Influence of erythropoiesis-stimulating agentson glycated hemoglobin in nondiabetic kidney diseases at the start of dialysis

Background: Although glycated hemoglobin (HbA 1C ) is apractical tool to assess long-term glucose control in the generalpopulation, it may underestimate glycemic control inchronic kidney disease (CKD) patients – especially those undergoingtreatment with erythropoiesis-stimulating agents(ESA). We evaluated the association of HbA 1C with other parametersof glucose homeostasis and tested its associationwith ESA use and mortality in nondiabetic incident dialysispatients. Methods: We studied 270 nondiabetic CKD stage 5patients referred to initiate dialysis therapy [median age: 54years (43–63), 154 males]. Patients were followed for up to 5years for survival analysis. Results: HbA 1C was positively correlatedwith age (Rho = 0.13; p = 0.031), C-reactive protein(Rho = 0.14; p = 0.024), total cholesterol (Rho = 0.19; p = 0.001),triglycerides (Rho = 0.21; p ! 0.001) and glucose (Rho = 0.21;p = 0.001), but it was negatively correlated with HDL-cholesterol(Rho = –0.22; p ! 0.001) and ESA dose (Rho = –0.27; p !0.001). Across increasing HbA 1C tertiles, increased glucoselevels and reduced use of ESA and dose of ESA were observed(p ! 0.001), but there were no differences in insulinand HOMA index. In a stepwise multivariate linear regressionanalysis, ESA dose was negatively associated with logHbA 1C .HbA 1C did not predict mortality. Conclusion: In nondiabeticCKD stage 5 patients, HbA 1C levels were associated with ESAdose. HbA 1C was not independently associated with surrogatemarkers of insulin resistance or mortality.