Structural Covariance-Based Morphometric Connectivity in Psychosis: Investigating Dysconnectivity and Cognition Across Disease Stages with Ultra-High Field MRI

Psychosis is among the most disabling disorders worldwide due to its broad spectrum of symptoms. Research has recognized brain dysconnectivity as a key feature of psychosis that is closely linked to cognitive impairments. Conventionally, brain connectivity is defined by structural and functional connectivity. This study employs a novel approach by using structural covariance, the correlation between morphometric changes across individuals, to define morphometric connectivity (MC). Although previous research has linked MC to developmental and cognitive processes, its role in psychosis and its relationship to cognition remain understudied, partly due to limitations associated with low-field magnetic resonance imaging (MRI) studies. This study investigates how MC and its correlation with cognitive functions is impacted by disease progression in psychosis. Data were collected using 7-Tesla MRI from 112 participants (Healthy Controls = 31, First-episode Psychotic Patients = 62, Clinical High-Risk Individuals = 10, Multiple-episode Psychotic Patients = 9). MC was computed using Graph Theory (strength and efficiency) from cortical thickness (62 regions) and hippocampal volume (18 subfields). Across disease stages, changes in MC were evaluated using a General Linear Model, while partial correlation matrices (correcting for age, gender, and total brain volume) compared shifts in the relationship between MC and cognitive performance. Preliminary analysis shows that, as disease progresses, MC is significantly impaired in certain regions (10/80 regions), and the correlation between MC and cognitive functions is weakened in specific regions (13/80 regions). These findings deepen our understanding of structural covariance-based MC as a hallmark of psychosis and its progression, offering insights into the cognitive decline in psychotic patients.