Exploratory analysis of continuous glucose monitoring patterns in first degree relatives with dysglycemia at risk of progression to stage 3 type 1 diabetes

Introduction: Continuous glucose monitoring (CGM) can characterize subtle glucose changes in early-stage type 1 diabetes (T1D). We assessed whether CGM metrics predict Stage 3 T1D in first-degree relatives (FDRs) with positive islet autoantibodies (IAb) and dysglycemia participating in INNODIA. Methods: We analysed IAb-positive FDRs with dysglycemia and available CGM data. CGM metrics were analyzed longitudinally and in a restricted analysis limited to the last assessment before Stage 3 T1D in progressors. CGM metrics were compared using linear mixed models. Predictive performance of single and combined CGM metrics models was evaluated using ROC analyses. Results: Data from 27 participants (14 females (52%), median [IQR] age: 16.8 [11.9-39.8] years) were analyzed. Nine participants progressed to Stage 3 T1D after a median of 163 [129–244] days from CGM assessment. In the restricted analysis, progressors had higher mean glucose (132 vs. 111 mg/dL, P=0.001), time >140 mg/dL (36 vs 10%, P=0.003), and time >180 mg/dL between 9pm-2am (8.1% vs 1.0%, P=0.005), whereas time <70 mg/dL between 4-7am was lower (0% vs 1.2%, P=0.001). Time >140 mg/dl yielded an area under the curve (AUC) of 0.83 (95% CI 0.66–1.00). Combining time >140 mg/dL with time-specific percentages >180 and <70 mg/dL increased AUC to 0.93 (95% CI 0.84–1.00). A model incorporating time between 70–140 mg/dL, mean glucose, early-morning <70 mg/dL and MAGE achieved 100% sensitivity and 86% specificity. Conclusions: CGM may help identify FDRs with IAbs and dysglycemia at higher risk of progressing to Stage 3 T1D. Combining CGM metrics improved prediction compared with time >140 mg/dL alone.