ageingageingAge and AgeingAge Ageing0002-07291468-2834Oxford University Press10.1093/ageing/afq022afq022Research PaperHandgrip strength as a predictor of functional, psychological and social health. A prospective population-based study among the oldest oldD. G. Taekema et al.Handgrip strength as a predictor of health declineTaekemaDiana G.12GusseklooJacobijn3MaierAndrea B.14WestendorpRudi G. J.14de CraenAnton J. M.141Department of Gerontology and Geriatrics, Leiden University Medical Center, Leiden, The Netherlands2Department of Geriatrics, Rijnstate Hospital, Arnhem, The Netherlands3Department of Public Health and Primary Care, Leiden University Medical Center, Leiden, The Netherlands4Netherlands Consortium for Healthy Aging, Leiden University Medical Center, Leiden, The NetherlandsAddress correspondence to: D. G. Taekema. Department of Gerontology and Geriatrics, Leiden University Medical Center, C2-R-133, PO Box 9600, 2300 RC Leiden, The Netherlands. Tel: +31 71 526 6640; Fax: +31 71 524 8159. Email: d.g.taekema@lumc.nl52010103201039333133715920092012200929122009© The Author 2010. Published by Oxford University Press on behalf of the British Geriatrics Society. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org2010Oxford University PressBackground: muscle wasting is associated with a detrimental outcome in older people. Muscle strength measurements could be useful as part of a clinical evaluation of oldest old patients to determine who are most at risk of accelerated decline in the near future.Objective: this study aimed to assess if handgrip strength predicts changes in functional, psychological and social health among oldest old.Design: the Leiden 85-plus Study is a prospective population-based follow-up study.Subjects: five-hundred fifty-five, all aged 85 years at baseline, participated in the study.Methods: handgrip strength was measured with a handgrip strength dynamometer. Functional, psychological and social health were assessed annually. Baseline data on chronic diseases were obtained from the treating physician, pharmacist, electrocardiogram and blood sample analysis.Results: at age 85, lower handgrip strength was correlated with poorer scores in functional, psychological and social health domains (all, P < 0.001). Lower baseline handgrip strength predicted an accelerated decline in activities of daily living (ADL) and cognition (both, P ≤ 0.001), but not in social health (P > 0.30).Conclusion: poor handgrip strength predicts accelerated dependency in ADL and cognitive decline in oldest old. Measuring handgrip strength could be a useful instrument in geriatric practice to identify those oldest old patients at risk for this accelerated decline.Keywordsdisabilityelderlyhandgrip strengthhealthsarcopeniaIntroductionMuscle wasting is a dominant feature of old age and is commonly referred to as sarcopenia. Estimates of the prevalence of sarcopenia range depending on the definition from 18% to over 60% in the general population of the oldest old [1]. Due to a rapid growth of the number of oldest old in our societies, sarcopenia will become a common health problem [1].Muscle wasting is associated with detrimental outcome in the elderly, such as disability and mortality [2]. Several recent cross-sectional studies have shown associations between muscle strength and physical fitness, disability or cognition [3–6]. A number of prospective studies have described the association of handgrip strength and health decline in the elderly, predominantly describing its association with functional disability [7–12] and mortality [13, 14]. A limited number of studies report on the associations between muscle strength and cognition [15–17].All these associations raise the question about the value of muscle strength as a potential predictor of declining health in the oldest old. Muscle strength measurements could be useful as part of a clinical evaluation of the oldest old patients in determining which patients are most at risk of accelerated decline in the near future. Therefore we have studied the impact of muscle weakness on three health domains, functional, psychological and social health. Handgrip strength was used as a proxy for muscle strength in this study [3].MethodsParticipants and proceduresThe Leiden 85-plus Study is a community-based prospective follow-up study of inhabitants of the city of Leiden, The Netherlands. Enrolment took place between 1997 and 1999. All inhabitants, including nursing home residents, who reached the age of 85 were eligible to participate. There were no selection criteria on health or demographic characteristics [18]. In total 599 persons participated, 87% of all eligible inhabitants. At baseline, a research nurse visited participants at their place of residence. During these visits, socio-demographic characteristics including gender, marital status and living situation were recorded, performance tests were conducted, blood samples collected and an electrocardiogram (ECG) was recorded. The medical history was obtained from the general practitioner or nursing home physician. Follow-up visits were performed annually. The Medical Ethical Committee of the Leiden University Medical Center approved the study. All participants gave informed consent. In case of severe cognitive impairment, a guardian gave informed consent.Handgrip strengthAt ages 85 and 89, handgrip strength was measured with a Jamar hand dynamometer (Sammons Preston Inc., Bolingbrook, IL). The participant was asked to stand up and hold the dynamometer in the dominant hand with the arm parallel to the body without squeezing the arm against the body. The width of the handle was adjusted to the size of the hand to make sure that the middle phalanx rested on the inner handle. The participant was allowed to perform one test trial. After this, three trials followed and the best score was used for analysis. Handgrip strength was expressed in kilogrammes (Kg). Only handgrip strength measurements that were assessed as reliable by the research nurse were included in the analysis. The research nurse judged the effort according to the following criteria: refusal of participation, physical impairment, cognitive impairment, inability to follow the instructions and technical difficulties.Items of functional healthCompetence in daily functioning was measured with the Groningen Activity Restriction Scale (GARS) [19]. The GARS is a questionnaire that assesses disabilities in competence in the area of nine basic activities of daily living (ADL) and nine instrumental activities of daily living (IADL). A sum score was calculated separately for ADL and IADL. Hence, each sum score ranged from nine (competent in all activities) to 36 (unable to perform any activity without help). Walking speed was assessed with a standardised 6-m walking test as used in other longitudinal ageing studies [20]. Participants were requested to walk 3 m back and forth as quickly as possible. The use of a walking aid was allowed during this test. The time for the walking test was measured in seconds.Items of psychological healthCognitive functioning was measured with the Mini-Mental State Examination (MMSE) [21]. The 15-item Geriatric Depression Scale (GDS-15) was used as a screening instrument for depression [22]. As the validity and reliability of the GDS-15 may be reduced in subjects with impaired cognitive function, this questionnaire was restricted to those with MMSE scores above 18 points (n = 482) [23, 24].Items of social healthSocial functioning was measured with the Time Spending Pattern questionnaire (TSP). The TSP lists involvement in social and leisure activities leading to a sum score for these activities [25]. The questionnaire consists of 23 items (e.g. bathing a spouse, cycling, gardening, reading a book or watching television) scored from 0 (no activities) to 4 (participating in activity on a daily basis).Feelings of loneliness were annually assessed by the Loneliness Scale [26], an 11-item questionnaire especially developed for use in elderly populations. Scale scores range from 0 (absence of loneliness) to 11 (severe loneliness). The Loneliness Scale was also restricted to those with MMSE scores above 18 points.Potential confoundersData on common chronic diseases were obtained from the general practitioner, pharmacist's records, ECG and blood sample analysis [27].Chronic diseases included stroke, angina pectoris, myocardial infarction, intermittent claudication, peripheral arterial surgery, diabetes mellitus, obstructive pulmonary disease, malignancy and arthritis. Multi-morbidity was defined as the sum score of these somatic diseases.Statistical analysisBaseline cross-sectional associations at age 85 were assessed between tertiles of handgrip strength and items of health, using ANOVA (Analyses of variance). Handgrip strength was ranked and divided into tertiles for men and women separately.The prospective association between handgrip strength and the items of health was analysed with linear mixed models. The flexibility of mixed models makes them the preferred choice for the analysis of repeated-measures data [28]. The used models included estimates for ‘handgrip strength’, ‘time’ and ‘handgrip strength * time’. The estimate for ‘handgrip strength’ indicates the baseline association between handgrip strength and health item scores (presented in Table 3 as ‘baseline difference’). This estimate indicates the change in health items per kilogramme increase in handgrip strength. The estimate for ‘time’ indicates the annual change in performance for those participants with mean handgrip strength levels (presented in Table 3 as ‘annual change’). The estimate for ‘handgrip strength * time’ indicates the accelerated annual decline in health items per kilogramme decrease of handgrip strength at baseline (presented in Table 3 as ‘accelerated decline’). All estimates were adjusted for gender, height, weight and income. Estimates were standardised per kilogramme change of grip strength by using the formula: (individual handgrip strength − mean handgrip strength in study population).SPSS 16.0 for Windows was used for all analyses. P-values < 0.05 were considered statistically significant.ResultsParticipants characteristicsReliable scores for handgrip strength were available for 555 (92.6%) participants at age 85. At baseline, there were 44 non-completed handgrip strength measurements due to refusal to participate (n = 3), physical impairment (n = 17), cognitive impairment (n = 9), inability to follow instructions (n = 5) and other reasons (n = 10). There were 73 (12.9%) participants with an MMSE score ≤ 18 points, being indicative of cognitive impairment. Depressive symptoms (GDS score ≥ 4 points) were present in 114 (20.5%) of the participants. The other baseline characteristics of the study population are shown in Table 1.Table 1Characteristics of participants (n = 555) at baseline (85 years)Men (%)194 (35.0)Widowed (%)314 (56.6)Living arrangements  Independent (%)319 (57.5) Sheltered (%)156 (28.1) Institutionalised (%)80 (14.4)General health  ≥3 chronic diseasesa (%)135 (24.3) Body mass index <20 (%)24 (4.5)Functional health domain (median, ITR)b  ADL disability (points)c10 (9.0–11.0) IADL disability (points)c17 (13.8–22.0) Walking speed (seconds)d11.6 (9.4–14.2)Psychological health domain (median, ITR)  Cognition (MMSE, points)e26 (24–28) Depression (GDS, points)f2 (1–3)Social health domain (median, ITR)  Time spending pattern (points)g48 (43–51) Loneliness (points)h1 (0–2)aChronic diseases included stroke, angina pectoris, myocardial infarction, intermittent claudication, peripheral arterial surgery, diabetes mellitus, obstructive pulmonary disease, malignancy and arthritis.bITR, intertertile range.cGARS, possible scores range from 9 to 36 points (best to worst).d6-m walking test, scores ranged from 4.16 to 76.47 s (best to worst).eMMSE, possible scores range from 0 to 30 points (worst to best).fGDS-15, possible scores range from 0 to 15 points (best to worst).gTSP, possible scores range from 0 to 92 points (worst to best).hde Jong-Gierveld Loneliness Scale, possible scores range from 0 to 11 points (best to worst).Functional, psychological, and social health domainThe cross-sectional analyses at age 85 of functional, psychological and social items of health are shown in Table 2 for each tertile of handgrip strength. Lower handgrip strength was significantly correlated with poorer health item scores at baseline (Table 2, all P ≤ 0.03).Table 2Items of health according to handgrip strength tertiles at age 85DomainHandgrip strengtha Highest tertileMiddle tertileLowest tertileP for trendb34–54 kg men20–33 kg men10–27 kg men21–32 kg women17–20 kg women1–16 kg women n = 194n = 177n = 184Functional health     ADL-disability (points)c10.2 (0.2)11.1 (0.2)14.1 (0.5)<0.001 IADL-disability (points)c21.8 (0.8)18.3 (0.5)23.6 (0.7)<0.001 Walking speed (seconds)d10.9 (0.7)14.8 (0.7)18.8 (1.1)<0.001Psychological health     Cognition (points)e26.3 (0.3)25.4 (0.3)22.3 (0.5)<0.001 Depression (points)f2.4 (0.3)2.4 (0.2)3.0 (0.2)<0.001Social health     Time spending pattern (points)g50.3 (0.5)48.3 (0.5)44.3 (0.5)<0.001 Loneliness (points)h1.6 (0.2)1.5 (0.2)2.1 (0.2)0.03aData presented as mean (SE, standard error). Handgrip strength was ranked and divided into tertiles for men and women separately.bANOVA.cGARS, possible scores range from 9 to 36 points (best to worst).d6-m walking test, scores ranged from 4.16 to 76.47 s (best to worst).eMMSE, possible scores range from 0 to 30 points (worst to best).fGDS-15, possible scores range from 0 to 15 points (best to worst).gTSP, possible scores range from 0 to 92 points (worst to best).hde Jong-Gierveld Loneliness Scale, possible scores range from 0 to 11 points (best to worst).To analyse the prospective association between baseline handgrip strength and changes in the various health domains, we used linear mixed models (Table 3). In line with the cross-sectional results, we confirmed the association between handgrip strength and health item scores at baseline as indicated by ‘baseline difference’. Over time all health items, except loneliness, declined as indicated by the estimate ‘annual change’. Finally, we assessed if lower handgrip strength at baseline predicted an accelerated decline in the health domains as assessed by the estimate ‘accelerated decline’. Where such an estimate is significant, this indicates a predictive relationship in the model. Lower handgrip strength predicted an accelerated decline in ADL-disability in the functional health domain (0.02 points increase in GARS score per kilogramme loss of handgrip strength, P ≤ 0.001) and cognition in the psychological health domain (0.01 points decline in MMSE score per kilogramme loss of handgrip strength, P = 0.001), but not in other items of health (all P > 0.30). Additional adjustments for baseline MMSE and GDS scores did not change the prospective results of the functional health items. The prospective results of the psychological health items did not change after adjustment for baseline scores of ADL disability, IADL disability and walking speed. The results of the social health items were not influenced by adjustment for baseline functional health items or baseline psychological items.Table 3Changes in items of health according to handgrip strength at 85 (per kg)aDomainBaseline differenceAnnual changeAccelerated decline Estimate (SE)bP-valueEstimate (SE)bP-valueEstimate (SE)bP-valueFunctional health       ADL-disability (points)c−0.27 (0.04)<0.0011.28 (0.05)<0.001−0.02 (0.01)<0.001 IADL-disability (points)c−0.46 (0.05)<0.0012.25 (0.06)<0.0010.01 (0.01)0.385 Walking speed (seconds)d−0.50 (0.08)<0.0010.35 (0.17)0.040.01 (0.02)0.471Psychological health       Cognition (points)e0.25 (0.04)<0.001−0.75 (0.04)<0.0010.01 (0.004)0.001 Depression (points)f−0.08 (0.02)<0.0010.29 (0.03)<0.0010.002 (0.003)0.626Social health       Time spending pattern (points)g0.40 (0.05)<0.001−1.38 (0.06)<0.001−0.004 (0.01)0.560 Loneliness (points)h−0.05 (0.02)<0.0010.02 (0.02)0.2520.001(0.002)0.968aLinear mixed model adjusted for gender, height, weight, income and multi-morbidity, n = 555. The estimate for ‘baseline difference’ indicates the baseline association between handgrip strength and health item scores. The estimate for ‘annual change’ indicates the annual change in performance for those participants with mean handgrip strength levels. The estimate for ‘accelerated decline’ indicates the accelerated annual change in health items per kilogramme change of handgrip strength at baseline.bSE, standard error.cGARS, possible scores range from 9 to 36 points (best to worst).d6-m walking test, scores ranged from 4.16 to 76.47 s (best to worst).eMMSE, possible scores range from 0 to 30 points (worst to best).fGDS-15, possible scores range from 9 to 36 points (best to worst).gTSP, possible scores range from 0 to 92 points (worst to best).hde Jong-Gierveld Loneliness Scale, possible scores range from 0 to 11 points (best to worst).DiscussionThe aim of the present study was to explore if handgrip strength predicts decline in functional, psychological and social health in the oldest old. Our findings show that lower handgrip strength predicted an accelerated decline in ADL-disability and cognition, and thus contributes to increasing dependency in old age.To our knowledge, our study is the first to report on the prospective associations between handgrip strength and three health domains in a cohort of oldest old participants. A number of other studies have reported on prospective associations between handgrip strength and functional ability, or cognition in the elderly, but the mean age of participants in these prospective studies was younger and none of these included all three health domains [7–12, 15–17]. Some of these studies were limited to men [7, 8] or women [12, 13].We confirmed the predictive value of handgrip strength in the functional health domain in oldest old participants. No predictive association was found between handgrip strength and IADL disability, which had been shown to be a predictor in Japanese community-dwelling elderly in participants aged 65 years and older [9]. For walking speed, we could not confirm a predictive value of handgrip strength, which were associated with each other in a comprehensive cross-sectional study [3].For the association between muscle strength and functional health, one would expect that interventions aimed at improving muscle strength are beneficial. A recent review [29] has assessed the effect of resistance training on physical functioning in subjects over 60 years old. High intensity strength training, three times a week, significantly improved muscle strength, and was associated with improvement in physical ability.Our finding that low handgrip strength predicts accelerated cognitive decline has been reported by others, but again in younger study participants [15–17]. Changes in handgrip strength did not predict changes in depression, possibly because of a process of psychological adaptation during ageing in elderly people [30].In the social health domain, no predictive association was found with the item loneliness. This might be explained by the fact that the need for care results in regular contact with caregivers, thereby stimulating psychological well being as suggested by a cross-sectional Scandinavian study among elderly nursing home residents [31].This study has several key strengths to studying consequences of sarcopenia in elderly people. The Leiden 85-plus Study is a longitudinal population-based cohort study with extensive measures for health and functioning. Therefore the results can be generalised to the western population of oldest old. Furthermore, the longitudinal design with repeated measurements of diverse items of health allowed us to demonstrate a temporal association.A possible weakness of our study could be that our participants appear to be relatively fit. For very frail elderly people, measuring handgrip strength might be difficult and the results could not be applicable to this group. But, only 44 (7.3%) measurements of handgrip strength were excluded from our study because these were deemed unreliable. Of which, 31 (5.2%) were the result of physical or cognitive impairment. We don't know if our participants are fitter compared to other oldest old. Comparison of participant characteristics of the Leiden 85-plus Study with other prospective studies on ageing is difficult because of age differences and different methodology. The Newcastle 85-plus Study started in 2006 [32] and is comparable in design to the Leiden 85-plus Study. The characteristics of the subjects from the Newcastle pilot study are similar to our study participants with regard to living arrangements, cognitive ability and depressive symptoms [33]. Another weakness of the study could be that the questionnaires on depression and loneliness were limited to those participants without cognitive decline which could have underestimated the associations between handgrip strength and the psychological indicators. However, only 73 (13%) of the participants were excluded due to an MMSE score of 18 points or lower. One could also argue that the chosen health domains are indirectly related to one another; however, further adjustment of the linear mixed model for this possible confounding did not change the results.Functional measurements, as walking or gait speed, chair stand test and balance, have also been shown to predict functional limitations of the lower body [12, 20, 34], and cognition [35] in older subjects. As yet it is unclear whether muscle strength or functional measurements are the stronger predictor, and which causal pathways are involved. An advantage of handgrip strength could be that it is easy to use in clinical practice.We conclude that poor handgrip strength is a predictor of accelerated dependency in ADL and cognitive decline in oldest old. Based on these findings, we conclude that measuring handgrip strength could be a useful instrument in geriatric clinical practice to identify those oldest old patients at risk for accelerated decline in ADL ability and cognition.Key pointsPoor handgrip strength predicts accelerated dependency in ADL and cognitive decline in oldest old.Measuring handgrip strength can be useful to identify those oldest old patients at risk for future decline.Handgrip strength measurement is an easy to use instrument in clinical geriatric practice.Conflicts of interestNone.This study was supported by unrestricted grants from the Netherlands Organisation of Scientific Research (ZonMw), the Ministry of Health, Welfare and Sports, the Netherlands Genomics Initiative/Netherlands Organization for scientific research and the Netherlands Consortium for Healthy Aging. (NGI/NWO; 05040202 and 050-060-810 NCHA).References1DohertyTJInvited review: aging and sarcopeniaJ Appl Physiol2003951717272FisherALOf worms and women: sarcopenia and its role in disability and mortalityJ Am Geriatr Soc2004521185903LauretaniFRussoCRBandinelliSAge-associated changes in skeletal muscles and their effect on mobility: an operational diagnosis of sarcopeniaJ Appl Physiol2003951851604JeuneBSkyttheACournilAHandgrip strength among nonagenarians and centenarians in three European regionsJ Gerontol A Biol Sci Med Sci200661707125HasegawaRIslamMMLeeSCThreshold of lower body muscular strength necessary to perform ADL independently in community-dwelling older adultsClin Rehabil200822902106TakataYAnsaiTSohIPhysical fitness and cognitive function in an 85-year-old community-dwelling populationGerontology200854354607GiampaoliSFerrucciLCecchiFHand-grip strength predicts incident disability in non-disabled older menAge Ageing19992828388RantanenTGuralnikJMFoleyDMidlife hand grip strength as a predictor of old age disabilityJAMA1999281558609IshizakiTWatanabeSSuzukiTPredictors for functional decline among nondisabled older Japanese living in a community during a 3-year follow-upJ Am Geriatr Soc20008424910RantanenTAvlundKSuominenHMuscle strength as a predictor of onset of ADL dependence in people aged 75 yearsAging Clin Exp Res20021410511Al SnihSMarkidesKSRayLHandgrip strength and mortality in older Mexican AmericansJ Am Geriatr Soc2002501250612OnderGPenninxBWFerrucciLMeasures of physical performance and risk for progressive and catastrophic disability: results from the Women's Health and Aging StudyJ Gerontol A Biol Sci Med Sci20056074913RantanenTVolpatoSFerrucciLHandgrip strength and cause-specific and total mortality in older disabled women: exploring the mechanismJ Am Geriatr Soc2003516364114Al SnihSMarkidesKSOttenbacherKJHand grip strength and incident ADL disability in elderly Mexican Americans over a seven-year periodAging Clin Exp Res200416481615ChristensenHKortenAEMackinnonAJAre changes in sensory disability, reaction time, and grip strength associated with changes in memory and crystallized Intelligence? A longitudinal analysis in an elderly community sampleGerontology2000462769216Alfaro-AchaAAl SnihSRajiMAHandgrip strength and cognitive decline in older Mexican AmericansJ Gerontol A Biol Sci Med Sci2006618596517BuchmanASWilsonRSBoylePAGrip strength and the risk of incident Alzheimer's diseaseNeuroepidimiology200729667318Bootsma-van der WielAvan ExelEde CraenAJMA high response is not essential to prevent selection bias: results from the Leiden 85-plus StudyJ Clin Epidemiol20025511192519KempenGIMiedemaIOrmelJThe assessment of disability with the Groningen Activity Restriction Scale. Conceptual framework and psychometric propertiesSoc Sci Med19964316011020GuralnikMSimonsickEMFerruciLA short physical performance battery assessing lower extremity function: association with self reported disability and prediction of mortality and nursing home admissionJ Gerontol199449M859421FolsteinMFFolsteinSEMcHughPR“Mini-Mental State”. A practical method for grading the cognitive state of patients for the clinicianJ Psychiatr Res1975121899822YesavageJABrinkTLRoseTLDevelopment and validation of a geriatric depression screening scale: a preliminary reportJ Psychiatr Res198217374923TombaughTNMcIntyreNJThe Mini-Mental State Examination. A comprehensive reviewJ Am Geriatr Soc1992409223524van ExelEde CraenAJRemarqueEJInteraction of atherosclerosis and inflammation in elderly subjects with poor cognitive functionNeurology200361169570125van EijkLActivity and Well-being in the Elderly1997The NetherlandsUniversity of GroningenThesis26de Jong-GierveldJKamphuisFThe development of a Rasch-type loneliness scaleAppl Psychol Measur198592899927Bootsma-van der WielAGusseklooJde CraenAJMCommon chronic diseases and general impairments as determinants of walking disability in the oldest-old populationJ Am Geriatr Soc20025014051028GueorguievaRKrystalJHMove over ANOVA: progress in analyzing repeated-measures data and its reflection in papers published in the Archives of General PsychiatryArch Gen Psychiatry200461310729Liu CJ, Latham NK. Progressive resistance strength training for improving physical function in older adults. Cochrane Database Syst Rev 2009, Issue 330von FaberMBootsma-van der WielAvan ExelESuccessful aging in the oldest old. Who can be characterized as successfully aged?Arch Intern Med2001161269470031DragesetJThe importance of activities of daily living and social contact for loneliness: a survey among residents in nursing homesScand J Caring Sci200418657132AdamsonAJCollertonJDaviesKNutrition in advanced age: dietary assessment in the Newcastle 85+ studyEur J Clin Nutr200963S61833CollertonJBarrassKBondJThe Newcastle 85+ study: biological, clinical and psychosocial factors associated with healthy ageing: study protocolBMC Geriatr200771434VasunilashornSCoppinAKPatelKVUse of the Short Physical Performance Battery Score to predict loss of ability to walk 400 meters: analysis from the in CHIANTI StudyJ Gerontol A Biol Sci Med Sci200964223935DeshpandeNMetterEJBandinelliSGait speed under varied challenges and cognitive decline in older persons: a prospective studyAge Ageing20093850914