Closed-loop insulin system improves T1D management in pregnancy

For pregnant women with type 1 diabetes (T1D), the use of a closed-loop insulin system helps increase time spent in target glucose range, according to the open-label CIRCUIT trial.

Between 16 and 34 weeks of gestation, women on closed-loop therapy spent 65.4 percent of time in the pregnancy-specific glucose range of 63–140 mg/dL. In contrast, those who received standard care spent 50.3 percent of the time in this range. [JAMA 2025;doi:10.1001/jama.2025.19578]

The mean adjusted between-group difference of 12.5 percentage points (95 percent CI, 9.5–15.6) was equivalent to 3 more hours per day spent in the range (p<0.001), which the investigators described as “clinically relevant.”

The increase in time spent in target glucose range was accompanied by less time below 63 mg/dL (1.3 percent vs 2.1 percent; adjusted treatment difference, –1.0 percentage point, 95 percent CI, –1.5 to –0.5) and less time above 140 mg/dL (33.3 percent vs 47.6 percent; adjusted treatment difference, –11.5 percentage points, 95 percent CI, –14.6 to –8.3).

Additionally, closed-loop therapy was associated with lower mean glucose (131.3 vs 148.4 mg/dL; adjusted mean difference, –10.7 mg/dL, 95 percent CI, –15.4 to –6.0) and less glycaemic variability (29.8 percent vs 35.1 percent; adjusted mean difference, –4.9 percent, 95 percent CI, –6.5 to –3.2).  

“Glycaemic improvements were found across all sites, baseline HbA_1c ranges, and modes of insulin delivery at enrolment,” the investigators noted.

In terms of safety, one episode of severe hypoglycaemia occurred in the closed-loop group, and three episodes of diabetic ketoacidosis were documented overall in the closed-loop (n=2) and standard care (n=1) groups. Seven device-related adverse events were reported in the closed-loop group, none of which resulted in a serious adverse event.

Opportunities for improvement

“The importance of the CIRCUIT study should be underscored,” said Dr Sarit Polsky from Barbara Davis Center for Diabetes and University of Colorado, Anschutz Medical Campus in Aurora, Colorado, US, in a linked editorial. [JAMA 2025;doi:10.1001/jama.2025.19201]

Polsky placed CIRCUIT within a larger context where every research is fundamental to ensuring the safest possible pregnancies for women with T1D. However, she noted that the findings of the study, while “encouraging,” is not “comprehensive.”

Treatment with the closed-loop insulin system took place at the end or after organogenesis, at an average of 10.1–12.4 weeks of gestation, although glycaemic levels by 10 weeks’ gestation predict large-for-gestational-age newborns, according to Polsky.

“Initiation of automated insulin delivery earlier in gestation could approach the potential maximum effects of optimizing time in the glucose range for pregnancy. Customization of automated insulin delivery systems to pregnancy-specific shifts in insulin requirements, not just inclusive of fasting glucose targets, may increase time in the glucose range for pregnancy by approximately an additional 10 percent,” she said.

Polsky also noted that the closed-loop systems studied so far “required high levels of user and clinician engagement” and emphasized that future systems need to simplify operation to be less burdensome for both the patient and the healthcare teams.

“Continued work is still needed to optimize automated insulin delivery systems for specific use in pregnant individuals with T1D and to ensure access to diabetes technologies to improve the health of both pregnant individuals with T1D and their infants,” she said.

CIRCUIT

CIRCUIT included 91 pregnant women (mean age 31.7 years, early pregnancy HbA1c 7.4 percent) with T1D who were recruited before 14 weeks of gestation and followed-up until 6 weeks postpartum. These women were randomly assigned to receive closed-loop therapy (n=46) or standard care comprising insulin delivery method used prior to randomization (n=45) with continuous glucose monitoring. Treatment commenced by week 16 of gestation.