Assessment for Predictors of Rise in Hemoglobin A1c During Extended Use of a Closed-Loop Control System.

We assessed predictors of rising hemoglobin A1c (HbA1c) during long-term use of closed-loop control (CLC) in children aged 6-13 years with type 1 diabetes. Participants used a CLC system during a 16-week randomization phase followed by a 12-week extension phase. We compared an "Increased-HbA1c" group (n = 17, ≥0.5% rise in HbA1c between randomization and extension phases) to a "Maintained-Improvement" group (n = 18, had stable or improved HbA1c). The Increased-HbA1c group had higher pre-CLC HbA1c (8.42% ± 0.80 vs. 7.45% ± 0.93, P = 0.002). Contrary to a-priori hypotheses, there were no differences in Δ-height-for-age z-score, a surrogate for a pubertal growth spurt (+0.16 vs. -0.15, P = 0.113), or number of carbohydrate boluses per day, a surrogate for missed boluses (4.4 ± 2.2 vs. 5.2 ± 2.1, P = 0.263). Both groups maintained high rates in closed-loop. Thus, some children exhibit meaningful rise in HbA1c after initial CLC use, likely from multiple contributing factors, and may benefit from added encouragement during ongoing use.