Aberrant modularity and lower resiliency of structural covariance network in first-episode antipsychotic-naive psychoses
Objective: Structural brain alterations are consistently observed in schizophrenia. However, divergent findings suggest that often-observed regions exist within a network of susceptible regions. We conducted structural covariance analysis of multiple morphometric features of 358 regions from 79 first-episode anti-psychotic-naive psychosis patients (FEAP) and 68 healthy controls to investigate network differences. Methods: Using graph theoretic methods, we investigated structural covariance network of Freesurfer-derived cortical gray matter volumes, thickness, curvature, and surface area using the Brain Connectivity Toolbox within MATLAB, compared network modularity using the Community Detection Toolbox within MATLAB, and examined the resilience of the network using simulated attacks. Results: FEAP showed decreased heterogeneity of cortical volumes compared to controls which was driven by decreased heterogeneity of cortical thickness but not surface area. Reduced morphological heterogeneity was associated with less differentiated community structure in FEAP compared to controls. FEAP patients, in general, showed less resilient networks that were more vulnerable to attacks on fewer nodes compared to healthy subjects. Conclusions: Our findings of decreased heterogeneity may be associated with FEAP-related pathology since the impact of illness chronicity and treatment are minimized. Contribution of cortical thickness but not surface area covariance network suggests that neurodevelopmental processes affecting the thickness rather than the surface area may be pathophysiologically more significant.