Prediabetes and GLP-1s: The A1c Numbers, What Works First, and What Wegovy and Zepbound Actually Showed

Can you take Wegovy or Ozempic for prediabetes?

Not for prediabetes alone. No GLP-1 is approved for it: Wegovy and Zepbound are approved for obesity, or overweight plus a weight-related condition, and Ozempic is a type 2 diabetes drug. If your weight qualifies, the trials are striking: most adults with prediabetes on semaglutide returned to normal blood sugar, and tirzepatide cut new diabetes about tenfold.

That is the short version. The longer one matters more for mothers, because a large share of us first meet the word "prediabetes" in a postpartum letter we never quite got around to acting on.

The numbers that make it prediabetes

The American Diabetes Association defines prediabetes by any one of three tests, in someone who is not pregnant¹:

The ADA adds a line worth taping to the fridge: for all three tests, risk is continuous, extending below the lower limit of the range and climbing faster toward the top¹. An A1c of 6.3% is not the same news as 5.7%, even though both carry the same label.

It is also common. The CDC says more than 2 in 5 American adults have prediabetes, and that it often goes undetected for years because it has no clear symptoms².

What works first: the trial that set the bar

Before any injection, there is the Diabetes Prevention Program. It randomized 3,234 adults with elevated fasting and post-meal glucose, 68% of them women, to placebo, metformin, or a lifestyle program aiming for 7% weight loss and 150 minutes of activity a week. Over an average of 2.8 years, lifestyle change cut new diabetes by 58% and metformin by 31%, compared with placebo³. About seven people had to do the lifestyle program, and about fourteen had to take metformin, to prevent one case over three years³.

The 15-year follow-up is the sobering half. The benefit shrank but held: diabetes incidence was 27% lower in the lifestyle group and 18% lower on metformin. By year 15, though, 55% of the lifestyle group had developed diabetes, against 62% on placebo⁴. Prevention delays the disease. For many people it does not erase the risk.

What the GLP-1 trials did to blood sugar

The weight-loss trials enrolled plenty of people with prediabetes, and two analyses looked at them directly.

Semaglutide (Wegovy). Across STEP 1, 3 and 4, 1,536 adults started with prediabetes. In STEP 1, 84.1% of them had normal glucose after 68 weeks on semaglutide, against 47.8% on placebo, with lifestyle counseling in both arms⁵. A1c, fasting glucose and insulin resistance all improved more than on placebo⁵.

Tirzepatide (Zepbound). SURMOUNT-1 kept 1,032 adults with obesity and prediabetes on treatment for 176 weeks, a little over three years. Type 2 diabetes was diagnosed in 1.3% on tirzepatide and 13.3% on placebo⁶. Seventeen weeks after stopping, the figures were 2.4% and 13.7%⁶. Average weight change at three years ran from −12.3% on the lowest dose to −19.7% on the highest, against −1.3% on placebo⁶.

A caution on reading that table: these are different trials with different people and different endpoints, so the rows are not a ranking. And the semaglutide numbers were measured while people were still on the drug. What happens after you stop is the question the three-year tirzepatide data only begins to answer.

Is it "for prediabetes"? The label and the coverage

No. Wegovy's label covers cardiovascular risk reduction in adults with heart disease, long-term weight management, and a liver indication (MASH)⁷. Zepbound's covers weight management and moderate to severe obstructive sleep apnea in adults with obesity⁸. Prediabetes appears in neither indication.

So the route to a prescription runs through weight. You need a BMI of 30 or more, or overweight plus at least one weight-related condition^7,8. Many prescribers treat prediabetes as that condition, but your insurer decides what counts, and some plans exclude weight-loss drugs altogether. Ask for the plan's prior-authorization criteria in writing before you start. If the answer is no, our board for coverage-friendly programs and our guide to what a GLP-1 costs without insurance set out the cash routes.

Metformin is the opposite case: generic, cheap and old. It has the longest prevention track record, even though it did less than lifestyle change in the DPP³. Some programs include it. The WeightWatchers Clinic membership covers metformin filled through its mail-order partner, and Allara Health treats insulin resistance and PCOS as its core practice.

If you had gestational diabetes

This is the part we most want mothers to read. A 2020 meta-analysis of more than 1.3 million women found that a history of gestational diabetes raised the risk of type 2 diabetes almost tenfold, with a relative risk of 9.51⁹. The authors stressed the early years after pregnancy as the window to act⁹.

The ADA's instruction is concrete: a 75-g glucose tolerance test at 4–12 weeks postpartum, then screening every 1–3 years for life¹. That second half is the one that slips. In a 2023 qualitative study of Australian women with prior gestational diabetes, researchers note that around half of such women go unscreened. Participants described why, in words many of us will recognize¹¹. One said: "I thought once I had the – the six week one, I'd be in the all-clear."¹¹ Another put the priority problem plainly: "…if it was for the health of my baby, I'll do it, but if it's just for me … it wasn't as high on my priority list."¹¹

There is good news here too. In the DPP, women with a gestational-diabetes history who got placebo developed diabetes at a 71% higher crude rate than other mothers with the same glucose. But lifestyle change and metformin each cut their risk by about half¹⁰. Metformin worked notably better in that group than in mothers without the history¹⁰. If you have the history and a borderline A1c, that is a specific, evidence-backed thing to raise. Our guide to what to tell your OB before starting a GLP-1 covers the rest of that conversation, and postpartum timing and breastfeeding covers when a GLP-1 can even be on the table.

What we would ask at the next appointment

Ask for your actual A1c number, not just "a little high." Ask whether your fasting glucose or a glucose tolerance test agrees with it. If you had gestational diabetes, ask when your next screen is due and put it in the calendar like a child's vaccine. Then ask the real question. Is the first step a structured lifestyle program, metformin, or, if your weight qualifies, a GLP-1? The evidence supports all three. Which one fits depends on your numbers, your plans for another pregnancy, and what your plan will pay for.