Weight regain after stopping GLP-1 treatment

Weight regain occurred after treatment withdrawal in studies of semaglutide and tirzepatide. The findings support planning follow-up and affordability before an interruption, but they do not predict exactly what will happen to one person.

Reviewed by Ellis Brandt. Sources checked October 2026.

Educational information, not individualized medical, insurance or legal advice. Discuss treatment decisions with your clinician.

In this guide

What the semaglutide follow-up found

The STEP 1 extension followed a subset of adults without diabetes after 68 weeks of treatment with semaglutide 2.4 mg or placebo. Both study treatment and the lifestyle intervention ended. In the semaglutide group, mean weight loss at week 68 was 17.3%; over the following year, participants regained 11.6 percentage points, leaving a mean loss of 5.6% from baseline.

The report describes this as regaining about two-thirds of the prior weight loss. That is a group average in an extension involving 327 participants in the analysis, not a statement that every patient regains two-thirds. It also cannot isolate the effect of stopping medication from stopping the structured lifestyle intervention.

Source: Wilding and colleagues: STEP 1 extension, Diabetes, Obesity and Metabolism (2022).

What the tirzepatide withdrawal trial found

In SURMOUNT-4, 670 participants who completed a 36-week tirzepatide lead-in were randomized to continue treatment or switch to placebo for 52 weeks. Mean weight change from randomization was a further 5.5% reduction with continued tirzepatide versus a 14.0% increase after switching to placebo. Lifestyle intervention continued in both groups.

Participants were adults with obesity or overweight and a weight-related complication, without diabetes. The randomized group had already completed the lead-in on a maximum tolerated dose of 10 or 15 mg. The result therefore does not describe everyone who starts treatment, including people who stop early or cannot tolerate it.

Source: Aronne and colleagues: SURMOUNT-4 randomized trial, JAMA (2024).

Why the percentages are not interchangeable

The STEP 1 extension's 11.6 percentage points refers to change relative to the original baseline. SURMOUNT-4's 14.0% is weight change from the later randomization point, after the lead-in loss. Different starting points, study designs and follow-up support make these numbers unsuitable for a simple claim that one drug causes more regain than the other.

Neither study tested every stopping strategy, every patient group or every modern formulation. They do not establish a tapering schedule for an individual, and they are not a reason to ignore a clinician's recommendation to stop for a medical concern. A treatment decision needs more information than the average weight curve.

Questions to discuss before an interruption

Ask your clinician what monitoring and follow-up would be appropriate if treatment ends, what changes should prompt contact and what options fit your medical situation. If cost is the problem, explain the timing before the next planned prescription so the discussion can include realistic access options.

Who will manage follow-up if the subscription ends?
What care is appropriate if insurance stops covering the prescription?
What records should you keep if you change providers?
What is the expected ongoing cost after promotional pricing ends?

Include follow-up in the budget

A first-month price cannot answer the financial question raised by long-term care. Compare the recurring medication and visit costs, the renewal terms and the cost of follow-up after leaving the service. Do not buy a longer commitment solely because a trial reported regain: a prepaid contract is a financial decision, while continued treatment is a clinical one.

These studies help frame a conversation about maintenance and access. They do not mean that regain is a personal failure or that everyone must use the same treatment indefinitely. The useful next step is a plan with your clinician that fits your health and circumstances.