If you've lost weight on semaglutide (Ozempic, Wegovy), tirzepatide (Mounjaro, Zepbound) or another GLP-1, the question that matters most isn't really about the medication. It's about the months after it — whether you stop by choice, by cost, or because your prescriber moves you off it. That's when the result either holds or it doesn't, and which one happens has almost nothing to do with willpower.
Why regain happens
Two things change when the medication stops, and they happen at the same time.
The first is appetite. GLP-1s work partly by quieting hunger signalling in the brain, and the leading explanation for what happens next is that this effect fades once the drug clears your system: a 2024 review in the Journal of Obesity describes rising ghrelin (a hunger hormone), falling leptin, and a general "deregulation" of appetite control in the brain once the medication's suppression is withdrawn. The review's own language is careful — it says this "has been posited," not proven outright — but it's the best current account of why hunger comes back, sometimes louder than before.
The second is arithmetic, and it needs no study to explain: you're lighter than you were, and a lighter body burns fewer calories. The eating pattern that once maintained your old weight will grow your new one. Combine that with appetite returning, and the gap between what you eat and what you burn opens from both sides at once.
How much of the loss actually comes back is documented in two randomized trials, though both are trial populations, not a guarantee about any one person. In the STEP 1 trial extension — adults with obesity or overweight, without diabetes, who'd lost an average 17.3% of body weight on semaglutide 2.4 mg — participants who stopped the medication regained a mean two-thirds of that loss over the following year, off treatment entirely. In SURMOUNT-4, a similar population completed a 36-week tirzepatide lead-in and was then randomized either to keep taking it or switch to placebo: over the next year, the group that kept taking it lost a further 5.5% on average, while the group switched to placebo regained 14.0% — only 16.6% of the placebo-switch group held onto at least 80% of what they'd lost, against 89.5% of the group that stayed on treatment.
Read those numbers as a warning about timing, not a verdict on you. Neither trial tells you what will happen to your own weight — they tell you that regain, at scale, is common enough that planning for it isn't pessimism. It's the realistic case, and it's exactly why the plan below exists: something has to replace the appetite suppression once it's gone, and "trying harder" isn't a plan, it's a hope.
This isn't about the dose
Whether, when, and how you stop — including whether a lower maintenance dose is right for you — is a conversation with your prescriber, not something this article or any calculator can answer. What follows is the other half: the eating and training decisions that are yours either way.
The four things that actually hold the result
None of this is exotic. It's the same handful of levers that hold any weight loss — they just matter more here, because appetite isn't doing the work for you anymore.
Know your new number. Your maintenance calories at your current weight are lower than they were before you lost the weight — that's the arithmetic above, not damage or a "broken metabolism." Knowing that number, roughly, turns "eat normally" from a guess into a target you can actually check yourself against. Our GLP-1 off-ramp calculator works out your new maintenance, a protein range, and a 12-week plan from a few numbers — it's a starting estimate to test, not a verdict, but it's a far better starting point than eating by feel while your appetite is still unpredictable.
Keep protein high. Rapid loss on a suppressed appetite tends to take more muscle with it than slower, deliberate loss does. Two independent meta-analyses of randomized trials converge on roughly the same range: Karakasis et al. (2025), pooling 22 trials and 2,258 participants, found lean mass made up about 25% of total weight lost on GLP-1 therapies; Beavers et al. (2025), pooling 9 placebo-controlled trials, put it around 30%. Call it a quarter to a third. Muscle is a large share of what you burn at rest, so losing more of it than you needed to means arriving at your new weight with an even lower maintenance number — right as appetite comes back. Protein is the lever that pushes against that.
Keep training. Protein without a reason to keep the muscle is raw material with nowhere to go; resistance training is the signal that tells your body it's still needed. Two or three sessions a week, nothing elaborate, is enough to protect what you have — and it also keeps your maintenance a little higher than it would otherwise be, which is exactly the number you're trying to hold.
Watch the weekly trend, not the day. Body weight bounces day to day for reasons that have nothing to do with fat — sodium, water, digestion, sleep. The only signal worth reacting to is the trend across a week or two, and the only useful reaction is a small one, caught early.
What the first twelve weeks look like
Trying to figure all of this out cold, the week you stop, is the hard way to do it. A simple three-phase shape covers most of the transition:
Weeks 1–4 — a small buffer. Eat a little under your new maintenance — roughly 200 calories below it, which is deliberately modest. It's not a diet; it's room to overshoot on a rough day without the scale climbing, while your appetite is still finding its post-medication level.
Weeks 5–8 — settle at maintenance. Move up to eating at your new maintenance number and hold there. This is where the trend-watching earns its keep: if your weekly average climbs for two weeks running, that's the tripwire — trim around 150 calories and watch the next couple of weeks before deciding whether to trim again. One rising week is noise. Two in a row is a trend worth acting on.
Weeks 9–12 — maintenance is the job. By now the shape of your new normal should be visible: a number you eat most days, protein you don't have to think hard about, training that's back to routine. The goal for this stretch isn't to lose anything further — it's proving to yourself that the number holds without the medication's help, which is the entire point of the previous eight weeks.
Twelve weeks isn't a finish line so much as the point where the plan stops needing daily attention and starts running on its own. None of the three phases requires the medication's appetite suppression to work — the buffer is small on purpose, the maintenance target is arithmetic, and the tripwire only asks you to notice a trend that's already visible if you're weighing in and averaging. What it does require is that the numbers stay visible, which is the one thing the medication was quietly doing for you without your having to look.
If you go back on the medication, that's not a failure
Some people stop and stay off. Some people stop, regain, and go back on — sometimes at the same dose, sometimes lower, sometimes a different drug entirely. All of that is a legitimate outcome, decided with your prescriber, not a verdict on your effort or your character.
GLP-1s are, for a lot of people, a long-term or intermittent tool rather than a one-time course — closer to how blood pressure or cholesterol medication gets managed than to a course of antibiotics you finish once. If your trend climbs despite doing the eating and training side well, that's information for a conversation with whoever prescribes for you, not evidence that you didn't try hard enough. This article — and the plan behind it — has nothing to say about dosing or timing. That decision was never going to be a calculator's to make.
What Sophia does with this
The hard part of this transition isn't knowing what to do — it's doing it consistently while your appetite is still unpredictable and the old shortcut is gone. That's the part worth handing off. Sophia plans your off-ramp from your own numbers, keeps your protein target honest even on the days food doesn't appeal much, and reads your weekly weight trend so a normal wobble gets caught as a wobble — not ignored for two months, and not panicked over after two days. She's built for exactly this stretch: not the losing, but the part after it, where the real work of keeping it actually happens.