RELATIONSHIP BETWEEN SERUM URIC ACID AND CARDIOVASCULAR PHENOTYPES IN THE BRISIGHELLA COHORT: A MENDELIAN RANDOMIZATION ANALYSIS

Objective: Cardiovascular disease (CVD) represents the most common and lethal chronic disease worldwide. High levels of serum uric acid (SUA) have been associated with CVD and related risk factors. However, a causal link between SUA and CVD has not been proven yet. We tested the hypothesis about causal relationship between SUA levels and susceptibility to CVD using the Mendelian Randomization (MR) approach. Design and method: We characterised 24 established SUA-associated DNA variants (SNPs) enriched for associations with cardiometabolic traits in a sample of 1978 individuals from the Genetics in Brisighella Heart Study (GBHS). We evaluated associations between these SNPs and SUA, creatinine, body mass index, blood pressure, coronary artery disease (CAD), hypertension (HTN) and hypercholesterolemia (HTC). We used weighted SUA-genetic risk score (GRS) based on the 5 SNPs nominally (p < 0.05) associated with SUA (rs12498742, rs2231142, rs1165151, rs1471633, rs2078267) as instrumental variable (IV) in a MR analysis of SUA levels on eight CVD-related phenotypes and compared IV estimators to SUA-trait associations from the same individuals. Results: We observed the most significant association for SUA at SLC2A9 (p = 7.56x10–16), while ABCG2, SLC16A9, SLC17A1, PRKAG2 loci showed nominally significant effect. All other, but three, loci showed consistent with established ones directions of SUA increasing allelic effects. Among genotyped variants, SLC2A9, GCKR, BAZ1B and INHBC loci were nominally associated with creatinine levels; SLC17A1, ORC4L, TRIM46 with HTC and A1CF with BMI. Five top SUA-associated variant GRS showed significant effect on CREA (p = 0.005643). We demonstrate a novel evidence for causal relationship between SUA levels and creatinine in the IV analysis using weighted 5-variant GRS (p = 0.007). We don’t observe causal relationship for other CVD-related phenotypes studied. Conclusions: MR analysis within GBHS does not demonstrate any causal relationship of SUA with subtypes of CVD (CAD) or its established risk factors (HTC, HTN), similarly to study by Palmer et al. 2013. We describe a causal relationship between SUA and creatinine levels and propose that SUA may influence cardiovascular traits by acting on serum creatinine. In fact, it is established that a high level of creatinine, index of chronic kidney disease, may lead to CVD.