Abstract
Background: Schizophrenia is characterized by both psychotic illness and cognitive impairment, but it is unclear whether they represent related yet distinct disease processes. There is evidence to suggest dissociation. For example, cognitive impairment occurs in schizophrenia patients during both active psychosis and symptom remission. However, the shared or nonshared neural underpinnings of cognition and psychotic psychopathology are also unclear despite findings of multi-focal cortical thinning in the illness. Accordingly, this study sampled patients and controls with a broad range of cognitive ability to examine relations between cortical thickness and cognitive performance with and without the presence of psychotic illness. Our basic questions were: do regional thickness values primarily index the psychotic disease process or cognitive performance and to what extent do disease and performance interact? Methods: Cognitive functioning of patients diagnosed with schizophrenia or schizoaffective disorder (n = 61) and healthy controls (n = 40) were assessed with the MATRICS Consensus Cognitive Battery (MCCB). Neuroimaging data were obtained with a 3T General Electric System MRI scanner, and cortical thickness was calculated using Freesurfer. General linear models were conducted to examine relations and interactions between cortical thickness, diagnosis, and cognition. Results: Cortical thickness and cognitive performance on MCCB subscales and overall composite score were positively correlated in 34 brain regions, predominantly in the frontal, parietal, and temporal brain areas, irrespective of diagnostic status. Patients showed the same cortical thickness-cognitive performance relationship as controls, but had significantly reduced thickness in 27/34 of these regions despite similar behavioral performance. An interaction of diagnosis, cognition, and cortical thickness was found in the parahippocampal and left caudal middle frontal gyri only. Lastly, there were several regions of reduced cortical thickness among patients with no corresponding relationship to cognitive performance. Conclusion: These findings suggest that despite their high rates of co-occurrence, cognitive impairment and psychosis may be partially independent pathologies of the schizophrenia disease process. Cortical thickness varies with cognition in both schizophrenia patients and healthy controls, but remains significantly reduced in patients. This occurs even when cognitive performance is largely equalized between patients and controls. These findings are consistent with recent neurogenetic research linking liability to schizophrenia with cortical abnormalities including thinning, reduced synaptic structure and excessive pruning. The results point to the importance of studying cognition and psychotic symptoms as potentially separable processes that may also represent independent treatment targets.