LAUSR

Metabolic acidosis is a common complication of CKD and may contribute to CKD progression. However, the association between serum bicarbonate and CKD progression has not yet been investigated using longitudinal measurements. We used data from CKD-REIN, a prospective cohort study conducted in 40 representative French nephrology clinics which included 3033 patients with CKD stages 3 to 5 between 2013 and 2016. Patients who had no serum bicarbonate or no creatinine measure at within 6 months of their inclusion date in the CKD-REIN study were excluded. All serum bicarbonate measurements were taken into account, from inclusion to the occurrence of renal replacement therapy (RRT), death, or end of follow-up, whichever came first. We used a shared random-effect model for the joint analysis of individual trajectories of serum bicarbonate and the hazard of the composite outcome (RRT or death before RRT) and each specific event separately. Hazard ratio were adjusted (aHR) for age, sex, primary kidney disease, metabolic syndrome, cardiovascular disease, proteinuria (< 30, 30-300, > 300 mg/day), and the CKD-EPI estimated glomerular filtration rate (eGFR) at baseline. A total of 2977 patients (65.4% men, median age 69 years) were included. At baseline, the median eGFR was 31.8 mL/min/1.73m2 and the median bicarbonate concentration was 25 mmol/L. During a median follow-up of 4.5 years, 615 patients received RRT and 378 died before RRT. At any time of follow-up, a decrease of 1 mmol/l of the current level of serum bicarbonate was associated with a significantly increased hazard of RRT or death before RRT (aHR 1.05, 95%CI 1.02-1.08). The estimated effect was stronger on death before RRT (aHR 1.05, 95% CI 1.00-1.09) than on RRT (aHR 1.02, 95% CI 0.99-1.06). A lower current level of serum bicarbonate is associated with an increased mortality in patients with CKD stage 3 to 5.