Persistent affective disturbance is a core, disabling feature of major depressive disorder (MDD), thought to stem from a dysfunctional interaction between emotional bias and cognitive control. However, the underlying neural dynamics are debated, with studies reporting both hyper- and hypoactivation. This study utilized high-temporal-resolution electroencephalogram (EEG) to resolve this discrepancy by examining distinct stages of emotional information processing.

We recruited 175 medication-free patients with MDD (Hamilton Depression Rating Scale-17 ≥ 14) and 101 healthy controls (HCs) who completed an emotional Stroop task while an EEG was recorded. We analyzed event-related potentials reflecting conflict monitoring (N250), inhibition (N450), and resolution (LSP) using a 2 (group) × 2 (valence) × 2 (congruency) analysis of variance.

Results revealed a stage-specific neural cascade. Compared to HCs, the MDD group showed: (1) hypoactivation during initial conflict monitoring (attenuated N250 amplitude); (2) compensatory hyperactivation during conflict inhibition (a significant N450 interaction revealed generalized conflict activity in MDD, unlike the context-specific response in HCs); and (3) subsequent hypoactivation during conflict resolution (reduced LSP amplitude for negative stimuli). Crucially, altered N450 correlated with depression severity, and the entire neural cascade predicted behavioral performance.

The apparent contradiction in the literature reflects a multistage process. MDD is characterized by an inefficient neural cascade: an initial deficit in conflict monitoring is followed by compensatory overactivation during inhibition, which ultimately proves insufficient, leading to impaired late-stage resolution. This temporally specific model advances our understanding of the pathophysiology of depression and identifies potential stage-specific targets for intervention.