Sleep Dysregulation and Circadian Electrodermal Activity in Children with Autism Spectrum Disorder

Background: Up to 80% of children with an autism spectrum disorder (ASD) experience sleep dysregulation. A breadth of research documents the behavioral correlates of sleep dysregulation—including increased ASD symptoms, aggression, and other challenging behaviors (e.g., Goldman et al., 2011). These associations are documented in both home and treatment settings (e.g., Abel et al., 2018). However, the physiological correlates of sleep dysregulation are poorly understood and likely to contribute to daytime behavior problems and potentially treatment progress.

Objectives: The goal of the present study is to understand the physiological correlates of sleep dysregulation in ASD by assessing child sleep and electrodermal activity during daytime center-based treatment.

Methods: Thirteen children (mean age of 6.19 years, SD = 2.24) with a medical diagnosis of ASD participated in this study for one treatment week. Children received treatment in an applied behavior analysis center for up to 8 hours per day from roughly 09:00 to 17:00. Each child wore an Empatica E4 multi-sensor during treatment, resulting in over 10,000 minutes of electrodermal activity (EDA) data. Skin conductance responses (SCRs) were extracted from the EDA data following guidelines from Kleckner et al. (2017). Actigraphy was used to measure sleep for five consecutive 24-hour periods (which corresponded with the EDA data collection). Each child’s sleep was classified as either regulated (n = 7) or dysregulated (n = 8) based on the following criteria: (1) slept less than the American Academy of Sleep Medicine recommendations for their age, (2) woke for more than an hour on a least two nights, and/or (3) had morning rise or bedtimes that varied by more than two hours. Child ASD symptoms was indexed with the Social Communication Questionnaire (SCQ).

Results: A series of regression analyses were conducted across the regulated and dysregulated sleep groups, predicting average SCRs per minute across each treatment hour. Children in the dysregulated sleep group had fewer SCRs as the treatment day progressed (Figure 1). When the covariates of child ASD symptoms, age, and maternal education were considered, these associations were attenuated and no longer significant. Children in the dysregulated sleep group were older and tended to have more ASD symptoms when compared to the regulated sleep group (Table 1). Due to small sample size, we were unable to fully detect how these covariates influence the relations between sleep and SCRs. However, we were able to document distinct circadian SCR patterns for a subgroup of children with ASD.

Conclusions: SCRs often reflect sympathetic arousal to environmental stimuli. Fewer SCRs may indicate less engagement with the environment (as indexed by lower arousal). This preliminary study requires replication before treatment or clinical recommendations can be drawn. However, our results clearly suggest that SCRs follow a circadian or daily cycle. The roles of child age, ASD symptoms, and sleep warrant further investigation to understand their relations with arousal and ultimately daytime behaviors during treatment.