Effects of structured exercise on executive function across childhood, adolescence, and later life: a systematic review and three-level meta-analysis examining exercise skill characteristics

Abstract

Background Structured exercise may benefit executive function across childhood, adolescence, and later life, but whether effects vary by executive-function domain, developmental stage, or qualitative exercise characteristics remains uncertain. Methods We conducted a systematic review and three-level meta-analysis of randomized controlled trials in non-clinical children, adolescents, and older adults. Five databases were searched from inception to May 20, 2026; eligibility was restricted during screening to formally published English-language journal articles dated from January 1, 2015 to May 20, 2026. Outcomes were classified as inhibitory control, working memory, or cognitive flexibility. Intervention arms were coded for environmental predictability, cognitive engagement, coordination complexity, and motor skill learning demand. Three-level random-effects models retained dependent effect sizes within trials, with Study_ID-clustered CR2 robust inference and sensitivity analyses using alternative sampling-covariance structures. The review was registered in PROSPERO (CRD420261402803). Results Forty-seven trials representing 49 independent samples and 4,195 actual participants contributed 202 effect-size estimates. Structured exercise was associated with a positive average effect on executive function (Hedges’ g = 0.412, robust 95% CI [0.284, 0.540], p < 0.001). Heterogeneity was substantial (total I 2 = 74.45%), and the 95% prediction interval ranged from −0.476 to 1.299. Positive pooled estimates were observed for all three executive-function domains and for children, adolescents, and older adults, but neither domain ( p = 0.060) nor developmental stage ( p = 0.825) significantly moderated effects. Environmental predictability, cognitive engagement, coordination complexity, motor skill learning demand, intervention duration, and comparator type were also not significant moderators. Main conclusions were generally robust to alternative sampling-covariance assumptions and study restrictions, although stage-specific effect magnitudes, particularly in adolescents, were less stable in leave-one-trial-out analyses. Small-study-effect diagnostics were inconsistent, publication bias remained uncertain, and GRADE certainty was low because of serious risk of bias and inconsistency. Conclusion Structured exercise was associated with better executive-function performance on average, but substantial heterogeneity and prediction intervals spanning no effect indicate that benefits are not uniform. Current evidence does not identify a specific executive-function domain, developmental stage, program duration, comparator type, or broadly defined exercise skill characteristic that reliably yields larger effects. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/view/CRD420261402803 , identifier PROSPERO (CRD420261402803).