Bilirubin levels and kidney function decline: An analysis of clinical trial and real world data

Abstract

Objective To evaluate if previously found associations between low serum bilirubin concentration and kidney function decline is independent of hemoglobin and other key confounders. Research design and methods Clinical trial data from the SAVOR-TIMI 53 trial as well as the UK primary care electronic healthcare records, Clinical Practice Research Datalink (CPRD), were used to construct three cohorts of patients at risk of chronic kidney disease (CKD). The randomized clinical trial (RCT) cohort from the subset of SAVOR-TIMI 53 trial consisted of 10,555 type-2 diabetic patients with increased risk of cardiovascular disease. The two observational data cohorts from CPRD consisted of 71,104 newly diagnosed type-2 diabetes (CPRD-DM2) and 82,065 newly diagnosed hypertensive (CPRD-HT) patients without diabetes. Cohorts were stratified according to baseline circulating total bilirubin levels to determine association on the primary end point of a 30% reduction from baseline in estimated glomerular filtration rate (eGFR) and the secondary end point of albuminuria. Results The confounder adjusted hazard ratios of the subpopulation with lower than median bilirubin levels compared to above median bilirubin levels for the primary end point were 1.18 (1.02–1.37), 1.12 (1.05–1.19) and 1.09 (1.01–1.17), for the secondary end point were 1.26 (1.06–1.52), 1.11 (1.01–1.21) and 1.18 (1.01–1.39) for SAVOR-TIMI 53, CPRD-DM2, CPRD-HT, respectively. Conclusion Our findings are consistent across all cohorts and endpoints: lower serum bilirubin levels are associated with a greater kidney function decline independent of hemoglobin and other key confounders. This suggests that increased monitoring of kidney health in patients with lower bilirubin levels may be considered, especially for diabetic patients.