Key points are not available for this paper at this time.
Abstract Introduction Structural changes emblematic of brain development, e.g., synaptic pruning/myelination, are theorized to impact sleep architecture, local sleep physiology, and sleep-dependent cognition. Conventional analyses of sleep in development, relying on absolute or relative power from electroencephalography (EEG) to study the function of sleep rhythms, have disregarded the 1/f EEG power law as “neural noise". However, evidence tying steepness (1/f exponent) and intercept (1/f offset) of the power law curve to excitatory-inhibitory (E/I) balance and asynchronous neuronal population spiking warrants a reexamination of developmental sleep EEG. Methods 19 healthy children and adolescents (ages 11-17, 11 girls) underwent overnight polysomnography with 128 channel EEG. Inhibitory control was assessed following sleep using inhibition scaled score (ISS) on a Color Word Interference Test. Spectral parameterization (fooof) was applied to median NREM multitaper absolute spectra (0.5-40Hz) to extricate oscillatory from aperiodic features (1/f exponent, 1/f offset, and residual oscillatory power). Topographical correlations between parameterized spectra, age and ISS were obtained with 5000-permutation threshold free cluster enhancement. Bootstrapped multiple regressions adjusting for age and sex modeled the association between ISS and parameterized neural spectra. Results Age showed widespread negative topographical correlations with broadband absolute spectral power, but these effects vanished in correlations with all oscillatory residuals except slow sigma (11-13Hz), which showed a modest negative trend over a central cluster. Remarkably, age showed widespread negative associations with 1/f offset, while age and 1/f exponent showed a negative trend over central channels. ISS was not significantly correlated with absolute power or with oscillatory residuals in canonical bands but was strongly negatively associated with 1/f exponent in a bilateral parietal arc. The cluster average of exponents from this parietal association significantly predicted next day ISS performance, adjusting for age and sex (B = 0.803, p = 0.013). Conclusion Aperiodic rather than oscillatory activity may drive broadband changes in developmental NREM spectra , and parietal E/I balance during sleep may partially explain (independent of age) developmental inhibitory control during wakefulness. More broadly, aperiodic features may complement neural oscillations in balancing localized and systemwide information processing demands to support developmental cognition. Support (if any) NIH NCATS Grant #UL1TR001414; Pediatric Exercise and Genomics Research Center Systems Biology Fund
Dave et al. (Sat,) studied this question.