Brain Behav. 2026 Aug;16(8):e71685. doi: 10.1002/brb3.71685.
ABSTRACT
BACKGROUND: Generalized anxiety disorder (GAD) is commonly accompanied by sleep disturbance in young adults. Whether clinically diagnosed GAD with self-reported habitual short sleep (HSS) is associated with altered resting-state large-scale functional connectivity remains unclear. We examined EEG functional connectivity in young adults with GAD + HSS and healthy controls using complementary connectivity measures and multilevel network analyses.
METHODS: Resting-state EEG data were analyzed from 74 college students, including 37 participants with clinically diagnosed GAD and self-reported HSS, defined as habitual weekday nocturnal sleep duration ≤ 6 h/night for at least 1 month, and 37 healthy controls. Functional connectivity was estimated using the weighted phase lag index (wPLI) and corrected amplitude envelope correlation (AEC-c) across delta, theta, alpha, and beta bands. Group differences were assessed using network-based statistics (NBS), node strength, and graph-theoretical analyses. Covariate-adjusted sensitivity analyses accounted for age and years of education, and graph metrics were tested across density thresholds of 10%-30%.
RESULTS: Significant group differences were confined to the alpha-band wPLI network. NBS identified one connected component comprising 434 suprathreshold edges (FWER-corrected p = 0.0002), predominantly showing stronger connectivity in the GAD + HSS group. All 30 scalp nodes showed higher node strength after FDR correction, including after covariate adjustment. The GAD + HSS group also showed higher global efficiency, lower characteristic path length, and higher weighted clustering coefficient, whereas transitivity did not differ. These findings were directionally stable across density thresholds.
CONCLUSIONS: Young adults with clinically diagnosed GAD and self-reported HSS showed distributed alterations in resting-state alpha-band phase-based connectivity. Convergent network-level, nodal, and graph-theoretical findings indicated stronger weighted integration and clustering. However, these results should not be interpreted as biomarkers of objectively verified sleep loss, and the independent contributions of anxiety and sleep status cannot be determined.
PMID:42638376 | DOI:10.1002/brb3.71685
