Peer-reviewed article

Dose-response relationship between exercise interventions and motor skills in children with autism spectrum disorder: a systematic review and multilevel meta-analysis.

Wang T, Dong X, Jia Y, Tian J, Hou X · 18 Sept 2026

Abstract

Autism spectrum disorder (ASD) is a significant public health concern, with most ASD children suffering from motor skill impairments. This study aimed to explore the dose-response relationship between exercise interventions and motor skill outcomes in children with ASD, thereby identifying the exercise dose associated with the largest intervention effects. Four databases (PubMed, Web of Science, Embase, and the Cochrane Library) were systematically searched from database inception to April 2026. Experimental studies evaluating the effects of exercise interventions on motor skills in children with ASD were included. The outcomes included overall motor skills, gross motor skills, and fine motor skills. Effect sizes were calculated as Hedges' g, and multilevel random-effect models were fitted using the metafor package. The multilevel meta-analysis and dose-response analysis were used to assess the dose-response relationship between exercise and motor skills. Moderation analyses were conducted to explore significant influencing covariates. A total of 34 studies from 12 countries were included, involving 1,082 children with ASD within the eligible age range of 3-12 years. Results of three-level meta-analysis showed that exercise interventions significantly improved motor skills in children with ASD (Hedges' g = 0.95, 95% CI [0.67, 1.23]), including gross motor skills (Hedges' g = 0.99, 95% CI [0.67, 1.30]) and fine motor skills (Hedges' g = 0.40, 95% CI [0.14, 0.66]). Dose-response analyses indicated an exploratory dose-response pattern, with the estimated effects reaching their highest values at approximately 46 sessions of 45 min each for overall motor skills and 44 sessions of approximately 47 min each for gross motor skills. Moderator analyses indicated that the time point of the test significantly influenced the effect of exercise on overall motor skills, with the immediate test showing a greater effect size than the delayed test (P = 0.044). No significant moderators were identified for gross motor skills or fine motor skills. Conclusion Exercise significantly improved overall, gross, and fine motor skills in children with ASD. Dose-response relationships were identified for overall and gross motor skills. What is known • Motor impairments are prevalent in children with ASD and represent an important barrier to functional independence and participation. What is new • This is the first multilevel and dose-response meta-analysis evaluating exercise effects on overall, gross, and fine motor skills in children with ASD. • Dose-response analyses identified exploratory dose-response patterns for overall and gross motor skills, with estimated peak effects at approximately 46 sessions (45 min/session) and 44 sessions (47 min/session), respectively. • Assessment time point moderated effects on overall motor skills, with larger effects observed immediately after intervention than at delayed assessment.