Semaglutide, tirzepatide and the rest of the GLP-1 drugs have completely transformed obesity treatment, with clinical trials showing losses of up to 20% of body weight. But a meta-analysis published in The BMJ, pooling 37 studies and more than 9,300 participants, puts exact numbers on what happens once people stop taking them, and the result is stark: the weight comes back, and it comes back fast.
The body regains lost weight at an average pace of 0.4 kilos a month. Researchers at the University of Oxford analyzed 37 clinical trials and observational studies (63 treatment groups, 9,341 people in total) that had tracked what happens to weight after stopping any obesity medication, from orlistat to the newest drugs. Using three different statistical models that arrived at consistent results, the conclusion was that at that pace, weight returns to baseline in an average of 1.7 years.
With semaglutide and tirzepatide, the most effective drugs, the rebound is even faster. While with older-generation drugs weight returns at a rate of 0.5 kilos a month, with semaglutide (Ozempic, Wegovy) and tirzepatide (Mounjaro, Zepbound) that figure rises to 0.8 kilos a month. These are precisely the drugs that produce the biggest losses (an average of 14.7 kilos during treatment), so they’re also the ones that bounce back fastest once stopped: in just 1.5 years, weight is back where it started.
It’s not just weight that reverses. The study also tracked metabolic markers linked to obesity: glycated hemoglobin, fasting glucose, total cholesterol, triglycerides and blood pressure. All of them improved during treatment, and all of them returned to baseline within 1.4 years of stopping the medication. In other words, the cardiometabolic benefits these drugs provide aren’t permanent, they depend on continuing to take them.
The comparison with diet and exercise leaves an uncomfortable finding. The same researchers had previously studied what happens after stopping a lifestyle-change program (diet and physical activity, no drugs): there, weight regain was just 0.1 kilos a month, and it took nearly 4 years for weight to return to baseline. Even though drugs produce a bigger initial loss (3.2 kilos more on average), the body regains it considerably faster than when the change comes from habits alone, even comparing people who lost the same amount of weight through either path.
Nearly half of people stop treatment before a year, in real-world practice. The authors themselves cite data from the US and Denmark placing the discontinuation rate at around 50% at twelve months, far below the duration of the clinical trials that got these drugs approved. And a survey of the US population revealed something telling: interest in using these medications fell from 45% to 14% once people were told the weight comes back after stopping.
The warning the authors raise isn’t alarmist, it’s about management. They’re not saying these drugs don’t work, the data show exactly the opposite while they’re being used. The important nuance is different: since obesity is a chronic condition, treating it as if it were something temporary (a few months of medication and done) isn’t backed by the evidence. The study found no evidence that more behavioral support during or after treatment slowed the pace of weight regain, so the question of how to sustain results long-term remains open.
Limitations worth noting. Only 12 of the 37 studies had a low risk of bias on the Cochrane scale, and the direct comparison between drugs and lifestyle programs is indirect, drawn from two separate reviews rather than a single trial comparing both groups at once. For semaglutide and tirzepatide, moreover, there’s barely any follow-up data beyond one year, so the 1.5-year projection relies on extrapolating the observed trend rather than having measured that exact point in most participants.
About this article. This piece is based on the scientific study “Weight regain after cessation of medication for weight management: systematic review and meta-analysis,” by Sam West, Jadine Scragg, Paul Aveyard, Jason L Oke, Lia Willis, Stella J P Haffner, Heather Knight, Danni Wang, Sarah Morrow, Laura Heath, Susan A Jebb and Dimitrios A Koutoukidis, published in 2026 in The BMJ, open access under a CC BY license. DOI: 10.1136/bmj-2025-085304.
