"Muscle growth and fat loss" is the fastest-rising phrase in this niche and the hardest promise in physiology. Two processes with opposite energy requirements, running at once, on purpose. This page collects what has actually been measured — including the one place it has been measured, which is not where the marketing points — and does the arithmetic on the drug class that genuinely produces large fat loss. It is research journalism about research compounds and contains no usage guidance, per our editorial standards.
The one measured result
Growth hormone. Not a secretagogue, not a fragment: the hormone itself, in healthy older adults, in randomised trials, pooled.
Liu and colleagues at Stanford screened the randomised controlled trials of growth hormone in the healthy elderly and found 31 articles describing 18 unique study populations — 220 participants who received growth hormone, 107 person-years of treatment, mean age 69, mean BMI 28, mean initial daily dose 14 micrograms per kilogram, mean treatment duration 27 weeks. The body-composition results: fat mass down 2.1 kg (95% CI -2.8 to -1.35), lean body mass up 2.1 kg (CI 1.3 to 2.9), both at p < 0.001, and body weight essentially unchanged at +0.1 kg (Liu et al., Annals of Internal Medicine 2007).
That is recomposition, and it is the ceiling for everything in this category — because every secretagogue on the market is sold as a way of producing endogenous growth hormone, and cannot plausibly exceed the effect of administering the hormone directly.
Three things about that ceiling belong in the same breath. Bone density and serum lipids did not change. Participants treated with growth hormone were "significantly more likely to experience soft tissue edema, arthralgias, carpal tunnel syndrome, and gynecomastia" and somewhat more likely to develop diabetes or impaired fasting glucose. And the review's own conclusion is unambiguous: "On the basis of this evidence, GH cannot be recommended as an antiaging therapy."
No strength or performance outcome was synthesised at all. Two kilograms of lean mass on a scan, in a population averaging 69 years old, over roughly six months, with no measurement of whether anything got stronger.
Why the 2 kg is smaller than it sounds
Lean body mass on DXA or bioimpedance is not muscle. It is everything that is not fat or bone mineral: contractile tissue, connective tissue, glycogen and — the relevant one here — water. Soft-tissue oedema was among the most common adverse effects in the trials above. A hormone that causes fluid retention will move a lean-mass number partly by moving fluid, and no meta-analysis of scan data can separate those without imaging.
This is not a technicality invented for this page. The 2026 systematic review discussed below deliberately reported CT and MRI results separately from DXA and bioimpedance results, and applied different benchmarks to each, precisely because the two methods measure different things. Our evidence ledger applies the same rule to every row: a lean-mass change on a scan is graded below a muscle measurement by imaging, which is graded below a strength outcome.
The secretagogues, on this specific question
None of them has a randomised trial with a body-composition endpoint in trained adults. What exists is pharmacology — evidence that they raise the hormone.
CJC-1295 produced dose-dependent 2- to 10-fold increases in mean plasma growth hormone lasting six days or more, and 1.5- to 3-fold increases in IGF-1 lasting nine to eleven days, in healthy adults across 28- and 49-day ascending-dose studies (Teichman et al., JCEM 2006). That is a well-conducted pharmacokinetic and pharmacodynamic study. It contains no body-composition outcome. Ipamorelin's only randomised human trial was in postoperative ileus and missed its endpoint. Sermorelin, GHRP-2 and GHRP-6 have human data of the same shape: hormone release measured, body composition not.
So the chain being sold runs: this compound raises growth hormone (measured), growth hormone changes body composition (measured, about 2 kg each way, in older adults, with adverse events), therefore this compound changes body composition (not measured). The middle link is real and the conclusion has never been tested.
Tesamorelin: the approved one, and what it was approved for
Tesamorelin is the growth-hormone-releasing analogue that made it through development, and its trial data are the best in this family. In HIV patients with central fat accumulation on antiretroviral therapy, 273 patients received 2 mg of tesamorelin daily by subcutaneous injection against 137 on placebo for 26 weeks, then were re-randomised for a further 26. Visceral adipose tissue fell 18% and the change was sustained across 52 weeks, with triglycerides down 51 mg/dl, and the drug was generally well tolerated with no clinically significant glucose changes (Falutz et al., AIDS 2008).
The paper's closing finding is the one that gets left out of the marketing: "Upon discontinuation of tesamorelin, VAT reaccumulated", and "these effects do not last beyond the duration of treatment."
This is a visceral-fat drug for a specific metabolic complication, tested in patients with that complication, with an effect that reverses on stopping. It is not a recomposition agent for healthy lifters, and no trial has tested it as one.
AOD-9604 and the fragment story
AOD-9604 is the C-terminal fragment of growth hormone marketed on the claim that it separates the fat-loss effect from the growth effects. PubMed lists 22 records of any kind for it as of 2026-09-06 — reviews, patent surveys and analytical papers, not a trial programme. The drug-development literature records the position: a 2004 review in Current Opinion in Investigational Drugs notes that the developer was pursuing it for obesity and that "by February 2002, phase IIa trials were underway" (Wilding 2004), and it still appears among anti-obesity candidates in a 2012 review of the field (Khan et al. 2012). More than two decades after those phase IIa trials, it has not been approved for obesity anywhere. The absence of a published pivotal result, in a compound that reached phase II and then stopped, is itself information.
The class that does produce fat loss, and what it costs in muscle
The honest answer to "what causes large fat loss in humans" in 2026 is the incretin drugs, and the recomposition question has now been asked of them properly.
Semaglutide at 2.4 mg weekly produced a mean 14.9% reduction in body weight over 68 weeks against 2.4% on placebo, in 1,961 adults — a mean 15.3 kg against 2.6 kg (Wilding et al., New England Journal of Medicine 2021). That is the effect size no peptide in this article approaches.
What it takes with it was quantified this year. Batsis and colleagues screened 8,102 titles and analysed 35 randomised studies of liraglutide, semaglutide, tirzepatide or dulaglutide reporting body composition, median duration 26 weeks. Within the incretin groups, "the median proportion of total weight loss attributable to reductions in muscle-based indices was 28.3% (IQR, 15.9% to 39.9%), with 65% exceeding the benchmark of about 25%". Measured by bioimpedance or DXA the median was about 29%; measured by CT or MRI, about 25.3%, with two thirds exceeding the 15% benchmark that applies to skeletal muscle specifically. The review notes plainly that "no study reported objective physical function outcomes" (Batsis et al., Annals of Internal Medicine 2026).
Combining those two findings gives the arithmetic that the phrase "muscle growth and fat loss" is usually hiding. On a 15.3 kg mean loss, a 28.3% muscle-based share is roughly 4.3 kg of fat-free mass. This is our arithmetic across two separate sources rather than a figure either paper reports, and the caveats are real — the incretin review pooled many trials rather than STEP 1 alone, and fat-free mass includes water and glycogen. But the direction is not in doubt, and it is the opposite of recomposition. The comparator matters too: the review's 13 studies with lifestyle or placebo comparators produced a median weight loss of just 2.5%, of which 38% exceeded the same benchmark — losing weight by any route costs some lean tissue, and the drugs simply lose much more weight.
MuscleLedger's position is that this makes the incretins a fat-loss story with a muscle problem, not a recomposition story. The trials testing whether resistance training closes that gap are running now; none of them involves any compound discussed elsewhere on this page.
What the evidence supports
Nothing in this category has demonstrated simultaneous muscle gain and fat loss in trained adults. The only measured example anywhere in the family is growth hormone in older adults, at about 2 kg each way, with a documented adverse-event profile and no strength outcome, and its own reviewers declined to recommend it. Everything sold as a route to that result is one untested inferential step away from it, and the compounds producing genuinely large fat loss are documented to remove muscle-related mass rather than add it.
The interventions with human evidence for gaining muscle while losing fat remain resistance training and adequate protein. The one legal supplement with a measured effect on muscle size is creatine, and that effect is small — our head-to-head on creatine and the research compounds reports how small, and who funded the studies that measured it.
Sources and dates
- Liu et al., Annals of Internal Medicine 2007 — growth hormone in the healthy elderly
- Falutz et al., AIDS 2008 — tesamorelin, 52-week extension
- Teichman et al., JCEM 2006 — CJC-1295
- Wilding 2004, Current Opinion in Investigational Drugs — AOD-9604 · Khan et al. 2012 — anti-obesity candidates
- Wilding et al., New England Journal of Medicine 2021 — STEP 1
- Batsis et al., Annals of Internal Medicine 2026 — incretin body-composition systematic review
- PubMed record counts run 2026-09-06. WADA 2026 Prohibited List, sections S0, S2.2 and S2.3.