Effect of Daily Physical Activity on Sleep Characteristics in Children with Autism Spectrum Disorder

Background. Sleep problems have been reported in children with autism spectrum disorder (ASD). One of the methods proposed to improve sleep characteristics is based on physical activity (PA). Objective. To assess characteristics of sleep and the effect of PA level on sleep quality in children with ASD compared to controls. Methods. Fifty boys with ASD (ASD; 10.8 +/- 2.6 years) and 18 controls (CONT, 10.1 +/- 2.2 years) wore an accelerometer device for five consecutive days to obtain measures of activity and sleep characteristics. Results. Some significant differences were reported between ASD and CONT groups. Wake-up time resistance was significantly higher (p < 0.05) in ASD. Total time for PA, and daily steps number were significantly lower in the ASD group (p < 0.05). Time for sedentary behavior was significantly higher (p < 0.01) in the ASD group. Using a principal component analysis and an agglomerative hierarchical analysis, we observed three clusters. Two showed the same poor-quality indices of sleep but two opposing profiles of PA, either an insufficient PA volume (cluster 1, Total time PA = 75.1 min; Daily steps: 7704) or an important PA volume (cluster 3, Total time PA = 222.1 min; Daily steps: 17,164). Cluster 2 was characterized by moderate participation in PA and children with the best sleep indices (Total time PA = 166.8 min; Daily steps: 12,718). Conclusion. The dose-response effect of exercise on sleep may indicate large individual differences but the present findings are important to prevent sedentary behaviors or to correct over-activity that could be detrimental to the sleep quality in children with ASD.

Automated summary

This study compared sleep characteristics and the impact of PA level on sleep quality in children with ASD and controls. Significant differences were found between the two groups, with ASD children showing higher wake-up time resistance, lower PA time, and higher sedentary behavior time. Three clusters were identified, with cluster 2 showing moderate PA participation and the best sleep indices.