Effects of different types of physical exercise interventions on aggressive behavior among children and adolescents: a network meta-analysis.

Introduction: Aggressive behavior is a common and important externalizing problem during childhood and adolescence, closely associated with poor social adjustment, subpar academic performance, and adverse mental health outcomes. Although physical exercise has been regarded as a viable nonpharmacological intervention to reduce aggressive behavior, the relative effects of different types of exercise interventions remain unclear. This study used a systematic review and network meta-analysis to compare how different types of physical exercise interventions affect aggression-related outcomes in children and adolescents. Methods: A systematic literature search across PubMed, Web of Science, Embase, and the China National Knowledge Infrastructure from database inception to December 10, 2025 identified randomized controlled trials (RCTs) and non-randomized controlled trials (non-RCTs) evaluating the effects of physical exercise interventions on aggression-related outcomes in children and adolescents. Based on motor skill characteristics and intervention composition, the interventions were classified into open-skill exercise (OSE), closed-skill exercise (CSE), and multicomponent exercise (MCE). Outcome measures comprised overall aggression, anger, hostility, physical aggression, and verbal aggression. Random effects models were used to conduct conventional pairwise meta-analyses and network meta-analyses, expressing effect sizes as standardized mean differences (SMDs) with 95% confidence intervals (CIs). Surface under the cumulative ranking curve (SUCRA) values determined intervention rankings, and consistency tests, meta-regression, and sensitivity analyses evaluated finding robustness. Results: This analysis included 25 studies involving 2,772 children and adolescents. Compared with control conditions, physical exercise interventions significantly mitigated negative aggression-related outcomes, including hostility (SMD = -0.37, 95% CI: -0.62 to -0.13), anger (SMD = -0.64, 95% CI: -1.00 to -0.27), overall aggression (SMD = -0.87, 95% CI: -1.36 to -0.39), and physical aggression (SMD = -0.70, 95% CI: -0.91 to -0.49). Conversely, the mitigation of verbal aggression lacked statistical significance (SMD = -0.05, 95% CI: -0.36 to 0.25). For outcomes permitting treatment network construction, the global consistency test indicated no significant inconsistency. SUCRA ranking results revealed that MCE ranked highest for hostility (79.4%) and physical aggression (74.7%), CSE ranked highest for anger (85.9%), and OSE ranked highest for overall aggression (84.4%). Meta-regression identified no significant moderators. Conclusion: Physical exercise interventions may help improve certain aggression-related outcomes in children and adolescents, although the effects are not entirely consistent across different outcomes. These findings suggest that exercise modalities possessing different skill characteristics may differ in their effects on hostility, anger, overall aggression, and physical aggression. Consequently, practitioners designing interventions for aggressive behavior in children and adolescents may consider the target outcome when selecting exercise programs. Given the substantial between-study heterogeneity and scarce head-to-head comparative evidence, additional high-quality RCTs are needed to validate these findings further. Systematic review registration: https://www.crd.york.ac.uk/PROSPERO/view/CRD42026133695, identifier (CRD420261336959).