Abstract
OBJECTIVES: There is a lot of variability between the results of studies reporting the pattern of gray matter volume changes in schizophrenia. Methodological issues may play an important role in this heterogeneity. The aim of the present study was to replicate the better performance of multivariate "source-based morphometry" (SBM) over the mass-univariate approach. EXPERIMENTAL DESIGN: Voxel-based morphometry of Jacobian-modulated gray matter volume images, using voxel and cluster level inference, and SBM were performed in a group of first-episode schizophrenia patients (N = 49) and healthy controls (N = 127). RESULTS: Using SBM we were able to find a significant reduction of gray matter volume in fronto-temporo-cerebellar areas whereas no significant results were obtained using voxel-based morphometry. CONCLUSION: Multivariate analysis of gray matter volume seems to be a suitable method for characterization of the pattern of changes at the beginning of the illness in schizophrenia subjects.