After stopping obesity drugs, weight returns about four times faster than after a diet

That people regain weight after stopping obesity medication has been known since the first withdrawal extensions of the semaglutide trials. What has not been available is a rate — how fast, compared with what, and whether the metabolic improvements go with it.

A systematic review and meta-analysis published in The BMJ on January 7 supplies one [s1].

The headline numbers

Across 37 studies comprising 63 intervention arms and 9,341 adults, weight increased by an average of 0.4 kg per month (95% CI 0.3 to 0.5) after weight-management medication was stopped [s1]. Mean treatment duration in the included studies was 39 weeks, with follow-up after cessation averaging 32 weeks [s1].

The rate is not uniform by drug class. For incretin mimetics as a group it was 0.5 kg per month (0.4 to 0.7); for the newer incretin mimetics — semaglutide and tirzepatide — it was 0.8 kg per month (0.7 to 0.9) [s1].

Projected forward, the models put return to pre-treatment weight at 1.7 years (95% CI 1.3 to 2.1) across all weight-management medications, 1.6 years (1.1 to 2.1) for incretin mimetics, and 1.5 years (1.0 to 1.9) for the newer ones [s1].

The comparison that has driven the coverage is with behavioural weight-management programmes. Regain was faster after medication by 0.3 kg per month (0.22 to 0.34), and the review reports this held independent of how much weight had been lost during treatment [s1]. Time to return to baseline after a behavioural programme was 3.9 years (95% CI 2.8 to 4.9), against 1.7 years after medication [s1].

The cardiometabolic markers move too

This is the part of the paper that goes beyond weight. All measured cardiometabolic markers were projected to return to baseline within 1.4 years of stopping [s1]. The monthly rates of change were: HbA1c up 0.05 mmol/mol (0.03 to 0.08); fasting glucose up 0.06 mmol/L (0.03 to 0.08); systolic blood pressure up 0.5 mm Hg (0.3 to 0.7); diastolic up 0.2 mm Hg (0.1 to 0.3); total cholesterol up 0.05 mmol/L (0.03 to 0.07); triglycerides up 0.03 mmol/L (0.01 to 0.04) [s1].

Restricting to randomised controlled trials only (28 studies), regain relative to control was 0.3 kg per month (0.3 to 0.4), with time to equivalence with control at 1.4 years (0.9 to 1.8) [s1].

What the analysis cannot carry

The authors are direct about several limitations, and the most important one concerns exactly the drugs readers will care about most.

Only eight of the 37 studies assessed the newer GLP-1-based drugs, with a maximum of 12 months of follow-up after cessation [s2], and only one study included follow-up beyond one year for that category [s1]. The 1.5-year projection for semaglutide and tirzepatide therefore extends past the observed data [s1]. A projection is not an observation.

The time-to-event models assume linear regain trajectories, though the authors report a sensitivity analysis found no evidence of departure from linearity [s1]. Regain in the real world could plateau or accelerate in ways a linear model would miss.

The comparison with behavioural programmes is indirect — the two literatures were reviewed separately and the populations, while similar, are not the same people [s1]. Indirect comparisons carry more uncertainty than the confidence intervals around either estimate suggest on its own.

Study quality was mixed: only 12 of 35 randomised trials were judged at low risk of bias [s1]. Funnel plot inspection showed little evidence of publication bias [s1].

How to read it

The finding is not that obesity drugs do not work. Within the treatment period they demonstrably reduce weight and improve the markers listed above; that is why those markers have somewhere to regress from.

The finding is about what happens when treatment stops, and the framing offered by the authors is that despite success in achieving initial weight loss, these drugs alone may not be sufficient for long-term weight control [s2].

That has a straightforward reading and a less obvious one. The straightforward reading is that obesity behaves like other chronic conditions — treatment suppresses the phenotype rather than curing it, and withdrawal returns it, in the same way blood pressure returns after antihypertensives stop. The less obvious reading concerns the faster regain after drugs than after behavioural programmes, which the review reports as independent of the amount lost [s1]. If that is robust, something about pharmacological weight loss differs from behavioural weight loss in what it leaves behind when it ends. This analysis identifies the pattern; it does not explain it.

What to watch

Longer withdrawal follow-up for semaglutide and tirzepatide specifically, since the current evidence base for those drugs runs to roughly a year after stopping [s2]; whether regain trajectories stay linear over longer periods; and whether trials of tapering, intermittent dosing, or medication-plus-behavioural-maintenance strategies change the curve.

This article describes a published meta-analysis and is informational only. It is not medical advice, and it does not recommend starting, continuing, changing, or stopping any medication. Those decisions belong with a clinician.

Sources

• [s1] West S, Scragg J, Aveyard P, et al., Weight regain after cessation of medication for weight management: systematic review and meta-analysis, The BMJ, 2026;392:e085304, published 2026-01-07.
• [s2] Stopping weight loss drugs linked to weight regain and reversal of heart health markers, BMJ Group press release, 2026-01-07.