Background: Schizophrenia (SZ) is characterized by progressive cortical thinning. However, the spatiotemporal trajectory of this structural decline across the illness stages and its underlying mechanisms remain unclear. Methods: We conducted a cross-staged neuroimaging study spanning clinical high risk (CHR), first-episode psychosis (FEP), early SZ, and chronic SZ. First, stage-specific epicenter of cortical thinning was identified using structural Magnetic Resonance Imaging (MRI) data. Next, hierarchical functional network was constructed by stepwise functional connectivity (SFC) analysis. We then tested whether alterations in SFC could predict shift of epicenter across stages. Additionally, we tested the repeatability and robustness of the epicenter results by several validating analyses. Findings: We found that epicenter progressively shift from occipital-temporal to fronto-parietal cortices across progressive stages of SZ. Additionally, epicenter likelihood was correlated with the severity of positive, negative, and cognitive symptoms. Moreover, the spatial progression of epicenter can be predicted by altered SFC. Notably, long-range connectivity emerged as stronger predictor of epicenter distribution. Meanwhile, alterations in SFC distance during earlier stages were also associated with epicenter likelihood in chronic stage. Finally, the robustness of epicenter findings was confirmed through replication in independent validation cohorts. The results also remained consistent after regressing out antipsychotic equivalent. Interpretation: Our work delineates a hierarchical progression of cortical pathology across stages of SZ, where epicenters progressively shift from lower to higher-order cortices. This spatiotemporal dynamic is actively shaped by the brain's intrinsic functional hierarchy. These findings offer a novel network-based staging framework for identifying phase-specific biomarkers and targeting therapeutic interventions.
原文链接
分享
引用
