Abstract:The neurobiological heterogeneity of schizophrenia is widely accepted, but it is unclear how mechanistic differences converge to produce the observed phenotype. Establishing a pathophysiological model that accounts for both neurobiological heterogeneity and phenotypic similarity is essential to inform stratified treatment approaches. In this cross-sectional diffusion tensor imaging study, we recruited 77 healthy controls, and 70 patients with DSM-IV diagnosis of schizophrenia. We first confirmed the heterogeneity in structural connectivity by showing a reduced between-individual similarity of the structural connectivity in patients compared to healthy controls. Second, at a system level, we found the diversity of the topographic distribution of the strength of structural connectivity was significantly reduced in patients (P = 7.21 × 10−7, T142 = 5.19 [95% CI: 3.37–7.52], Cohen’s d = 0.91), and this affected 65 of the 90 brain regions examined (False Discovery Rate <5%). Third, when topographic diversity was used as a discriminant feature to train a model for classifying patients from controls, it significantly improved the accuracy on an independent sample (T99 = 5.54; P < 0.001). These findings suggest a highly individualized pattern of structural dysconnectivity underlies the heterogeneity of schizophrenia, but these disruptions likely converge on an emergent common pathway to generate the clinical phenotype of the disorder.

Topographic diversity of structural connectivity in schizophrenia