Background: Population-based cohort studies provide opportunities to study chronic disease and aging; but older cohorts inevitably shrink. We describe an enrollment strategy that preserves a nondemented base population as well as selected studies. Methods: We originally established a dementia registry (1987) from 23,000 persons over age 65 enrolled in Group Health, a stable, longstanding HMO. We then enrolled the ACT cohort: 2581 persons, in 1994-1996; a replacement cohort of 811 in 2000-2002; and we started continuously enrolling in 2004, creating a constant base population of about 2000, all randomly selected and enrolled when known not to be demented (total N = 3915, mean age 75.1 +/- 6.4, range 65-101). Results: Incidence rates of dementia and Alzheimer's disease (AD) in the ACT cohort resemble those in other cohort studies. The population has enabled: randomized trials of exercise for frailty and mild cognitive impairment (MCI); ongoing trials including interventions to improve sleep and cognitive and physical well-being; studies of the predictive value of sensory and neuropsychological testing for early AD; and traditional epidemiologic risk factor studies establishing exercise, but not antioxidant vitamin use, is protective. Computerized pharmacy data (since 1977), lab data (since 1981), and over 120,000 person-years of medical record data create unique opportunities to study pharmacoepidemiologic and cardiovascular risk factors, which appear increasingly important, and possibly modifiable, for late-life dementias. ACT is now the source for: studies of the epidemiology of MCI and insulin resistance; population-based neuropathologic studies (n = 1105 autopsy consents, n = 283 completed autopsies); and genome-wide analyses (GWAS). With a median 30 years of medical record data before enrollment, we can also relate mid-life treatments and cardiac risk factors like hypertension to late-life brain function, neuroimaging, and pathologic changes. Conclusions: Seattle's ACT study establishes proof of concept of a “living laboratory” for studies of aging and dementia in older adults and has demonstrated protective effects of exercise, importance of vascular risk factors, and population-based studies of MCI and late-life neuropathologic phenomena. Achieving the ACT study's full potential will require support of infrastructure needs beyond the study's data and unique population base and setting.
FULL TEXT