Sleep in adults with Autism Spectrum Disorder: a systematic review and meta-analysis of subjective and objective studies.

Background/objective: Sleep problems are commonly reported by individuals with Autism Spectrum Disorder (ASD). However, to date, no quantitative evidence synthesis of available studies has been performed to quantify sleep alterations in adults with ASD. We performed a systematic review and meta-analysis of objective (ie, based on actigraphy or polysomnography [PSG]) and subjective (ie, based on sleep diaries/questionnaires) studies comparing sleep parameters in adults with ASD and in a typically developing (TD) control group.

Methods: PubMed, OVID databases and Web of Knowledge were systematically searched up to February 2019 with no language restrictions. Original studies including adults with a diagnosis of ASD according to DSM, ICD, or based on standard diagnostic tools (eg, ADOS), and a TD control group were included. Random-effects models were used. Study quality was evaluated with the Newcastle Ottawa Scale (NOS). Analyses were conducted using Comprehensive Meta-Analysis.

Results: From initial pool of 1948 references, 14 publications including 8 datasets, (194 ASD and 277 controls) met the inclusion criteria. Compared to controls, individuals with ASD were significantly more impaired in six out of 11 subjective parameters, including lower sleep efficiency (SE, SMD = -0.87, CI = -1.14 - 0.60) and in 10 out of 17 objective outcomes, including longer sleep onset latency (PSG) (SMD = 0.86, CI = 0.29-1.07) and wake after sleep onset (WASO, actigraphy) (SMD = 0.57, CI = 0.28-0.87). The mean NOS score was 4.88/6.

Conclusions: Individuals with ASD demonstrated impaired sleep compared to controls in most subjective and objective measures.