Sleep Traits and Autism Spectrum Disorder: A Two-sample Mendelian Randomization Study

Abstract

Purpose: The correlation between autism spectrum disorder (ASD) and sleep has been found in previous observational research. Whether these relationships are causal or not is still unclear. This study employing two-sample Mendelian randomization (MR) attempted to research the causal relationship between ASD and sleep. Methods: GWAS results were obtained for traits of sleep from the UK Biobank, consisting of Insomnia, Daytime sleepiness, and Sleep duration. The sleep duration data was split into two datasets: short sleep duration and long sleep duration.The GWAS result for ASD from Psychiatric Genomics Consortium (PGC). For the primary method of this MR analysis, inverse-variance weighting (IVW) was used. Results: Long sleep duration (sleep duration ≥ 9 hours) was nominally connected with an enhanced likelihood of autistic spectrum disorder (OR = 24.018; 95%CI 1.830-315.096; P = 0.016), no reverse causality existing (OR = 0.996; 95%CI 0.971-1.021; P = 0.762). Our results showed no causal relationship between daytime sleepiness, insomnia, sleep duration, sleep duration < 7 hours, and autism spectrum disorders. Conclusion: This study suggested evidence for an adverse effect of long sleep duration on ASD risk. Improving sleep problems could alleviate the occurrence of autism spectrum disorders.