Longitudinal Study of Sleep and Co-Occurring Problems in Autism Spectrum Disorder

Background: Autism spectrum disorder (ASD) is characterized by high rates of medical and behavioral comorbidities. Sleep problems are particularly detrimental, contributing to significant distress and impairment over and above the effects of primary ASD symptoms. Prior studies have found that sleep problems are associated with anxiety, stereotypic and challenging behavior, and sensory reactivity among children with ASD. However, the relations among these symptoms are not well understood. Studies to date have been primarily cross-sectional, meaning that directionality and causality cannot be determined. A better understanding of the longitudinal course and correlates of sleep problems in children with ASD would provide important insights into potential mechanisms and targets for treatment.

Objectives: The primary objective of this study was to describe the chronicity of sleep disturbance in children with ASD over time. The secondary objective was to understand the relations among sleep disturbance, internalizing symptoms, behavior problems, and sensory problems over time.

Methods: The data for this study included 514 children and adolescents with ASD (81% male, age M = 5.9, SD = 3.2) who were enrolled in the Autism Treatment Network Registry. Of the 2400 children assessed in this registry, 514 had baseline and longitudinal follow-up data on the variables of interest (M = 3.8 years later, SD = .62). Primary measures included the Children’s Sleep Habit Questionnaire, Child Behavior Checklist, Aberrant Behavior Checklist, and Short Sensory Profile. Structural equation models were used to examine the interrelationships among sleep problems and co-occurring symptoms over time, while controlling for IQ, age, gender, sleep medications, parent education, household income, and time between visits.

Results: Sleep problems at baseline were predictive of future sleep problems, even after controlling for covariates. Over time, 36% of the sample showed intensification of sleep problems and 44.4% showed improvement in sleep problems. Children were more likely to improve if their initial sleep problems were less severe. Neither age [F(2, 343) = .088, p =.915], IQ [F(2, 343) = 1.511, p = .222], gender [χ2 (2) = 1.285, p = .526], nor time between assessments [F(2, 343) = .977, p = .377] were related to the progression of sleep problems. In addition, parent education, household income, and use of sleep medications were unrelated to change in sleep problems (ps > .05). However, individuals were more likely to improve if their initial sleep problems were minor (r = -.490, p < .001). Initial anxiety predicted future sleep problems, and initial sleep problems predicted future attention problems and hyperactivity. Sleep and somatic complaints were interrelated over time; whereas, several other internalizing and externalizing symptoms were independent of sleep problems.

Conclusions: These longitudinal analyses provide insights into the nature, course, and development of sleep and other co-occurring problems in children with ASD, offering new directions for targeted interventions. Although further research is needed, the results suggest that addressing anxiety may improve sleep problems for some children with ASD, and that improving sleep may reduce attention problems and hyperactivity.