Self‐reported affective biases, but not all affective performance biases, are present in depression remission

Objective Individuals with active major depressive disorder (MDD) have shown affective biases in cognitive flexibility and memory, particularly for negatively valenced stimuli. We evaluated whether impairments in affective flexibility would remain even during remission (rMDD), potentially representing trait‐ or scar‐like effects of illness. Method: Participants completed the Emotion Card Sort Test (ECST), a measure of cognitive flexibility containing emotionally valenced stimuli, and the Emotion Word Stimulus Test (EWST), a measure of affective biases in delayed recall and recognition memory, and several self‐report measures. Results Healthy controls (HCs; n = 35) and individuals with rMDD (n = 93) did not differ on performance for any of the three word types on the ECST or EWT. However, individuals with rMDD demonstrated greater negative bias on EWT recognition trials relative to HCs (d = .36). On self‐report measures, individuals with rMDD exhibited greater levels of neuroticism, problems with attentional control, pessimistic attributional style, and negative automatic thoughts compared to HCs. Conclusions These results provide initial evidence that some performance, but not self‐reported, indices of affective bias may improve during remission from MDD. Results of this study could suggest that some components of affective bias may represent state feature of illness and others trait‐like risk or scar features. Practitioner points This study suggests that self‐reported affective biases may persist in remission of major depressive disorder (rMDD). Affective attentional biases and affective memory biases were not demonstrated in individuals with rMDD, with the exception of a bias for recognizing negatively versus neutrally valenced stimuli. Cautions or limitations A limitation of this study was its cross‐sectional design. Under ideal conditions, the same individuals would be studied in both the active and remitted phases of illness. Another limitation of this study was the smaller number of healthy controls relative to individuals with rMDD.