Effects of open-skill and closed-skill exercise on subdomains of executive function in healthy older adults: a systematic review and meta-analysis

Background Executive-function decline can compromise independence in later life. Exercise may support executive function, but whether open-skill exercise confers greater benefits than closed-skill exercise across domains remains uncertain. Methods PubMed, Embase, the Cochrane Library, Web of Science, and Scopus were searched from inception to 3 May 2026 without language restrictions. We included randomized controlled trials of structured exercise lasting at least 4 weeks in cognitively healthy older adults. Outcomes were classified as inhibitory control, working memory, or cognitive flexibility and separated into reaction-time and performance-based measures. Benefit-oriented Hedges’ g values were synthesized using dependency-aware three-level random-effects models with trial-level CR2 cluster-robust inference. Exercise category was examined as an exploratory moderator. Risk of bias was assessed with RoB 2 and certainty with GRADE. Results Twenty-seven independent trials contributed 1,598 distinct analysed participants across eligible intervention and comparator arms. Exercise significantly improved inhibitory-control reaction time ( g = 0.346, 95% CI 0.025 to 0.667), inhibitory-control performance ( g = 0.466, 95% CI 0.089 to 0.844), working-memory performance ( g = 0.251, 95% CI 0.046 to 0.455), cognitive-flexibility reaction time ( g = 0.411, 95% CI 0.074 to 0.748), and cognitive-flexibility performance ( g = 0.383, 95% CI 0.025 to 0.741). Working-memory reaction time favoured exercise but remained inconclusive ( g = 0.410, 95% CI − 0.013 to 0.833). Exercise category did not significantly moderate any outcome. Sensitivity analyses preserved positive pooled-effect directions, although statistical significance was not uniformly robust. Certainty was moderate for working-memory performance and low for the other outcomes. Conclusions Structured multi-week exercise may improve several components of executive function in cognitively healthy older adults. Current evidence does not establish that open-skill exercise is superior to closed-skill exercise, and non-significant moderator findings should not be interpreted as equivalence. Better-powered head-to-head trials with transparent classification, standardized outcomes, and longer follow-up are needed. Trial registration PROSPERO CRD420261393695.