Nobody has measured this. Not one published trial or meta-analysis has reported the fat-versus-muscle split of the weight that returns after stopping a GLP-1 medication. The idea that it "comes back as fat" is repeated widely and confidently, and it rests on a different body of research, in a very different physiological state, in twelve men. This article covers what is genuinely known, where the assumption came from, why the distinction matters, and what can be done given that the answer is not in.
Body composition: the proportions of fat mass and lean mass that make up body weight. Two people at the same weight can be composed very differently.
What is actually known
Two things are well established, and they are about the weight coming off, not going back on.
Weight loss includes lean mass. In the SURMOUNT-1 body-composition sub-study, approximately 75% of the weight lost was fat mass and 25% was lean mass — in both the tirzepatide arm and the placebo arm (Look 2025). In ordinary dieting, the median share lost as fat-free mass ran from about 14% on standard low-calorie diets to about 23% on very-low-calorie diets (Chaston 2007).
Weight returns when treatment stops. Pooled trial data put about 60% of lost weight back within a year of stopping, with the curve decelerating and modelled to plateau below the original starting weight (Budini 2026).
Put those together and an obvious question follows: if a quarter of what came off was lean tissue, what is the returning weight made of? That question has not been answered.
Why the answer is missing
Because the trials did not measure it.
The largest meta-analysis of GLP-1 withdrawal trials pooled eight randomised studies and 2,372 participants. The outcomes it pooled were body weight, waist circumference and BMI. It contains no fat-mass or lean-mass data at all (Berg 2025).
That is not an oversight by the reviewers — it reflects what the underlying trials collected. Body-composition scanning is expensive and was typically done in small sub-studies during treatment, not in the follow-up period after treatment ended. So the field knows a great deal about the composition of weight lost and almost nothing about the composition of weight regained.
Where the "comes back as fat" idea comes from
It comes from a real and respectable literature — just not this one.
In a reanalysis of the Minnesota Starvation Experiment, twelve men were refed after 24 weeks of severe semi-starvation. Fat and fat-free mass recovered on separate, independently regulated trajectories, with hyperphagia inversely correlated with both fat recovery (r = −0.6) and fat-free-mass recovery (r = −0.5) (Dulloo 1997). The phenomenon has since been described across several contexts: refeeding after experimental semi-starvation, nutritional rehabilitation of malnourished adults and children, catch-up growth, and recovery from anorexia nervosa (Dulloo 2015).
Read the list of contexts again. Every one of them is a severe catabolic state. The Minnesota participants had been reduced to roughly three-quarters of their normal body weight over six months of near-starvation. Medically supervised weight loss on a GLP-1 medication, with adequate protein and food available, is not that.
The mechanism may well carry across. It is biologically coherent, and it would explain something people report. But "biologically coherent" is not the same as "measured", and the extrapolation from severe semi-starvation to medically supervised weight loss has not been tested.
Why the distinction matters
Because the two possible answers point to different things, and only one of them has been checked.
If regain were mostly fat, the consequence would fall on body composition rather than on body weight. If it were proportional, it would not. Nobody has looked, so the question stays open — and it is a question worth keeping open rather than closing with an assumption. Work an example through and the shape of the concern becomes clearer — as does what it is not. Take someone starting at 100 kg with 35% body fat: 35 kg of fat, 65 kg of lean mass. They lose 20 kg, of which 5 kg is lean, ending at 80 kg with 20 kg of fat and 60 kg of lean. They then regain 12 kg — the pooled one-year figure — and suppose all of it returns as fat. They finish at 92 kg with 32 kg of fat and 60 kg of lean.
That is 8 kg lighter than they started, and their body-fat percentage is 34.8% against 35.0% at baseline — essentially unchanged. So the worst-case version of this does not leave someone fatter than they began, and anyone claiming it does is overstating it.
What it does leave them with is 5 kg less muscle than they started with, at a lower body weight. That is the real concern, and it is a concern about the weight that was lost, not the weight that came back. It is also a modelled illustration rather than a measured outcome.
One thing worth being clear about, because it is easy to misread. The lean-mass share of weight loss was the same in the placebo arm as in the drug arm of the trial sub-study (Look 2025). This is a property of losing weight, not a property of the medication.
What to do while the answer is missing
Work on the part that is well evidenced, which is the composition of the weight coming off.
Resistance training combined with caloric restriction kept lean mass statistically unchanged across 114 trials and 4,184 people (Lopez 2022). That is a large and consistent body of evidence, and it addresses the problem at the point where it is known to occur. Muscle that was never lost does not need to be regained.
The practical version is covered in the minimum effective training routine and hitting protein on a suppressed appetite. Neither is complicated. Both take months rather than weeks to matter.
At GetLean, our philosophy is that the medication is the catalyst and what you keep is the result. The case for protecting muscle rests on what happens during weight loss, which has been measured properly — not on what happens afterwards, which has not.
Individual results vary, and clinical-trial figures describe the populations studied. Any decision about starting, continuing or stopping treatment is one to make with your doctor.
Common questions
Does weight regained after GLP-1 come back as fat?
Nobody has measured it. The largest meta-analysis of GLP-1 withdrawal trials pooled body weight, waist circumference and BMI only, with no fat-mass or lean-mass data (Berg 2025). The idea is biologically plausible and has not been demonstrated in this context.
Where does the idea that regain is mostly fat come from?
From research on refeeding after severe semi-starvation, where fat and lean tissue recovered on different trajectories (Dulloo 1997), and from related work on malnutrition rehabilitation and recovery from anorexia (Dulloo 2015). That research involved 12 men, in far more extreme physiological states than medically supervised weight loss.
Why does the composition of regain matter?
Because weight lost includes lean mass — about 25% of it in one trial sub-study (Look 2025). If the returning weight were mostly fat, someone could end up lighter than they started while carrying less muscle than they started with. That is the concern rather than a measured finding.
What can you actually do about it?
Protect muscle while the weight is coming off, which is the well-evidenced part. Resistance training combined with caloric restriction kept lean mass statistically unchanged across 114 trials and 4,184 people (Lopez 2022).
Does GLP-1 medication make this worse than ordinary dieting?
The sub-study evidence says no. The lean-mass share of weight lost was about 25% in both the tirzepatide arm and the placebo arm (Look 2025), so this is a feature of losing weight itself rather than of the medication.