Long-term morphometric similarity gradients relate to cortical hierarchy and psychiatric symptoms in schizophrenia

Abstract Schizophrenia spectrum disorders (SSD) are characterized by altered brain structure, reflecting widespread dysconnectivity across brain-specific networks. However, the role of hierarchical organization on cortical morphometric networks in shaping clinical outcomes over the course of the disease remains unclear. Connectome-derived gradients have increasingly been used to investigate spatial transitions in brain organization. Here, we computed cortical and subcortical Morphometric INverse Divergence (MIND) similarity networks from 1293 structural MRI data of 193 healthy controls (HC) and 350 individuals with SSD followed for up to 20 years. MIND features were calculated for each subject-specific network by computing regional averages and performing gradient decomposition. We found that MIND in SSD was longitudinally associated with treatment duration and medication. These associations were co-localized with hierarchical axes of cortical organization and schizophrenia epicenters. Moreover, psychiatric symptoms were associated with these alterations in structural similarity, which were also related to treatment duration. Collectively, these findings advance our understanding of how brain organization, treatment duration, and medication shape clinical symptoms throughout the course of SSD.