What Happens When You Stop a GLP-1

Disclosure: Metabolic Media is an independent educational publisher. We may have a financial affiliation with some entities featured on this site. Rank positions are determined by our published methodology. We do not provide medical care, diagnoses, or treatment.

The short answer: In the STEP 1 extension, 228 adults who had taken semaglutide regained about two-thirds of their prior weight loss within a year of stopping, as the authors describe it. In SURMOUNT-4, 335 adults switched from tirzepatide to placebo regained 14.0% of body weight over 52 weeks (670 were randomized). Both were trials in selected groups; individual results varied. The two trials differ in design and aren’t directly comparable.

This research summary covers what withdrawal trials and cohort studies report, how often people stop in real-world data, and what hasn’t been tested. Every number is attributed to its source. You’ll find more in our After GLP-1 coverage.

Why do researchers study what happens after stopping?

Most weight-management trials follow people while they take a medicine. Withdrawal designs ask what happens next. Some trials stop treatment for everyone at a set point, and others randomize people to continue or to switch to placebo.

This piece summarizes research. It isn’t medical advice, it doesn’t say whether anyone should stop or continue a medicine, and it doesn’t cover dosing, side effects, or price.

What did the STEP 1 extension find after semaglutide was stopped?

Design. STEP 1 was a randomized trial in 1,961 adults with overweight or obesity and without diabetes. Its off-treatment extension followed a subset of participants for a further year after treatment and structured lifestyle support both ended at week 68 (Wilding et al., 2022). The extension analyzed 327 adults (228 in the semaglutide arm, 99 in the placebo arm) from selected sites in five countries. All of its analyses were exploratory.

Weight. Mean weight loss at week 68 was 17.3% with semaglutide and 2.0% with placebo. By week 120, the semaglutide arm had regained 11.6 percentage points and the placebo arm 1.9, leaving net losses from baseline of 5.6% and 0.1%. The authors’ conclusion says participants “regained two-thirds of their prior weight loss.” In the semaglutide arm, observed mean body weight was 105.6 kg at baseline, 87.5 kg at week 68, and 99.0 kg at week 120 (197 participants measured). At week 120, 48.2% of the semaglutide arm (95 of 197) and 22.6% of the placebo arm (21 of 93) had kept off at least 5% of their baseline weight.

Metabolic markers. In the semaglutide arm, observed mean systolic blood pressure was 129, 121, and 131 mmHg at weeks 0, 68, and 120. HbA1c was 5.7%, 5.2%, and 5.6% at the same points. Among participants with prediabetes at baseline, 93.6% of the semaglutide arm and 41.5% of the placebo arm had a normal HbA1c at week 68. By week 120 the figures were 43.3% and 34.0%. The authors report that most cardiometabolic improvements moved back toward baseline by week 120, while a few measures (HDL cholesterol, VLDL cholesterol, triglycerides, and C-reactive protein) stayed better than in the placebo arm. For background on how these blood-sugar measures work, see Insulin Resistance, Explained.

Limits the authors state. The extension was a small subset of the main trial, the choice of sites could introduce selection bias, no power calculation was done, no formal significance tests were run, and participation in lifestyle programs after week 68 wasn’t recorded. STEP 1 and its extension were funded by Novo Nordisk, and several authors are Novo Nordisk employees.

What did SURMOUNT-4 find after tirzepatide was stopped?

Design. SURMOUNT-4 was a randomized withdrawal trial at 70 sites in four countries (Aronne et al., 2024). Adults without diabetes entered a 36-week open-label lead-in on tirzepatide (783 adults). At week 36, 670 were randomized to continue (335) or to switch to placebo (335) for 52 weeks. Among the 670 randomized, mean age was 48, 71% were women, and mean weight was 107.3 kg.

Results. Mean weight reduction before randomization was 20.9%. From week 36 to week 88, weight changed by -5.5% (-4.7 kg) in the group that continued and by +14.0% (+11.1 kg) in the group switched to placebo. The reported difference was -19.4 percentage points (95% CI -21.2 to -17.7). Measured from week 0 to week 88, the changes were -25.3% and -9.9%. At week 88, 89.5% of the continuing group (300 participants) and 16.6% of the placebo group had kept at least 80% of their lead-in weight loss.

Limits the authors state. The design didn’t allow adjustments to treatment after randomization, the trial didn’t evaluate intensive behavioral support, and people who tolerated the lead-in may not represent everyone. The trial was sponsored by Eli Lilly, which took part in its design, analysis, and manuscript, and several authors are Lilly employees.

How to read these two sections. The STEP 1 extension and SURMOUNT-4 are separate studies. They differ in population, length, design (an extension in which both arms stopped, versus a randomized withdrawal after a lead-in), and whether lifestyle support continued, so their numbers aren’t comparable.

What did the STEP 4 withdrawal trial show?

STEP 4 is a separate trial with its own design, reported in JAMA in 2021 and not directly comparable with the two above (Rubino et al., 2021). It randomized 803 adults, after a 20-week run-in in which mean weight loss was 10.6%, to continue semaglutide or switch to placebo for 48 weeks, with a lifestyle intervention in both groups. From week 20 to week 68, weight changed by -7.9% in the continuing group and +6.9% in the placebo group (difference -14.8 percentage points; 95% CI -16.0 to -13.5). Novo Nordisk funded the trial, and its representatives took part in the design, analysis, and manuscript.

What does pooled research say about weight regain?

A systematic review and meta-analysis published in the BMJ in January 2026 pooled 37 studies (63 arms, 9,341 participants) of people who stopped weight-management medicines (West et al., 2026). Average treatment lasted 39 weeks and average follow-up 32 weeks, with searches through February 2025. Many arms came from non-randomized or observational studies.

All medicines pooled: average regain of 0.4 kg per month (95% CI 0.3 to 0.5), or an estimated 4.8 kg in the first year. Average loss at stopping was 8.3 kg.
Incretin-based medicines (32 arms, 4,757 participants): 0.5 kg per month, 6.0 kg in the first year. Average loss at stopping was 10.1 kg.
A subgroup the authors call newer, more effective incretin medicines (10 arms, 1,776 participants): 0.8 kg per month, 9.9 kg in the first year. Average loss at stopping was 14.7 kg.

These are the authors’ groupings, reported as groups rather than as named medicines, and they aren’t a ranking.

The authors’ projection. The authors project a return to baseline weight at about 1.7 years (95% CI 1.3 to 2.1) across all medicines, and at 1.5 years (1.0 to 1.9) for the newer-medicines subgroup. They also project that cardiometabolic markers return to baseline within 1.4 years. These are the authors’ projections, not observed results. Data for the newer medicines came from 8 studies with at most 12 months of follow-up, so that projection is extrapolated. Only one included study followed people for two years after stopping, and the authors rated the regain rate as moderate certainty.

Among 28 randomized trials, the authors’ pooled result was that weight did not differ between treatment and control groups 1.4 years after stopping. The review was funded in part by the NIHR Oxford Biomedical Research Centre, and the paper says the funder had no role.

How often do people stop in real-world data?

Rodriguez and colleagues studied 125,474 US adults with overweight or obesity who started liraglutide, semaglutide, or tirzepatide between 2018 and 2023, using electronic health records from Truveta, a data collective, in a retrospective cohort (Rodriguez et al., 2025). Several of the authors are affiliated with Truveta Inc. Within one year, 64.8% (95% CI 64.4 to 65.2) of adults without type 2 diabetes had stopped, as had 46.5% (46.2 to 46.9) of adults with it. Among 41,792 people who stopped and had a weight measurement, 36.3% of those without type 2 diabetes and 47.3% of those with it restarted within a year.

Moderate or severe gastrointestinal events were associated with a higher hazard of stopping (hazard ratios 1.19 without type 2 diabetes and 1.38 with it). The authors note that ongoing use may be limited by issues including tolerability, access, and cost, and, among adults with type 2 diabetes, they found higher income associated with a lower rate of stopping.

Limits: the study didn’t model medicine shortages, it defined stopping as 60 days without medicine on hand, and some people may be misclassified. Not every stop is a choice.

What are the limits of the evidence?

Group averages hide variation. Withdrawal trials include people who completed treatment, and individual courses varied widely. In the STEP 1 extension, the standard deviation of regain was 7.7 percentage points.
Follow-up is short. It runs from about one to two years at most, so long-run patterns are unknown. The STEP 1 extension authors say longer follow-up is needed to tell whether weight plateaus below baseline or simply regains more slowly.
Support differed. Trials differ in whether lifestyle support continued after stopping.
Trials aren’t real life. Stopping in a trial isn’t the same as stopping because of cost, supply, or side effects.
No strategy was tested. No trial cited here tested a way to limit regain after stopping.
Markers aren’t outcomes. Average changes in blood pressure, HbA1c, lipids, and CRP are markers, not proof of future health events.
Funding matters. Novo Nordisk funded the STEP 1 extension and STEP 4, and Eli Lilly sponsored SURMOUNT-4.

For a separate look at research on protein and lean mass during GLP-1 weight loss, see Protein on a GLP-1.

What questions can you bring to your care team?

Questions that can start the conversation:

What is my starting point, and how will we track changes over time?
How would we check blood pressure, blood sugar, and lipids?
What do these averages mean, and not mean, for my situation?
What support is available if cost, supply, or side effects come up?

A clinician who knows your history can help answer them.

Frequently asked questions

Do people regain weight after stopping a GLP-1?

In trials, on average, yes. In the STEP 1 extension (228 adults in the semaglutide arm), participants regained 11.6 percentage points of body weight in the year after stopping semaglutide. In SURMOUNT-4 (335 adults switched to placebo), those regained 14.0%. Individual results varied.

Do people regain all of the weight?

Not on average in these trials. In the STEP 1 extension, the semaglutide arm’s mean weight was 5.6% below baseline at week 120. In SURMOUNT-4, the placebo-switch group was 9.9% below baseline at week 88. A 2026 BMJ review projected a return to baseline weight at about 1.5 to 1.7 years, which the authors say is an extrapolation.

What happens to blood pressure and blood sugar after stopping?

In the STEP 1 extension, average systolic pressure and HbA1c in the semaglutide arm rose back toward starting values by week 120, while a few other markers stayed better than in the placebo arm. Markers are not health outcomes.

How many people stop taking GLP-1 medicines?

In one US cohort of 125,474 adults, 64.8% of those without type 2 diabetes and 46.5% of those with it had stopped within a year. Reasons vary, and not every stop is a choice.

Should I stop or keep taking mine?

This page can’t answer that. It summarizes research, and that decision belongs to you and your clinician.

Looking for more? Browse more metabolic health explainers on the Read hub.

Sources

Wilding JPH, Batterham RL, Davies M, et al. Weight regain and cardiometabolic effects after withdrawal of semaglutide: The STEP 1 trial extension. Diabetes Obes Metab 2022;24(8):1553-1564. doi:10.1111/dom.14725. (PubMed) Funded by Novo Nordisk; several authors are Novo Nordisk employees.
Aronne LJ, Sattar N, Horn DB, et al. Continued Treatment With Tirzepatide for Maintenance of Weight Reduction in Adults With Obesity: The SURMOUNT-4 Randomized Clinical Trial. JAMA 2024;331(1):38-48 (published Jan 2, 2024). doi:10.1001/jama.2023.24945. (PubMed) Sponsored by Eli Lilly, which was involved in design, analysis, and manuscript; several authors are Lilly employees.
Rubino D, Abrahamsson N, Davies M, et al. Effect of Continued Weekly Subcutaneous Semaglutide vs Placebo on Weight Loss Maintenance in Adults With Overweight or Obesity: The STEP 4 Randomized Clinical Trial. JAMA 2021;325(14):1414-1425 (published Apr 13, 2021). doi:10.1001/jama.2021.3224. (PubMed) Funded by Novo Nordisk, whose representatives took part in design, analysis, and manuscript.
West S, Scragg J, Aveyard P, et al. Weight regain after cessation of medication for weight management: systematic review and meta-analysis. BMJ 2026;392:e085304 (published Jan 7, 2026). doi:10.1136/bmj-2025-085304. (PubMed) Funded in part by the NIHR Oxford Biomedical Research Centre; the funder had no role. Return-to-baseline figures are the authors’ projections.
Rodriguez PJ, Zhang V, Gratzl S, et al. Discontinuation and Reinitiation of Dual-Labeled GLP-1 Receptor Agonists Among US Adults With Overweight or Obesity. JAMA Netw Open 2025;8(1):e2457349. doi:10.1001/jamanetworkopen.2024.57349. (PubMed) The data came from Truveta, and several authors are affiliated with Truveta Inc; the paper’s conflict-of-interest disclosures are listed in the article. No study funder is named in the text we reviewed.

This article is for general education. It isn’t a substitute for personal medical advice. If you have questions about weight-management medicines or your own health, talk with a licensed clinician who knows your health history.

Medical disclaimer: Content on Metabolic Media is for educational purposes only and is not medical advice, diagnosis, or treatment. We are not a medical provider, pharmacy, or telehealth clinic. Always consult a licensed clinician. If you have a medical emergency, call 911 (US) or your local emergency number.