EEG Microstate Alterations in Anhedonia Subtypes of Major Depressive Disorder and Their Modulation by Intermittent Theta‐Burst Stimulation

Abstract

Background Anhedonia is a core symptom of major depressive disorder (MDD) and is associated with functional impairment, treatment resistance, and relapse. Emerging evidence suggests that anhedonia severity reflects distinct neural alterations; however, current neuromodulation protocols have not yet accounted for this clinical heterogeneity. We investigated whether intermittent theta‐burst stimulation (iTBS) differentially modulates electroencephalography (EEG) microstate dynamics across anhedonia subtypes and aimed to identify electrophysiological signatures for treatment stratification. Method Fifty MDD patients were stratified into high‐anhedonia (HA, n = 25) and low‐anhedonia (LA, n = 25) groups based on Snaith–Hamilton pleasure scale (SHAPS) scores, alongside 25 healthy controls (HCs). Clinical assessment and resting‐state EEG data were collected at baseline and after a 2‐week course of iTBS targeting the left dorsolateral prefrontal cortex (DLPFC). Microstate parameters and transition probabilities (TPs) were analyzed. Results At baseline, both patient groups showed reductions in microstates, MS_A and MS_D relative to HCs. The HA group additionally exhibited MS_C deficits, whereas the LA group showed MS_B reductions. TPs were decreased in both groups, with broader deficits in the HA group. Following iTBS, most microstate parameters of LA group no longer differed significantly from those of HCs, whereas the HA group exhibited only partial changes, with persistent abnormalities in MS_B duration, MS_C coverage, and MS_D parameters. Correlation analyses further revealed that, in the HA group, greater anhedonia improvement was associated with increased MS_A duration and decreased MS_B duration. Conclusion iTBS exerts subtype‐specific modulation of large‐scale network dynamics in MDD with anhedonia. The LA group demonstrates treatment‐associated resolution, whereas the HA group displays enduring integration deficits involving the dorsal attention network (DAN), sensory network (SN), and default mode network (DMN). Notably, in the HA group, clinical improvement correlates with SN reorganization, suggesting that microstate changes may serve as a treatment response marker. MS_D dynamics represent a candidate signature for subtype classification, supporting stratification‐guided neuromodulation.