Glycated albumin as an indicator of glycemic control in pre-dialysis diabetic chronic kidney disease patients

Abstract

Background Good glycemic control in diabetic patients is important to prevent progression of diabetic nephropathy and reduce cardiovascular events. Recently, several studies suggested that hemoglobin A1c (HbA1c), a widely used glycemic control marker, has serious limitations by falsely underestimating glycemic state in diabetic patients on hemodialysis. In contrast, glycated albumin (GA), which is not influenced by diseases of shortened RBC life span, and use of iron supplements and erythropoietin (EPO) therapy, is thought to more accurately reflect glycemic state in ESRD patients. Study aim was to validate this finding in pre-dialysis diabetic chronic kidney disease (CKD) patients.Methods Clinically stable pre-dialysis type 2 diabetic patients were enrolled between March 2009 and August 2012. A total of 497 patients were enrolled and stratified into 5 groups according to 2012 KDIGO CKD guideline that subdivided stage 3 into 3a and 3b stages. Parameters of glycemic control were investigated along with other biochemical and clinical informations.Results The numbers of patients according to CKD stages 1 to 4-5 were consisted of 168, 151, 76, 47, and 55 subjects, respectively. The mean serum glucose concentrations did not differ significantly among 5 groups. The HbA1c and GA showed positive correlations at all CKD stages, however, increased steepness in slope of regression line between HbA1c and GA with decline in renal function was shown (control group vs. other CKD stages: p < 0.05). The GA/HbA1c and serum glucose/HbA1c in CKD stage 3b and 4-5 were significantly higher than in controls, and the GA/HbA1c increased (r = 0.22, p < 0.001) with progression of renal failure. In contrast, the GA and glucose/GA remained constant throughout the all CKD stages. In multivariate analysis, weekly erythropoietin dose (p = 0.02) was associated with HbA1c and, the cut off value in ROC curve of weekly EPO dose altered HbA1c level was 6000 U/week (AUC = 0.91, 95% CI 0.78 - 1.00).Conclusion Our results indicate that the HbA1c may underestimate glycemic control state even in pre-dialysis diabetic CKD patients, especially those who are on EPO treatment and the GA might be an useful indicator of glycemic control in pre-dialysis diabetic CKD patients.