Obesity drugs now remove a quarter of body weight or more. The newest trials ask what, exactly, is coming off.
Ask PeptideBlog about any peptide
Last reviewed: October 30, 2026. Updated to add Biohaven’s October 29 taldefgrobep alfa announcement and the published TRIUMPH-1 paper.
What happens when obesity drugs become extraordinarily effective at reducing body weight, but researchers still need to determine how much of that reduction comes from fat versus muscle?
That question became urgent in 2026. In May, Eli Lilly announced that adults on the highest dose of retatrutide in its phase 3 TRIUMPH-1 trial lost an average of 28.3 percent of body weight over 80 weeks in the company’s efficacy analysis, and 45.3 percent reached a 30 percent loss, a level Lilly described as long associated with bariatric surgery. Those figures come from the company announcement. The peer-reviewed New England Journal of Medicine paper, published on September 29, reports a more conservative intention-to-treat average of 25.0 percent at 12 mg against 3.9 percent on placebo. Either way, the figures extend a pattern: each generation of drug moves the scale further than the last.
A scale reports one number. It cannot separate a kilogram of fat from a kilogram of lean tissue, and the two do different work. Excess fat drives diabetes and cardiovascular risk. Muscle clears glucose, supports the skeleton, and decides whether a 70-year-old can stand from a chair unaided.
This feature follows the evidence trial by trial. It separates peer-reviewed results from company announcements, marks what is still animal research, and spends as much space on what the studies cannot show as on what they can.
The weight-loss paradox in plain terms
The tension is simple. The stronger the drug, the more the composition of the loss matters and the less the scale can say about it. Ten percent of body weight lost, a quarter of it lean, is a modest tissue cost. Twenty-eight percent lost in the same proportion is a large one.
Drugs can also win on one measure and tie on another. Apitegromab, covered below, left total weight loss unchanged and shifted only the mix. CagriSema produced far more weight loss than semaglutide in Novo Nordisk’s REDEFINE 1 trial, yet lean soft tissue made up 33.1 percent of its loss in a DXA subset against 30.3 percent for semaglutide. A drug can be better at weight lost, at quality of weight lost, or at function. Trials have measured the first two far more often than the third.
What the semaglutide and tirzepatide data show
Semaglutide’s numbers come from an exploratory DXA analysis of STEP 1, which scanned 140 of 1,961 participants. At 68 weeks, semaglutide 2.4 mg cut body weight 15.0 percent (placebo 3.6), fat mass 19.3 percent and lean body mass 9.7 percent. Because fat fell faster, lean tissue’s share of body mass still rose 3.0 percentage points.
Tirzepatide’s come from a SURMOUNT-1 DXA substudy of 160 of 2,539 participants, published in 2025. At 72 weeks, tirzepatide cut weight 21.3 percent, fat mass 33.9 percent and lean mass 10.9 percent; placebo cut them 5.3, 8.2 and 2.6 percent. About three-quarters of the weight lost was fat and one-quarter lean in both groups. Six of the eight authors work for Lilly, the sponsor.
The matching split in the placebo arm fits the view that lean tissue loss tracks the amount of weight lost, whatever the cause. Eric Ravussin of Pennington Biomedical Research Center puts lifestyle-driven weight loss at about 25 percent fat-free mass and GLP-1 therapy at 25 to 40 percent. A 2024 review by Neeland and colleagues found lean mass at 40 to 60 percent of weight lost in some GLP-1 trials and 15 percent or less in others, citing differences in population, drug and comorbidity. Both substudies are small slices of larger trials, and both measure lean mass, which is not the same as muscle.
Lean mass is not the same as skeletal muscle
DXA, the scan behind nearly every figure above, sorts the body into fat, bone mineral and everything else. Everything else is lean mass, which Neeland’s review notes includes organs, bone, fluids and water held inside fat tissue alongside muscle. Heavier bodies also carry more lean tissue to move them, so some loss is expected as they shrink. Ravussin estimates fat-free mass at about 75 percent of body weight in a lean person and closer to 50 percent in severe obesity.
MRI measures muscle volume directly. A post hoc analysis of the SURPASS-3 trial in type 2 diabetes scanned 246 participants at week 52. Pooled tirzepatide reduced thigh muscle volume by 0.64 liters, a drop in line with what UK Biobank data predict for that much weight loss, while fat inside the muscle fell more than predicted. The authors read this as an adaptive response. Neeland’s review agrees, while warning that older age and disease severity may raise the risk of sarcopenia, the loss of muscle with age.
Lean mass on a scan is a biomarker. Strength and mobility are the outcome.
Function sits above both. It is measured with grip dynamometers, chair-stand tests and questionnaires, and obesity trials have used them sparingly. A 2024 Lancet Diabetes & Endocrinology commentary by Prado, Phillips, Gonzalez and Heymsfield carries the debate in its title: Muscle matters. Keep three tiers in mind below: lean mass on DXA, muscle volume on MRI, and what a person can do. Nearly every trial reports the first.
Taking the brakes off muscle: the antibody programs
Muscle growth is held in check by signaling proteins. Myostatin (GDF8) is the best known, and activins are relatives that share some receptors. Blocking that signaling should, in theory, help muscle hold on during weight loss. Bimagrumab blocks the receptors themselves, the activin type II receptors. Trevogrumab and apitegromab neutralize myostatin, apitegromab by binding its inactive precursor. Garetosmab neutralizes activin A. A fourth program, Biohaven’s taldefgrobep alfa, has its own section below.
These are antibodies, large proteins that fall outside the peptide class, tested as add-ons to peptide drugs like semaglutide and tirzepatide. Taldefgrobep alfa is an engineered fusion protein rather than a classic antibody.
Bimagrumab and the BELIEVE trial
Bimagrumab began as a treatment for muscle-wasting conditions and reached Lilly through its 2023 purchase of Versanis Bio. An earlier 48-week trial in 75 adults with type 2 diabetes and obesity reported fat mass down 20.5 percent and lean mass up 3.6 percent, which set up BELIEVE. PeptideBlog’s earlier analysis of the BELIEVE trial covers that study in more depth.
BELIEVE snapshot
- Source
- Heymsfield et al., Nature Medicine 2026; Lilly-funded; NCT05616013
- Design
- 507 adults, nine arms. Bimagrumab by infusion every 12 weeks and weekly semaglutide, alone or combined, plus placebo. 48 randomized weeks, then an open-label extension to week 72.
- Week 48
- Weight change of −9.3 kg (bimagrumab 30 mg/kg), −14.2 kg (semaglutide 2.4 mg), −17.8 kg (combination) and −3.3 kg (placebo).
- Week 72
- Weight −22.1% combination, −15.7% semaglutide, −10.8% bimagrumab. Lean mass −2.9%, −7.4% and +2.5%. Fat was 92.8% of weight lost with the combination and 71.8% with semaglutide.
- Limits
- DXA, not MRI. The extension was open label. Muscle spasms, diarrhea and acne were common with bimagrumab.
The combination stands out: weight loss well beyond semaglutide alone, with lean mass falling far less. Two details temper it. Bimagrumab by itself produced less weight loss than semaglutide, so the better mix comes from adding a second drug with its own tolerability cost. And the functional result highlighted at the 2025 American Diabetes Association presentation was a self-reported physical-functioning score. A questionnaire is not a strength test. Heymsfield, the lead investigator, notes that activin receptors also sit on fat cells, a possible reason bimagrumab cuts fat as well as sparing lean tissue.
Lilly is now testing bimagrumab with tirzepatide. A 252-person phase 2 study (NCT06643728) has finished enrolling, a 63-person Massachusetts General Hospital study (NCT05933499) is recruiting, and a diabetes study (NCT06901349) was withdrawn before enrolling anyone. ClinicalTrials.gov lists only phase 2 bimagrumab obesity trials.
Trevogrumab, garetosmab and the COURAGE program
Regeneron’s COURAGE program (NCT06299098) is the largest muscle-preservation trial discussed here, with 975 adults across two parts. Results were presented at European Association for the Study of Diabetes meetings in 2025 and 2026, and a Lancet paper is in press, so the figures below remain company-reported. Our COURAGE trial page tracks the registry record.
COURAGE snapshot
- Source
- Regeneron releases of September 2025 and October 2026; EASD presentations
- Design
- Part A (599 adults): semaglutide 2.4 mg with placebo, trevogrumab 200 or 400 mg, or trevogrumab plus garetosmab for 26 weeks, then semaglutide stopped. Part B (376 adults): trevogrumab 25 or 75 mg for 52 weeks, with an MRI substudy.
- 26 weeks, DXA
- Lean mass −6.6% with semaglutide alone and about −3.4% to −3.9% with trevogrumab. Lean mass was 33% of semaglutide weight loss.
- 52 weeks, DXA
- Lean mass −7.3% with semaglutide alone, −5.8% with trevogrumab 25 mg and −4.2% with 75 mg.
- 52 weeks, thigh MRI
- Regeneron calculates that 72% (25 mg) and 69% (75 mg) of muscle-volume loss was avoided.
- After semaglutide stopped
- Lean mass 0.4% above baseline with trevogrumab, 2.4% below with placebo.
Two findings stand out. The MRI figures move the claim from lean mass toward muscle, though they come from 76 participants and the meeting abstract calls the substudy exploratory. And the lower trevogrumab doses did not meaningfully change weight loss, so the benefit appears as composition without extra weight lost.
Safety divided the arms. On the lower doses, 77 percent of participants reported an adverse event versus 82 percent on placebo, mostly nausea, constipation, diarrhea and vomiting. Adding garetosmab preserved more lean mass but brought more serious adverse events, discontinuations and muscle spasms, and two participants in that triplet group died within 26 weeks: one of undetermined cause with several cardiovascular risk factors, one of cardiac arrest with cardiovascular disease. Regeneron reports it identified no causal link, and it will not advance the triplet.
Garetosmab is a real, approved drug, just for something else. The FDA approved it in August 2026 as Pasatru for adults with fibrodysplasia ossificans progressiva, a rare disease in which bone forms in soft tissue, and its label warns of skin infections, nosebleeds and fetal harm. Regeneron plans a new phase 2 study of trevogrumab in older adults with obesity and reduced muscle mass or strength, the population where function matters most.
Apitegromab and the EMBRAZE trial
Apitegromab, from Scholar Rock, binds the inactive precursor of myostatin. The FDA approved it in September 2026 as Isembyld for spinal muscular atrophy. Use in obesity remains experimental.
EMBRAZE snapshot
- Source
- Pratley et al., Nature Medicine 2026; Scholar Rock-funded; NCT06445075
- Design
- 102 adults. Tirzepatide plus apitegromab (10 mg/kg by infusion every four weeks) or plus placebo, for 24 weeks.
- Result
- 1.9 kg less lean mass lost (80% CI 1.2 to 2.7 kg; P = 0.001), described as 54.9% retention. Weight loss was similar between arms.
- Mix of loss
- Lean mass was 14.6% of weight lost with apitegromab and 30.2% with placebo.
- Safety and limits
- Adverse events in 76% versus 71%, one serious event per arm. More than 80% women, mean age 43.
The paper reports an 80 percent confidence interval, looser than the 95 percent interval most readers expect. In expert reaction gathered by the Science Media Centre, commentators noted that the trial showed no clear improvement in physical function or cardiometabolic outcomes. One observed that the paper found no significant differences in grip strength or sit-to-stand tests, which could reflect a short trial or lean mass failing to track true muscle change. Unlike bimagrumab, apitegromab did not add to weight loss. Its effect here was confined to composition.
Biohaven’s New GLP-1 Strategy: Bigger Ambitions, Longer Wait
On October 29, 2026, Biohaven used its third-quarter earnings release to change the plan for its phase 2 obesity study of taldefgrobep alfa. The study will add a cohort that pairs the drug with a GLP-1-based therapy, and the company moved its expected topline data to the second half of 2027 for both the original monotherapy portion and the new combination portion. In its March 2026 release and again in its August report, Biohaven had guided to the second half of 2026.
The release disclosed no efficacy or safety data. Everything about how the drug performs in people with obesity is still open, so this section treats the announcement as news about trial design and timing, not trial results.
The biology: two brakes on muscle
Myostatin (GDF8) and the activins are signaling proteins from the same family. They attach to activin type II receptors on muscle cells and tell the cells to hold back on growth. The programs above intervene at different points on that route. Bimagrumab sits on the receptors, so myostatin, activins and related signals cannot use them. Trevogrumab captures myostatin itself, and apitegromab captures its inactive precursor so less active myostatin is released. Garetosmab captures activin A.
Taldefgrobep alfa, known earlier as BMS-986089, was developed by Bristol Myers Squibb and licensed to Biohaven, which first pursued it in spinal muscular atrophy. It is built on a fibronectin-based scaffold called an adnectin. Biohaven describes it as a myostatin-activin pathway inhibitor that targets free myostatin and, according to the company, also affects signaling by activins. Whether covering more of the pathway preserves more muscle, or only adds side effects, is a hypothesis to test. The bimagrumab and garetosmab data above show that broad blockade can bring muscle spasms and other tolerability costs, and Biohaven has released no safety data from its obesity study. The doses reported so far are collected on our taldefgrobep alfa trial page.
What changed in the study
Taldefgrobep alfa snapshot
- Source
- Biohaven releases of March 19, August 10 and October 29, 2026; registry entry NCT07281495
- Design
- Randomized, double-blind, placebo-controlled, dose-ranging study in adults with overweight or obesity, about 150 randomized at 20 US sites. Once-weekly and once-monthly subcutaneous dosing by autoinjector. A 24-week double-blind period, then a 24-week open-label extension. The registry excludes people with diabetes.
- Endpoints
- Primary: percent change in body weight at week 24. Secondary: percent change in total fat mass and total lean mass, plus safety. The registry lists no muscle-volume or strength outcome.
- Timeline
- Enrollment completed in the first quarter of 2026. Topline data were guided to the second half of 2026 in March and August. On October 29, Biohaven moved them to the second half of 2027.
- Change
- The study “is being amended” to add a GLP-1-based combination cohort. The release names no GLP-1 drug, cohort size or amendment date. When reviewed on October 30, 2026, the registry entry still listed only taldefgrobep alfa and placebo, with a September 2026 primary completion date.
- Results disclosed
- None.
Why the GLP-1 combination matters
Most people losing weight on medication take a GLP-1-based drug, so the muscle question is mostly a question about them. A myostatin-pathway drug tested alone can show what it does to weight, fat and lean mass by itself. It cannot show whether it protects muscle during the steep weight loss a GLP-1 drug produces, which is the claim the rest of the field is making. Bimagrumab, trevogrumab and apitegromab were each tested on top of semaglutide or tirzepatide for that reason. A combination cohort moves taldefgrobep alfa toward the same question, and Biohaven’s own framing of the drug as a treatment for body recomposition points the same way.
The cost is time. A cohort added after the monotherapy portion finished enrolling has to be designed, approved, enrolled and followed, and Biohaven has not said why it chose this route or how large the cohort will be. A later date is a scheduling fact. By itself it says nothing about whether the drug works, and nobody outside the study has seen the data.
How it compares with the other myostatin and activin programs
| Program | Where it acts | Human obesity data | Evidence type |
|---|---|---|---|
| Bimagrumab | Activin type II receptors | BELIEVE, 507 adults, phase 2, with semaglutide | Peer-reviewed |
| Trevogrumab | Myostatin | COURAGE, 975 adults, phase 2, with semaglutide | Company-reported; paper in press |
| Apitegromab | Inactive precursor of myostatin | EMBRAZE, 102 adults, 24 weeks, with tirzepatide | Peer-reviewed |
| Taldefgrobep alfa | Myostatin and related activins, per Biohaven | None disclosed; GLP-1 cohort being added | Sponsor-reported design only |
Biohaven’s subcutaneous weekly or monthly dosing also differs from the infusion schedules used for bimagrumab and apitegromab in the trials above. Whether that matters to patients will depend on results nobody has yet.
Two measurement points keep expectations grounded. The registry’s body-composition outcomes are total lean mass and total fat mass, which is lean mass and not skeletal muscle. Muscle volume on MRI and strength or mobility tests, the measures that speak to function, are not among its listed outcomes. Even a clean lean-mass result would leave the function question where EMBRAZE left it. What stays unproven is whether taldefgrobep alfa changes lean mass or muscle in people with obesity, whether it does so on top of a GLP-1 drug, what it costs in tolerability, and whether any of that translates to strength. Nothing in the October 29 release answers those questions.
The next generation: more weight lost, unclear muscle effects
Meanwhile the weight-loss drugs themselves keep strengthening, and each one reopens the composition question.
Retatrutide
Retatrutide activates GIP, GLP-1 and glucagon receptors. In TRIUMPH-1 (2,339 adults), the NEJM paper reports 80-week weight loss of 17.6, 23.7 and 25.0 percent at 4, 9 and 12 mg against 3.9 percent on placebo (treatment-regimen estimand). Lilly’s earlier efficacy-analysis announcement gave 19.0, 25.9 and 28.3 percent. The published body-composition data come from a phase 2 substudy in type 2 diabetes, where fat mass fell 23.2 percent at the top dose by week 36 and lean loss as a share of weight lost resembled other obesity drugs.
CagriSema
Cagrilintide mimics amylin, a fullness hormone, and CagriSema pairs it with semaglutide. REDEFINE 1 (3,417 adults) found 20.4 percent weight loss against 3.0 percent on placebo. In a 252-person DXA subset presented at ObesityWeek 2025, lean soft tissue made up 33.1 percent of the loss. A 2026 European Congress on Obesity report of the Novo Nordisk-funded analysis described better questionnaire-based function and no decline in muscle strength versus placebo.
Eloralintide
Lilly’s amylin receptor agonist cut weight 9 to 20 percent across doses at 48 weeks in a 263-person Lancet trial (placebo 0.4 percent; a separate research site summarizes the phase 2 results), with nausea and fatigue most common. An expert commenting at ObesityWeek put the fat share of weight lost at 60 to 70 percent, and a Lilly rat study reported a better mix than cagrilintide. Neither is peer-reviewed human evidence.
Across these three, more potency has not visibly changed the mix of what is lost. Percent-of-loss figures also misbehave at the extremes: in the REDEFINE 1 DXA placebo group, where little weight came off, lean tissue made up 43.3 percent of it. That is one more reason to ask for MRI and function data.
ZFP423: attacking the fat side instead
Not every strategy protects muscle directly. A Northwestern University team led by Joseph Bass published a mouse study in PNAS in September 2026 that targets fat. An antisense RNA therapy silences Zfp423, a gene that restrains white fat from turning into calorie-burning beige fat. Alone, it raised energy expenditure and improved glucose control in mice. Paired with semaglutide, mice lost more weight and fat than on semaglutide alone and lost 5.5 percent of lean mass versus about 10 percent, according to Northwestern.
This is the earliest evidence tier. The therapy was designed against the mouse gene, not its human counterpart ZNF423, and the team says it is now testing similar therapies in human cells. Northwestern has filed a provisional patent. Nothing in human trials supports the approach yet. A different non-GLP-1 route to energy expenditure, a creatine-driven heat pathway, is covered in PeptideBlog’s SANA research review.
The evidence side by side
Read the last two columns first. They show where each result came from and what limits it. Biohaven’s entry has a design and a timeline, and no result.
| Approach | Stage and size | Body-composition result | Evidence type | Key limit |
|---|---|---|---|---|
| Semaglutide, STEP 1 DXA subset | Phase 3 subset, n=140, 68 weeks | Lean −9.7%, fat −19.3% | Peer-reviewed abstract | Exploratory and small |
| Tirzepatide, SURMOUNT-1 DXA substudy | Phase 3 substudy, n=160, 72 weeks | Lean −10.9%, fat −33.9%; about 25% of loss lean | Peer-reviewed | 160 of 2,539; sponsor authors |
| CagriSema, REDEFINE 1 DXA subset | Phase 3 subset, n=252, 68 weeks | 33.1% of loss lean soft tissue (semaglutide 30.3%) | Conference-reported | Sponsor-funded analysis |
| Bimagrumab + semaglutide, BELIEVE | Phase 2, n=507, 48 weeks plus extension to 72 | Lean −2.9% versus −7.4% on semaglutide | Peer-reviewed | Infusions; DXA; open-label extension |
| Trevogrumab + semaglutide, COURAGE | Phase 2, n=975, 26 and 52 weeks | Lean −4.2% versus −7.3% at 52 weeks (75 mg); thigh MRI loss cut about 69% | Company-reported; paper in press | MRI n=76; triplet deaths |
| Apitegromab + tirzepatide, EMBRAZE | Phase 2, n=102, 24 weeks | 1.9 kg less lean loss (54.9% retained) | Peer-reviewed | No clear function gain; mostly young women |
| Taldefgrobep alfa, phase 2 obesity study (NCT07281495) | Phase 2, about 150 randomized, 24 weeks plus 24-week extension; GLP-1 cohort being added | No human obesity efficacy or safety results disclosed | Sponsor-reported design; registry | No data yet; topline now expected 2H 2027 |
| Retatrutide | Phase 2 substudy (type 2 diabetes, 189 enrolled); phase 3 weight data (n=2,339) | Fat −23.2% at top dose; lean share similar to other drugs | Peer-reviewed | No muscle-sparing claim; no phase 3 body-composition result identified |
| Eloralintide | Phase 2, n=263, 48 weeks | Fat 60 to 70% of loss (congress remark) | Not published | No human DXA paper |
| ZFP423 RNA therapy + semaglutide | Mice | Lean −5.5% versus about −10% | Peer-reviewed, animal | Mouse-specific; human untested |
What current research cannot establish
Each result above answers a narrow question. Several larger ones remain open.
- Whether preserved lean mass means better function. EMBRAZE found no clear grip-strength or chair-stand gains over 24 weeks, BELIEVE’s highlighted functional signal is a questionnaire, and the COURAGE summaries released so far report body composition.
- Who benefits. Older adults and people with sarcopenia face the most risk from muscle loss, yet EMBRAZE enrolled mostly younger women. COURAGE’s prespecified low-lean-mass subgroup lost more lean mass on semaglutide and numerically preserved more with trevogrumab, a hint and not a result.
- Whether any benefit lasts. The longest placebo-controlled data run 52 weeks, and many patients may take these medicines for years.
- What safety cost comes with it. Muscle spasms recurred in the bimagrumab and garetosmab data, the COURAGE triplet saw two deaths and heavy dropout, and bimagrumab and apitegromab were given by infusion, while the sources reviewed do not state how trevogrumab was given.
- How the drugs compare with exercise and nutrition. The COURAGE presenters say trevogrumab’s lean-mass preservation exceeded what has been reported for a GLP-1 drug plus moderate-to-vigorous exercise, while noting that trial differences rule out a direct comparison. No trial reviewed here pits a drug against a supervised resistance-training program. For the non-drug side, the Weight Loss Atlas keeps a plain-language protein and strength reference.
- How the newest entrant performs. Biohaven has disclosed no human obesity efficacy or safety results for taldefgrobep alfa, its GLP-1 combination cohort is still being added, and the readout now sits in the second half of 2027.
- What regulators will accept. No drug here is approved to preserve muscle during weight loss; the antibody approvals so far cover spinal muscular atrophy and fibrodysplasia ossificans progressiva.
Frequently asked questions
Do GLP-1 drugs cause muscle loss?
Trials show lean tissue falls along with fat. In the SURPASS-3 MRI analysis, thigh muscle volume fell in line with the weight lost and fat inside muscle fell more than expected. Whether that matters for strength depends on factors such as age and disease severity.
How much of GLP-1 weight loss is lean mass?
About 25 percent in the SURMOUNT-1 tirzepatide substudy, and 25 to 40 percent across GLP-1 therapies by Ravussin’s estimate. A 2024 review found a spread from roughly 15 to 60 percent across trials.
Is lean mass the same as muscle?
No. DXA lean mass covers muscle plus organs, fluid and other tissue. MRI measures muscle volume more directly, and strength tests measure function.
Are the muscle-preserving antibodies approved for weight loss?
No. Bimagrumab and trevogrumab have no approvals, and taldefgrobep alfa remains investigational. Apitegromab is FDA-approved for spinal muscular atrophy and garetosmab for fibrodysplasia ossificans progressiva, and neither approval covers obesity.
Did these drugs improve strength?
Not clearly. EMBRAZE found no clear grip-strength or chair-stand differences, and BELIEVE’s highlighted functional result was a questionnaire score. Regeneron has planned an older-adult trial to test function.
What is the weight-loss paradox?
As obesity drugs remove more weight, the mix of tissue lost matters more, yet scales and most trials cannot separate fat from muscle. Drugs can also differ on total weight lost, quality of weight lost and function, and few trials measure all three.
What would prove these drugs protect muscle?
Longer trials in older and frailer adults that measure strength, mobility and muscle volume by MRI as primary endpoints, report safety over years, and compare drugs with resistance training and protein, not only with placebo.
What is ZFP423?
A gene that restrains white fat from becoming heat-producing beige fat. A Northwestern mouse study silenced it with an RNA therapy and saw less lean-mass loss alongside semaglutide. Human data do not exist yet.
What is taldefgrobep alfa?
An investigational fusion protein that Biohaven describes as a myostatin-activin pathway inhibitor. Bristol Myers Squibb developed it first, and Biohaven is testing it in a phase 2 obesity study. No human obesity efficacy results have been released.
Why did Biohaven move its taldefgrobep alfa obesity data to 2027?
On October 29, 2026, Biohaven said it is amending the study to add a GLP-1 combination cohort and now expects topline data from both portions in the second half of 2027, after earlier guidance of the second half of 2026. The company gave no results and no reason, so the change cannot be read as evidence about whether the drug works.
Where this leaves the paradox
The most defensible reading of the 2026 data is a measured one. Three antibody programs working on two biological mechanisms each cut DXA lean-mass loss by about half or more in phase 2, two with peer-reviewed papers and one with a paper in press. That agreement across sponsors carries weight. It also carries a caution, because every one of those results is a body-composition result.
Patients do not live inside a DXA scan. They climb stairs, carry groceries and try to avoid falls. The trials that settle this will enroll older and frailer adults, measure strength and mobility as primary endpoints, run longer than a year, and compare drugs with resistance training and protein, not only with placebo. Regeneron’s planned older-adult study, the ongoing bimagrumab-tirzepatide trials and Biohaven’s expanded taldefgrobep alfa study point that way. Their results are pending, and Biohaven’s now sits in the second half of 2027.
Until then, “lean mass preserved” is a promising biomarker whose clinical meaning is still being tested. The paradox holds: the better these drugs get at removing weight, the more the unseen composition of that weight decides whether the result counts as a win.
This article is educational and summarizes published and company-reported research on approved and investigational drugs. It is not medical advice and does not recommend any compound or dose. Talk with a licensed clinician about your own health and treatment. Trial statuses change, so check ClinicalTrials.gov for current information.
Sources
Peer-reviewed papers
- Look M, et al. Body composition changes during weight reduction with tirzepatide in the SURMOUNT-1 study. Diabetes Obes Metab 2025;27(5):2720-2729. doi:10.1111/dom.16275
- Wilding JPH, et al. Impact of semaglutide on body composition in adults with overweight or obesity: exploratory analysis of STEP 1. J Endocr Soc 2021 (abstract). PMC8089287; parent trial: N Engl J Med 2021
- Neeland IJ, Linge J, Birkenfeld AL. Changes in lean body mass with GLP-1-based therapies and mitigation strategies. Diabetes Obes Metab 2024;26(Suppl 4):16-27. doi:10.1111/dom.15728
- Sattar N, et al. Tirzepatide and muscle composition changes in people with type 2 diabetes (SURPASS-3 MRI): post hoc analysis. Lancet Diabetes Endocrinol 2025. PubMed 40318682
- Prado CM, Phillips SM, Gonzalez MC, Heymsfield SB. Muscle matters: the effects of medically induced weight loss on skeletal muscle. Lancet Diabetes Endocrinol 2024;12(11):785-787. doi:10.1016/S2213-8587(24)00272-9
- Heymsfield SB, et al. Bimagrumab plus semaglutide alone or in combination for the treatment of obesity: a randomized phase 2 trial (BELIEVE). Nat Med 2026;32(3):869-882. doi:10.1038/s41591-026-04204-0
- Heymsfield SB, et al. Effect of bimagrumab vs placebo on body fat mass among adults with type 2 diabetes and obesity. JAMA Netw Open 2021;4(1):e2033457. JAMA Network Open
- Pratley RE, et al. Apitegromab for lean mass preservation during tirzepatide-induced weight loss (EMBRAZE). Nat Med 2026;32(7):2673-2678. doi:10.1038/s41591-026-04440-4
- Coskun T, et al. Effects of retatrutide on body composition in people with type 2 diabetes: a phase 2 substudy. Lancet Diabetes Endocrinol 2025;13(8):674-684. doi:10.1016/S2213-8587(25)00092-0
- Garvey WT, et al. Coadministered cagrilintide and semaglutide in adults with overweight or obesity (REDEFINE 1). N Engl J Med 2025;393(7):635-647. doi:10.1056/NEJMoa2502081
- Billings LK, et al. Eloralintide, a selective amylin receptor agonist for the treatment of obesity: a 48-week phase 2 trial. Lancet 2025;406:2631-2643. doi:10.1016/S0140-6736(25)02155-5
- Thorne AK, et al. An RNA thermogenic therapy to preserve lean mass and enhance metabolic health during GLP-1 weight loss (mouse study). PNAS 2026. doi:10.1073/pnas.2618845123
- Jastreboff AM, et al. Retatrutide, a triple hormone receptor agonist, for treatment of obesity (TRIUMPH-1). N Engl J Med 2026 (published September 29). doi:10.1056/NEJMoa2604169
Trial registries
- ClinicalTrials.gov NCT05616013 (BELIEVE). Registry entry
- ClinicalTrials.gov NCT06643728 (bimagrumab and tirzepatide, Lilly). Registry entry
- ClinicalTrials.gov NCT05933499 (tirzepatide and bimagrumab, Massachusetts General Hospital). Registry entry
- ClinicalTrials.gov NCT06299098 (COURAGE). Registry entry
- ClinicalTrials.gov NCT06445075 (EMBRAZE). Registry entry
- ClinicalTrials.gov NCT05929066 (TRIUMPH-1). Registry entry
- ClinicalTrials.gov NCT06230523 (eloralintide phase 2). Registry entry
- ClinicalTrials.gov NCT07281495 (taldefgrobep alfa phase 2 obesity study, Biohaven). Registry entry
Company announcements and conference reports (not peer-reviewed)
- Biohaven. Third-quarter 2026 financial results and recent business developments (Exhibit 99.1), October 29, 2026. SEC filing
- Biohaven. Second-quarter 2026 financial results (Exhibit 99.1), August 10, 2026. SEC filing
- Biohaven. Phase 2 obesity study of taldefgrobep alfa, March 19, 2026. Press release
- Cure SMA. Biohaven licenses an anti-myostatin adnectin for spinal muscular atrophy (background on origin and structure). Cure SMA
- Regeneron. Phase 2 COURAGE trial confirms trevogrumab prevents lean mass and muscle loss during GLP-1 receptor agonist-induced weight loss. October 2026. Press release
- Regeneron. Results from Phase 2 COURAGE trial presented at EASD. September 2025. Press release
- Regeneron interim COURAGE results with death details, as reported by Patient Care. Article
- Summary of the EASD 2026 COURAGE presentation. News-Medical; epocrates; decision not to advance the triplet: EMJ
- Lilly. Retatrutide TRIUMPH-1 phase 3 topline results. May 2026. Press release; efficacy-analysis figures: ConsultantLive
- BELIEVE presentation at the ADA Scientific Sessions 2025. ConsultantLive; ADA meeting news; Pennington Biomedical release
- REDEFINE 1 DXA analysis. Healio, ObesityWeek 2025; Ravussin interview; ECO 2026 coverage
- Eloralintide ObesityWeek 2025 discussion. Healio; Lilly rat study abstract: Diabetes 2025;74(Suppl 1):849-P; Lilly release: November 2025
- Northwestern University release on the ZFP423 mouse study. News-Medical; Inside Precision Medicine
Regulatory news and expert reaction
- FDA approval of apitegromab-mstn (Isembyld) for spinal muscular atrophy, September 2026. FDA
- FDA approval of garetosmab-grts (Pasatru) for fibrodysplasia ossificans progressiva, August 2026. Healio; American Pharmaceutical Review
- Expert reaction to the EMBRAZE paper. Science Media Centre; population details: Deutsches Aerzteblatt (German)
