Background: Sleep macro and microstructure has been suggested to reflect brain changes that support cognitive and behavioral maturation and contain potential markers to track typical and atypical development. In present study, we use two large samples from the National Sleep Research Resource to investigate how sleep characteristics change during normal development as well as in the subgroups with neurodevelopmental diagnoses (NDD).
Methods: We used polysomnography data from the NCH sleep biobank (2841 individuals, 9 [2.5-17.5] years) and the Child Adenotonsillectomy Trial (CHAT, 1232 individuals, 7 [4.5-10] years). Using clinical records available for the NCH dataset we selected clinical subgroups of patients with NDD: autism spectrum disorder, attention deficit and hyperactivity disorder, expressive language disorder, developmental coordination disorder, intellectual disability, Down syndrome, and cerebral palsy (all n>100). For all preprocessing steps we used an open-source software Luna. Linear regression was employed to investigate effects of age, diagnosis and their interaction including sex and race as covariates.
Results: Congruently with previous studies, we found increase in N2 proportion and decrease in duration of N3 and Rapid Eye Movement (REM) (in NCH all p-values<10-50). Confirming reports of sleep issues, the NDD subgroups displayed a unified pattern of alterations indicating shorter, less efficient sleep and a reduction of REM sleep. While absolute power uniformly decreased with age, developmental changes in relative power were more stage-, band- and channel specific. The strongest age effect was observed for relative sigma power during N2 (R=0.6 in NCH and R=0.28 in CHAT, both p-values< 10-20). Both slow and fast spindle density and inter-spindle deceleration increase with age (p-values < 10-10) but only slow spindle density and chirp were altered in the NDD subgroups and were significantly associated with degree of intellectual disability.
Conclusions: Sleep macro and micro-architecture examination can facilitate better understanding of the atypical neurodevelopment and help in search for the potential targets for NDD therapeutic interventions.