Shared cortical morphometry and molecular signatures across depression and physical diseases

Major depressive disorder (MDD) is highly comorbid with physical diseases, suggesting shared neurobiological mechanisms. However, the common structural and molecular substrates underlying this comorbidity remain poorly understood. We analyzed multimodal MRI data from 577 patients (MDD, asthma, fibromyalgia, gastroesophageal reflux disease, type 2 diabetes) and 275 healthy controls, constructed individual morphometric similarity networks (MSN) to identify shared cortical alterations, and applied principal component analysis with multiscale molecular annotation (transcriptomic, cell-type, neurotransmitter, and cognitive mapping) to characterize underlying mechanisms. All patient groups exhibited regional MSN abnormalities. A shared MSN pattern emerged, featuring increased similarity in visual and somatomotor regions and decreased similarity in ventral attention, limbic, and default mode networks, and was replicated in independent cohorts. This pattern was significantly associated with spatial gene expression enriched for synaptic signaling, membrane potential regulation, and neuronal architecture, with predominant expression in excitatory/inhibitory neurons and astrocytes. It further mapped onto specific neurotransmitter systems, with exploratory cognitive decoding suggesting possible relevance to emotion and social cognition. Related genes overlapped with known risk loci for Alzheimer's disease and bipolar disorder. These findings identify a robust across diseases cortical MSN phenotype linked to convergent molecular and neurochemical correlates, offering a framework for studying shared vulnerability and transdiagnostic mechanisms across depression and physical diseases.