Losing fat, not muscle, on a GLP-1: the part the scale doesn't show

TL;DR

• Trial-derived estimates put lean tissue at roughly 25–40% of the weight lost on caloric restriction and incretin therapies.
• A 2026 real-world cohort (n=486) reads less alarmingly - 80–85% of weight lost was fat mass, and relative skeletal muscle mass was preserved or increased in more than 70% of patients.
• Evidence class does the work here - that cohort is retrospective, single-centre and not randomised, so the clinic's nutrition guidance can't be separated from the drug.
• The headline trials are reasonable on the ratio - SURMOUNT-1 showed tirzepatide producing roughly three times more fat loss than lean loss.
• Muscle is the metabolic floor - losing it drops resting metabolic rate and makes regain, preferentially as fat, easier.
• Protein, resistance training, a sane pace and body-composition monitoring are what decide which side of that 70% you land on.

Updated September 2026: adds a 2026 real-world cohort (n=486) reporting that 80-85% of weight lost was fat mass and that relative skeletal muscle mass was preserved or increased in more than 70% of patients, a less alarming picture than the trial-derived 25-40% lean-tissue figure. Both are reported, with their evidence classes stated. ⚠️ New material added after the June 2026 MAB clearance; re-review required before publication.

MAB-reviewed and cleared for the Malaysian market (Jun 2026). Educational content, not medical advice. Compound and trial data lead; investigational agents are labelled as such. No doses or protocols appear here - exercise, nutrition and any medication are physician-directed.

---

Every rapid weight loss costs some muscle - dieting, surgery and GLP-1 drugs alike - and the trials put a number on it: roughly a quarter to a third (about 25–40%) of the weight lost on caloric restriction and incretin therapies is lean tissue, not fat. That matters because muscle is the engine that burns calories and the tissue that makes regain harder. The good news, up front: the levers that protect muscle are simple and largely in your control - enough protein, resistance training and a sane pace - and the drug's job (fat loss) and your job (keeping muscle) are separable.

Here's the detail.

The 2026 correction: the number is probably less alarming than you've been told

The 25–40% figure above comes largely from controlled trials using DXA or similar imaging, and it has driven two years of "GLP-1s eat your muscle" commentary. A 2026 retrospective real-world cohort presented at the European Congress on Obesity puts a different shape on it, and honesty requires reporting it.

The cohort followed 486 adults with obesity (mean BMI 37.7, mean age 49.9, 82% female) treated at a specialist obesity clinic in Vienna between 2022 and 2025 with a GLP-1 or GIP/GLP-1 receptor agonist - 82% semaglutide, 8% tirzepatide, 8% liraglutide. Its findings:

• 80–85% of the weight lost was fat mass.
• Mean fat mass reduction was around 18%; mean skeletal muscle mass loss around 5%.
• Relative skeletal muscle mass was preserved or increased in more than 70% of patients.
• Fat mass and absolute skeletal muscle mass moved together, which the authors read as proportional body-composition change rather than disproportionate muscle wasting.

State the evidence class plainly, because it does the work here. This is a real-world retrospective cohort from a single specialist clinic, not a randomised trial. Patients attending a specialist obesity service receive nutrition and activity guidance that the general population does not, which plausibly flatters the result; the design cannot separate the drug's effect from the clinic's. It is genuinely useful evidence and it is not the same tier as an RCT.

So how do the two numbers coexist? They are measuring in different settings and, partly, different things - absolute lean tissue lost during weight loss versus relative muscle mass at the end of it. Both can be true: some lean tissue does come off, and in a supervised setting the body-composition result still comes out favourable. What the cohort undercuts is the strong version of the claim - that muscle loss is the default, dominant outcome of incretin therapy. On current evidence it is not. What it does not undercut is the reason the rest of this article exists: the levers below are what put you in the 70% rather than outside it.

Why losing muscle is a bigger deal than it sounds

Muscle is metabolically active tissue - it burns calories even at rest. Lose it, and your resting metabolic rate drops, which feeds straight into the rebound problem: a lower-muscle body regains weight more easily, and tends to regain it as fat (covered in keeping the weight off). Run a few careless loss-and-regain cycles and you can land at the same scale weight with less muscle, more fat and a slower metabolism than you started with. The scale says "success"; your body composition says otherwise.

This is why the smarter framing isn't "how much did I lose" but "how much of what I lost was fat."

The ratio with GLP-1s - and where it can go wrong

The headline trials are actually reasonable on this. In SURMOUNT-1, tirzepatide produced roughly three times more fat loss than lean loss - a sensible ratio. The problem isn't that these drugs are uniquely catabolic; it's that rapid loss without protein and training pushes the lean share toward the bad end of that 25–40% range, and the effect bites hardest in people who start older or already lean.

What you actually control

None of this is exotic. It's the unglamorous scaffolding that decides whether your weight loss is quality weight loss - and it's exactly what supervised care is supposed to build in rather than leave to chance.

The frontier: drugs designed to spare muscle

This is where the science is moving, and it's worth knowing - clearly labelled as investigational and not offered here. A class of activin-receptor inhibitors (the lead candidate is bimagrumab) is being studied specifically to preserve or build muscle while cutting fat. In the phase 2 BELIEVE trial (Nature Medicine 2026), the contrast was striking: semaglutide alone reduced lean mass by about 7.4%, bimagrumab alone increased it by about 2.5%, and the combination held lean loss to ~2.9% while pushing total weight loss higher - with about 92% of the weight lost coming from fat. Earlier work (Heymsfield et al., JAMA Network Open 2021) had already shown bimagrumab cutting fat mass ~20% while raising lean mass.

The point for a patient today isn't to chase an unapproved drug - it's that the field now takes "quality of weight loss" seriously, and you don't need an investigational antibody to act on it. Protein and lifting do the same job you can start on now.

The bottom line - and where to start

The drug handles fat. You and your physician handle muscle - with protein, resistance training, a reasonable pace, and monitoring to confirm it's working. That's the difference between losing weight and losing it well, and it's most of what separates a result that lasts from one that rebounds worse than baseline.

Explore it with verified supply. Every Vivre Labs lot is independently tested and the certificate of analysis is published, so you know exactly what is in the vial. Explore tirzepatide.

Related reading

• Keeping the weight off after you stop - why muscle loss feeds weight regain
• Why your weight loss stalls on a GLP-1 - the metabolic-floor link to plateaus
• "Ozempic face" and hair loss, explained - the same "lose it well" levers for skin and hair
• What to eat on a GLP-1 - protein-first eating to protect muscle
• How sarcopenia is actually diagnosed - why the diagnosis turns on strength and function rather than on the lean-mass number a scan reports
• Looksmaxxing, the evidence-based version - body composition as the biggest aesthetic lever · GLP-1s and bone density · Function as a biological-age marker · Loose skin after rapid weight loss

---

Educational content, not medical advice. Exercise, nutrition and medication decisions must be made with a qualified clinician. Bimagrumab is investigational, not approved, and not offered by Vivre Labs; it is referenced here only as research context. Trial figures describe study populations; individual results vary. Claims for the Malaysian market were reviewed and cleared by MAB (Jun 2026).

Primary references

• Jastreboff AM, et al. Tirzepatide once weekly for the treatment of obesity (SURMOUNT-1). N Engl J Med. 2022;387:205–216. (≈3× fat-vs-lean reduction)
• Heymsfield SB, et al. Effect of bimagrumab vs placebo on body fat mass among adults with type 2 diabetes and obesity: a phase 2 randomized clinical trial. JAMA Netw Open. 2021;4(1):e2033457.
• Real-world retrospective cohort (n=486) on body composition during GLP-1RA / GIP-GLP-1RA therapy, specialist obesity outpatient clinic, Vienna, 2022–2025; presented at the European Congress on Obesity 2026. 80–85% of weight lost was fat mass; mean fat mass reduction ~18%, mean skeletal muscle mass loss ~5%; relative skeletal muscle mass preserved or increased in >70% of patients. Evidence class: real-world retrospective cohort, single centre, not randomised. Summary reporting: EurekAlert (https://www.eurekalert.org/news-releases/1128021) · News-Medical (https://www.news-medical.net/news/20260513/GLP-1-obesity-drugs-mainly-reduce-fat-while-preserving-muscle-mass.aspx). Figures to be re-verified against the full publication when it appears.
• BELIEVE phase 2 trial (bimagrumab ± semaglutide). Nature Medicine 2026 (NCT05616013). (lean-mass: −7.4% semaglutide vs −2.9% combination; ~92% of loss from fat)
• Stefanakis K, et al. The impact of weight loss on fat-free mass, muscle, bone and hematopoiesis. Metabolism 2024;161:156057. (25–40% lean-loss context)