Neural correlates of device-based sleep characteristics in adolescents

Abstract

Understanding the brain mechanisms underlying objective sleep patterns in adolescents and their implications for psychophysiological development is a complex challenge. Here, we applied sparse canonical correlation (sCCA) analysis on 3300 adolescents from Adolescent Brain Cognitive Development (ABCD) study, integrating extensive device-based sleep characteristics and multimodal imaging data. We revealed two sleep-brain dimensions: one characterized by later being asleep and shorter duration, linked to decreased subcortical-cortical network functional connectivities; the other showed higher heart rate and shorter light sleep duration, associated with lower brain volumes and decreased functional connectivities. Hierarchical clustering based on brain dimension associated with sleep characteristics revealed three biotypes of adolescents, marked by unique sleep profiles: biotype 1 exhibited delayed and shorter sleep, coupled with higher heart rate during sleep; biotype 3 with earlier and longer sleep, accompanied by lower heart rate; and biotype 2 with intermediate pattern. This biotypic differences also extended to cognition, academic attainment, brain structure and function in a gradient order. Longitudinal analysis demonstrated consistent biotypic differences from ages 9 to14, highlighting enduring cognitive and academic advantages in biotype3. The linked sleep-brain dimensions and the associated biotypes were well replicated in a longitudinal sample of 1271 individuals. Collectively, our novel findings delineate a linkage between objective sleep characteristics and developing brain in adolescents, underscoring their significance in cognitive development and academic attainment, which could serve as references for individuals with sleep difficulties and offer insights for optimizing sleep routines to enhance better cognitive development and school achievement.