Electroencephalography (EEG) microstates have emerged as potential biomarkers of large-scale brain network dynamics. However, their role in depression remains unclear due to inconsistent findings and limited replication. This study investigated whether microstate parameters can differentiate depression. Resting-state EEG was analyzed from 122 young adults (76 controls, 23 subclinical depression, 23 major depressive disorder (MDD)). Microstate analysis was conducted using standardized pipelines (MICROSTATELAB in EEGLAB), with duration, occurrence, and coverage extracted for six-class solutions. Group differences were assessed using Bayesian ANCOVA, while associations with depressive and anxiety symptoms (BDI, STAI) were evaluated with Bayesian regression. Replication analyses were performed using a second independent EEG recording. Test–retest reliability was assessed with intraclass correlation coefficients. Microstate G duration was reduced in both subclinical and high-symptom groups compared with controls and showed negative associations with BDI and STAI scores. These effects are partially replicated in the second dataset. Microstate G also demonstrated high test–retest reliability (ICC = 0.842). In contrast, microstate A showed weaker and less reliable associations with depressive symptoms. Microstate G represents a reliable electrophysiological marker of depressive symptomatology. These findings highlight EEG microstate analysis as a promising approach for developing objective, dimensional biomarkers of depression.
Jinwon Chang (Wed,) studied this question.