
Blog · Nutrition & Diet
What Happens When You Stop a Weight-Loss Drug?
By Vitae Team ·
Most of the weight returns, fast at first then slowing. How much you hold onto appears to track how much of the health benefit you keep.
Survey evidence suggests around one in fourteen UK adults has tried one of the newer weight-loss injections, and separate estimates indicate that most people using them pay privately. In January a large review put a number on what happens afterwards: around eighteen months back to where you started. Three further studies published since have complicated that, and the most useful of them suggests the question is not whether the weight returns but how much of it, because that appears to track how much of the health benefit returns with it.
TL;DR
- Regain is fast at first. Across 37 studies and 9,341 adults, weight returned at 0.4 kg a month after any weight-loss medicine and 0.8 kg a month after the newer ones.
- That review projected a return to starting weight in about 1.7 years, assuming regain carries on in a straight line.
- A second model says it levels off instead. Pooling six trials, about 60% of the lost weight was back at one year, with the curve flattening towards roughly 75%.
- Both numbers are projections. Nobody has followed people beyond 52 weeks on the current drugs.
- Regaining less went with keeping more of the benefit. In a tirzepatide withdrawal analysis, people who regained least generally held onto waist, triglyceride and insulin improvements. Those who regained most lost them.
- Blood pressure crept back in every group, including the people who held their weight.
- What prevents regain is barely tested. The big review found no evidence that behavioural support changed the rate, and the two trials that found something were small.
What the pooled evidence says
The review that set off the coverage came from Oxford's Nuffield Department of Primary Care Health Sciences and was published in the BMJ on 7 January 2026. It pooled 37 studies and 63 intervention arms covering 9,341 adults who came off a weight-management medicine after an average of 39 weeks, and followed them for an average of 32 weeks.
Weight returned at 0.4 kg a month across all the drugs, and 0.8 kg a month for semaglutide and tirzepatide. Projected forward, that put people back at their starting weight in about 1.7 years overall and 1.5 years for the newer drugs. The comparison that gave the review its headline was with behavioural weight-loss programmes: the medicines produced about 3.2 kg more loss to begin with, but regain afterwards ran 0.3 kg a month faster.
Only one of the 37 studies followed anyone beyond a year after stopping, and the newer-drug evidence comes from eight studies with at most twelve months of follow-up. Everything past that is a model that assumes regain continues at the same rate. Independent reviewers made the point plainly at the time: the full-regain estimate rests on extrapolation, and many included studies were not at low risk of bias.
A second model says it levels off
In March, a different group published a nonlinear meta-regression in eClinicalMedicine that asked what shape the regain curve takes rather than assuming a straight one. It reviewed 48 studies and used six randomised trials covering 3,236 people, with up to 52 weeks of follow-up after stopping. A straight line has no end point short of baseline. A curve that flattens does.
The pattern it found was an exponential recovery curve: rapid at first, then slowing. About 60% of the weight lost during treatment was back at one year, which rests on real follow-up rather than projection. The curve then flattens towards a modelled maximum of 75.3% regained (95% CI 68.9 to 81.6), which would leave roughly a quarter of the loss in place. For someone who lost 15 to 20% of their body weight the authors put that at about 4 to 5% kept, which they describe as sitting just below the 5 to 10% usually treated as clinically significant while still offering modest long-term benefit.
The authors are clear that the plateau is an extrapolation past 52 weeks, and the independent commentary was cautious. Adam Collins at Surrey called it an optimistic reading, noting that regain varied widely and that the people who regained least tended to have some ongoing support. Marie Spreckley at Cambridge's MRC Epidemiology Unit pointed out that only six trials informed the modelling and that body composition was not modelled at all, so whether the regained weight is fat, lean tissue or both is unknown.
Regaining less went with keeping more of the benefit
The most informative study here is a post hoc analysis of SURMOUNT-4, published in JAMA Internal Medicine. It followed 308 adults who had lost at least 10% of their body weight on tirzepatide over 36 weeks and were then switched to placebo for a further 52 weeks. By the end, 254 of them, or 82.5%, had regained at least a quarter of what they had lost.
Grouping people by how much they regained produces a clear gradient. Those who regained less than a quarter of their loss saw waist circumference, triglycerides, non-HDL cholesterol, fasting insulin and insulin resistance hold roughly where they had been at the end of treatment. Those who regained three quarters or more put 14.7 cm back on their waist, saw HbA1c rise by 0.35 percentage points and insulin resistance climb sharply, with the middle groups in between. This was a post hoc grouping by what happened rather than a randomised comparison, so it shows regain travelling with marker deterioration rather than proving it caused it.
One result did not follow the pattern. Systolic blood pressure rose in every group, by 6.8 mmHg even among those who held their weight best, which suggests some of the benefit depends on the treatment itself rather than on the weight. Worth knowing who ran it: several of the authors are employees of the drug's manufacturer, which is direct industry involvement rather than an advisory role.
The real-world picture looks milder
In March the Cleveland Clinic published a retrospective cohort in Diabetes, Obesity and Metabolism covering 7,938 adults in Ohio and Florida who had started injectable semaglutide or tirzepatide and stopped within three to twelve months. Those treated for obesity had lost 8.4% of their body weight by the time they stopped, and a year later had regained an average of 0.5%, with 45% still losing or holding steady.
That looks like a flat contradiction of everything above, and the study's own figures explain most of it. Of the people who stopped, 27% switched to a different medicine and 20% restarted the one they had been on. Close to half the cohort were not medication-free over the year being measured. Cost was the most common reason for stopping, ahead of side effects.
So the cohort describes what happens to people who stop one prescription, which is a different question from what happens to people who stop taking anything. It is still worth knowing, because stopping one prescription and finding another is what a lot of people do, and the trial literature has almost nothing to say about that path.
What has been tested to prevent it
Very little, and the results are mixed. The Oxford review looked for an effect of behavioural support and did not find one. Once initial weight loss was accounted for, the rate of regain did not differ by how much support people received during treatment, and the handful of studies comparing support with nothing afterwards showed no difference either. The authors were direct about it: they found no evidence that such support leads to slower regain. That rests on two studies, so it is thin rather than conclusive.
Two small trials point the other way. In October, a pilot randomised trial in Obesity Science & Practice put 39 adults who had just come off a GLP-1 drug into one of three groups: delivered meals, a commercial weight-management app, or usual care. Over four months the usual care group regained about 11% of body weight against about 3.6% in the two intervention arms, a difference between groups of roughly seven percentage points. The trial was designed to test whether people would stick with the programmes rather than to measure weight, the weight findings were exploratory, and the meals and subscriptions were free.
The other is Danish and older. After an eight-week diet producing an average 13.1 kg loss, 166 adults were randomised to a year of supervised exercise, a GLP-1 drug, both, or placebo, then followed for a year with no treatment at all. Regain in that untreated year was 6.0 kg larger after the drug alone than after supervised exercise, and the odds of still holding a 10% loss were 3.7 times placebo for exercise and 7.2 times for the combination. The drug was liraglutide, which is weaker than what people take now, and the exercise was supervised.
What This Actually Means
The honest summary is that the studies show substantial regain, concentrated in the first year, and that the straight-line version everyone quoted in January probably overstates the end point. One model suggests roughly a quarter of the loss might stay off in the longer term, but nobody has followed people far enough to establish that the plateau happens. That is a more hopeful reading than the all-or-nothing framing the coverage settled on, and it is still a projection.
What the SURMOUNT-4 analysis adds is that partial success may be worth something. People who held onto most of their loss also held onto most of the metabolic improvement, and those who regained most lost most of it. That is an association rather than a demonstrated cause, and blood pressure did not follow it, but it shifts the question from whether you can keep it all off, which almost nobody does, to how much you can hold.
On what helps, the evidence is unsettled and anyone claiming otherwise is ahead of it. The largest analysis found nothing for behavioural support. Two small trials found something for structured support and for exercise, both with real limitations. NICE quality standards recommend post-treatment support for at least a year, which is a sensible precaution rather than a demonstrated result, and most UK users are buying privately without any of it attached.
Frequently Asked Questions
Do you put all the weight back on?
Probably not all of it. One model projects a return to starting weight in about 1.7 years; another finds the curve flattening at around 75% of the loss regained, leaving roughly a quarter held. Neither has been checked past 52 weeks on the current drugs.
How fast does it come back?
Fastest in the first few months. Pooled trial data gives 0.8 kg a month for semaglutide and tirzepatide, and the meta-regression found about 60% of the loss back by one year before the curve starts to flatten.
Do the health improvements go too?
Largely in proportion to the weight. People who regained least generally held onto waist, triglyceride and insulin improvements, and those who regained most lost them. That is an association from a post hoc analysis rather than a controlled comparison, and blood pressure was the exception: it crept up in every group.
Does anything stop the regain?
Nothing is established. The largest review found no effect of behavioural support. A 39-person pilot found less regain with delivered meals or an app, and an older Danish trial found less regain after supervised exercise. All three are small or indirect.
Why do people stop?
In the Cleveland Clinic cohort, cost was the most common reason, ahead of side effects. Real-world data suggests around half of people stop within twelve months, and many of those start something else.
The Bottom Line
The eighteen-month figure that travelled in January is a projection from twelve months of data, and a second model published two months later suggests regain slows rather than running to completion, leaving perhaps a quarter of the loss in place. Both are extrapolations. What is not in doubt is that appetite returns when the thing suppressing it is withdrawn, and that most people regain most of what they lost.
The finding worth carrying away is the one from the tirzepatide withdrawal data: metabolic improvement largely tracked the weight, so holding onto part of the loss went with holding onto part of the benefit. What reliably helps people do that has not been established, which is an uncomfortable answer but the accurate one.
This article is for general information and is not medical advice. Do not start, stop or change a prescribed medicine without speaking to your GP or prescriber.
Sources
- 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. https://doi.org/10.1136/bmj-2025-085304
- Budini B, Luo S, Tam M, Stead I, Lee A, Akrami A, Vidal-Puig A, Park A. Trajectory of weight regain after cessation of GLP-1 receptor agonists: a systematic review and nonlinear meta-regression. eClinicalMedicine, 2026. (No funding source declared.) https://doi.org/10.1016/j.eclinm.2026.103796
- Horn DB, Linetzky B, Davies MJ, et al. Cardiometabolic parameter change by weight regain on tirzepatide withdrawal in adults with obesity: a post hoc analysis of the SURMOUNT-4 trial. JAMA Internal Medicine, 2026;186(2):157-167. https://doi.org/10.1001/jamainternmed.2025.6112
- Gasoyan H, et al. Obesity treatments and weight changes in clinical practice after discontinuation of semaglutide or tirzepatide. Diabetes, Obesity and Metabolism, 2026. https://doi.org/10.1111/dom.70660
- Hollis-Hansen K, Vonderschmidt A, Haskins C, Kane RM, Almandoz JP, Lingvay I. Preventing weight regain after glucagon-like peptide-1 therapy discontinuation: a pilot randomized controlled trial. Obesity Science & Practice, 2026;12(5):e70194. https://doi.org/10.1002/osp4.70194
- Jensen SBK, Blond MB, Sandsdal RM, et al. Healthy weight loss maintenance with exercise, GLP-1 receptor agonist, or both combined followed by one year without treatment: a post-treatment analysis of a randomised placebo-controlled trial. eClinicalMedicine, 2024;69:102475. https://doi.org/10.1016/j.eclinm.2024.102475
- Science Media Centre. Expert reaction to meta-analysis and modelling study on weight regain after stopping GLP-1 receptor agonists, 4 March 2026. https://www.sciencemediacentre.org/expert-reaction-to-meta-analysis-and-modelling-study-on-weight-regain-after-stopping-glp-1-receptor-agonists/
- The Food Foundation. The GLP-1 Revolution: implications for health, inequality and the food system, April 2026. (YouGov survey of 6,055 UK adults, fieldwork 23 January to 3 February 2026.) https://foodfoundation.org.uk/sites/default/files/2026-04/The%20GLP-1%20Revolution_briefing_v4_0.pdf