The timing and magnitude of age-related cognitive decline, some of the oldest questions in cognitive aging research, remain open questions of major importance. Although many studies have been conducted, we often cannot compare results because they use different cognitive tests and are not directly comparable. The goal of this proposal is to conduct an integrative data analysis of cognitive aging, meaning drawing together data collected in multiple studies, made comparable by harmonizing traits for general cognitive function, memory, and executive function. We will use several large epidemiologic studies with longitudinal data that administered different neuropsychological test batteries. Cognitive performance will be scaled across study to a nationally representative sample of older adults. To approach this task, we have neuropsychological performance data from 10 epidemiologic and clinical studies of older adults, for a combined sample size of 39,832 adults between the ages of 55 and 110 who have been followed up to 20 years. The studies are a mix of representative, volunteer, and specially selected samples. Our goals in this proposal are to address the timing and magnitude of age-related cognitive decline, and to obtain support for continued development of the composites and eventual preparation of an R01 to address other major questions in cognitive aging. Before we can address such substantive research questions, substantial work is needed to evaluate bias and measurement properties of the harmonized composites. Our study team possesses necessary experience and unique capabilities to complete this work. We have already published the basics of our approach.
We are proposing to determine the trajectory of cognitive decline using the largest sample of older adults ever assembled for that purpose of which we are aware. Understanding the magnitude of and contributors to cognitive decline may help identify and target interventions to delay or prevent decline into dementia. Beyond the current scope of this research, some of the biggest questions we have for neuropsychological data, such as how to measure cognition in pharmacological interventions that will improve patient outcomes in dementia research, require harmonization efforts combining data in meaningful ways across studies.
