EEG microstate analysis between patients with major depressive disorder, subclinical depression, and healthy controls.

Introduction: 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. Methods: 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. Results and discussion: 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.