P.095 Increased peripheral inflammation in schizophrenia is associated with worse cognitive performance and related cortical thickness reductions.

Abstract

Background: Schizophrenia is characterised by positive, negative, and cognitive symptoms and neuropathology, such as cortical thinning. While the pathogenesis of schizophrenia remains unclear, increasing evidence implicates the immune system and inflammation. No schizophrenia biomarkers currently exist and the potential contribution of peripheral inflammation is yet to be determined. C-Reactive protein (CRP) is widely accepted as a marker of peripheral inflammation, which is increased in schizophrenia and is associated with poorer cognitive performance. Further, a recent study in a smaller cohort of schizophrenia patients and controls found that CRP was associated with reduced cortical thickness. Aim(s): In a large case-control study, we addressed three aims: 1. to confirm that schizophrenia cases have increased peripheral CRP levels compared to healthy controls 2. to determine if, and which, cognitive abilities were related to peripheral CRP levels in schizophrenia and controls 3. to determine if cortical thickness reductions in brain regions associated with cognitive deficits are also predicted by peripheral CRP levels Methods: Concentration of CRP in blood serum was determined using high sensitivity ELISA assays in the Australian Schizophrenia Research Bank cohort of 499 schizophrenia patients and 644 healthy controls who also had neuropsychological testing using the Repeatable Battery for Neuropsychological Status (RBANS). Structural MRI scans were obtained from a subset of 194 schizophrenia cases and 99 controls and cortical thickness data were extracted using FreeSurfer. We statistically controlled for multiple comparisons, using False Discovery Rate (FDR), and covariates such as age, sex and education. Result(s): CRP was significantly higher in schizophrenia cases than controls (p<0.0001). Worse performance on the RBANS attention domain was significantly correlated with higher CRP in schizophrenia (r=-0.15 p=0.001), but not controls (r=-0.053, p=0.181). There were no other significant correlations between CRP levels and the 5 other cognitive domains. We then determined if CRP predicted thinner grey matter in 11 regions comprising the prefrontal and cingulate cortices likely involved with attention, using backward linear regression, including CRP, age, diagnosis and sex as predictors. Higher CRP predicted thinner medial orbitofrontal, pars opercularis and posterior cingulate cortex; with pars opercularis and posterior cingulate remaining significant after FDR correction. In the same regression models, the diagnosis of schizophrenia predicted medial orbitofrontal and pars opercularis but not posterior cingulate cortex thinning. In all other prefrontal and cingulate cortex regions, CRP was not a significant predictor of thickness although, as expected, schizophrenia diagnosis was predictive of cortical thickness reductions (all p's<0.003 survived FDR) except in the caudal anterior cingulate cortex (p=0.083). Conclusion(s): In one of the largest cohort studies to date, we confirm that CRP is increased in schizophrenia and find that CRP is inversely correlated with performance on the RBANS attention score in schizophrenia. Our analysis suggests CRP predicts cortical thinning in regions of the prefrontal and cingulate cortices, which may be involved in attention deficits and thus warrants further investigation. Together these findings suggest CRP may be a contributing factor to the cognitive deficits of schizophrenia and imply its potential use as a peripheral biomarker.Copyright © 2019