Data-driven clustering differentiates subtypes of major depressive disorder with distinct brain connectivity and symptom features

Background Major depressive disorder (MDD) is a clinically and biologically heterogeneous syndrome. Identifying discrete subtypes of illness with distinguishing neurobiological substrates and clinical features is a promising strategy for guiding personalised therapeutics. Aims This study aimed to identify depression subtypes with correlated patterns of functional network connectivity and clinical symptoms by clustering patients according to a weighted linear combination of both features in a relatively large, medication-naïve depression sample. Method We recruited 115 medication-naïve adults with MDD and 129 matched healthy controls, and evaluated all participants with magnetic resonance imaging. We used regularised canonical correlation analysis to identify component mapping relationships between functional network connectivity and symptom profiles, and K -means clustering was used to define distinct subtypes of patients. Results Two subtypes of MDD were identified: insomnia-dominated subtype 1 and anhedonia-dominated subtype 2. Subtype 1 was characterised by abnormal hyperconnectivity within the ventral attention network and sleep maintenance insomnia. Subtype 2 was characterised by abnormal hypoconnectivity in the subcortical and dorsal attention networks, and prominent anhedonia symptoms. Conclusions Our study identified two distinct subtypes of patients with specific neurobiological and clinical symptom profiles. These findings advance understanding of the biological and clinical heterogeneity of MDD, offering a pathway for defining categorical subtypes of illness via consideration of both biological and clinical features.