Background: Hemoglobin A1C’s accuracy and precision decrease in CKD, and there is an increased interest in using CGM as an alternative method for glycemic monitoring. However, most CGM data stems from type 1 or insulin-treated type 2 diabetes (T2D) patients with normal eGFR. We conducted a pilot study to evaluate CGM metrics in CKD, including patients with T2D not using insulin and those without diabetes.
Methods: We enrolled 10 CKD patients (6 with T2D and 4 without diabetes) who wore blinded CGM (DexCom G6) for 20 days. A1C was measured at the end. Discordance was defined as an absolute difference between A1C and glucose management indicator (GMI) >=0.5%.
Results: Participants had a mean age of 68 and mean eGFR was 35 mL/min/1.73m². In participants with T2D (A1C range: 5.4-6.9%), there was wide range of CGM metrics: range of mean glucose was 115-215mg/dL, GMI 6.1-8.4%, CV 16-29%, TIR 13-96%. In participants without diabetes (A1C range: 5.2-6.1%), the range of mean glucose was 117-158mg/dL, GMI 6.1-7.1%, CV 16-29%, time-in-tight-range (70-140mg/dL) 37-86%. For several participants with similar A1C values, glycemic patterns differed substantially. Seven (70%) participants had A1C-GMI discordance, with A1C underestimating GMI (Fig).
Conclusion: Among CKD patients, large variations in CGM metrics existed and A1C-GMI discordance was common. Our findings warrant further investigation into the prevalence and predictors of such discordance.
M. Tang: Research Support; Dexcom, Inc. I. de Boer: Consultant; Boehringer-Ingelheim, Lilly Diabetes, AstraZeneca, Novo Nordisk, Alnylam Pharmaceuticals, Inc., George Clinical. Research Support; Dexcom, Inc., Novo Nordisk. S. Kalim: Advisory Panel; Alnylam. Speaker's Bureau; Fresenius Kabi.
American Heart Association (23POST1010825); DexCom (IIS-2023-007)
