Changes in brain function during negative emotion processing following cognitive-behavioural therapy in depressive disorders.

Abstract

Cognitive-behavioural therapy (CBT) is a first-line treatment for depressive disorders, but research on its neurobiological mechanisms is limited. Given the heterogeneity in CBT response, investigating the neurobiological effects of CBT may improve response prediction and outcomes. To examine brain functional changes during negative emotion processing following naturalistic CBT. In this case-control study, 59 patients with depressive disorders were investigated before and after 20 CBT sessions using a negative-emotion-processing paradigm during functional magnetic resonance imaging, clinical interviews and depressive symptom questionnaires. Healthy controls (<i>n</i> = 60) were also assessed twice within an equivalent time interval. Patients were classified into subgroups based on changes in diagnosis according to DSM-IV criteria (<i>n</i> = 40 responders, <i>n</i> = 19 non-responders). Brain activity changes were examined using group × time analysis of variance for limbic areas, and at the whole-brain level. Analyses yielded a significant group × time interaction in the hippocampus (<i>P</i> family-wise error [<i>P</i>_FWE] = 0.022, <i>η</i> _<i>P</i> ² = 0.101), and a significant main effect of time in the dorsal anterior cingulate cortex (<i>P</i>_FWE = 0.043, <i>η</i> _<i>P</i> ² = 0.098), resulting from activity decreases following CBT (<i>P</i>_FWE ≤ 0.024, <i>η</i> _<i>P</i> ² ≤ 0.233), with no changes in healthy controls. Hippocampal activity decreases were driven by responders (<i>P</i>_FWE ≤ 0.020, <i>η</i> _<i>P</i> ² ≤ 0.260) and correlated with symptom improvement (<i>r</i> = 0.293, <i>P</i> = 0.024). Responders exhibited higher pre-treatment hippocampal activity (<i>P</i>_FWE = 0.017, <i>η</i> _<i>P</i> ² = 0.189). Following CBT, reduced activity in emotion-processing regions was observed in patients with depressive disorders, with hippocampal activity decreases linked to treatment response. This suggests successful CBT could correct biased emotion processing, potentially by altering activity in key areas of emotion processing. Hippocampal activity may function as a predictive marker of CBT response.