Glycated albumin, not hemoglobin A1c, predicts cardiovascular hospitalization and length of stay in diabetic patients on dialysis.

Background: The utility of glycated hemoglobin (HbA1c) and glycated albumin (GA) in diabetic dialysis patients remains unknown. GA was previously associated with all-cause hospitalization and patient survival. Relationships between GA, HbA1c, and casual plasma glucose (PG) with cause-specific cardiovascular (CV) disease, infectious disease (ID), and vascular access- (VA) related hospitalization rates and length of stay (LOS) were assessed.

Methods: 444 prevalent diabetic dialysis patients had monthly PG, quarterly GA, and all HbA1c values recorded for 2.33 years; hospitalizations within 17 and 30 days of testing were evaluated. Best-fit, time-dependent Cox models were constructed in unadjusted, case-mix-adjusted (age, sex, race, BMI, diabetes duration, dialysis vintage), and case-mix- plus lab-adjusted (hemoglobin, albumin, phosphorus) models.

Results: Mean ± SD diabetes duration was 18.5 ± 10.8 years and dialysis vintage 2.9 ± 2.6 years. In fully adjusted models, CV hospitalization rates were associated with increasing GA (HR 1.32; 95% CI 1.11-1.57; p = 0.002 at 17 days; HR 1.21; p = 0.02 at 30 days) and PG (HR 1.10; 95% CI 1.02-1.17; p = 0.01 at 17 days; HR 1.07; p = 0.03 at 30 days), not HbA1c (HR 1.24; 95% CI 0.89-1.73; p = 0.21 at 17 days; HR 1.26; p = 0.10 at 30 days). LOS for CV admissions was positively associated with GA (HR 1.18; 95% CI 1.01-1.39; p = 0.03), not PG (HR 1.04; 95% CI 0.99-1.10; p = 0.15) or HbA1c (HR 1.03; 95% CI 0.92-1.15; p = 0.21). Admissions due to ID and VA complications (and LOS) did not correlate with these assays.

Conclusions: Improved glycemic control based on GA and PG predicted CV-related hospitalizations; GA also predicted CV hospitalization LOS. HbA1c did not predict cause-specific hospitalizations in dialysis populations.