GLP-1 Drugs for Prediabetes: What Semaglutide and Tirzepatide Did to Diabetes Risk

In people with obesity and prediabetes, semaglutide and tirzepatide pushed most blood-sugar readings back into the normal range and sharply cut the number who went on to type 2 diabetes. Over three years of tirzepatide, 1.3% developed diabetes against 13.3% on placebo.^[1] No GLP-1 is approved by the FDA for prediabetes itself, though: the drugs are prescribed for weight, and the diabetes-prevention effect lasts as long as the treatment does.^[5]

What counts as prediabetes

The ADA defines prediabetes by any one of three blood tests: an HbA1c of 5.7% to 6.4%, a fasting glucose of 100 to 125 mg/dL, or a two-hour glucose of 140 to 199 mg/dL during an oral glucose tolerance test.^[6] About 97.6 million US adults had it in 2021, and most have no symptoms.^[10] It matters because it is the stage before type 2 diabetes, and the question for a drug is whether it changes that path.

Trials use slightly different cut-offs. STEP 10 used the UK definition, an HbA1c of 6.0% to 6.4% or a fasting glucose of 5.5 to 6.9 mmol/L, which is a narrower and higher band than the ADA’s.^[2] That difference makes the trial results hard to compare line for line, but the direction is the same in all of them.

What the trials found

The tirzepatide result is the largest and longest. In the SURMOUNT-1 participants who had prediabetes at the start, the drug was given for 176 weeks and then stopped for 17. On treatment, about 1 person in 77 developed type 2 diabetes, against about 1 in 8 on placebo.^[1] The trial page for SURMOUNT-1 covers its main 72-week weight results.

The semaglutide trials measured the other end: how many people moved from prediabetes back to normal glucose. In STEP 10, 103 of 127 people on semaglutide with a week-52 result got there, against 9 of 64 on placebo.^[2] The pooled STEP analysis found the same pattern at 68 weeks, and it also found fewer people with normal glucose slipping into prediabetes on the drug.^[3] The placebo groups in those trials also improved, because everyone received diet and activity counseling.

The only head-to-head comparison is a post hoc look at SURMOUNT-5, which pitted tirzepatide against semaglutide. Among its 425 participants with prediabetes, more returned to normal blood sugar on tirzepatide, which also took off more weight.^[11] It used the ADA’s wider prediabetes definition and was not designed to answer this question, so it shows a direction rather than a settled ranking.

Liraglutide, the older daily drug sold as Saxenda, showed a smaller effect over three years.^[4] The trial had a weakness that matters for reading it: about half the participants withdrew before week 160, and people who stopped were not followed afterwards, so the diabetes counts cover only those who stayed on treatment.

What happens when the drug stops

SURMOUNT-1 is the only one of these trials that reported diabetes onset after treatment ended. Seventeen weeks after the last dose, 2.4% of people who had taken tirzepatide had type 2 diabetes, against 13.7% on placebo.^[1] The gap was still wide, but the tirzepatide figure had nearly doubled within four months, and the trial did not follow people long enough to show where it settled.

The semaglutide trials point the same way. In STEP 4, everyone took semaglutide for 20 weeks before half were switched to placebo; at week 68, 70.4% of the switched group still had normal glucose, against 89.8% of those who stayed on the drug.^[3] A normal reading while on the drug shows the treatment working; it does not mean the underlying tendency has gone. The same distinction runs through our article on why normal blood sugar on tirzepatide is not remission, and the regain data are in what happens when you stop a GLP-1.

What the ADA’s 2026 Standards of Care say

The ADA’s 2026 chapter on preventing diabetes is plain about the regulatory position: no drug is approved by the FDA specifically to prevent type 2 diabetes.^[5] It acknowledges that several obesity drugs, liraglutide, semaglutide and tirzepatide among them, have lowered the incidence of type 2 diabetes in people with prediabetes.^[5]

“There are currently no long-term data to support the use of pharmacologic treatments other than metformin for the sole purpose of preventing type 2 diabetes.”

— ADA Standards of Care in Diabetes 2026, section 3

So metformin is the drug the ADA names for prevention alone. Its recommendation 3.7 says metformin should be considered in adults at high risk, especially those aged 25 to 59 with a BMI of 35 or more, a fasting glucose around 110 mg/dL or higher, an HbA1c around 6.0% or higher, or a history of gestational diabetes.^[5] GLP-1 drugs enter through a different recommendation, 3.15, which says drug treatment for weight management, for slowing the rise in blood sugar and for cardiovascular risk should be considered to support the person’s own care goals.^[5]

Read together, the guidance supports a GLP-1 for someone with prediabetes who also has obesity and wants to treat it, with diabetes prevention as one of the benefits. It does not support starting one for the blood-sugar number alone. People already on metformin who add a GLP-1 will find the combination was common in the diabetes trials; our article on metformin and a GLP-1 together covers the side-effect overlap.

Does prediabetes help with insurance?

Not directly. Zepbound and Wegovy are labeled for long-term weight management when someone has obesity, or is overweight and has a weight-linked health problem, plus a few named conditions; neither label mentions prediabetes.^[7][8] A prescriber can list prediabetes as the weight-related condition, and a commercial plan that covers weight-loss drugs may accept it, but employer plans choose whether to cover these drugs at all, and the member’s own plan documents decide.

Medicare is the one place where prediabetes is written into the rules. Under the Medicare GLP-1 Bridge, which began on July 1, 2026, prediabetes qualifies anyone whose BMI is at least 27, and the copay is $50 a month. Type 2 diabetes rules a person out of the Bridge, since diabetes coverage for these drugs runs through ordinary Part D.^[9] For a commercial plan, the prior-authorization letter tool lays out what a request usually needs.

Who these results apply to

• People with obesity. STEP 10 needed a BMI of at least 30; SURMOUNT-1 and SCALE also took people from a BMI of 27 if they had a weight-linked health problem. Nobody at a normal weight was studied.^[1][2][4]
• People willing to stay on treatment. The protection was measured on the drug. The one off-treatment reading showed it fading within months.^[1]
• People who can tolerate it. Stomach side effects were the most common problem, mostly in the first 20 weeks while the dose was being raised.^[1] The titration planner shows that schedule.

Two things remain unknown. No trial has compared a GLP-1 head to head with metformin or with an intensive lifestyle program for preventing diabetes, and no trial has shown that preventing the diagnosis this way lowers heart attacks, kidney disease or deaths in people with prediabetes. The longer-term blood-sugar picture also links to insulin resistance, covered in what GLP-1 drugs do to insulin resistance.