What Happens When You Stop Taking Zepbound? What SURMOUNT-4 and SURMOUNT-MAINTAIN Found
The two randomized trials that tested stopping tirzepatide — what people regained, how fast, and what happened to the group that lowered their dose instead of quitting.
Semaglutides.org is for information only and is not medical advice. Always talk to a licensed healthcare provider about your own care. Some links to telehealth services are affiliate links, labeled where they appear.
Two randomized trials have tested what happens when tirzepatide stops. One tested stopping outright. The other, published in 2026, tested lowering the dose instead. Together they are the clearest evidence available on a question that a lot of people face for reasons that have nothing to do with medicine — insurance changes, cost, supply, side effects, or simply reaching a goal.
This article describes what the trials found. It is not advice about whether to start, stop, continue or change a dose. That is a conversation for a healthcare provider.
What was SURMOUNT-4 and what did it find?
SURMOUNT-4 was built as a randomized withdrawal trial. It ran in two phases [1]:
Phase one (weeks 0 to 36). All 670 participants — adults with obesity or overweight plus a weight-related condition, excluding type 2 diabetes — took open-label tirzepatide at their maximum tolerated dose. Everyone knew they were on the drug. Average weight loss across this lead-in was 20.9%.
Phase two (weeks 36 to 88). Participants were randomized either to keep taking tirzepatide or to switch to a placebo injection, for 52 more weeks. Nobody knew which.
The results at week 88:
| Outcome | Continued tirzepatide | Switched to placebo |
|---|---|---|
| Weight change from week 36 | -5.5% | +14.0% |
| Kept 80% or more of lead-in loss | 89.5% | 16.6% |
| Total weight change from week 0 | -25.3% | -9.9% |
The difference between the groups over the withdrawal period was 19.4 percentage points.
Two details worth pulling out. First, the group that kept taking the drug kept losing — another 5.5% on top of an already substantial 20.9%. Second, the placebo group did not return all the way to their starting weight over those 52 weeks. They ended at -9.9% from baseline, which is still meaningful, but they were clearly moving in the wrong direction and the trial ended before we could see where they landed.
How does that compare with semaglutide?
The pattern is the same drug-class behavior, not something specific to tirzepatide.
STEP 4 ran the equivalent design with semaglutide: a 20-week lead-in on the active drug, then randomization to continue or switch to placebo. The continued group kept losing; the withdrawn group regained.
What differs is the scale. SURMOUNT-4 participants had further to fall because they had lost more to begin with — 20.9% over 36 weeks. A bigger loss followed by regain is not evidence that tirzepatide is worse at maintenance; it is evidence that there was more weight loss to lose.
What did SURMOUNT-MAINTAIN add?
SURMOUNT-MAINTAIN, published in The Lancet in May 2026, is the first randomized trial in this class to test a middle option: keep taking the drug, but at a lower dose.
The design [2][3]:
- 441 US adults with obesity or overweight plus a weight-related condition, excluding type 2 diabetes, enrolled at 20 sites. Average starting weight 113.8 kg, average BMI 40.1, 65% female, average age 46.6.
- Everyone took open-label tirzepatide at their maximum tolerated dose (10 mg or 15 mg) for 60 weeks.
- At week 60, participants who had lost at least 5% and tolerated the drug were randomized 3:3:2 to continue at maximum tolerated dose, drop to 5 mg, or switch to placebo — for 52 more weeks. 378 were randomized.
- From week 84 onward, anyone who regained more than half of their lost weight could receive rescue tirzepatide.
- 345 of 378 participants (91%) completed.
Weight change from baseline to week 112:
| Group | Weight change | Difference vs placebo |
|---|---|---|
| Maximum tolerated dose | -21.9% (95% CI -23.5 to -20.3) | -12.0 points (95% CI -13.8 to -10.1) |
| Stepped down to 5 mg | -16.6% (95% CI -18.0 to -15.1) | -6.6 points (95% CI -8.3 to -5.0) |
| Switched to placebo | -9.9% (95% CI -11.1 to -8.8) | — |
All comparisons reached p<0.0001.
Lilly’s own release translated this into absolute weights, which is easier to picture. Participants entered at about 112 to 113 kg. After 60 weeks on their maximum tolerated dose they were at roughly 89 kg. A year later, the group that stayed on that dose was at 88.7 kg — essentially unchanged. The group that dropped to 5 mg was at 94.6 kg, an average regain of about 5.6 kg [4].
How many people needed rescue treatment?
This is the number that makes the placebo result concrete.
Among participants who regained at least half their lost body weight, rescue tirzepatide was given to [2]:
- 11 of 138 (8%) in the maximum tolerated dose group
- 35 of 142 (25%) in the 5 mg group
- 60 of 90 (67%) in the placebo group
Two thirds of the people who stopped hit the regain threshold that triggered rescue within roughly six months of it becoming available.
It also explains how the placebo arm’s -9.9% figure should be read. The primary analysis used a modified treatment-regimen estimand that assumed people who started rescue treatment gained no further benefit from their assigned arm. Two thirds of that group took rescue tirzepatide, so -9.9% is not a picture of untreated regain. Without rescue, the placebo group’s number would very likely have been worse.
How fast does weight come back, in general?
SURMOUNT-4 gives one data point: 14.0% over 52 weeks. Broader evidence fills in the shape.
A 2026 meta-analysis in The BMJ pooled 37 studies with 63 treatment arms and 9,341 participants who stopped weight-management treatment. Average regain worked out to about 0.4 kg per month. Cardiometabolic markers — blood pressure, lipids, glucose measures — were projected to return to baseline within roughly 1.4 years. Regain was faster after stopping a drug than after finishing a behavioral program, by about 0.3 kg per month, and crucially the rate was independent of how much weight had been lost in the first place [5].
That last finding is worth sitting with, because it contradicts an older and intuitive idea that people who lose more regain proportionally more. The BMJ analysis says the rate is similar; it just takes longer to give back a larger loss.
What happens to blood sugar after stopping?
The SURMOUNT-1 prediabetes extension is the only randomized tirzepatide data on this.
Participants with obesity and prediabetes were treated for 176 weeks, then followed for 17 weeks off treatment. During treatment, 1.3% of the tirzepatide group had developed type 2 diabetes versus 13.3% on placebo (hazard ratio 0.07; 95% CI 0.0 to 0.1). After the 17-week off-treatment window, the figures were 2.4% versus 13.7% (hazard ratio 0.12; 95% CI 0.1 to 0.2) [6].
The protective gap narrowed but held over that short follow-up. Nobody has published what it looks like at one, two or five years off treatment.
Is there a way to taper off?
No randomized trial has tested one — for tirzepatide, for semaglutide, or for anything else in the class. Every withdrawal trial stopped the drug outright.
SURMOUNT-MAINTAIN is the closest thing available, and it is not a taper. It tested a single fixed step down from 10 or 15 mg to 5 mg, held for a year. The result was meaningfully better than stopping (-16.6% versus -9.9%) and meaningfully worse than staying put (-21.9%). That is useful information, but it does not tell you whether a slower, stepwise reduction would do better.
This is a genuine gap in the evidence, and the honest framing is that anyone considering a change should be working it out with their prescriber rather than with a trial table.
What about switching to a pill?
ATTAIN-MAINTAIN, reported alongside SURMOUNT-MAINTAIN in May 2026, tested a different exit route. It took participants from SURMOUNT-5 who had reached a weight plateau on maximum tolerated tirzepatide or semaglutide and moved them to once-daily oral orforglipron (brand name Foundayo, FDA-approved April 1, 2026) or placebo for 52 weeks [4][10].
Results:
- Switching from semaglutide at maximum tolerated dose: maintained all but 0.9 kg of prior weight loss.
- Switching from tirzepatide at maximum tolerated dose: maintained all but 5.0 kg.
The trial met its primary and all key secondary endpoints under both estimands: the switch groups held 82.4% (from semaglutide) and 78.0% (from tirzepatide) of the weight they had lost in SURMOUNT-5 [10]. The gap between the two switch groups is consistent with everything else here: people coming off tirzepatide had lost more, so a switch to a less potent agent gave back more.
How many people actually stop?
More than the trials would suggest, and usually not because a doctor told them to.
In a US commercial claims analysis of 33,607 adults without diabetes who started a weight-loss-indicated GLP-1 drug, one-year persistence on tirzepatide was 64.0% in 2023 and 64.8% in the first half of 2024. Semaglutide ran at 39.8% and 58.6% in the same years. Overall persistence across both drugs had nearly doubled since 2021, when it was 33.2% — the authors credit the end of supply shortages plus better handling of dose escalation and side effects [7].
A different picture emerges when coverage is guaranteed. In a Military Health System cohort of 10,649 active-duty service members, one-year persistence was 81.9% for tirzepatide, 70.7% for semaglutide and 34.2% for liraglutide [8]. That population has no insurance friction and close medical follow-up. The contrast suggests that a large share of real-world stopping is about access, not about the drug.
What does all this mean in practice?
Four things the trials support:
- Stopping outright reverses most of the benefit. SURMOUNT-4 found 14.0% regain over 52 weeks, with only one in six people holding most of their loss.
- Continuing holds it. In SURMOUNT-MAINTAIN, people who stayed on their maximum tolerated dose were at essentially the same weight a full year after the weight-loss phase ended.
- A lower dose is better than nothing but not as good as staying put. The 5 mg group gave back about 5.6 kg on average, and a quarter of them needed rescue treatment.
- Most people reach a plateau within about six to nine months anyway. In SURMOUNT-1 and SURMOUNT-4 pooled, median time to plateau ranged from 24.3 to 36.1 weeks by starting BMI, and 88% to 90% had plateaued by week 72 [9]. Continued treatment after that is about holding ground, not about losing more.
What none of it tells you is what you should do. That depends on why you are considering stopping, what else is going on medically, and what your options are — all of which belong in a conversation with a healthcare provider.
Sources
- SURMOUNT-4, JAMA 2024 — https://pubmed.ncbi.nlm.nih.gov/38078870/
- SURMOUNT-MAINTAIN, The Lancet 2026 — https://pubmed.ncbi.nlm.nih.gov/42119587/
- SURMOUNT-MAINTAIN abstract, The Lancet — https://www.thelancet.com/article/S0140-6736(26)00656-2/abstract
- Lilly release on SURMOUNT-MAINTAIN and ATTAIN-MAINTAIN, 2026-05-12 — https://investor.lilly.com/news-releases/news-release-details/lillys-foundayo-and-lower-dose-zepbound-helped-people-maintain
- Weight regain after cessation of medication for weight management, The BMJ 2026 — https://pubmed.ncbi.nlm.nih.gov/41500720/
- SURMOUNT-1 three-year prediabetes analysis, New England Journal of Medicine 2025 — https://pubmed.ncbi.nlm.nih.gov/39536238/
- GLP-1 persistence trends, Journal of Managed Care & Specialty Pharmacy 2026 — https://pubmed.ncbi.nlm.nih.gov/41760566/
- Military Health System GLP-1 cohort, Military Medicine 2026 — https://pubmed.ncbi.nlm.nih.gov/42424303/
- Time to weight plateau in SURMOUNT-1 and SURMOUNT-4, Clinical Obesity 2025 — https://pubmed.ncbi.nlm.nih.gov/39800653/
- ATTAIN-MAINTAIN, Nature Medicine 2026;32(7):2679-2687 — https://pubmed.ncbi.nlm.nih.gov/42120723/
Questions people ask
Do you gain the weight back after stopping Zepbound?
In SURMOUNT-4, participants who switched to placebo after 36 weeks of tirzepatide regained an average of 14.0% of body weight over the following 52 weeks. Those who stayed on the drug lost a further 5.5%. Only 16.6% of the placebo group kept at least 80% of their earlier weight loss, compared with 89.5% of those who continued.
How fast does the weight come back?
SURMOUNT-4 measured 14.0% regain over 52 weeks, which averages out to a steady climb rather than a sudden rebound. A large 2026 meta-analysis across 37 weight-management studies estimated average regain after stopping any weight-loss drug at about 0.4 kg per month.
Can you take a lower dose instead of stopping?
SURMOUNT-MAINTAIN tested exactly that. After 60 weeks of open-label treatment, participants who dropped to 5 mg were at -16.6% from baseline at week 112, compared with -21.9% for those who stayed on their maximum tolerated dose and -9.9% for those switched to placebo. The step-down group regained about 5.6 kg on average. Whether any dose change is appropriate is a decision for a prescriber.
Did anyone in the trials need rescue treatment?
Yes. SURMOUNT-MAINTAIN allowed rescue tirzepatide from week 84 for anyone who regained more than half their lost weight. It was used by 8% of the group that stayed on the full dose, 25% of the 5 mg group and 67% of the placebo group.
Is there a safe way to taper off Zepbound?
No randomized trial has tested a gradual taper for tirzepatide or for any drug in this class. Every withdrawal trial stopped the drug outright. SURMOUNT-MAINTAIN tested a single fixed step down to 5 mg, which is the closest evidence available. This is an evidence gap, not a recommendation either way.
How many people stop Zepbound on their own?
In US commercial claims data covering 33,607 adults without diabetes, one-year persistence on tirzepatide was 64.0% in 2023 and 64.8% in the first half of 2024. So roughly a third stopped within a year, even though the trials show what happens when treatment ends.
Does blood sugar go back up after stopping?
In the SURMOUNT-1 prediabetes extension, 17 weeks after treatment ended 2.4% of former tirzepatide participants had type 2 diabetes versus 13.7% of the placebo group. The gap narrowed compared with the on-treatment period but did not disappear over that short follow-up.
Can you switch to a pill instead of stopping?
ATTAIN-MAINTAIN tested moving SURMOUNT-5 participants from injections to once-daily oral orforglipron. Those switching from semaglutide maintained all but 0.9 kg of their weight loss over 52 weeks; those switching from tirzepatide maintained all but 5.0 kg.
This article summarizes FDA labeling, published research and company information current as of September 14, 2026. It is not medical advice and does not replace a conversation with your own healthcare provider. How we research and verify.