Effects of aquatic exercise on symptoms, behaviors, and motor function in children with autism spectrum disorder: a systematic review and meta-analysis

Abstract

Background Aquatic exercise combines structured physical activity, sensorimotor training, and social participation, and may be an acceptable adjunctive intervention for children with autism spectrum disorder (ASD). However, existing trials are limited by small samples, heterogeneous outcomes, and complex data structures. Objective To evaluate the effects of aquatic exercise on ASD-related symptoms or behaviors, motor or physical function, autism severity, stereotyped behaviors, social function, and other outcomes in children with ASD. Methods Controlled studies comparing structured aquatic exercise with passive or active controls were included. Hedges’ g was calculated using post-intervention values, with positive values favoring aquatic exercise. Random-effects models were fitted using restricted maximum likelihood with Knapp–Hartung confidence intervals. For outcomes with only 2 studies, fixed-effect sensitivity analyses were also reported. Change-score analyses and sensitivity analyses using different pre–post correlation assumptions were performed. Multiple motor outcomes from the same study were combined within study. Results Six studies involving 159 children were included for overall ASD-related symptoms or behaviors. Aquatic exercise showed a favorable effect compared with control conditions (g = 1.31, 95% CI 0.63–1.99, p = 0.004; I2 = 45.5%), and the result remained stable in leave-one-out and change-score sensitivity analyses. Four studies involving 119 children were included for broad motor or physical function. The pooled effect favored aquatic exercise but did not reach statistical significance and showed very high heterogeneity (g = 2.35, 95% CI −0.11 to 4.80, p = 0.056; I2 = 90.2%). After excluding the study with estimated data, the effect decreased to g = 1.60 (95% CI −0.71 to 3.91). CARS/CARS-2, stereotyped behaviors, and social function each included only 2 studies, with unstable model-based inferences. Subgroup analyses did not identify significant differences by control type, intervention duration, or training frequency. Conclusion Aquatic exercise may improve overall ASD-related symptoms or behaviors in children with ASD. Evidence for motor function and other specific outcomes remains limited, heterogeneous, and exploratory. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/, identifier CRD420261436475.