Stage-specific EEG spectral signatures of primary insomnia: regional and sex-related differences across NREM and REM sleep.

NREM EEG patterns separated the insomnia and healthy-sleeper groups more clearly than REM patterns in this 100-person study. The result improves understanding of sleep physiology but is not yet a validated diagnostic test.

Premakumara N, Yang CY, Sudantha BH, Chen KW.
Physiological measurement · Publication date: 2026-09-18
Added to site: 2026-09-19
DOI: 10.1088/1361-6579/aea9ee

Online-ahead-of-print publication date verified in PubMed.

Modalities

QEEG, EEG, Sleep EEG

Clinical focus

Primary insomnia, Sleep stages, Explainable machine learning

Population

Adults with primary insomnia and healthy sleepers

Sample

100 adults: 50 healthy sleepers and 50 with primary insomnia

Intervention / method

Regional spectral-power analysis across frontal, central and occipital sites during NREM and REM, with LASSO, random forest, CatBoost and XGBoost classifiers and SHAP feature interpretation

Comparator

Primary-insomnia versus healthy-sleeper recordings; NREM-versus-REM and sex-stratified analyses

Clinician summary

Reduced NREM delta and higher beta activity, especially centrally, were the dominant model features. REM models emphasized prefrontal beta1 and occipital alpha, and feature importance differed by sex.

How to interpret the findings

The abstract does not report external validation, recruitment and medication controls, full classifier metrics, leakage safeguards or prospective diagnostic performance. Group classification and SHAP importance do not establish causal mechanisms or clinical utility.

Equipment and disclosures

EEG was analyzed at F3, F4, C3, C4, O1 and O2. The acquisition hardware and polysomnography system were not specified in the accessible abstract.

Funding and competing-interest details were not available in the checked indexed record.

Related evidence

Compare findings from the same clinical area, including mixed results.