The Evidence on Teaching

Systematic review and meta-analysis investigating moderators of long-term effects of exercise on cognition in healthy individuals

Ludyga S, Gerber M, Pühse U, Looser VN, Kamijo K · 2020

grade Cmeta-analysisindependentfailed
Population
Healthy individuals across lifespan; 80 RCTs
Design
Meta-regression of 80 RCTs of chronic exercise interventions in healthy participants spanning children to older adults; moderators tested included exercise type (coordinative vs aerobic/resistance), session duration, intervention length, age and sex.
Key findings
The main pro-exercise meta-analysis and the source of the popular 'coordinative/motor-rich exercise is best for the brain' claim. Two deflations apply: the summary effect is small and domain-general rather than a targeted cognitive gain; and Ciria 2023's umbrella review shows this class of meta-analysis rests on underpowered, publication-biased RCTs, with the pooled effect collapsing to ~0.05 once corrected. Include as the steelman, grade C, treat the coordinative-exercise moderator as hypothesis-generating.
Genetic confound
RCTs, so randomization handles confounding; but no analysis of heritable trainability or responder variance despite HERITAGE-style evidence that response to training is itself heritable.
Replication notes
Headline conclusion directly challenged by Ciria et al. 2023 (bias-corrected null) and by the null Fit to Study RCT.

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
Global cognition (post-intervention)Summary standardized effectSmall, and did not differ across cognitive domains (exact g unverified from accessible abstract; magnitude comparable to Ciria's uncorrected d~0.2)standardizedtransfer (far); post-intervention; lab cognitive testsunclearend-of-treatmentfar-transfer
Coordinative/motor-demanding exercise vs other exercise typesModerator contrastLarger benefit for coordinative exercisestandardizedtransfer; post-intervention; labactive-alternativeend-of-treatmentfar-transfer
Sex moderationModerator contrastSmaller effects in females; dose-response differed by sexstandardizedtransfer; post-intervention; labunclearend-of-treatmentfar-transfer

Cited by