Umbrella systematic review and meta-analysis: physical activity as an effective therapeutic strategy for improving psychosocial outcomes in children and adolescents

Objective: Physical activity (PA) interventions are part of many interdisciplinary programs for the management of children and adolescents with or without physical or psychological conditions or disabilities. Aiming to summarize the available evidence, we conducted an umbrella review of meta-analyses of PA interventions that included psychosocial outcomes in populations of children and adolescents. Method: Literature searches were conducted in PubMed, Cochrane Central, Web of Science, Medline, SportDiscus, and PsychInfo from Jan 1, 2010 to May 6, 2022. Meta-analyses of randomised and quasi-randomised studies investigating the efficacy of PA interventions for psychosocial outcomes in children and adolescents were included. Summary effects were recalculated using common metric and random-effects models. We assessed between-study heterogeneity, predictive intervals, publication bias, small-study effects, and whether the results of the observed positive studies were more than expected due to the chance. On the basis of these calculations, strength of associations was assessed using quantitative umbrella review criteria, and credibility of evidence using the GRADE approach. Quality was assessed using the AMSTAR-II tool. This study is registered with the Open Science Framework, https://osf.io/ap8qu. Results: A total of 112 studies from 18 meta-analyses generating 12 new meta-analyses with 21232 children and adolescents with conditions or disabilities as ADHD, cancer, cerebral palsy, chronic respiratory diseases, depression, neuromotor impairment, obesity and in general populations were included. PA interventions were efficacious in reducing psychological symptoms in all meta-analyses across the different population groups using random-effects models. However, umbrella review criteria suggested a weak strength of association for this outcome, and GRADE credibility of evidence ranged from moderate to very low. For psychological wellbeing, three out of five meta-analyses identified significant effects, but the strength of these associations was weak, and GRADE credibility of evidence ranged from moderate to very low. Similarly, for social outcomes, meta-analyses reported a significant summary effect, but the strength of association was weak, and GRADE credibility of evidence ranged from moderate to very low. For self-esteem, one meta-analysis in children with obesity failed to show any effect. Conclusion: Even though existing meta-analyses suggested a beneficial effect of PA interventions on psychosocial outcomes across different population groups, the strength of associations was weak, and the credibility of evidence was variable, depending on the target population, outcome, and condition or disability. Randomised studies of PA interventions in children and adolescents with and without different physical and psychological conditions or disabilities should always include psychosocial outcomes as an important dimension of social and mental health.