GLP-1s vs. Metformin for Prediabetes in 2026: New Study of 256,000 Patients Links GLP-1s to Lower Diabetes and Heart Risks

Researchers at the Mayo Clinic in Scottsdale analyzed electronic health records from 256,492 adults with both prediabetes and obesity, drawn from the TriNetX network of 104 health care institutions between 2010 and 2023. Half took a GLP-1 medication and half took metformin, with the groups matched on age, blood pressure, sleep apnea, smoking, and other medications. Average age was 52, about three-quarters were women, and patients were followed for roughly three years.

According to the American Heart Association's summary of the findings, GLP-1 users had, compared with metformin users:

• About 1.5 times higher odds of returning to normal blood sugar
• A 31% lower risk of progressing to type 2 diabetes
• A 41% lower risk of death from any cause
• 15% fewer heart attacks, 20% fewer strokes, 25% fewer heart failure cases, and 8% fewer cases of atrial fibrillation

The findings will be formally presented at the AHA Scientific Sessions, held November 6–9, 2026.

Prediabetes is remarkably common: the AHA's 2026 statistics cited in the release estimate nearly 40% of U.S. adults have it, and NIH figures suggest up to 10% of adults with prediabetes progress to type 2 diabetes every year. Most people in that group are told to change their diet, move more, and often to start metformin — an inexpensive generic that has been the standard prevention drug for decades.

What this new analysis adds is a direct, if observational, comparison at scale. As coverage in the American Journal of Managed Care notes, prior evidence already hinted GLP-1s could reverse prediabetes: in the STEP 10 randomized trial published in 2024, 81% of patients on semaglutide 2.4 mg returned to normal blood sugar at one year versus 14% on placebo. But STEP 10 never tested semaglutide against metformin. This week's data is among the first large head-to-head looks — and the GLP-1 group came out ahead on nearly every outcome measured, including the cardiovascular ones.

Study author Dr. Vinicius de Sousa Barzon Serra said the findings "may add to the case for considering GLP-1 medications earlier for certain patients," while emphasizing that metformin remains a first-line choice and lifestyle change remains the foundation of prediabetes care.

This is a retrospective records study, not a randomized trial, and the abstract has not yet been peer-reviewed. That means it can show association, not proof of cause and effect. People prescribed GLP-1s may have differed from metformin users in ways the matching could not fully capture. Diagnosis codes can be incomplete, the records do not show whether patients actually took their medication consistently or at what dose, and the four GLP-1 drugs were analyzed as one group rather than separately.

Dr. Robert Eckel, a past president of the American Heart Association who was not involved in the research, offered a grounded read: the benefit "likely relates to the patients losing weight," and cost and access remain real barriers. In other words, GLP-1s probably are not magic for blood sugar specifically — they work in large part because they produce far more weight loss than metformin does.

Metformin costs a few dollars a month, has a 60-year safety record, and modestly reduces diabetes risk. Its downsides are gastrointestinal upset in some patients and relatively little weight loss.

GLP-1 medications like semaglutide and tirzepatide produce substantially more weight loss — and, based on this and other recent data, larger metabolic improvements — but they cost more, are injected weekly (oral options now exist), and carry their own side effects. The most common are nausea, vomiting, constipation, and diarrhea, especially during dose increases; many patients manage this with slower titration or anti-nausea support. GLP-1s are not appropriate for everyone: people with a personal or family history of medullary thyroid carcinoma or MEN2, a history of pancreatitis, or who are pregnant or planning pregnancy should not take them, which is why screening with a licensed provider matters before starting.

Neither drug is officially FDA-approved for "prediabetes" as a standalone label — GLP-1s are approved for weight management and type 2 diabetes — so treating prediabetes with them is a decision made individually with a clinician, usually anchored to a weight-management indication when BMI qualifies.

If you have been told you have prediabetes — an A1C of 5.7% to 6.4% — this study is not a reason to abandon metformin if it is working for you. It is a reason to have a broader conversation. The key questions worth raising with a provider: What is my BMI and do I qualify for GLP-1 therapy on weight-management grounds? How much weight loss would meaningfully change my diabetes risk? What will each option cost me month to month? And what does my cardiovascular risk picture look like, since that is where this new data showed some of the most striking differences?

For many patients, the honest answer will still be lifestyle change first, metformin where appropriate, and a GLP-1 when weight is the central driver of their risk.

If prediabetes showed up on your labs and weight is part of the picture, JoeyMed's licensed providers can evaluate whether GLP-1 treatment is right for you — including compounded tirzepatide options like Tirz HD — through a straightforward online visit. Your provider reviews your history, screens for contraindications, and builds a plan that fits your goals and budget, with ongoing check-ins as your treatment progresses. Learn more or start your free consultation at joeymed.com/weight-loss, or explore our broader wellness services if you are focused on prevention and long-term health.