The Association Between Activity Engagement and Cognitive Performance Amongst Healthy Older Adults: A Meta-Analysis

This project is a systematic review and meta-analysis examining the association between activity engagement and cognitive performance in healthy older adults. Activity engagement is conceptualized as a diverse behavioral repertoire encompassing participation across cognitive (e.g., reading, playing games, learning languages), physical (e.g., walking, sports), and social (e.g., volunteering, club participation) activities. Drawing on the cognitive reserve construct (Scarmeas & Stern, 2003; Stern, 2002), which posits that individual differences in resilience to brain aging arise from lifelong engagement in stimulating activities, this review estimates the overall magnitude of the engagement–cognition association and examines whether it varies as a function of sample and assessment characteristics. Despite generally positive evidence linking activity engagement to cognitive outcomes, findings in the literature are mixed, with inconsistencies across activity types, cognitive domains, and study populations. To our knowledge, no prior meta-analysis has synthesized this association across cognitive, physical, and social activities as a composite construct in healthy older adults. This review addresses that gap. Three databases (PsycInfo, Web of Science, and Science Direct) were systematically searched following PRISMA guidelines, supplemented by a manual backward search of reference lists. Eligible studies reported zero-order Pearson correlations between self-reported activity engagement and at least one cognitive performance measure in samples averaging 65 years or older. A correlated and hierarchical effects (CHE) model with robust variance estimation was used to account for the non-independence of multiple effect sizes within studies. Meta-regressions examined six moderators: age, education, activity type, timeframe of retrospective assessment, performance index, and cognitive domain. The final sample comprised 52 studies, 60 independent samples, 52,621 participants, and 682 effect sizes. We expected to find a small but significant positive overall association, with moderation by activity type, age, and education. This project contributes to the optimal aging literature by providing a quantitative synthesis of the engagement–cognition association and identifying the conditions under which it is most robust, with implications for recommendations regarding modifiable lifestyle behaviors in later life. All data, analysis code, and supplementary materials are publicly archived on the Open Science Framework.