Follistatin sells on its animal data: mice engineered to make extra follistatin carry far more muscle than normal mice, because the protein neutralises myostatin, the body's brake on muscle size. The human evidence is two small gene-therapy trials in patients with muscle-wasting disease, run by one Ohio group. What lifters buy as "follistatin 344" is not what those patients received, and no trial has tested it. This page counts what exists; it reports research and gives no usage guidance, per our editorial standards.
What it is — and why "344" is a gene
Follistatin is a glycoprotein that binds members of the TGF-β family — myostatin and the activins among them — and keeps them from reaching the activin type IIB receptor. Human follistatin is made as a 344-amino-acid precursor; the first 29 amino acids are a signal sequence that the cell cuts off, leaving a mature chain of 315 amino acids, about 35 kDa (UniProt P19883, read 2026-09-24). A shorter splice form lacks the last 27 amino acids.
So "follistatin 344" names the precursor gene product, which is why it appears in gene-therapy papers: the Ohio group chose the FS344 sequence so muscle cells would secrete the circulating form. As a molecule, follistatin is a protein roughly 27 times the mass of kisspeptin-10, not a short peptide — a point that matters for anyone assuming it behaves like the small peptides it is sold beside. The plain-language entry is Peptide Lexicon's follistatin page.
The human muscle evidence: twelve patients, one group
| Trial | Who | What was given | Result |
|---|---|---|---|
| Mendell 2015, Mol Ther (NCT01519349) | 6 men with Becker muscular dystrophy | Virus (AAV1) carrying FS344, injected into both quadriceps; two dose cohorts | Six-minute walk: +58, +125, 0 m (low dose); +108, +29, 0 m (high dose). Biopsies: less fibrosis, larger fibres. No adverse effects reported |
| Mendell 2017, Mol Ther | 6 people with sporadic inclusion body myositis, plus 8 untreated matched patients | Same vector, 6 × 10^11 vg/kg, both quadriceps, with an exercise regimen | Walk distance +56.0 m/year treated vs −25.8 m/year untreated (p = 0.01); four of six improved 58–153 m, two by 5–23 m |
Three features limit what these trials can say about a lifter. They are disease populations — muscle already breaking down, where the brake is not the main problem. They are not placebo-controlled: the Becker trial had no comparison group and the myositis trial compared against eight untreated patients matched for age, sex and baseline measures — not a randomised control group. And they were challenged: Steven Greenberg, a neurologist at Brigham and Women's Hospital, published a critique in the same journal titled "Unfounded claims of improved functional outcomes attributed to follistatin gene therapy in inclusion body myositis" (PMID 28927986), and a reply followed in the same issue.
The dose in these trials is a count of virus particles per kilogram per leg. It says nothing about how much follistatin the muscle then made, and there is no published way to turn it into a quantity of protein.
The census: nobody has injected the protein
ClinicalTrials.gov, 2026-09-24. An intervention search for follistatin returns 15 records. Read one by one, the ones that give follistatin all give it as a gene:
- NCT01519349 and NCT02354781 — the Nationwide Children's Hospital AAV1 trials (Becker, myositis, and a three-patient Duchenne study).
- NCT06411366 — Minicircle, "injectable follistatin plasmid gene therapy", 43 people, Honduras. Start date 2022-08-18, first submitted 2024-05-08 — about 21 months after it began. Listed as completed on 2023-08-31. No results posted.
- NCT07285629 — Minicircle again, klotho plus follistatin gene therapy, 14 people, United States and Honduras, started 2025-12-16.
- NCT07443826 — Unlimited Biotechnology, AAV9-follistatin with a VEGF plasmid for age-related muscle loss, 12 people, Honduras, planned start 2026-06-01.
The rest measure follistatin in blood as a marker — in obesity, cancer cachexia, Sjögren's syndrome, after exercise or omega-3 supplements. None injects follistatin protein, in anyone.
PubMed, 2026-09-24. follistatin[tiab] returns 2,963 records; with trial publication types, 90, most of which measure it as an outcome of exercise or diet. Restricted to trial-tagged human gene-therapy papers, 1 — the Becker trial.
None of that is evidence the therapy fails. It does mean the most recent human follistatin work was registered late, sits at sites in Honduras, and has published nothing a reader can check.
What the lifter's version would need to show
A vial of follistatin protein would have to survive injection, reach muscle in active form, and outlast its own clearance — none of which has been measured in a person. The closest human evidence for blocking this pathway comes from drug candidates rather than follistatin — ACE-031 among them, whose development was halted. That file, and the evidence score this site gives follistatin (1 of 5), is on peptides for muscle growth. Peptifact's page on the follistatin figures that circulate covers the numbers sold with the vials.
Status in sport and at FDA
- WADA: S4.3, agents preventing activin receptor IIB activation — follistatin, myostatin propeptide, apitegromab, bimagrumab, ACE-031 — prohibited at all times, non-Specified. The map is on WADA status by compound.
- FDA: no follistatin protein or follistatin gene therapy is approved for any use. A gene therapy given outside a trial is an unapproved biologic.
The honest summary
- "Follistatin 344" is a 344-amino-acid precursor; the active protein is 315 amino acids, about 35 kDa.
- Human muscle data: two gene-therapy trials, six patients each, both in muscle disease, neither placebo-controlled, one publicly challenged.
- No human trial has injected follistatin protein, and none has studied healthy or trained adults.
- The newest trials are plasmid and viral gene therapies registered from Honduras; the completed one was registered 21 months late and has posted no results.
- Prohibited in sport at all times (S4.3); not approved.
Limits of this page
The registry and PubMed counts depend on query wording and were taken on one date; a trial registered elsewhere, or published without a registry record, would not appear. The gene-therapy results come from the group that developed the vector, and the Becker trial's six results are reported individually rather than as a statistical test. Nothing here tests whether follistatin protein sold as a research chemical contains what its label says.
Sources and dates
- UniProt P19883, FST_HUMAN — 344 aa, signal peptide 1–29, chain 30–344, isoform lacking 318–344 — read 2026-09-24
- Mendell JR, Sahenk Z, Malik V, et al. A phase 1/2a follistatin gene therapy trial for Becker muscular dystrophy. Mol Ther 2015;23:192–201 — PMID 25322757
- Mendell JR, et al. Follistatin gene therapy for sporadic inclusion body myositis improves functional outcomes. Mol Ther 2017;25:870–9 — PMID 28279643
- Greenberg SA. Unfounded claims of improved functional outcomes attributed to follistatin gene therapy in inclusion body myositis. Mol Ther 2017;25:2235–7 — PMID 28927986
- ClinicalTrials.gov API v2, read 2026-09-24: intervention search
follistatin→ 15 records; NCT01519349, NCT02354781, NCT06411366 (Minicircle, start 2022-08-18, first submitted 2024-05-08, no results), NCT07285629, NCT07443826. - PubMed, searched 2026-09-24:
follistatin[tiab]→ 2,963; with trial publication types → 90; trial-tagged human gene therapy → 1. - World Anti-Doping Agency, 2026 Prohibited List, S4.3 — as mapped on this site's WADA status by compound.