Comparison of glycated albumin and hemoglobin A1c concentrations in diabetic subjects on peritoneal and hemodialysis.

Background: Relative to hemoglobin A(1c) (HbA(1c)), percentage of glycated albumin (GA%) more accurately reflects recent glycemic control in diabetic hemodialysis (HD) patients.

Methods: To determine the accuracy of glycemic assays in a larger sample including patients on peritoneal dialysis (PD), HbA(1c) and GA% were measured in 519 diabetic subjects: 55 on PD, 415 on HD, and 49 non-nephropathy controls.

Results: Mean +/- SD serum glucose levels were higher in HD and PD patients relative to non-nephropathy controls (HD 169.7 +/- 62 mg/dL, PD 168.6 +/- 66 mg/dL, controls 146.1 +/- 66 mg/dL; p = 0.03 HD vs controls, p = 0.13 PD vs controls). GA% was also higher in HD and PD patients (HD 20.6% +/- 8.0%, PD 19.0% +/- 5.7%, controls 15.7% +/- 7.7%; p < 0.02 HD vs controls and PD vs controls). HbA(1c) was paradoxically lower in dialysis patients (HD 6.78% +/- 1.6%, PD 6.87% +/- 1.4%, controls 7.3% +/- 1.4%; p = 0.03 HD vs controls, p = 0.12 PD vs controls). The serum glucose/HbA(1c) ratio differed significantly between dialysis patients and controls (p < 0.0001 HD vs controls, p = 0.002 PD vs controls), while serum glucose/GA% ratio was similar across groups (p = 0.96 HD vs controls, p = 0.64 PD vs controls). In best-fit multivariate models with HbA(1c) or GA% as outcome variable, dialysis status was a significant predictor of HbA(1c) but not GA%.

Conclusions: The relationship between HbA(1c) and GA% differs in diabetic patients with end-stage renal disease who perform either PD or HD compared to those without nephropathy. HbA(1c) significantly underestimates glycemic control in peritoneal and hemodialysis patients relative to GA%.