Key points are not available for this paper at this time.
Abstract Introduction Sleep electroencephalography (EEG) patterns exhibit intricate variations influenced by multiple factors, including sex-specific disparities. While sex effects in sleep neurophysiology are relatively well-described in adults, information is more limited in pediatric samples, even less so in specific cohorts such as neurodevelopmental disorders (NDD). In this study we aimed to describe sex differences in a comprehensive set of EEG-derived sleep features in non-NDD individuals and compare those to the NDD cohorts including autism spectrum disorder(ASD), attention deficit hyperactivity disorder(ADHD), and intellectual disabilities(ID). Methods We used whole-night polysomnography data from a large National Sleep Research Resource sample, spanning from 2.5 to 17.5 years and Nationwide Children’s Hospital Sleep Databank, a pediatric sleep clinic cohort. Using ICD codes, we have defined the NDD subsets for ASD (N=196), ADHD (N=525), intellectual disabilities (N=167), and no-NDD sample (N=1523). The open-source package Luna was used to derive multiple sleep features including sleep spindles, slow oscillations (SOs) and spectral power across standard frequency bands. Linear regression models controlling for age were used to investigate sex effects with standardized b-coefficients (bstd) and p-values reported. Results Almost half of sleep EEG features expressed nominally significant sex differences in the no-NDD cohort (165 out of 351 tested EEG metrics), although their effect sizes were rather small. The strongest sex differences were observed in fast spindle density across all channels (largest effect at C4, bstd=-0.22,p=7x10-6 at C3, higher in girls), SO duration across all channels (largest effect at F3, bstd=0.24,p=1x10-5, longer in boys), occipital spectral power in alpha and sigma bands during REM sleep (bstd=-0.23,p=7x10-6 at O1, higher in girls). Generally, the NDD groups expressed similar pattern of sex differences evidenced by significant correlation between the standardized beta coefficients of each disorder and no-NDD group across all metrics (Pearson’s r=0.32 in ASD, 0.34 in ADHD and 0.27 in ID) although we also found evidence of disorder-specific sex differences. Conclusion Our findings indicate that multiple sleep EEG metrics express significant sex differences in childhood which are also observed in NDD populations. Such sex differences could potentially relate to disorder-relevant factors such as the differential prevalence of sleep issues in boys and girls or symptom manifestation. Support (if any)
Kozhemiako et al. (Sat,) studied this question.