Background: Major depressive disorder (MDD) is a leading cause of disability worldwide and exhibits high heterogeneity, particularly in cognitive impairments. Understanding the biological mechanisms underlying cognitive impairment in MDD is crucial for developing individualized treatment.
Methods: This study included 80 first-episode medication-naïve (FEDN) MDD patients and 40 healthy controls (HC). Sparse multiple canonical correlation analysis (smCCA) and K-means clustering were used to identify MDD subtypes based on individualized functional connectivity (FC), cognitive performance, and clinical symptoms.
Results: Two data-driven MDD subtypes were identified: the global impairment subtype (giMDD), characterized by widespread cognitive deficits across memory, processing speed, and executive function; and the memory-specific impairment subtype (miMDD), showing memory-specific impairments. giMDD exhibited unique hyperconnectivity between the dorsal attention network (DAN) and frontoparietal network (FN), and between the ventral attention network (VAN) and FN.
Conclusions: This study highlights the heterogeneity of cognitive impairment in MDD and identifies two data-driven MDD subtypes associated with distinct brain network connectivity patterns. These findings provide insights into the biological mechanisms underlying MDD heterogeneity and offer a potential framework for personalized treatment strategies. Future studies are needed to evaluate the clinical utility of these subtypes in improving individualized care for MDD.
Keywords: Cognition; Individualized functional connectome; Major depressive disorder; Subtype; fMRI.
Copyright © 2026 Elsevier B.V. All rights reserved.