One of the most common questions about GLP-1 medications is also one of the least comfortable: what happens when treatment ends? Researchers at the University of Cambridge tried to put numbers on it, and the result is worth reading closely, including the parts the headline version leaves out.
What the analysis did
The team, publishing in eClinicalMedicine in March 2026, searched the literature and identified 48 studies. Only six randomized trials, covering 3,236 participants, reported data in a form that could be used for modeling. The authors fit a nonlinear curve to those trials to describe how weight changed after people stopped a GLP-1 receptor agonist.
The authors did not find a cliff. They found a predictable, decelerating curve:
- About 60% of the weight lost during treatment had returned by 52 weeks after stopping.
- The model’s half-life for regain was roughly 23 weeks, meaning the gap closes fastest early and slows down.
- Extrapolating the curve forward, it leveled off at about 75% of the original loss regained (95% confidence interval 68.9% to 81.6%).
In plain terms: on average, in these trials, most of the change reversed, and the reversal was gradual rather than immediate.
What it does not establish
The 75% figure is an extrapolation. The trial data stop at 52 weeks after cessation. Where the curve flattens is a model’s estimate, not something anyone observed over several years.
Six trials is a small base. The authors note they could not formally assess publication bias with so few studies. They also limited the analysis to trials where average weight loss was at least 3 kg, which may leave out people who responded modestly.
An average is not your number. The curve describes groups. Individual trajectories vary, and the analysis could not account for things that plausibly matter, such as other medications, other health conditions, or what people did with diet and activity after stopping, because the trials did not report them consistently.
Trial participants are not everyone. People in randomized trials are selected, monitored, and often receiving structured lifestyle support. That can cut either way for how well the curve transfers to real life.
It does not say regain is inevitable or that treatment “failed.” Obesity is widely treated by medical bodies as a chronic condition, and in that framing, weight returning when a treatment stops looks less like a surprise and more like what happens with many long-term therapies. The study describes a pattern. It does not judge anyone.
Other signals point the same direction, with nuance
A large real-world analysis from Cleveland Clinic, published in Diabetes, Obesity and Metabolism and summarized in a March 2026 release, reported that many patients who discontinued a GLP-1 were able to stabilize their weight over the following year, often with other obesity treatments or structured lifestyle interventions. That is observational data from one health system, so it cannot show that any particular approach caused the stabilization. But it is a useful counterweight to reading the Cambridge curve as fixed.
Behavioral programs aimed at maintenance after stopping are also being studied. Early-stage trial registrations exist, but results are not in yet, so treat any claim that a specific program prevents regain as unproven.
The part that applies to compounded medications
Every trial in this analysis studied a branded, manufactured GLP-1 product under its manufacturer’s protocol. That evidence does not transfer to a compounded preparation.
Compounded semaglutide and tirzepatide, which NoBsRx provides provider-guided access to, are prepared by state-licensed pharmacies for individual patients. They are not FDA-approved, and they have not been through these trials. Nobody has measured a regain curve for a compounded product. What the research supports is a narrower statement: for the ingredient class studied in branded form, weight tended to return gradually after treatment ended, on average.
What is reasonable to take from it
None of this is advice to stay on or come off any medication. But a few questions follow naturally, and they are worth raising with a licensed provider before, not after, a decision to stop:
- What is the plan if treatment ends, whether by choice, cost, or supply?
- Which habits and supports are in place to carry forward, and which were the medication doing alone?
- Is a gradual change or a different maintenance approach appropriate for your situation?
This is part of why how a program is structured matters, and why we lay out what it costs and commits you to in writing. You should know before starting what follow-up looks like, not learn it at the end.
This post is general information, not medical advice. It does not describe an outcome you should expect from any NoBsRx program. Compounded medications are not FDA-approved. Treatment is never guaranteed. Eligibility and every treatment decision, including starting, changing, or stopping a medication, are made by independent licensed providers based on an individual clinical assessment, and availability varies by state. Talk to a licensed clinician before making changes to your care.