GLP 1 drugs in prediabetes tied to less diabetes and fewer heart events than metformin
A records study of more than 256,000 adults with prediabetes and obesity found GLP 1 users reverted to normal blood sugar more often than metformin users. The abstract is not peer reviewed and cannot show cause and effect.
What's new
An AHA 2026 records study of 256,492 matched adults with prediabetes and obesity linked GLP 1 drugs to more reversion to normal blood sugar and less diabetes than metformin.
Why it matters
Most GLP 1 research has been in people who already have diabetes, and the study also linked the drugs to lower death and heart failure risk.
What's next
The abstract is not peer reviewed, and the authors say randomized trials are needed to confirm cause and effect.
Adults with prediabetes and obesity who started a GLP 1 drug were more likely to return to normal blood sugar and less likely to develop diabetes than similar people who started metformin, a records study presented at the American Heart Association meeting in Chicago reports. The analysis of GLP 1 drugs in prediabetes covered more than 256,000 matched patients. It also linked GLP 1 use to lower death and heart failure risk. The abstract has not been peer reviewed, and its authors say trials must confirm cause and effect.
How the prediabetes records study worked
The team, led by Vinicius de Sousa Barzon Serra, a research fellow in cardiovascular medicine at the Mayo Clinic in Scottsdale, Arizona, used the TriNetX network of electronic health records from 104 institutions. They selected adults with prediabetes and obesity who started semaglutide, tirzepatide, liraglutide, dulaglutide or metformin between January 2010 and December 2023. People who already had diabetes, stroke, heart failure, atrial fibrillation or ischemic heart disease were left out. Each GLP 1 user was matched with a metformin user on age, other conditions, medicines and laboratory values. After matching, 256,492 patients remained, with a mean age of 52.4 years, and 76.1% were women. Average follow up was about three years.
The study had two main outcomes, reversion to normal blood sugar and progression to type 2 diabetes. Secondary outcomes were death from any cause and major cardiovascular events, analysed with standard survival models. Mean body mass index was 39.5 in the GLP 1 group and 38.5 in the metformin group, so the GLP 1 users were slightly heavier at the start despite matching. The four GLP 1 drugs were analysed as one group, so the study cannot say whether any single drug did better than another. The American Heart Association released the findings on 2 November, ahead of the meeting, as abstract MDP1524, and HCPLive reported the confidence intervals given in it.
GLP 1 drugs in prediabetes against metformin
Blood sugar returned to normal in 37.3% of GLP 1 users and 29.3% of metformin users, a hazard ratio of 1.53. Diabetes developed in 25.1% of GLP 1 users against 37.0% of metformin users, a hazard ratio of 0.69, or about 31% lower risk. Death from any cause was 41% lower with GLP 1 drugs, a hazard ratio of 0.59 with a 95% confidence interval of 0.55 to 0.63. The GLP 1 group also had lower rates of acute heart failure, ischemic stroke, heart attack, cardiac arrest and atrial fibrillation. Rates of ventricular tachycardia, a fast heart rhythm, did not differ between the groups. The abstract gave no confidence intervals for the individual heart outcomes.
Diabetes developed in 25.1% of GLP 1 users against 37.0% of metformin users, a hazard ratio of 0.69, or about 31% lower risk.
Serra framed the result as a reason to talk about options, not to switch drugs. "The great news with prediabetes is that it's still possible to reverse the progression of the disease," he said in the association news release. He added that "metformin is still a first-line treatment choice. It is safe, inexpensive and has a long track record." Robert Eckel, a former volunteer president of the association who was not involved in the study, said the benefits likely relate to weight loss and that "costs and access to these medications are important factors." Most earlier GLP 1 research focused on people who already have type 2 diabetes.
| Outcome, GLP 1 against metformin | GLP 1 users | Metformin users | Hazard ratio |
|---|---|---|---|
| Return to normal blood sugar | 37.3% | 29.3% | 1.53 |
| Progression to type 2 diabetes | 25.1% | 37.0% | 0.69 |
| Death from any cause | not reported | not reported | 0.59 |
| Acute heart failure | not reported | not reported | 0.74 |
| Ischemic stroke | not reported | not reported | 0.80 |
| Heart attack | not reported | not reported | 0.84 |
Limits of an observational comparison
The study cannot show that the drugs caused the differences. People prescribed a GLP 1 drug may differ from those given metformin in ways that matching cannot capture, such as insurance coverage, access to care, general health and willingness to take medicines. The presenters themselves raised these points, the association said. Outcomes were found through diagnosis codes, which can miss or misclassify events. Prescription records show what was prescribed, not whether people took it or at what dose, and the abstract reports no treatment duration. The death result is large for a three year follow up in a group without heart disease, which is a further reason to read it with care until a full paper appears.
The comparison also sets newer, costlier drugs against a cheap generic. Metformin, alongside diet and exercise, remains the standard first medicine for prediabetes, and no GLP 1 drug is approved in the United States for prediabetes as such. The study does not tell doctors which patients would gain most, and it gives no information on side effects, stopping rates or cost. These are the questions a randomized trial would need to answer. Findings on GLP 1 drugs in prediabetes drawn from routine records also cannot capture harms that were never coded. Our editorial standards describe how this site labels conference abstracts and other evidence that has not been peer reviewed.
What comes next for prediabetes care
The investigators say prospective randomized trials are needed to confirm whether GLP 1 drugs prevent diabetes and heart events better than metformin in this group. Until then, the result adds to observational evidence and to the question of whether these drugs should be used earlier, before diabetes is diagnosed. Any such shift would also depend on price and coverage, which the authors did not study. Serra said the main effect could be on shared decision making between doctors and patients, and that the work is "less about replacing one medication with another and more about discussing additional options with specific patients," according to the association.
Readers should not change treatment on the basis of an abstract. People with prediabetes who are interested in a GLP 1 drug should discuss it with a prescriber and use only approved medicines from licensed pharmacies, since compounded and online products vary in content and quality. The full manuscript, when published, should report the methods and confidence intervals in more detail and may change the estimates. New readouts and how to read them appear on the trials desk, alongside other comparisons that rely on health records rather than randomized trials. Whether the death and heart results hold up will be the main test of this work.
Conclusion
A large records study of GLP 1 drugs in prediabetes found more people returning to normal blood sugar and fewer developing diabetes than with metformin, along with lower death and heart failure rates. The findings come from an abstract that is not peer reviewed and cannot separate the drugs' effect from differences between patients. Metformin remains the standard first choice, and trials are needed before practice changes.