How much muscle do you lose on a GLP-1 — and how much of that is avoidable?
The scariest GLP-1 headline of 2025–2026 is also the most fixable one. The data on lean-mass loss is real; so is the data on preventing most of it.
In body-composition substudies of the major trials, roughly 25 to 40 percent of total weight lost on GLP-1 therapy was lean mass rather than fat — a ratio similar to diet-induced weight loss generally, not unique to the drugs. The two countermeasures with real evidence are protein intake around 1.2–1.6 g per kilogram of body weight per day and progressive resistance training two to three times weekly, which in weight-loss studies preserves most at-risk muscle.
- What the trial substudies actually measured: In the STEP-1 DEXA substudy of semaglutide 2.4 mg, lean mass accounted for roughly 39 percent of total mass lost, though as a proportion of remaini....
- Why lean mass is the number to protect: Muscle is the body's glucose sink and metabolic reserve.
- The protein number, made concrete: Weight-loss research converges on 1.2 to 1.6 grams of protein per kilogram of body weight daily during active loss — for a 100 kg person, 120 to 16....
- Resistance training: Trials combining resistance exercise with significant caloric deficit consistently preserve the majority of at-risk lean mass, and the effective do....
- How to know if it is happening to you: Scale weight cannot distinguish fat from muscle.
What the trial substudies actually measured
In the STEP-1 DEXA substudy of semaglutide 2.4 mg, lean mass accounted for roughly 39 percent of total mass lost, though as a proportion of remaining body weight, lean mass actually rose. SURMOUNT-1's body-composition analysis of tirzepatide put fat loss at about three times lean loss — roughly a quarter of the total coming from lean tissue. Those two framings — “you lost muscle” and “you became proportionally more muscular” — are both true, and most alarming headlines quote only the first.
Context matters: bariatric surgery and calorie restriction show similar or worse ratios. The GLP-1-specific concern is scale and speed. Losing 20 percent of body weight in a year means the absolute lean-mass number is large, and rapid loss gives less time for training and protein to compensate.
| Trial / drug | Total weight loss | Lean share of loss | Note |
|---|---|---|---|
| STEP-1 DEXA (semaglutide 2.4) | ~15% | ~39% | Lean % of body weight rose |
| SURMOUNT-1 (tirzepatide) | ~21% top dose | ~25% | Fat loss ~3× lean loss |
| Diet-only comparators | varies | ~25–35% | Similar ratio to drugs |
| With resistance training | varies | Substantially reduced | Across weight-loss RCTs |
Why lean mass is the number to protect
Muscle is the body's glucose sink and metabolic reserve. Lower lean mass predicts weaker glycemic control, lower resting energy expenditure — which worsens regain when treatment stops — and, in older adults, frailty and fall risk. For patients over 60, where sarcopenia is already the background condition, an unmanaged 8–10 kg lean loss is a genuine clinical harm, not a cosmetic one.
It also shapes the rebound problem. Our article on stopping GLP-1s covers the regain data; the mechanism connects here, because weight regained after unstructured loss returns disproportionately as fat. Preserving muscle during loss is the single best hedge against ending up worse-composed than you started.
The protein number, made concrete
Weight-loss research converges on 1.2 to 1.6 grams of protein per kilogram of body weight daily during active loss — for a 100 kg person, 120 to 160 grams, which is two to three times a typical American intake pattern concentrated at dinner. On a GLP-1, appetite suppression makes this genuinely hard: patients routinely report satiety after a few hundred calories.
The practical adaptations that work in clinic: protein-first eating at every meal, distributing intake across three to four feedings of 30–40 grams, and using dairy, eggs, fish and — where needed — protein supplements to hit targets inside a small appetite budget. During dose-escalation weeks when nausea peaks, liquid protein tolerates best. None of this requires a specific product; it requires arithmetic.
Resistance training: the dose that shows up in studies
Trials combining resistance exercise with significant caloric deficit consistently preserve the majority of at-risk lean mass, and the effective dose is modest: two to three sessions weekly, covering major muscle groups, progressing load over time. Walking is excellent for health and adherence, but it does not send the muscle-retention signal; progressive overload does.
For patients new to training, the realistic on-ramp is bodyweight and machine-based work with a focus on legs, hips and back — the muscle groups whose loss drives functional decline. The window that matters most is the first six months of therapy, when the rate of weight loss, and therefore lean-mass risk, is highest.
How to know if it is happening to you
Scale weight cannot distinguish fat from muscle. Reasonable monitoring, in ascending order of precision: grip strength and simple performance markers (chair stands, stair climbing), circumference measures, bioimpedance smart scales (noisy but trend-usable), and DEXA scans, which many imaging centers sell for $50–$150 cash and which give a true lean/fat breakdown worth doing at baseline and every six months during active loss.
Red flags worth raising with your prescriber: strength loss out of proportion to weight loss, especially in older adults; and weight loss continuing past goal without a maintenance plan. Dose reduction to a maintenance level is a legitimate clinical tool once targets are reached.
The supplement industry has noticed
“Muscle-preservation stacks” marketed at GLP-1 users — HMB, creatine, essential amino acid blends, and peptide products — are now a category of their own. The honest ranking: adequate total protein and resistance training carry nearly all of the evidence. Creatine monohydrate has decent general support for training performance and is cheap. HMB's evidence in this context is thin. Unregulated “recomposition peptides” sold online belong in the same risk bucket as all gray-market peptides we cover: unverified, untested, and unnecessary.
If a telehealth program bundles a proprietary supplement stack into its GLP-1 plan price, that is a pricing decision, not a clinical one — compare the plan against our verified pricing with the stack stripped out.
Frequently asked questions
Do GLP-1s cause more muscle loss than dieting?
The lean-to-fat ratio of loss is broadly similar to caloric restriction. The difference is magnitude: the drugs produce much larger total losses, so the absolute lean-mass number is bigger without countermeasures.
How much protein do I need on a GLP-1?
Research supports 1.2–1.6 g per kg of body weight daily during active loss — for most adults, 100–160 g spread across the day. Appetite suppression makes deliberate protein-first planning necessary.
Is a DEXA scan worth it?
At $50–$150 cash in most metros, a baseline plus six-month DEXA is the cheapest way to see your actual fat/lean split rather than guessing from the scale.