1925-P: High Discordance between A1C and CGM-Derived Glucometrics in Adults Receiving Dialysis

Introduction and Objective: A1C is difficult to interpret in adults receiving dialysis because of altered red blood cell turnover and treatment-related factors. Continuous glucose monitoring (CGM) provides complementary measures of glycemia, including glucose management indicator (GMI) and glycemia risk index (GRI). We quantified A1C-GMI discordance and evaluated relationships among A1C and CGM-derived metrics, including race/ethnicity and erythropoiesis-stimulating agent (ESA) associations. Methods: This cross-sectional analysis used data from the BLOSSOM study and included adults with kidney failure treated with maintenance hemodialysis or peritoneal dialysis who wore a blinded Dexcom G6 Pro CGM for ten days and had paired A1C and CGM data (N=259). Discordance was defined as the absolute difference between A1C and GMI. Pearson correlations assessed associations among glycemic metrics. Multivariable linear regression with robust standard errors adjusted for mean CGM glucose was used to assess factors associated with A1C bias. Results: Mean A1C was 6.8%, while mean GMI was higher at 8.0%. Median A1C-GMI discordance was 1.3%, with discordance of at least one percent in 64.5% and at least two percent in 17.0% of participants. Discordance was predominantly characterized by A1C values lower than corresponding GMI. Associations between A1C and CGM-derived metrics were strong (A1C-GMI r=0.85; A1C-GRI r=0.79; A1C-TIR r=−0.78), with high concordance among CGM metrics. Higher mean CGM glucose was associated with greater discordance, while discordance did not differ by race/ethnicity or ESA exposure. Conclusion: Marked discordance between A1C and CGM-derived glycemia is common in adults receiving dialysis, with A1C frequently underestimating CGM-measured glucose exposure. CGM-derived glucometrics, including glucose management indicator and risk-weighted measures such as glycemia risk index, provide complementary information for glycemic assessment in dialysis populations where A1C is biologically biased. Disclosure P. Panday: None. I. de Boer: Consultant; Current; Boehringer Ingelheim International GmbH, Novo Nordisk, GlaxoSmithKline plc., Lilly. Consultant; Ended; Dexcom, Inc. Research Support; Current; Novo Nordisk, Dexcom, Inc. Consultant; Ended; AstraZeneca. Consultant; Current; Roche Diabetes Care. Consultant; Ended; Lexicon Pharmaceuticals, Inc. L. Zelnick: None. I. Hirsch: Research Support; Current; MannKind Corporation, Sequel Tech. Consultant; Current; Abbott Diabetes, Roche Diabetes Care, Hagar.