Neural mechanisms of cognitive behavioral therapy for depression: A multimodal systematic review

Abstract

Major depressive disorder involves disturbances in neural systems supporting emotion regulation, cognitive control, reward processing, and self-referential thought. Cognitive behavioral therapy (CBT) is an established treatment for depression, yet the neural mechanisms through which different CBT modalities exert therapeutic effects remain unclear. This review aimed to synthesize multimodal neuroimaging evidence to identify consistent and modality-specific neural changes associated with CBT for depression. A systematic search was conducted in a major biomedical database using predefined keywords. Eligible studies included peer-reviewed human research examining a CBT-based intervention and reporting pre- or posttreatment neural outcomes measured using functional magnetic resonance imaging, resting-state imaging, structural magnetic resonance imaging, or diffusion imaging. Study characteristics and neural results were extracted using a structured template. Findings were organized by CBT modality to compare neural outcomes across intervention types. Eleven studies met inclusion criteria. Across modalities, CBT was associated with changes in prefrontal, limbic, cingulate, and striatal systems. Standard CBT produced changes in prefrontal activation, limbic gray-matter volume, and connectivity between cognitive control and reward-related regions. Internet-based CBT was linked to enhanced reward-circuit engagement, reduced dominance of default-mode network states, and greater transitions toward central-executive network states. Rumination-focused CBT influenced posterior midline regions involved in repetitive negative thinking. Computer-assisted CBT increased connectivity between cognitive control and limbic circuits. Neurofeedback delivered prior to CBT enhanced amygdala activation to positive imagery and strengthened downstream reward-related engagement. Structural and diffusion findings demonstrated therapy-related gray- and white-matter plasticity consistent with functional reorganization. Across modalities, CBT produced convergent effects within neural systems aligned with the cognitive and emotional processes targeted by each intervention. These findings underscore the importance of large-scale network integration, reward sensitivity, and reduced self-focused rumination in symptom improvement. CBT influences neural systems at multiple organizational levels, with modality-specific patterns reflecting distinct therapeutic mechanisms. These results support the potential usefulness of neural markers in guiding personalized psychological treatment and highlight the value of considering intervention modality when interpreting neural outcomes