Comparative effects of different exercise modalities on executive function in children with attention-deficit/hyperactivity disorder: a systematic review and network meta-analysis

BackgroundExercise interventions have been increasingly identified as potential non-pharmacological approaches to ameliorate cognitive and behavioural performance among children with Attention-Deficit/Hyperactivity Disorder (ADHD). Nonetheless, the relative efficacy of various exercise modalities is still uncertain.MethodsThis network meta-analysis and systematic review combined 19 trials (N = 845) to compare the impact of structured exercise on executive function and other outcomes. Four databases were searched: Scopus, PubMed, Web of Science, and SportDiscus. Randomised and controlled trials published between 2010 and 2025 were included. The main outcomes included working memory, inhibitory control, and cognitive flexibility; secondary outcomes included ADHD symptoms and motor performance. Random-effects models were used to pool standardised mean differences (SMDs). Only inhibitory control accuracy could be analysed with a Bayesian network meta-analysis (k = 14). I2, prediction intervals, design-by-treatment models, and GRADE were used to evaluate heterogeneity, inconsistency, and certainty of evidence.FindingsEffects were mixed across domains of executive functions. The overall effect of inhibitory control reaction time was not significant (SMD = −0.64, 95% CI: −1.72 to 0.44; I2 = 94%). The accuracy of working memory was also non-significant (SMD = 0.34, 95% CI: −0.26 to 0.94; I2 = 78%), but it was highly variable (prediction interval −1.22 to 1.90). There was little evidence of benefit in cognitive flexibility accuracy (SMD = 0.20, 95% CI: −1.51 to 1.91; I2 = 93%). The network meta-analysis found coordinative training to be the most effective modality for inhibitory control accuracy, performing better than the seven comparators (mean SMD ≈ 1.77). Direct comparisons indicated large, significant effects compared to waitlist (SMD = 2.21, 95% CI: 1.14–3.27) and attention control (SMD = −2.12, 95% CI: −3.51 to −0.72). There was a small, marginally significant improvement in ADHD symptoms (SMD = −0.43, 95% CI: −0.89 to 0.02; I2 = 6%). Moderation analyses revealed no significant dose–response or age effects, but moderate-dose programmes (481–2,880 min) demonstrated the most consistent benefits (SMD = 1.06, 95% CI: 0.32–1.79).ConclusionCoordinative training produces the most consistent gains in inhibitory control among children with ADHD, and gains in more widely studied executive functions are inconclusive due to high variability and sparse networks. Modality-specific recommendations should be reinforced by more powerful head-to-head trials.