31. The Association Between Psychosis and Nonneurological Autoimmune Disorders: A Systematic Review and Meta-Analysis of the Evidence to Date

Abstract

Background: Epidemiological studies conducted over the past 50 years have observed an increased prevalence of psychosis among individuals with nonneurological autoimmune disorders relative to the general population, with the notable exception of rheumatoid arthritis. Whether this association refects shared risk factors (e.g., infections or genetic liability) or a causal relationship (mediated by infammation or brain-reactive antibod-ies) is yet to be determined. Quantifying the extent to which psychosis is associated with individual nonneurological autoimmune disorders may help to address this issue. A systematic review of studies investigating the comorbidity of psychosis and nonneurological autoimmune disorders was therefore conducted to clarify the nature of this association; meta-analyses were performed for individual autoimmune disorders where possible. Methods: PubMed, PsycINFO, and EMBASE were systematically searched, and supplemented with manual searching, to identify studies examining (1) the prevalence of psychotic disorders among individuals with nonneuro-logical autoimmune disorders, or (ii) the prevalence of nonneurological autoimmune disorders among individuals with psychotic disorders. Studies were assessed using the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist. Raw (unadjusted) data were extracted from studies and used to compute odds ratios (OR). Results: Fourteen studies were suitable for inclusion, the majority of which met STROBE criteria for methodology reporting. These studies provided useable data for 35 non-neurological autoimmune disorders, from which 60 individual effect sizes were computed. Overall, individuals with non-neurological autoimmune disorders were signifcantly more likely to have a psychotic disorder (OR = 1.51; 95% CI: 1.47-1.56; P