Abstract
BACKGROUND: Cross-sectional studies have reported elevated concentrations of inflammatory markers in psychosis and depression. However, questions regarding temporality and specificity of association, crucial for understanding the potential role of inflammation, remain. METHODS: Based on 2224 ALSPAC birth cohort participants, we used regression analyses to test associations of interleukin-6 (IL-6) and C-reactive protein (CRP) levels at age 9 with risks for psychosis (psychotic experiences; negative symptoms; psychotic disorder), and depression (depressive episode; symptom score) at age 24. Regression models were adjusted for sex, ethnicity, social class and body mass index. We tested for linearity (using quadratic terms) and specificity (using bi-variate probit regression) of association, and used multiple imputation to explore the impact of missing data. RESULTS: After adjustments, higher IL-6 levels at age 9 were associated with increased risk of psychotic disorder (OR = 1.56; 95% C.I., 1.09-2.21 per SD increase in IL-6; OR=2.60; 95% C.I., 1.04-6.53 for the top compared with bottom third of IL-6) and depressive episode (OR = 1.14; 95% C.I., 0.99-1.32 per SD increase in IL-6; OR = 1.49; 95% C.I., 1.02-2.18 for the top compared with bottom third of IL-6). IL-6 was associated with negative symptoms after adjusting for depression (β = 0.09; 95% C.I., 0.01-0.22). There was no evidence for outcome-specific associations of IL-6. Childhood CRP was not associated with adult psychosis or depression. CONCLUSIONS: Evidence for similar, longitudinal, dose-response associations suggest that elevated childhood IL-6 could be a shared risk factor for adult psychosis and depression. The IL-6 pathway may represent a novel target for treatment and prevention of these disorders.