Structured physical activity for social communication and repetitive behaviour in children with autism spectrum disorder: a systematic review and exploratory dose-effect meta-regression

Background This systematic review evaluated the effects of structured physical activity on social communication and repetitive behaviour outcomes in children and adolescents with autism spectrum disorder (ASD). It also examined study-level associations between prescribed exercise volume and intervention effects. Methods PubMed, Embase, Web of Science, CENTRAL and EBSCO-PsycINFO were searched from database inception through 5 April 2026. We included controlled intervention studies of structured physical activity in participants aged 18 years or younger with ASD. Continuous outcomes were pooled using Hedges’ g, and prescribed exercise volume was examined using exploratory study-level meta-regression. Results Twenty studies involving 23 intervention arms were included. The studies enrolled 685 participants at baseline, and 664 contributed endpoint data. Social communication outcomes included 14 effect sizes from 12 studies and showed a favourable pooled effect (Hedges’ g = -1.13, 95% CI -1.63 to -0.63; I² = 78.7%). Repetitive behaviour outcomes included 16 effect sizes from 13 studies and showed a favourable pooled effect (Hedges’ g = -0.63, 95% CI -0.84 to -0.42; I² = 0%). Weekly prescribed exercise volume was not clearly associated with either outcome. An exploratory association between total prescribed exercise volume and repetitive behaviour outcomes was observed, but this did not establish a clinical dose threshold. Conclusions Structured physical activity may be associated with improvements in repetitive behaviour outcomes, whereas its effects on social communication remain less certain. The certainty of evidence was very low for social communication and low for repetitive behaviour. No optimal weekly exercise volume or clinically meaningful dose threshold could be identified. Structured physical activity may be a useful adjunct to established behavioural, educational or rehabilitation interventions. Systematic review registration PROSPERO, CRD420261342183.