The short version: in STEP 1's DXA substudy [DXA — dual-energy X-ray absorptiometry, the scan that splits body mass into fat, bone and lean tissue], roughly 39% of the weight lost on semaglutide was lean mass; a SURMOUNT-1 substudy put tirzepatide's share at roughly a quarter.
Lean mass is not all muscle — it includes water and organ tissue — and lean loss accompanies all rapid weight loss, including dieting and bariatric surgery. The GLP-1-specific issue is speed and scale.
The two counterweights with real evidence are resistance training and adequate protein, commonly cited around 1.6-2.2 g/kg/day [grams of protein per kilogram of bodyweight, per day]. Older adults and already-lean users have the least margin.
- The numbers are real: roughly 39% of weight lost was lean mass in STEP 1's DXA substudy [semaglutide]; roughly a quarter in the SURMOUNT-1 substudy [tirzepatide].
- Lean mass is not a synonym for muscle. The DXA "lean" compartment includes water, organ tissue, and everything that is neither fat nor bone — the true muscle figure is smaller than the headline share.
- This is not a GLP-1 curse. Diets and bariatric surgery also take lean tissue. What the drugs change is how fast and how far the loss runs, in people who mostly are not training for it.
- Two counterweights have real evidence: resistance training preserves lean mass in a deficit, and protein targets around 1.6-2.2 g/kg/day are the commonly cited range in sports-nutrition literature.
- Margin is unevenly distributed. Older adults and already-lean users sit closest to functional thresholds — they should take this section of the label most seriously.
Two markets, one missing answer
The drug makers stay quiet and the supplement sellers shout. Neither one starts from the body-composition substudies.
Ask what GLP-1s do to muscle and the answer you get depends entirely on who profits from your reaction. The pharmaceutical marketing around semaglutide and tirzepatide barely mentions body composition. The story is pounds and A1C [the blood-sugar marker used to track diabetes control].
Meanwhile an entire secondary economy of supplements, coaching programs, and gray-market compounds has built its pitch on the word sarcopenia. That word means the age-related loss of muscle mass and strength. The pitch wields the scariest available number as if it described pure muscle evaporating.
One side needs you calm. The other needs you frightened. Neither starts with the substudies.
The substudies exist, they used DXA — dual-energy X-ray absorptiometry, the standard body-composition scan — and their numbers are specific. That is where we start.
What the trials measured
The STEP 1 substudy put the lean share of weight lost near 39%. SURMOUNT-1 put it near a quarter. Both were scans of a subset, not the full trial.
STEP 1 was semaglutide's pivotal obesity trial: 68 weeks, 2.4 mg weekly, mean weight reduction approximately 14.9%. It was published in the New England Journal of Medicine in 2021.
Within it, a DXA substudy scanned a subset of participants before and after. It found that roughly 39% of total weight lost was lean body mass.
SURMOUNT-1 was tirzepatide's counterpart: 72 weeks, with the 15 mg arm averaging approximately 20.9% weight reduction, published in 2022. Its body-composition subanalysis put the lean share of weight lost at roughly a quarter.
| trial | compound | duration | mean weight loss | share of loss that was lean mass |
|---|---|---|---|---|
| STEP 1 [NEJM 2021] | semaglutide 2.4 mg | 68 wk | ~14.9% | ~39% [DXA substudy] |
| SURMOUNT-1 [NEJM 2022] | tirzepatide 15 mg | 72 wk | ~20.9% | ~25% [substudy — roughly a quarter] |
Resist the urge to read that table as a verdict that tirzepatide "spares muscle" and semaglutide does not. These are substudies of different trials, on subsets of participants, with different populations and durations. That is the same cross-trial caution we apply everywhere else on this site.
What the two numbers establish together is the honest range. When these drugs take a large amount of weight off, somewhere between a quarter and two-fifths of it, by DXA, is not fat.
What "lean mass" actually includes
The DXA lean compartment is muscle plus water, organ tissue, skin and connective tissue. The muscle number is smaller than the lean number, always.
Here is where the scare copy cheats. DXA divides you into three compartments — fat, bone mineral, and everything else. That third compartment, "lean mass," is skeletal muscle plus water, organ tissue, skin, and connective tissue.
When total body mass drops fast, some of the lean-compartment decline is water and the general shrinkage of a smaller body. Organs and support tissue scale down with the frame they serve.
Skeletal muscle is only part of the lean number. That means "39% of the loss was lean mass" and "39% of the loss was muscle" are different sentences. Only the first one is supported.
That correction cuts the alarm down — but not to zero. Muscle is in there and the decline is real. Pretending the water content of the lean compartment makes the whole number an artifact is the calm-side version of the same cheat.
Three reasons it matters
Regain returns as fat, physical function rides on muscle, and resting metabolic rate drifts down. In that order.
Regain quality. This is the one the market talks about least and the data supports most directly. Weight regain after stopping these drugs is common — and regain arrives disproportionately as fat.
Take a person who loses 25% of body weight at a 39% lean share, stops, and regains half of it. They can end the round-trip lighter but with a worse body composition than they started with. The lean tissue lost on the way down is not automatically restored on the way back up.
Function. Strength and mobility ride on muscle, and the relationship gets unforgiving with age. For a 35-year-old with substantial fat mass, a few kilograms of lean loss is recoverable noise. For a 68-year-old, the same loss can be the difference between climbing stairs comfortably and not.
Metabolic rate. Lean tissue is the metabolically expensive part of you. Lose it and resting energy expenditure drifts down — modestly per kilogram, but in the direction that makes maintenance harder. This is the most commonly cited reason and the least dramatic in magnitude. We rank it third on purpose.
The missing context
Dieting and bariatric surgery take lean tissue too. What the drugs changed is the speed, the scale, and who is now doing it.
Lean mass loss is not a GLP-1 invention. It is a property of losing weight fast. Caloric-restriction diets take lean tissue. Very-low-calorie protocols take more.
Bariatric surgery — the previous benchmark for rapid, large-scale loss — takes a substantial lean share too. The body in a deep deficit does not politely burn only fat. It never has.
What the drugs genuinely changed is distribution and scale. Surgical-magnitude weight loss used to happen to a small population, screened and followed by clinical teams.
GLP-1s deliver it to millions. Most of them received a prescription, not a training program. Many are eating well under their protein needs precisely because the drug has deleted their appetite.
The compound-specific effect and the context-specific effect point the same direction, and the context is the part you control.
The two counterweights
Resistance training and adequate protein. Nothing else sold into this gap has comparable evidence behind it.
The fix-selling economy offers you dozens of answers. The exercise-science literature offers two, and neither is for sale in a dropper bottle.
Resistance training. The finding that lifting preserves lean mass during a caloric deficit is one of the most consistently replicated results in the field. It holds across diet studies, across populations, across decades.
It does not require heroic volume. It requires progressive load, done regularly, through the loss phase. This is the single highest-leverage variable a person losing weight on an incretin controls.
Protein. Appetite suppression suppresses protein intake along with everything else, which is exactly the wrong nutrient to lose.
Intakes around 1.6-2.2 g per kg per day are the commonly cited targets in the sports-nutrition literature for preserving lean mass in a deficit. A person eating 40% fewer total calories will not hit that range by accident. Protein has to be planned first, not left to whatever appetite remains.
Neither counterweight is exotic, which is precisely why the market underweights them. There is no margin in telling you to lift and eat protein. There is enormous margin in telling you the muscle is melting and the answer costs $89 a month.
reader skillsLooking at any of these compounds regardless? Read the certificate before the claim.The five measures a real COA carries, what most certificates skip, and how to verify one with the lab — in five minutes.→Who should care most
Older adults and people who are already lean. Both sit closest to a functional threshold and rebuild the slowest.
Two groups sit closest to the edge. Older adults start nearer to functional thresholds and rebuild muscle more slowly. They are the population where the word sarcopenia stops being marketing and starts being medicine. For them, the resistance-training and protein counterweights are not optimization. They are the core of doing this safely. Already-lean users are the second group — people reaching for these drugs to drop a modest amount. Weight lost from an already-lean body draws proportionally more from lean tissue.
The person with the least fat to give has the most muscle at stake. That is exactly the user profile the gray market has been growing fastest.
Building the sequel
Muscle-preservation drugs are in trials now, designed to be paired with these compounds. None is approved for that use.
Nothing confirms the lean-mass question like the drug industry's own pipeline. Muscle-preservation compounds designed to pair with incretin therapy are in active clinical development across several companies. The explicit bet is that the next generation of weight-loss treatment is a combination: one drug for the fat, one to hold the muscle.
None of these agents is approved for that use today. We are not going to launder pipeline press releases into trial results by naming candidates here.
The gray market, predictably, is already selling its own versions of that story, well ahead of the evidence. When those programs publish, we will cover the data.
Until then, the two counterweights above are what exists. How this class's broader side-effect profile fits around them is covered in our side-effects fundamentals.
One more note on where this collides with the research-compound economy: the muscle-loss numbers above come from pharmaceutical-grade drugs in supervised trials.
The gray-market versions of these molecules add quality variance on top of everything this article describes. They are covered across our compound index and in our COA guide. Questions about how research dosing is even discussed belong in our dosing fundamentals, not here.
FAQ
How much muscle do you actually lose on a GLP-1?
By DXA, roughly 39% of the weight lost in STEP 1's substudy [semaglutide 2.4 mg, 68 weeks] was lean mass. The SURMOUNT-1 substudy [tirzepatide 15 mg, 72 weeks] put it at roughly a quarter. Read those as lean mass, not as muscle.
The DXA lean compartment is skeletal muscle plus water, organ tissue, skin and connective tissue. So 39% of the loss was lean mass and 39% of the loss was muscle are different sentences. Only the first one is supported.
The caveat on the comparison matters too. These are substudies of different trials, scanning subsets of participants, with different populations and durations. So the gap between the two numbers is not evidence that tirzepatide spares muscle.
What the pair establishes together is the honest range. When these drugs take a large amount of weight off, somewhere between a quarter and two-fifths of it, by DXA, is not fat.
Is this unique to GLP-1 drugs?
No. Lean mass loss is a property of losing weight fast, not a GLP-1 invention. Caloric-restriction diets take lean tissue. Very-low-calorie protocols take more.
Bariatric surgery — the previous benchmark for rapid, large-scale loss — takes a substantial lean share too. The body in a deep deficit does not politely burn only fat. It never has. What the drugs genuinely changed is distribution and scale.
Surgical-magnitude weight loss used to happen to a small population, screened and followed by clinical teams. GLP-1s deliver it to millions. Most of them received a prescription rather than a training program. Many are eating well under their protein needs precisely because the drug has deleted their appetite.
So the honest framing is not that these drugs attack muscle. It is that they hand a diet-magnitude deficit to people who are not set up for it, and the setup is the part you control.
What actually prevents it?
Two things, and neither is for sale in a dropper bottle. Resistance training first. The finding that lifting preserves lean mass during a caloric deficit is one of the most consistently replicated results in the field. It holds across diet studies, across populations, across decades.
It does not require heroic volume. It requires progressive load, done regularly, through the loss phase. It is the single highest-leverage variable a person losing weight on an incretin controls.
Protein second. Intakes around 1.6-2.2 g per kilogram of bodyweight per day are the commonly cited targets in the sports-nutrition literature. That is the range for preserving lean mass in a deficit. The caveat is that both have to be deliberate.
Appetite suppression suppresses protein intake along with everything else, and a person eating 40% fewer total calories will not hit that range by accident. Protein has to be planned first, not left to whatever appetite remains.
Should older adults avoid these drugs because of muscle loss?
That is a medical decision, not an editorial one, and the conversation belongs with a physician. What the data supports is narrower. Older adults start nearer to functional thresholds and rebuild muscle more slowly. They are the population where the word sarcopenia stops being marketing and starts being medicine.
The function argument is concrete. For a 35-year-old with substantial fat mass, a few kilograms of lean loss is recoverable noise. For a 68-year-old, the same loss can be the difference between climbing stairs comfortably and not.
So for older adults the resistance-training and protein counterweights are not optimization. They are the core of doing this safely.
The caveat is that margin, not muscle loss on its own, is what the evidence describes. Already-lean users sit in the same exposed position, because weight lost from an already-lean body draws proportionally more from lean tissue.
Do the muscle-preservation drugs in development work?
Unknown, and that is the whole answer today. Muscle-preservation compounds designed to pair with incretin therapy are in active clinical development across several companies. The explicit bet is that the next generation of weight-loss treatment is a combination: one drug for the fat, one to hold the muscle.
None of these agents is approved for that use. We are not going to launder pipeline press releases into trial results by naming candidates here — when those programs publish, we will cover the data.
The predictable part is already happening. The gray market is selling its own versions of that story well ahead of the evidence. Gray-market versions of these molecules add quality variance on top of a mechanism nobody has demonstrated in a published trial.
Until the trials read out, resistance training and adequate protein are the two counterweights that exist.
References
- Wilding JPH, et al. Once-Weekly Semaglutide in Adults with Overweight or Obesity [STEP 1]. N Engl J Med. 2021;384:989-1002 — PMID 33567185. The lean-mass share quoted here comes from the DXA body-composition substudy, which scanned a subset of participants rather than the full trial population.
- Jastreboff AM, et al. Tirzepatide Once Weekly for the Treatment of Obesity [SURMOUNT-1]. N Engl J Med. 2022;387:205-216 — PMID 35658024. Body-composition figures come from the DXA substudy reported separately: Diabetes Obes Metab. 2025 — PMID 39996356, n=160.
- Jastreboff AM, et al. Triple-Hormone-Receptor Agonist Retatrutide for Obesity — A Phase 2 Trial. N Engl J Med. 2023;389:514-526 — PMID 37366315.
- Morton RW, et al. A systematic review, meta-analysis and meta-regression of the effect of protein supplementation on resistance training-induced gains in muscle mass and strength. Br J Sports Med. 2018 — PMID 28698222. This is the source of the roughly 1.6 g/kg/day plateau.
- Jäger R, et al. International Society of Sports Nutrition Position Stand: protein and exercise. J Int Soc Sports Nutr. 2017 — PMID 28642676. The position stand is where the upper end of the 1.6 to 2.2 g/kg/day range comes from, and it is expressed per kilogram of bodyweight.
- Inside Your Peptides, Retatrutide: What the Data Actually Shows — the evidence review
- Inside Your Peptides, Side Effects Fundamentals — the incretin class in practice
disclosure: inside your peptides is published by the owners of the next lab, the only vendor we are paid by elsewhere on this site — this article carries none. grades and rankings follow the published criteria in our editorial policy — never referral terms. research + education only · not medical advice.