Critical states in sleep EEG: developmental evolution across 3192 clinical children aged 0-18 years

Sleep is a crucial process that assists the brain in returning to its critical state with optimal information processing capacity. How does the ability of sleep to restore the critical states develop with age? This study proposes to analyze the sleep EEG dynamics in children using long-range temporal correlation, to investigate the evolution of the criticality in brain system during sleep.

In this study, we screened the sleep EEG of 3192 clinical children aged 0 to 18 years in the Nationwide Children's Hospital Sleep Databank and analyzed the developmental evolution of long-range temporal correlation of EEG in sleep stages (wake, N1, N2, N3, and REM) using detrended fluctuation analysis (DFA).

In synchronized N3 sleep, the DFA exponent remains at ~1.5 across all ages, indicating highly regular, Brownian-like activity far from criticality. In light (N1/N2) and REM sleep, the exponent falls progressively from 1.5 in neonates to ~1.0 by late adolescence, marking the emergence of scale-free, network-critical dynamics.

This study provides the first population-scale evidence linking sleep architecture to the developmental approach toward neural criticality. The stage-specific trajectories align with known periods of synaptic pruning and network refinement, offering a biomarker of cortical maturation and highlighting sleep's role in tuning the brain toward optimal adaptability.