The association between C-reactive protein, mood disorder, and cognitive function in UK Biobank

Background Systemic inflammation has been linked with mood disorder and cognitive impairment. The extent of this relationship remains uncertain, with the effects of serum inflammatory biomarkers compared to genetic predisposition toward inflammation yet to be clearly established. Methods We investigated the magnitude of associations between C-reactive protein (CRP) measures, lifetime history of bipolar disorder or major depression, and cognitive function (reaction time and visuospatial memory) in 84,268 UK Biobank participants. CRP was measured in serum and a polygenic risk score for CRP was calculated, based on a published genome-wide association study. Multiple regression models adjusted for sociodemographic and clinical confounders. Results Increased serum CRP was significantly associated with mood disorder history (Kruskal-Wallis H = 196.06, p < 0.001, eta(2) = 0.002) but increased polygenic risk for CRP was not (F = 0.668, p = 0.648, eta(2) < 0.001). Compared to the lowest quintile, the highest serum CRP quintile was significantly associated with both negative and positive differences in cognitive performance (fully adjusted models: reaction time B = -0.030, 95% CI = -0.052, -0.008; visuospatial memory B = 0.066, 95% CI = 0.042, 0.089). More severe mood disorder categories were significantly associated with worse cognitive performance and this was not moderated by serum or genetic CRP level. Conclusions In this large cohort study, we found that measured inflammation was associated with mood disorder history, but genetic predisposition to inflammation was not. The association between mood disorder and worse cognitive performance was very small and did not vary by CRP level. The inconsistent relationship between CRP measures and cognitive performance warrants further study.