Slower than a crash, and faster than most people expect. The clearest timeline in the literature comes from the STEP 1 trial extension, published by John Wilding and colleagues in Diabetes, Obesity and Metabolism in 2022. Participants had lost an average of 17.3% of their body weight over 68 weeks. One year after treatment and the accompanying lifestyle support ended, they had regained about two-thirds of that loss. Not all of it, and not overnight, but a substantial majority of it within twelve months.
The short answer
Roughly two-thirds of lost weight returns within a year once the thing that was driving the loss stops. Spread across twelve months, that works out to a slow monthly drift rather than a dramatic reversal, which is precisely the problem: a gradual change is much harder to see from the inside than a sudden one.
What the STEP 1 extension measured
The original STEP 1 trial randomised 1,961 adults to 68 weeks of treatment or placebo, alongside a lifestyle intervention. Average weight loss in the treatment group was 17.3%, against 2.0% in the placebo group. The extension study then followed a subset of participants for a further year after everything stopped, both the medication and the structured lifestyle support.
What it found was a consistent, gradual return. Cardiometabolic markers moved back in step with the weight, which is worth noting because it means the regain was a genuine reversal rather than a scale artefact. The mechanism behind that reversal, and why appetite in particular comes back the way it does, is covered in why people regain weight after stopping a GLP-1.
Two-thirds of the loss, back within a year. Spread over twelve months, that is a slow drift, not a crash.
The shape of the curve, not just the endpoint
The headline number is the endpoint, and the endpoint is the least useful part of it. What matters day to day is the shape of the line getting there, which was a steady upward slope rather than a step change. Nothing dramatic happened in month two that would have announced itself.
Work out what two-thirds of a year's loss looks like divided across twelve months and you get a monthly change smaller than the amount your weight already swings between a Tuesday and a Saturday. That's the entire difficulty. The signal is real and it is smaller than the noise around it, which means single weigh-ins cannot detect it and only the direction of a trend can. How much weight fluctuates within a single day is larger than most people assume, and it is what hides the early part of this curve.
Why the first few weeks look worse than they are
There's a countervailing effect right at the start that trips people up in the other direction. In the first week or two after a period of restriction ends, the scale often jumps by several pounds quickly, and that early movement is mostly glycogen and the water bound to it rather than fat. Eating more carbohydrate refills muscle glycogen stores, and each gram of stored glycogen holds roughly three grams of water with it.
So the first fortnight tends to overstate what's happening, and months three through twelve tend to understate it. That combination is unhelpful: people panic at the part that isn't real and miss the part that is. The practical rule is to ignore the first two weeks after any change in eating pattern entirely, and to start reading the trend after that. A single reading in that window tells you almost nothing, which is also why rebound weight gain is worth defining properly before deciding you're in it.
What actually changes the timeline
The trial's design contains its own answer. The regain followed the withdrawal of both the medication and the structured lifestyle support, and only one of those two is something you can keep. Whatever structure was doing the work during the losing phase does not automatically survive into the phase after it, and if nothing replaces it, the default is the curve above.
In practice that means the things worth keeping are the ones that can survive without a programme around them: a protein floor you can hit without planning your week around it, movement that doesn't depend on motivation, and regular enough weigh-ins to see a direction. None of that is exciting, and the boringness is the point. Anything that requires sustained effort to maintain is precisely what stops when the structured phase ends.
Timing matters too. A drift caught at three weeks is a small correction. The same drift caught at eight months is most of the way back, and it takes far more effort to reverse than it would have taken to interrupt. Early detection is worth more here than almost any other intervention, purely because of the arithmetic.
How OffRamp helps
The Regain Radar exists specifically for this curve. It plots weigh-ins as a trend line rather than as individual numbers, which is the only way a change smaller than daily fluctuation becomes visible at all. In practice that means a genuine upward direction separates itself from noise in a couple of weeks rather than a couple of months, which is the difference between a small adjustment and a long climb back. The daily protein floor is the other half: a minimum to hit rather than a limit to stay under, so it's a structure that survives when the losing phase ends. Anything to do with your treatment itself, including any change to it, is a question for your prescriber rather than something an app should weigh in on.
