Continuous Glucose Monitoring Predicts Progression to Diabetes in Autoantibody-Positive Children

Abstract

Children with islet autoantibodies (Ab+) are at high risk for developing clinical type 1 diabetes (T1D) and need accurate surveillance for prediction and diagnosis of T1D. Monitoring typically utilizes repeated 2-hour oral glucose tolerance tests. The objective of this study was to explore whether continuous glucose monitoring (CGM) could be an alternative measure for T1D prediction. In the Diabetes Autoimmunity Study in the Young (DAISY), 23 Ab+ subjects with a baseline CGM were followed for development of T1D for a mean of 28 months (8 of these progressed to T1D at a mean age of 15.2 years). Compared to non-progressors, subjects who progressed to T1D had significantly increased glycemic variability (SD 29 vs. 21 mg/dl, p=0.047), daytime sensor average (122 vs. 1mg/dl, p=0.02), and daytime sensor AUC (p=0.047). They spent 24% of time above 140 mg/dl and 12% above 160 mg/dl compared to, respectively, 8% and 3% (both p=0.005) for non-progressors; time spent over 200 mg/dl was minimal and not different between groups. Performance of prediction by receiver operating curve (ROC) analyses showed AUC of 0.85 for both %time spent above 140 mg/dl and 160 mg/dl. The cutoff of 18% time spent above 140 mg/dl had 100% specificity and 100% PPV with good sensitivity for diabetes prediction (Table). In conclusion, ≥18% CGM time spent above 140 mg/dl accurately predicts progression to clinical diabetes in islet Ab+ children.TableIndexCutoffPPVNPVSensitivitySpecificity% time > 12036%0.7140.8120.6250.867% time > 14018%1.0000.8830.7501.000% time > 1606%0.8330.8240.6250.933% time > 1802%0.7140.8130.6250.867% time > 2000.2%0.5460.8130.7500.667 Disclosure A. Steck: None. F. Dong: None. I. Taki: None. M.R. Hoffman: None. K. Simmons: None. B.I. Frohnert: None. M. Rewers: None.