Abnormal Structural Connectivity Patterns in Large-Scale Brain Networks in Schizophrenia

Background: While cognitive impairment is a core feature of schizophrenia, a minority of patients demonstrate average to superior ability on many standard cognitive measures with no attenuation of the psychotic disease process (Heinrichs et al. 2008; Muharib et al., 2014). The data imply a dissociation of cognitive and psychosis-generating neural mechanisms whereby patients share a disease process that leads to psychosis but vary in terms of the pathophysiology that causes cognitive impairment. Furthermore, current views hold that schizophrenia involves abnormalities in the connectivity of large-scale brain networks [default mode (DMN), salience (SN), central executive (CEN), and social brain (SBN)]. However, these findings may reflect pathophysiology related to both the cognitive and psychotic features of schizophrenia. Therefore, we asked: Are aberrations in cortical thickness and/or structural connectivity within and between networks associated with cognitive impairment and/or the severity of psychotic psychopathology?

Method: Structural magnetic resonance (MRI) and diffusion tensor imaging (DTI), cognitive, and clinical data were collected from 121 participants, which include 16 cognitively-intact and 48 cognitively-impaired schizophrenia patients as well as 36 cognitively normal and 21 below-normal controls. Between-group comparisons and region-of-interest analyses of cortical thickness and structural integrity in the DMN, SN, CEN, and SBN were performed on MRI and DTI data.

Results: Cognitively normal controls had greater DMN and SN cortical thickness than both cognitively normal and below-normal patients. Structural integrity of the genu of the corpus callosum was significantly different between cognitively normal controls and both patient groups. Superior longitudinal fasciculus connectivity patterns differed between cognitively normal controls and below-normal patients. Lastly, the inferior longitudinal and inferior fronto-occipital fasciculi combined were significantly different between cognitively normal controls and patients.

Conclusions: The results suggest that cortical thinning may represent the presence of psychotic psychopathology independent of cognitive impairment. However, tract integrity may index cognitive status, the psychotic disease process, or both. The similarities in white matter integrity associations with cognition among cognitively normal patients and controls suggest shared neurocognitive processes, and the dissimilarities may point to cortical structure aberrations that give rise to psychotic psychopathology. Taken together, this study contributes to the advancement of the literature by providing evidence for dissociable or partially dissociable disease processes in psychotic illness.