GLP-1 for Prediabetes: What the New Study Found

Taking a GLP-1 for prediabetes was linked to better blood-sugar outcomes than taking metformin in a large new study. That doesn't prove the medicine caused the difference. Researchers compared existing care. The findings are preliminary and haven't been reviewed as a full journal paper. They also combine four medicines, so they don't establish what semaglutide alone would do.

Prediabetes means blood sugar is above normal but below the diabetes range. GLP-1 medicines act like a gut hormone involved in appetite and blood sugar. Metformin is a long-used diabetes medicine.

What the GLP-1 for prediabetes study found

As of November 5, 2026, the day this article was written, the American Heart Association's November 2 release describes findings from 256,492 adults with prediabetes and obesity. Their average age was 52; 76% were women.

Researchers used TriNetX, a network of medical records, covering 104 institutions during 2010–2023. Follow-up averaged roughly three years. They compared people prescribed semaglutide, tirzepatide, liraglutide or dulaglutide with people prescribed metformin, matching records on measured health characteristics.

The conference abstract, MDP1524, reports:

That's about 37 versus 29 people in every 100 returning to normal blood sugar. The “1.5 times” headline comes from a hazard ratio, a comparison of event rates over time. It isn't the ratio of those two percentages.

The reported hazard ratios also corresponded to 31% lower diabetes and 41% lower death rates. Heart attacks, strokes, heart failure and atrial fibrillation, an irregular heartbeat, were less frequent too. These are associations in records, not proof of treatment effects.

What it means for a prescribing discussion

The practical question is whether treating obesity could also help someone whose blood sugar is rising. It isn't a reason to assume metformin has failed.

Study author Vinicius de Sousa Barzon Serra, a Mayo Clinic research fellow, said in the AHA release: “Metformin is still a first-line treatment choice.” First-line means a usual initial option.

Metformin has randomized evidence too. In the Diabetes Prevention Program, NEJM, 2002, 3,234 adults at high diabetes risk were followed for an average 2.8 years. Diabetes incidence was 31% lower with metformin and 58% lower with intensive lifestyle support, each compared with placebo. That 31% describes a different comparison from the new AHA findings.

For context, CDC defines prediabetes as an A1C of 5.7–6.4%. A1C estimates average blood sugar over roughly three months. Below 5.7% is the usual normal range. The conference summary doesn't specify its exact testing threshold or how long a normal result had to last.

The ADA's 2026 prevention guidance supports considering metformin for selected adults at high risk. It also discusses GLP-1 treatment for weight management in people with excess weight. Nutrition and activity support remain part of care.

The insulin-resistance evidence guide explains the broader treatment research. The HOMA-IR guide covers a blood-test estimate of how readily the body responds to insulin, the hormone that helps cells use sugar.

What randomized trials add

Randomized trials assign treatment by chance, making the groups more comparable. Earlier trials support studying these medicines before diabetes develops, although they don't answer the new metformin comparison.

In STEP 10, published in The Lancet Diabetes & Endocrinology in 2024, 207 adults had obesity and prediabetes. At 52 weeks, about four in five semaglutide participants with measurements met the trial's normal-blood-sugar definition: 81%, versus 14% with placebo, an inactive comparison treatment. Both groups received food and activity counseling. The abstract reports those figures.

That trial used a different definition from the usual U.S. A1C threshold: both A1C below 6.0% and fasting blood sugar below 5.5 mmol/L. Fasting means after a period without food; mmol/L is a blood-sugar measurement unit.

Tirzepatide has separate evidence. The SURMOUNT-1 analysis in the New England Journal of Medicine in 2025 followed 1,032 participants with obesity and prediabetes. At 176 weeks, diabetes was diagnosed in about one in 100 on tirzepatide versus 13 in 100 on placebo: 1.3% versus 13.3%. The paper's abstract reports this comparison, which did not involve metformin.

What the new findings don't establish

The AHA findings come from an observational study: researchers watched existing care rather than assigning treatment. They are reported in a conference abstract, a short research summary. A full paper hasn't undergone peer review, the independent expert check before journal publication. The clinical-trial phases guide explains how planned trials build evidence.

Matching can balance recorded details. It cannot fully account for why someone received one medicine, insurance access, follow-up care or differences in weight. The release also notes incomplete diagnosis codes and uncertainty about doses and whether prescriptions were actually taken.

A return to normal blood sugar during follow-up doesn't establish permanent protection after treatment stops. Nor does pooling four medicines show that one beats another. The choice between metformin and a GLP-1 remains an individual prescribing decision.

As of November 5, 2026, the day this article was written, branded GLP-1 medicines have defined uses, but no FDA-approved indication specifically for prediabetes. The FDA's Ozempic label, Revised: 05/2026, includes type 2 diabetes; its Wegovy label, Revised: 06/2026, includes weight management. The ADA guidance states that no medication has a specific diabetes-prevention indication.

Through Promise, semaglutide and tirzepatide are dispensed by a licensed U.S. compounding pharmacy as compounded medications. These are different from an FDA-approved product: the formulations offered here are not FDA-approved. Neither the conference comparison nor the brand trials establish identical results for a compounded preparation.

A licensed provider may still prescribe — that decision is between you and your doctor.

What to watch next

The AHA program schedules this poster for November 8. A full paper could clarify the blood-sugar definitions, treatment changes, individual-drug results and side effects.

Access also matters. As of November 5, 2026, the day this article was written, MassHealth's October 2026 prescriber letter announces coverage changes for January 4, 2027. Semaglutide and tirzepatide used for prediabetes will require prior authorization, meaning insurer permission before coverage. That is a policy of MassHealth, Massachusetts's Medicaid program, not a nationwide prescribing rule. MassHealth's prior-authorization requests must document results with metformin, Trulicity and an SGLT2 inhibitor first.

Medicaid's weight-loss drug coverage rules reflect a federal framework that lets states exclude medicines used solely for weight loss.

Meanwhile, tolerability belongs in the discussion. In STEP 10, side effects led 6% on semaglutide to stop treatment, versus 1% on placebo. The provider weighs that possibility alongside blood sugar, weight, medical history and the other options.

At Promise, a licensed provider reviews every request, and not everyone qualifies.