Abstract

Multiple brain imaging parameters have been used to study the pathophysiology of schizophrenia; however, the difficulties in data collection often limit the size of the sample. The purpose of this study was to compare the brain networks of schizophrenia patients and healthy controls using structural covariance analysis of datasets from multiple sites. We obtained a total sample of 652 patients and 415 healthy controls. When comparing schizophrenia patients and controls, 36 correlation coefficients between two ROIs exhibited significant group difference, including 23 out of the 68 brain regions defined in this study. The brain regions that demonstrated between-group differences in structural covariance were also associated with the symptom severity of schizophrenia and deficits in neurocognitive function in the patient group. Rather than focusing solely on brain deterioration and reduction to understand schizophrenia, the results of this study emphasize the need to consider the complex patterns of brain networks and explore how these structural relationships correlate with the clinical symptoms and cognitive impairment in schizophrenia patients.