Glycated hemoglobin (A1c) is the gold-standard metric of average glycemia and quality of glycemic control, and a key predictor of long-term diabetes complications. We present a method for tracking A1c from continuous glucose monitoring (CGM)-derived time-in-range (TIR), which provides daily estimates of A1c (eA1c). The method uses a dynamical model of hemoglobin glycation/clearance and is calibrated with a single lab A1c value. CGM data collected by the DIAMOND study from 304 individuals with type 1 and type 2 diabetes (T1D/T2D) over 24 weeks were used in the analysis. Lab A1c samples were collected at weeks 12 and 24. Three population parameters were estimated by training the model on data from 50/50 randomly selected T1D/T2D subjects; the remaining 103/101 subjects were used for model validation. One individual glycation parameter was estimated from the week 12 lab A1c for each subject (model calibration), and was fixed thereafter. The week 24 lab A1c samples were used to test model performance (Table 1). The TIR-A1c relationship is mediated by glycation parameters, 3 population and 1 personal, the latter estimated using a single lab A1c value. Following this personalization, TIR provides accurate estimates of A1c in both T1D and T2D for (at least) 3 months post calibration. Thus, CGM-derived TIR can be a viable clinical alternative to A1c, allowing daily tracking of changes in average glycemia via a dynamical model of hemoglobin glycation.


C. Fabris: Consultant; Self; Abbott. R. Beck: None. B. Kovatchev: Advisory Panel; Self; Dexcom, Inc. Research Support; Self; Dexcom, Inc., Tandem Diabetes Care. Speaker’s Bureau; Self; Dexcom, Inc., Sanofi, Tandem Diabetes Care. Other Relationship; Self; Dexcom, Inc., Johnson & Johnson, Sanofi.


University of Virginia Strategic Investment in Diabetes (88); Dexcom, Inc.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at