Effects of automated insulin delivery systems on glucose control in subgroups of adults with type 1 diabetes in clinical practice over 2 years in Sweden.

Several studies have demonstrated improvements in glucose control after initiation of automated insulin delivery (AID). This study aimed to evaluate glycemic outcomes and treatment-related barriers in subgroups of adults with type 1 diabetes (T1D) using AID systems in routine clinical practice. We analyzed data from 142 adults with T1D using Tandem Control IQ or MiniMed 780G. Correlation analyses were conducted to evaluate how changes in glycemic outcomes varied in relation to baseline characteristics. Mean age was 42 years, 40.8% were females, and median follow-up time was 1.7 years. Individuals with higher hemoglobin A1c (HbA1c) had less time below range (TBR) before starting AID (P < 0.001). Effects on glucose control were similar irrespective of sex, prior insulin therapy, body mass index, and diabetes duration, but varied in relation to glucose control. Following AID initiation, the estimated reduction in HbA1c was 1.3% (14.3 mmol/mol) in individuals with a baseline HbA1c of 8.6% (70 mmol/mol), compared to 0.3% (2.9 mmol/mol) in those with a baseline HbA1c of 6.7% (50 mmol/mol) (P < 0.001). Among individuals with a baseline TIR of 50% and 70%, the estimated increase in TIR following AID initiation was 19.0% and 7.2%, respectively (P < 0.001). Greater reductions in HbA1c were associated with smaller reductions in TBR and vice versa (r = - 0.29, P = 0.003). Among MiniMed 780G users, older age correlated with greater improvements in TIR (P = 0.005). Skin reactions tended to be more frequent in women (41%) than men (26%), P = 0.057. In conclusion, AID system use was associated with substantial reductions in HbA1c in individuals with poor glycemic control. Clinically meaningful reductions in HbA1c were achieved also in individuals with an HbA1c closer to target who additionally gained greater reductions in TBR.