GLP-1 Muscle Loss: What the DEXA Data Shows

The question arrives in the same shape every time. Retatrutide produced the largest weight-loss numbers anyone has reported in this class, so does it take the largest bite out of muscle? The answer sits in a small pile of body-composition substudies that almost nobody reads past the abstract. Where a share of total weight loss was actually reported, it did not rise with the size of the weight loss. Those substudies are also small, they report on different denominators, and not one of them measured whether anybody got weaker.

What a DEXA scan actually measures

Dual-energy X-ray absorptiometry splits the body into fat mass, lean soft tissue and bone. Lean mass is not muscle. It takes in water and organ tissue as well, and it moves when hydration changes, which is what happens on medicines that alter eating and gastric emptying. Papers in this area swap between fat-free mass, lean body mass and lean soft tissue almost interchangeably, and the three are not identical quantities. So when a headline says a drug costs you a quarter of your weight loss in muscle, what was measured was lean mass on a scanner, and those are not the same claim.

The tirzepatide number, because it is the cleanest

The SURMOUNT-1 body-composition substudy was published in Diabetes, Obesity and Metabolism in 2025, and it is the dataset most people end up quoting. 160 of the 2539 participants in SURMOUNT-1 had scans at baseline and at week 72: 124 on pooled tirzepatide doses and 36 on placebo, 73 percent female, mean body weight 102.5 kg, mean body mass index 38.0. From baseline to week 72, body weight changed by -21.3 percent, fat mass by -33.9 percent and lean mass by -10.9 percent on tirzepatide, against -5.3 percent, -8.2 percent and -2.6 percent on placebo. Of the body weight lost, roughly 75 percent was fat mass and roughly 25 percent was lean mass, and that split was the same for tirzepatide and for placebo. It held across most of the subgroups the authors examined, including sex, age band and how much weight each person lost.^[1]

The retatrutide substudy, and which population it was in

The retatrutide body-composition data comes from a substudy of the phase 2 trial in people with type 2 diabetes, not the obesity trial, and it was published in The Lancet Diabetes and Endocrinology in 2025. Of 281 people in the main trial, 189 entered the substudy, 155 had a baseline scan, and 103 completed the trial with scans at both baseline and week 36. Percent reduction from baseline in total fat mass ran 4.9 percent in the lowest retatrutide group, 15.2 percent in the pooled middle group, 26.1 percent in the pooled higher group and 23.2 percent in the top group, against 2.6 percent for dulaglutide and 4.5 percent for placebo. On lean mass the authors state their conclusion in plain words: the proportion of lean mass loss to weight loss was similar to other obesity treatments, and they present that as reassurance that a greater proportion of lean mass is not lost with retatrutide despite the larger overall weight loss. That sentence is what the retatrutide muscle question turns on, and it comes from an Eli Lilly author team reporting 103 completed scan pairs.^{[2, 8]}

The semaglutide number sits on a different denominator

This is where most comparisons fall over. The STEP 1 body-composition substudy covered 140 participants, 95 on semaglutide and 45 on placebo, mean body weight 98.4 kg, mean body mass index 34.8, 76 percent female, scanned at baseline and week 68. Body weight changed by -15.0 percent on semaglutide against -3.6 percent on placebo. Total fat mass fell 19.3 percent and visceral fat mass 27.4 percent. Total lean body mass fell 9.7 percent, while lean body mass as a proportion of total body mass rose by 3.0 percentage points, and the lean to fat ratio improved by 0.23 from a baseline of 1.34. Now read those numbers again. They are percent changes within each compartment, not shares of the weight that came off. Setting 9.7 percent beside the 25 percent lean share from SURMOUNT-1 compares two different denominators, and that is how the internet arrived at the belief that semaglutide is far kinder to muscle than tirzepatide. The substudy says nothing of the sort. It was also published as a conference abstract in a journal supplement rather than as a full paper.^[3]

What happens when you pool everything

A systematic review in Annals of Internal Medicine in 2026 went through 35 randomised trials of liraglutide, semaglutide, tirzepatide or dulaglutide that reported body composition, with a median duration of 26 weeks and a median of 78 participants each. Across agents and measurement methods, the median proportion of total weight loss attributable to reductions in muscle-based indices was 28.3 percent, with an interquartile range from 15.9 to 39.9 percent, and 65 percent of the interventions exceeded the prespecified benchmark of about 25 percent. Among studies using bioelectrical impedance or DEXA the median was about 29 percent; among those using CT or MRI it was about 25.3 percent against a 15 percent benchmark. The comparator arms are the part people skip past. Of the 13 studies reporting weight loss in lifestyle or placebo groups, 38 percent also exceeded their respective benchmark. Losing lean mass alongside fat is what weight loss does to a body. The medicines did not invent it.^[4]

So does more potency mean more muscle loss?

On the evidence that reports a share on a comparable denominator, no. SURMOUNT-1 put the split at roughly three quarters fat and one quarter lean, and got the same split in its own placebo arm. The retatrutide authors report their proportion as similar to other obesity treatments despite the larger weight loss. The pooled median across 35 trials sits at 28.3 percent, in the same neighbourhood as both. What the pooled review adds is the variation, and it is unflattering: an interquartile range from 15.9 to 39.9 percent is enormous, the underlying studies were small and short, and the measurement methods varied so much that the reviewers could not run a meta-analysis. The honest answer to whether retatrutide is worse for muscle than semaglutide or tirzepatide is that no trial has been designed to answer it, and the ratios that have been measured do not track potency.^{[1, 2, 4]}

The one place anyone measured function

Across all 35 trials in that systematic review, no study reported an objective physical function outcome. Not one measured whether people could do anything they could not do before. The nearest published thing is SEMALEAN, a prospective single-centre study in France of 115 patients starting semaglutide, of whom 106 finished, with a mean body mass index of 46.3 and scans at baseline, seven months and twelve months. Mean weight reduction was 10 percent at seven months and 13 percent at twelve. Total fat mass fell 14 percent and then 18 percent. Lean mass declined by about 3 kg by month seven and then held steady. Handgrip strength rose by 4.5 kg at twelve months, and the prevalence of sarcopenic obesity in the group fell from 49 percent at baseline to 33 percent at twelve months. That is one uncontrolled study at one hospital, and it points the opposite way to the panic. It is not a randomised trial.^[5]

What the evidence does not show

Four gaps. The retatrutide substudy was in people with type 2 diabetes over 36 weeks with 103 participants completing both scans, so it is not a readout on the obesity population the headline weight-loss figures came from. No randomised trial in this class has reported strength, gait speed or any other function endpoint. Nothing published follows lean mass after people stop, so whether the weight that returns comes back as fat or as lean is simply unknown. And the amylin analogues, the compounds most likely to sit alongside these medicines next, have no published body-composition substudy at all; our amylin versus GLP-1 guide covers where that class currently stands.^{[2, 4]}

Why this matters more outside the trials than inside them

Every number above came out of a trial where participants had a lifestyle programme running alongside the medicine, scheduled visits, and a research team watching them. None of that comes with material bought from a website. The resistance-training argument that surfaces in every one of these discussions is reasonable, and it is still a hypothesis: a 2024 review in Diabetes Care proposed tailored resistance exercise as an adjunct to incretin therapy specifically to protect lean mass, and a proposal is not a result. If lean mass is the thing worrying you, the person to raise it with is a clinician who can measure yours.^[6]

Regulatory status

Tirzepatide is the active molecule in the branded prescription medicines Mounjaro and Zepbound, and semaglutide in Ozempic and Wegovy. Retatrutide is neither: it is investigational, and it is not approved by the FDA for any use. Research-grade material sold online under any of these names is not the prescription product and was never tested to the same standard; our page on research peptides versus prescription drugs sets out the difference.

References & sources

1. Look M, Dunn JP, Kushner RF, et al. Body composition changes during weight reduction with tirzepatide in the SURMOUNT-1 study of adults with obesity or overweight. Diabetes Obes Metab. 2025 May;27(5):2720-2729. PMID 39996356. PMCID PMC11965027. 160 of 2539 SURMOUNT-1 participants scanned at baseline and week 72; roughly 75 percent fat and 25 percent lean of weight lost, for tirzepatide and placebo alike. An erratum was published in the same journal in November 2025.
2. Coskun T, Wu Q, Schloot NC, et al. Effects of retatrutide on body composition in people with type 2 diabetes: a substudy of a phase 2, double-blind, parallel-group, placebo-controlled, randomised trial. Lancet Diabetes Endocrinol. 2025 Aug;13(8):674-684. PMID 40609566. ClinicalTrials.gov NCT04867785. Eli Lilly. 189 entered the substudy, 155 had a baseline scan, 103 completed both scans at week 36.
3. Wilding JPH, Batterham RL, Calanna S, et al. Impact of Semaglutide on Body Composition in Adults With Overweight or Obesity: Exploratory Analysis of the STEP 1 Study. J Endocr Soc. 2021;5(Supplement_1):A16-A17. doi:10.1210/jendso/bvab048.030. A conference abstract published in a journal supplement, not a full paper. 140 participants scanned at baseline and week 68.
4. Batsis JA, Gavras A, Gross DC, et al. Effect of Incretin-Based and Nonpharmacologic Weight Loss on Body Composition: A Systematic Review. Ann Intern Med. 2026 Jul;179(7):996-1013. doi:10.7326/ANNALS-25-00478. PMID 41996180. 35 primary studies; median 28.3 percent of weight loss attributable to muscle-based indices, interquartile range 15.9 to 39.9 percent; no study reported an objective physical function outcome. An erratum was published in the same issue.
5. Alissou M, Demangeat T, Folope V, et al. Impact of Semaglutide on fat mass, lean mass and muscle function in patients with obesity: The SEMALEAN study. Diabetes Obes Metab. 2026 Jan;28(1):112-121. PMID 41068996. PMCID PMC12673431. Prospective single-centre study, 115 enrolled and 106 completing, scans at baseline, seven months and twelve months.
6. Locatelli JC, Costa JG, Haynes A, et al. Incretin-Based Weight Loss Pharmacotherapy: Can Resistance Exercise Optimize Changes in Body Composition? Diabetes Care. 2024 Oct 1;47(10):1718-1730. PMID 38687506. Narrative review proposing tailored resistance exercise training as an adjunct to incretin therapy.
7. Wilding JPH, Batterham RL, Calanna S, et al. Once-Weekly Semaglutide in Adults with Overweight or Obesity. N Engl J Med. 2021 Mar 18;384(11):989-1002. PMID 33567185. STEP 1, ClinicalTrials.gov NCT03548935. 1961 adults, 68 weeks; the parent trial for the body-composition substudy above.
8. Jastreboff AM, Kaplan LM, Frias JP, et al. Triple-Hormone-Receptor Agonist Retatrutide for Obesity: A Phase 2 Trial. N Engl J Med. 2023 Aug 10;389(6):514-526. PMID 37366315. ClinicalTrials.gov NCT04881760. 338 adults, 48 weeks; the obesity trial that produced the headline retatrutide weight figures, and a separate trial from the body-composition substudy above.