Three Years of Tirzepatide: What the Trial Found

Most weight loss studies run for a few months. That is long enough to show that something works and far too short to answer the question people actually have, which is what happens if you stay on it.

A phase 3 trial published in the New England Journal of Medicine in March 2025 ran considerably longer. It followed participants for 176 weeks, roughly three and a half years, and then took them off treatment for another 17 weeks to see what happened next. It is one of the longer looks we have at this class of medication.

What the trial actually did

The trial enrolled 2,539 adults with obesity, of whom 1,032 also had prediabetes. It was double blind and randomized, meaning neither participants nor investigators knew who was getting what, and assignment was by chance rather than choice.

Participants were split evenly into four groups: tirzepatide at 5 mg, 10 mg, or 15 mg once weekly, or placebo. Everyone stayed on their assigned arm for 176 weeks. Then treatment stopped for 17 weeks, which is the part that answers the question about coming off.

What it found on diabetes progression

Among the participants who started with prediabetes, the difference in who went on to develop type 2 diabetes was large.

At 176 weeks, 1.3 percent of those on tirzepatide had progressed to type 2 diabetes, compared with 13.3 percent on placebo. The hazard ratio was 0.07, with a 95 percent confidence interval of 0.0 to 0.1.

After the 17 week period off treatment, the figures were 2.4 percent and 13.7 percent, with a hazard ratio of 0.12. So some of the separation narrowed once treatment stopped, and most of it did not, at least over those 17 weeks.

What it found on weight

Mean body weight reduction from baseline at 176 weeks came out by dose:

• 5 mg: 12.3 percent
• 10 mg: 18.7 percent
• 15 mg: 19.7 percent
• Placebo: 1.3 percent

Those are averages across a large group. An average is not a forecast. Individual results in a trial of this size vary widely around the mean in both directions, and your own biology, starting point, and everything else you are doing all sit outside what a number like this can tell you.

The distinction that matters most here

This trial studied tirzepatide as an FDA approved medication, manufactured and tested as such. That is what was in the syringes.

Compounded tirzepatide is a different thing. A compounded preparation is made by a licensed pharmacy for an individual patient. It is not FDA approved, and it has not been evaluated by FDA for safety or effectiveness. Results from a trial of an approved product are not evidence about a compounded one, and anyone presenting them as interchangeable is telling you something the data does not support.

That is not a knock on compounding, which exists for real clinical reasons. It is a statement about what this particular evidence covers.

Who ran it

Worth knowing: this trial was sponsored by the manufacturer of the medication, which is standard for phase 3 work and is also a reason to read the result with the same care you would apply to any industry funded study. The trial was randomized, double blind, and placebo controlled, which are the design features that guard against sponsor influence on the outcome. It was published in a peer reviewed journal with the funding disclosed.

How to read this if you are deciding

A trial like this answers one question well: over three and a half years, in a large randomized group, what happened on average. It does not tell you what happens to you.

The questions it cannot answer are the ones worth bringing to a provider. Where do your labs actually sit. Whether prediabetes is part of your picture at all. What the side effect profile looks like against your history. What happens to muscle mass alongside fat, which this readout does not separate out. What the plan is if you stop.

Those are the questions that decide whether any of this applies to you, and none of them are answered by a percentage.