A multi-metric profiling of OFC subregion connectivity identifies system-level alterations in major depressive disorder.

BACKGROUND: Orbitofrontal cortex (OFC) dysfunction contributes to major depressive disorder (MDD), particularly through disrupted network interactions. Current studies lack whole-brain coverage and fine-grained OFC mapping. We employed OFC parcellation and multi-metric connectivity to comprehensively profile OFC-related network disturbances in MDD. METHODS: OFC was parcellated into medial (MOFC), anterior middle (aMidOFC), posterior middle (pMidOFC), and lateral (LOFC) subregions via community analysis of resting-state fMRI from 586 healthy individuals. Using these subregions and whole OFC as seeds, we computed whole-brain functional connectivity (FC), total interdependence (TI), and Granger causality (GC) from 83 MDD patients and 80 healthy controls. Group differences were evaluated, and a support vector machine integrated multi-metric indicators to identify key network alterations. RESULTS: FC revealed altered coordination between OFC subregions and distributed networks in MDD. GC and TI showed imbalanced anterior/posterior default mode network (DMN) interactions with right MOFC, LOFC, and pMidOFC; enhanced ventral attention network influence to left MOFC and pMidOFC; increased right MOFC and left pMidOFC interactions with sensorimotor and visual systems; and reduced dorsal anterior cingulate cortex input to left LOFC. Machine learning identified that alterations involving bilateral LOFC, MOFC, and left pMidOFC with DMN, ventral attention network, and dACC contributed most to classification. CONCLUSION: These findings reveal widespread OFC-subnetwork dysregulation in MDD, reflecting impaired coordination with distributed neural systems that may underlie the disorder's pathophysiology.