Comparison of Glycemic Markers in Chronic Hemodialysis Using Continuous Glucose Monitoring.

BACKGROUND: Glycated hemoglobin A1c (HbA1c) among diabetic hemodialysis patients continues to be the standard of care, although its limitations are well recognized. This study evaluated glycated albumin (GA) and glycated serum protein (GSP) as alternatives to HbA1c in detecting glycemic control among diabetic hemodialysis patients using continuous-glucose-monitoring (CGM)-derived glucose as reference standard.

METHODS: A CGM system (iPRO) was applied for 7 days in 37 diabetic hemodialysis patients to determine glycemic control. The accuracy of GA and GSP versus HbA1c in detecting a 7-day average glucose ≥184 mg/dL was evaluated via receiver-operating-characteristic (ROC) analysis.

RESULTS: CGM-derived glucose exhibited strong correlation (r = 0.970, p < 0.001) and acceptable agreement with corresponding capillary glucose measurements obtained by the patients themselves in 1,169 time-points over the 7-day-long CGM. The area under ROC curve (AUC) for GA, GSP, and HbA1c to detect poor glycemic control was 0.976 (0.862-1.000), 0.682 (0.502-0.862), and 0.776 (0.629-0.923) respectively. GA levels >20.3% had 90.9% sensitivity and 96.1% specificity in detecting a 7-day average glucose ≥184 mg/dL. The AUC for GA was significantly higher than the AUC for GSP (difference between areas: 0.294, p < 0.001) and the AUC for HbA1c (difference between areas: 0.199, p < 0.01).

CONCLUSION: Among diabetic hemodialysis patients, GA is a stronger indicator of poor glycemic control assessed with 7-day-long CGM when compared to GSP and HbA1c.