Sleep differences among children with autism spectrum disorders and typically developing peers: a meta-analysis

Abstract

Objective: Sleep problems such as difficulties in sleep initiation, nighttime awakening, and shortened sleep time are often subjectively reported in children with autism spectrum disorder (ASD). However, results of objective studies have been mixed. Our goal was to evaluate the existing data from objective measures using a systematic approach to identify and describe the differences in sleep parameters by comparing total sleep time (TST), sleep latency (SL), and sleep efficiency (SE) in children with ASD with those of typically developing (TD) peers.

Methods: Studies that used objective measures such as actigraphy or polysomnography (PSG) to describe the sleep parameters of TST, SL, and SE in children with ASD compared with children with TD were identified. A meta-analysis was performed for the 10 studies that met inclusion criteria with evaluation of differences in means using random effects models. A total of 343 children with ASD and 221 children with TD were included. Assessments for sources of heterogeneity and publication bias were undertaken.

Results: TST for children with ASD was on average 32.8 minutes less per day (95% confidence interval [CI]: 16.6-49.0 minutes) than their TD peers. Average SL was 10.9 minutes longer (95% CI: 6.7-15.0 minutes), and average SE was 1.9% less (95% CI: 0.7%-3.1%) than their TD peers. Notable heterogeneity was found within studies for TST, and mild heterogeneity was found for SE. Concurrent intellectual disability was a moderator of TST. Children with ASD and normal intelligence had a small and nonsignificant decrease in TST as compared with TD peers, whereas those with ASD and intellectual disability (ID) had a significant decrease in TST as compared with TD peers. The magnitude of the difference in mean SL and SE increased as compared with TD peers as age increased. Studies that used PSG and those that did not include children on medications were more likely to report mean decreases in SE.

Conclusions: Children with ASD have small but measurable objective differences in their sleep parameters that are consistent with subjective reporting. Children with ASD have shorter TST, longer SL periods, and decreased SE as compared with TD peers. Concurrent ID, medication use, method of data collection, and age of subjects significantly moderated these results. The decrease in TST in children with ASD and normal intelligence was not significant as compared with TD peers, suggesting that ID may help explain the shortened TST in children with ASD.