Comparison of power spectra from overnight electroencephalography between patients with Down syndrome and matched control subjects

Abstract

SummaryElectroencephalograms can capture brain oscillatory activities during sleep as a form of electrophysiological signals. We analysed electroencephalogram recordings from full‐night in‐laboratory polysomnography from 100 patients with Down syndrome, and 100 age‐ and sex‐matched controls. The ages of patients with Down syndrome spanned 1 month to 31 years (median 4.4 years); 84 were younger than 12 years, and 54 were male. From each electroencephalogram, we extracted relative power in six frequency bands or rhythms (delta, theta, alpha, slow sigma, fast sigma, and beta) from six channels (frontal F3 and F4, central C3 and C4, and occipital O1 and O2) during five sleep stages (N3, N2, N1, R and W)—180 features in all. We examined differences in relative power between Down syndrome and control electroencephalograms for each feature separately. During wake and N1 sleep stages, alpha rhythms (8.0–10.5 Hz) had significantly lower power in patients with Down syndrome than controls. Moreover, the rate of increase in alpha power with age during rapid eye movement sleep was significantly slower in Down syndrome than control subjects. During wake and N1 sleep, delta rhythms (0.25–4.5 Hz) had higher power in patients with Down syndrome than controls. During N2 sleep, slow sigma rhythms (10.5–12.5 Hz) had lower power in patients with DS than controls. These findings extend previous research from routine electroencephalogram studies demonstrating that patients with Down syndrome had reduced circadian amplitude—the difference between wake alpha power and deep sleep delta power was smaller in Down syndrome than control subjects. We envision that these brain oscillatory activities may be used as surrogate markers for clinical trials for patients with Down syndrome.