C-reactive protein polygenic risk is associated with obesity-related traits in schizophrenia spectrum disorders
Background Patients with schizophrenia spectrum disorders (SSDs) are at increased risk of cardiometabolic dysregulations, which substantially contribute to cardiovascular morbidity and reduced life expectancy. Chronic low-grade inflammation is a key factor in the development of cardiometabolic outcomes. Polygenic risk scores (PRS) for inflammatory biomarkers like for C-reactive protein (CRP) and interleukin 6 (IL-6) may help clarify the genetic contribution to this risk, yet evidence in SSDs populations remains limited. Methods We investigated the associations of PRSes for CRP and IL6 with cardiometabolic outcomes in 671 patients with SSDs from the longitudinal Dutch Genetic Risk and Outcome in Psychosis (GROUP) study. Seven PRSCRP and seven PRSIL-6 were constructed using clumping and threshold method at seven P-value thresholds (Pt_x) based on large, independent genome-wide association studies. We examined 11 cardiometabolic outcomes measured at three years after diagnosis, including body mass index (BMI), waist circumference (WC), lipid levels, blood pressures, glycaemic markers, and a metabolic composite score. Linear regression models adjusted for age, sex, and population substructure tested the associations between PRSes, with multiple testing correction and bootstrapping for validation. Regression coefficients (βP-threshold) and 95% confidence intervals and unadjusted R2 (variance explained) were reported. Sensitivity analyses were conducted by further adjusting for smoking and antipsychotic medication use. Results Higher standardized PRSCRP was significantly associated with increased BMI (βPt_0.5 = 0.64, 95%CI=0.21-1.02, Pbootstapping = 0.003) and WC (βPt_0.5 = 2.25, 1.00-3.53, Pbootstapping < 0.001), explaining up to 1.85% variance in BMI, and 2.52% in WC. Nominal associations were also observed between PRSCRP and triglycerides (TG) levels (βPt_0.2 = 0.13, 0.01-0.26, Pbootstapping = 0.036), and metabolic composite score (βPt_0.2 = 0.14, 0.04-0.24, Pbootstapping = 0.006), and between PRSIL-6 and HbA1c level (βPt_5e06=-0.66, -1.26 to -0.05, Pbootstapping = 0.033); however these associations did not remain significant after correction for multiple testing. Conclusions Higher genetic susceptibility for low grade inflammation as captured by PRSCRP is modestly but robustly associated with increased levels of obesity-related traits in SSDs independent from antipsychotics use. These results support CRP-related genetics pathways as potential contributors to risk of cardiometabolic vulnerability in SSDs and may inform genetic-based personalized risk stratification and prevention strategies in SSDs patients.