Effects of high-intensity interval training on cognitive function in children and adolescents: a three-level meta-analysis

Cognitive impairments are prevalent among children and adolescents. Although high-intensity interval training (HIIT) is a time-efficient exercise intervention with substantial potential to improve cognitive function, current empirical findings remain equivocal. This study uses a three-level meta-analysis to comprehensively synthesize existing evidence. We aim to evaluate the overall impact of HIIT on cognitive function in this population and systematically explore potential moderators. We systematically searched the Web of Science, PubMed, Embase, and Cochrane databases for eligible randomized controlled trials and quasi-experimental studies. Effect sizes were calculated using Hedges’ g with a correction for small-sample bias. Data were analyzed by constructing a three-level hierarchical linear model using the metafor package in R. Seventeen studies contributed 59 eligible effect sizes. The primary three-level random-effects meta-analysis including all eligible effect sizes yielded a positive average effect of HIIT on cognitive task performance (Hedges’ g = 0.45, 95% CI [0.23, 0.66], p < 0.001). However, the 95% prediction interval was wide and crossed zero ([-0.40, 1.29]), and substantial heterogeneity was observed (Q(58) = 418.73, p < 0.001), with 98.54% of the total variance attributable to between-study heterogeneity. A sensitivity analysis excluding two statistically influential effect sizes included 57 effect sizes from 16 studies and yielded a smaller pooled estimate (g = 0.35, 95% CI [0.20, 0.50], 95% PI [-0.17, 0.87], p < 0.001). None of the examined moderators showed a statistically significant between-subgroup effect. Domain-specific estimates varied descriptively, but the between-domain omnibus test was not statistically significant (F(6, 52) = 2.20, p = 0.058); therefore, these differences were exploratory and did not establish differential effects across cognitive domains. HIIT was associated with a positive average effect on pooled cognitive task performance in children and adolescents. However, the pooled estimate was partly sensitive to two influential effect sizes, the prediction interval crossed zero, between-study heterogeneity was substantial, and none of the examined moderators showed a statistically significant between-subgroup effect. The findings should therefore be interpreted cautiously and do not support definitive recommendations regarding specific populations, cognitive domains, or HIIT prescriptions. The protocol for this systematic review is registered with the International Prospective Register of Systematic Reviews (registration number: CRD420261346288; registration date: March 20, 2026).