Automated Insulin Delivery in Young Children with Type 1 Diabetes: A Systematic Review and Meta-Analysis.

Abstract

Automated insulin delivery (AID) in young children (<7 years) with type 1 diabetes has not yet been systematically evaluated. Web of Science, PubMed, Scopus, CENTRAL, and ClinicalTrials.gov were searched from inception to December 9, 2025. Studies reporting glycemic outcomes in children younger than 7 years were included in this meta-analysis. Studies not published in English were excluded. The primary outcome was the change in time-in-range (TIR; 70-180 mg/dL). Pooled estimates were calculated using random-effects models and expressed as mean changes (MCs) with 95% confidence intervals (CIs). This study included 30 studies (9 randomized controlled trials, 7 single-arm studies, and 14 cohort studies) involving 1,155 young children. AID systems were associated with a significant increase in TIR (MC, 9.88% [95% CI 9.14 to 10.62], I2 = 8%, p < 0.0001, moderate certainty), equivalent to 2.37 h/day. Favorable improvement was also observed during the daytime (MC, 6.88% [95% CI 5.70 to 8.07]) and overnight (MC, 16.85% [95% CI 13.48 to 20.22]). TIR improvements were comparable across devices: MiniMed 670G (9.65%), MiniMed 780G (10.04%), CamAPS FX (10.58%), Control-IQ (9.51%), Omnipod 5 (10.25%), and Open Source (6.77%). AID use was also associated with reduced hyperglycemia exposure and modest reductions in glycated hemoglobin, without significant changes in hypoglycemia exposure. Episodes of severe hypoglycemia and diabetic ketoacidosis were infrequent. This systematic review with meta-analysis revealed that AID systems have greater benefits for improving glycemic outcomes and have good safety profiles in young children with type 1 diabetes.