Semaglutide discontinuation erodes time in range in type 1 diabetes

Adults with type 1 diabetes and obesity lost part of their glucose gains within three months of stopping the drug, a follow up of the ADJUST-T1D trial shows. The semaglutide discontinuation analysis found time in range fell in former users while it edged up in the placebo group. The results appeared in Diabetes, Obesity and Metabolism and were presented at the Breakthrough T1D Clinical and Research Congress in Philadelphia, held 9 to 11 October. The study is small and post hoc. It does not show how long the decline lasts or whether it reverses.

What the extension study measured

ADJUST-T1D was a 26 week, double blind trial that randomly assigned 72 adults with type 1 diabetes and a body mass index of 30 or more to weekly semaglutide or placebo. Every participant already used automated insulin delivery, a pump that adjusts insulin from a sensor reading. When the trial closed, the drug stopped. Investigators then gathered continuous glucose monitor data over a 12 week extension phase. The paper compares 16 people who had received semaglutide and then came off it with 22 people from the placebo arm who took no drug of that class during follow up. Healio reported that records for 58 participants were reviewed in total. The analysis tracked the change in sensor metrics from week 26 to the extension window, with statistical adjustment for testing many measures at once.

The author list includes Montaser, Kanchan Bhasin and senior investigator Viral Shah. Bhasin, a postdoctoral fellow in endocrinology at Indiana University School of Medicine, presented the work at the congress. The paper was posted online on 31 August and is assigned to the November issue of the journal. The central question was simple. If a person with type 1 diabetes gains better control on semaglutide, does that control hold once the weekly shot is withdrawn? The extension design cannot answer that for every patient. It covers one short window, a few dozen people and a single care setting. Still, it gives clinicians published numbers on semaglutide discontinuation in this group, where earlier data covered only the treatment period.

Time in range and glycemic variability after semaglutide discontinuation

Time in range is the share of the day a sensor reads between 70 and 180 milligrams per decilitre. In the former semaglutide group it fell by a median of 3.6 percentage points over the follow up. In the placebo group it rose by a median of 1.5 points. The gap was statistically significant after adjustment, with a p value of 0.044. Two measures of glycemic variability moved the same way. The standard deviation of glucose rose by 6.0 units in former users against 1.2 in the placebo group. The coefficient of variation rose by 3.1 points against 1.1. Each of those differences also carried an adjusted p value of 0.044. Time above range rose in former users, but that change lost significance after adjustment.

"One particularly interesting finding was that glycemic variability increased despite all participants in the ADJUST-T1D trial using automated insulin delivery," Bhasin told Healio.

The paper found no difference between groups in sensor measures of low glucose. That matters because hypoglycemia is the main day to day risk in type 1 diabetes. Healio reported that, by the end of follow up, both variability measures had become similar between the two groups. In plain terms, former users drifted back toward where the placebo group stood. "One particularly interesting finding was that glycemic variability increased despite all participants in the ADJUST-T1D trial using automated insulin delivery," Bhasin told Healio. The authors concluded that benefits gained during treatment may not be fully sustained after withdrawal. They did not claim the drug caused lasting harm. The numbers describe a loss of earlier gains, measured over twelve weeks in a small sample, rather than control that ended up worse than before treatment began.

What the parent trial had shown

The parent results were published in NEJM Evidence in June 2025. Over 26 weeks, 36% of semaglutide users met a composite goal of more than 70% time in range, less than 4% time below range and at least 5% weight loss. No placebo user met it. Semaglutide added 8.8 percentage points of time in range and 8.8 kilograms of weight loss compared with placebo. Glycated hemoglobin fell by an extra 0.3 points. Each arm recorded two severe hypoglycemia events, and no diabetic ketoacidosis was reported. The trial was funded by Breakthrough T1D and is registered as NCT05537233. A separate post hoc analysis in Diabetes Care reported that total daily insulin fell by 22.6% over the trial in the semaglutide group, mostly through lower mealtime doses.

Those gains frame the new finding. A drug that adds close to nine points of time in range during treatment and then gives back several points within three months behaves much as it does in type 2 diabetes, Bhasin said. Readers who follow trial coverage will know the pattern from weight studies, and our guide on how to read a weight-loss trial headline sets out the questions worth asking of a result like this one. The comparison here is also uneven in size. Only 16 former users were analysed, and the follow up began when the blinded phase ended. Patients and doctors knew the drug was stopping, which could change eating, activity and pump settings in ways the data cannot separate.

Off label use and the limits of the data

Semaglutide is not approved for type 1 diabetes. Its labels as Ozempic and Wegovy cover type 2 diabetes, weight management and related heart and kidney uses, so use in type 1 is off label. That status is one reason the data matter. Some people with type 1 diabetes and obesity already receive the drug from their doctors, and the extension study suggests the transition off treatment needs planning. "Clinicians should anticipate that glucose control and insulin requirements may change after semaglutide is discontinued," Bhasin told Healio. She recommended closer review of sensor data, a check of insulin settings when needed and timely follow up. She also stressed that semaglutide does not replace insulin, and that insulin therapy must continue for every person with type 1 diabetes.

The study has clear limits. It is a post hoc analysis, which means the question was not part of the original trial plan. Groups were small and the window was twelve weeks. The results do not say whether time in range keeps falling, settles or returns to its pretreatment level. They also cannot show whether weight regain drove the change. Bhasin said larger and longer studies are needed. She is an investigator on ATT1D, a real world study of GLP 1 drugs and SGLT2 inhibitors as add on therapy in type 1 diabetes. Our trials desk tracks studies of this kind. Until larger data arrive, the extension study is the main published guide to what happens when treatment ends.

Conclusion

The ADJUST-T1D follow up gives new published numbers on semaglutide discontinuation in type 1 diabetes. Former users lost a few points of time in range and saw more glucose swings within twelve weeks, while low glucose measures held steady. The sample is small and the analysis post hoc. For now, the data support careful monitoring when treatment stops, not conclusions about long term harm or benefit.