Automated Insulin Delivery in Women with Pregnancy Complicated by Type 1 Diabetes. AiDAPT triaN Engl J Med Oct 5



Hybrid closed-loop insulin therapy has shown promise for management of type 1 diabetes during pregnancy; however, its efficacy is unclear.


In this multicenter, controlled trial, we randomly assigned pregnant women with type 1 diabetes and a glycated hemoglobin level of at least 6.5% at nine sites in the United Kingdom to receive standard insulin therapy or hybrid closed-loop therapy, with both groups using continuous glucose monitoring.

The primary outcome was the percentage  of  time in  the  pregnancy-specific  target glucose  range  (63 to  140  mg per  deciliter  [3.5 to  7.8  mmol  per liter])  as  measured by  continuous  glucose monitoring from 16 weeks’ gestation until delivery.

Analyses were performed ac-ording to the intention-to-treat principle. Key secondary outcomes were the per-centage of time spent in a hyperglycemic state (glucose level >140 mg per deciliter), overnight time in the target range, the glycated hemoglobin level, and safety events


A total of 124 participants with a mean (±SD) age of 31.1±5.3 years and a mean baseline glycated  hemoglobin  level of  7.7±1.2%  underwent randomization.  

The  mean percentage of time that the maternal glucose level was in the target range was  68.2±10.5%  in the  closed-loop  group and  55.6±12.5%  in the  standard-care  group (mean adjusted difference, 10.5 percentage points; 95% confidence interval [CI], 7.0 to 14.0; P<0.001). 

Results for the secondary outcomes were consistent with those  of  the primary  outcome;  participants in  the  closed-loop group  spent  less time in a hyperglycemic state than those in the standard-care group (difference, −10.2 percentage points; 95% CI, −13.8 to −6.6); had more overnight time in the target  range  (difference, 12.3  percentage  points; 95%  CI,  8.3 to  16.2),  and had  lower  glycated hemoglobin  levels  (difference, −0.31  percentage  points; 95%  CI,  −0.50 to −0.12). 

Little time was spent in a hypoglycemic state. 

No unanticipated safety problems associated with the use of closed-loop therapy during pregnancy occurred (6 instances of severe hypoglycemia, vs. 5 in the standard-care group; 1 instance  of diabetic  ketoacidosis  in each  group;  and 12  device-related  adverse events in the closed-loop group, 7 related to closed-loop therapy)


Hybrid closed-loop therapy significantly improved maternal glycemic control during pregnancy complicated by type 1 diabetes. 


Editorial in summary

In addition, major barriers  exist  to implementing  use  of a  closed-loop system or continuous glucose monitoring, including biases related to socioeconomic status, ethnicity, and implicit racial bias. 

Clearly, closed-loop systems have changed the landscape of diabetes care in nonpregnant populations. Although more studies are needed, the AiDAPT trial provides  hope  that this  landscape  may also  be  altered for the better for pregnant persons with type 1 diabetes.


From the article


However, despite advancements in insulin therapy, continuous glucose monitoring, and high motivation  among pregnant  persons  to manage  their diabetes, most pregnant persons with diabetes do not have glucose levels in the pregnancy-specific  glucose  target range  of  63 to  140  mg  per deciliter (3.5 to 7.8 mmol per liter), which is lower  than  the target  range  of 70  to  180 mg  per  deciliter (3.9 to 10.0 mmol per liter) for nonpreg-nant persons.

The  use  of hybrid  closed-loop  therapy is  associated with improved glucose control in non-pregnant adults  and  in children, but whether  the  more stringent  glucose  targets required  for  optimal pregnancy  outcomes  can be  achieved  with this therapy is unknown. 

The CamAPS FX is a hybrid closed-loop system that enables automatically adjusted insulin delivery from an insulin pump according to real-time glucose-sensor measurements. This  system  was  approved  for use during pregnancy in the United Kingdom on the  basis of  results  from two  feasibility  stud-ies.

Subsequently, the  system  was updated,  leading to two key changes: first, glucose measurements  from  continuous glucose  monitors  can now  be  used to  inform  user-initiated  pre-meal boluses of insulin; second, additional features allow the user to intensify or relax closed-loop insulin delivery and to specify personalized glucose  targets,  which the  user  can adjust  dur-ing pregnancy. 

We tested whether hybrid closed-loop therapy initiated before 16 weeks’ gestation would improve  maternal  glucose levels  during  pregnancy complicated by type 1 diabetes.



We found that the percentage of time that glucose  levels were  within  the pregnancy-specific  target range of 63 to 140 mg per deciliter from 16 weeks’ gestation until delivery was 10.5 percents points  higher, an additional  2.5  hours per day, among participants assigned to closed-loop  therapy than among  those assigned  to  continuous glucose  monitoring  alongside their  usual insulin-delivery method. 

The time-in-range benefits were achieved by a reduction in maternal hyperglycemia and an increase in nocturnal time  in the  target  range. Improvements  in  glucose outcomes  were  consistent across  baseline  maternal age, glycated hemoglobin levels, clinical  sites,  and pretrial  insulin-delivery  method. Furthermore,  there  was  no  increase in gestational  weight gain  or  maternal insulin  doses  with closed-loop therapy. 

The incidence of hypoglycemia was low at baseline, and apart from a lower incidence of  nighttime   hypoglycemia events in  the  closed-loop group than  in the  standard-care  group, did not  differ between  groups.  

Among patients  receiving  closed-loop therapy, an  increase of  5  percentage points  in  the time in the target range was apparent by the end  of  the first  trimester,  which suggests  that  the benefits  occurred  soon after  initiation  of closed-loop therapy (which occurred at approxi-mately  12 weeks’  gestation);  this time  frame  is crucially  important  for women  and  clinicians considering  therapeutic  changes during  early  pregnancy.

The trial was conducted during the Covid-19 pandemic, which particularly affected pregnant persons, and necessitated rapid implementation of virtual training and visits. Nonetheless, use of the closed-loop system was high (>95%) through-out pregnancy and without apparent safety prob-lems, including  among  participants who  were  new to  insulin-pump  therapy. 

Participants who  continued  standard care  had  more clinic  visits  and more unscheduled contacts, which suggests that  beyond  initial training,  use  of the  closed-loop  system did  not  require additional  input  from health care professionals.

Recent trials  have  shown the  benefits  of closed-loop therapy in persons with newly diagnosed  type 1  diabetes  and young  children,  and these  results  extend the  evidence  to pregnant  women.

Alongside the  participants’ motivation to minimize pregnancy complications, closed-loop therapy  facilitated  attainment of  glucose  levels in the pregnancy-specific target range 70% of the time. 

Given the rapid increase in the time in the target range observed within 1 week after initiation of  therapy  in this  trial,  and within  1 day in a recent trial,we speculate that further benefits may be obtained from starting closed-loop  therapy before  pregnancy  or as  soon  as possible  after  pregnancy is  confirmed.  

Participants were offered the option to continue closed-loop therapy during the inpatient admission for labor and delivery (results not reported here).

The participants in the current trial gained an additional  10  percentage points  of  time in  the  target range  above  the 10  percentage-point  increment seen with continuous glucose monitor-ing  and  standard insulin  therapy  during preg-nancy.  Previous  studies have  shown  that every  increase  of 5  percentage  points of  the  time in  the  target range  is  associated with  improved  obstetrical and  neonatal  outcomes.

Our  trial  was not  powered  for pregnancy  outcomes,  but  we infer that this additional 10 percentage points of  time in  the  pregnancy-specific  target range  would  be expected  to  have additional  health  benefits for pregnant persons and their babies.

Strengths of our trial include its randomized, controlled  design;  the generalizability  of  our population,  which  included participants  who  had not previously received insulin-pump therapy; a large percentage of participants who initiated  therapy during  the  first trimester;  and  a  flexible  trial protocol  that  facilitated virtual  or  in-person visits.  

We observed  no  increase in clinical contacts, which is often observed in trials of investigational devices.

This trial had certain limitations. The sample size did not provide definitive data on maternal and neonatal health outcomes. Most of the participants  (93%)  were white, participants  were excluded if  they did not  have a  glycated  hemoglobin level  of 10% or  less  by  the  time of  randomization, and 56% of the participants had an undergraduate or  equivalent  education. 

First-trimester data were limited because participants underwent  randomization  at a median of 11  weeks’  gestation. We  did  not record  the  use of  the CamAPS Boost and Ease-Off features of the closed-loop system, and data cannot be extrapolated to systems with higher glucose-level targets.In this trial, closed-loop therapy was effective during pregnancy complicated by type 1 diabetes, accounting  for the  marked  gestational changes  in  insulin doses  in  trial participants,  and  pro-vided a clinical advantage beyond that achieved with  continuous  glucose monitoring  and  insu-lin-pump therapy.  

These  results support  the in Pregnancy recommendations,  proposed in  the  guideline from  the  National Institute  for  Health Care  Excellence (NICE), that hybrid  closed-loop  therapy should be offered to all pregnant persons with type 1 diabetes.


På amerikanska diabetesmötet ADA presenterades studien i juni så som en stor highlight i form av e-poster. Helen B Murphy presenterade dessa data. ”Start automatic insulin pump therapy as early as possible in patients with pregnant women with T1DM.”



www red Diabetolognytt