MULTIMODAL STRUCTURAL BRAIN ALTERATIONS IN SCHIZOPHRENIA: GRAY MATTER LOSS AND DISRUPTED STRUCTURAL COVARIANCE NETWORKS

Schizophrenia (SCZ) is a heterogeneous neuropsychiatric disorder. Multimodal structural approaches may help integrate regional and network-level alterations. We investigated structural brain alterations using multimodal MRI data from the INSCRIPTION project in 75 young adults (35 SCZ patients, 40 controls; mean age 32.8±9.2 years). Voxel-based morphometry (VBM) and Freesurfer quantified regional gray matter volume. Additionally, individual structural covariance networks were constructed using Morphometric INverse Divergence (MIND). General linear models were adjusted for age, sex, laterality, education attainment, and total intracranial volume. VBM results were considered significant if surviving a voxelwise threshold of p<0.001 (k=100). Freesurfer and MIND results were adjusted using FDR. VBM revealed gray matter reductions in patients, in the bilateral middle cingulate, anterior cingulate, and insular cortices. Freesurfer analyses showed reduced volume in the whole brain (p_FDR=3𝗑10^-6), mid posterior (p_FDR=0.002) and central corpus callosum (p_FDR=0.010), and left thalamus (p_FDR=0.020), in patients. MIND identified decreased strength and eigenvector centrality in the left fusiform and parahippocampal lobe, and increased degree in the right superior frontal cortex (p_FDR<0.05), indicating altered centrality and similarity within the structural covariance network (Figure 1). Including medication use as a model covariate confirmed the significance of the findings in key regions. These results suggest multiscale structural disruption in young SCZ patients, involving regional gray matter reductions and altered structural network topology. Multimodal approaches such as MIND provide complementary insights into disease-related structural network changes. Figure 1. Structural brain alterations and network disruptions in young SCZ patients. A) VBM results; B) Freesurfer regional volumes; C) MIND results.