What Happens When You Stop a GLP-1: What the Withdrawal Trials Measured

Weight comes back. This is one of the few questions about GLP-1 drugs with a direct randomized answer, because three separate trials took people who had lost weight on semaglutide or tirzepatide and deliberately took the drug away. A year after semaglutide and its lifestyle program were withdrawn, participants had regained two-thirds of what they lost.^[1] In the tirzepatide withdrawal trial the placebo group put on 14.0 percent of body weight over the 52 weeks that followed.^[3] That is what the evidence says. What you do about it is not this page's decision, and the honest reading of those results is less bleak than it first sounds.

The three trials that took the drug away on purpose

Most drug trials measure what happens when you take something. These measured what happens when you stop, and they were designed that way from the start. Two used semaglutide, one used tirzepatide, and their designs differ in ways that matter when you read the numbers.

Three trials, two molecules, one direction. The STEP 1 extension deserves a caveat its headline rarely gets: its analyses were exploratory, its extension covered a subset of sites rather than the whole trial, and the lifestyle program was withdrawn along with the drug, so it measures the removal of a whole program rather than of a molecule.^[1] STEP 4 and SURMOUNT-4 kept lifestyle support running in both arms, which isolates the drug more cleanly, and they point the same way.

What comes back besides the number on the scale

This is the part that is usually left out, and it is also where the picture stops being uniform. A post hoc analysis of SURMOUNT-4 sorted the people who came off tirzepatide by how much of their loss returned, then looked at what else moved.^[4]

Between week 36 and week 88, waist circumference grew by less than a centimeter on average in those who regained under a quarter of what they had lost, and by close to fifteen in those who regained three quarters or more. Systolic blood pressure, glycated hemoglobin and non-HDL cholesterol climbed on the same gradient, each one rising further in the bands that regained more.^[4] Fasting insulin did not sort as tidily, and we are not going to smooth that over: its largest rise landed in the second-heaviest regain band rather than the heaviest.^[4] In the group whose weight barely moved, waist circumference, non-HDL cholesterol and fasting insulin at week 88 were not significantly different from where they had been at week 36.^[4]

Across the wider literature the direction is the same. A 2025 systematic review pooled 18 randomized trials with at least twelve weeks of follow-up after treatment ended, covering 3,771 participants, and found weight and glycated hemoglobin both moving back in adults with obesity and in adults with type 2 diabetes, with the regain larger where follow-up ran past 26 weeks.^[6] We would rather flag its limitation than print its pooled figures as though they were clean: heterogeneity across the included trials was extremely high on several of those estimates, which means the average conceals a wide spread of individual trial results.

Appetite: what we can say, and what nobody published

Readers describe hunger returning before the weight does, and the pharmacology makes that plausible: these drugs act at receptors involved in appetite and caloric intake, and that action ends when the drug is no longer there. What we could not find is a withdrawal trial reporting appetite or hunger on a validated scale as a published endpoint. So we are not going to give you a number for it, or a week at which it happens. The measured outcomes in these trials are body weight and cardiometabolic markers, and those are what this page reports.

That gap is worth naming rather than filling. If your appetite has changed since stopping and it is distressing you, that is a thing to describe to a clinician, who has options this page cannot list.

The drug leaves faster than the effect does

People often assume the regain curve is the clearance curve. It is not, and the labels make the difference obvious. For semaglutide the label puts the elimination half-life at roughly a week, and says the drug will still be circulating some five to seven weeks after a last injectable 2.4 mg or 7.2 mg dose, or a 25 mg oral one.^[9] Tirzepatide's half-life is approximately five to six days in people with overweight or obesity.^[10]

So the molecule is gone within a couple of months, while the trials measured weight moving over a year. Whatever is happening after that first stretch is no longer the drug clearing. Our article on how long GLP-1 side effects last takes the pharmacokinetics apart properly, including what the long tail means for symptoms that persist after a last dose.

If you have type 2 diabetes, this is a different conversation

Two of these products are indicated for diabetes rather than for weight. Ozempic's approved use is blood-sugar control in adults who have type 2 diabetes, taken alongside diet and exercise, and Mounjaro carries that same use for adults and for pediatric patients aged 10 and over.^[11][12] Stopping one of them removes a glucose-lowering medicine from a regimen built around it, which is a different kind of event from regaining weight.

We have to be careful about the evidence base here, because it is thinner than the weight evidence. The three withdrawal trials above all excluded diabetes. What exists is the meta-analysis subgroup, where glycated hemoglobin rose after discontinuation in people with type 2 diabetes and rose by more than it did in people with obesity alone, again with high heterogeneity behind the pooled estimate.^[6] Alongside it sits one long-term signal that runs the other way: in the three-year SURMOUNT-1 report, participants with obesity and prediabetes who then spent seventeen weeks off treatment were still far less likely to have received a diabetes diagnosis than the placebo group.^[5] Not everything measured reverts on the same timetable.

Why people actually stop

The trials answer what happens. They do not answer why anyone is in this position, and the real-world data on that is unflattering to the market rather than to the patients. Researchers reviewed the records of 288 adults who had started injectable semaglutide or tirzepatide for obesity and stopped inside the first year, and read the reason recorded for each.^[7]

• Cost or an insurance problem: 47.6 percent. Nearly half. Not tolerability, not disappointment with the results.^[7]
• Could not tolerate the side effects: 14.6 percent.^[7]
• Could not fill the prescription because of a shortage: 11.8 percent.^[7]
• Switched to a compounded medication: 2.4 percent.^[7]
• Unsatisfactory weight loss: 1.7 percent. The reason people assume dominates this list is close to the bottom of it.^[7]

That matters for how the results above should be read. A separate cohort of 7,881 patients from the same health system found one-year weight reduction of 3.6 percent among those who stopped early, 6.8 percent among those who stopped later in the year, and 11.9 percent among those who did not stop, with most patients on low maintenance dosages throughout.^[8] The authors' own explanation for why their average fell short of the trial results was discontinuation and dosing rather than anything about the drug.^[8]

If cost is what is pushing this decision, say that to your prescriber before you stop rather than after, and check the dated figures on our price tracker against whatever you are being quoted. Where a compounded program is the thing ending, we walk through the remaining routes in what to do when a compounded GLP-1 program ends, and the cash routes for the approved products in what semaglutide costs without insurance.

The framing that actually fits the evidence

None of this is a story about willpower, and the trial authors did not treat it as one. The STEP 1 extension investigators put their own conclusion in the paper's abstract.

Findings confirm the chronicity of obesity and suggest ongoing treatment is required to maintain improvements in weight and health.

— Wilding et al., STEP 1 trial extension, Diabetes Obes Metab 2022

Read plainly, that is a statement about a class of medicine rather than about the people taking it. Blood pressure returns when an antihypertensive stops. Nobody calls that a relapse. What the withdrawal trials establish is that these drugs work while they are being taken and that the condition they treat is still there underneath, which is an argument about how treatment should be planned and paid for, not about anyone's discipline.

It also means the decision in front of you is genuinely a decision, with a real trade-off on both sides and no answer this register can supply. The multi-year safety record that belongs on the other side of that scale is in tirzepatide long-term side effects. The person who can weigh both against your own history is your prescriber.