| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
|
|
|
|||||||
|
|||||||||
a Departments of Internal Medicine, University of Michigan, Ann Arbor
b Departments of Health Management and Policy, University of Michigan, Ann Arbor
c Institute for Social Research, University of Michigan, Ann Arbor
d Institute of Gerontology, University of Michigan, Ann Arbor
Caroline S. Blaum, Center on Aging and Health, 2024 E. Monument Street, Suite 2-700, Baltimore, MD 21205 E-mail: cblaum{at}mail.jhmi.edu.
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionAppendixReferences |
|---|
Methods. Data were from the first wave of the Assets and Health Dynamics Among the Oldest-Old Study, a national panel survey of individuals age 70 and older (n = 6600 age-eligible self-respondents). Low cognitive performance (LCP) was defined as scores in the lowest (poorest performing) 25th percentile of a cognitive performance scale. The associations of LCP with prevalent chronic diseases and conditions and with limitations in 14 tasks (strength and mobility, instrumental activities of daily living, and activities of daily living) were evaluated. Associations of LCP and task limitations were adjusted for potential modifiers and confounders, including demographic characteristics (age, gender, race), educational attainment, chronic diseases, depressive symptoms, and sensory impairments. Data were weighted to account for complex sample design and nonresponse.
Results. More than one third of people with LCP had three or more coexisting diseases and conditions. The unadjusted associations of LCP with task functioning were attenuated after covariate adjustment, but even after adjustment, LCP remained significantly and independently associated with functioning problems in 9 of 14 tasks (borderline with four more), including mobility tasks.
Conclusions. Low cognitive performance, regardless of its relationship to clinical dementia, coexists with multiple chronic diseases and conditions. It is independently associated with a broad array of functioning difficulties, even after controlling for demographic characteristics, educational attainment, and chronic conditions. Chronic diseases and conditions, however, attenuate the relationship between LCP and some task difficulties. LCP should be considered an important comorbid condition associated with both chronic diseases and disability that substantially increases the health burden of many older adults who are poorly equipped to handle it.
AS the population ages, dementia and its associated disability are emerging as major threats to the health and well-being of older adults (1)(2)(3)(4). The prevalence of dementia increases with age (5), paralleling the increased incidence and prevalence with age of Alzheimer's disease, other neurodegenerative conditions (6), and stroke, the major causes of clinical dementia (7).
The relationship of disability to dementia has been examined in some detail. Presumed or diagnosed dementia is associated with difficulties in prevalent activities of daily living (ADLs) (1)(8)(9)(10)(11)(12)(13)(14) and instrumental activities of daily living (IADLs) (3)(4)(15) and predicts ADL decline (3)(4)(11)(15)(16). Findings have been mixed as to whether dementia or presumed dementia is related to difficulty in tasks associated with mobility and strength, such as walking, climbing stairs, or carrying heavy objects (3)(11)(15).
Recently, it has become apparent that some older adults have detectable, mild cognitive impairments, but cannot be classified as "demented" by current diagnostic criteria. Impairment in memory (17) and/or executive function (18) have been considered key cognitive domains affected in those with mild or subclinical cognitive impairment. Those who demonstrate mild cognitive impairment are reported to be at increased risk for progression to Alzheimer's disease (19)(20) and cerebrovascular disease (21). Some people with mild cognitive deficits may have early dementia, while cognitive deficits in others may be due to medical disease burden (22)(23)(24). Many demented elders living in the community remain undiagnosed (25).
So far, few population-based studies have evaluated the health status correlates of low cognitive performance (LCP), whether it is due to mild cognitive impairment, subclinical dementia, or undiagnosed dementia. Nationally representative surveys concerned with chronic disease and disability have not always measured cognitive functioning due to time constraints, concerns over respondent burden, and methodological challenges. Even if cognitive functioning has been measured, it is rarely measured across several domains. In addition, poor cognitive performance is difficult to define in large population-based studies because it is a composite of several types of cognitive problems (undiagnosed dementia, early or subclinical dementia, mild cognitive impairment, temporary cognitive impairment, or longstanding low cognitive performance) (25)(26). Despite all these difficulties, LCP is important to study because the few population-based studies available suggest that its public health impact in older people may be significant (21)(26)(27).
The goal of our research was to specifically evaluate the health burden associated with LCP in older people in the United States who live in the community. Our research strategy had several components: (i) We studied a sample of older adults representative of the entire U.S. population aged 70 and older living in the community in 1993. (ii) We defined LCP empirically, using respondent scores on a unique measure of cognitive impairment designed specifically for a population-based survey. Our survey data do not allow us to define "mild cognitive impairment" clinically. Rather, we identified those who performed poorly relative to other older adults sampled, excluding those with the most severe cognitive and medical problems (see detail in methods). We fully expect that our LCP group contains people with clinical dementia, early or subclinical dementia, mild cognitive impairment (as currently defined [17]), temporary cognitive impairment, and lifelong poor performance. (iii) We studied the relationship of LCP to a broad spectrum of chronic diseases/conditions and disability tasks to evaluate the hypothesis that LCP would show pervasive co-occurrence with both comorbid disease and functioning difficulty. (iv) We controlled for co-occurring chronic diseases and conditions, sensory impairments, depressive symptoms, and demographic and social factors in order to clarify the independent association of LCP with disability.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionAppendixReferences |
|---|
Baseline AHEAD interviews were conducted with sampled respondents and their spouses, resulting in a sample of 7442 people aged 70 years and older. When the eligible respondent was unable to be interviewed himself or herself (often due to significant cognitive or medical problems), a proxy respondent was interviewed. Our research used only age-eligible self-respondents (n = 6660) and data from the first wave (1993). Descriptive characteristics of the AHEAD sample have been published by Soldo and colleagues (28). Of the 7442 age-eligible respondents, the mean age was 77 ± 4.6 years, 62.4% were women, 13.2% were African American, and 13.4% had some education beyond high school (28).
Model Development and Variables
The dependent variables in this research were 14 of the task-specific measures of functional limitations and disabilities that are available in AHEAD. These functional tasks were modeled separately to assess how associations between functional limitations on one hand and LCP and other diseases and impairments on the other varied across the tasks. All 14 dependent variables, described below, were categorized as completely able to be performed versus able to be performed with any amount of difficulty or with any help.
Functional limitation and mobility tasks..-- Difficulties with the three mobility tasks (walking several blocks, walking across a room, and climbing one flight of stairs) and one strength task (lifting 10 pounds).
IADL tasks..-- Difficulties with the three "cognitive" IADLs (using the telephone, managing money, and taking medications) and two complex IADLs (grocery shopping and preparing meals).
ADL tasks..-- Difficulties with five personal care tasks (transferring in/out of bed, dressing, bathing, toileting, or eating; walking was included with the mobility/functional limitation items).
The study's independent variables included the following:
Low cognitive performance (LCP)..-- Age-eligible self-respondents who scored in the lowest quartile of a cognitive impairment scale, which was a composite of 35 questions testing cognitive function, were empirically defined as the LCP group. There was no exact cutoff score for the 25th percentile, so the closest, lowest cutoff score was chosen. For the 25th percentile, that number was 15 correct (LCP group n = 1383, scores of 15 or below). For the LCP group, the mean number of correctly answered items from this 35-item composite measure was 11.53 (standard error [SE] 0.07, range 1–35); for the top three quartiles, the mean number correct was 22.13 (SE 0.09). The common epidmiologic technique of comparing the lowest quartile to higher quartiles was chosen because our research interest specifically concerned respondents who perform poorly relative to other older adults, and thus represent a broad spectrum of community-living people with low cognitive performance. The LCP group does not represent significantly demented older people, because most of those who are highly impaired are either institutionalized and therefore out of scope of the survey, nonrespondents due to refusal or inability to contact, or interviewed by proxy and thus excluded from the analysis.
The 35 questions measuring cognitive impairment administered in AHEAD include items from a mental status questionnaire and tests of memory and reasoning (see A, Note 1). Taken together, these measures are intended to differentiate across the full range of cognitive functioning (see A, Note 2). Because the cognitive impairment tests had to be suitable for administration over the telephone, many were drawn from the Telephone Interview for Cognitive Status, which is modeled after the Mini-Mental State Exam (7) and has been validated specifically for phone use (29).
Chronic diseases and conditions..-- The causal relationship between disability and disease (30) and the association of comorbidity with disability (31) are well known. Therefore, models evaluating the relationship of LCP and functional limitation/disability tasks controlled for chronic diseases and conditions. We also explored the extent to which LCP is associated with co-occurring chronic diseases and conditions. Although some chronic diseases may be causally associated with LCP (diabetes, stroke, pulmonary disease), our cross-sectional data cannot evaluate this issue. Also, several diseases with which LCP may be causally associated, such as Alzheimer's disease and Parkinson's disease, are unmeasured in this and other self-report surveys.
Chronic diseases and conditions were handled as follows: heart and lung diseases were combined into the cardiovascular group; arthritis, back problems, and hip fracture were combined into the musculoskeletal group; diabetes and stroke were considered separately. A variable indicating the presence of miscellaneous other diseases not consistently associated with task difficulty (i.e., hypertension and cancer) was also used.
Sensory impairments..-- Self-reported problems with vision and hearing were modeled separately. Impaired vision was defined as a respondent's assessment of his or her vision as fair, poor, or blind. Similarly, impaired hearing was defined as a respondent's assessment of his or her hearing as fair or poor.
Depressive symptoms..-- Some researchers have modeled depressive symptoms as resulting from chronic diseases and conditions (32) and others as contributing to disability, presumably resulting from severe and milder psychiatric disease (3)(4)(33). Depressive symptoms are also related to cognitive functioning, although the exact nature of the relationship is not yet clear. Therefore, we controlled for the independent, direct cross-sectional association of depressive symptoms to disability. Depressive symptoms were measured by a composite variable formed from eight items of the Center for Epidemiological Studies–Depression scale (34) adapted specifically for the AHEAD (28). Depressive symptoms were considered to be present if the respondent affirmed depressive symptoms in six of the eight questions.
Demographic factors..-- Also included in our models of task-specific functioning difficulty were social and demographic variables well known to be associated with cognitive functioning, functional limitations, and disability (35). Although the contribution of these variables has varied in different studies, they clearly must be considered in models that evaluate functioning difficulty (32)(36). Demographic variables were controlled in all models and included age (continuous variable), race (African American or not), educational attainment (dichotomized as education beyond high school or not), and gender. Mode of administration, by telephone or face-to-face, was also controlled in all adjusted analyses. This was done because face-to-face administration depended upon age and respondent preferences. Published analyses, however, have not found any evidence that mode of administration introduces significant bias into the AHEAD interview results (27).
Statistical Analysis
The chi-square of association was used to test the crude, prevalent associations of low cognitive performance to chronic diseases and conditions, and then to the 14 task-specific functioning and disability measures. To examine the adjusted association of LCP to the task-specific measures, logistic regression models were adjusted for demographic characteristics, educational attainment, chronic diseases and conditions, sensory impairments, and depressive symptoms. Mode of administration (i.e., in person or by telephone) was also controlled. All analyses were weighted to adjust for differential probability of selection, nonresponse, and complex sample design using appropriate weights and the statistical package Stata (Stata, College Station, TX) (37). The logistic regression models were evaluated by standard techniques to assess significance and goodness of fit (38). Standard techniques for evaluating significance and overall goodness of fit in logistic regression models assume simple random samples and cannot be used when design effect is considered. However, because estimates are virtually the same, variances did not vary widely, and standard techniques showed appropriate models in those weighted only for selection probability, it was assumed that the fully weighted models were adequate.
To evaluate potential problems with multiple comparisons due to estimating 14 models, the sample was randomly divided in half to form an analytic sample and a validation sample. The samples were compared for any statistical difference in key variables, and none was found. These models were compared with each other and with full sample models. For four task difficulties, associations were significant in one subsample and not the other, so these associations are considered to be borderline significant. All other associations were nearly identical in the analytic and validation samples. Therefore, results of the whole sample are presented for both the adjusted and unadjusted analyses for all tasks.
Item nonresponse was not a major problem in the models (<5% missing) except for the composite cognitive impairment variable. Two hundred ninety respondents did not respond to any of the 35 questions. Although these respondents are likely to be similar to respondents in the LCP group, in the absence of all information on their cognitive performance, they were excluded from analyses. However, we did evaluate these respondents in a sensitivity analysis, which was done to assess both our empirical use of the lowest 25th percentile cognitive score as the cut-point for the LCP group and the impact of the 290 people who did not respond to any cognitive measure. We compared the log-likelihoods of the logistic regression models described above using four different definitions of the LCP variable: (i) as a continuous variable (35 levels); (ii) dichotomized at two lower cut-points, the 10th and 5th percentiles, because estimates of dementia prevalence in an independent-living community population suggest that somewhere between 5% and 10% of older adults would be expected to have clinical dementia (39); (iii) model in which the LCP group included the 290 nonresponders; and (iv) model chosen to report, in which the 290 nonresponders were dropped from the analyses. For all models, our choice of the model that used a cutoff point at the 25th percentile and excluded the 290 nonresponders was associated with a better or equal overall model fit as determined by log-likelihood analyses.
To assess the relationship of LCP in terms of its defined prevalence in this study (lowest 25th percentile) and its odds of being related to difficulty in a particular task, the adjusted attributable fraction was calculated using the formula
AF = [(p)(aOR - 1)/aOR]
where p is the proportion of those with a particular task difficulty who have LCP and aOR is the adjusted odds ratio (3).
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionAppendixReferences |
|---|
|
|
Table 2 illustrates the relationship of LCP and task-specific disability measures of mobility and strength. LCP was strongly associated with difficulty in all functioning tasks before adjusting for confounders; many proportions were doubled for those in the LCP group versus all others.
|
Table 3 shows prevalences, crude associations, and the adjusted odds ratios and 95% confidence intervals for the relationships of LCP with IADLs. Although the odds ratios decreased when the model was fully adjusted, LCP remained independently and strongly associated with difficulty in all these tasks. The strongest associations were noted with difficulty in handling medications (3.2, 2.1–5.0) and money (2.7, 2.2–3.3).
|
|
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionAppendixReferences |
|---|
Low cognitive performance was independently associated with limitations in multiple functional tasks. Unadjusted associations of LCP with difficulty in all 14 tasks studied were strong and significant. After full multivariate adjustment, only one task difficulty had no significant association with LCP; four had a borderline association; and nine remained strongly and significantly associated. For more than half of these tasks, 20% to 40% of the functioning difficulty in the older population is specifically associated with low cognitive performance and is not explained by demographic characteristics, educational attainment, or prevalent chronic diseases and conditions, including sensory problems and depressive symptoms.
Disabilities associated with low cognitive performance represented a broad array of tasks and included those represented in multiple disability domains. Tasks such as managing money and medications, traditionally considered tasks related to cognition, were highly associated, as were complex tasks such as grocery shopping and preparing meals. We noted somewhat more attenuation of the association between LCP and these IADL tasks by adjustment for chronic diseases and conditions, suggesting a potentially larger role for chronic conditions in mediating these relationships. Future research and longitudinal evaluation would be necessary to test this hypothesis.
Two tasks associated with mobility, walking across a room and climbing stairs, were also significantly associated with low cognitive performance. The relationship of LCP to mobility limitations has been noted before (3)(4)(15), but often the mobility item "walking across a room" is included within an ADL scale. The cognition to mobility relationship noted in this study is not likely due to major strokes because all models controlled for reported stroke. Of course, early cerebrovascular disease and strokes that are not noted by respondents would be unmeasured and have been associated with LCP in other population-based studies (21). In addition, LCP may be associated with other neurological impairments such as peripheral neuropathies, unsteady gait, and balance problems that are related to mobility limitations. Such problems, also unmeasured in this data, could be associated with early degenerative neurological diseases or severe medical diseases, in turn associated with LCP.
LCP was not uniformly associated with limitations in personal care (ADL limitations). A potential explanation for this is the low prevalence of difficulties with some of these tasks in the sample and the fact that the most impaired respondents, who required proxies, were excluded.
The two major findings of this research are important to consider together. Despite the co-occurrence of LCP with multiple diseases and conditions, LCP remains independently associated with most task limitations after adjusting for chronic diseases and conditions. This independent association of LCP with task limitations is not surprising because LCP should be related to early degenerative neurological disease (20) or preclinical atherosclerotic brain disease (21). However, some respondents may have LCP related to nonneurological chronic diseases, because adjustment for chronic diseases and conditions does attenuate the odds ratios for the associations of LCP and some task limitations. While self-reported chronic medical diseases have generally explained only a small part of the variance in cognitive impairment in other studies (27)(40), LCP may be associated with severe medical disease and other disease-related variables, such as medications, which are unmeasured in these data. For example, although we controlled for heart disease, diabetes, and depressive symptoms (among other conditions), our data do not describe how severe these conditions were or how many medications people reporting these conditions were taking.
Given our reliance on self-reported data, it is important to consider the potential influence of low cognitive impairment on reports of chronic disease and functional limitations. As noted by Belli and colleagues (41), studies that have compared older and younger respondents in terms of the accuracy of self-reports of factual information and retrospective behaviors (e.g., automobile ownership, voting behavior, health care utilization) have either found no age differences (42)(43) or some underreporting on the part of older respondents (44). A recent report on the accuracy of self-reported health services use among older adults found that while older adults underreported utilization, no specific characteristics were correlated with inaccurate reports (45). There have been a few studies on self-report of medical conditions, which have generally found that report accuracy, when compared to the medical record (which may also be inaccurate) or objective data (46), depends more on the specific disease than the characteristics of the older respondent (46)(47)(48)(49)(50)(51). Although older adults may underreport certain diseases, particularly cardiovascular diseases, variations in cognition have not been shown to be related (47). For other diseases or conditions (i.e., rheumatoid arthritis, hip fracture), self-report by older adults has been found to compare favorably to the medical record (48) or to the general adult population (49).
This study has several limitations besides our use of self-reported information. Our data were cross-sectional so we cannot test hypotheses about causation and the direction of relationships. However, because longitudinal data from the AHEAD are now available, this research forms the first step in the study of the antecedents and outcomes of LCP and the relationship of LCP to disability transitions. The cutoff point for our measure of LCP was arbitrary, although we had a substantive basis for that choice, and it was based on a standard epidemiological technique. In addition, our sensitivity analysis suggested there was reasonable statistical criteria for using that cutoff. Also, our chosen prevalence is similar to prevalences reported using other methods in other populations. For example, the prevalence of mild cognitive impairment measured using completely different methods has been reported as 16% (26) and 24% among those aged 65 to 79 (52).
This research makes several contributions to our knowledge about the relationship of low cognitive performance to functioning. First, we examined a broad array of task-specific limitations and disabilities. Although it is often reasonable to aggregate different tasks, from a practical and clinical point of view, an individual's problems with specific tasks are highly relevant in order to provide assistance. From an analytic point of view, aggregating tasks can obscure the relationship of LCP across the full spectrum of functional difficulty.
Second, the data used in this analysis come from a nationally representative sample of older adults in the United States and contain information about many diseases, conditions, and disabilities. As such, it joins the few other population-based studies such as the Established Populations for Epidemiologic Studies of the Elderly (53), the Canadian Study of Health and Aging (54), and the Kungsholmen Project (26) to complement more clinically oriented studies of volunteers, studies without the full spectrum of disability or cognitive measures, and studies of people who are already disabled.
Finally, the cognitive measures used have been designed specifically to capture a wide range of functioning. The rationale for the choice of these measures and their psychometric properties and construct validity have been described in detail (27). Unfortunately, the few population-based studies available have used different measures to define low cognitive performance (26)(52)(54), so there is not yet a standard definition. However, one suggested definition of "mild cognitive impairment" specifically states that ADLs are not impaired (17). In our study, we found that ADLs were not strongly associated with LCP as we defined it, suggesting at least some empirical comparison between definitions. Regardless, our LCP group should be considered to include people with multiple types of cognitive problems, including clinical dementia, early or subclinical dementia, mild cognitive impairment, temporary cognitive impairment, and lifelong poor performance.
This research demonstrated that, regardless of the ultimate diagnosis, low cognitive performance is a marker for difficulty in a wide variety of functional tasks. While LCP coexists with multiple chronic diseases and impairments, it is independently associated with difficulty in most functional tasks. Therefore, LCP should be considered an important comorbidity with a significant impact on disability in older adults. Our results also point out the need for longitudinal population-based studies to answer several clinically related questions about LCP: What proportion of LCP is related to early or preclinical dementia and what proportion is related to chronic disease burden? If LCP is related to chronic disease burden, what mediates that relationship—disease severity, medications, disease interactions? Would management of comorbid conditions improve functioning in people with LCP? Answers to such questions will have important implications for clinical practice. Already, routine evaluation of cognitive impairment in older adults is becoming more common as preclinical dementia and/or early cognitive decline is increasingly recognized by physicians, patients, and families. Interventions such as medication review, appropriate management of chronic disease, use of cholinesterase inhibitors, and stroke prevention measures are now offered to some people with LCP. Unfortunately, older adults with LCP would be expected to have significant difficulty handling medical and lifestyle instructions associated with comorbid chronic diseases, on the one hand, and adapting to disabilities, on the other hand. Future research is needed to evaluate whether and how LCP can be treated and whether such treatment will result in improved functioning for many older adults.
|
Acknowledgments |
|---|
We thank Tisha Moore for her manuscript preparation and assistance with development of figures and tables.
Received October 18, 2001
Accepted February 1, 2002
|
Appendix |
|---|
|
TopAbstractMethodsResultsDiscussionAppendixReferences |
|---|
a. Immediate recall test: Interviewers read a list of 10 short, concrete, high-frequency nouns to the respondent. As soon as the list had been read, respondents were asked to repeat as many words as possible. Respondents were given 1 point for each word that they correctly recalled for a maximum of 10 points. Together with the delayed recall test described below, this test is thought to measure memory impairment.
b. Serial 7s test: Respondents were asked to start with the number 100 and subtract by 7 for five trials. Respondents were given 1 point for each correct subtraction for a maximum of 5 points. This is thought to test working memory.
c. Backwards count: Respondents were asked to count backwards from 20 for 10 continuous numbers. If a respondent failed on the first try, he or she was given a second try. Two points were given for a correct outcome on the first try, 1 point for the correct outcome on the second try, and 0 points if the respondent failed on both tries. This test is thought to be a test of the speed of processing.
d. Naming tests: Respondents were asked to name the day of the week and the date (day, month, year), two objects that were described by the interviewer ("what people usually use to cut paper" and "kind of prickly plant that grows in the desert"), and the current President and Vice-President of the United States. One point was given for a correct response on each naming test, for a maximum of 8 points. These tests are thought to measure orientation and knowledge.
e. Delayed recall test: At the end of the cognition section, respondents were asked to recall as many as possible of the 10 nouns that had been previously read approximately 5 minutes earlier by the interviewer. Respondents were given 1 point for each word that was correctly recalled for a maximum of 10 points.
2. Analysis of the AHEAD cognitive impairment measures has indicated that the measures display desirable psychometric properties. As part of their evaluation of the AHEAD cognition measures, Herzog and Wallace (27) found that the measures contain a memory dimension and a second dimension that was labeled mental status. They also found that the measures formed associations with sociodemographic and health correlates that replicate well-established relationships in other studies. Thus, the measures appear to have strong internal and construct validity.
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionAppendixReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. L. Fitzpatrick, C. K. Buchanan, R. L. Nahin, S. T. DeKosky, H. H. Atkinson, M. C. Carlson, J. D. Williamson, and for the Ginkgo Evaluation of Memory (GEM) Study In Associations of Gait Speed and Other Measures of Physical Function With Cognition in a Healthy Cohort of Elderly Persons J. Gerontol. A Biol. Sci. Med. Sci., November 1, 2007; 62(11): 1244 - 1251. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. T. Cigolle, K. M. Langa, M. U. Kabeto, Z. Tian, and C. S. Blaum Geriatric Conditions and Disability: The Health and Retirement Study Ann Intern Med, August 7, 2007; 147(3): 156 - 164. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Inzitari, M. Baldereschi, A. D. Carlo, M. D. Bari, N. Marchionni, E. Scafato, G. Farchi, D. Inzitari, and for the ILSA Working Group Impaired Attention Predicts Motor Performance Decline in Older Community-Dwellers With Normal Baseline Mobility: Results From the Italian Longitudinal Study on Aging (ILSA) J. Gerontol. A Biol. Sci. Med. Sci., August 1, 2007; 62(8): 837 - 843. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. H. Dodge, Y. Du, J. A. Saxton, and M. Ganguli Cognitive Domains and Trajectories of Functional Independence in Nondemented Elderly Persons J. Gerontol. A Biol. Sci. Med. Sci., December 1, 2006; 61(12): 1330 - 1337. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Littbrand, E. Rosendahl, N. Lindelof, L. Lundin-Olsson, Y. Gustafson, and L. Nyberg A High-Intensity Functional Weight-Bearing Exercise Program for Older People Dependent in Activities of Daily Living and Living in Residential Care Facilities: Evaluation of the Applicability With Focus on Cognitive Function Physical Therapy, April 1, 2006; 86(4): 489 - 498. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Liang, J. W. Brown, N. M. Krause, M. B. Ofstedal, and J. Bennett Health and Living Arrangements Among Older Americans: Does Marriage Matter? J Aging Health, June 1, 2005; 17(3): 305 - 335. [Abstract] [PDF]
|
|
|
|
|
|
D. R. Thomas The Relationship Between Functional Status and Inflammatory Disease in Older Adults J. Gerontol. A Biol. Sci. Med. Sci., November 1, 2003; 58(11): M995 - 998. [Full Text] [PDF]
|
|
|
|
 |
|
 V K Podichetty, D J Mazanec, and R S Biscup Chronic non-malignant musculoskeletal pain in older adults: clinical issues and opioid intervention Postgrad. Med. J., November 1, 2003; 79(937): 627 - 633. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Morley Editorial: Sarcopenia Revisited J. Gerontol. A Biol. Sci. Med. Sci., October 1, 2003; 58(10): M909 - 910. [Full Text] [PDF]
|
|
|
|
|
|
J. E. Morley Editorial. Mobility Performance: A High-Tech Test for Geriatricians J. Gerontol. A Biol. Sci. Med. Sci., August 1, 2003; 58(8): M712 - 714. [Full Text] [PDF]
|
|
|
|
|
|
W. A. Banks and J. E. Morley Memories Are Made of This: Recent Advances in Understanding Cognitive Impairments and Dementia J. Gerontol. A Biol. Sci. Med. Sci., April 1, 2003; 58(4): M314 - 321. [Full Text] [PDF]
|
|
|
|
|
|
J. E. Morley Editorial: Hot Topics in Geriatrics J. Gerontol. A Biol. Sci. Med. Sci., January 1, 2003; 58(1): M30 - 36. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
