Infant sleep characteristics in children with autism spectrum disorder: a population-derived Australian birth cohort study.

Abstract

To examine the prospective associations between infant sleep patterns and subsequent autism characteristics and diagnosis in a population-derived sample. Population-derived birth cohort study in Victoria, Australia's Barwon region, with 1074 mother-infant pairs recruited from June 2010 to 2013. Infant sleep characteristics via the Brief Infant Sleep Questionnaire at 6 months (n=925) and 12 months (n=885). Parent-reported autism characteristics measured using the Child Behaviour Checklist for Ages 1½-5 (CBCL/1½-5; n=676) at 2 years and Strengths and Difficulties Questionnaire for report for ages 4-10 (SDQ/P4-10; n=791) at 4 years. Autism Diagnostic and Statistical Manual for Mental Disorders fifth edition (DSM-5) diagnoses (n=64) were confirmed by 11.5 years. At 6 months, each 10% increase (~1 hour) in night sleep duration was associated with fewer autism characteristics at 2 years (4.5% decrease, CBCL/1½-5) and 4 years (4.5% decrease, SDQ/P4-10) and 22% lower autism DSM-5 diagnosis odds by 11.5 years (adjusted mean difference (AMD): -0.02, 95% CI: -0.04 to -0.01; AMD: -0.02, 95% CI: -0.03 to -0.007; adjusted OR (AOR): 0.78, 95% CI: 0.65 to 0.94). At 12 months, each 25% increase in sleep latency (~5 min) was associated with more autism characteristics (1.5% increase, CBCL/1½-5, AMD: 0.006, 95% CI: 0.002 to 0.01) and 7.7% higher autism diagnosis odds (AOR: 1.08, 95% CI: 1.03 to 1.13). Among diagnosed school-aged children, 42% used melatonin in the past month. Poor infant sleep quality was linked to increased autism characteristics and diagnosis odds in a population-derived Australian sample. The extent to which this reflects common determinants of poor sleep and autism is not known. These findings suggest early monitoring of sleep issues as a potential autism indicator.