LAUSR

Up to 80% of autistic children experience sleep disturbances. Sleep hygiene guidelines advise patients to maintain consistent bedtimes and waketimes for the purpose of regulating one’s circadian rhythm. However, the relationship between bedtime and waketime variability and sleep outcomes in autistic children is unclear, and so the current study examines these associations. One hundred thirty autistic children (Mage = 8.81 yrs, SD = 1.92 yrs, range = 6-12 yrs, 71% male) completed two weeks of daily sleep diaries with parental assistance and concurrently wore actigraphic wrist watches (Actiwatch-2, Respironics). Rest intervals in Actiware software were set using visual inspection while referencing self-reported bed/waketime. Boundaries of rest intervals were defined as the time when the child first attempted sleep to the time when the child vacated bed. Subjective and objective estimates of bedtime, waketime, total wake time during attempted sleep periods, and total sleep time were obtained. Bayesian multi-level models obtained individual estimates of average variability in bedtime and waketime. Multiple linear regressions were then used to examine associations between subjective and objective bedtime and waketime variability and total wake time and total sleep time, controlling for age and average levels of bed/waketime variability. Models were separated out by subjective/objective sleep outcomes (i.e., subjective waketime variability predicting subjective total sleep time). Greater objective waketime variability was associated with less objective total sleep time (B=-38.80, p<.001, partial eta-squared = .071). Greater subjective waketime variability was associated with longer subjective total wake time (B = 18.58, p=.02, partial eta-squared = .058) and shorter subjective total sleep time (B=-70.60, p<.001, partial eta-squared = 0.227). While results do not possess causal implications, greater subjective and objective waketime variability are associated with poorer sleep outcomes in autistic children. Effect sizes ranged from approximately medium to large. Results suggest that bedtime variability may not be related to sleep outcomes in autistic children. Future studies should examine whether reductions in waketime variability mediate the effect of behavioral insomnia interventions on improved sleep in autistic children. MU Research Board Grant (McCrae, PI); Department of Defense Autism Research Program (McCrae, PI; W81XWH2010399); Thompson Center for Autism and Neurodevelopmental Disorders.