Reinhard W. Holl, Germany
University of Ulm Institute of Epidemiology and Medical Biometry, ZIBMTPresenter of 1 Presentation
GLUCOSE MANAGEMENT INDICATOR (GMI) BASED ON SENSOR DATA AND LABORATORY HBA1C IN PEOPLE WITH TYPE 1 DIABETES FROM THE DPV DATABASE: DIFFERENCES BY SENSOR TYPE
Abstract
Background and Aims
The Glucose Management Indicator (GMI) was developed using data from real-time continuous glucose monitoring (rtCGM). We aimed to compare GMI and laboratory HbA1c using both rtCGM and intermittent scanning CGM (iscCGM) profiles collected during routine care in people with type 1 diabetes (T1D).
Methods
We analyzed 132,361 CGM days from N=1,973 individuals with T1D duration ≥1 year from the German/Austrian DPV Registry. As measurement ranges of CGM devices differ, we truncated glucose values to the same range (40-400 mg/dl). GMI was calculated from a median number of 77 [IQR: 46-89] days/individual as GMI (%) = 3.31+0.02392*[mean glucose, mg/dL]. Differences between GMI and laboratory HbA1c were illustrated using boxplots for rtCGM vs. iscCGM stratified by glucose variability (coefficient of variation [CV] </≥36%), normal weight vs. overweight, and HbA1c </≥7.5%.
Results
Mean GMI and HbA1c were similar in rtCGM users (n=405, 7.6±0.7% vs. 7.6±1.1%), whereas iscCGM users (n=1,568) had higher mean GMI than HbA1c (7.9±0.9% vs. 7.6±1.2%). Overall and stratified by glucose variability or weight, differences between GMI and laboratory HbA1c were almost symmetrically distributed around 0 in rtCGM users, whereas GMI was higher than HbA1c in almost three fourth of all iscCGM users. For both sensor types, most individuals with HbA1c <7.5% had higher GMI than HbA1c, whereas three fourth of the individuals with HbA1c ≥7.5% and rtCGM had lower GMI than HbA1c (Figure).
Conclusions
As measurement ranges, distributions of glucose values, and calibration methods differ between CGM sensors, it may be necessary to use formulas specific for sensor type to calculate GMI.