O

1 CLINICAL INVESTIGATIONS. cognitive , Extrapyramidal, and Magnetic Resonance Imaging Predictors of Functional Impairment in Nondemented Older Community Dwellers: The Sydney Older Person Study Hayley P. Bennett, PhD, w Olivier Piguet, PhD, w David A. Grayson, PhD,wz Helen Creasey, MBBS, FRACP,w Louise M. Waite, PhD, MBBS, FRACP,w Tanya Lye, PhD,w Alastair J. Corbett, MD, ChB, FRACP, Michael Hayes, MBBS, FRACP, . G. Anthony Broe, MBBS, FRACP, and Glenda M. Halliday, PhD . (See editorial comments by Dr. Donald Royall on pp 163 165). OBJECTIVES: To identify the clinical correlates of func- CONCLUSION: This study has shown that many nonde- tional incapacity in the community living old-old.'' mented very old people living in the community are losing DESIGN: Cross-sectional. capacity to perform IADL functions and that areas of in- SETTING: Community-based.

2 Capacity are associated with the presence of EP signs and impaired cognition. These results highlight the need for PARTICIPANTS: One hundred six nondemented people health workers to include an assessment of EP and cognitive aged 80 to 94. status in their evaluation of older persons living in the MEASUREMENTS: Participants were medically and cog- community, even in the context of a lack of dementia di- nitively assessed, underwent magnetic resonance imaging agnosis. Furthermore, it signifies the need to directly eval- scanning (MRI), and were interviewed regarding their func- uate IADL function to identify need for intervention and tional status: activities of daily living (ADLs), instrumental support if required. This group of old-old individuals may ADLs (IADLs), and the complex IADL functions of reading, now be considered the survivors'' of their cohort, and early hobbies, and socializing.

3 Detection of the difficulties they are experiencing will enable RESULTS: Dependency in IADLs, but not ADLs, was clinicians to respond appropriately, thus providing them a present. After controlling for age, sex, and education, ex- higher quality of life for their years to come. J Am Geriatr trapyramidal (EP) signs were significantly associated with Soc 54:3 10, 2006. two of the three IADLs, with EP signs comprising a com- Key words: functional impairment; normal elderly;. posite score of 10 EP signs ( , resting tremor) and a cognition; mild cognitive impairment 5-meter timed walk. cognitive test performance on a range of tests was also associated with functional status. A hier- archical model confirmed the association between the EP. signs and cognitive test performance and functional scores, but no pattern'' of cognitive association emerged.

4 Hippo- campal volume was associated with socializing. O ld-old'' individuals are the fastest-growing seg- ment of the population in developed countries. The incidence of dementia, as well as milder forms of cog- From the Prince of Wales Medical Research Institute and University of New South Wales, Sydney, Australia; wCentre for Education and Research on nitive impairment, is greatest in this age 4 Although Aging, Concord Hospital and University of Sydney, Australia; zSchool presence of dementia is well recognized as being associated of Psychology, University of Sydney, Australia; and Department of with functional incapacity, and its presence is usually re- Neurology, Concord Hospital, Australia. quired for diagnosis,5 the functional capacity of older peo- Funded by the National Health and Medical Research Council of Australia ple who do not suffer from dementia yet may be in need of and an Infrastructure Stream C grant from the Department of Health of New support and services remains unclear.

5 South Wales, Australia. OP is supported by a National Health and Medical Research Council Neil Hamilton Fairley postdoctoral fellowship (ID In past decades, research on aging and functional in- 222909). capacity has centered on dementia and its clinical associ- Address correspondence to Dr. Hayley P. Bennett, Prince of Wales Medical ations. Dementia research has more recently turned to the Research Institute, Barker Street Randwick NSW 2031, Sydney, Australia. detection of cognitive changes that precede clinical demen- E-mail: tia, a condition broadly referred to as mild cognitive im- DOI: pairment (MCI).6,7 Common to most MCI definitions is the JAGS 54:3 10, 2006. r 2005, Copyright the Authors Journal compilation r 2006, The American Geriatrics Society 0002-8614/06/$ 4 BENNETT ET AL.

6 JANUARY 2006 VOL. 54, NO. 1 JAGS. absence, or minimal effect, in them of any cognitive changes Of the 299 SOPS participants assessed at 6 years, 167. on functional capacity in terms of activities of daily living (56%) agreed to participate in the more-detailed MRI. (ADLs; , dressing, feeding, toileting).8 Such definitions study. Participants were only excluded if they were not ca- of MCI have implications for the interpretation of research pable of giving consent for the cognitive testing or MRI. on the functional status of older individuals, because people procedures, if they were practically unable to participate with MCI will typically be included in any sample of non- ( , sickness), met MRI exclusion criteria ( , cardiac demented elderly people. In this context, there have been a pacemaker or claustrophobia), were demented according to number of studies that have identified functional impair- the Clinical Dementia Rating scale (CDR; , score of ments in some groups of nondemented older people.)

7 1),19 or had a Mini-Mental State Examination (MMSE)20. For example, a review of the literature described a link score of less than 22. This resulted in a sample of 106 com- between instrumental activities of daily living (IADLs; , munity-living, nondemented individuals. Qualified clini- shopping, cooking, housework) and cognition in nonde- cians conducted all medical examinations and cognitive mented elderly This association has been re- assessments. ported in unselected older populations8 and in clinically The CDR was selected as the method to exclude par- defined groups with Population-based studies of ticipants with dementia, because it has been shown to be a aging have found extrapyramidal (EP) signs,11 depres- sensitive method of distinguishing normal aging from early sion,11 13 and executive or visuospatial deficits8,14,15 to be dementia;21 it has been used in previous studies evaluating associated with functional impairment.

8 One study of MCI functional status and cognition;22 and in a previous study groups10 found impairments in IADLs, as well as other on an earlier review of this same sample, a correlation of higher order'' or advanced'' activities such as preparing was found between the informant CDR and clinical tax records and writing checks, and another identified mild (geriatrician) dementia diagnosis23 using Diagnostic and impairments in capacity to manage their Statistical Manual of Mental Disorders, Fourth Edition cri- Thus, there is evidence of a relationship between cog- Given the close relationship demonstrated between nitive and motor deficits, as well as presence of depression, CDR and geriatrician dementia diagnosis in this sample, a and functional impairment in nondemented individuals, in MMSE score of less than 22 was chosen as the additional particular for IADL functions, although the independent cutoff to retain a representation of the full range of com- relationship between these variables and functional status munity-living elderly.

9 Such a span of MMSE scores for remains unclear. Furthermore, of the studies cited above, all nondemented older people is consistent with the findings examined subjects with average ages in their 70 s at most from other community-based and only examined a limited range of variables. Whether Retention of participants across waves was reasonable this relationship holds in the old-old, where the implica- for a longitudinal study of this age group, with 47% of the tions for this association are greatest, is uncertain. original sample being available for assessment at 6 years, of The first aim of this study was to determine the prev- whom 35% were able to participate in the MRI procedures alence of functional incapacity in a group of old-old com- (17% of original sample). The dropout from this study was munity-living nondemented individuals.

10 The second aim mainly due to illness or death. Those who had died by the 6- was to identify the clinical correlates of different types of year follow-up were older, more likely to be male, and more functional incapacity. Aspects of functional capacity eval- cognitively impaired at the baseline SOPS assessment. Re- uated included ADLs and IADLs, as well as other complex fusal and noncontacts did not differ in age or MMSE score activities such as reading and socializing. To assist clinicians at in their evaluation and support of older people, a broad range of indicators potentially associated with functional Participant Interview and Functional Capacity decline were examined. These indicators conceptually en- Participants were interviewed on a range of health measures compassed information available at clinical assessment, and questionnaires, including medical history, lifestyle, such as cognitive and psychological status, as well as EP.