 415, 1035–1038 (2002)Mombaerts, P. Therapeutic cloning in the mouse. Proc. Natl Acad. Sci. USA 100, 11924–11925 (2003); published online 29 August 2003Wakayama, S. et al. Equivalency of nuclear transfer-derived embryonic stem cells to those derived from fertilized mouse blastocysts. Stem Cells 24, 2023–2033 (2006); published online 11 May 2006Brambrink, T., Hochedlinger, K., Bell, G. & Jaenisch, R. ES cells derived from cloned and fertilized blastocysts are transcriptionally and functionally indistinguishable. Proc. Natl Acad. Sci. USA 103, 933–938 (2006); published online 17 January 2006Lanza, R. P. et al. Generation of histocompatible tissues using nuclear transplantation. Nature Biotechnol. 20, 689–696 (2002); published online 3 June 2002Martin, M. J. et al. Skin graft survival in genetically identical cloned pigs. Cloning Stem Cells 5, 117–121 (2003)Yang, J. et al. Epigenetic marks in cloned rhesus monkey embryos: comparison with counterparts produced in vitro. Biol. Reprod. 76, 36–42 (2007)Wakayama, S. et al. Establishment of mouse embryonic stem cell lines from somatic cell nuclei by nuclear transfer into aged, fertilization-failure mouse oocytes. Curr. Biol. 17, R120–R121 (2007)Byrne, J., Mitalipov, S. & Wolf, D. Current progress with primate embryonic stem cells. Curr. Stem Cell Res. Ther. 1, 127–138 (2006)Zelinski-Wooten, M. B., Hutchison, J. S., Hess, D. L., Wolf, D. P. & Stouffer, R. L. Follicle stimulating hormone alone supports follicle growth and oocyte development in gonadotrophin-releasing hormone antagonist-treated monkeys. Hum. Reprod. 10, 1658–1666 (1995)Bavister, B. D. & Yanagimachi, R. The effects of sperm extracts and energy sources on the motility and acrosome reaction of hamster spermatozoa in vitro. Biol. Reprod. 16, 228–237 (1977)McKiernan, S. H. & Bavister, B. D. Culture of one-cell hamster embryos with water soluble vitamins: pantothenate stimulates blastocyst production. Hum. Reprod. 15, 157–164 (2000)Kuo, H. C. et al. Differentiation of monkey embryonic stem cells into neural lineages. Biol. Reprod. 68, 1727–1735 (2003)Pearson, P. L. et al. Report of the committee on comparative mapping. Cytogenet. Cell Genet. 25, 82–95 (1979)Rogers, J. et al. An initial genetic linkage map of the rhesus macaque (Macaca mulatta) genome using human microsatellite loci. Genomics 87, 30–38 (2006)Domingo-Roura, X., Lopez-Giraldez, T., Shinohara, M. & Takenaka, O. Hypervariable microsatellite loci in the Japanese macaque (Macaca fuscata) conserved in related species. Am. J. Primatol. 43, 357–360 (1997)St John, J. C. & Schatten, G. Paternal mitochondrial DNA transmission during nonhuman primate nuclear transfer. Genetics 167, 897–905 (2004)Brazma, A. et al. Minimum information about a microarray experiment (MIAME)—toward standards for microarray data. Nature Genet. 29, 365–371 (2001)
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0BAE243B509F5F3E735D4774B38A8EF81EABBEC1.txt b/test/dataset/in/resources/corpus/Clean_0BAE243B509F5F3E735D4774B38A8EF81EABBEC1.txt
deleted file mode 100644
index f1e41e6..0000000
--- a/test/dataset/in/resources/corpus/Clean_0BAE243B509F5F3E735D4774B38A8EF81EABBEC1.txt
+++ /dev/null
@@ -1 +0,0 @@
- gerona J Gerontol A Biol Sci Med Scigerona The Journals of Gerontology Series A: Biological Sciences and Medical Sciences J Gerontol A Biol Sci Med Sci 1079-5006 1758-535X Oxford University Press 9902310.1093/gerona/55.5.B220 Journal of Gerontology: Biological Sciences Increase of Oxidatively Modified Protein Is Associated With a Decrease of Proteasome Activity and Content in Aging Epidermal Cells Petropoulos Isabelle a Conconi Mariangela a Wang Xin b Hoenel Batya b Brégégère François a b Milner Yoram b Friguet Bertrand a aLaboratoire de Biologie et Biochimie Cellulaire du Vieillissement, Université Denis Diderot, Paris, France bDepartment of Biological Chemistry, The Hebrew University, Jerusalem, Israel Bertrand Friguet, Laboratoire de Biologie et Biochimie Cellulaire du Vieillissement, Universit\|[eacute]\| Denis Diderot-Paris 7, CC 7128, 2 place Jussieu, 75251 Paris Cedex 05, France E-mail: bfriguet@paris7.jussieu.fr. 1 5 2000 55 5 B220 B227 20 9 1999 1 3 1999 The Gerontological Society of America 2000 For the process of aging in epidermal cells to be characterized, the status of oxidized and damaged protein accumulation and removal by the proteasome has been investigated. Modified protein content and proteasome activity were assayed in lysates of epidermal cells from donors of different ages. Increased levels of oxidized proteins, glycated proteins, and proteins modified by the lipid peroxidation product 4-hydroxy-2-nonenal were observed in cells from old donors. At the same time, a decline of chymotrypsin-like and peptidylglutamyl-peptide hydrolase activities of the proteasome was found in aging keratinocytes. This age-related decline of the proteasome peptidase activities can be explained, at least in part, by a decreased proteasome content as observed by immunoblotting and enzyme-linked immunosorbent assay. In keratinocyte cultures, a decrease of proteasome activity and content was observed upon serial passaging. In cultures, as well as in skin, an inverse relationship was found between the aging marker β-galactosidase and the proteasome content. These results suggest that proteasome is downregulated during replicative senescence as well as in aged cells in vivo, possibly resulting in the accumulation of modified proteins. hwp-legacy-fpage B220 hwp-legacy-dochead RESEARCH ARTICLE Decision Editor: Jay Roberts, PhD KERATINOCYTES, which are the main cell type in the epidermis, not only constitute the physicochemical barrier that protects the body from environmental assaults (e.g., by pathogens and toxins) but also have more complex immune functions, such as the expression of class II antigens and interaction with T lymphocytes during inflammation (1). The epidermis is also very sensitive to age-related alterations leading to various major pathologies, including hyperproliferative and autoimmune diseases (2)(3)(4). Numerous theories have been proposed to explain the mechanisms of cellular aging. Among them, the free-radical hypothesis suggests that oxygen free radicals are causal factors in aging, inflicting molecular damages, some of which are not repaired and accumulate with age (5). Indeed, it has been shown that aging cells do accumulate damaged DNA, lipids, and proteins, leading to cellular functional alterations that are strongly believed to contribute to the aging process (6)(7). Among biological macromolecules, proteins constitute a particular target for the reactive oxygen species attacks that generate protein carbonyl derivatives. There is an age-related increase in the carbonyl content of protein in the human brain (8), gerbil brain (9), rat hepatocyte (10), human red blood cells (11), and in the whole body of flies (12). Moreover, the carbonyl content of protein in cultured human fibroblasts increases exponentially as a function of donor age (11). There are other protein modifications, such as glycation, that cause alterations to cellular and organ function (13). Aging is also characterized by a decline of the specific activity and thermostability of enzymes, which is due to oxidative damage as well as to other modifications (11)(13)(14)(15). The accumulation of modified proteins reflects a dysregulation of the steady-state level of damaged proteins in aged tissues, which can be explained by a higher modification rate, or a lower degradation rate of these proteins, or both. The half-lives of proteins in senescent animals have been shown to be significantly extended as compared with those in younger ones (16). It is now well established that the main proteolytic system, responsible for the intracellular degradation of oxidized or misfolded proteins, is the multicatalytic proteinase (MCP), also called 20 S proteasome (17)(18)(19). The 20 S proteasome is a high-molecular-mass (700 kDa) proteolytic complex, made of 28 subunits arranged as four stacked rings (see ref. (20) for a review). The two outer rings are formed by α subunits and the two inner rings are formed by the β subunits that carry the proteolytic activities (21)(22). Recently, we and other investigators have shown that the peptidylglutamyl-peptide hydrolase (PGPH) activity of rat liver proteasome exhibited the most remarkable reduction with age (23)(24)(25)(26). These results strongly suggest a potential participation of the MCP in the slower turnover of proteins, thus contributing to the age-related accumulation of altered proteins. In this study, we present evidence for an impaired proteasome function in human keratinocytes upon aging. We have found a decrease of proteasome peptidase activities in epidermis extracts from old donors as compared with young ones and, as expected, an age-related increase of damaged proteins. We also found a similar decrease in senescent keratinocyte cultures. This can be explained, at least in part, by the decrease of proteasome content found in old cells. Materials and Methods Reagents and Chemicals Leu-Leu-Val-Tyr-7-amido-4-methylcoumarin (LLVY-AMC), N-Cbz-Leu-Leu-Glu-β-naphthylamine (LLE-NA), 7-aminomethylcoumarin, and β-naphthylamine were purchased from Sigma-Aldrich (L'Isle d'Abeau, France). N-Cbz-Leu-Leu-Leucinal (MG 132) was kindly supplied by Dr. F. Arenzana-Seisdedos. The rabbit anti-glycation products polyclonal antibodies (anti-GP) (27), a gift from Dr. H. Bakala, were raised against advanced glycated endproducts conjugated RNase. Rabbit immune sera raised against pure rat MCP and against HNE (4-hydroxynonenal)-modified keyhole limpet hemocyanin (anti-HNE) were obtained from Assay Research (Silver Spring, MD) as previously described (23)(28). Epidermis and Keratinocyte Extracts Preparations Epidermal cells were prepared from skin biopsies or plastic surgery procedures (mammary gland reductions) from healthy donors from 17 to 67 years of age. Skin samples were incubated 2 hours at 4°C in 2 M of NaCl and 20 mM of ethylenediamine tetra-acetic acid (EDTA). Then, the epidermis was separated from the dermis and homogenized with a Dounce homogenizer (Prolabo, Fontenay, France). Primary cultures of keratinocytes were derived from newborn foreskins and cultivated in Dulbecco's modified eagle medium nutrient mixture F-12 ham (1:1) at 37°C and 5% CO2 by using a 3T3 fibroblast feeder layer, as previously described (29). Cells were usually passaged up to six times. Cells were collected with a rubber policeman and homogenized in an extraction buffer (20 mM of Tris, pH 7.4, 250 mM of sucrose, 1 mM of EDTA, and 5 mM of 2-mercaptoethanol) with a Dounce homogenizer. Cell debris and organelles were removed from the crude extracts by centrifugation for 1 hour at 10,000 × g at 4°C. Protein and Peptidase Assays Protein concentrations were determined by a bicinchoninic acid assay (Pierce). Chymotrypsin-like and PGPH activities of the MCP in crude extracts were assayed with fluorogenic peptides LLVY-AMC and LLE-NA, respectively, as previously described (30). Typically, assay mixtures contained 75-μg proteins of epidermal extract, or 50-μg proteins of keratinocyte extract, in extraction buffer with the appropriate peptide substrate (10 μM of LLVY-AMC or 200 μM of LLE-NA) in a final volume of 200 μl. The mixtures were incubated 30 minutes at 37°C and analyzed by spectrofluorometry for release of aminomethylcoumarin (AMC) with excitation and emission wavelengths at 350 nm and 440 nm, respectively, or for release of β-naphthylamine (NA) with excitation and emission wavelengths at 333 nm and 410 nm, respectively. MCP proteolytic activity was determined as the difference between the total activity of crude extracts and the remaining activity in the presence of 20-μM proteasome inhibitor MG 132 (31). Gel Electrophoresis and Western Blots Sodium dodecyl sulfate/polyacrylamide gel electrophoresis (SDS/PAGE) was done by using the Laemmli method (32) on a 12% acrylamide (wt/vol) separating gel. Immunoblot experiments using anti-MCP, anti-GP, or anti-HNE antibodies were performed after SDS/PAGE separation of 20 μg of total cellular proteins from keratinocyte cultures or epidermis followed by electrotransfer onto Hybond nitrocellulose membrane (Amersham-Pharmacia-Biotech, Les Ulis, France). Immunoblot detection of carbonyl groups was performed with the OxyBlot oxidized protein detection kit (Appligene-Oncor, Strasbourg, France), according to the manufacturer. Briefly, 10 μg of proteins were incubated for 15 minutes at room temperature with 2,4-dinitrophenylhydrazine (DNPH) to form the carbonyl derivative dinitrophenylhydrazone before SDS/PAGE separation. After transfer onto nitrocellulose, modified proteins were revealed by anti-DNP antibodies. The detection of bands in immunoblots was carried out with the ECL (Enhanced ChemiLuminescence) Western blotting analysis system of Amersham (Les Ulis, France), using peroxidase conjugated anti-rabbit, secondary antibodies. The densitometric analyses of the autoradiographies were performed by using Lecphor software on a Biocom system (Biocom, Les Ulis, France). Enzyme-linked Immunosorbent Assays and β-Galactosidase Assay A competitive enzyme-linked immunosorbent assay (ELISA) was performed to measure the amount of modified glycated products in skin epidermis upon aging. Microplates were coated with 1 μg/ml of modified glycated bovine serum albumin (BSA) diluted in a buffer carbonate (20 mM, pH 9.6) for 14 hours at 4°C. After they were washed with 0.05% Tween in phosphate-buffered saline (PBS) and blocked with 6% milk (wt/vol) in PBS, epidermis extracts from various donors were added and incubated with polyclonal anti-GP antibodies (27) for 2 hours at room temperature. After washing, peroxidase-conjugated secondary antibodies were added and incubated for 2 hours at room temperature. After washing, the substrate 2, 2′Azino-bis-3-ethylbenz-thiazoline-6-sulfonic acid was added and the absorbance of the reaction product was measured at 405 nm in a SLT Rainbow microplate reader (SLT Labinstruments, Salzburg, Austria). A calibration curve was obtained by using glycated BSA. For the proteasome amount to be determined, cell suspensions were obtained by trypsinization of skin epidermis from donors of different ages, or cultured keratinocytes after different passages. For ELISA tests, 5 × 105 suspended cells were fixed in 3% paraformaldehyde in Eppendorf tubes for 15 minutes at room temperature, then incubated for 15 minutes with 0.1 M of glycine (pH 7.0), and permeabilized by 1% Triton X-100 for 30 minutes at room temperature. Suspended cells were pretreated for 10 minutes with 3% H2O2 and blocked with 5% BSA in PBS for 2 hours at 37°C. They were incubated with anti-MCP antibodies, properly diluted in PBS containing 3% BSA and 0.05% NaN3, overnight at 4°C, and then washed four times with 0.05% Tween in PBS for 10 minutes each. Peroxidase-conjugated secondary antibodies were diluted in PBS, 3% BSA, 0.05% NaN3, and added to the cells for 2 hours at room temperature. Finally, the cells were washed and incubated with 0.4 mg/ml of tetramethyl benzidine and 1.3 mM of H2O2 in 0.1 M of citric acid Na2HPO4 buffer (pH 5.5). The reaction was stopped in 0.33 M of H3PO4 and the absorbance was measured at 450 nm in a BIO-TEK (BIO-TEK Instruments, Winooski, VT) microplate reader (33)(34). For β-galactosidase activity to be measured, 5 × 105 suspended cells were fixed with 3% paraformaldehyde in PBS for 5 minutes, washed, and incubated in 0.5 ml of the following substrate solution: 0.4 mg/ml of p-nitrophenyl-β-galactopyranoside, 2 mM of MgCl2, 150 mM of NaCl, and 0.1% NaN3 in 50 mM of Tris buffer (pH 8.0) supplemented with a mixture of protease inhibitors at 37°C. The reaction product (p-nitrophenol) was measured spectrophotometrically at 405 nm. Statistical Analysis Differences for proteasome and β-galactosidase activities and proteasome antigen data were evaluated by using analysis of variance (ANOVA) procedures with the StatView software (Abacus Concepts, Berkeley, CA). A p value of <.05 was considered significant. Results Increase of Modified Proteins in Aging Epidermis Protein oxidation has been implicated in different biological processes, both physiological, such as aging or apoptosis, and pathological, such as cancer and neurodegenerative diseases (see ref. 6 for a review). In order to investigate the status of oxidatively modified proteins in keratinocytes during aging, protein carbonyl content was monitored by using the OxyBlot detection kit as described in the Materials and Methods section. Proteins of epidermis extracts from young donors (17, 20, and 26 years old), middle-aged donors (39 and 42 years old), and old donors (50, 60, and 67 years old) were treated with DNPH to derivatize their carbonyl groups into 2,4-dinitrophenylhydrazone. Western blots revealed oxidized polypeptides by using anti-DNP antibodies. As shown in Fig. 1, an age-related increase of carbonyl content was observed in some proteins. In samples from old cells in particular, high-molecular-weight polypeptides appeared on the top of the gel (molecular weights above 100 kDa), possibly resulting from cross-linking reactions. As a way to assess whether other modifications occur and to determine the origin of carbonyl groups in these proteins, Western blots were performed by using anti-HNE and anti-GP polyclonal antibodies. A general increase in modified protein adducts was visible with age, and anti-HNE antibodies recognized modified groups in high-molecular-weight proteins (Fig. 1 and Fig. 1). Because the patterns of glycated proteins did not show qualitative changes, a competitive ELISA was performed to measure their amounts and confirmed their increase with age (Fig. 1). Taken together, these results show that glycation and modification by the lipid peroxidation product HNE are implicated in the age-related accumulation of damaged proteins. Age-Related Decrease of Proteasome Activity and Content in Epidermis There is considerable evidence that the 20 S proteasome is responsible for the degradation of oxidatively modified intracellular proteins (35)(36) and that an age-related accumulation of these proteins may be due to a decline of proteasome peptidase activities (23)(24)(25)(26). For this hypothesis to be examined in epidermal cells, the chymotrypsin-like and PGPH activities of proteasome were monitored in epidermis extracts of various donors of from 17 to 67 years of age. As shown in Fig. 2, both activities decreased gradually with the age of donors. As a way to determine whether the age-related decline in the proteasome peptidase activities was due to decreased proteasome content, the amounts of proteasome were estimated by Western blot analysis of epidermis extracts from donors of different ages, using anti-rat MCP. Several proteasome subunits from human keratinocytes were recognized by these antibodies. A representative experiment is shown in Fig. 2, where the proteasome content was indeed reduced by ∼10% and 60% in the epidermis from 50- and 67-year-old donors, respectively, when compared with the sample from the 17-year-old donor. ELISA measurements of MCP antigens performed on epidermal cell suspensions from various donors (see Fig. 5 below) showed a similar age-related decrease. Increase of Protein Damage and Decrease of Proteasome in Keratinocytes Upon Serial Passaging Normal diploid cells undergo numerous physiological and biochemical changes during serial passaging in vitro. These changes may reflect aging events that occur in organisms (37)(38). We used keratinocytes in culture to examine the possible modification of proteasome activity and content in relation to protein damage upon aging. To investigate the status of oxidatively modified proteins in keratinocytes during serial passages, we monitored protein carbonyl content by using the OxyBlot detection kit as described in the Materials and Methods section. Cellular extracts were prepared at different passages and treated with DNPH; the derived products were revealed by an anti-DNP antiserum after SDS PAGE separation and Western blotting. As shown in Fig. 3, an increase of high-molecular-weight modified proteins was observed in passages 3 and 4 as compared with passages 1 and 2. Western blots were also performed by using an anti-HNE polyclonal antibody. This antibody also reacted with more proteins of high molecular weight in later passages than in earlier ones (Fig. 3). The accumulation of such modified high-molecular-weight polypeptides presumably results from protein cross-linking events. Proteasome activity was measured by monitoring the chymotrypsin-like and the PGPH activities in cellular extracts of keratinocytes. The results, summarized in Fig. 4, show a drop in chymotrypsin-trypsin-like activity in passages 4 and 5, compared with passages 1 and 2. A similar profile is observed for the PGPH activity. As a way to determine if the decline of activity can be correlated to a decreased proteasome content in different keratinocytes extracts, a Western blot analysis was performed. Fig. 4 clearly shows a decrease of proteasome level during keratinocyte serial passages. In the passages 3 and 4, the residual level was approximately 30% and 20% respectively, of that observed in the first passage. Decrease of Proteasome and Increase of β-Galactosidase in Aging Epidermis and Keratinocyte Cultures Dimri and colleagues described β-galactosidase as a senescence marker in human fibroblasts and keratinocytes (39). The intracellular β-galactosidase activity was measured in suspensions of keratinocytes prepared from skin epidermis or from cell cultures of different ages, as described in the Materials and Methods section. The amount of proteasome was measured in similar suspensions obtained from the same samples, by ELISA tests using anti-rat MCP antibodies (see the Materials and Methods section). As shown in Fig. 5, β-galactosidase activity increased with serial passages of keratinocytes, while proteasome antigen, detected by ELISA, decreased. An age-related increase of β-galactosidase activity and a decrease of proteasome antigen were also observed in skin epidermal cells of old donors (Fig. 5). Discussion The accumulation of damaged or oxidized protein reflects not only the rate of protein modification but also the rate of damaged or oxidized protein degradation that is dependent, at least in part, on the activity of intracellular proteases that preferentially degrade damaged proteins. The proteasome is the major protease that degrades oxidatively modified cytosolic proteins (35)(36), and it has been reported that an age-related accumulation of damaged or oxidized proteins may be explained by an impaired activity of the proteasome (23)(24)(25)(26). In this work, we have studied the fate of the proteasome in the cellular aging of human epidermis, a tissue very sensitive to aging alterations. We first showed that there is an age-related increase in the protein carbonyl content in the epidermis and in cultured keratinocytes, the main cellular component of the epidermis. This result corroborates previous reports that have established such a correlation in the mammalian brain (8)(9), in rat hepatocytes (10), and in human fibroblasts (11). As demonstrated in this study, in epidermis, as well as in serial passages of keratinocytes, protein carbonyl groups that can be generated by direct oxidation of amino acids are also produced by protein glycation and reactions with peroxidation products of polyunsaturated fatty acids such as HNE (see Fig. 1 and Fig. 3). Interestingly, these modifications seem to be highly selective, because only a few proteins are the target of oxidative damage. This finding is in agreement with previous observations in aging flies, which showed that two mitochondrial proteins, aconitase and adenine nucleotide translocase, are selectively oxidized during aging (40)(41). Furthermore, among these modified proteins that accumulate with age, some are high-molecular-weight proteins. Such high-molecular-weight species might result from cross-linking reactions that occur upon production of bityrosine by the hydroxyl radical (42), of intermolecular lysyl-HNE-lysyl linkages (43)(44), or of Maillard reaction adducts (45). We then addressed the question of whether the accumulation of damaged proteins reflects an alteration in the activity of the proteolytic machinery in charge of damaged or oxidized protein degradation, that is, the 20 S proteasome. Modifications of proteasome activities and content have been extensively studied in rat liver during aging by us and by other groups (23)(24)(25)(26). Shibatani and colleagues (24) reported a significant decrease of the SDS-treated proteasome PGPH activity in cytosolic extracts upon aging. Using purified proteasome preparations, we demonstrated that the PGPH activity declined in old animals to ∼50% compared with young animals (23)(25). More recently, Hayashi and Goto (26) also reported an age-related decline of the PGPH activity and of the chymotrypsin-like activity, although to a lesser extent. In addition, Wagner and Margolis showed an age-related decrease in proteasome activity in the bovine lens (46). We assayed both the chymotrypsin-like and the PGPH activities of proteasome in extracts of epidermal cells, and we found that they were notably reduced in aged cells. Using the specific proteasome inhibitor MG 132, we determined that proteasome accounts for ∼90% of both activities (data not shown). Therefore, this decrease represents a net reduction of proteasome activity. To determine whether the decline of proteasome activity may be due to a decrease in proteasome content upon aging, we performed both ELISA and Western blots by using rabbit polyclonal antibodies raised against rat proteasome. These antibodies recognize the human proteasome and reveal two distinct major bands. The intensities of the two major bands decreased with age (Fig. 2). Similarly, Fig. 4 shows that, in keratinocyte cultures, more than one proteasome subunit decreased with age. However, we cannot yet conclude whether there is a general decline of proteasome content, or a differential reduction in some subunits. It would be of interest to identify among proteasome subunits those that show an age-related decrease. A depletion of proteasome subunits with age has not been detected in rat liver (25)(26), leading us to propose that the activity decline resulted from posttranslational modifications of proteasome subunits. To our knowledge, this is the first time that an age-related decline of proteasome subunit content has been observed. However, it has been shown that the β-subunits X (also called MB 1 or epsilon) and Y (also called delta) are downregulated under pathologic conditions such as infection (47) and after γ-interferon treatment (48). It is interesting to note that these subunits have been involved in chymotrypsin-like and PGPH activities, respectively (49). Furthermore, decreased proteasome content is likely to explain the observed decline in activity but does not rule out the possibility that some subunits are also posttranslationally modified. Moreover, the proteasome activity was tested in crude extracts, and it is possible that the age-related alteration of proteasome activity could also be due to a loss of specific activators, the presence of endogenous inhibitors, or both. In fact, it has been pointed out that HNE cross-linked proteins may act as proteasome inhibitors (43)(50). Thus, further work is needed to ascertain why the proteasome is declining in activity with age. A decrease of proteasome activity and content was also observed in serial passages of keratinocytes. Cells were tested at various passages for β-galactosidase activity, which was used as a marker of replicative senescence in human fibroblasts, keratinocytes (39), and more recently in endothelial and smooth muscle cells from humans and rabbits (51). The inverse relationship between β-galactosidase activity and proteasome content strongly suggests that proteasome is downregulated in replicative senescence. Interestingly, Bosman and colleagues, in pioneering studies in replicative senescence of WI-38 fibroblasts, have shown a correlated increase of the β-galactosidase activity and a loss of a “critical neutral protease” that might be the proteasome (52). The accumulation of oxidized and modified proteins and the correlative decrease in proteasome activity and content during epidermial aging take place in the context of complex cellular events leading to skin alterations. The fact that a similar status of protein modification and of proteasome decline was found in senescent keratinocytes in culture and in aged epidermal cells in vivo establishes cell culture as a valuable tool for research on proteasomes in keratinocyte aging. Further investigations will be necessary to elucidate the actual role of proteasomes in the aging process and may lead to an original approach of aging control in skin. Figure 1. Modification of cytosolic proteins upon epidermal cell aging. A: The OxyBlot was performed with 10 μg of cytosolic protein extracts from epidermal cells from young, middle-aged, and old donors. Extracts were first treated with 2,4-DNPH, then subjected to SDS/PAGE on a 12% polyacrylamide gel, and electrotransferred onto a nitrocellulose membrane. The blots were processed with anti-DNP antibodies as described by the manufacturer. For a densitometric analysis, the measurements corresponding to the 17-year-old donor were taken as 100%. The intensity of the major band was increased to ∼300% in the oldest sample, and the high-molecular-weight bands appeared only in epidermal extracts from middle-aged donors and to a larger extent in those from old-aged donors. B and C: 20 μg of cytosolic protein extracts were subjected to SDS/PAGE on a 12% polyacrylamide gel and electrotransferred onto a nitrocellulose membrane. The blots were processed with anti-HNE (B) or anti-GP (C) polyclonal antibodies at a dilution of 1:2000 or 1:1000, respectively. For a densitometric analysis, the measurements corresponding to the 17-year-old donor were taken as 100%. The intensities of the major bands, reacting with anti-HNE antibodies and ranging from 50 kDa to 80 kDa in molecular weights, were increased to 800% for the 67-year-old donor, and a band of higher molecular weight was only present in the three old samples. The band with a molecular weight ∼80 kDa reacting with anti-GP was increased to 130% for the 50-year-old donor and 160% for the 67-year-old donor, respectively. D: A competitive ELISA was performed as described in the Materials and Methods section to determine the amount of glycated adducts in epidermal extracts. The assay was calibrated by using modified glycated BSA. The increase of glycated adduct is significant (\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(F\ =\ 12.073\) \end{document}, p < .05) between the young (\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(n\ =\ 3\) \end{document}) and the old (\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(n\ =\ 3\) \end{document}) donors. Figure 2. Effect of age on proteasome peptidase activities and content in epidermal extracts. A: The chymotrypsin-like (▴) and PGPH (○) activities of the proteasome were assayed in epidermal extracts by using LLVY-AMC and LLE-NA as fluorogenic peptide substrates, respectively. The enzymatic assays were performed as described in the Materials and Methods section. For the chymotrypsin-like activity, a significant difference (\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(F\ =\ 9.534\) \end{document}, p < .05) was found between the young donors, from 17 to 23 years old (n = 4) and the old donors, from 50 to 67 years old (\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(n\ =\ 5\) \end{document}). A similar trend to an age-related decline was observed for the PGPH activity. B: 20 μg of soluble proteins from epidermal extracts from 17, 50, and 67-year-old donors were subjected to SDS/PAGE on a 12% polyacrylamide gel, electrotransferred onto a nitrocellulose membrane, and immunoblotted with rabbit anti-MCP polyclonal antibodies at a dilution of 1:5000, as described in the Materials and Methods section. Figure 3. Modification of cytosolic proteins in keratinocytes upon serial passages. A: The OxyBlot was achieved with 10 μg of cytosolic protein extracts from passage 1 to passage 4 keratinocyte cultures. Protein extracts were first incubated with 2,4-DNPH followed by SDS/PAGE, electrotransferred, and Western blotted with anti-DNP antibodies as described by the manufacturer. For a densitometric analysis, the measurement corresponding to the first passage was taken as 100%. The intensity of the high-molecular-weight band reacting with anti-DNP antibodies was increased to 180% for passage 4. B: 20 μg of protein cytosolic extracts from passage 1 to passage 4 keratinocyte cultures were subjected to SDS/PAGE, electrotransferred, and Western blotted as described in the Materials and Methods section. The blots were processed with anti-HNE polyclonal antibodies at a dilution of 1:2000. For a densitometric analysis, the measurement corresponding to the first passage was taken as 100%. The intensity of the high-molecular-weight band reacting with anti-HNE antibodies was increased to 620% for passage 4. The experiments, reported in each panel, were done twice with cells obtained from two different cultures. Figure 4. Effect of serial passaging on the proteasome chymotrypsin-like activity and content in keratinocytes in culture. A: The chymotrypsin-like (♦) and PGPH (○) activities of the proteasome were assayed in cytosolic extracts by using 10 μM of LLVY-AMC and 200 μM of LLE-NA as the peptide substrate with 75 μg of total protein as described in the Materials and Methods section. A significant difference in the chymotrypsin-like activity (\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(F\ =\ 18.622\) \end{document}, p < .05) was found between early passages 1 and 2 (\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(n\ =\ 3\) \end{document}) and late ones 4 and 5 (\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(n\ =\ 3\) \end{document}). A similar trend to a decline of activity in serial passages is observed for the PGPH activity. B: 20 μg of cytosolic protein extracts from keratinocytes at passages 1 to 4 were subjected to SDS/PAGE on a 12% polyacrylamide gel, electrotransferred onto a nitrocellulose membrane, and immunoblotted with rabbit anti-MCP polyclonal antibodies at a dilution of 1:5000, as described in the Materials and Methods section. The experiments, reported in each panel, were done twice with cells obtained from two different cultures. Figure 5. Proteasome antigen vs β-galactosidase activity in serial keratinocyte cultures and epidermis donors. Cell suspensions were obtained from the epidermis of various age donors or from keratinocytes in serial culture passages. For each sample from skin A or cell culture B, ELISA tests were performed for MCP, and β-galactosidase was assayed as described in the Materials and Methods section. The two series of result values were plotted on a single graph. In A, the symbols used are ♦, young donors; ○, middle-aged donors; and ▴, old donors. In B, the symbols used are ♦, early passages 1 and 2; ○, intermediate passage 3; and ▴, late passages 4 and 5. The decrease of proteasome antigen as well as the increase of β-galactosidase activity are significant between the young donors from 17 to 23 years old (\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(n\ =\ 4\) \end{document}) and the old donors from 50 to 67 years old (\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(n\ =\ 6\) \end{document}), p < .001 A, and in passages 4 and 5 (\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(n\ =\ 5\) \end{document}) as compared with passages 1 and 2 (\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(n\ =\ 6\) \end{document}), p < .001 B. This work was supported by funds from the MENSR (Institut Universitaire de France and Université Paris 7), the Fondation pour la Recherche Médicale, and the Myers Foundation. 1 Salmon JK, Amstrong CA, Ansel JC, 1994. The skin as an immune organ. West J Med.160:146-152. 2 Grossman D, Leffell DJ, 1997. The molecular basis of non melanoma skin cancer: new understanding. Arch Dermatol.133:1263-1270. 3 Valdimarsson H, Baker BS, Jonsdottir I, Powles A, Fry L, 1995. Psoriasis: a T-cell mediated autoimmune disease induced by streptococcal superantigens?. Immunol Today.16:145-149. 4 Nousari H, Anhalt GJ, 1995. Bullous skin diseases. Curr Opin Immunol.7:844-852. 5 Harman D, 1956. Aging: a theory based on free radical and radiation chemistry. J Gerontol.11:298-300. 6 Berlett BS, Stadtman ER, 1997. Protein oxidation in aging, disease, and oxidative stress. J Biol Chem.272:20313-20316. 7 Beckman KB, Ames BN, 1998. The free radical theory of aging matures. Physiol Rev.78:547-581. 8 Smith CD, Carney JM, Starke-Reed PE, et al. 1991. Excess brain protein oxidation and enzyme dysfunction in normal aging and in Alzheimer disease. Proc Natl Acad Sci USA.88:10540-10543. 9 Carney JM, Starke-Reed PE, Oliver CN, et al. 1991. Reversal of age-related increase in brain protein oxidation, decrease in enzyme activity, and loss in temporal and spatial memory by chronic administration of the spin-trapping compound N-tert-butyl-alpha-phenylnitrone. Proc Natl Acad Sci USA.88:3633-3636. 10 Starke-Reed PE, Oliver CN, 1989. Protein oxidation and proteolysis during aging and oxidative stress. Arch Biochem Biophys.275:559-567. 11 Oliver CN, Ahn BW, Moerman EJ, Goldstein S, Stadtman ER, 1987. Age-related changes in oxidized proteins. J Biol Chem.262:5488-5491. 12 Sohal RS, Agarwal S, Dubey A, Orr WC, 1993. Protein oxidative damage is associated with life expectancy of houseflies. Proc Natl Acad Sci USA.90:7255-7259. 13 Vlassara H, Bucala R, Striker L, 1994. Pathogenic effects of advanced glycosylation: biochemical, biologic, and clinical implications for diabetes and aging. Lab Invest.70:138-151. 14 Dreyfus JC, Kahn A, Schapira F, 1978. Posttranslational modifications of enzymes. Curr Top Cell Regul.14:243-297. 15 Rothstein M, 1977. Recent developments in the age-related alteration of enzymes: a review. Mech Aging Dev.6:241-257. 16 Goto S, Hasegawa A, Nakamoto H, Nakamura A, Takahashi R, Kurochkin IV 1995. Age-associated changes of oxidative modification and turnover of proteins. Cutler RG, Packer L, Bertram J, Mori A, , ed.Oxidative Stress and Aging151-158. Birkhaüser Verlag, Basel. 17 Rivett AJ, 1985. Purification of a liver alkaline protease which degrades oxidatively modified glutamine synthetase: characterization as a high molecular weight cysteine proteinase. J Biol Chem.260:12600-12606. 18 Pacifici RE, Salo DC, Davies KJ, 1989. Macroxyproteinase (M.O.P.) a 670 kDa proteinase that degrades oxidatively denatured proteins in red blood cells. Free Rad Biol Med.7:521-536. 19 Grune T, Reinheckel T, Davies KJ, 1997. Degradation of oxidized proteins in mammalian cells. FASEB J.11:526-534. 20 Coux O, Tanaka K, Goldberg AL, 1996. Structure and functions of the 20S and 26S proteasomes. Annu Rev Biochem.65:801-847. 21 Lowe J, Stock D, Jap B, Zwickl P, Baumeister W, Huber R, 1995. Crystal structure of the 20S proteasome from the archaeon T. acidophilum at 3.4 A resolution. Science.268:533-539. 22 Groll M, Ditzel L, Lowe J, et al. 1997. Structure of 20S proteasome from yeast at 2.4 A resolution. Nature.386:463-471. 23 Conconi M, Szweda LI, Levine RL, Stadtman ER, Friguet B, 1996. Age-related decline of rat liver multicatalytic proteinase activity and protection from oxidative inactivation by heat-shock protein 90. Arch Biochem Biophys.331:232-240. 24 Shibatani T, Nazir M, Ward WF, 1996. Alterations of rat liver 20 S proteasome activities by age and food restriction. J Gerontol Biol Sci.51A:B316-322. 25 Anselmi B, Conconi M, Veyrat-Durebex et al. 1998. Dietary self-selection can compensate an age-related decrease of rat liver 20 S proteasome activity observed with standard diet. J Gerontol Biol Sci.53A:B173-179. 26 Hayashi T, Goto S, 1998. Age-related changes in the 20 S and 26 S proteasome activities in the liver of male F344 rats. Mech Aging Dev.102:55-66. 27 Verbeke P, Perichon M, Borot-Laloi C, Schaeverbeke J, Bakala H, 1997. Accumulation of advanced glycation endproducts in the rat nephron link with circulating AGEs during aging. J Histochem Cytochem.45:1059-1068. 28 Uchida K, Szweda LI, Chae HZ, Stadtman ER, 1993. Immunochemical detection of 4-hydroxynonenal protein adducts in oxidized hepatocytes. Proc Natl Acad Sci USA.90:8742-8746. 29 Rheinwald JG, Green H, 1975. Formation of a keratinizing epithelium in culture by a cloned cell line derived from a teratoma. Cell.6:317-330. 30 Conconi M, Petropoulos I, Emod I, Turlin E, Biville F, Friguet B, 1998. Protection from oxidative inactivation of the 20 S proteasome by heat-shock protein 90. Biochem J.333:407-415. 31 Read MA, Neish AS, Luscinskas FW, Palombella VJ, Maniatis T, Collins T, 1995. The proteasome pathway is required for cytokine-induced endothelial-leukocyte adhesion molecule expression. Immunity.2:493-506. 32 Laemmli UK, 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature.227:680-685. 33 Stoscheck CM, 1990. Enzyme-linked immunosorbent assay for the epidermal growth factor receptor. J Cell Biochem.43:229-241. 34 Madersbacher S, Berger P, 1991. Double wavelength measurement of 3,3′,5,5′-tetramethylbenzidine (TMB) provides a three-fold enhancement of the ELISA measuring range. J Immunol Meth.138:121-124. 35 Grune T, Reinheckel T, Joshi M, Davies JA, 1995. Proteolysis in cultured liver epithelial cells during oxidative stress: role of the multicatalytic proteinase complex, proteasome. J Biol Chem.270:2344-2351. 36 Grune T, Reinheckel T, Davies JA, 1996. Degradation of oxidized proteins in K562 human hematopoietic cells by proteasome. J Biol Chem.271:15504-15509. 37 Hayflick L, 1992. Aging, longevity, and immortality in vitro. Exp Gerontol.27:363-368. 38 Rattan SIS, Stacey GN, 1994. The uses of diploid cell strains in research in aging. Griffiths JB, Doyle A, Newell DG, , ed.Cell and Tissue Culture: Laboratory Procedure2.1-2.12. Wiley, Chichester, UK. 39 Dimri GP, Lee X, Basile G, et al. 1995. A biomarker that identifies senescent cells in culture and in aging skin in vivo. Proc Natl Acad Sci USA.92:9363-9367. 40 Yan LJ, Levine RL, Sohal RS, 1997. Oxidative damage during aging targets mitochondrial aconitase. Proc Natl Acad Sci USA.94:11168-11172. 41 Yan LJ, Sohal RS, 1998. Mitochondrial adenine nucleotide translocase is modified oxidatively during aging. Proc Natl Acad Sci USA.95:12896-12901. 42 Davies KJ, Delsignore ME, 1987. Protein damage and degradation by oxygen radicals. III. Modification of secondary and tertiary structure. J Biol Chem.262:9908-9913. 43 Friguet B, Stadtman ER, Szweda LI, 1994. Modification of glucose-6-phosphate dehydrogenase by 4-hydroxy-2-nonenal: formation of cross-linked protein that inhibits the multicatalytic protease. J Biol Chem.269:21639-21643. 44 Cohn JA, Tsai L, Friguet B, Szweda LI, 1996. Chemical characterization of a protein-4-hydroxy-2-nonenal cross-link: immunochemical detection in mitochondria exposed to oxidative stress. Arch Biochem Biophys.328:158-164. 45 Grandhee SK, Monnier VM, 1991. Mechanism of formation of the Maillard protein cross-link pentosidine. J Biol Chem.18:11649-11653. 46 Wagner BJ, Margolis JW, 1995. Age-dependent association of isolated bovine lens multicatalytic proteinase complex (proteasome) with heat-shock protein 90, an endogenous inhibitor. Arch Biochem Biophys.323:455-462. 47 Maksymowych WP, Ikawa T, Yamaguchi A, et al. 1998. Invasion by Salmonella typhimurium induces increased expression of the LMP, MECL, and PA28 proteasome genes and changes in the peptide repertoire of HLA-B27. Infect Immunol.66:4624-4632. 48 Akiyama K, Kagawa S, Tamura T, et al. 1994. Replacement of proteasome subunits X and Y by LMP7 and LMP2 induced by interferon gamma for acquirement of the functional diversity responsible for antigen processing. FEBS Lett.343:85-88. 49 Dick TP, Nussbaum AK, Deeg M, et al. 1998. Contribution of proteasomal beta-subunits to the cleavage of peptide substrates analyzed with yeast mutants. J Biol Chem.273:25637-25646. 50 Friguet B, Szweda LI, 1997. Inhibition of the multicatalytic proteinase (proteasome) by 4-hydroxy-2-nonenal cross-linked protein. FEBS Lett.405:21-25. 51 Van der Loo B, Fenton MJ, Erusalimsky JD, 1998. Cytochemical detection of a senescence-associated beta-galactosidase in endothelial and smooth muscle cells from human and rabbit blood vessels. Exp Cell Res.15:309-315. 52 Bosmann HB, Gutheil RL, Case KR, 1976. Loss of a critical neutral protease in ageing WI-38 cells. Nature.261:499-501.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0BAFDDE8709B9A764528AAD5B1AEF1DEA250AE6F.txt b/test/dataset/in/resources/corpus/Clean_0BAFDDE8709B9A764528AAD5B1AEF1DEA250AE6F.txt
deleted file mode 100644
index 0128912..0000000
--- a/test/dataset/in/resources/corpus/Clean_0BAFDDE8709B9A764528AAD5B1AEF1DEA250AE6F.txt
+++ /dev/null
@@ -1 +0,0 @@
-jpartjopartJournal of Public Administration Research and Theory1477-98031053-1858Oxford University Press10.1093/jopart/mun016ArticlesLeadership, Service Reform, and Public-Service Networks: The Case of Cancer-Genetics Pilots in the English NHSMartinGraham P.CurrieGraemeFinnRachaelUniversity of NottinghamAddress correspondence to the author at graham.martin@nottingham.ac.uk.1020091382008194769794© The Author 2008. Published by Oxford University Press on behalf of the Journal of Public Administration Research and Theory, Inc. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org2009In attempting to reform public services, governments worldwide have sought to effect change through policies aimed at both transforming structures of public-service provision and facilitating the agency of public servants working within these. Various obstacles have been found, however, to impede the effectiveness of such efforts. In this article, the authors examine the role of organizational networks and distributed leadership—two prominent policies aimed at structure and agency, respectively—in the establishment and consolidation of service reform in the English National Health Service. Using a comparative case-study approach, they contrast the trajectories of two attempts to introduce and gain acceptance for service reform, noting important differences of context, process, and outcome between the sites. The findings indicate the importance of dispersed, as well as distributed, leadership in achieving change in a networked public-service setting. Effective leaders may indeed achieve change through the structures and processes of the network. However, the coexistence alongside the network of other organizational forms constrains the ability of leaders to achieve change without complementary action beyond the boundaries of the network.INTRODUCTIONAcross the economically developed world over the last 10–15 years, governments have pursued increasingly interlaced strategies aimed at reforming public-service provision. Transformative efforts to “Reinvent Government” (Osborne and Gaebler 1992) in the United States and “modernize public services” (Newman 2001) in the United Kingdom are exemplary of policies seeking to improve both the effectiveness and efficiency of public-service delivery. In pursuit of this aim, policies addressing the “structures” of public-service delivery—the way public services are organized—have accompanied policies addressing the “agency” of public servants—the skills and attributes of the individuals employed in public services. Often, these two approaches to public-service reform are seen as mutually dependent. Organizational reforms may give rise to new roles necessitating new skills sets; the broadening of public-service staff's skills may in turn allow them to act more dynamically and autonomously in identifying opportunities for service improvement and acting upon them.This apparent synergy between structural reconfigurations and upskilling individuals as agents of reform, however, is not always so straightforward in practice. Many commentators have noted the contradictions inherent in certain organizational changes within the public services and the inhibitive effect these have on the possibilities for collaboration and joined-up service provision in reformed public services. In particular, there is the well-documented tension between hierarchies, markets, and networks. The collaborative network, as an organizing principle for public-service provision which might bridge gaps between state agencies and assist in addressing social problems that defy the efforts of a single agency (Huxham and Vangen 2000; Jennings and Ewalt 1998; Provan and Sebastian 1998), clashes with traditional top-down hierarchical organizational forms, and with the increasingly prominent role of markets and quasi-markets in national contexts subject to the Anglo-American influence (Currie and Suhomlinova 2006; Ferlie et al. 2003; Hood et al. 2000). In the United Kingdom, a policy focus on networks has coexisted with the reinforcement of a hierarchical performance-management regime and the increasing marketization of public-service provision, mirroring developments elsewhere (Kickert and Koppenjan 1997; Tenbensel 2005).In this article, we empirically examine the degree of convergence (or otherwise) between a particular “structural” reform—the centrally mandated network—and a particular facet of individual “agency”—leadership of change—which have both been central to government reform policies, in the United Kingdom and elsewhere. Networks and leadership have been advocated as means of reinvigorating state-run services, especially in the United States, the United Kingdom, and developed-world countries subject to Anglo-American influences such as Canada, Australia, and New Zealand (Agranoff and McGuire 2001; Hennessey 1998; Kakabadse et al. 2003; Kickert et al. 1997; O'Toole 1997; Shortell et al. 2000). In theory, leadership and networks should complement each other, with the looser organization and less hierarchical ordering logic of the network allowing leadership of change, distributed among network members rather than led from a single organizational apex, to flourish. This would affirm the findings of Denis et al. (1996) on the potential of distributed leadership in complex organizations with plural, ambiguous aims (cf. Bryson and Kelley 1978). In practice, as we set out in the sections that follow, this may not be the case. Moreover, as we explore in the empirical section of this article, in the ever more complex and unstable context of networks in the National Health Service (NHS) in England,1 both the possibilities of distributed leadership and the barriers it faces are increasing. Drawing on a wider comparative study of the implementation of pilot cancer-genetics services in the NHS, we focus on the divergent courses of two case-study sites in which similar service reforms were introduced, highlighting the importance of differences in network organization and leadership style, and both the synergies and incongruities between these. Our analysis suggests that in itself, the structural reform of the network does little to achieve public-service reconfiguration. It does, however, create a “space” within which a certain enactment of leadership might be effective in linking diverse actors across organizational boundaries in pursuit of joint agenda. To the extent that the network competes with other organizational forms and modes of governance, however, this space is a constricted one; consequently, there are limits to the effectiveness of the leaders and their agency, at the points where the network overlaps with other organizational forms (hierarchies and markets), which have their own norms of influence and authority which differ from those which are effective within the network.We continue first by reviewing the policy and academic literature on networks and leadership, before describing the specific empirical field in which our study was located. We then present our methods and findings before discussing the theoretical and policy implications of these.NETWORKS AND LEADERSHIP: THEORY AND POLICY IN THE PUBLIC SERVICESThe increasing attention paid to the role of networks and leadership in public administration has its roots in the waves of reforms to public services over recent decades often grouped under the heading “New Public Management” (Ferlie et al. 1996; Hood and Peters 2004; Peters and Pierre 1998). Though all characterized by efforts to improve the effectiveness and efficiency of public services, these reforms were heterogeneous in nature. The evolution of the approach, exemplified in approaches such as “Reinventing Government” and “modernization” that aim for a more totalizing, transformative effect on public services, has precipitated an increasing need for new organizational forms and new forms of leadership. As Ferlie et al. (1996, 13) have it, this model of New Public Management that emerged from the 1990s onward moves away from hierarchical forms and leadership approaches towards “new management styles, such as management by influence; an increased role for network forms of organization; stress on strategic alliances between organizations as a new organizational form.” Collaborative organizational forms and associated new forms of leadership have gained currency among governments of many developed-world countries (Agranoff and McGuire 2001; Hall and O'Toole 2004; Kickert et al. 1997; Meier 2004; Provan and Milward 1995; Rhodes 1997; Scharpf 1993). In the United Kingdom, the public-service network has been seen as a means of “joining up” service provision even as hierarchy and market have remained, and indeed been strengthened, as modes of governance. Though the notions of the intra- and interorganizational networks originate in the private sector (see Nohria and Eccles 1992; Reed 1992), there is a growing literature on their role in public-service provision (Ferlie and Pettigrew 1996; O'Toole 1997; Peters and Pierre 1998; Provan and Milward 1995; Rashman and Hartley 2002). Private-sector networks are largely explained in terms of reduced transaction costs and interdependence between firms, but for public-service organizations, network-level outcomes for clients are additionally seen as a motivation for integration (Provan and Sebastian 1998). In the literature, moreover, networks are seen as more flexible and adaptable in the face of organizational turbulence, as opposed to hierarchies that orient toward stability (Thompson et al. 1991), though empirical evidence for this is not strong (Bate 2000; Rashman and Hartley 2002).Networks have had a particular appeal for policymakers under “new Labour” governments in the United Kingdom since 1997 as an adaptable organizational form which combined the dynamism of the market with the co-ordination of the hierarchy in delivering the “modernization” of public services. By facilitating joined-up working across professional and organizational boundaries and by assisting the sharing of knowledge, good practice might be translated into rapid service improvement, spread across the public services. As we will see in the next section, though, the reality of the implementation and practice of organizational networks is not quite so straightforward.Alongside this focus on the reconfiguration of organizations and the relationships between them, efforts at public-service reform have also been based on complementary attempts to transform the workforce and its capacities: empowering public-service employees to act intelligently and autonomously within the enabling structures of networked organizations to effect change. The “Reinventing Government” initiative of the Clinton-led administration of the United States is exemplary of the expectations placed on this synergy, as a means of “[changing] the culture of our national bureaucracy away from complacency and entitlement toward initiative and empowerment.”2 In the United Kingdom, new Labour's ambitions for “modernization” similarly envisaged a public-service workforce empowered to act intelligently and autonomously to effect change within the new organizational forms (Currie et al. 2005).Within such policies, “leadership” has been an ongoing emphasis. Research on leadership in the private sector has tended to give precedence to transformational leadership, embodied in a single, heroic figure at the apex of an organization, as both a subject of study and a prescription for effectiveness (Bryman 1992; 1999). Such norms and ideals of leadership have to some extent permeated public-service leadership policy (Currie et al. 2005), but there is recognition also (among academics and policymakers) of the limits to this model in the more pluralistic and political context of the public services. For Heifetz (1994), leadership theory premised on the idea of a single inspirational leader is inapplicable to the public services, and attempts to conceptualize public-service leadership in this way serve only to perpetuate this myth. Rather, leadership is about “adaptive work” which requires buy-in and enactment by individuals at all levels of an organization. In the policy and practice of leadership in the public services, then, there is an onus on “distributed” leadership, “from apex to front line” (Hartley and Allison 2000, 39), even as power and responsibility are formally concentrated in the roles of the heads of public-service organizations. Arguably, the need for distributed leadership is all the more pronounced in public-service networks since “the notion of a leader with a hierarchical relationship to followers does not apply in collaborations, so the potential for exercising ‘decisive leverage’ by virtue of a formal position is reduced” (Huxham and Vangen 2000, 1167; cf. O'Toole 1997).Although the need for an understanding of leadership as distributed within public-service organizations, and especially networks, is accepted, exactly what constitutes distributed leadership is less obvious, and the subject of ongoing clarification in the literature. Van Wart (2005) offers a typology of distributed leadership, variously emphasizing the deliberate delegation of leadership from the top-down through “substitutes for leadership” and “superleadership,” and the more “organic” growth of “self-leadership” and “self-managed teams.” In all cases, though, shared leadership ultimately requires both a willingness to cede leadership to others on the part of organizational heads and the capacity of other actors to take it on (van Wart 2005, 372–3). As Crosby and Bryson (2005, 29) suggest, “potential for effective leadership lies alike with those who do and do not have formal positions of power and authority. Indeed, this view of leadership may be most useful in reminding those with little authority how powerful they can be through collaboration […] and in reminding those in a supposedly powerful position just how much they rely on numerous stakeholders for any real power they have.”Thus, even if we accept Bennett et al.’s (2003, 7) notion of distributed leadership as “the product of concertive or conjoint activity, […] an emergent property of a group or network,” it is important to recognize that such emergence occurs within the boundaries set by other parties, including those in whom leadership is traditionally concentrated. Furthermore, distribution of leadership emphatically does not imply equal spread of leadership among all parties. Power and influence may remain concentrated in certain “nodes” of leadership, who “lead up and out rather than down” (Crosby and Bryson 2005, 32). This is, indeed, the finding of Vangen and Huxham (2003), who find that the leadership qualities required by network managers are as much about “thuggery”—manipulation and politicking—as about facilitation and mobilization. These contours of distributed leadership—in particular, its relation to traditional top-down authority and its potential concentration in new nodes of power—are important to acknowledge in an understanding of the role of leadership in networks. As with transformational leadership, distributed leadership is seen in the literature as having the potential to improve organizational performance (Denis et al. 1996; 2000; Pearce and Conger 2003; Rainey and Steinbauer 1999).Government policy in England has emphasized the centrality of leadership in the modernization of all fields of public services (Cabinet Office 1999). Calls for individualistic leadership from the top and distribution of leadership coexist in government policy, so that in health, for example, programs supporting transformational leadership from organizational heads are accompanied by efforts to empower staff and “lead change through people.”3Together, structural reforms such as the introduction of networks and reforms aimed at harnessing the agency of staff, such as leadership development, are seen by policymakers as having the potential to contribute positively to reforms in the United States, the United Kingdom, and elsewhere. Furthermore, they are seen not just as complementary but to some extent as mutually dependent. Without equipping and empowering staff with the leadership and other skills to inseminate, manage and adapt to change, organizational reforms to public services are bound to fail. To this extent, then, the logic of these approaches to public service reform is evident. As we see in the next section, however, the practice is not so clear-cut.NETWORKS AND LEADERSHIP: OBSTACLES AND IMPEDIMENTSIn implementing networks in the public services, certain practical difficulties come to light. First, and most obviously, the resilience of the hierarchy and the increasing role of the market in many national contexts mean that networks have coexisted alongside two other organizational forms, with certain consequent tensions. Top-down performance management and quasi-market–based competition have the potential to drive a wedge between individuals and organizations in a way that is inimical to the collaborative ideals of the network (Currie and Suhomlinova 2006; Ferlie and Pettigrew 1996). To the extent that their performance is managed according to their core duties and judged in competition with their colleagues, professionals will tend to orient toward their own silos rather than reach out to foster collaborative relationships which may not benefit them. The same is true of organizations (Ferlie et al. 2003; Hood et al. 2000; McNulty and Ferlie 2002) and is reflected in various national contexts where network-based reforms coexist with other organizational forms, such as Australia (Ryan and Walsh 2004), the Scandinavian countries (Christensen et al. 2007), the Netherlands (Brandsen and van Hout 2006; Kickert et al. 1997), and the United States (Provan et al. 2004). Competition between organizations in contexts governed by market or quasi-market relations, as exhibited, for example, in managed care networks in the United States, will have a similarly divisive effect (Shortell et al. 2000), with the risk that agencies relapse into relationships based on contracting rather than collaboration (Brandsen and van Hout 2006; Keast and Brown 2006).Furthermore, many public-service fields, such as health and education, are professional bureaucracies in which these kinds of difficulties are compounded (Riccucci 2005). Each profession has its own pervasive socialization, which might work against a network orientation in a number of ways: a professional knowledge domain that is esoteric, to some extent tacit, and difficult to share (Currie and Suhomlinova 2006); a career pathway that tends to be based on specialization rather than broadening of knowledge (Nancarrow and Borthwick 2005); and a place in a well established and institutionalized hierarchy of professions and subprofessions (Bate 2000). Consequently, professions have long been acknowledged as averse to managerial interventions (e.g., Lipsky 1980; Riccucci 2005). Notably for our empirical field, physicians are often characterized as the quintessential autonomous profession (e.g., Riccucci 2005) and tend to be individualistic in their practice and attitude toward integration (Shortell et al. 2000). Furthermore, the status, power, and expertise of the medical profession mean that any set of organizational reforms requires the compliance of physicians for its effectiveness (Greco and Eisenberg 1993; Shortell et al. 1998). This is arguably all the more so in the NHS, where in contrast to the relative erosion of medical power and autonomy in the United States (Hafferty and Light 1995), physicians have retained an authority that gives them power of veto over NHS administrators in regard of any organizational reconfigurations that they do not welcome (Bate 2000; Ferlie and Pettigrew 1996; Harrison et al. 1992).Moreover, there is the risk that, in insisting that organizations adopt a more networked form, policymakers undermine the very essence that makes the network a functional organizational form in the first place. If they rely for their effectiveness and dynamism upon their voluntaristic, informal, and organic nature, then efforts to impose networks may have perverse consequences. In particular, as Kickert and Koppenjan (1997, 39) point out, in networks, “government is not the single dominant actor that can unilaterally impose its will. Hierarchicial, central top-down steering does not work in networks.” Rather, governance mechanisms must fit within the network structure (Bruijn and Heuvelhof 1997). Yet the ongoing influence of top-down performance management regimes, so strong in countries such as the United Kingdom where vertical accountability remains characteristic of public services, may militate against this need, to the extent that the network may become little more than an instrument of surveillance and performance management, to the detriment of collaborative efforts by network managers and members (Addicott et al. 2007).The necessity of more lateral means of governing and guiding networks toward common aims creates particular challenges for the leader within the public-service network. The widely noted ambiguity of objectives, power and accountability that characterizes the public sector as compared to the private sector (Denis et al. 1996; 2000; Heifetz 1994) are compounded, as we have seen, by the network. This suggests a need for the distribution of leadership across individuals in collaborating organizations. However, organizational objectives are frequently set by stakeholders “external” to networks, such as policymakers. This complicates any such effort to pluralize change agency and distribute leadership since actors across networks will be constrained in their actions by the parameters set by these external stakeholders (Currie and Lockett 2007), which may not encourage collaborative activity. Furthermore, the hierarchical order of professions—in health, for example, the ongoing subservience of nursing to medicine, at least in the view of many physicians, and the endurance of intraprofessional hierarchies and power relationships (e.g., between surgery and medicine) (Currie and Suhomlinova 2006)—is also likely to, at best, complicate any effort to distribute leadership beyond traditional leaders.NETWORKS AND LEADERSHIP: TRANSFORMING STRUCTURE AND AGENCY?The preceding section indicates the difficulties of establishing both networks and corresponding modes of leadership in public services characterized by the coexistence of other organizational forms and various tensions between professions and policy aims. Ostensibly, distributed leadership and networked organizational forms seem well suited to each other, but the complications of the public-services context mean that their effectiveness and synergy in effecting service reform are compromised. Equally, however, each policy might have the potential to surmount the difficulties presented by the contingencies of the public services by reconfiguring power relationships to foster a more dynamic organization than one governed by hierarchy or competition.What is suggested in particular is the need for an enactment of leadership that seeks to engage a multiplicity of stakeholders, facilitating and consolidating change by concentrating on processes and outcomes together (Huxham and Vangen 2000; Vangen and Huxham 2003). Even as this suggests a distribution of leadership, it also implies personal investment by those charged with network leadership tasks: as Huxham and Vangen (2000, 1171) have it, “it is paradoxical that the single-mindedness of leaders appears to be central to collaborative success.” At the same time, though, in a professional bureaucracy characterized by looseness of accountability and practitioner autonomy, there is a need to engage professionals for organizational change to be achieved and consolidated (Riccucci 2005; Shortell et al. 2000). Any change needs to ensure that “organizational players are empowered enough to be able to avoid the frustration and chaos that leads to the reassertion of pervious forms and processes” (Ferlie et al. 1996, 14).In other words, the “structural change” of the network is reliant on the “agency” of leaders—distributed or otherwise—in making it functional since “formal bureaucratic structure, as defined by rules and regulations, has little impact on the behaviors of front-line workers” (Riccucci 2005, 64). For all the collaborative rhetoric of the network form, then, it will have little tangible impact on practice in the absence of effective leadership. In making this rhetoric reality, leaders can, however, draw on the resources of the network: in the language of Huxham and Vangen (2000, 1171), the “leadership media” of “structures, processes and participants.” It is through the properties of the network that leadership might be enacted, and although these may well be outside the control of any one leader, they are also to some extent malleable. Structures, for example, “normally emerge out of the practical reality of the tasks that they tackle” (Huxham and Vangen 2000, 1167), so that even within an overarching organizational structure, certain “substructures” might be assembled in pursuit of the leader's task. Similarly, O'Toole (1997, 48) suggests to network managers the possibilities of a number of routes to influence: “find ways to shift network membership toward more supportive coalitions; locate key allies at crucial nodes; try to alter agreements between the parties to heighten program salience; and buffer well-functioning arrays to limit uncertainty and complexity.” Crosby and Bryson (2005) indicate the potential for influence in networks through “forums,” “arenas,” and “courts,” each subject to their own rules of conduct and therefore amenable to particular enactments of leadership. These are the concrete opportunities within interorganizational collaborations for leaders to be effective, for example, through visionary leadership, “creating and communicating shared meaning” in formal and informal network forums that inseminate ideas, and political leadership, “making and implementing decisions” in the legislative and executive arenas that determine formal policy (Crosby and Bryson 2005, 35).Although the nature of the structures, processes, and participants of the network will vary, and with them the opportunities for influence, what these ideas do encompass is an emergent image of what the effective leader within a network might look like, in terms of skills, style, and personality. The remainder of this article considers this question, as well as the more general question of the interaction between the structural reform of the network and the agency of leaders in achieving change. In the next section, we introduce the specific field in which we conducted our fieldwork and describe the service reform that the pilots studied attempted to introduce into this context.CANCER NETWORKS IN THE NHS, MODERNIZATION, AND SERVICE REFORMAlthough in the United Kingdom networked public services are often characterized by commentators as being a particular feature of policies of new Labour governments since 1997, the form has earlier roots. Reflecting wider trends for networked organization in the policies of various developed-world countries, UK public services were already moving towards networked governance in the early 1990s under the previous Conservative administration (Newman 2001). To this extent, networks can be seen as a logical development of the New Public Management policies prevalent across developed-world governments, especially those subject to Anglo-American influences (Ferlie et al. 1996; Kickert et al. 1997; O'Toole 1997). In the United Kingdom, health policy was an early adopter of the network principle, with “managed clinical networks” introduced in a variety of specialties, including cancer, in the NHS from the mid-1990s, mirroring trends toward networked health-service delivery in Europe (Wijngaarden et al. 2006) and the United States (Shortell et al. 1994).Though funded from general taxation and offering universal access to British citizens, the NHS is far from a monolithic public-service organization. In common with other fields of public-service provision, the introduction of the internal market to the NHS in the early 1990s created an increasingly fragmented organizational scene. From a system based on geographically determined organizational divisions, the NHS was subdivided into purchaser and provider “trusts,” each effectively an organization in its own right, with autonomy over its human resources policies, contracting arrangements, and finances, and the potential (in theory) for bankruptcy (though the political unacceptability of hospital closures ensured that in practice this was rarely threatened, with mergers and takeovers a more palatable alternative). As with public-service networks elsewhere in Europe and the world (Kickert et al. 1997), then, the network represented one means of mediating such organizational fragmentation: effectively, managed clinical networks in the NHS were to be interorganizational in character (between autonomous NHS trusts), albeit technically within a single overarching organization, the NHS.In relation to cancer, the 1995 Calman-Hine report envisaged new service structures in England and Wales “based on a network of expertise in cancer care reaching from primary care through Cancer Units in district hospitals [acute secondary-care hospitals] to Cancer Centres [specialist tertiary-care facilities]” (Department of Health 1995, 7). The vision here was of a network that would coexist with the emerging internal market in healthcare by encouraging competing providers to work together with purchasers to reduce duplication of services and promote clearer patient pathways. Good practice was to spread across the network through leadership from the medical lead in collaboration with colleagues across the network. In the interests of better clinical outcomes, though, concentration of expertise was also on the agenda, with cancer units needing a critical mass of patient throughput to maintain surgical subspecialization, and expertise in rarer cancers and specialist diagnostic and treatment processes to be concentrated in cancer centers. Consequently, although they were intended to be collaborative ventures between the different partners in the networks, “big players” in the networks—hospitals with much to gain or lose from the concentration of resources—have in many places come to dominate their agenda over smaller hospitals and primary care organizations. From the start, cancer networks were intended to promote both service rationalization and collaboration—with all the potential tensions this could create. However, as Addicott et al. (2007, 95) point out, in recent years the role of cancer networks has been significantly managerialized, becoming “largely focused on the structural configuration of services, performance targets and workforce planning.” In four of the five cancer networks they study, a preoccupation with meeting top-down performance-management agenda meant that collaboration was limited.Despite their government-driven mandate, cancer networks mirrored the function of two rather different network forms in the United States. First, they reflected the role of integrated networks of health-care providers that emerged in response to the market pressures created by the advent of managed care, as a means of rationalizing services and offering cost-effectiveness through economies of scale (Shortell et al. 1994). Second—and as Addicott et al. (2007) point out, in tension with the first role—they also serve a function similar to that of the community clinical oncology programs and special populations networks funded by the National Cancer Institute, with their aim of fostering knowledge-sharing relationships between research institutions and community-based intervention, education, and research projects (Baquet et al. 2005; McKinney et al. 1993).This was the context faced by the pilot program of cancer-genetics services cosponsored by the Department of Health and Macmillan Cancer Support,4 which sought to improve provision in this field by introducing a structured care pathway for those with a suspected family history of cancer. The Harper report (Department of Health 1998) identified inequalities, inefficiencies, and poor practice in cancer-genetics provision and suggested a division of labor and flow of knowledge between primary care, cancer units, and cancer centers which mirrored the set-up of cancer networks. This would result in smoother patient experiences and the elimination of poor practice, such as inaccurate risk assessments resulting in unnecessary screening (if genetic risk is overestimated) or inappropriate reassurance (if underestimated). An expert group convened by Macmillan and the Department of Health operationalized this in a framework known as the “Kenilworth model,” and in turn, this was adapted to local conditions and piloted in seven locations across England. With endorsement from an expert committee comprising policymakers, hospital clinicians, family physicians (GPs), nurses, service users, and others, it had credibility and a degree of “top-down push,” though it was not mandated; rather, the expectation was that each pilot would shape the care pathway in accordance with the particularities of its own locality.In principle, then, the Kenilworth model represented just the kind of improvement to and rationalization of service organization that cancer networks were designed to help to achieve and reproduce. In terms of innovation diffusion, the overall setting augured well. With the change agency of Macmillan and the Department of Health behind it, the innovation of the Kenilworth model seemed to have some leverage behind it and addressed an identifiable problem (Rogers 2003). Given the preceding discussion, however, there are clear obstacles to the operation of networked collaboration and to leadership that might confound uptake and establishment of the new care pathway. In the remainder of this article, we compare the trajectories of two of the Kenilworth pilots, considering the roles played by leadership and by the cancer networks in which they were located in their course. The degree of success of the two pilots was divergent, as assessed in terms of two key outcomes: throughput of patients, and whether the model was sustained beyond the pilot period. “Derton” had a throughput that was several times that of “Nottley”; the project (and therefore the Kenilworth-based care pathway) in Derton was sustained with local money following the pilot period, whereas Nottley's project and pathway were not.5METHODSThese two cases were among the 11 that were selected as case studies for a wider evaluation of a program of pilot genetics services in the NHS. This program encompassed 28 government-funded pilots working in various clinical areas, including the seven cancer-genetics pilots working to implement the Kenilworth model in different localities. The pilots followed a genetics white paper (Secretary of State for Health 2003), which set out the aim of “mainstreaming” genetics knowledge and practice in everyday NHS provision. The pilots were to be one means of achieving this ambition.The brief of the evaluation was to identify the barriers and facilitators faced by the pilots in realizing this aim. In pursuit of this, a comparative case-study approach was adopted, using qualitative methods to facilitate intra- and intercase analysis to illuminate key differences of context and mechanism that give rise to differences of outcome. This enables generalization through process analysis and the development of theory (Eisenhardt 1989; Yin 2003). More generally, where quantitative approaches permit the statistically based identification of apparent correlations between cause and effect, qualitative methods such as these are particularly useful in elucidating the mechanisms which give rise to such correlations (Lee 1999; Silverman 2004). As such, they are well suited to the study of “processes” within organizations (Langley 1999; Pettigrew 1997): in this case, the establishment of service reform within a network and the enactment of leadership in facilitating this. Our study thus addresses the gap identified by Bryman (2004) in relation to research on leadership, by adopting a qualitative approach which does not simply attempt to replicate quantitative research in identifying the key variables on which effectiveness rests, but rather focuses on the mechanisms in given contexts (such as the network) that give rise to certain outcomes.Our selection of 11 case studies from 28 funded pilots was theoretically guided (Eisenhardt 1989). The intention was to select a sample that was representative of the variation present among the 28 pilots in terms of a number of variables. These included clinical field, host organization (primary care organization [PCT], “mainstream” clinical department in a hospital, specialist genetics center), profession of pilot lead (medical geneticist, mainstream physician, GP, nurse, manager), characteristics of the area served (urban/rural, socioeconomic profile, ethnic profile), and so on. Potential cases were assessed by members of the research team on the basis of documentary materials provided by the pilots and preliminary interviews with pilot leads in every site. Following this, the research team selected its sample of 11, which included four (of a total of seven) cancer-genetics pilots aiming to implement the Kenilworth model. The focus of this article is on the pilots in Derton and Nottley, which were both based in hospitals and served areas covered by their cancer networks. The other two cancer-genetics cases were considerably smaller in scale, each serving the population of the single PCTs which hosted them and therefore engaging less with the wider cancer networks. Whereas the issues relating to cancer networks in this article are specific to Derton and Nottley—since no other pilot among the 11 cases was embedded within a network in this way—other findings presented (e.g., the range of leadership behaviors exhibited) have much in common with the wider findings in the other case-study sites.In each case-study site, our research included a number of qualitative methods: in-depth interviews with pilot stakeholders, participant, and non-participant observation of meetings at pilot and program level, and documentary analysis. The interviews covered a number of issues, grouped under the following headings: collaborative working, information technology, human resources management, leadership, user involvement, knowledge management, organizational culture, and policy. Although the interview schedule included a number of possible questions under each heading, in line with in-depth interview methodology, the interviewers tended to use these as prompts only, deviating from the list of questions to discuss issues as they were relevant to respondents and their organizational contexts, but seeking to frame questions as openly as possible so as not to “lead” interviewees (Silverman 2004). In the course of observational research, we paid particular attention to exchanges of knowledge between involved parties and barriers identified in the course of meetings to collaboration and service reform in their “mainstreaming genetics” ambitions. Documentary analysis encompassed bid documents, reports to sponsors and evaluation reports, and focused on the same issues as those explored through interviews and observation, with particular attention paid to change through time.Some 88 interviews have been conducted across the 11 sites, including 12 each in Derton and Nottley. Interviews were tape recorded and transcribed in their entirety and lasted between 40 min and 3 hours. In Derton and Nottley, similar respondents were interviewed, reflecting the similar aims and organizational contexts of the two pilots. They included pilot leads (medical geneticists in both cases), nurses and genetic counselors involved in delivering the pilots, primary-care–based practitioners such as GPs, managers, and clinicians within the cancer networks and the host hospitals, and service users involved as members of the projects’ steering groups. Data drawn from observation and documentary analysis activities both informed these interviews and complemented them by illuminating the approaches taken to establishing service reform within wider networks and the barriers encountered in attempting this.We undertook an iterative analysis process, rereading and coding transcripts, notes and documents, generating themes, and cross-checking these through discussions between authors. Thematically related parts of the embedded analysis in each data source were grouped together. Each of the three authors had engaged in fieldwork, and all three engaged in analysis. In Derton, GPM developed an initial analysis that was checked by GC; in Nottley, RF developed an initial analysis that was checked by GPM. We discussed the coding of transcripts, document, and meeting notes with each other, ensuring inter-researcher reliability of interpretation and enhancing analysis. Subsequently, the analysis agreed across the authorial team for each case was considered against the over-arching research questions. As a means of triangulating this analysis, findings were presented to both the commissioners of and participants in the research, providing a check on the authenticity of our analysis (Yin 2003).In presenting qualitative analysis, a number of approaches are possible, each seeking in its way to produce a summary that is both comprehensible and explicitly derived from primary data and thus amenable to critical appraisal by the reader (Eisenhardt 1989). Our approach in this article involves presenting the narratives from the two cases examined in turn: in-depth descriptions that deploy the primary data to illustrate the dimensions of leadership in the establishment of service reform they emerged in the cancer networks of Derton and Nottley, which reflect many of the key findings from the other nine case-study sites.RESULTS: DERTONThis pilot was led by a specialist geneticist, but he was keen to ensure that the service was delivered through and “owned” by the cancer network, which covered three acute hospitals (comprising two cancer units and one cancer center) and 10 PCTs. Genetics provision here had previously been delivered from another hospital, some 40 miles from the cancer center: consequently, patients and clinicians in the network had seen genetics as a remote specialty, its perceived esoteric knowledge compounded by its geographical distance. Recently, however, a new genetics “satellite unit” had been set up in the hospital that hosted the cancer center, and the geneticist leading the project had been appointed to a post here with a view to improving genetics coverage in this underserved area. The prior inaccessibility of cancer-genetics expertise had led to the setting up of a plethora of local protocols within the network for dealing with suspected inherited cases of cancer.This variation, together with the arrival of the genetics satellite unit 2 years earlier, had created the political will within the cancer network for a more coherent way of dealing with suspected family histories of cancer and an opportunity to do something about it. A colorectal surgeon (and, notably, the current medical director of the cancer network) described the dissatisfaction with the incumbent system:The ones with the really high risk we would have sent to [the genetics service], obviously. But there's a lot of them who didn't fit the high-risk criteria, for example for bowel cancer, and we give them advice which—you're not quite, 100 percent sure whether it is scientifically based.Colorectal surgeonClinicians in cancer-related specialties outside genetics had taken on responsibility for genetic risk assessment. But they encountered criteria for assessing genetic risk in breast, bowel, and ovarian cancers which were far from universally agreed upon and limited access to specialist clinical knowledge which might assist adjudication. Here, then, was a potentially receptive structure in terms of the attitudes of the network's members, with political will exemplified by the attitude of the colorectal surgeon and medical director of the network quoted above.Having taken up responsibility for genetic risk assessment reluctantly, existing services had nevertheless grown into an important part of cancer-service provision locally. Growing patient numbers and dissatisfaction over the variation in protocols saw the issue rise in importance within the cancer network, for practitioners and managers alike. Increasing numbers of patients were being admitted to screening services, without a “proper” (i.e., universally agreed) system of weighing the benefits and risks that might accrue from this radiation exposure. Moreover, this was without the explicit agreement of the primary-care purchasers who paid for the service, and this was a concern for the hospitals providing the service, who feared that in the absence of an agreed contract, funding might be withdrawn. As with the attitudes of clinicians, the organizational context too, then, was potentially receptive to the kind of change in care pathways augured by the Kenilworth model.Consequently—and in contrast to the experience of Nottley below, and indeed of other case-study sites not detailed here—the cancer network was proactive in seeking to establish the pilot as an integral part of local cancer care pathways, as a means of solving these issues. The pilot's lead clinician, meanwhile, was acutely aware of the need to ensure ownership by the network, given his status as an “outsider” associated with the distant genetics service. He sought to ensure “ownership” of the service by the network and adopted a highly dialogical and diplomatic stance in putting his ideas to colleagues within the network:Lead: [Previously] there was no continuity, and no discussion between genetics and the surgeons. I think if any lesson's come out of this project, it's that this sort of very detailed day-to-day discussion underpins a successful service. If you don't have that networking, then you can't sustain a service.GPM: The discussion between geneticists and their referrers?Lead: Yes, the stakeholders, the ability to sit down with somebody and explain exactly what you want to do, and for them to be able to pick up the phone and say, ‘I've got this patient: what do you think?’, any different niggles that happen. If you've got someone who's not happy, it's easy to go and see them or pick up the phone: it's just simple human communication.Lead (clinical geneticist)Through these kinds of “forums”, in the lexicon of Crosby and Bryson (2005), the clinical lead provided a vision in a politically sensitive way. These dialogical, quiet leadership skills had some success in paving the way for the service reform, as confirmed by the geneticist's new colleagues:It sounds a bit like I'm blowing [his] trumpet, but seriously, he was a knight in shining armor, and was really swamped with demands from people like me, who wanted to have this pathway set up, like, yesterday. It was very efficient very quickly. And [he] is a particularly good public-relations individual: his ability to make complicated genetics understandable is almost unique.Gynecological oncologistIn persuading physicians in these forums, then, a “quiet” approach to leadership was important, alongside firm, quantitative evidence of the potential benefits of the service reform in terms of patient outcomes and conforming to national guidance on risk assessment: “speaking their language,” as the pilot lead put it. In the terminology by Crosby and Bryson (2005), this was about visionary leadership using forums adroitly in order to create a compelling, shared vision of service reform and utilizing existing “norms of relevance and pragmatic communication, media and modes of argument” (Crosby and Bryson 2005, 120) to good effect.Credibility with physicians, through quiet leadership and the language of clinical audit, was however only one aspect of consolidating the place of the service within the network. Equally important was winning over the nurses who had taken on responsibility for risk assessment under the extant system, on whom the pilot relied for much of its throughput. This rather different professional group necessitated a different approach, requiring that the entire pilot team be versed in the arts of quiet leadership and negotiation. So one of the pilot's nurses took on the task of delicately persuading other nurses about the improvements that the new system might bring, deploying her understanding of “nursing culture,” as she put it.Thus, the pilot team members adapted their leadership to the different constituencies to be persuaded. “Objective evidence” was a key means of convincing medical specialists. Nurses, as the delivers of the service, were generally more concerned about the effects of the new model on the patient experience. Persuading each group, though, was crucial to the pilot's success, and each group could be engaged through its own informal and formal forums, with their own unwritten rules of engagement. And once key physicians and nurses elsewhere in the network had been convinced of the worth of the new system, they themselves were likely to spread acceptance among their colleagues. A breast-care nurse specialist in one of the cancer units, who had previously taken on much of the local risk-assessment work, explained how her colleagues had had concerns that the new care pathway was “just a way of getting people out of the system.” Following her own discussions with the lead nurse on the pilot, she was now convinced of its worth and had been instrumental in persuading other nurses in the unit, acting as a “link” between the pilot and her colleagues. In this way, the initial visionary work of the pilot team members as leaders gave rise to a more diffuse leadership, in the terms of Huxham and Vangen (2000) enacted through the structures and processes of the network as well as embodied in the wider participants beyond the core leaders from the pilot team.The success of this process also rested on the effective structure of the cancer network, recognized by respondents as being functional, democratic, and consensual, at least compared with others.6 Prior to the start of the pilot, the network had convened a “visioning event” attended by stakeholders (physicians, nurses, managers, patients, others) from every hospital and PCT in the network, and this was seen to have promoted the sense of “network ownership” so desired by the pilot lead. This, then, had provided an important forum for the development of a collective vision. More generally, the pilot lead highlighted the role of the network in establishing the service across the cancer units, citing the leadership of one network manager in particular:When I met her, I thought, ‘Goodness, this woman is so networked!’ She's been a senior nurse for years and she knows everyone. She knows the system: she's got an amazing knowledge of how to exert mild pressures here and there to achieve different aims.Lead (clinical geneticist)Alongside the receptive structural context provided by the network and its forums, and the delicate, visionary leadership work of the pilot team, the pilot's establishment also rested on the slightly more directive leadership enacted by this manager. Occasionally, individual clinicians were reluctant to accept the implications of the pilot for their own practice, despite the efforts of pilot staff through quiet leadership and audit evidence. On such occasions, the consensus of the wider network, once achieved, could be deployed as a means of persuasion in its own right, as the manager explained:The objectivity of the network is useful to say, ‘You're not supposed to be doing that kind of work on patients’: we bring intelligence from elsewhere in the network. […] Often just by bringing that objectivity, it can be made to work, by bringing in other clinical views.Cancer-network managerIn these ways, the collaborative rhetoric of the notional “network” was made real. Where collegiate, professional leadership through the network's structures and processes did not succeed, a certain amount of more directive, manipulative, even “thuggish” (Vangen and Huxham 2003) leadership could fill in, using this “objectivity” of the network as a tool of persuasion of more reluctant actors.The network itself, then, and the shrewd and varied leadership tactics of various agents within it, was effective at gaining sign-up willingly or reluctantly from a range of hospital practitioners. Engaging with primary-care practitioners, though, was less straightforward. Crucially, primary-care stakeholders were not such central actors in the network, with its mandated focus on service rationalization. The pilot had held various educational and publicity events for GPs and community nurses, but ensuring longer-term engagement was a challenge. Notably, there were fewer forums at which primary-care actors might be engaged; furthermore, there were fewer obvious common interests between the pilots and GPs, on which visionary leadership might draw. To this extent, primary-care practitioners were on the fringes of the network, both in terms of its “infrastructure” and in terms of the commonalities of interest on which it was built.Similar difficulties were present in the pilot's efforts to obtain ongoing funding, which involved dealing with purchasers who were also based in primary care. Though it had sign-up “in principle” from purchasers from the start, making post-pilot financial arrangements was not straightforward. Notably, although formally included in the network, those in charge of purchasing decisions in primary care were also responsible for funding a range of services beyond cancer. Consequently, negotiations about ongoing funding were subject to a range of considerations emanating from outside the cancer network. In the terms of Crosby and Bryson (2005), this “arena” was one that was subject to a range of different authorities and influences, including not only competing services but also centrally determined guidance on funding priorities. Furthermore, reorganization and management changes within primary care meant that not all purchasers were willing to follow through on their original commitment to the pilot, and indeed, only about half the PCTs in the network eventually funded the service and care pathway on an ongoing basis. This partial success highlights the fact that the structure of the network, and the styles of leadership this facilitates, is limited where it meets other principles of organization and governance, a theme which we will take up in the discussion. And as we discover in the next section, this space of operation can be even more constricted by the overlapping imperatives of other organizational forms.RESULTS: NOTTLEYThis was a joint pilot led by two geneticists at neighboring hospitals, part of a large cancer network containing six hospital trusts and covering a total population of around two million (twice that of the network in Derton). The pilot involved two, largely separate, projects aimed at increasing the accessibility of cancer genetics in primary care. Here we focus on one of these—which used advertising in GPs’ surgeries to attract self-referrals from concerned individuals, especially those from “hard-to-reach” ethnic-minority backgrounds—though many of the issues we raise apply to both. In common with Derton, both projects followed the core principles of the Kenilworth model to rationalize care pathways. Given the primary-care focus, however, the principal constituencies to whom they needed to appeal were quite different. The catchment of each project was also smaller than in Derton, partly because of concerns about the risk of overwhelming the genetics service with new referrals by “opening the door” in primary care.As in Derton, the pilot had obtained sign-up from the cancer network, but here the network was somewhat less proactive in encouraging and spreading the work of the two projects. There was a sense that the cancer network was preoccupied with conforming to central-government targets and with the need to find funding for existing services from cash-strapped purchasers:Because it hasn't got a definitive work stream or target attached to it, no-one has said to me in the three-and-a-half years I've been here, ‘Account for what you're doing on cancer genetics’—and they ask me that about lots of other things. I've got to have action plans on a million things, but cancer genetics isn't one of them. […] If I go to the [purchasers] with any service development, unless it is linked to a target or is completely unavoidable, they just say, ‘You must be mad.’ What they say—which is true—is: ‘We haven't got the money to run the services we've got: why would we fund new services?’Cancer network managerCompared to Derton, the network seemed much more governed by hierarchical principles—especially conformity to centrally determined priorities—and this was reflected in the nature of the leadership enacted by this network manager and others.The pilot thus sought to forge its own relationships with primary-care practitioners and the target communities, but with mixed success. The project successfully negotiated space in the premises of a number of GP practices to carry out its surgeries and to display advertising materials. But contact with GPs went little further than this, due to difficulties similar to those experienced by the pilot staff in Derton, around the lack of buy-in from primary-care professionals. So pilot staff decided to concentrate on self-referrals, exempting GPs from the care pathway and avoiding the need to bring them onboard. However, engaging the target communities in the pilot also proved problematic, and so efforts to increase uptake from minority-ethnic groups had only marginal success:Administrator: It's always been the aim but it hasn't necessarily come to the outcome we would like.RF: In terms of?Administrator: Because gaining access to some ethnic minorities is very difficult. I still think there's a lot of understanding yet to be achieved as to why we don't get these people coming forward.With little experience of such community-engagement strategies, inseminating interest among the public was problematic.Efforts, then, to engage the key stakeholders—potential patients and GPs—were marginal, by accident and design, respectively. This was in contrast to Derton, where the cancer network enabled the pilot to gain buy-in from clinicians “within hospitals” whose direct importance to the service (as referrers) matched their indirect influence (in creating a critical mass of consensus and support among the powerful constituents of the network).7 An absence of common interest was exacerbated by the absence of the opportunities for effective leadership that the structures and processes of the network might provide. Over time in Nottley, the realization dawned that given a relatively low throughput of self-referrals, buy-in from primary-care practitioners (especially GPs, with their influence on purchasing decisions) was something that the project needed after all:GPM: Is it an issue if GPs don't know about it? Do they need to know?Genetic counselor: Yeah, they do. For any possibility of extending the service, they do. Every person we've seen has had a letter, […] copied to their GP, so the GP knows they've been seen as part of the project. But it's just along with the other thousand [things] they look at per day! That's the side that has been a little disappointing, I thought we might get more involvement and a bit of ownership of the project amongst the practice nurses, and the other staff in the GP surgeries.Pilot genetic counselorThe pilot staff had little experience of engagement with primary care, with its rather different set of pressures, policies, and relationships. The same culture shock had affected Derton to some extent, but there the project could rely on the proactive work of the cancer network or amenable groups of hospital-based stakeholders. Within Nottley, the structure of the network did not support the pilot team's forlorn efforts at influencing practitioners. The managerial priorities of the network were compounded by the fact that primary-care practitioners, though formally part of the network, were also subject to many other pressures and priorities.Furthermore, the pilot staff lacked experience of the competitive funding mechanisms for this kind of project, being based in a specialist genetics department which received its funding through separate arrangements. Bidding for money from mainstream purchasers was thus a foreign experience for the project lead, and the departmental business manager. So, as a cancer-network manager explained, the pilot was “outside the mainstream cancer community” of the wider hospital and therefore marginal to cancer clinicians, the cancer network, and purchasers. The manager continued that although she wanted the project to work, “in terms of my overall job [in the cancer network]—the things I am meant to lose sleep about—it is quite a small part.” The pilot, then, found itself falling between stools. To use the lexicon of Crosby and Bryson, it was estranged from the core concerns of the cancer network, and the opportunities for effecting visionary leadership through the network's “forums.” Like Derton, it was also estranged from the primary-care–based purchasers. Unlike Derton, though, it lacked the critical mass of support from stakeholders within the cancer network, which had provided Derton's project with sufficient political leadership and leverage to succeed in the “arenas” outside the cancer network which determined decisions on ongoing funding.Consequently, despite various applications, neither this part of the pilot nor the other obtained ongoing funding. The absence of money or political will across the cancer network led the project lead to surmise, rather gloomily, that proactive efforts to lead this kind of reform were doomed to failure. It would only be when genetics centers began to place directive limits on the numbers of referrals that purchasers would be forced, reactively, to engage with the need for reformed care pathways and effective demand management.DISCUSSIONThe divergent trajectories of these ostensibly similar service reforms, based on a common care-pathway model, can be traced to a number of interacting factors. Both sought to introduce service-delivery reforms to improve client focus, and differences in the adaptation of this ambition to existing provision were significant in this divergence in themselves. However, the divergent trajectories also related to differences in the contexts of the two pilots—particularly the networks—and the agency of those seeking to implement them—notably, the presence or absence of particular forms of leadership among pilot staff and wider stakeholders.In both cases, the pilot staff attempted to draw on the resources of networks that should, in principle, be supportive of organizational reforms to rationalize care pathways, improve patient outcomes, and reduce inconsistency of provision. In practice, though, and reflecting the findings of Addicott et al. (2007), the network in Nottley was dysfunctional in this regard, with a focus on performance management and efficiencies that squeezed out any nonmandated initiatives. In Derton, the cancer network was more supportive of the pilot. In part, this seemed due to the more productive relationships already present, but it was notable also that the issue addressed by the pilot was already high on the political agenda—not least, due to financial concerns. This confirms the important point made by Van de Ven and Rogers (1988) that we should not overemphasize the role of organizational resistance in impeding innovation diffusion. Rather, it may be more instructive to view organizational resistance as the aggregate resistance of members or the inappropriateness of the innovation to the core concerns of the organization, as seemed to be the case in Nottley (Van de Ven and Rogers 1988). In Derton, in contrast, the pilot was aligned with the managerial agenda of the wider network, indicating that the susceptibility of networks to top-down pressures (Addicott et al. 2007; Keast and Brown 2006) is not necessarily an obstacle to collaboration within them (cf. Moynihan and Ingraham 2003; Olsen 2006; Provan and Milward 1995)—provided those concerned know which “mild pressures” to exert, as the lead here put it. Rather, there was complementarity between the collegiate, “quiet” leadership embodied by the pilot team and the slightly more directive leadership deployed by the network manager, who recognized the synergy between the aims of the network and those of the pilot. Here we echo the findings of various authors about the diversity of skills needed by leaders in networked settings (Huxham and Vangen 2000; Riccucci 1995; Vangen and Huxham 2003), the importance of influential champions to innovation diffusion (Rogers 2003), and indeed the need for leaders with different skills and organizational positions in relation to different leadership tasks (Crosby and Bryson 2005; Shortell et al. 2000). Furthermore, we highlight that public-service networks in particular, with diffuse objectives and loci of power, may require this combination of leadership styles and all the more so if they are professional bureaucracies with the extant complexities of power and accountability that these entail. Especially important in this context was the ability of the pilot team to make their service relevant to the concerns of others, increasing the likelihood of acceptance (Ferlie et al. 2005; Fitzgerald and Dopson 2005; Harrison 1998) by creating, as Crosby and Bryson (2005, 115) have it, “communal stories that help diverse stakeholder groups develop a sense of what they have in common with each other and what they might do to tackle common problems and create a better future.” This was a crucial way in which heterophilous links, in Rogers’ (2003) terms, were created to ensure that the Kenilworth model was adopted and accepted by diverse stakeholders.Evident in Derton was the way in which the structure of the network facilitated effective leadership through formal and informal forums which provided pilot members with the opportunity to foster a collective vision with other actors, mirroring closely the role for visionary leadership articulated by Crosby and Bryson (2005). Through the work of the pilot team in engaging with others through these forums and the directive leadership of the network manager, they were able to make effective use of the three “leadership media” identified by Huxham and Vangen (2000): the structure of the network provided a normative tool that could be used to bring “deviant” practitioners into line; the processes of the network offered the opportunity for the development of a shared vision; and, crucially, through other participants in the network, the vision was spread further and became established across diverse stakeholders. Central to this, then, was not just the quiet, “distributed” leadership within the pilot, but also the quiet, “dispersed” leadership (Buchanan, Caldwell, et al. 2007; cf. Hartley 2005) of stakeholders in the wider network. The former was to some extent under the control of the project, whereas the latter was more diffuse, but nevertheless important to widespread uptake. To this extent, then, the agency of the leaders relied on the structures (and processes and participants) of the network. This finding also reflects the intimations of Alexander et al. (2001) on the importance of what they call “collateral leadership” among both staff and community groups in the community-care networks they study, as a complement to—not a substitute for—more vision-based leadership. Our findings here converge with and develop existing literature in Europe and North America on distributed leadership in health-care settings (Buchanan, Addicott, et al. 2007; Denis et al. 2000; Huxham and Vangen 2000), local government (Hartley and Allison 2000), and education (Leithwood et al. 1999; Spillane et al. 2001). These studies illustrate the importance of subtle, distributed forms of leadership in settings characterized by the ambiguity of objectives and power relationships that typifies public-service organizations, particularly professional bureaucracies. Our study shows how, with the added complication of the interorganizational network, multiple forms of distributed and dispersed leadership are required to engage and inculcate diverse, powerful stakeholders, given the powerlessness to achieve change of any single actor or group of actors. This combination of leadership styles—and, perhaps more importantly, “leaders”—seems from our study crucial to achieve transformative change in such a context.Thus, the structures, processes, and participants of the network were essential resources for the effectiveness of leadership. It is in this sense, perhaps, that, as O'Toole (1997, 49) has it, there is a need for “a consideration of networks as causal forces in the administrative setting.” Certainly, the efficacy of leadership in Derton relied in part on the opportunities and normative framework provided by the network. However, what is also clear, from the comparison with Nottley, in particular, is the impotence of the network form in itself. Without the will and the agency of its actors, as expressed through leadership at various organizational levels, the network has no causal force. Formal structures, as scholars such as Lipsky (1980), Riccucci (2005), and Sandfort (2000) have found, have little determining effect on the behavior of front-line staff, especially where those staff enjoy professional autonomy. As Sandfort (2000) suggests, it is the “informal” structures of professionals’ collective schemas of understanding that are determinant of behavior, not the reconfiguration of formal, organizational structures. As such, transformative agency is crucial, embodied in leadership of some form, even as leaders employ the structures, processes, and participants of the network in achieving their ends. To this extent, the policymakers seem correct in their assessment of the need for complementarity.However, what our findings also illustrate are the limits to the space in which this synergistic relationship between structure and agency can operate. In Nottley especially, but to some extent also in Derton, the influence of coexisting structural forms was always present. This was most obvious in relation to physicians and purchasers in primary care. The division between primary-care and hospital physicians has been noted as a significant barrier to integration elsewhere (Grumbach et al. 1999; Shortell et al. 2000). The purchaser-provider split is a ubiquitous one in networks in any field which is at least partly characterized by a contractual, market or quasi-market relationship (Kickert et al. 1997). Though formally part of the network, it was clear in both sites that physicians and purchasers were susceptible to neither the principles of collaboration nor the performance of leadership in the way that the network's more central stakeholders were. It is evident, then, that the space created for effective leadership by the structural intervention of the network is a partial, constricted one, that is less open in the margins of the network where there are competing pressures on behavior from other organizational forms, such as hierarchical performance management and competition between providers. Crosby and Bryson (2005) recognize this when they identify the rather different leadership skills required in arenas and courts as compared to forums. Notable in relation to their example of the African-American Men Project in Minnesota was the administrative, decision-making power borne by a key leader as an elected official, enabling him to enact political leadership in the executive arenas concerned (Crosby and Bryson 2005). Nottley's lack of such political leadership saw it fail in its bid for resources in the arena of the purchasing decision-makers; Derton's project almost met a similar fate, saved only by the collective political clout of the network's powerful stakeholders. The more general, theoretical point in relation to the categorization of Crosby and Bryson (2005) of effective leadership strategies is that whereas forums may be susceptible to the shaping influence of leaders within a network, arenas and courts frequently are not. As such they require not just a different style of leadership but potentially also a different set of leaders altogether, with legitimacy and leverage within their domains.Our contribution to theory may thus be summarized as follows. The need for distributed leadership, highlighted in the literature in relation to various public-service professional bureaucracies, is compounded by the interorganizational network form. Leadership needs to be dispersed as well as distributed to engage diverse, powerful stakeholder groups, and this implies a variety of leaders as well as leadership styles. The interorganizational network itself is ineffective as a means of achieving change, but it does provide the space and the media for leaders to effect change. However, this space is a constricted one, and at its boundaries are other organizational forms, with their pressures on actors and corresponding norms of behavior. Consequently distributed and dispersed leadership are often not enough to achieve change, if change rests also on the decisions and behaviors of those governed by the pressures of these other organizational forms. The arenas and courts outside the network, in terms of Crosby and Bryson (2005), may require not only a different form of leadership but also a different tranche of leaders. This last point may relate especially to countries such as the United Kingdom and much of continental Europe, where the remnants of traditional, centralized systems coexist alongside the network form, though as we have seen, it is also relevant in relation to the coexistence of networks and markets.On a practical level, certain specific lessons for the health sector also have broader implications for the operation of public-service networks more generally. In the health services of the United Kingdom, United States, and other countries, “the relative power of specialists has diminished as primary care physicians have assumed a more central role as the initial contact point with patients and as major decision makers regarding referrals to specialists and sites of care throughout the system” (Shortell et al. 2000, 79). In the United Kingdom, policymakers are currently seeking to consolidate this new-found power of primary-care physicians and purchasers, so that they might become the drivers of health-service change. However, evident from both these studies was the degree to which primary-care actors were sidelined within the networks and in service reform. The marginality of this locus of power was instrumental in Nottley's failure and almost derailed Derton's project too, and in more general terms, the need for policymakers to align incentives and levers, through integration or centralization within the network, is strong (Keast and Brown 2006; Milward and Provan 1998; Provan and Milward 1995; Provan et al. 2004). Client-level networking between practitioners (Provan and Sebastian 1998) needs to be supported by organizational integration that supports practitioners’ efforts and enables the establishment of change in the face of prevailing institutional pressures. “Inter alia,” this means that notwithstanding their contractually based relationship, organizations on both sides of purchaser-provider relationships need to be incorporated in the structures, and inculcated in the ethos, of the network.CONCLUSIONOur study is based on relatively thin empirical data, and so should be interpreted cautiously, especially in terms of the additions to existing theory suggested. Many of the findings, though, confirm existing evidence; moreover the findings on the importance of distributed leadership from these two case studies are mirrored in data from our wider study of 11 sites. Our findings indicate that despite the well-documented obstacles to networked governance and distributed leadership in the public services, a careful alignment of objectives with managerial agenda can bring success to reforms in service provision. A combination of clear benefits to wider stakeholders within the network, and distributed and dispersed leadership, can give rise to effective collaboration and establishment of reforms, through structural integration and the harnessing of agency. To this extent, policies invoking structure and agency were indeed synergistic. However, the relationship was a contingent one, illustrating how network-based reforms to organizational structure are both potentially powerful and simultaneously impotent: powerful in creating a space within which certain enactments of leadership might flourish and achieve service reform; impotent as instruments of that reform in themselves, as a means of changing practice in the face of the much more powerful, informal structures that are determinant of behavior (Riccucci 2005; Sandfort 2000). Furthermore, the space for effective leadership created was a limited one, which became more and more constricted in the face of coexistent organizational forms and modes of governance. Networks compound the need for distributed leadership in public-service contexts since they introduce new loci of power that might be influenced by a more dispersed form of leadership. However, the coexistence of networks alongside other organizational forms constricts the effectiveness of this mode of leadership, and it is likely that in many settings, complementary modes of leadership that conform with the requirements of these other organizational forms may be required.Our study adds to a growing literature on the practice of leadership and the role of networks in public-service organizations and brings these literatures together to highlight the potentials for and impediments to leadership in public-service networks. As we noted in our opening section, as policies focused on transforming the structures of public-service provision and the agency of public servants respectively, networks and leadership seem to some extent complementary. The tangible success of distributed, quiet leadership, and dispersed change agency, in one case, indicates how these two policies can indeed work synergistically. The less successful efforts of the second case, illustrate the fragility of this complementarity.In terms of policy, our study reaffirms the importance of carefully aligned policy levers that ensure that organizations within networks have drivers which incline them toward networked collaboration rather than operation within their silos. In health, the gap between hospital and primary care would seem to be a particularly important one in terms of divergent drivers and ill-aligned incentives, reflecting the uneven influence of network-based organizational forms in other sectors, especially those characterized by market or quasi-market relationships. For organizations and practitioners, then, the key message is around ensuring the functionality of networks and the importance of locating those crucial change agents whose influence might be harnessed toward service reform. Finally, we hope our study illustrates the distinctive contribution of qualitative study (cf. Bryman 2004), even as it highlights the need for further investigation of the processes by which individual agency interacts with structural reconfiguration to succeed or fail in consolidating modernizing service reforms of this kind.FUNDINGUK Department of Health.AddicottRachaelMcGivernGerryFerlieEwanThe distortion of a managerial technique? The case of clinical networks in UK health careBritish Journal of Management20071893105AgranoffRobertMcGuireMichaelBig questions in public network management researchJournal of Public Administration Research and Theory200111295326AlexanderJeffery AComfortMaureen EWeinerBryan JBogueRichardLeadership in collaborative community health partnershipsNonprofit Management and Leadership20011215975BaquetClaudia RMackKelly MBrambleJoyDatcherDeloresSavoyMervin AHummelKeryMishraShiraz IBrooksSandra EBoykin-BrownStephanieMaryland's special populations cancer network: Cancer health disparities reduction modelJournal of Health Care for the Poor and Underserved200516192206BatePaulChanging the culture of a hospital: From hierarchy to networked communityPublic Administration200078485512BennettNigelWiseChristineWoodsPhilipHarveyJanet ADistributed leadership2003Nottingham, UKNational College for School LeadershipBrandsenTacovan HoutEelcoCo-management in public service networks: The organizational effectsPublic Sector Management2006853749BruijnJ. A. deten HeuvelhofE. F.KickertWalter J MKlijnErik-HansKoppenjanJoop F MInstruments for network managementManaging complex networks: Strategies for the public sector1997LondonSage11936BrymanAlanCharisma and leadership in organizations1992LondonSage———CleggStuart RHardyCynthiaNordWalter RLeadership in organizationsManaging organizations: Current issues1999LondonSage2642———Qualitative research on leadership: A critical but appreciative reviewLeadership Quarterly20041572969BrysonJohnKelleyGeorgeA political perspective on leadership emergence, stability, and change in organizational networksAcademy of Management Review1978371323BuchananDavid AAddicottRachaelFitzgeraldLouiseFerlieEwanBaezaJuan INobody in charge: Distributed change agency in healthcareHuman Relations200760106590BuchananDavid ACaldwellRaymondMeyerJulienneStoreyJohnWainwrightCharlesLeadership transmission: A muddled metaphor?Journal of Health Organization and Management20072124658Cabinet OfficeModernising government1999LondonHMSOChristensenTomLise FimreiteAnneLaegriedPerReform of the employment and welfare administrations: The challenges of co-coordinating diverse public organizationsInternational Review of Administrative Sciences200773389408CrosbyBarbara CBrysonJohn MLeadership for the common good: Tackling public problems in a shared-power world2005San FranciscoJossey-BassCurrieGraemeLockettAndyA critique of transformational leadership: Moral, professional and contingent dimensions of leadership within public services organizationsHuman Relations20076034170CurrieGraemeBoyettIngerSuhomlinovaOlgaTransformational leadership within secondary schools in England: A panacea for organizational ills?Public Administration20058326596CurrieGraemeSuhomlinovaOlgaThe impact of institutional forces upon knowledge sharing in the UK NHS: The triumph of professional power and the inconsistency of policyPublic Administration200684130DenisJean-LouisLangleyAnnCazaleLindaLeadership and strategic change under ambiguityOrganization Studies19961767399DenisJean-LouisLangleyAnnPineaultMarcBecoming a leader in a complex organizationJournal of Management Studies200037106399Department of HealthA policy framework for commissioning cancer services1995LondonDepartment of Health———Genetics and cancer services1998LondonDepartment of HealthEisenhardtKathleen MBuilding theories from case study researchAcademy of Management Review19891453250FerlieEwanPettigrewAndrewManaging through networks: Some issues and implications for the NHSBritish Journal of Management19967S8199FerlieEwanHartleyJeanMartinSteveChanging public service organizations: Current perspectives and future prospectsBritish Journal of Management200314S114FerlieEwanFitzgeraldLouiseWoodMartinHawkinsChrisThe nonspread of innovations: The mediating role of professionalsAcademy of Management Journal20054811734FerlieEwanAshburnerLynnFitzgeraldLouisePettigrewAndrewThe new public management in action1996OxfordOxford Univ. PressFitzgeraldLouiseDopsonSueDopsonSueFitzgeraldLouiseKnowledge, credible evidence, and utilizationKnowledge to action? Evidence-based health care in context2005OxfordOxford Univ. Press13254GrecoPeter JEisenbergJohn MChanging physicians’ practicesNew England Journal of Medicine199332912714GrumbachKevinSelbyJoe VDambergCherylBindmanAndrew BQuesenberryCharlesTrumanAlisonUratsuConnieResolving the gatekeeper conundrum: What patients value in primary care and referrals to specialistsJournal of the American Medical Association19992822616HaffertyFrederic WLightDonald WProfessional dynamics and the changing nature of medical workJournal of Health and Social Behavior199535Extra Issue13253HallThad EO'TooleLaurence JShaping formal networks through the regulatory processAdministration and Society200436186207HarrisonStephenThe politics of evidence-based medicine in the United KingdomPolicy and Politics1998261531HarrisonStephenHunterDavid JMarnochGordonPollittChristopherJust managing: Power and culture in the National Health Service1992Basingstoke, UKMacmillanHartleyJeanInnovation in governance and public services: Past and presentPublic Money and Management2005252734HartleyJeanAllisonMariaThe role of leadership in the modernization and improvement of public servicesPublic Money and Management2000203540HeifetzRonald ALeadership without easy answers1994CambridgeHarvard Univ. PressHennesseyJ. Thomas‘Reinventing’ government: Does leadership make the difference?Public Administration Review19985852232HoodChristopherPetersGuyThe middle aging of New Public Management: Into the age of paradox?Journal of Public Administration Research and Theory20041426782HoodChristopherJamesOliverScottColinRegulation of government: Has it increased, is it increasing, should it be diminished?Public Administration200078283304HuxhamChrisVangenSivLeadership in the shaping and implementation of collaboration agendas: How things happen in a (not quite) joined-up worldAcademy of Management Journal200043115975JenningsEdward TEwaltJo Ann GInterorganizational coordination, administrative consolidation, and policy performancePublic Administration Review19985841728KakabadseAndrewKorac-KakabadseNadaKouzminAlexanderEthics, values and behaviours: Comparison of three case studies examining the paucity of leadership in governmentPublic Administration200381477508KeastRobynBrownKerryAdjusting to new ways of working: Experiments with service delivery in the public sectorAustralian Journal of Public Administration20066544153KickertWalter J. MKlijnErik-HansKoppenjanJoop F N.Managing complex networks: Strategies for the public sector1997LondonSageedKickertWalter J. MKoppenjanJoop F N.KickertWalter JMKlijnErik-HansKoppenjanJoop F NPublic management and network management: An overviewManaging complex networks: Strategies for the public sector1997LondonSage3561LangleyAnnStrategies for theorizing from process dataAcademy of Management Review199924691710LeeThomas WUsing qualitative methods in organizational research1999Thousand OaksSageLeithwoodKennethJantziDorisSteinbachRosanneChanging leadership for changing times1999Philadelphia, PAOpen Univ. PressLipskyMichaelStreet-level bureaucracy: Dilemmas of the individual in public services1980New YorkRussell Sage FoundationMcKinneyMartha MMorrisseyJoseph PKaluznyArnold DInterorganizational exchanges as performance markers in a community cancer networkHealth Services Research19932845978McNultyTerryFerlieEwanRe-engineering health care: the complexities of organizational transformation2002OxfordOxford Univ. PressMilwardH. BrintonProvanKeith GPrinciples for controlling agents: The political economy of network structureJournal of Public Administration Research and Theory1998820321MoynihanDonald PIngrahamPatricia WLook for the silver lining: When performance-based accountability systems workJournal of Public Administration Research and Theory20031346990NancarrowSusan ABorthwickAlan MDynamic professional boundaries in the healthcare workforceSociology of Health and Illness200527897919NewmanJanetModernizing Governance: New Labour, Policy and Society2001LondonSageNohriaNitinEcclesRobert GNetworks and organizations1992BostonHarvard Business School PressOlsenJohan PMaybe it is time to rediscover bureaucracyJournal of Public Administration Research and Theory200616124OsborneDavidGaeblerTedReinventing government1992Reading, MAAddison-WesleyO'TooleLaurence JTreating networks seriously: Practical and research-based agendas in public administrationPublic Administration Review1997574552MeierKenneth JPublic management in intergovernmental networks: Matching structural networks and managerial networkingJournal of Public Administration Research and Theory20041446994PearceCraig LCongerJay APearceCraig LCongerJay AAll those years ago: The historical underpinnings of shared leadershipShared leadership: Reframing the hows and whys of leadership2003Thousand OaksSage118PetersB. GuyPierreJohnGovernance without government? Rethinking public administrationJournal of Public Administration Research and Theory1998822343PettigrewAndrew MWhat is processual analysis?Scandinavian Journal of Management19971333748ProvanKeith GMilwardH. BrintonA preliminary theory of interorganizational network effectiveness: A comparative study of four community mental health systemsAdministrative Science Quarterly199540133ProvanKeith GSebastianJuliann GNetworks within networks: Service link overlap, organizational cliques, and network effectivenessAcademy of Management Journal19984145363ProvanKeith GRoussin IsettKimberleyMilwardH. BrintonCooperation and compromise: A network response to conflicting institutional pressures in community mental healthNonprofit and Voluntary Sector Quarterly200433489514RaineyHal GSteinbauerPaulaGalloping elephants: Developing elements of a theory of effective government organizationsJournal of Public Administration Research and Theory19999132RashmanLyndsayHartleyJeanLeading and learning? Knowledge transfer in the Beacon Council schemePublic Administration20028052342ReedMichaelThe disorganized society and the future of organizations1992LondonSageRhodesRAWUnderstanding governance: Policy networks, governance, reflexivity and accountability1997Philadelphia, PAOpen Univ. PressRiccucciNorma MUnsung heroes: Federal execucrats making a difference1995Washington, DCGeorgetown Univ. Press———How management matters: Street-level bureaucrats and welfare reform2005Washington, DCGeorgetown Univ. PressRogersEverett MDiffusion of innovations2003New YorkFree PressRyanChristineWalshPeterCollaboration of public sector agencies: Reporting and accountability challengesInternational Journal of Public Sector Management20041862131SandfortJodi RMoving beyond discretion and outcomes: Examining public management from the front lines of the welfare systemJournal of Public Administration Research and Theory20001072956ScharpfFritz W.Games in hierarchy and networks: Analytical and empirical approaches to the study of governance institutions1993Frankfurt, GermanyCampus VerlagSecretary of State for HealthOur inheritance, our future: Realising the potential of genetics in the NHS2003LondonThe Stationery OfficeShortellStephen MGilliesRobin RAndersonDavid AThe new world of managed care: Creating organized delivery systemsHealth Affairs1994134664ShortellStephen MGilliesRobin RAndersonDavid AMorgan EricksonKarenMitchellJohn BRemaking health care in America: The evolution of organized delivery systems2000San FranciscoJossey-BassShortellStephen MWaltersTeresa MClarkeKenneth W B.BudettiPeter PPhysicians as double agents: Maintaining trust in an era of multiple accountabilitiesJournal of the American Medical Association199828011028SilvermanDavidQualitative research: Theory, method and practice2004LondonSageSpillaneJames PHalversonRichardDiamondJohn BInvestigating school leadership practice: A distributed perspectiveEducational Researcher200130238TenbenselTimMultiple modes of governance: Disentangling the alternatives to hierarchies and marketsPublic Management Review2005726788ThompsonGrahameFrancesJenniferLevacicRosalindMitchellJeremyMarkets, hierarchies and networks: The coordination of social life1991LondonSageVan de VenAndrew HRogersEverett MInnovations and organizations: Critical perspectivesCommunication Research19881563251van WartMontgomeryDynamics of leadership in public service: theory and practice2005LondonSharpeVangenSivHuxhamChrisEnacting leadership for collaborative advantage: Dilemmas of ideology and pragmatism in the activities of partnership managersBritish Journal of Management200314S6176WijngaardenJeroen D. H. vande BontAntoinette AHuijsmanRobbertLearning to cross boundaries: The integration of a health network to deliver seamless careHealth Policy20067920313YinRobert KCase study research: Design and methods2003LondonSage1We refer throughout this article to the public services in England only since policies in the devolved administrations of the other three nations of the United Kingdom have diverged significantly from English public-service reforms, particularly in relation to health-service reorganization.2Bill Clinton, Announcement of the initiative to streamline government, March 1993. http://govinfo.library.unt.edu/npr/library/speeches/030393.html, accessed January 18, 2008.3See, for example, http://www.nhsleadershipqualities.nhs.uk/3, accessed September 18, 2007.4Macmillan is a British cancer charity with a particular interest in contributing to improving the management of cancer and ensuring an improved patient experience by working toward more joined-up cancer provision within the NHS and between NHS and other providers of care.5We use pseudonyms and do not give exact figures regarding patient throughput here, in order to protect the identities of our case-study sites, in accordance with the terms of ethical approval for the wider evaluation from which this article derives.6As one cancer network manager here confided: “I've been to other networks where they haven't got relationships with hospitals: they're not allowed in through the doors.”7It was also in contrast to another case-study site not detailed here, where mobilization of involved service users had seen a proactive publicity campaign resulting in large numbers of self-referrals and sustained media coverage for the service.
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0BB74813D12C151A76D786E766714354809B160E.txt b/test/dataset/in/resources/corpus/Clean_0BB74813D12C151A76D786E766714354809B160E.txt
deleted file mode 100644
index e4e4076..0000000
--- a/test/dataset/in/resources/corpus/Clean_0BB74813D12C151A76D786E766714354809B160E.txt
+++ /dev/null
@@ -1 +0,0 @@
-]>EXG5327S0531-5565(00)00139-X10.1016/S0531-5565(00)00139-XElsevier Science Inc.Fig. 1Possible causes of aging in the mechanisms of life, and the target of “Gerontology Proteomics”. Cascade of cellular aging may proceed in an intracellular environment that is made with cell-type-specific “proteome”. Many primary causes of cellular aging are commonly speculated for all cell types including mitotic and post-mitotic cells. Genetic background is one of the most important factor that determine the quality of the defense system against environmental attacks: Cosmic rays, UV light, reactive oxygen species (ROS), and reducing sugars are significant elements of environmental attacks. Results of the battle against these environmental attacks may be revealed in their proteome profiles.Fig. 2Detection of age-related protein alteration by 2D gel electrophoresis and quantitation of relative intensity by image processing in proteome analysis. The relative intensity of ssp7001 shows a transitional increase around 65PDL, which is the border between phases 2 and 3 of cellular aging of normal human diploid fibroblasts TIG-3.Fig. 3Identification of spot protein by MS. The spot protein was transferred to a PVDF membrane, and digested with lysylendoproteinase Lys-C. The mass spectrum of peptide fragments was obtained using a Micromass Q-Tof MS system. The mass fingerprint database search was executed on the internet MS-FIT site. The ssp7001 was identified as phosphoprotein stathmin.Fig. 4Determination of methionine oxidation in human amyloid beta peptide (1–40) by MALDI-TOF-MS. The sample solution was prepared as described in Methods. One-μl aliquot was applied to the MALDI target plate. The mass spectrum was recorded with a Voyager-DE STR MALDI-TOF-MS system (PE Biosystems, Foster City, CA) in a reflection mode. The spectrum indicated that the authentic preparation of human amyloid beta peptide (1–40) contained both native and Met35-oxidized forms. Detected ions were all monovalent and the 15.96-Da mass shift was derived from the methionine oxidation.Current status and perspectives of proteomics in aging researchTToda*ttoda@tmig.or.jpDepartment of Gene Regulation and Protein Function, Tokyo Metropolitan Institute of Gerontology, 35-2 Sakaecho, Itabashi-Ku, Tokyo 173-0015, Japan*Tel. +81-3-39643241; fax: +81-3-35794776AbstractThe accumulation of non-enzymatic modifications on both DNA and protein molecules under the attack of reactive oxygen species (ROS), is one of the most possible factors responsible for the functional deterioration in aged cells. Direct protein modifications as well as DNA damages may be detectable, in part, by proteome analysis if the gene expression is affected by the damages on DNA. The novel term “proteome”, which is a compound of “protein” and “genome”, means a whole set of proteins expressed in a tissue or a cell strain to be investigated. Proteomics is a methodology for analyzing proteomes. In proteomics, two-dimensional gel electrophoresis is performed primarily to separate constitutive proteins, followed by mass spectrometry to identify each protein of interest and to determine a possible post-translational modification. Proteomics has offered us an innovative tool for investigating the molecular mechanisms of cellular aging.KeywordsProteomeTwo-dimensional gel electrophoresisMass spectrometryProtein oxidation1IntroductionThe age-dependent deterioration of cell function appears in both replicative and post-mitotic cells, in living organisms. The major function of stem cells is cell division in the regeneration of cell population and tissues. Most replicative blast cells play another role: in the excretion of cytokines for inter-cellular signal transduction. Post-mitotic cells in the brain, heart, kidney and many other organs have tissue-specific functions. Although a downstream of the cascade in cellular aging may be cell-type specific, and may proceed in a given intracellular environment that is made with all constituents of the proteome, there may be common factors in the upstream of the cascade. The accumulation of molecular damages on DNAs and proteins under environmental attacks including oxidative stress, may be one of the major events that trigger the cascade of cellular aging as described in Fig. 1.Oxidative protein modification may alter methionine residue to sulfoxide, phenylalanine to diphenyl, and lysyl residue to carbonyl (Smith et al., 1997; Wells Knecht et al., 1997; Stadtman and Berlett, 1998). In proteomics, such protein altered by post-translational modification, may be separated from the original form by high resolution two-dimensional gel electrophoresis (2-DE), and the modification may be determined by mass spectrometry (MS).2Materials and methodsImmobiline DryStrip and Pharmalyte were purchased from Amersham Pharmacia Biotech KK (Tokyo, Japan). Sequi-Blot PVDF membrane was obtained from the Nippon Bio-Rad Laboratories (Tokyo, Japan). Acrylamide, N,N′-methylenebisacrylamide and TEMED were from Daiichi Pure Chemicals (Tokyo, Japan). Sequencing grade endoproteinase Lys-C was from Roche Molecular Biochemicals (Indianapolis, IN, USA). Trizma base, Tricine, SDS, Triton X-100, iodoacetamide, acetonitrile, trifluoroacetic acid and alpha-cyano 4-hydroxy-cinnamic acid were from Sigma (St. Louis, MO, USA). Quick CBB staining reagent, silver staining reagent kit “Wako”, urea, and other chemicals of reagent grade were obtained from Wako Pure chemicals (Osaka, Japan).2.1Two-dimensional gel electrophoresisConstitutive proteins in a cell extract were separated into isolated spots by 2-DE in an “immobilized pH gradient isoelectric focusing (IPG-IEF)/sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) system” according to the standard protocol (Toda and Kimura, 1997), with minor modifications. The details of the modified protocol are shown in our web homepage: http://proteome.tmig.or.jp/2D/2DE_method.html. In brief, the first-dimensional IPG-IEF was carried out on a re-swollen Immobiline DryStrip, pH 4–8, 18-cm long (Code No. 18-1004-34) in the CoolPhoreStar IPG-IEF Type P horizontal IEF apparatus (Anatech, Tokyo, Japan). A 20-μl aliquot of the sample solution, which was absorbed in a small piece of filter paper, was applied near the cathode wick on the IPG gel. Spot proteins on a gel plate for “differential protein display” were subsequently visualized by silver staining. For the preparative use of 2-DE gel, the IPG gel was swollen in a solution containing 0.2ml of cell extract. After 46,700-Vh electrofocusing, the IPG gel was equilibrated with the SDS-treatment solution (50mM Tris–HCl, pH 6.8, 6M urea, 0.5% (w/v) dithiothreitol, 2% (w/v) SDS, 0.005% (w/v) BPB, 25% (v/v) glycerol), initially for 30min, and then followed by carbamoylmethylation in an iodoacetamide-containing buffer (50mM Tris–HCl, pH 6.8, 6M urea, 4.5% (w/v) iodoacetamide, 2% (w/v) SDS, 0.005% (w/v) bromophenol blue (BPB), 25% (v/v) glycerol) for 10min. An equilibrated gel strip was placed on top of a gel slab (7.5%T, 3%C, 20×18cm), and firmly contacted to the top of a gel slab by pressing the gel strip with a shark-teeth comb. SDS-PAGE was run vertically in the CoolPhoreStar SDS-PAGE Tetra-200 vertical slab gel electrophoresis apparatus (Anatech, Tokyo, Japan) using a tricine buffer system.2.2Protein staining and image processingProtein spots were visualized on the gel slab for a differential protein display analysis by silver staining, using the original version of the reagent kit “Wako” as it showed the widest dynamic range in optical density among all the commercially available reagents tested. The 2-DE gel images were acquired using a Sharp JX-330 scanner. Noise reduction, background subtraction, spot detection, spot quantification, gel-to-gel matching and differential protein analysis were carried out using a PDQuest software (Bio-Rad Laboratories, Hercules, CA, USA).2.3Endoproteinase digestion and peptide mass finger printingAfter scanning, Coomassie-stained proteins in the gel slab were re-solubilized and transferred onto a PVDF membrane as follows. The stained gel was rinsed in pure water thrice, each time for 30min, incubated in a re-solubilization buffer (0.2% (w/v) SDS, 0.3% (w/v) Tris, 0.7% (w/v) glycine) for 10min and mounted on an electrotransfer blotting chamber. The electrotransfer was carried out overnight at 4V/cm in 0.1% (w/v) SDS, 0.3% (w/v) Tris and 1.5% (w/v) glycine.Protein spots were excised from the PVDF membrane and the Coomassie dye was removed by rinsing in 60% (v/v) methanol. For the MS analysis, the piece of membrane was incubated with 4.5% (w/v) polyvinyl pyrolidone 25 in 2.5mM Tris–HCl, pH 7.7, for 30min, followed by a brief rinsing in 5% (v/v) acetonitrile. The digestion was carried out overnight with 0.1μg of sequencing grade endoproteinase Lys-C in 30μl of 8% (v/v) acetonitrile, 2.5mM Tris–HCl, pH 7.7, at 37°C. After digestion, the reaction mixture was supplemented with 10μl of acetonitrile and sonicated for 5min.After removing the acetonitrile by blowing with nitrogen gas, peptides in the digestion mixture were trapped on C18 resin packed in a ZipTipC18 (Millipore Corporation, MA, USA), and eluted in 5μl of 75% (v/v) methanol, 1% (v/v) formic acid. Nano-ESI-MS for peptide-mass fingerprinting was performed using a Micromass Q-Tof system, equipped with a NanoFlow Probe Tip Type F (Micromass UK Ltd., Manchester, UK). The peptide solution was put into a bolosilicate capillary tip and subjected to electrospray ionization (ESI) at a flow rate of 10nl/min. The MS spectrum was analyzed with the ProteinLynx software. Protein identification through a database search was carried out using the MS-FIT proteomic tool at the UCSF web server, through the internet.3Results3.1Screening of age-related protein alteration by proteome profiling and differential protein displayAlterations in the relative intensity of protein spots appearing on proteome profiles, may be the results of various molecular alterations that occur during aging. Quantitation of the integrated optical density of each protein spot is required for the 2D gel electrophoresis of protein, to screen out the molecular events that may be responsible for the functional deterioration in senescent cells.Coomassie Blue staining is the most reliable method for the quantitative demonstration of proteins on a gel slab. However, the sensitivity of Coomassie staining is not high enough to detect minor components of cellular proteins. Autoradiography of [35S]methionine-labeled proteins shows the highest sensitivity. However, it is not applicable to the detection of post-translational modification of proteins. This is due to the fact that, only the fresh proteins that are de novo synthesized during the period of labeling incubation are detected by autoradiography. Cypro ruby fluorescent staining shows almost the same sensitivity as silver staining, and quantitativity as Coomassie staining. Therefore, fluorescent staining will be included in the standard protocol of proteomics in the future. Silver staining is a practical method for spot protein visualization in proteome analysis at present, because it shows the highest sensitivity and practical reliability in relative quantitation for differential protein display, as shown in Fig. 2.In this 2D gel area, the transitional increase of spot protein ssp7001 was observed around 65PDL; at that level, the normal human diploid fibroblasts (TIG-3) transfer into phase 3 of the replicative cell aging and the doubling time was delayed.3.2Identification of spot protein by mass spectrometry in proteomicsIn general, western blotting has been performed for the identification of proteins in spots for the last several years. However, the immunochemical technique was not applicable to the unknown protein for which a specific antibody was not available. Co-electrophoresis is another way to assign a spot to a candidate protein when its authentic protein standard protein is available. In modern proteomics, a sequence database search queried with peptide mass fingerprint data and/or partial sequence tag data obtained by MS, is generally performed to identify proteins on 2D gel patterns. Fig. 3 shows an example of identification of ssp7001 spot protein by peptide mass fingerprinting. The mass spectrum of Lys-C digests of ssp7001 was recorded using a Micromass ESI-Q-Tof-MS system (MICROMASS UK, Manchester, UK). The database search was queried to MS-FIT in the Protein Prospector Server at California University, and subsequently it was assigned to a microtubule associating phosphoprotein stathmin. When no candidate protein is hit by the database search, because it is really novel, molecular cloning should be done to clarify the meaning of the protein alteration in the aging process. In that case, Edman-degradation microsequencing has an advantage over mass spectrometry in sequencing longer peptide fragments of the protein in order to design proves and/or primers in molecular cloning of its corresponding cDNA.3.3Determination of protein modification by mass spectrometry in proteomicsAge-related protein alterations in heat stability and specific activity have been reported by many groups (Holliday and Tarrant, 1972; Gershon and Gershon, 1973; Chetsanga and Liskiwskyi, 1977; Pigeolet and Remacle 1991). Further, the accumulation of detergent-insoluble protein was also observed in aged cells and tissues (Yang and Wang, 1994). It was suggested that these age-related alterations reflect post-translational modifications such as oxidation by attacking of reactive oxygen species (Gordillo et al., 1989). Oxidative modification may produce carbonyl groups, ortho-tyrosine structures and methionine sulfoxide residues that increase the hydrophobicity of the protein surface (Chao et al., 1997). In the proteome analysis, structures of oxidative modification on spot proteins, isolated by 2D gel electrophoresis, can be determined by MS. Human amyloid beta (Aβ)1–40 peptide contains a methionine, and its oxidation induces the alteration of the 3D structure. The Met35-oxidized Aβ shows a higher molecular mass of 16Da than native Aβ, as shown in Fig. 4. Carbonylation does not result in a sufficient shift in molecular mass (−1Da); however, theoretically its dinitrophenylhydrazine derivatives gives an increase of 180Da.4DiscussionWe have obtained an excellent methodology, proteomics, which is most suitable for investigating protein factors in molecular mechanisms of cellular aging. Alterations in gene expression that are the results of DNA damages, and accumulation of altered proteins that are made by oxidative modification can be detected by methods in proteomics. The advanced method of two-dimensional gel electrophoresis, in which isoelectric focusing is carried out on an immobilized pH gradient for the first dimensional separation, offers the highest resolution of proteins. More than thousands of spots are detected on a gel slab in an optimized condition. The proteome database of normal human diploid fibroblasts TIG-3 has been established, and is displayed on our Gerontology Informatics database: the PDL-dependent alterations of protein spots are demonstrated here. Although the resolution of 2D gel electrophoresis is still not high enough to separate all the proteins in a cell extract, it will be improved by partitioning the pH and molecular mass range using narrow pH range IPG-IFE gel strips and various concentrations of acrylamide gel slabs for SDS-PAGE. The multiplication of 2D gel electrophoresis for partitioning the range of separation may yield better results than the simple enlargement of a gel size to cover the overall range of separation. Identification of the protein by making inquiries on the proteomic database with a peptide mass fingerprint data, will be more successful after completion of the genome-sequencing projects. Most post-translational modifications that are accumulated in aged cells under attacks of reactive oxygen species (ROS), are easily detected by proteomics. The extensive application of proteomics in the investigation of altered proteins in aged cells has led to the next stage of research: the molecular mechanisms of aging.ReferencesChao et al., 1997C.CChaoY.SMaE.RStadtmanModification of protein surface hydrophobicity and methionine oxidation by oxidative systemProc. Natl. Acad. Sci. USA7199729692974Chetsanga and Liskiwskyi, 1977C.JChetsangaMLiskiwskyiDecrease in specific activity of heart and muscle aldolase in old miceInt. J. Biochem.81977753756Gershon and Gershon, 1973HGershonDGershonAltered enzyme molecules in senescent organisms, mouse muscle aldolaseMech. Aging Dev.219733341Gordillo et al., 1989EGordilloAAyalaJBautistaAMachadoImplication of lysine residues in the loss of enzymatic activity in rat liver 6-phosphogluconate dehydrogenase found in agingJ. Biol. Chem.26419891702417028Holliday and Tarrant, 1972RHollidayG.MTarrantAltered enzymes in aging human fibroblastsNature23819722630Pigeolet and Remacle, 1991EPigeoletJRemacleAlteration of enzymes in aging human fibroblasts in culture. V. Mechanisms of glutathione peroxidase modificationMech. Aging. Dev.58199193109Smith et al., 1997J.BSmithXJiangE.CAbrahamIdentification of hydrogen peroxide oxidation sites of alpha A- and alpha B-crystallinsFree Radic. Res.261997103111Stadtman and Berlett, 1998E.RStadtmanB.SBerlettReactive oxygen-mediated protein oxidation in aging and diseaseDrug Metab. Rev.301998225243Toda and Kimura, 1997TTodaNKimuraStandardization of protocol for Immobiline 2-D page and construction of 2-D page protein database on World Wide Web home pageJpn J. Electroph.4119971319Wells Knecht et al., 1997M.CWells KnechtT.JLyonsD.RMcCanceS.RThorpeJ.WBaynesAge-dependent increase in ortho-tyrosine and methionine sulfoxide in human skin collagen is not accelerated in diabetes. Evidence against a generalized increase in oxidative stress in diabetesJ. Clin. Invest.1001997839846Yang and Wang, 1994GYangEWangTerminin (Tp 63/60), a novel cell senescence-related protein, is present in the aging human hippocampusBrain Res.6441994188196
\ No newline at end of file
diff --git a/test/dataset/in/resources/corpus/Clean_0BCBE54EBC6305C3FAC6AD0B9E7B93DBB6F31E5B.txt b/test/dataset/in/resources/corpus/Clean_0BCBE54EBC6305C3FAC6AD0B9E7B93DBB6F31E5B.txt
deleted file mode 100644
index 49b7e9a..0000000
--- a/test/dataset/in/resources/corpus/Clean_0BCBE54EBC6305C3FAC6AD0B9E7B93DBB6F31E5B.txt
+++ /dev/null
@@ -1 +0,0 @@
-AMGP61210S1064-7481(12)61210-X10.1097/00019442-199905000-00009American Association for Geriatric PsychiatryTABLE 1Comparison of demographics between the early-onset-of-depression (EOD) and the late-onset-of-depression (LOD) samples (Student's t-test and chi-square)Onset of DepressionTotal Group (TG) EOD and LOD (N = 245)EOD (n = 114)LOD (n = 131)PAge, years, mean ± SD75.28±7.1372.19±6.9477.96±6.170.001Length of stay, days, mean ± SD118.69±63.42125.75 ±60.52112.53±65.440.100Ham-D, baseline, mean ± SD19.89 ±6.8420.03 ±7.0919.76 ±6.630.759Ham-D, discharge, mean ± SD10.30±6.599.35 ±6.4911.28±6.580.048Sex, n (%)0.604 Male64 (26.1%)28 (24.6%)36 (27.5%) Female181 (73-9%)86 (75.4%)95 (72.5%)Level of education, n (%)0.062 No formal education7 (2.9%)1 (0.9%)6 (4.6%) Elementary, partial or complete90 (36.7%)36 (31.6%)54 (41.2%) High school or technical school, partial45 (18.4%)24 (21.1%)21 (16%.0) High school, complete68 (27.8%)34 (29.8%)34 (26.0%) College/University, partial10 (4.1%)6 (5.3%)4 (3.1%) College/University, complete20 (8.2%)10 (8.8%)10 (7.6%) Graduate studies, partial or complete5 (2.0%)3 (2.6%)2 (1.5%)DSM-IV diagnoses, n (%) Primary Axis I0.187 Major depression240 (98.0%)110 (96.5%)130 (99.2%) Othera5 (2.0%)4 (3.5%)1 (0.8%) Axis II0.224 No diagnosis217 (88.6%)104 (91.2%)113 (86.3%) Personality disorderb28 (11.4%)10 (8.8%)18 (13.7%)Note: EOD: onset