Impact of Early Initiation of Automated Insulin Delivery on Glycemia Two Years from Type 1 Diabetes Diagnosis

Background and Aims: Recently, our clinical trial of youth randomized shortly after T1D diagnosis to intensive diabetes management with automated insulin delivery (AID) reported mean time in range (TIR) of 78% at 52 weeks. This analysis investigated whether high TIR persisted over 104 weeks. Methods: Participants were initially randomly assigned to intensive care or a control group receiving standard care through 52 weeks. During this randomized period, the intensive care group was intensively managed with AID and frequent contacts and the control group used CGM with injections or pumps without AID. At 52 weeks all participants could select their insulin delivery modality, and diabetes management returned to clinical teams. This secondary analysis compared TIR between groups in this observational extension phase (52- to 104-weeks post-diagnosis). Results: Thirty-nine of 61 RCT intensive care and 30 of 52 control participants enrolled in the extension and provided 104-week CGM data. Three RCT intensive care participants discontinued AID and 25 control participants initiated AID. From 52 to 104 weeks, mean %TIR remained stable (65% ± 21% to 65 ± 15%) in the RCT control group and declined in the RCT intensive care (79 ± 13% to 67 ± 15%) with no significant between-group difference at 104 weeks (mean difference 4% (95% CI: -5%, 13%) P=0.72). Forty-nine percent of RCT intensive care and 40% of control participants had TIR >70% at 104-weeks (P=0.73). Mean %time <54 mg/dl remained very low (<0.4%) in both groups at 104 weeks. Mean hemoglobin A1c was 6.7% ± 0.7 in the RCT intensive care group and 7.2% ± 1.5 in the controls at 104 weeks (mean difference -0.4 (-1.1, 0.4) P=0.72). Conclusion: Very high mean TIR at 52-weeks in youth randomized to intensive care with AID soon after T1D diagnosis waned during the 52-week observation period after the randomized trial.