Descriptions of IALSA Affiliated Studies

The longitudinal studies listed here have rich data in terms of both health and cognitive outcomes. Principal investigators from each of these studies will be actively engaged in the analyses proposed here and have given written permission for the use of these datasets in this proposed collaborative project. The preceding table provides a summary of study characteristics for longitudinal studies that will be analyzed as part of this project. Further description of each study and major findings follow. A detailed listing of common and specific study variables is provided in Appendix B. References to published longitudinal findings from these and other studies are included in Appendix C.

Australian Longitudinal Study of Ageing (ALSA). The South Australian Electoral Roll was used as a sampling frame to identify households with residents over 70 years of age (Hugo, Healy, & Luszcz, 1987). The sample was stratified by age and gender into three five-year cohorts: 70 - 74, 75 - 79, 80 - 84, and a fourth cohort of individuals older than 85 years of age. Randomly sampled individuals within these cohorts were invited to participate in the ALSA on a voluntary basis. The participation rate for the baseline data collection (Wave 1) was 55%. The study comprises seven waves of data collection: baseline, between September 1992 and March 1993, and three subsequent waves at approximately 12 month intervals; the most recent three waves were at 2 to 2.5 year intervals. Further waves are planned, pending funding. Waves 2, 4 and 5 comprised telephone interviews and did not include a clinical or cognitive assessment, therefore, data from Waves 1, 3, 6 and 7 only will be used in this study. At these waves, a comprehensive two-hour home interview was followed by an optional individual clinical assessment conducted approximately two weeks later. The home interviews yielded demographic data and information on self-rated health, depression, medical conditions, cognitive status, memory and subjective measures of well-being, vision, audition and physical performance. Individual clinical assessments provided objective cognitive and sensory data. At Wave 1, 2087 (1056 males) participants were interviewed and 1620 (77.6%, 843 males) underwent the clinical assessment. At Wave 3, 1679 participants were interviewed (84.8%, 824 males) and 1423 underwent the clinical assessment (707 males). At Wave 6, 791 participants were interviewed (334 males) and 525 (66.4% 223 males) underwent the clinical assessment. At Wave 7, 486 participants were interviewed (156 males) and 396 (81%) underwent the clinical assessment. Deaths have been censored (monitored) comprehensively, most recently to June 2004.

This study aims to determine physical, mental, behavioral and social features of ‘healthy’ or ‘active’ aging (Andrews, et al., 2002). Initial cross-sectional analysis of individual differences in memory performance found that processing speed was the major mediator of age-related variance in memory with measures of depression, activity, gender and health having minimal influence (Luszcz, et al., 1997).  Initial longitudinal analysis of the ALSA data (Luszcz, 1998) showed decline in speed, picture memory, depression, morale, and self-esteem over the first two years of the study.  Subsequent cognitive studies have expanded to examine the role of sensory and mediators (Anstey, Luszcz & Sanchez, 2001a, 2001b), to model change in (Anstey, et al., 2003b) and dedifferentiation of, cognitive abilities over 8 years (Anstey, Hofer & Luszcz, 2003a) and to explore attrition and possible dementia (Anstey & Luszcz, 2002). Work on mortality (or longevity) has shown protective effects from cognition, when health, sensory and affective factors are controlled.  Gender differences examined in relation to mortality show that perceived control over one’s life is protective of survival for both men and women (Anstey et al., 2002). In looking at depressive symptoms, while they are more common among women in ALSA, their presence does not increase the risk of death among women but incident depression in men is strongly related to reduced survival (Anstey & Luszcz, 2002; Anstey, Luszcz & Andrews, 2002). Women also derive benefit from maintenance of social relationships; the sparseness of these relationships in men may contribute to their not deriving similar protective effects from them. The role of social networks has also been explored across the first 6 waves of ALSA; by differentiating kin and non-kin networks (Giles et al., 2002); we have shown that non-kin networks are particularly protective of longevity (Giles et al., in press). Additional work has been done on anthropometric and activity indices, blindness and lifestyle factors implicated in ageing well.

Bonn Longitudinal Study of Aging (BOLSA). The Bonn Longitudinal Study on Aging (BOLSA, Lehr & Thomae, 1987; Thomae, 1976, 1993) was the first 'global' longitudinal study in Germany to examine psychological and medical issues of aging. The aim of the study was to assess the psychological, social, and physical status of the elderly population and its developmental changes. Based on the assumption of increasing interindividual variability with increasing age no specific hypotheses were developed and tested. Instead, a large number of variables were collected to identify patterns of aging. BOLSA followed a cross-sequential design. Two cohorts, born between 1890-1895 and 1900-1905, respectively, were tested up to eight times during 1965 and 1984. During the nineteen years of data collection the number of participants decreased from n = 221 to n = 34. The dropout in participants (85%) was primarily due to severe illness or death rather than a lack of motivation. At each measurement point extensive medical and psychological examinations were conducted. Medical evaluation included assessment of health status, physical capacity, physical independence, and numerous physiological parameters. Moreover, a specialist for Internal Medicine assessed overall health status, the capacity of the organs, and the etiology and severity of diseases. Psychological analysis included assessment of intelligence, learning, memory, psychomotor speed, and personality. Extensive biographical interviews were carried out, focusing on individual's life history, on the present situation (including interests, activities, social network, and social contacts, family, health, housing, and economic situation), and on future-oriented perspectives (plans, expectations, hopes, fears). Since the participants stayed in Bonn for one week, it was possible to observe and rate the individual's behavior.

The main results of the study demonstrate that based on large interindividual variability aging is a highly individualized process. They also indicate that older individuals are quite capable of coping with challenging situations and that they use a number of different coping behaviors to successfully deal with these challenges. A large number of different patterns of aging could be identified. The analysis of intellectual development clearly demonstrated a considerable decline in mechanical functions but much stability in pragmatic functions during the 'third' age. Although data collection was completed over twenty years ago, the study continues to be valuable from the perspective of developmental psychology and for interdisciplinary analyses. Second and third generation researchers have started in the late 1990s to introduce new topics, such as sensory functioning and cognition (Rott, 1995) and new methodologies such as latent growth curve models and factor-of-curve models (Zimprich, 1998, 2002).

The Bronx Aging Study (BAS): In 1980 Robert Katzman, M.D. established the BAS with the aim of prospectively following a volunteer group of ambulatory, active, non-demented Bronx residents aged 75 to 85 years in order to track the incidence of dementia. The study enrolled 488 community volunteers whose baseline assessments occurred between1980-1983; follow-up testing ended in 1998 (the cohort is now closed). Yearly assessments of each participant included a neuropsychological test battery, a neurological examination, blood tests, and social and behavioral questionnaires. CT and EEG were performed when subjects developed cognitive change, defined by a cumulative increase of 4 points from baseline on the Blessed Information Memory Concentration test (BIMC), a BIMC score of more than 8 errors at any point, or changes in behavior suggestive of dementia reported by significant others or by study personnel. A diagnosis of dementia was made according to modified DSM-III criteria (American Psychiatric Association, 1980). Follow-up continued until an outcome was obtained for each participant. Possible outcomes included: 1) clinical diagnosis of dementia, 2) dropout from study participation, and 3) death. Follow-up ranged from none to 16 years. The average length of follow-up for noncases was 5 years, and the average time to diagnoses was 4.7 years for the preclinical cases. Diagnoses were made without using information about performance or decline on a memory test . A total of 98 (25%) participants from the initially nondemented sample were eventually diagnosed with some form of dementia. Of the 98 incident cases, 67% were subtyped with possible or probable Alzheimer’s disease, 24% with vascular dementia, and 9% with other etiologies (e.g., Lewy body disease).

Longitudinal studies in the BAS have demonstrated that cognitive changes associated with normal aging are significantly influenced by the contamination of normative samples by individuals with unidentified preclinical dementia (Sliwinski, Lipton, Buschke, & Stewart, 1996; Sliwinski & Buschke,1997). Others findings demonstrate that while processing speed mediates longitudinal changes in memory and cognition to some degree (6% to 29%), longitudinal mediation accounts for much less variance compared to cross-sectional age-mediation models (70%-100%; Sliwinski & Buschke, 1999). Sliwinski, Hofer, Hall, Lipton, & Buschke (2003; see also Sliwinski, Hofer, & Hall, 2003) examined memory loss in a sample of 391 initially non-demented older adults who were tested annually for up to 16 years. Analyses decomposed total memory loss into decline associated with the presence and progression of preclinical dementia, study attrition, terminal decline, and chronological age. Modeling memory as a function of only chronological age failed to provide an adequate representation of cognitive change while disease progression accounted for the virtually all of the memory loss in the 25% of the sample that developed diagnosable dementia. In the remainder of the sample, both chronological age and study attrition contributed to observed memory loss.

 Caerphilly Cohort Study of Older Men (CCS). The Caerphilly study has been called the ‘UK’s Framingham’ with over 170 published papers. Caerphilly was begun in 1979 as a study of heart disease, with a complete sample of 2512  (89%) men aged 45-59 years living in and around the town of Caerphilly, South Wales. The fifth wave was completed in 2004. A broad range of risk factors have been studied including lipids, diet, psychosocial and lifestyle factors and more recently, outcomes of stroke, cognitive decline and dementia.

“Baseline cognitive function was established for a study of pre- symptomatic cognitive decline in 1870 men from the general population aged 55–69 years as part of the third examination of the Caerphilly Study. Cognitive assessment included the AH4, a four choice serial reaction time task, a modified CAMCOG, MMSE, NART and various memory tests. Distributions and relationships with age, social class, education and mood at time of testing are presented for a younger population than has previously been available. Multiple linear regression showed cognitive function to be independently associated with all four factors. The age effect was equivalent to one half of a standard deviation (SD) in CRT and AH4 scores. Only the NART score was not associated with age, supporting the use of NART score as an estimate of pre-morbid IQ” Gallacher et al., (1999).

Canberra Longitudinal Study (CLS). This is a probability sample of 897 people aged 70 years and older drawn from the electoral roll for Canberra and the neighboring town of Queanbeyan, Australia. The sample is predominantly native English speaking (86%) and predominantly Caucasian, commensurate with the characteristics of people living in the region (Australian Bureau of Statistics, 1989). Four occasions of measurement were obtained, the first completed in 1991, the second in 1994, the third in 1998, and the fourth in 2002, with an average between-occasion span of 3.5 years. Of the initial sample of 897, 294 completed all cognitive tests at each of the three occasions. Further demographic, diversity and dispersion data on the sample are published elsewhere (Christensen, Mackinnon, Jorm, Henderson, Scott, & Korten, 1994; Korten, Henderson, Christensen, Jorm, Rodgers, Jacomb, & Mackinnon, 1994). Of these, 294 completed all cognitive tests at each of the three occasions.

A series of investigations focus on individual differences in rates and patterns of change, with emphasis on predictors of interindividual differences in intraindividual change and the structure of such changes in late life. Longitudinal results include the finding that education is associated with long-term individual differences in cognitive functioning but has little influence on timing or amount of cognitive change in late life (Christensen, Hofer, Mackinnon, Korten, Jorm, & Henderson, 2003). Changes in memory and speed functioning over 3.5 years were associated with decline in grip strength, more illnesses, and higher depression (Christensen, Mackinnon, Korten, Jorm, Henderson, & Jacomb, 1999). Hofer et al., (2002) evaluated the influence of a genetic risk factor, apolipoprotein (apoE) 4 variant. Previous findings relate apoE 4 to an increased risk for Alzheimer’s disease (AD) and risk for lower performance and greater decline in samples of individuals not diagnosed with AD. Results from latent growth curve analyses of cognitive abilities, based on three occasions over a seven-year period, are reported for individuals genotyped for apoE at the second occasion (n=601) and a subsample of individuals not diagnosed with probable or definite dementia during the first or second measurement occasions (n=434). ApoE 4 status was a significant predictor of both level and change in memory performance, and change in speed performance in the full sample. ApoE 4 status remained predictive of initial level and change in memory performance in the nondemented subsample. Mackinnon et al., (2003) found a moderalte association between activity and cognitive performance.

Cardiovascular Health Study (CHS). “The Cardiovascular Health Study (CHS) is a population-based, longitudinal study of coronary heart disease and stroke in adults aged 65 years and older. The main objective of the study is to identify factors related to the onset and course of coronary heart disease and stroke. CHS is designed to determine the importance of conventional cardiovascular disease (CVD) risk factors in older adults, and to identify new risk factors in this age group, especially those that may be protective and modifiable. The study design called for enrollment of 1250 men and women in each of four communities: Forsyth County, North Carolina; Sacramento County, California; Washington County, Maryland; and Pittsburgh, Pennsylvania. Eligible participants were sampled from Medicare eligibility lists in each area. Extensive physical and laboratory evaluations were performed at baseline to identify the presence and severity of CVD risk factors such as hypertension, hypercholesterolemia and glucose intolerance; subclinical disease such as carotid artery atherosclerosis, left ventricular enlargement, and transient ischemia; and clinically overt CVD. These examinations in CHS permit evaluation of CVD risk factors in older adults, particularly in groups previously under-represented in epidemiologic studies, such as women and the very old. The first of two examination cycles began in June 1989. A second comprehensive examination will be repeated three years later. Periodic interim contacts are scheduled to ascertain and verify the incidence of CVD events, the frequency of recurrent events, and the sequellae of CVD” (Fried et al., 1991).

Einstein Aging Study (EAS), first funded by the NIA in 1993, has since its inception focused on remediable risk factors and biological pathways that lead to cognitive decline, dementing disorders, and the intermediate states that precede them. Since 1993 the EAS has used systematic recruiting methods to assemble a cohort of 1769 elderly individuals from the Bronx, 26% of whom are African American. The EAS sample is broadly representative of the elderly population in one of the poorest and most racially and ethnically diverse urban counties in the United States. Furthermore, the population-based recruitment strategy partially circumvents the selection biases that inevitably arise in referral samples recruited from memory disorder clinics. The EAS has developed, tested, and applied strategies designed to meet the recruitment and retention challenges that may arise when conducting research studies with older adults. In 2004, the EAS began using the Registered Voter Lists (RVL) for Bronx County for recruitment efforts. Individuals were randomly selected and sent a letter followed by a screening telephone call. Persons who completed the telephone screening battery and agreed to participate in clinical follow-up were invited to enroll. The telephone screening battery developed by the EAS has facilitated recruitment and enrollment procedures. Participants consent to annual visits that include a collection of medical, neuropsychological, physical, and epidemiological data and assignment of clinical cognitive outcomes and significant life transitions. Consensus Case Conferences assign clinical cognitive outcomes for each subject at each annual visit. Cognitive outcomes assigned include: 1) Diagnosis of DSM-IV ‘Dementia’ versus ‘No Dementia’, 2) for subjects with dementia, subtypes are diagnosed using standard criteria, 3) Intermediate States of Cognitive Impairment (amnestic Mild Cognitive Impairment, non-amnestic Mild Cognitive Impairment) and 4) documentation of significant life transitions (e.g., nursing home placement, death). The EAS collects and banks biological specimens for current and future assays.

EAS researchers have made many contributions: 1) Developed novel approaches for recruiting systematic, racially diverse, community-based samples using telephone screening and assessment to recruit and retain the sample. 2) Refined neuropsychological measurement of preclinical disease by developing sensitive cognitive tests of memory and other domains 3) Demonstrated the importance of robust normative data and the utility of combining cognitive tests to improve prediction and determine future risk of cognitive decline. 4) Demonstrated that cognitive decline and locomotor performance are intimately linked, and that locomotor impairments predict increased risk of cognitive decline and dementia. 5) Shown that engagement in cognitively-stimulating leisure activities is associated with reduced risk of dementia and mild cognitive impairment syndrome. 6) Extended statistical procedures that include both change point and transition models to improve assessment of cognitive decline in relation to risk factors. 7) Shown that amyloid pathology may be present in the absence of cognitive impairment in a condition called pathological aging, that cytoskeletal pathology is crucial to Alzheimer’s disease, and that hippocampal sclerosis is common and clinically significant in elderly individuals. 8) Collaborated with longevity researchers to show that specific longevity genes protect against cognitive decline and dementia.

The Gerontological and Geriatric Population Studies in Göteborg, Sweden (H-70). The H-70 study is a prospective longitudinal population study, initiated by Dr. A. Svanborg at the University of Göteborg, Sweden. H-70 is a multidisciplinary examination of intraindividual stability and change in biological, physical, psychological, and social functioning. The latter two domains were the responsibility of Dr. Stig Berg. A random sample of subjects, 70 years of age in the county of Göteborg was examined in 1971-72. A second age cohort, also 70 years old, was examined initially five years later in 1976-77. Both age cohorts were re-examined at the ages of 75 and 79. The first age cohort has been investigated up to the age of 88. The H-70 study represents a normative approach to late life and has the advantage of a longitudinal design. About 250 papers have been published from the project and the results generally contrast strikingly with previous stereotyped conceptions of aging and age-related illness, which depicted aging as a more-or-less continuous functional decline which affects cells, organs, and individuals. The results from H-70 have shown that the functional consequences of aging (including specifically measures of intelligence, memory and personality) are more variable and much more open to influence than was previously thought to be the case. Of particular interest is the observation of "terminal drop" in cognitive performance presaging death in the next 4 to 5 year interval (Berg, 1987; Johansson & Berg, 1989). The aims of the H-70 study include description of “normal” aging, and prevalence and incidence of disease, and the evaluation of the potential for preventing functional decline in late life (Rinder, Roupe, Steen, & Svanborg, 1975). The examination of aging-related changes in cognitive functioning related to education, disease, and survival (terminal drop) has been a central focus of psychologically-oriented analyses (Berg, 1987; Maxson, Berg, & McClearn, 1996).

Health and Retirement Study (HRS; 1992- ) and Asset and Health Dynamics Among the Oldest Old (AHEAD). The University of Michigan Health and Retirement Study (HRS) surveys more than 22,000 Americans over the age of 50 every two years. Supported by the National Institute on Aging (NIA U01AG09740), the study paints an emerging portrait of an aging America's physical and mental health, insurance coverage, financial status, family support systems, labor market status, and retirement planning (excerpt from HRS website).

Healthy Older Person Edinburgh Study (HOPE). For the first wave of data collection (carried out from 1990 to 1991), people who were 70 years and older were identified from the registers kept by general medical practitioners in the city of Edinburgh, Scotland. Over 10,000 potential subject’s medical case notes, from 67 general medical practices in the Edinburgh area were scrutinized, first by trained research nurses and then by each patient’s own doctor, to ensure that there was no significant disease and that each person was in good health. 1,467 people agreed to take part in the study and, after, being interviewed at home by a specially trained research nurse for any medical problems or current medications, 603 (237 men, 366 women) were found to have no health problems and to be on no regular medications. The mean age of these subjects was 75.7 years, with a range of 70-88 years.

Interdisciplinary Longitudinal Study of Adult Development (ILSE). This study focuses on predictors and determinants of successful physiological and mental aging (Martin & Martin, 2000; 2002; Rudinger & Minnemann, 1997). The sample includes 1384 participants from two age cohorts (1930-32 and 1950-52 born before and after WWII) and four years between measurement occasions. Reports include changes in cognitive functioning and its predictors (Martin & Zimprich, 2003; Zimprich, Martin & Kliegel, 2003), occurrence of depressive disorders and cognitive impairments (Pantel et al., 2002; Voss et al., in press), adjustment to retirement (Lehr, Juchtern, Schmitt, Sperling, Fischer, Grunendahl, & Minnemann, in press), changes in cognitive ability across participants from former East and West Germany (Oswald, Rupprecht, & Hagen, 1997), and age differences in stress, social resources, and well-being (Martin, Grunendahl, & Martin, 2001).

Long Beach Longitudinal Study (LBLS). This study, which focuses on mechanisms and models of change in cognitive functioning, was initiated in 1978. Recent results include comparison of cross-sectional and longitudinal age trends on memory performance (Zelinski, Gilewski, & Schaie, 1993). Zelinski & Burnight (1997) report age decline in list and text recall with no evidence for differential decline across cohorts over a 16-year period. No reliable decline was observed for recognition memory (Zelinski & Stewart, 1998).

Longitudinal Aging Study Amsterdam (LASA). The central objective of LASA is to “enhance the autonomy and quality of life of older persons". It is guided by questions concerning changes in physical, cognitive, emotional, and social components of aging, predictors of changes with age, association between aging-related changes, and consequences of such changes in terms of quality of life, adjustment, and need for care (Deeg, Beekman, Kriegsman, & Westendorp-de Serière, 1998, Deeg et al 2002). Findings from this study demonstrate a link between depression and cognitive decline specifically in well-educated elders (Geerlings et al 2000). Moreover, memory complaints were shown to be predictive of cognitive decline (Dik et al 2001), but also to be associated with depressive symptoms and personality (Comijs et al 2001, 2002). Dik et al. (2003) showed that early life physical activity affected cognitive performance in older age. Smits, Deeg, Kriegsman, & Schmand (1999) found that measures of cognitive functioning (i.e., processing speed, fluid intelligence) independently predicted mortality after controlling for health, age, depressive symptoms and other covariates. In contrast to other studies, social engagement was not shown to influence cognitive decline (Aartsen et al 2002).

Longitudinal Study of Cognitive Change in Normal, Healthy Old Age (LSCC). This study includes data on approximately 6400 self-selected volunteers aged between 49 and 96 years and approximately 75% female (Rabbitt, 1990; 1993; Rabbitt, Donlan, Bent, McInnes & Abson, 1993). Age, distance and cause of death, and self-reports of health status and recent medical care, and activities of daily living were found to be predictive of cognitive ability (Rabbitt, Bent, & McInnes, 1997). An analysis of four trials of a letter-letter coding test, similar to the WAIS Digit-Symbol Substitution Subtest, given in succession within a single test occasion suggests that memory plays an important role in substitution coding tests (Piccinin & Rabbitt, 1999).

National Survey of Health and Development (NSHD). The UK Medical Research Council (MRC) NSHD study is the oldest of the UK birth cohort studies (1946). The initial study included all 16,500 births that occurred in England, Scotland and Wales during one week of March, 1946. Lifecourse data are currently available on 5362 individuals from birth to age 60 years. The focus of this project is age related change in physical and cognitive function and on the biological, psychological and social processes across the life course that drive these changes. The publications from this study cover wide ranging topics in child and adulthood (Richards, Jarvis, Thompson & Wadsworth, 2003), and the links between them (e.g., Hotopf, Mayou, Wadsworth & Wessely, 1999; Kuh, Hardy, Langenberg, Richards & Wadsworth, 2002; Richards, Hardy, Kuh & Wadsworth, 2001).

Nordic Research on Aging (NORA). The NORA study (Nordic Research on Ageing; Schroll, Steen, Berg, Heikkinen, & Viidik, 1993; Heikkinen, Berg, Schroll, Steen, & Viidik, 1997) is a comparative study on functional capacity and health in 75-year-old men and women. Population-based random samples of 75-year-old residents were obtained in Glostrup, Denmark, and Gothenburg, Sweden, and an invitation to participate was given to all residents of Jyvaskyla in Finland. The participation rates for the targeted samples were high: Jyväskylä, Finland (n=355; 93%), Gothenburg, Sweden (n=368; 83%); and Glostrup, Denmark (n=481; 85%). For further details of the sample see Era, Schroll, Ytting, Gause-Nilsson, Heikkinen, & Steen (1996a). The total sample size available for this study was 1041 and is based on the following sample sizes: Jyväskylä (n=309); Gothenburg (n=322); and Glostrup (n=410). There were 445 males and 596 females in the study.

Normative Aging Study (NAS). This study was established as an intramural research program within the Department of Veterans Affairs and focused on describing the biomedical, psychosocial, and disease-related changes associated with aging (Bosse Silbert & Ekerdt, 1984). The initial sample was composed of 2,280 men with an average age of 72 years (mean age at entry was 42 years), most of whom were veterans from WWII and the Korean War. Clinical health data are collected at 3-5 year intervals and supplemented with periodic mail surveys, interviews, and examinations.

Aldwin et al. (1989) found distinct patterns of change in physical and psychological symptoms across age, with physical symptoms showing an increase over time, while psychological symptoms were relatively stable. Aldwin et al. (2001) clustered physical and psychological health trajectories and found that high, increasing physical health symptoms were associated with higher hostility, anxiety, overweight, and smoking behavior. Trajectories with lower health symptoms were related to being emotionally stable, educated, nonsmoking, and thin. Achat et al. (2000) found that depressive symptomatology was associated with reduced levels of functioning across all SF-36 domains, but that optimism was associated with higher levels of general health perceptions, vitality, and mental health, and lower levels of bodily pain, but not to physical functioning, social functioning, or role limitations due to physical or emotional problems. Mroczek and Spiro (2003) reported distinct trajectories for extraversion and neuroticism and found that birth cohort and age-graded life events such as marriage or remarriage, death of spouse, and memory complaints significantly predicted interindividual differences in change. Brady et al. (2001) reported the associations among risk factors for stroke and specific decline in verbal fluency, a indicator of frontal lobe functioning.

Oregon Brain Aging Study (OBAS). The purpose of this study is to obtain information on the effects of aging on the brain. Healthy community volunteers 55 years or older with normal cognition, blood chemistry and brain MRI were recruited: 258 between 1989 and 2000. Participants are followed every six months for the rest of their life. It is expected that many of the participants’ brains will donate their brains to allow an examination of brain tissue after death.

Origins of Variance in the Oldest-Old: Octogenarian Twins (OCTO-Twin). Potential candidates for participation in this study were 351 twin pairs, including 149 identical or monozygotic (MZ) pairs, and 202 same-sex fraternal or dizygotic (DZ) twin pairs, aged 80 and older and who were assessed in the first wave of the ongoing longitudinal study ”Origins of Variance in the Old‑Old” (OCTO-Twin Study; McClearn et al., 1997). The sample was drawn from the oldest-cohort of the population-based Swedish Twin Registry (Cederlöf & Lorich, 1978) which was comprised of all complete twin pairs, born 1913 and earlier, who were both alive when contacted for potential participation if they were, or became, 80 years of age during the three year period of data collection that started in 1991 (737 pairs in 1474 individuals). Of these pairs, some were excluded because one or both were deceased before they were scheduled for examination (188 pairs), or because one or both declined participation in the study for other reasons (198 pairs). The total number of participants for this study was 702 individuals from 351 complete twin pairs. Other than for reasons of death, the pairwise cooperation rate at the initiation of this study was 65% (corresponding to an individual response rate. The participants were assessed four times at 2-year intervals. The first wave (T1) was initiated from 1991-1993, the second wave (T2) from 1993-1995, the third wave (T3) from 1995-1997, and the fourth wave from 1997-1999. For the present analyses, all available individuals from the twin sample with at least partial data on the cognitive tests were included. The full sample of individuals with partial cognitive test data numbered 648 and included individuals who met the DSM-III-R (American Psychiatric Association, 1987) clinical criteria for dementia at the first occasion. The non-demented subsample (n=508) included only those individuals without diagnosis of dementia across four waves of measurement. Type of illness and cause of death is known and independently verified across three sources of information.

A broad spectrum of bio-behavioral measures of health and functional capacity, personality, well being, interpersonal functioning, as well as memory and cognition were obtained in this study. Findings include evidence for substantial genetic influence on cognitive capabilities late in life (McClearn, Johansson, Berg, Pederson, Ahern, Petrill, & Plomin, 1997) and somewhat less genetic influence on measures of memory (Johansson, Whitfield, Pedersen, Hofer, Ahern, & McClearn, 1999). Results from latent growth curve analysis show that chronological age and time to death are consistent predictors of decline. Intra-pair differences in concordance for dementia across measurement occasions demonstrate that decline trajectories are influenced by non-shared environmental influences associated with survival and compromised health due to dementia in the individual (Johansson, Hofer, et al., 2004).

Seattle Longitudinal Study (SLS). The Seattle Longitudinal Study (SLS) has been a major resource for monitoring age and cohort trends in adult cognitive development, providing normative data for assessment instruments used with older adults, exploring the causes of individual differences in aging, and assessing the effects of targeted cognitive interventions within the context of a longitudinal study (Schaie, 1996). The Seattle Longitudinal Study began in 1956 and has operated continuously with new participant recruitment at each wave of measurement at seven year intervals. Extensions of the study include longitudinal data on second-generation family members and assessment of grandchildren of the original participants. Bosworth and Schaie (1999) reported results that decedents exhibit lower levels of crystallized knowledge, spatial abilities, verbal abilities, and perceptual and psychomotor speed at their last occasion of measurement and showed greater declines in psychomotor speed and verbal meaning across 7 and 14 years. Bosworth, Schaie, Willis, & Siegler reported that time to death accounted for individual differences in the last occasion measured (1% to 3% of the variance) but not of the decline in functioning after controlling for age, education, gender, and survivorship.

Swedish Adoption/Twin Study of Aging (SATSA). SATSA was started in 1984 and is comprised of several longitudinal components. A comprehensive questionnaire was sent in the first component to all twins separated at an early age and reared apart and a control sample of twins reared together from the Swedish Twin Registry. The questionnaire included items concerning rearing, adult, and working environment, health status, health related behaviors (e.g. alcohol, tobacco, and dietary habits) as well as attitude and personality measures. The questionnaire phase is repeated every third year: More than 2,000 twins have responded to questionnaires sent in 1984, 1987, 1990 and 1993. In the second component a subsample of approximately 150 twin pairs reared apart and 150 twin pairs reared together have participated in four waves of in person testing including a health examination, structured interviews/tests of functional capacity, cognitive abilities, and memory. The first three waves were at 3 year intervals; the fourth was initiated in 1999 and fifth in 2002. Data from SATSA are analyzed for the purpose of examining the relative importance of genetic and environmental factors for individual differences in aging related processes. Longitudinal changes as well as the relationships within and among domains (e.g. the importance of genetic effects for mediating the relationship between physical health and cognitive decline) for the elderly are of primary interest (excerpt from SATSA website)

Swiss Interdisciplinary Longitudinal Study on the Oldest-Old (SWILSO-O). SWILSO-O, the first longitudinal Swiss research on the oldest old, examines the life and health trajectories, the social and support networks, the events, and the regulatory processes adopted by the oldest old and by their entourage (Lalive d'Epinay, Pin, & Spini, 2001). The study is multidisciplinary and involves sociology, medico-social problems of elderly (e.g., frailty), economics, psychology and is particularly interested in the elderly’s health (overall rating, habits, accidents, aids, etc.), the factors associated (familial situation, social network, engagement in activities, social participation, ), and its consequences in terms of well-being and quality of life. Two cohorts were assessed, both stratified by age (80-84 years), sex (approximately 50% females), and housing region (50% urban, 50% rural). The first cohort was initially composed of 340 individuals, the second of 376.

UNC Alumni Heart Study (UNCAHS).  The UNC Alumni Heart Study is a retrospective-prospective cohort study that started with MMPI data at college entry in 1964-66.  In 1986 these individuals were located and invited to join the prospective study, in 1992, spouses of participants were invited to join the study.  We have collected 11 waves of data on these three cohorts.  In the college cohort, 95% of the sample was aged 16-21 and 5% was aged up to age 50. The majority of the college cohort was born 1942-1949. Spouses are both younger and older.  The 3 college entry classes differ in gender composition as women did not enter the university as Freshmen until 1996.  Spouses were recruited to help correct this gender imbalance and to test hypotheses about spouse characteristics as part of the environment of the subject. Depending on the particular analysis, women represent 15-30% of the respondents.  The study was designed to test the hypothesis that hostility as measured by the MMPI is a predictor of CHD in current college educated cohorts. Hostility was measured at ages 18, 42, 55 and 60.  Personality was measured with the NEO-PI in 1988 and items reflecting the revised version were added in 1991. Spouses had NEO –PIR measured in 1992.  All Ss had a second NEO PIR in 1997 at approximately age 50.  This is a mail survey.  All data are self-report, but reports of heart disease and death are verified with medical records reviews and death certificates.  Thus, the strength of the data set is the college personality data, with detailed measures from age 40-60 of personality and life events during middle age.

Victoria Longitudinal Study (VLS). The objective of the VLS is to examine profiles and predictors related to late-life changes in cognition, neurocognition, health, biological status, adaptivity, and psychosocial characteristics. Three independent samples of initially healthy, community-dwelling older adults (55-85 years old) are followed at 3-year intervals (Dixon & de Frias, 2004; Hultsch, Hertzog, Dixon,& Small, 1998). Findings from the VLS include the association between intellectually demanding lifestyle activities and maintenance of cognitive functioning and evidence that changes in cognition may lead to reduced intellectual activities (Hultsch, Hertzog, Small, & Dixon, 1999). Gradual declines across six years in the VLS samples (e.g., Small, Dixon, Hultsch, & Hertzog, 1999), with the rate of decline approximately equivalent to that observed in a population-based study of very old adults (Dixon, Wahlin, Maitland, Hultsch, Hertzog, & Backman, 2004).

Wisconsin Longitudinal Study (WLS). The WLS is a long-term study of a random sample of 10,317 men and women who graduated from Wisconsin high schools in 1957 and of their randomly selected brothers and sisters. Survey data were collected from the original respondents or their parents in 1957, 1964, 1975, and 1992, and from a selected sibling in 1977 and 1993. These data provide a full record of social background, youthful aspirations, schooling, military service, family formation, labor market experiences, and social participation of the original respondents. The survey data from earlier years have been supplemented by mental ability tests (of primary respondents and 2000 of their siblings), measures of school performance, and characteristics of communities of residence, schools and colleges, employers, and industries. The WLS records for primary respondents are also linked to those of three same-sex high school friends within the study population. In 1977 the study design was expanded with the collection of parallel interview data for a highly stratified sub-sample of 2000 siblings of the primary respondents. In the 1992-93 round of the WLS the sample was expanded to include a randomly selected sibling of every respondent with at least one brother or sister, and the content was extended to obtain detailed occupational histories and job characteristics; incomes, assets, and inter-household transfers; social and economic characteristics of parents, siblings, and children and descriptions of the respondents’ relationships with them; and extensive information about mental and physical health and well-being. The WLS sample is broadly representative of white, non-Hispanic American men and women who have completed at least a high school education (excerpt from WLS website).