Accurate evaluation of glycemia in patients receiving maintenance dialysis remains challenging because glycated hemoglobin (HbA1c) may show measurement bias in kidney failure. A prospective community-based cohort study published in Diabetes Care assessed the accuracy, variability, and covariate bias of three glycemic biomarkers compared with glycemia measured using continuous glucose monitoring (CGM). 

The study included 251 adults receiving maintenance dialysis, including individuals with and without diabetes. Participants wore a CGM device for 10 days. The analysis compared HbA1c, glycated albumin (GA), and fructosamine with CGM-derived mean glucose.

Participants (43% women; 63% with diabetes) contributed a median of 9.3 valid CGM days (interquartile range 8.5-9.4). Mean (standard deviation) values were 6.2% (1.4%) for HbA1c, 19.6% (6.3%) for GA, and 351 (99) µmol/L for fructosamine. Mean CGM glucose was 170 (63) mg/dL. Correlations with CGM glucose were 0.85 for HbA1c, 0.87 for GA, and 0.70 for fructosamine. In participants with diabetes, correlations were 0.84 for HbA1c, 0.84 for GA, and 0.64 for fructosamine.

Compared with CGM glucose, HbA1c values were influenced by erythropoiesis-stimulating agent dose, body mass index, hemoglobin, and serum albumin. Glycated albumin and fructosamine values were influenced by dialysis modality, dialysis duration, residual kidney function, and body mass index.  The findings indicate that HbA1c and glycated albumin correlate with CGM-derived average glucose in patients receiving maintenance dialysis, although clinical characteristics may influence biomarker interpretation.