Myostatin and activin: more tissue, more function?

Updated 10 Oct 2026

Removing a biological brake on muscle growth is a compelling idea. Human trials show that this can change body composition substantially. The harder question is whether the added or preserved tissue improves strength, mobility or disease outcomes enough to justify the tradeoff. The answer now includes both disappointing trials and a recent, indication-specific success.

The target and the product matter

Myostatin is one signal that limits skeletal-muscle growth. The activin receptor system handles several related signals, so blocking receptors and selectively preventing myostatin activation are different interventions. Bimagrumab is an antibody against type II activin receptors. Apitegromab binds inactive forms of myostatin and prevents their activation; it is a large monoclonal antibody, not a short peptide. Neither can be equated with an unspecified vial labeled myostatin inhibitor. Original bimagrumab study, FDA apitegromab mechanism and product description

The older lesson: size and strength can separate

In a randomized trial of 180 adults aged at least 70 with sarcopenia, six months of bimagrumab increased lean mass by about 7%, versus 1% with placebo. Both groups received nutrition and home exercise support. The primary physical-performance score improved in both groups, without a significant difference between them (P=0.13); walking-distance and gait-speed comparisons were also not significant. The investigators found no clinically relevant advantage in strength or mobility despite the composition change. Original sarcopenia trial

A separate 75-person trial in adults with obesity and type 2 diabetes found fat mass down 20.5% and lean mass up 3.6% with bimagrumab at 48 weeks, compared with little change under placebo. Handgrip strength did not show a treatment effect. Muscle spasms and diarrhea were more frequent; five active-treatment participants stopped because of adverse events, versus none under placebo. This was a small phase 2 trial using 80% confidence intervals, not proof of durable performance enhancement. Original diabetes–obesity trial

These null functional results do not prove the whole pathway cannot improve function. They establish that increasing tissue alone is an insufficient demonstration of that benefit.

BELIEVE: a stronger body-composition signal in obesity

The March 2026 BELIEVE report randomized 507 adults with obesity or qualifying overweight, without diabetes, across nine bimagrumab, semaglutide, combination and placebo groups. At 48 weeks, the highest-dose combination reduced fat mass 33.7%, versus 21.1% with semaglutide alone. Across combination groups, lean mass fell 0.8–2.3%, compared with 4.7–6.9% across semaglutide groups. This is meaningful evidence of better preservation of the measured lean compartment alongside fat reduction. Original BELIEVE report

Function was less decisive: one combination group improved grip strength relative to placebo at week 48; the other groups were similar to placebo. Grip was exploratory, comparisons were not adjusted for multiplicity, and semaglutide was open label. Only 377 of 507 completed the primary period; adverse-event discontinuations were particularly frequent in bimagrumab monotherapy groups. The 72-week extension is follow-up of the same trial. These qualifications matter when turning an impressive scan into a claim about useful muscle. Original BELIEVE methods, outcomes and limitations

EMBRAZE: preserving lean tissue during tirzepatide treatment

The June 2026 EMBRAZE report randomized 102 adults receiving tirzepatide to apitegromab or placebo. Its primary analysis used 87 completers with evaluable scans. At 24 weeks, apitegromab preserved 1.9 kg more lean mass (80% CI 1.2–2.7); lean losses were 1.6 versus 3.5 kg. Supportive intention-to-treat analyses agreed directionally. Total weight loss was similar between groups. Original EMBRAZE trial

The exploratory force-production and chair-stand assessments showed no notable group differences. Most participants were women, activity was not objectively measured, and the statistical plan used a two-sided 20% significance level for this exploratory phase 2 programme. The narrow-looking interval therefore has 80%, not 95%, coverage. One serious adverse event occurred in each arm, and overall adverse-event rates were similar. This is promising composition evidence with remaining functional and long-term questions, rather than a demonstrated benefit for trained young men. Original EMBRAZE outcomes and statistical plan

A real functional success, in a specific disease

It would be outdated to say this approach has produced no approved muscle-targeted therapy. FDA approved Isembyld, apitegromab-mstn, on September 11, 2026, for spinal muscular atrophy in people aged at least two who already receive an SMN2-targeted treatment. That indication addresses a neuromuscular disease; obesity preservation and healthy training are separate uses. FDA approval action, FDA Purple Book record

The SAPPHIRE phase 3 study enrolled 188 nonambulatory participants with type 2 or type 3 SMA, all receiving background treatment. In its 156-person primary age group, two to twelve, pooled apitegromab groups improved the motor-function scale relative to placebo by 1.8 points (95% CI 0.30–3.32; P=0.019) at a year. The higher-dose-only comparison was not significant (P=0.11); more drug did not produce a clearer result. The smaller older-age cohort and later subgroup reports are not independent replications. Original SAPPHIRE report

The FDA label reports the approved-dose group separately: 2.2 points relative to placebo (95% CI 0.49–3.95), with a nominal P value. That is a different comparison within SAPPHIRE, rather than a second trial contradicting its pooled 1.8-point result. FDA efficacy table

The approved label also warns about fractures. In the younger trial population at the approved dose, fractures occurred in 5 of 53, versus 1 of 50 under placebo; many people with fractures in the broader programme had pre-existing skeletal vulnerabilities. This does not estimate fracture risk for healthy users, but it directly challenges the assumption that increasing muscle signaling must protect every nearby tissue. FDA label and fracture warning

Broader inhibition has produced different tradeoffs

ACE-031, a soluble receptor-based intervention, was tested in boys with Duchenne muscular dystrophy. The trial stopped because of nosebleeds and dilated small vessels. Walking and composition findings were trends, without a significant walking benefit. These were safety concerns even though the report did not describe serious or severe adverse events. They are evidence about that intervention, not proof that every selective myostatin antibody has the same risk. Original ACE-031 trial

The COURAGE programme combines semaglutide with the anti-myostatin antibody trevogrumab, with or without the anti-activin A antibody garetosmab. Sponsor-reported 26-week results in 599 participants showed less DXA lean loss with combinations. The three-drug group had substantially more tolerability-related discontinuations and two deaths; the sponsor reported that it had not identified a causal association with treatment. The reported efficacy analysis excluded data after discontinuation and was not adjusted for multiplicity. These qualifications should accompany the promising composition figures. Its 2026 conference subgroup report is further analysis of the same programme. Sponsor’s COURAGE results, Original 2026 conference analysis

Follistatin gene transfer is a different exposure

An early six-person Becker muscular dystrophy study delivered a follistatin gene through an AAV vector to muscle. Four participants improved walking distance and two showed little or no improvement, without a randomized comparator. This provides a human research signal, but it tested gene transfer in a particular disease, not injection of a retail follistatin protein. The initial paper and a later author review describe the same six people and cannot be counted as independent successful trials. Original follistatin gene-transfer study

What would change the practical conclusion?

For obesity treatment, the next useful evidence is sustained functional benefit or prevention of meaningful disability alongside composition, with tolerability and durability assessed in an appropriate comparative study. For my own recovery question, these studies do not establish treatment for anterior shoulder discomfort, tendon healing or safe escalation of training. The recovery investigation and incretin investigation explain the directly relevant endpoints. The research is worth following because the biology is real and outcomes are becoming more specific—not because every intervention in the pathway is interchangeable.

Back to investigations