Affective burden and resting-state EEG microstate dynamics in sleep-severity-matched chronic insomnia

Clinical heterogeneity in chronic insomnia disorder may partly reflect affective symptom burden, but its neurophysiological correlates are difficult to distinguish from those of insomnia severity. We examined whether resting-state electroencephalographic microstate dynamics differed between affective-burden phenotypes with comparable subjective sleep severity.

Among 125 treatment-naive adults with chronic insomnia disorder, 21 pairs (n = 42) with low versus high affective burden, defined using prespecified Hamilton Anxiety and Depression Rating Scale thresholds, were matched 1:1 on standardized Pittsburgh Sleep Quality Index and Insomnia Severity Index scores. Resting-state electroencephalographic microstate features were compared between groups using linear models adjusted for residual subjective sleep severity. Further exploratory analyses of eyes-open features examined symptom associations and full-cohort dimensional associations, and evaluated classification using leave-one-pair-out cross-validation.

During eyes-open rest, the high-affective-burden phenotype showed greater microstate C coverage and lower microstate D coverage, occurrence, and duration. Microstate C coverage correlated positively with a composite affective score, whereas microstate D occurrence correlated negatively with anxiety, depression, and composite affective scores. Eyes-closed analyses showed a similar trend, but the differences did not reach statistical significance. A four-feature linear support vector classifier yielded a cross-validated area under the curve of 74.8%. Corresponding effects were directionally concordant but attenuated to non-significance in full-cohort dimensional models.

Although limited by the small matched sample and reliance on questionnaire-based sleep severity without objective sleep monitoring, these findings suggest that C/D microstate dynamics may be associated with affective heterogeneity in chronic insomnia. This potential association warrants further investigation in larger cohorts incorporating objective sleep measures, while the exploratory classification findings require independent external validation.