Journal of Gerontology: SOCIAL SCIENCES
2008, Vol. 63B, No. 4, S219–S228

Copyright 2008 by The Gerontological Society of America

Physical Limitations and Depressive Symptoms:
Exploring the Nature of the Association
Mathew D. Gayman,1 R. Jay Turner,2 and Ming Cui3
1
Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill.
Department of Sociology, Center for Demography & Population Health, Florida State University, Tallahassee.
3
Department of Family & Child Sciences, Florida State University, Tallahassee.

2

Objectives. Although evidence suggests that physical disability and depression may be reciprocally related, questions
of causality versus spuriousness and the direction of causality remain to be conﬁdently answered. This study considered
the hypothesis of reciprocal inﬂuence; the possibility of spuriousness in relation to pain, stress, and lifetime major
depression; and the possible mediating effects of pain and social stress.
Methods. We analyzed data from a two-wave panel study of Miami-Dade County residents (n ¼ 1,455) that included
a substantial oversampling of individuals reporting a physical disability.
Results. Results indicated that, although prior levels of physical limitations predicted changes in depressive symptoms,
there was no evidence of the reverse association. Results also indicated that part of the association between prior physical
limitations and changes in depressive symptoms was explained by intervening level of pain and, to a lesser extent, by the
day-to-day experience of discrimination.
Discussion. Much of whatever causation may be involved in the linkage between physical limitations and depressive
symptomatology ﬂows from limitations to depression rather than in the reverse direction. Results also make clear that this
linkage is not an artifact of shared associations with pain, social stress, or lifetime major depression.
Key Words: Physical limitations—Depression—Stress—Pain.

T

HE presence of physical limitations, by deﬁnition, implies
a circumstance of limited capacity to meet the requirements of core social, familial, and occupational roles and thus
to obtain associated rewards and satisfactions. From this
perspective the now-abundant evidence linking physical health
and depression (Bruce, Seeman, Merrill, & Blazer, 1994;
Gurland, Wilder, & Berkman, 1988; Katon & Ciechanowski,
2002; Lenze et al., 2001) might have been anticipated.
However, an alternative interpretation has been proposed
based on the major correlates and consequences of depression. This perspective suggests that the reduced interest in
activities and relationships, sleep problems, and fatigue that
characterize depression tend to foster declines in physical
performance.
Most prior research has limited consideration to only one
of these alternative interpretations, presenting evidence that
physical disability is a risk factor for depressive symptoms
(Kennedy, Kelman, & Thomas, 1990; Roberts, Kaplan,
Shema, & Strawbridge, 1997; Zeiss, Lewinsohn, Rohde, &
Seeley, 1996) or that depressive symptomatology is a risk factor for physical disability (Bruce et al., 1994; Penninx, Leveille,
Ferrucci, van Eijk, & Guralnik, 1999; Tinetti, Inouye, Gill, &
Doucette, 1995). An early exception is the work of Aneshensel,
Frerichs, and Huba (1984). In a study of a community-based
sample of adults, they found evidence that physical illness and
depression exert reciprocal effects over time. More recently,
additional studies have considered the temporal and reciprocal
associations between physical limitations and depressive
symptoms, presenting results largely consistent with those of
Aneshensel and colleagues (Geerlings, Beekman, Deeg,

Twisk, & Van Tilburg, 2001; Meeks, Murrell, & Mehl, 2000;
Ormel, Rijsdijk, Sullivan, van Sonderen, & Kempen, 2002).
Although there are no clear grounds for doubting the
reciprocity of associations between physical limitations and
depressive symptomatology, several issues of both theoretical
and practical signiﬁcance remain to be resolved. Still in
question are (a) whether the observed associations are causal or
artifactual and (b) the relative strengths of the directional
predictions. Plausible sources of possible spuriousness include
pain, social stress, and major depressive disorder. In addition, it
is important to evaluate the extent to which intervening levels
of pain and stress exposure mediate the physical limitation–
depressive symptom relationships over time. This article explores the nature of the association between physical limitations
and depressive symptoms over a 3-year interval.
Studies have consistently observed an association between
physical limitations and depressive symptoms (Bruce et al.,
1994; Everson-Rose et al., 2005; Gurland et al., 1988; Katon &
Ciechanowski, 2002; Lenze et al., 2001; Turner, Lloyd, &
Taylor, 2006; Turner & McLean, 1989). However, because
many of these early investigations were cross-sectional, they
could not conﬁdently assess temporal order. Other recent
studies have employed longitudinal designs, allowing for the
speciﬁcation of the time ordering of physical functioning and
depressive symptoms.
For example, evidence has suggested that the presence and
severity of physical disability increases risk for depression
over time (Geerlings, Beekman, Deeg, & Van Tilburg, 2000;
Kennedy et al., 1990; Schieman & Plickert, 2007; Turner &
Noh, 1988; Yang & George, 2005). There is also substantial

S219

S220

GAYMAN ET AL.

evidence indicating that depressive symptoms can lead to
problems with physical health and functioning (Armenian,
Pratt, Gallo, & Eaton, 1998; Bruce et al., 1994; Cronin-Stubbs
et al., 2000; Dunlop, Manheim, Song, Lyons, & Chang, 2005;
Jiang, Tang, Futatsuka, & Zhang, 2004; Lenze et al., 2005;
Penninx, Deeg, van Eijk, Beekman, & Guralnik, 2000; Penninx
et al., 1998, 1999). These lagged analyses have provided evidence consistent with the conclusion that physical limitations
increase the risk of depression and vice versa. None of these
studies, however, speciﬁcally considered reciprocal associations between physical limitations and depressive symptoms.
Aneshensel and colleagues (1984) conducted one of the ﬁrst
investigations to evaluate this possibility. They found that
physical illness can lead to depression and that depression can
lead to physical illness over a 1-year interval, and more recent
ﬁndings have appeared to conﬁrm such a reciprocal relationship
(Geerlings et al., 2001; Ormel et al., 2002). It is of interest to
note that prior studies have also found that physical health and
limitations are stronger predictors of depressive symptomatology than vice versa (Aneshensel et al., 1984; Meeks et al.,
2000; Ormel et al., 2002). An important recent study by KelleyMoore and Ferraro (2005) examined reciprocal effects across
three waves of data, distinguishing disease from disability.
They were able to conﬁrm reciprocal effects across both
intervals between disease onset and depression. However, in
analyses that included assessments of both disability and
disease, they found no evidence of a predictive association
between disability and depression or vice versa.
Taken together, the bulk of available evidence suggests a
reciprocal association between physical limitations and depressive symptoms. However, researchers cannot answer with conﬁdence the question of whether these observed linkages imply
either mutual or directional causation. Not well considered in
prior work is the possibility that the association between physical
limitations and depressive symptoms might arise from their
shared association with one or more inﬂuential third variables.
For example, level of pain is associated with both physical
limitations (Bryant, Scarbro, & Baxter, 2001; Grigsby et al.,
1999; Lichtenstein, Dhanda, Cornell, Escalante, & Hazuda,
1998) and depressive symptoms (Bair, Robinson, Katon, &
Kroenke, 2003; Benjamin, Morris, McBeth, Macfarlane, &
Silman, 2000; Corruble & Guelﬁ, 2000; Ruoff, 1996). Certainly,
bodily pain limits the extent to which an individual can
engage in physical activity, and the evidence that high levels
of chronic pain are emotionally distressing conﬁrms general
expectation. The established link between pain and depression
noted above may also arise partly from the fact that bodily pain
and depression share physiological attributes (Bair et al., 2003).
Social stress is also associated with physical health limitations (Farmer & Ferraro, 1997; Lin & Ensel, 1989; Pavalko,
Mossakowski, & Hamilton, 2003; Ross & Mirowsky, 2001) and
is a major predictor of depressive symptoms (Bromberger &
Matthews, 1996; Chiriboga, Black, Aranda, & Markides, 2002;
Kessler, Mickelson, & Williams, 1999; Pearlin, Lieberman,
Menaghan, & Mullan, 1981; Taylor & Turner, 2002). There has
been substantial progress toward an understanding of the
mechanisms by which stress exposure gets translated into
increased risk for both physical and emotional health problems.
For example, evidence from research on allostatic load (e.g.,
Geronimus, Hicken, Keene, & Bound, 2006; McEwen &

Seeman, 1999; Seeman, Singer, Rowe, Horwitz, & McEwen,
1997) has speciﬁed a range of biomarkers that are affected by
social stress and that mediate the demonstrated linkage between
stress exposure and both morbidity and mortality. In the case
of depression, stress exposure inﬂuences neuroendocrine responses that are correlates or expressions of depressive symptoms (e.g., Nierop, Bratsikas, Zimmermann, & Ehlert, 2006;
Robles, Glaser, & Kiecolt-Glaser, 2005).
As noted elsewhere, most of what is known about the
signiﬁcance of social stress for health is based on studies that
have estimated differences in stress exposure solely in terms of
recent life events (Turner & Wheaton, 1995). In this context,
researchers argued more than a decade ago that the stress
hypothesis had not been effectively tested because stress exposure had not been adequately estimated (Turner, Wheaton, &
Lloyd, 1995). Recent research has demonstrated the utility of
going beyond recent life events to achieve more comprehensive
assessments by revealing more powerful predictions of depressive symptomatology (Turner & Avison, 2003; Turner &
Lloyd, 1999). Thus, an effective evaluation of possible
spuriousness requires consideration of a range of different
sources or types of social stress.
Finally, the observed relationship between major depressive
disorder and higher levels of depressive symptoms (Breslau,
1985; Cuijpers & Smit, 2004; Horwath, Johnson, Klerman, &
Weissman, 1992) is hardly surprising. When this fact is taken
with the well-established cross-sectional relationship between
depressive disorder and the presence of physical disability
(Armenian et al., 1998; Turner et al., 2006), the need to
consider the issue of spuriousness with respect to the physical
limitation–depressive symptom relationship seems apparent.
Assuming some degree of causality in the observed physical
limitation–depressive symptom relationships, a second question
that, aside from speculation (Bruce et al., 1994), has largely
escaped attention is that of what factors may mediate these
directional or bidirectional relationships. The present analyses
evaluated the reciprocal association between physical limitations and depressive symptoms over time. We assessed the
independent signiﬁcance of prior level of physical limitations in
predicting changes in depressive symptoms, and vice versa,
considering the mediating effects of intervening measures of
pain and stress exposure.

METHODS

Sample
The data employed in this study were from a two-wave (W1
and W2) panel study of Miami-Dade County residents that
included a substantial oversampling of individuals with a physical disability. A total of 10,000 randomly selected households
were screened with respect to gender, age, ethnicity, disability
status, and language preference. Using this sampling frame, the
study sample was drawn such that there were even numbers of
women and men, even numbers of people screened as having
a physical disability and those not, and equivalent numbers of
the four major ethnic groups comprising more than 90% of all
Miami-Dade County residents (non-Hispanic Whites, Cubans,
non-Cuban Hispanics, and African Americans). These proportions departed only slightly from those for the county as

PHYSICAL LIMITATIONS AND DEPRESSIVE SYMPTOMS

a whole. Well-trained and predominantly bilingual interviewers
administered computerized questionnaires in either English or
Spanish as preferred by each participant. The majority of
interviews took place in the homes of participants. However,
a small number of interviews were conducted at alternative sites
or by telephone when requested by participants.
A total of 1,986 ﬁrst-wave interviews were completed in
2000–2001 out of 2,420 potential participants released to ﬁeld
study staff. Using participant payments of $50, the study
achieved a success rate of 82%. Included were 1,086 adults
who screened as having no physical disability and 900 individuals who screened as having a disability. The oversampling
of individuals with a physical disability and the fact that the
nondisabled participants were group matched on race/ethnicity,
gender, and age resulted in a greater proportion of older
respondents than in the general population. Ages in the sample
ranged from 18 to 93, with a median of 59 (the median age of
the general population of Miami-Dade County in 2000 was
35.6; see Census, 2000, Summary File 1, Table P13). Given
this discrepancy in ages and the dramatic oversampling of individuals with a physical disability, we can hardly claim that our
sample was representative of the Miami-Dade County population. However, the sample was generally representative of
physically disabled individuals in the county and of their
gender, race/ethnicity, and age counterparts without a disabling
condition. Of the 900 who, within the screening process, were
reported by a family member as having activity limitations,
only 559 conﬁrmed this status within the actual interview.
Presumably, this discrepancy arose from differing views about
the level of activity limitation that deﬁnes disability.
Second-wave interviews of 1,495 W1 participants were completed approximately 3 years later. Excluding 100 W1 participants who had died and 59 who were too ill to be interviewed,
the second-wave success rate was 82.5%. Offering, when
necessary, participant fees of up to $100 dollars clearly
facilitated this achievement. Funding problems dictated the
necessity of stopping ﬁeldwork as soon as we were conﬁdent in
achieving an 80% success rate. As a result, the cases lost to
interview included 65 that had been located but interviews not
yet scheduled, and 53 who were still in the search process. Only
7.7% refused to participate in a second interview, and we gave
up on ﬁnding only 5%. A comparison between completed
participants and all others regardless of reason for exclusion
revealed several statistically signiﬁcant but rather small differences. Cases lost to W2 interviews, together with those left over
when ﬁeldwork was stopped, had previously reported lower
levels of social stress, socioeconomic status (SES), and pain,
and slightly higher levels of both physical limitations and
depressive symptoms. The magnitude of these differences
varied from 0.10 SD to slightly more than 0.30 SD. In our view,
the nature and magnitude of these differences do not seriously
challenge the representativeness of subpopulation employed in
these analyses or in any way suggest that selection biases may
have accounted for the results presented.
A total of 1,455 study participants provided valid data across
all study variables employed in the current analysis. This
included 785 women (54%), 343 non-Hispanic Whites (24%),
349 Cuban Americans (24%), 320 non-Cuban Hispanics (22%),
and 443 African Americans (30%). The average age of the
study respondents was 57 years (SD ¼ 17 years; range ¼ 18–93

S221

years). Forty W2 participants had missing data on one or more
variables considered in the analyses (3%). These participants
differed from those included only in reporting slightly higher
levels of depressive symptoms within the W2 interviews.

Measures
Physical limitations. —Problems with physical limitations
were indexed at both W1 and W2 by using a measure that
combined indicators of physical mobility, instrumental daily
activities, and basic activities of daily living. This approach
provided a relatively comprehensive picture of physical
abilities and limitations, capturing variations at both the severe
end and the more able end of the limitations spectrum. Drawing
from well-established scales (see Fries, Spitz, Kraines, &
Holman, 1980; Nagi, 1976; Rosow & Breslau, 1966), nine
items were used to assess physical mobility. Problems with
instrumental daily activities were assessed using two items
from Lawton and Brody’s (1969) scale, and activities of daily
living were assessed using eight items drawn from the work of
Katz and colleagues (Katz, Downs, Cash, & Grotz, 1970; Katz,
Ford, Moskowitz, Jackson, & Jaffee, 1963) and (Jette and
Deniston 1978; Jette, 1980). We should note that of these 19
items, 7 were asked only of persons who had conﬁrmed the
presence of a physical disability. These items referred to one’s
ability to do such things as dress oneself, take a bath/shower,
get out of bed, prepare one’s meals, and climb stairs. We
assigned individuals not presented with these items responses
of ‘‘easy to do’’ on all seven items. In order to assess the
potential impact of this decision, we ran all analyses using
only items that were asked of all study respondents; the results
were substantively identical to the results presented here. The
Appendix presents all items used to index physical limitations.
We coded the 19 items such that higher scores indicated greater
physical limitations. Because response categories for some
items involved a 4-point scale and those for other items
involved a 5-point scale, we transformed responses to each item
into standard scores, which we summed into our measure of
physical limitations. We then restandardized these scores for
ease of interpretation so that they had a mean of 0 and a
standard deviation of 1. We also standardized all other continuous measures in order to maintain consistency in interpretation. The internal reliability coefﬁcient for physical limitations
was .95 within both waves of data. We employed this measure
of the presence and severity of physical limitations for all but
descriptive analyses, which we report as unstandardized scores.
Depressive symptomatology. —Depressive symptoms were
assessed at both data points using the Center for Epidemiologic
Studies–Depression scale (CES-D), which is a highly reliable
and widely used measure of depressive symptomatology
(Radloff, 1977). From the original 20-item CES-D index, 6
items involving somatic problems were omitted from all
analyses to avoid measurement confounding between physical
limitations and depressive symptoms (Ormel et al., 2002).
These omitted items included problems with eating, keeping
your mind on what you are doing, effort, restlessness, talking,
and getting going. In completing the CES-D items, participants
were asked about their experiences over the past month rather
than the past week using response categories of ‘‘not at all,’’

S222

GAYMAN ET AL.

‘‘occasionally,’’ ‘‘frequently,’’ and ‘‘almost all the time.’’ We
coded items such that higher scores indicated higher levels of
depressive symptoms, then summed and standardized scores for
analysis (W1 a ¼ .81; W2 a ¼ .84).
Pain. —Study participants were asked whether they had
experienced any bodily pain over the preceding 4 weeks. Those
who answered afﬁrmatively were questioned on the frequency
and average intensity of such pain. Response categories for the
intensity item ranged from ‘‘very mild’’ (1) to ‘‘very severe’’
(5). Those for frequency ranged from ‘‘once or twice’’ (1) to
‘‘everyday or almost everyday’’ (5). We coded respondents
who reported no pain over the past 4 weeks as ‘‘none’’ (0) for
intensity of pain and ‘‘never’’ (0) for frequency of pain. We
coded both items so that higher values indicated greater
intensity and frequency of pain. We derived the pain measure
by multiplying scores on the two dimensions and then
standardizing the resulting products.

Table 1. Description of Study Variables
Wave 1
Variable

M (SD)

Physical limitations
9.64 (12.34)
Depressive symptoms
21.60 (6.61)
Lifetime major depressiona 0.09
Pain
4.34 (6.90)
Chronic stress
4.40 (4.92)
Daily discrimination
13.62 (5.02)
Recent life events
1.05 (1.50)

Wave 2
Range

M (SD)

Range

1–83
0–53

9.85 (12.49) 1–82
21.50 (6.65) 4–49

0–25
0–27
0–39
0–12

2.83
5.93
13.63
0.89

(4.93)
(5.35)
(4.92)
(1.33)

0–20
0–31
4–39
0–9

Demographic controls
Femalea
0.54
Non-Hispanic Whitea
0.24
Cubana
0.24
Non-Cuban Hispanica
0.24
African Americana
0.30
Age
57.03 (16.75)
18–93
Socioeconomic statusb
0 (1)
À2.72–2.60
Notes: Shown are unstandardized mean scores (SD) N ¼ 1,455.
Mean scores should be interpreted as proportion of individuals who met
criteria.
b
Standardized score for summed occupational prestige, education, and
household income.
a

Social stress. —The role of stress exposure in changes in
physical limitations and depressive symptoms was evaluated
using measures of chronic stress, daily discrimination, and
recent life events as occurring during the approximately 3-year
interval between interviews. The chronic stress measure was an
adaptation Wheaton’s (1991, 1994) scale, modiﬁed to better
capture the kinds of enduring stressors older individuals are
likely to experience. Daily discrimination (9 items) and recent
life events (32 items, past 12 months) were indexed by
previously employed measures (Turner & Avison, 2003;
Williams, Yu, Jackson, & Anderson, 1997). We standardized
all measures of social stress for analysis.
Lifetime major depression. —The lifetime occurrence of
major depression was measured using the depression module of
the World Mental Health version of the Composite International Diagnostic Interview (Kessler & Ustun, 2004). This
instrument has been described in greater detail elsewhere
(Turner et al., 2006). In the present analyses we employed
diagnostic estimates derived from the W1 interviews.
Sociodemographic variables. —We used W1 data to control
for age, gender, race/ethnicity, and SES. We measured age as a
continuous variable. We based estimates of SES on a composite score that equally weighted the occupational level
(Hollingshead, 1957), educational attainment, and household
income of each participant. To avoid problems of missing data,
particularly on the variable of household income, we
standardized scores on each of these dimensions, summed
them, and divided them by the number of dimensions on which
data were available. Table 1 provides descriptive statistics for
all study variables.

Analysis Plan
Our analyses began with a correlation matrix in order to
show bivariate associations for key study variables. Using
structural equation modeling (SEM), we then tested for
reciprocal effects between physical limitations and depressive
symptoms over a 3-year period. We then employed this
modeling process in order to rule out the possibility of
spuriousness stemming from prior pain, stress exposure, and

lifetime major depression as assessed at W1. Finally, we used
SEM to evaluate the potential mediating effects of pain and
social stress assessed at W2 on the physical limitation–
depressive symptom relationship.

RESULTS
Table 2 provides the correlations among the study variables,
including physical limitations, depressive symptoms, lifetime
major depression, pain, and social stress. The table reveals
several important ﬁndings. First, both physical limitations and
depressive symptoms showed signiﬁcant stability across time
(i.e., r ¼ .70, p , .001, for physical limitations from W1 to
W2). Second, the associations between physical limitations and
depressive symptoms were signiﬁcant for each wave as well as
across waves (i.e., r ¼ .16, p , .001 between physical
limitations at W1 and depressive symptoms at W2). Furthermore, the measures of lifetime major depression, pain, and
stress were signiﬁcantly associated with both depressive
symptoms and physical limitations. These correlations provided
initial support for our hypotheses involving reciprocity,
spuriousness, and mediation. Based on these ﬁndings, we
then estimated the SEM.
We used AMOS software for the SEM (Arbuckle, 2005).
Figure 1 shows the results of our SEM analysis relating to the
hypothesized reciprocal effects between physical limitations
and depressive symptoms. We included age, gender, race/
ethnicity, and SES as control variables but do not show them so
that we can more clearly convey the primary ﬁndings. Among
the control variables in the model, SES and age were negatively
associated with depressive symptoms at W2. SES was also
negatively associated with physical limitations at W2, and
age was positively associated with physical limitations at
W2. Being female was positively associated with physical
limitations at W2, and both Hispanic groups reported
signiﬁcantly more depressive symptoms at W2 compared to
their non-Hispanic White counterparts.

PHYSICAL LIMITATIONS AND DEPRESSIVE SYMPTOMS

S223

Table 2. Correlation Matrix for Study Variables
Variable
a

DS
DSb
PLa
PLb
LMDa
Paina
Painb
CSa
CSb
DDa
DDb
RLEa
RLEb

DS

a

—
.35**
.21**
.14**
.24**
.11**
.12**
.25**
.07**
.13**
.02
.18**
.07**

b

a

DS

PL

PLb

LMDa

Paina

Painb

CSa

CSb

DDa

DDb

RLEa

RLEb

—
.16**
.23**
.15**
.09**
.14**
.11**
.03
.00
.06*
.04
.05*

—
.70**
.16**
.57**
.40**
.04
À.15**
.09**
.02
.04
.00

—
.08**
.38**
.53**
À.03
À.19**
.00
.04
.01
.02

—
.17**
.11**
.29**
.08**
.12**
.02
.24**
.09**

—
.41**
.17**
À.05*
.12**
.02
.13**
.05*

—
.05*
.01
.03
.05*
.09**
.13**

—
.29**
.33**
.11**
.46**
.18**

—
.16**
.29**
.14**
.37**

—
.34**
.28**
.14**

—
.13**
.17**

—
.17**

—

Notes: N ¼ 1,455. DS ¼ depressive symptoms (Center for Epidemiologic Studies–Depression scale); PL ¼ physical limitations; LMD ¼ lifetime major depression;
Pain ¼ intensity* 3 frequency of bodily pain; CS ¼ chronic stress; DD ¼ daily discrimination; RLE ¼ recent life events.
a
Measured at wave 1.
b
Measured at wave 2.
*p .05; **p .01.

For our model, the chi-square was v2 (5 N ¼ 1,455) ¼ 12.26,
p ¼ .03. The Tucker–Lewis Index (TLI) was .98, the
comparative ﬁt index (CFI) was .99, and the root mean square
error of approximation was .03. These ﬁt indexes all indicated
a good ﬁt of the model to the data (Kline, 2005).
Figure 1 illustrates several ﬁndings. First, the stability
coefﬁcients for depressive symptoms and physical limitations
from W1 to W2 were both signiﬁcant (b ¼ .27, p , .001, for
depressive symptoms; and b ¼ .66, p , .001, for physical
limitations). Second and more important, although the level of
physical limitations at W1 predicted changes in depressive
symptoms (b ¼ .11, p , .001), these results offered no evidence
that prior depressive symptoms were of signiﬁcance in
predicting changes in physical limitations. We thus concluded
that most of whatever causation may have been involved in the
clear relationship between physical limitations and depressive
symptomatology over a 3-year period ﬂowed from physical
limitations to depressive symptoms rather than in the reverse
direction. Although the magnitude of the cross-lagged coefﬁcient for W1 physical limitations was rather modest, the
twin facts that this relationship was net of variance shared with
depression cross-sectionally and that the prediction involved
spans of approximately 3 years suggested substantive as well as
statistical signiﬁcance.
Based on the ﬁndings from Figure 1, the remaining analyses
focused on W1 physical limitations predicting changes in

depressive symptoms. To test the possibility that the physical
limitation–depression relationship may have been partially or
largely an artifact of associations between these variables and
prior levels of pain, stress exposure, and lifetime major depression, we added these variables to the model (see Figure 2)
to predict both physical limitations at W1 and depressive
symptoms at W2. Results indicated that with pain, stress, and
major depression all in the model, the path from physical
limitations at W1 to depressive symptoms at W2 remained
statistically signiﬁcant and virtually unchanged (b ¼ .10, p ,
.001). We concluded that the association between prior physical
limitations and changes in depressive symptoms was not an
artifact of their shared association with pain, stress exposure, or
lifetime major depression.
Next, we tested the mediating hypothesis. We proposed that
the association between physical limitations and depressive
symptoms would be mediated by our indicators of pain and
social stress measured at W2. Figure 3 presents these ﬁndings.
As in Figure 1, we included control variables in the model but
do not show them here. The model also indicated a reasonable
ﬁt to the data v2 (19 N ¼ 1,455) ¼ 55.91, p ¼ .00 (TLI ¼ .95;
CFI ¼ .99; RMSEA ¼ .04).
Results indicated that physical limitations at W1 predicted
pain, chronic stress, and daily discrimination at W2. Pain and
daily discrimination also both predicted changes in depressive
symptoms. The direct path from physical limitations at W1

Figure 1. Reciprocal effects between depressive symptoms and physical limitations. Standardized regression coefﬁcients are shown. N ¼ 1,455.
Signiﬁcant effects for age, gender, race/ethnicity, and socioeconomic status are controlled for in the model but not shown. Model ﬁt: v2 ¼ 12.26,
df ¼ 5; Tucker–Lewis Index ¼ .98; comparative ﬁt index ¼ .99; root mean square error of approximation ¼ .03. W ¼ wave; e1 ¼ error 1; e2 ¼ error 2.

S224

GAYMAN ET AL.

Figure 2. Physical limitations predicting changes in depressive symptoms: spurious effects of pain, social stress, and lifetime major depression.
Standardized regression coefﬁcients are shown (v2 ¼ .00, df ¼ 0). N ¼ 1,455. W ¼ wave. *p .05; **p .01.

to depressive symptoms at W2 was reduced from .11 (see
Figure 1) to .09 (p , .01, in Figure 3). With respect to
mediating effects, the Sobel test (Sobel, 1982) yielded a highly
signiﬁcant z score of 3.03 for pain. However, neither chronic
stress nor recent life events were signiﬁcant mediators, whereas
the mediation effect of daily discrimination was marginally
signiﬁcant statistically (z ¼ 1.87, p , .10). We tested this
mediating effect after controlling for depressive symptoms at
W1 (b ¼ .25, p , .001). Taken together, the results
demonstrated support for the mediating effects of pain and,
to a lesser extent, daily discrimination on the association
between prior physical limitations and changes in depressive
symptoms.

DISCUSSION
This study evaluated the contention that the association
between physical limitations and depressive symptoms arises
from causal reciprocity, assessed the possibility that this
relationship is artifactual, and assessed potential mediating
effects of related third variables. Consistent with prior research
(Aneshensel et al., 1984; Ormel et al., 2002), we found that
prior physical limitations predict changes in depressive
symptoms. However, we observed no such symmetry in the
capacity of W1 depressive symptoms to predict changes in
physical limitations over a 3-year period. That is, although W1
physical limitations predicted changes in depressive symptoms,
there was no hint of a reverse association. Accordingly, we

Figure 3. Physical limitations predicting changes in depressive symptoms: mediating effects of pain and social stress. Standardized regression
coefﬁcients are shown. N ¼ 1,455. Signiﬁcant effects for age, gender, race/ethnicity, and socioeconomic status are controlled for in the model but
not shown. Model ﬁt: v2 ¼ 55.91, df ¼ 19; Tucker–Lewis Index ¼ .95; comparative ﬁt index ¼ .99; root mean square error of approximation ¼ .04.
Mediation effects using Sobel test: pain, z ¼ 3.03**; chronic stress, z ¼À0.74; daily discrimination, z ¼ 1.87; recent life events, z ¼ 0.73. W ¼ wave.
*p .05; **p .01.

PHYSICAL LIMITATIONS AND DEPRESSIVE SYMPTOMS

argue that much of whatever causation may be involved in the
linkage between physical limitations and depressive symptomatology over a 3-year period ﬂows from limitations to
depression rather than in the reverse direction. These ﬁndings
are consistent with the view that physical limitations represent
a source of chronic or enduring stress (Turner & McLean, 1989;
Turner & Noh, 1988) and the observation that chronic stress is
a particularly strong predictor of depressive symptoms (Turner
et al., 1995).
The failure to observe either an association between prior
depressive symptomatology or prior depressive disorder and
changes in physical limitations might be seen as contradicting
an array of previous studies (see Armenian et al., 1998; Bruce
et al., 1994; Cronin-Stubbs et al., 2000; Dunlop et al., 2005;
Jiang et al., 2004; Lenze et al., 2005; Penninx et al., 1998,
1999, 2000). However, most such studies focused on the
capacity of prior depressive symptoms (or depressive disorder)
to predict the onset of disability rather than changes in the level
of physical limitations. Failing to conﬁrm this body of evidence
may well arise from the fact that the variance shared by prior
physical limitations and depressive symptoms is held constant
in analyses addressing changes in limitations over time. In
addition, the relatively high stability coefﬁcient observed for
physical limitations means that there was more variability in
depressive symptoms remaining to be explained. Although this
difference in stability of measures may have impacted results to
some extent, it does not, in our view, challenge the clear
ﬁndings presented. The results presented here are consistent
with those reported from a population study of elderly biracial
participants in which prior depressive symptoms were not
associated with declines in physical performance (EversonRose et al., 2005). However, readers should bear in mind that
the process by which depression may get translated into an
increase in physical limitations may take longer than the 3-year
follow-up period studied here. As Kelley-Moore and Ferraro
(2005) argued, ‘‘Health decline is a process that occurs over
several years rather than instantaneously’’ (p. 386).
Although a large body of research has demonstrated a robust
correlation between physical limitations and depressive symptoms, few studies have considered the extent to which this
apparent linkage might be spurious—that is, driven by factors
known to be associated with both indicators of health. We
examined the relationship between prior levels of physical
limitations and changes in depressive symptoms in the context
of measures of pain, stress exposure, and prior lifetime
occurrence of major depressive disorder. After controlling for
these potential sources of spuriousness, we found that the
magnitude of the coefﬁcient representing the association
remained strong and statistically signiﬁcant. This result allows
for increased conﬁdence that this frequently observed relationship is real rather than artifactual in nature.
The study also presents evidence of possible mechanisms by
which physical limitations are translated into increased mental
health risk. We found strong evidence to support the conclusion
that the relationship between prior physical limitations and
subsequent changes in depressive symptoms is mediated by
intervening levels of pain and, to a lesser extent, daily
discrimination. These results suggest that physical limitations
may partially be translated into increased risk for depressive
symptoms through a process of stress generation analogous to

S225

that described by Hammen (1991, 2006). That is, in addition to
having direct effects, physical limitations appear to indirectly
inﬂuence depressive symptoms by elevating risk for higher
levels of pain and exposure to social stress. Among the stress
measures, daily discrimination appears the most salient. This
may be a reﬂection of experiences of stigma associated with
being physically limited (Crandall & Moriarty, 1995) or may
arise from the fact that approximately three quarters of the
sample was drawn from racial/ethnic minority populations.
Several limitations to the current investigation are worthy of
note. For one thing, it employed self-report physical limitation
measures rather than objective assessments of physical
performance. Although it is possible that objective measures
of physical limitations might yield different results, prior research provides moderate to strong evidence for the validity of
such self-reports evaluated with respect to objectively assessed
physical limitations (Cress et al., 1995; Judge, Schechtman, &
Cress, 1996; Kempen, Sullivan, van Sonderen, & Ormel, 1999;
Myers, Holliday, Harvey, & Hutchinson, 1993). Also reassuring is the fact that these results are generally consistent with
prior research employing objective measures of physical
performance (see Everson-Rose et al., 2005).
Our consideration of the possibility that the physical
limitation–depressive symptom linkage may arise from shared
association with inﬂuential third variables focused on several
factors that prior research has shown to be associated with both
physical limitations and depressive symptoms. Clearly, these
analyses are far from exhaustive, being limited to self-reports of
bodily pain, stress exposure, and lifetime major depression. We
also considered only a limited number of potential mediators
(i.e., pain, chronic stress, daily discrimination, and recent life
events). Our analyses omitted the evaluation of nondisabling
medical conditions, cognitive functioning (Bruce, 2001),
availability of instrumental support (Kennedy et al., 1990),
and several personal resources such as mastery and self-esteem
that are relevant to depression (Turner & Lloyd, 1999; Turner,
Taylor, & Van Gundy, 2004).
Such limitations notwithstanding, the present study builds
on prior research in several important ways. First, it offers
a prospective assessment of the potential reciprocal relationships between physical limitations and depressive symptoms
based on a community-based sample controlled on gender, age,
SES, racial/ethnic variations, and the prior lifetime occurrence
of one or more major depressive episodes. Although several
prior studies assessing the temporal and reciprocal associations
between physical limitations and depressive symptoms have
involved community samples (see Geerlings et al., 2001;
Meeks et al., 2000; Ormel et al., 2002), these studies focused on
participants of a limited age range and a limited range of racial/
ethnic diversity. Second, this study contributes to increased
conﬁdence that the physical limitation–depression relationship
is not spurious, being an association of potential theoretical and
practical signiﬁcance.
In the context of the long-standing argument that physical
limitations and pain represent important dimensions of chronic
stress (Gurland et al., 1988; Turner & Noh, 1988), it seemed
crucial to evaluate the mental health signiﬁcance of physical
limitations in the context of other sources of social stress. Our
results leave little doubt that over a 3-year period, physical
limitations represent a signiﬁcant risk factor for depressive

S226

GAYMAN ET AL.

symptomatology independent of the effects of other sources of
stress exposure, and that this presumably causal linkage
accounts for much of the often reported cross-sectional and
longitudinal relationship between physical limitations and
depressive symptoms.
ACKNOWLEDGMENTS
An earlier version of this article was presented at the 102nd Annual
Meeting of the American Sociological Association, New York, in August
2007. We are grateful to the reviewers for their conscientious and helpful
critiques of our paper. This work was supported by Grants R01DA13292
and R01DA16429 from the National Institute on Drug Abuse to R. Jay
Turner.
M. D. Gayman conducted all original data analysis and wrote the paper.
R. J. Turner contributed to revising the paper. M. Cui contributed to the
ﬁnal data analysis.
CORRESPONDENCE
Address correspondence to Mathew D. Gayman, Cecil G. Sheps Center
for Health Services Research, University of North Carolina at Chapel Hill,
725 Martin Luther King Jr Boulevard, CB#7590, Chapel Hill, NC 275997590. E-mail: mgayman@schsr.unc.edu
REFERENCES
Aneshensel, C. S., Frerichs, R. R., & Huba, G. J. (1984). Depression and
physical illness: A multiwave, nonrecursive causal model. Journal of
Health and Social Behavior, 25, 350–371.
Arbuckle, J. L. (2005). AMOS 6.0 user’s guide. Chicago, IL: SPSS.
Armenian, H., Pratt, L. A., Gallo, J., & Eaton, W. W. (1998).
Psychopathology as a predictor of disability: A population-based
follow-up study in Baltimore, Maryland. American Journal of
Epidemiology, 148, 269–275.
Bair, M. J., Robinson, R. L., Katon, W., & Kroenke, K. (2003). Depression
and pain comorbidity: A literature review. Archives of Internal
Medicine, 163, 2433–2445.
Benjamin, S., Morris, S., McBeth, J., Macfarlane, G. J., & Silman, A. J.
(2000). The association between chronic widespread pain and mental
disorder. Arthritis & Rheumatism, 43, 561–567.
Breslau, N. (1985). Depressive symptoms, major depression, and
generalized anxiety: A comparison of self-reports on CES-D and
results from diagnostic interviews. Psychiatry Research, 15(3),
219–229.
Bromberger, J. T., & Matthews, K. A. (1996). A longitudinal study of the
effects of pessimism, trait anxiety, and life stress on depressive
symptoms in middle-aged women. Psychology and Aging, 11, 207–213.
Bruce, M. L. (2001). Depression and disability in late life: Directions for
future research. American Journal of Geriatric Psychiatry, 9(2),
102–112.
Bruce, M. L., Seeman, T. E., Merrill, S. S., & Blazer, D. G. (1994). The
impact of depressive symptomatology on physical disability: MacArthur Studies of Successful Aging. American Journal of Public Health,
84, 1796–1799.
Bryant, L. L., Scarbro, S., & Baxter, J. (2001). Pain, chronic conditions,
and worsening physical performance. The Gerontologist, 41(Special
Issue I), 290–291.
Chiriboga, D. A., Black, S. A., Aranda, M., & Markides, K. (2002). Stress
and depressive symptoms among Mexican American elders. Journal of
Gerontology: Psychological Sciences, 57B, P559–P568.
Corruble, E., & Guelﬁ, J. D. (2000). Pain complaints in depressed
inpatients. Psychopathology, 33, 307–309.
Crandall, C. S., & Moriarty, D. (1995). Physical illness stigma and social
rejection. British Journal of Social Psychology, 34, 67–83.
Cress, M. E., Schechtman, K. B., Mulrow, C. D., Fiatarone, M. A., Gerety,
M. B., & Buchner, D. M. (1995). Relationship between physical
performance and self-perceived physical function. Journal of the
American Geriatrics Society, 43, 93–101.
Cronin-Stubbs, D., Mendes de Leon, C. F., Beckett, L. A., Field, T. S.,
Glynn, R. J., & Evans, D. A. (2000). Six-year effect of depressive

symptoms on the course of physical disability in community-living
older adults. Archives of Internal Medicine, 160, 3074–3080.
Cuijpers, P., & Smit, F. (2004). Subthreshold depression as a risk indicator
for major depressive disorder: A systematic review of prospective
studies. Acta Psychiatrica Scandinavica, 109, 325–331.
Dunlop, D. D., Manheim, L. M., Song, J., Lyons, J. S., & Chang, R. W.
(2005). Incidence of disability among preretirement adults: The impact
of depression. American Journal of Public Health, 95, 2003–2008.
Everson-Rose, S., Skarupski, K. A., Bienias, J. L., Wilson, R. S., Evans,
D. A., & Mendes de Leon, C. F. (2005). Do depressive symptoms
predict declines in physical performance in an elderly, biracial
population? Psychosomatic Medicine, 67, 609–615.
Farmer, M. M., & Ferraro, K. F. (1997). Distress and perceived health:
Mechanisms of health decline. Journal of Health and Social Behavior,
38, 298–311.
Fries, J. F., Spitz, P., Kraines, R. G., & Holman, H. R. (1980).
Measurement of patient outcomes in arthritis. Arthritis and Rheumatism, 23(2), 137–145.
Geerlings, S. W., Beekman, A. T. F., Deeg, D. J. H., Twisk, J. W. R., & Van
Tilburg, W. (2001). The longitudinal effect of depression on functional
limitations and disability in older adults: An eight-wave prospective
community-based study. Psychological Medicine, 31, 1361–1371.
Geerlings, S. W., Beekman, A. T. F., Deeg, D. J. H., & Van Tilburg, W.
(2000). Physical health and the onset and persistence of depression in
older adults: An eight-wave prospective community-based study.
Psychological Medicine, 30, 369–380.
Geronimus, A. T., Hicken, M., Keene, D., & Bound, J. (2006). ‘‘Weathering’’
and age patterns of allostatic load scores among blacks and whites in the
United States. American Journal of Public Health, 96, 826–833.
Grigsby, J., Morgenstern, N. E., Kaye, K., Shetterly, S. M., Baxter, J., &
Hamman, R. F. (1999). Effects of pain on physical functioning and ADL/
IADL capacity: Results from the San Luis Valley Health and Aging
Study [Abstract]. Journal of the American Geriatrics Society, 47, S13.
Gurland, B. J., Wilder, D. E., & Berkman, C. (1988). Depression and
disability in the elderly: Reciprocal relations and changes with age.
International Journal of Geriatric Psychiatry, 3, 163–179.
Hammen, C. (1991). The generation of stress in the course of unipolar
depression. Journal of Abnormal Psychology, 100, 555–561.
Hammen, C. (2006). Stress generation in depression: Reﬂections on origins,
research, and future directions. Journal of Clinical Psychology, 62,
1065–1082.
Hollingshead, A. B. (1957). Two factor index of social position. New
Haven, CT: Author.
Horwath, E., Johnson, J., Klerman, G. L., & Weissman, M. M. (1992).
Depressive symptoms as relative and attributable risk-factors for 1stonset depression. Archives of General Psychiatry, 49, 817–823.
Jette, A. M. (1980). Functional status index: Reliability of a chronic disease
evaluation instrument. Archives of Physical Medicine and Rehabilitation, 61, 395–401.
Jette, A. M., & Deniston, O. L. (1978). Inter-observer reliability of
a functional status assessment instrument. Journal of Chronic Diseases,
31, 573–580.
Jiang, J., Tang, Z., Futatsuka, M., & Zhang, K. (2004). Exploring the
inﬂuence of depressive symptoms on physical disability: A cohort study
of elderly in Beijing, China. Quality of Life Research, 13, 1337–1346.
Judge, J. O., Schechtman, K., & Cress, E. (1996). The relationship between
physical performance measures and independence in instrumental
activities of daily living. Journal of the American Geriatrics Society,
44, 1332–1441.
Katon, W., & Ciechanowski, P. (2002). Impact of major depression on
chronic medical illness. Journal of Psychosomatic Research, 53,
859–863.
Katz, S., Downs, T. D., Cash, H. R., & Grotz, R. C. (1970). Progress in
development of the index of ADL. The Gerontologist, 10, 20–30.
Katz, S., Ford, A. B., Moskowitz, R. W., Jackson, B. A., & Jaffee, M. W.
(1963). Studies of illness in the aged: The index of ADL. A
standardized measure of biological and psychosocial function. Journal
of the American Medical Association, 185, 914–919.
Kelley-Moore, J. A., & Ferraro, K. F. (2005). A 3-D model of health
decline: Disease, disability, and depression among black and white
older adults. Journal of Health and Social Behavior, 46, 376–391.
Kempen, G. I. J. M., Sullivan, M., van Sonderen, E., & Ormel, J. (1999).
Performance-based and self-reported physical functioning in low-

PHYSICAL LIMITATIONS AND DEPRESSIVE SYMPTOMS

functioning older persons: Congruence of change and the impact of
depressive symptoms. Journal of Gerontology: Psychological Sciences,
54B, P380–P386.
Kennedy, G. J., Kelman, H., & Thomas, C. (1990). The emergence of
depressive symptoms in late life: The importance of declining health
and increasing disability. Journal of Community Health, 15(2), 93–104.
Kessler, R. C., Mickelson, K. D., & Williams, D. R. (1999). The
prevalence, distribution, and mental health correlates of perceived
discrimination in the United States. Journal of Health and Social
Behavior, 40, 208–230.
Kessler, R. C., & Ustun, T. B. (2004). The World Mental Health (WMH)
survey initiative version of the World Health Organization (WHO)
Composite International Diagnostic Interview (CIDI). International
Journal of Methods in Psychiatric Research, 13, 93–121.
Kline, R. B. (2005). Principles and practice of structural equation
modeling (2nd ed.). New York: Guilford Press.
Lawton, P. M., & Brody, E. M. (1969). Assessment of older people: Selfmaintaining and instrumental activities of daily living. The Gerontologist, 9, 179–186.
Lenze, E. J., Rogers, J. C., Martire, L. M., Mulsant, B. H., Rollman, B. L.,
Dew, M. A., et al. (2001). The association of late-life depression and
anxiety with physical disability—A review of the literature and
prospectus for future research. American Journal of Geriatric
Psychiatric, 9(2), 113–135.
Lenze, E. J., Schulz, R., Martire, L. M., Zdaniuk, B., Glass, T., Kop, W. J.,
et al. (2005). The course of functional decline in older people with
persistently elevated depressive symptoms: Longitudinal ﬁndings from
the Cardiovascular Health Study. Journal of the American Geriatrics
Society, 53, 569–575.
Lichtenstein, M. J., Dhanda, R., Cornell, J. E., Escalante, A., & Hazuda,
H. P. (1998). Disaggregating pain and its effect on physical functional
limitations. Journal of Gerontology: Medical Sciences, 53A, M361–
M371.
Lin, N., & Ensel, W. M. (1989). Life stress and health: Stressors and
resources. American Sociological Review, 54, 382–399.
McEwen, B. S., & Seeman, T. (1999). Protective and damaging effects of
mediators of stress: Elaborating and testing the concepts of allostasis and
allostatic load. Annals of the New York Academy of Sciences, 896, 30–47.
Meeks, S., Murrell, S. A., & Mehl, R. C. (2000). Longitudinal relationships
between depressive symptoms and health in normal older and middleaged adults. Psychology and Aging, 15, 100–109.
Myers, A. M., Holliday, P. J., Harvey, K. A., & Hutchinson, K. S. (1993).
Functional performance measures: Are they superior to self-assessments? Journal of Gerontology: Medical Sciences, 48, M196–M206.
Nagi, S. Z. (1976). An epidemiology of disability among adults in the
United States. Milbank Memorial Fund Quarterly: Health and Society,
54, 439–467.
Nierop, A., Bratsikas, A., Zimmermann, R., & Ehlert, U. (2006). Are stressinduced cortisol changes during pregnancy associated with postpartum
depressive symptoms? Psychosomatic Medicine, 68, 931–937.
Ormel, J., Rijsdijk, F. V., Sullivan, M., van Sonderen, E., & Kempen, G. I.
J. M. (2002). Temporal and reciprocal relationship between IADL/ADL
disability and depressive symptoms in late life. Journal of Gerontology:
Psychological Sciences, 57B, P338–P347.
Pavalko, E. K., Mossakowski, K. N., & Hamilton, V. J. (2003). Does
perceived discrimination affect health? Longitudinal relationships
between work discrimination and women’s physical and emotional
health. Journal of Health and Social Behavior, 44, 18–33.
Pearlin, L. I., Lieberman, M., Menaghan, E., & Mullan, J. (1981). The
stress process. Journal of Health and Social Behavior, 22, 337–356.
Penninx, B. W. J. H., Deeg, D. J. H., van Eijk, J. T. M., Beekman,
A. T. F., & Guralnik, J. M. (2000). Changes in depression and physical
decline in older adults: A longitudinal perspective. Journal of Affective
Disorders, 61, 1–12.
Penninx, B. W. J. H., Guralnik, J. M., Ferrucci, L., Simonsick, E. M., Deeg,
D. J. H., & Wallace, R. B. (1998). Depressive symptoms and physical
decline in community-dwelling older persons. Journal of the American
Medical Association, 279, 1720–1726.
Penninx, B. W. J. H., Leveille, S., Ferrucci, L., van Eijk, J. T. M., &
Guralnik, J. M. (1999). Exploring the effect of depression on physical
disability: Longitudinal evidence from the Established Populations for
Epidemiologic Studies of the Elderly. American Journal of Public
Health, 89, 1346–1352.

S227

Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for
research in the general population. Applied Psychosocial Measurement,
1, 385–401.
Roberts, R. E., Kaplan, G. A., Shema, S. J., & Strawbridge, W. J. (1997).
Does growing old increase the risk for depression? American Journal of
Psychiatry, 154, 1384–1390.
Robles, T. F., Glaser, R., & Kiecolt-Glaser, J. K. (2005). Out of balance: A
new look at chronic stress, depression, and immunity. Current
Directions in Psychological Science, 14(2), 111–115.
Rosow, I., & Breslau, N. (1966). A Guttman Health Scale for the aged.
Journal of Gerontology, 21, 556–559.
Ross, C. E., & Mirowsky, J. (2001). Neighborhood disadvantage, disorder,
and health. Journal of Health and Social Behavior, 42, 258–276.
Ruoff, G. E. (1996). Depression in the patient with chronic pain. Journal of
Family Practice, 43(6), S25–S33.
Schieman, S., & Plickert, G. (2007). Functional limitations and changes in
levels of depression among older adults: A multiple-hierarchy stratiﬁcation perspective. Journal of Gerontology: Social Sciences, 62B, S36–S42.
Seeman, T. E., Singer, B. H., Rowe, J. W., Horwitz, R. I., & McEwen, B. S.
(1997). Price of adaptation—allostatic load and its health consequences.
Archives of Internal Medicine, 157, 2259–2268.
Sobel, M. E. (1982). Asymptotic intervals for indirect effects in structural
equations models. In S. Leinhart (Ed.), Sociological methodology (pp.
290–312). San Francisco: Jossey-Bass.
Taylor, J., & Turner, R. J. (2002). Perceived discrimination, social stress,
and depression in the transition to adulthood: Racial contrasts. Social
Psychology Quarterly, 65, 213–225.
Tinetti, M. E., Inouye, S. K., Gill, T. M., & Doucette, J. T. (1995). Shared
risk factors for falls, incontinence, and functional dependence: Unifying
the approach to geriatric syndromes. Journal of the American Medical
Association, 273, 1348–1353.
Turner, R. J., & Avison, W. R. (2003). Status variations in stress exposure:
Implications for the interpretation of research on race, socioeconomic
status, and gender. Journal of Health and Social Behavior, 44, 488–505.
Turner, R. J., & Lloyd, D. A. (1999). The stress process and the social
distribution of depression. Journal of Health and Social Behavior, 40,
374–404.
Turner, R. J., Lloyd, D. A., & Taylor, J. (2006). Physical disability and
mental health: An epidemiology of psychiatric and substance use
disorders. Rehabilitation Psychology, 51(3), 214–223.
Turner, R. J., & McLean, P. D. (1989). Physical disability and
psychological distress. Rehabilitation Psychology, 34(4), 225–242.
Turner, R. J., & Noh, S. (1988). Physical disability and depression: A
longitudinal analysis. Journal of Health and Social Behavior, 29, 23–37.
Turner, R. J., Taylor, J., & Van Gundy, K. (2004). Personal resources and
depression in the transition to adulthood: Ethnic comparisons. Journal
of Health and Social Behavior, 45, 34–52.
Turner, R. J., & Wheaton, B. (1995). Checklist measures of stressful life
events. In L. Gordon, S. Cohen, & R. Kessler (Eds.), Measuring stress:
A guide for health and social scientists (pp. 29–58). Oxford, England:
Oxford University Press.
Turner, R. J., Wheaton, B., & Lloyd, D. A. (1995). The epidemiology of
social stress. American Sociological Review, 60, 104–125.
U.S. Census Bureau. (2000). Miami-Dade County, Summary File 1, Table 13.
Wheaton, B. (1991). The speciﬁcation of chronic stress: Models and
measurement. Cincinnati, OH: Annual Meeting of the Society for the
Study of Social Problems. Paper Presentation, August.
Wheaton, B. (1994). Sampling the stress universe. In W. R. Avison & I. H.
Gotlib (Eds.), Stress and mental health (pp. 77–115). New York:
Plenum Press.
Williams, D. R., Yu, Y., Jackson, J. S., & Anderson, N. B. (1997). Racial
differences in physical and mental health: Socioeconomic status, stress
and discrimination. Journal of Health Psychology, 2, 335–351.
Yang, Y., & George, L. K. (2005). Functional disability, disability
transitions, and depressive symptoms in late life. Journal of Aging and
Health, 17, 263–292.
Zeiss, A. M., Lewinsohn, P. M., Rohde, P., & Seeley, J. R. (1996).
Relationship of physical disease and functional impairment to depression in older people. Psychology and Aging, 11, 572–581.
Received June 19, 2007
Accepted April 16, 2008
Decision Editor: Kenneth F. Ferraro, PhD

GAYMAN ET AL.

S228

Appendix
Items for Physical Limitation Scale
Items asked of all respondents
1. Reach up and get a 5-pound object (such as a bag of sugar) from just above
your head
2. Bend down to pick up an object (like a piece of clothing) from the ﬂoor
3. Turn faucets on/off
4. Walk ¼ mile
5. Stoop or crouch down
6. Lift 10 pounds
7. Sit for more than 2 hr
8. Stand for long periods, such as 30 min
9. Stand up from sitting
10. Walk more than a mile
11. Moderate activities such as moving a table, pushing a vacuum cleaner,
bowling, or playing golf
12. Vigorous activities such as running, lifting heavy objects, or participating
in strenuous sports
Items asked only of those screened as disabled
13.
14.
15.
16.
17.
18.
19.

Can
Can
Can
Can
Can
Can
Can

you
you
you
you
you
you
you

prepare your own meals?
do your housework?
dress/undress self?
get in/out of bed?
take bath/shower?
get to the bathroom on time?
climb up stairs?

Errata
In the article entitled, ‘‘Divergent Pathways? Racial/Ethnic Differences in Older Women’s Labor
Force Withdrawal,’’ published in the Journal of Gerontology: Social Sciences, 63B (3), S122–
S134, by authors Tyson H. Brown and David F. Warner, the following changes should be made:
Text:
 On page S122, the ﬁrst true paragraph in the second column should read: ‘‘A major theme of life
course research is that transitions, such as retirement, are shaped by the intersection of biography
and history.’’
 On page S132, second column, the ﬁrst sentence in the penultimate paragraph should read:
‘‘Finally, although the HRS made efforts to oversample racial/ethnic minorities, the data contain
a relatively small number of Hispanic respondents, which limited our ability to detect signiﬁcant
differences between groups.’’
Tables:
 The title of Table 1 (p. S126) should read: ‘‘Model Variables, Coding, and Characteristics at First
Interview (Means and t Tests), by Race/Ethnicity (Weighted Estimates). ’’
 The second sentence in the Table 1 notes (p. S127) should read: ‘‘Welch-Satterthwaite t-tests
computed for difference in means with unequal variances.’’