Schedule and Presentations 2013
This 2013 theme will be “Aging Brains in Aging Bodies.” We will use data from the Adult Changes in Thought (ACT) study. ACT is a prospective longitudinal cohort study of older adults from Group Health Cooperative, a Seattle-area Health Maintenance Organization. ACT data include data collected by the ACT study, data generated by clinical care, genetic data, and neuropathology data from autopsies on 470 people, over 25% of the decedents. Since 1994, the study has enrolled some 4,700 participants. Enrollment has been in three phases. An initial cohort of 2,581 was enrolled in 1994-1996. An expansion cohort of 811 was enrolled in 2000-2002. Continuous enrollment began in 2005, with a target to keep 2,000 people actively enrolled in the study and at-risk to develop dementia.
The overall training goal for the 2013 workshop is to consider ways of integrating information on health to improve our understanding of late-life brain outcomes. Because of its setting within a healthcare delivery system, ACT includes data sources that are either completely unavailable or difficult to obtain from other studies. ACT study data include extensive epidemiological data with large overlap with other studies. Group Health data available for the entire cohort include pharmacy data for every prescription medication since 1977, and all clinical laboratory data since 1988. For example, diabetes is an important medical condition, and there is great interest in potential brain effects of diabetes treatments. Of the 4,700 ACT participants, there were 9 who were treated for diabetes in 1977. Every other ACT participant’s complete lifetime diabetes treatment regimen is completely captured with pharmacy data, and all treatments since 1977 are also captured for the 9 people who were already treated for diabetes in 1977.
ACT has also begun an ambitious medical records abstraction project to collect research-quality data from clinical records. The study is up to date with the autopsy cohort of 470 people. The mean (SD) years of data available is 33 (11) years for this cohort. Data such as weights, blood pressures, ejection fractions, laboratory tests preceding 1988, medications preceding 1977, a variety of conditions, and so on – over 100 data elements are captured in the medical records abstraction project. These data facilitate unprecedented opportunities for clinical-pathological correlation studies.
Traditionally, the ACT study has focused on three types of late-life brain outcomes: time to dementia and Alzheimer’s disease; global cognition as measured by the Cognitive Abilities Screening Instrument (CASI) scored using item response theory (IRT); and neuropathological findings at autopsy. The analytic team has experience with time-varying covariates, integrating clinical data captured at sporadic intervals into risk models, sensitivity analyses considering a variety of concerns, mixed effects models to account for missing data, and use of marginal structural models to address selection in the autopsy cohort.
The first day of the conference will include a scientific summary of the ACT study and talks about the broader literature on how health affects cognition and the brain in late life.