Within cohorts, NREM slow-wave power did not decrease with age in 42 patients recorded with SEEG

This study tested whether NREM slow-wave power (0.5–2 Hz) recorded at clinically normal bipolar depth pairs decreases with age, as scalp EEG studies report. We analyzed 90 s (30–120 s) from the first NREM-sleep SEEG file for each of 78 subjects with SEEG channels in OpenNeuro ds007118, ds007119, and ds007120 v1.0.0. The main sample comprised 42 epilepsy patients aged 6–45 years (median 21), with at least five normal-normal pairs (671 pairs). Welch spectra were calculated in 4 s windows. Patient values were medians across normal-normal pairs; the cohort-stratified statistic was the cohort-size-weighted mean within-cohort Spearman correlation, tested by 20,000 within-cohort age permutations. Log 0.5–2 Hz power showed no age decrease (S = −0.001, p = 0.997; 95% bootstrap CI −0.30 to +0.30). Relative slow-wave power also did not decrease (S = +0.156, p = 0.362; Holm p = 0.723). Neither primary hypothesis was supported. No band passed Holm correction, and within-cohort correlations differed in sign. An exploratory subset of nine subjects under 18 in one cohort had S = −0.745 (uncorrected p = 0.025); this result is not a finding. The study mainly tests late-adolescent and adult ages and is weak for ages 6–15. NREM stage, electrode site, medication, and cohort hardware effects were not controlled; planned adjustment for pair distance and temporal-lobe fraction was not feasible because data were mostly missing.

NREM sleep · slow-wave power · SEEG · intracranial EEG · age · adolescence · epilepsy · Spearman correlation · OpenNeuro