Frailty and Disability in Older Adults with Intellectual Disabilities: Results from the Healthy Ageing and Intellectual Disability Study

  • Published on

  • View

  • Download


Frailty and Disability in Older Adults with IntellectualDisabilities: Results from the Healthy Ageing and IntellectualDisability StudyHeleen M. Evenhuis, PhD,* Heidi Hermans, MSc,* Thessa I. M. Hilgenkamp, MSc,*Luc P. Bastiaanse, MSc,* and Michael A. Echteld, PhD*OBJECTIVES: To obtain first insight into prevalence andcorrelates of frailty in older people with intellectual dis-ability (ID).DESIGN: Population-based cross-sectional study in per-sons using formal ID services.SETTING: Three Dutch care provider services.PARTICIPANTS: Eight hundred forty-eight individualswith borderline to profound ID aged 50 and older partici-pating in the Healthy Ageing and Intellectual Disability(HA-ID) Study.MEASUREMENTS: All participants underwent an exten-sive health examination. Frailty was diagnosed accordingto Cardiovascular Health Study criteria. Associationsbetween frailty and participant characteristics were investi-gated using multivariate logistic regression analysis.RESULTS: Prevalence of frailty was 11% at age 50 to 64and 18% at age 65 and older. Age, Down syndrome,dementia, motor disability, and severe ID were signifi-cantly associated with frailty, but only motor disabilityhad a unique association with frailty. In a regressionmodel with these variables, 25% of the variance of frailtywas explained.CONCLUSION: At age 50 to 64, prevalence of frailty isas high as in the general population aged 65 and older (79%), with a further increase after the age of 65. Motordisability only partially explains frailty. Future studiesshould address health outcomes, causes, and preventionof frailty in this population. J Am Geriatr Soc 60:934938, 2012.Key words: intellectual disability; frailty; prevalenceAs a result of normalizing life expectancy in the popula-tion with intellectual disabilities (IDs),1 the numberof older persons with IDs is rapidly increasing. A majorityof these persons now live in the community and make useof regular health care. Apart from people with Downsyndrome, with early sensory losses and dementia,2,3 thereare no signs that primary aging is premature in this group.Nevertheless, professionals working for this population,have generally accepted the idea of early aging for severaldecades, primarily based on limited life expectancy andearly deterioration of activities of daily living.4,5 Evidence-based insight into conditions underlying this supposedearly aging may contribute to health policies aimed atprevention, timely detection, and intervention in thispopulation or specific subgroups. Tentative prevalencedata regarding frailty in this population were published in2010, based on a questionnaire assessment based on twodistinct approaches of frailty, which hampers comparisonwith studies in the general population.6Recognizing that validity of internationally establishedfrailty criteria might be different in a population withlifelong disability, as a first step, a cross-sectionalinventory of frailty, diagnosed according to CardiovascularHealth Study (CHS) criteria,7 was performed in a popula-tion that was nearly representative of older clients ofDutch ID care provider services. It was hypothesized that,in this group, frailty would be present at a younger agethan in the general older population and that frailty wouldoverlap with but not be identical to disability. Therefore,frailty prevalence and its associations with age, sex, Downsyndrome, dementia, care setting, motor disability, andsevere ID were studied.From the *Department of General Practice, Intellectual DisabilityMedicine, Erasmus University Medical Centre, Rotterdam, theNetherlands; Amarant Center for Intellectual Disability, Tilburg, theNetherlands; Abrona Center for Intellectual Disability, Huis ter Heide,the Netherlands; and Ipse de Bruggen Center for Intellectual Disability,Zwammerdam, the Netherlands.Address correspondence to Prof. Heleen M. Evenhuis, Department ofGeneral Practice, Erasmus University Medical Center, IntellectualDisability Medicine, PO Box 2040, 3000 CA Rotterdam, the Netherlands.E-mail: h.evenhuis@erasmusmc.nlDOI: 10.1111/j.1532-5415.2012.03925.xJAGS 60:934938, 2012 2012, Copyright the AuthorsJournal compilation 2012, The American Geriatrics Society 0002-8614/12/$15.00METHODSStudy Design and ParticipantsThis study was part of a large cross-sectional study:Healthy Ageing and Intellectual Disability (HA-ID). Thestudy population consists of clients aged 50 and older in aconsortium of three care provider services in the Nether-lands offering specialized support varying from ambulatorysupport or day care for clients living independently or withfamily to residential settings providing support of indepen-dency or more or less intensive care. In most high-incomecountries, regional formal ID services serve comparablegroups as in the Netherlands, meaning that the largergroup of people with borderline or mild IDs not using anyspecialized services or living in general geriatric settingshas not been included.Informed consent was obtained from 1,069 clientsaged 50 and older or their legal representatives; 1,050 par-ticipated in the assessments. The study population is nearlyrepresentative of the consorts total client population aged50 and older, with a slight underrepresentation of men,individuals living independently, and individuals aged 80and older. Details regarding design, recruitment, and rep-resentativeness of the sample and diagnostic methods havebeen published previously.8 The Medical Ethics Committeeof the Erasmus Medical Center Rotterdam (MEC 2008234) and the ethics committees of the participating careprovider services provided ethics approval. The studyadheres to the Declaration of Helsinki for research involv-ing human subjects.Diagnostic MeasurementsAll participants underwent an extensive diagnostic assess-ment, consisting of a physical assessment, a fitness test bat-tery, measurement of physical activity using pedometersfor 2 weeks, standardized mealtime observations of swal-lowing, and standardized psychiatric assessment of depres-sion and anxiety.8Diagnoses of Down syndrome, confirmed by karyotyp-ing, were retrieved from participant files. Dementia wasdefined as possible or probable dementia. Diagnoses wereobtained from participants physicians and behavioral spe-cialists and were included in the analysis only in the caseof consensus between these professionals. Care setting wasclassified as centralized setting (primarily care), commu-nity-based group home (primarily support), or ambulatorysupport (living independently or with family, participatingin daycare or getting specific support).Mobility was judged before the fitness assessment andclassified as independent mobility, mobility with aids, andmobility in wheelchair. Motor disability was defined asmobility with aids or wheelchair. Information was notavailable on age of onset of motor disability.Psychologists or behavioral scientists had performedintelligence quotient (IQ) testing after admission to thecare organizations; severity of ID had been classified asborderline (IQ 7080), mild (IQ 5569), moderate (IQ 3554), severe (IQ 2534), or profound (IQ < 25). This infor-mation was retrieved from the files. For the analysis,severe and profound ID were combined as severe ID.Frailty and DisabilityAccording to the CHS criteria, participants with at leastthree of the following five criteria were defined as frail:weight loss, poor grip strength, slow walking speed, lowphysical activity, and poor endurance or exhaustion. Par-ticipants with one or two criteria were considered prefrailand participants with no criteria as robust.7 The criteriawere operationalized as follows. Weight loss: an item of the Mini Nutritional Assessment,9weight loss during the past 3 months was assessed on a4-point rating scale. Losses >3 kg were scored. Grip strength was measured using a Jamar Hand Dyna-mometer (#5030J1, Sammons Preston Rolyan, Dolge-ville, NY) as part of the fitness assessment. Poor gripstrength was defined according to CHS criteria, withstratification according to sex and body mass index.7 Comfortable walking speed was tested as part of the fit-ness assessment and was measured as the average ofthree recordings of the time to complete a distance of5 m. Slow walking speed was defined according to CHScriteria, with stratification according to sex and height.7All participants in a wheelchair and all participants whocould not participate in the walking speed assessmentbecause of physical limitations were scored as having aslow walking speed. Physical activity was tested using pedometers (NL-1000;New Lifestyles, Lees Summit, MO) for 14 days. Theminimal comfortable walking speed required for a reli-able measurement was 3.2 km/hour. All participantswalking fewer than 5,000 steps/day (sedentary life-style)10 were scored positive on low physical activity, aswere all participants in a wheelchair and all participantswho could not engage in the walking speed assessmentbecause of physical limitations. Poor endurance or exhaustion was assessed using theitem Lacks energy of the Anxiety, Depression andMood Scale.11 This item has a 4-point rating scale. Theno problem and mild problem answers were recordedinto no and the moderate problem and severe problemanswers into yes.AnalysisBecause not all participants were able to participate in allassessments,8 frailty criteria could be missing. All partici-pants with at least three criteria measured were includedin the analyses. Characteristics of included and excludedparticipants were compared using chi-square (v2) statistics.Prevalences of frailty and prefrailty and 95% confi-dence intervals were calculated for the total includedgroup and for subgroups according to age, sex, care set-ting, Down syndrome, dementia, severity of ID, and motordisability. Prevalences of frailty and prefrailty in subgroupswere compared using chi-square statistics.A multivariate logistic regression analysis was per-formed to determine whether Down syndrome, dementia,motor disability, and severe ID are independently associ-ated with frailty, controlling for the effect of age (5064vs 65) and gender. Age and sex were entered into theequation first, after which Down syndrome, dementia,JAGS MAY 2012VOL. 60, NO. 5 INTELLECTUAL DISABILITY AND FRAILTY 935motor disability, and severe ID were entered simulta-neously. Correlations of frailty and motor disability withage were tested using Spearman correlation coefficients.RESULTSEight hundred forty-eight of the 1,050 participants hadthree or more criteria assessed and were included in theanalyses. The excluded subgroup (n = 202) had a sex andage distribution similar to that of this group (data notshown) but included significantly more persons with severeID (v2 = 51.6, P = .01) and motor disability (v2 = 22.5,P = .01). In the included group, cause of ID was unknownin 673, Down syndrome in 120, other genetic syndromesin 13, and varying other diagnoses in 42 participants.Cerebral palsy was not scored as a causal diagnosis of ID.The distribution of age, sex, care setting, Down syn-drome, dementia, severity of ID, and motor disabilityis presented in the second column of Table 1. Of 120participants with Down syndrome, 47 had a diagnosis ofdementia.Of the total included population, 27% was robust(95% confidence interval (CI) = 2430), 60% prefrail(95% CI = 5763), and 13% frail (95% CI = 1115).Prevalence of frailty was 11% (95% CI = 814%) in thegroup aged 50 to 64, 18% (95% CI = 1323) in the groupaged 65 and older, and 21% (95% CI = 1232) in thegroup aged 70 and older.Thirty-seven of 833 participants (4%) lost weight, and373 of 701 (53%) had poor grip strength, 138 of 827(17%) exhaustion, 263 of 795 (33%) slow walking speed,and 254 of 421 (60%) low physical activity. Because ofmissing data, total numbers were never 848. Physicalactivity could not be established in a majority of partici-pants, largely because of unreliable pedometer results dueto a walking speed slower than 3.2 km/hour (n = 256 inthe total HA-ID study population) and limited understand-ing or noncooperation (n = 204).12Prevalences of frailty and prefrailty in subgroups andrelationships with other participant characteristics are pre-sented in Table 1, showing significant associations with allcharacteristics except sex. Down syndrome was associatedonly with prefrailty. Multivariate logistic regression analy-sis confirmed that, if age and sex were entered into theequation, age was positively associated with frailty, withparticipants aged 65 and older being 1.7 times as likely tobe frail as the younger group. After entering Down syn-drome, dementia, motor disability, and severity of ID intothe equation, motor disability was independently and verystrongly associated with frailty; people using walking aidsTable 1. Participant Characteristics (N = 848) and Distribution of FrailtyCharacteristicTotal Robust Prefrail FrailDifference (Pre)Frailty, Chi-Squaren (%)Total 848 230 (27) 508 (60) 110 (13)Age5064 582 (69) 178 (31) 341 (59) 63 (11) 15.33c 65 266 (31) 52 (20) 167 (63)d 47 (18)d5069 777 (82) 221 (28) 461 (59) 95 (12) 10.45b 70 71 (18) 9 (13) 47 (66)d 15 (21)dSexFemale 412 (49) 108 (26) 251 (61) 53 (13) 0.39Male 436 (51) 122 (28) 257 (59) 57 (13)Care settingCentralized 436 (51) 86 (20) 270 (62)d 80 (18)d 48.56cCommunity 363 (43) 118 (33) 215 (59) 30 (8)Ambulatory support 41 (5) 23 (56) 18 (44) 0Causes of IDDown syndrome 120 (14) 19 (16) 84 (70)d 17 (14) 7.25aOther 621 170 (27) 367 (59) 84 (14)DementiaYes 74 9 (12) 49 (66)d 16 (22)d 10.09bNo 659 179 (27) 396 (60) 84 (13)Severity of IDBorderline to mild 218 (26) 91 (42) 113 (52) 14 (6) 37.95cModerate 425 (50) 101 (24) 263 (62)d 61 (14)dSevere to profound 185 (22) 35 (19) 116 (63)d 34 (18)dMotor disabilityNo 613 (72) 216 (35) 365 (59.5) 32 (5) 207.39cWalking aid 121 (14) 10 (8) 86 (71)d 25 (21)dWheelchair 107 (13) 0 54 (50.5)d 53 (49.5)da P < .05.b P < .01.c P < .001.d Subgroup with significantly more (pre)frailty.Frail, 3/5 Cardiovascular Health Study criteria; prefrail, 12/5 criteria; robust, 0/5 criteria.ID = intellectual disability.936 EVENHUIS ET AL. MAY 2012VOL. 60, NO. 5 JAGSor a wheelchair were 9.8 times as likely to be frail as thosewho walked independently.The proportion of explained variance was 0.25 (Nagel-kerke R2), and the overall model fit the data adequately,as indicated by the Hosmer and Lemeshow test (v2 = 8.98;P = .34). The correlation of frailty and motor disabilitywith age is shown in Figure 1.DISCUSSIONThis is the first large-scale study of frailty in a populationthat was nearly representative of Dutch people with IDreceiving formal care aged 50 and older applying CHS cri-teria for frailty. Persons with all levels of ID and supportneeds were included and comprehensively tested usinginternationally accepted diagnostic methods. Thirteen per-cent of the ID population is frail and 60% prefrail,whereas 28% to 31% of those aged 50 to 69 were labelledrobust, declining to 13% after age 70.Whereas frailty in the general older populationbecomes apparent mainly after age 75, this appears to becompletely different in the population with ID; the preva-lence of 11%, found in the subgroup aged 50 to 64, issimilar to published prevalences in large-scale generalpopulations aged 65 and older measured using the same cri-teria as in the present study and ranging approximately 7%to 9%,7,1316 confirming the first hypothesis of early frailty.Frailty prevalence becomes larger with older age, with indi-viduals aged 65 and older being 1.6 times as likely to befrail as the younger group. With 18%, the prevalence inthat age group is two times as high as in the general popula-tion, whereas for aged 70 and older, 21% is frail, comparedwith 15% to 17% of the general population.1720Apart from age, univariate analyses show significantassociations between (pre)frailty and Down syndrome,dementia, care setting, motor disability, and severe ID.The finding that motor disability is uniquely associatedwith frailty, with people using walking aids and wheel-chairs being 10 times as likely to be frail as people whowalk independently, together with the low explainedvariance (R2 = 0.25) of the regression model, confirms thesecond hypothesis that frailty and motor disability arestrongly overlapping but that motor disability and severeID only partially explain frailty.In geriatric practice and research, disability is primar-ily considered a late consequence of frailty. A limitation ofthe present study was that information on the age of onsetof motor disabilities was lacking. Figure 1 shows a highstarting prevalence of more than 15% motor disabilityat age 50 to 54, suggesting early impairments but alsolarger prevalences with older age. Although the cross-sectional design precludes conclusions about the directionof causality, it may be hypothesized that, in the populationwith ID, the relationship between frailty and disabilitymay partially be the other way around; childhood motorimpairment, complicated by early arthrosis and other age-related motor pathology, may lead to early muscle wasteand loss of muscle strength and function (sarcopenia),2124which is a strong risk factor for frailty. Moreover, in arecent analysis of physical activity in the total HA-ID pop-ulation, high prevalences of low activity levels, as mea-sured with pedometers, were found in the ambulantsubgroup; 64% walked fewer than 7,500 steps/day and39% were sedentary.12 This finding leads to the specula-tion that, next to long-standing motor disability, the lowlevel of physical activity in persons with independentmobility may make this population prone to early sarcope-nia and subsequent frailty. These hypotheses need to besupported by empirical prospective studies.Another limitation of this study is the presence ofselective missing observations. Participants who had to beexcluded from the analyses because of missing data weremore disabled, and not all variables could be assessed inmany included participants (specifically physical activitywith pedometers). Therefore, the frailty prevalence foundis likely to be an underestimation rather than an overesti-mation of the true frailty prevalence.The high prevalence of frailty and motor disability(Figure 1) in the group aged 50 to 64 suggests frailtybefore age 50. Whether the individual frailty items weremore common with advancing age was not analyzed.Analyses in the HA-ID cohort of physical activity and fit-ness components and their associations are in progress andwill be published as part of reports on fitness and itshealth outcomes in the cohort.The validity of frailty criteria to predict subsequentnegative health outcomes and increasing dependency isbeing evaluated for this population in a prospective fol-low-up. Long-standing disability might modify effects ofCHS-defined frailty on health, dependency, and mortalityin a negative way (e.g., greater comorbidity) or in a posi-tive way (e.g., because the availability of high-level special-ized ID care from a young age advances earlyrehabilitation and habituation).An outline of clinical implications of this first explor-atory study can not yet be made. Apart from health out-comes (validity), causes and prevention of frailty shouldfirst be examined in future research.ACKNOWLEDGMENTSWe thank Marieke van Schijndel-Speet, Sandra Mergler,Ellen van Dijk, and Channa de Winter, all PhD students inthe HA-ID study, for their active participation in thediscussions about findings and implications of this study;all have provided written consent for publication. We alsoFigure 1. Relationships between motor disability and frailtyand age in participants with intellectual disability. Motor dis-ability Spearman correlation coefficient (r) = 0.17, P < .001;frailty r = 0.13 P < .001.JAGS MAY 2012VOL. 60, NO. 5 INTELLECTUAL DISABILITY AND FRAILTY 937thank the care provider services involved in the HA-IDconsortium for their active support and collaboration, allparticipants and their legal representatives for theirpatience and enthusiasm, and all involved professionalstaff for their motivation and help with the diagnosticmeasurements.Conflict of Interest: None of the authors has anyfinancial, personal, or potential conflict of interest.Author Contributions: Evenhuis HM: Initiated theHA-ID study, obtained financial support, supervised allstages of the research project, and wrote this paper. Dr.Echteld MA: Supervises some of the PhD students, hasbeen specifically involved in methodological issues andanalysis, and commented on drafts of the paper. HermansH, Hilgenkamp T, and Bastiaanse L as PhD students eachdeveloped one subtheme of the HA-ID study and wereactively involved in study concept and design, recruitmentof participants and diagnostic measurements, creation ofthe data file, and performance of all analyses and com-mented on drafts of the paper.Sponsors Role: The involved care provider serviceshave actively contributed in the research infrastructure,supported recruitment of clients for the study, and contin-uously created all necessary conditions for data collection.They were not involved in study concept and design,methods, analysis, or publication.The Netherlands Organisation for Health Researchand Development provided a government grant for thestudy (ZonMw 57000003) and supervises its quality andprogress on a basis of annual reports. It was not involvedin study concept and design, methods, analysis, orpublication.REFERENCES1. Patja K, Iivanainen M, Vesala H et al. Life expectancy of people with intel-lectual disability: A 35-years follow-up study. J Intellect Disabil Res2000;44:591599.2. Garcez ME, Peres W, Salvador M. Oxidative stress and hematologic andbiochemical parameters in individuals with Down syndrome. Mayo ClinProc 2005;80:16071611.3. Thomas P, Harvey S, Gruner T et al. The buccal cytome and micronucleusfrequency is substantially altered in Downs syndrome and normal ageingcompared to young healthy controls. Mutat Res 2008;638:3747.4. Maaskant MA, van den Akker M, Kessels AG et al. Care dependence andactivities of daily living in relation to ageing: Results of a longitudinalstudy. J Intellect Disabil Res 1996;40:535543.5. Salvatori P, Tremblay M, Sandys J et al. Aging with an intellectual disabil-ity: A review of Canadian literature. Can J Aging 1998;17:249271.6. Brehmer B, Weber G. Frailty vs disability distinctions in people with intel-lectual disabilities. J Policy Pract Intellect Disabil 2010;7:4958.7. Fried LP, Tangen CM, Walston J et al. Cardiovascular Health Study Col-laborative Research Group. Frailty in older adults: Evidence for a pheno-type. J Gerontol A Biol Sci Med Sci 2001;56A:M146M156.8. Hilgenkamp TIM, Bastiaanse LP, Hermans H et al. Study healthy ageingand intellectual disabilities: Recruitment and design. Res Dev Disabil2011;32:10971106.9. Guigoz Y, Vellas B, Garry PJ. Mini Nutritional Assessment: A practicalassessment tool for grading the nutritional state of elderly patients. FactsRes Gerontol 1994; (Suppl 2):1560.10. Tudor-Locke C, Hatano Y, Pangrazi RP et al. Revisiting how many stepsare enough? Med Sci Sports Exerc 2008;40:S537S543.11. Esbensen AJ, Rojahn J, Aman MG et al. Reliability and validity of anassessment instrument for anxiety, depression, and mood among individu-als with mental retardation. J Autism Dev Disord 2003;33:617629.12. Hilgenkamp TIM, Reis D, van Wijck R, Evenhuis HM. Low physicalactivity levels in older adults with intellectual disabilities. Res Intell DevelDisabil 2011;33:477483.13. Hirsch C, Anderson ML, Newman A et al. The association of racewith frailty: The Cardiovascular Health Study. Ann Epidemiol 2006;16:545553.14. Kulminski AM, Ukraintseva SV, Kulminskaya IV et al. Cumulative deficitsbetter characterize susceptibility to death in elderly people than phenotypicfrailty: Lessons from the Cardiovascular Health Study. J Am Geriatr Soc2008;56:898903.15. Syddall H, Roberts HC, Evandrou M et al. Prevalence and correlates offrailty among community-dwelling older men and women: Findings fromthe Hertfordshire Cohort Study. Age Ageing 2010;39:197203.16. Hubbard RE, Lang IA, Llewellyn DJ et al. Frailty, body mass index, andabdominal obesity in older people. J Gerontol A Biol Sci Med Sci2010;65A:377381.17. Rockwood K, Andrew M, Mitnitski A. A comparison of two approachesto measuring frailty in elderly people. J Gerontol A Biol Sci Med Sci2007;62A:738743.18. Ensrud KE, Ewing SK, Taylor BC et al. Comparison of 2 frailty indexesfor prediction of falls, disability, fractures, and death in older women. ArchIntern Med 2008;168:382389.19. Ensrud KE, Ewing SK, Cawthon PM et al. A comparison of frailty indexesfor the prediction of falls, disability, fractures, and morbidity in older men.J Am Geriatr Soc 2009;57:492498.20. Rochat S, Cumming RG, Blyth F et al. Frailty and use of health and com-munity services by community-dwelling older men: The Concord Healthand Ageing in Men Project. Age Ageing 2010;39:228233.21. Bauer JM, Sieber CC. Sarcopenia and frailty: A clinicians controversialpoint of view. Exper Gerontol 2008;43:674678.22. Topinkova E. Aging, disability and frailty. Ann Nutr Metabol 2008;52(Suppl 1):611.23. Cruz-Jentoft AJ, Baeyens JP, Bauer JM et al. Sarcopenia: European consen-sus on definition and diagnosis. Report of the European Working Groupon Sarcopenia in Older People. Age Ageing 2010;39:412423.24. Morley JE, Abbatecola AM, Argiles JM et al. Sarcopenia with limited mobil-ity: An international consensus. J Am Med Dir Assoc 2011;12:403409.938 EVENHUIS ET AL. MAY 2012VOL. 60, NO. 5 JAGS