Follistatin-344 Is a Gene Therapy Payload, Not a Peptide — and Myostatin Inhibition Adds Mass Without Function

"Follistatin-344" names a 344-residue human glycoprotein of about 38,000 daltons. It is the payload gene therapy trials deliver as DNA, not a peptide in a vial. And myostatin inhibition, the reason anyone wants it, keeps adding mass without adding function.

Those are two independent findings and either one is separately disqualifying. This article resolves what follistatin actually is before it discusses what it might do, inside the growth hormone peptides category, using UniProt sequence data, five registered trials with posted or conspicuously absent results, and Peptigrity's own Purity Index testing record for follistatin.

What is Follistatin-344, and is it a peptide?

Follistatin-344 is a glycoprotein, not a peptide. UniProt P19883 records the full-length human follistatin precursor at 344 residues and 38,007 daltons, with residues 1–29 forming a signal peptide that is cleaved during secretion. The forms that circulate in blood are FS-315 and FS-288. A vial containing an intact 344-residue chain would hold the uncleaved precursor, which is not the species found in human circulation.

The biology behind the interest is real. Follistatin binds and neutralises activins and myostatin (GDF-8), and neutralising myostatin is the mechanism every muscle claim in this category rests on.

What is not real is the implied product. The number 344 in the product name is the precursor count, and the first 29 of those residues are removed by the cell on the way out. Nothing with 344 residues intact circulates in a human being. That is a fact about the molecule, available from a public sequence database, and it is the first thing a buyer should check.

Can a 344-residue glycoprotein be made by peptide synthesis?

No. Follistatin-344 cannot be built by solid-phase peptide synthesis, which works routinely up to roughly 50 residues and degrades in yield as chains lengthen, because every coupling step is imperfect and errors compound. A 344-residue chain would require over 340 sequential couplings. Follistatin is also glycosylated, and solid-phase synthesis cannot glycosylate anything, so a genuine protein product would have to be expressed recombinantly in a cell line, folded, glycosylated and purified.

Solid-phase peptide synthesis is the process that produces almost everything else on this platform, and its limits are well understood — see how peptides are made. Sugar chains, by contrast, are added by cellular machinery in the endoplasmic reticulum and Golgi. There is no synthetic bench route to them.

So a genuine follistatin protein product would be a biologics manufacturing operation, not peptide synthesis, and biologics carry a different quality regime entirely — the distinction our guide to GMP vs non-GMP peptide manufacturing sets out. Whether that is what any vendor is selling is not something a buyer can determine from a certificate of analysis.

Note the sequence of facts here. The 344 residue count, the signal peptide at 1–29, the 38 kDa mass and the glycoprotein classification are all directly verifiable from UniProt. The inference about synthesis limits follows from them, and it is the question a buyer should be asking rather than a claim about any specific product.

Has follistatin ever been given to humans as a protein?

We found no registered trial of follistatin protein administration. The two human follistatin studies we identified both deliver genes: NCT01519349, a 15-patient intramuscular transfer of rAAV1.CMV.huFollistatin344 at Nationwide Children's Hospital, and NCT06411366, a 43-participant plasmid study run by Minicircle in Roatán, Honduras. Neither has posted results on the registry, and the one publication reporting a functional benefit was formally challenged in print.

The second trial's own description is explicit about the mechanism: the plasmid is "designed to express and secrete bioidentical human follistatin into serum circulation." Both approaches deliver genetic instructions for the body to make the protein. Neither delivers the protein in a vial.

The publication question is where this gets contested. A paper reported positive findings — Al-Zaidy, Sahenk, Rodino-Klapac, Kaspar & Mendell, "Follistatin Gene Therapy Improves Ambulation in Becker Muscular Dystrophy," Journal of Neuromuscular Diseases 2015;2:185–192 (PMID 27858738). And that paper was directly challenged. Greenberg SA, "Unfounded Claims of Improved Functional Outcomes Attributed to Follistatin Gene Therapy in Inclusion Body Myositis," Molecular Therapy 2017;25:2235–2237 (PMID 28927986). Anyone citing the 2015 result should cite the rebuttal alongside it.

Does myostatin inhibition actually build usable muscle?

Myostatin inhibition reliably increases muscle mass and reliably fails to improve function. In RESILIENT (NCT01925209, phase 2b/3, n=251), bimagrumab lifted lean body mass to 102.8% of baseline against 97.2% on placebo, while every six-minute-walk comparison was non-significant at p=0.2210, p=0.1909 and p=0.9263. Four large pharmaceutical companies pursued this pathway; domagrozumab's Duchenne trial was terminated on 30 August 2018 for "Lack of Efficacy."

This is the most important evidence in this article, and it comes from pharmaceutical trials with posted results rather than from anyone selling anything.

Bimagrumab in sporadic inclusion body myositis — RESILIENT, NCT01925209, phase 2b/3, n=251. Primary endpoint, change in 6-minute walking distance at week 52 (least-squares means, metres):

All non-significant. Meanwhile lean body mass at week 52, as a ratio to baseline: 102.8% on 10 mg/kg (95% CI 101.4–104.2) versus 97.2% on placebo (95% CI 95.9–98.5). Mass up. Walking, unchanged.

Bimagrumab in sarcopenia — InvestiGAIT, NCT02333331, phase 2, n=217. This is the cleanest published demonstration of the dissociation, because both outcome families sit inside the same trial.

Every functional endpoint non-significant. Lean body mass significant at p<0.001 in two of three arms.

Two further programmes close the pattern. Domagrozumab in Duchenne muscular dystrophy — NCT02310763, phase 2, 121 enrolled. TERMINATED on 30 August 2018. Reason given: "Lack of Efficacy." MYO-029 (stamulumab) — NCT00104078, Wyeth, phase 1/2, n=108 in Becker, facioscapulohumeral and limb-girdle muscular dystrophy, completed January 2007, no results posted. Landogrozumab (LY2495655) ran five registered trials, including phase 2 studies in pancreatic cancer, in older people who had fallen with muscle weakness, and around elective hip replacement.

Four large pharmaceutical companies. Multiple indications. Real money, real trials, posted results. Muscle mass goes up. Function does not.

That is not a failure of dosing or a series of underpowered studies. It is a consistent finding across programmes, and the most plausible reading is that myostatin inhibition adds tissue without adding the neural drive, tendon adaptation and motor control that turn tissue into performance. It is the single result anyone comparing muscle-growth compounds — including genuinely synthesisable ones such as IGF-1 LR3 and PEG-MGF — should weigh first.

Which follistatin claims survive the evidence?

One of nine. Of the nine claims routinely made for follistatin, exactly one survives: follistatin does bind and neutralise activins and myostatin (GDF-8), which is real biology. The other eight fail on identity, manufacturing, trial record, function, or regulation — including "not banned in sport", which is false, since WADA names follistatin by name at S4.3, prohibited at all times, in and out of competition.

Read the verdict column downward and the structure of the problem is visible: the only surviving row is the one about what follistatin does inside a body that made it itself.

Is follistatin approved, and is it banned in sport?

Follistatin sits on none of FDA's compounding lists and is prohibited in sport by name. We enumerated both tables on the page current 22 April 2026 — 14 substances in active Category 2, 17 nominated but withdrawn — and follistatin is on neither. That absence is not clearance: it has no USP monograph, is not a component of any approved drug, and is not on the 503A bulks list.

Those are the three statutory routes for compounding, and follistatin satisfies none of them. The wider regulatory picture is in our FDA peptide regulation timeline.

WADA names follistatin explicitly, and not in the section most people check. It is not in S2 with the growth factors. It is in S4, "Hormone and Metabolic Modulators," at S4.3, "Agents Preventing Activin Receptor IIB Activation," prohibited at all times. The listing, verbatim, covers "Activin A-neutralizing antibodies," "decoy activin receptors (e.g. ACE-031)," "anti-activin receptor IIB antibodies (e.g. bimagrumab)," and "myostatin inhibitors such as: agents reducing or ablating myostatin expression; myostatin-binding proteins (e.g. follistatin, myostatin propeptide); myostatin- or precursor-neutralizing antibodies (e.g. apitegromab, domagrozumab, landogrozumab, stamulumab)."

Follistatin, by name. Prohibited at all times, in and out of competition, alongside every failed pharmaceutical programme described above. This is not an S0 catch-all inference — it is a named substance in a named category, and anyone subject to testing should read our note on whether peptides show up on drug tests with that placement in mind.

What do independent lab tests show about follistatin?

Follistatin has been tested 7 times across 6 laboratories, averaging 99.02% purity against 217 verified shops (verified August 2026), with individual results running 97.57% to 99.80%. One of those seven is a 40% fill shortfall: a May 2026 sample labelled 1 mg tested at 0.6 mg. Others ran from +2% to +32% over label, and only one test reports an endotoxin result.

The Purity Index record comes from Freedom Diagnostics, Janoshik, Vanguard, Bioviridian, Ethos Analytics and Liquilabs. Across the platform as a whole, Peptigrity tracks 530 shops, 11,852 independent lab tests, 118 peptides and 1,283 community reviews (verified August 2026), with trust scores weighted 50% community reviews and 50% independent lab purity. Seven tests is a thin dataset by that standard, and one of the seven missing its stated content by 40% is a large proportion of it — the mechanics are in our guide to why 10 mg isn't 10 mg, and the single endotoxin figure in our guide to endotoxin testing and the LAL assay.

But the deeper problem is what those tests measure. HPLC purity establishes homogeneity against a declared analyte. For a glycoprotein, "purity" is not a single number in the way it is for a synthetic peptide — glycosylation produces a family of glycoforms with different masses, and correct disulfide folding is invisible to both HPLC and mass spectrometry. A follistatin purity percentage is answering a question the molecule does not really pose.

Follistatin price data shows 1 shop in stock at $79.99/mg on a 1 mg vial (verified 11 August 2026), which is simultaneously the low, the high and the notional median. Those figures exclude shipping, taxes and customs, coupon codes, bulk tiers, multi-vial kits and account-gated pricing. A median computed from one listing is not a median, and we are flagging that on our own page. Compare vendors on the follistatin compound page and against independent lab tests.

How does follistatin compare with the antibody programmes?

Follistatin has two registered human trials (n=15 and n=43), no posted results, and one publication that was formally challenged. Bimagrumab, an anti-activin receptor IIB antibody, has phase 2/3 data in 251 and 217 patients with confirmed target engagement — and every function endpoint was null. Domagrozumab reached phase 2 in 121 Duchenne patients and was terminated for lack of efficacy. Stamulumab enrolled 108 and posted nothing.

The pharmaceutical programmes are the most useful evidence here precisely because they were not selling anything to consumers. Four companies pursued myostatin pathway inhibition in real indications with proper endpoints, and the result was consistent: the tissue responds, the person does not.

If antibody-based inhibition with confirmed target engagement in 251 patients cannot move a six-minute walk distance, the prior on an unverified glycoprotein injected subcutaneously is low.

What trial would settle the follistatin question?

No trial can settle follistatin until the product is characterised, because nobody can currently say which species is in a vial — 344-residue precursor, FS-315 or FS-288, glycosylated or not. The trial that would settle this injects protein rather than delivering a gene, measures whether it reaches muscle, and takes function as its primary endpoint: strength, gait speed, walking distance. Mass is already known to respond.

The fourth row is where this category has repeatedly failed, and it is the only row that matters to anyone buying it.

Frequently Asked Questions

What is follistatin-344?

The full-length 344-residue human follistatin precursor — a glycoprotein of about 38,000 daltons. It is the payload used in gene therapy trials, delivered as a viral vector or plasmid so the body produces the protein itself. Residues 1–29 are a signal peptide cleaved during secretion, so the circulating protein is FS-315 or FS-288, not a 344-residue chain.

Can follistatin be made by peptide synthesis?

Solid-phase peptide synthesis works routinely to around 50 residues and cannot glycosylate anything. A 344-residue glycoprotein would require recombinant expression in cells, folding and purification — biologics manufacturing rather than peptide synthesis. What is actually in a research vial is not something a purity certificate can tell you.

Has injected follistatin been tested in humans?

We found no registered trial of follistatin protein administration. The two human follistatin trials we identified — a 15-patient AAV gene transfer study and a 43-participant plasmid study in Honduras — both deliver genes, and neither has posted results.

Does myostatin inhibition build usable muscle?

The pharmaceutical evidence says no. Bimagrumab in 251 patients showed lean body mass rising while every six-minute-walk comparison was non-significant. In a separate 217-patient sarcopenia trial, lean body mass rose at p<0.001 while SPPB, gait speed and six-minute walk were all non-significant. Domagrozumab was terminated for lack of efficacy.

Is follistatin banned in sport?

Yes, by name. WADA section S4.3, "Agents Preventing Activin Receptor IIB Activation," lists "myostatin-binding proteins (e.g. follistatin, myostatin propeptide)" as prohibited at all times. Note that it is in S4, not S2 where most peptides sit.

What about the gene therapy study that showed improved walking?

The 2015 report of improved ambulation in Becker muscular dystrophy was formally challenged in Molecular Therapy in 2017 in a paper titled "Unfounded Claims of Improved Functional Outcomes Attributed to Follistatin Gene Therapy in Inclusion Body Myositis." Both should be read together.

Follistatin fails on two independent grounds, and either one would be sufficient.

The first is what is in the vial. "Follistatin-344" names a 344-residue glycoprotein that gene therapy trials deliver as DNA because that is how you get a cell to make it. The first 29 residues are a signal peptide the body removes on secretion, so the number in the product name describes a species that does not circulate. Whether a research vial contains recombinant glycosylated protein, an unglycosylated bacterial expression product, a fragment, or something else is not determinable from any test performed in this market.

The second is whether it would matter if it were real. Four pharmaceutical companies pursued myostatin pathway inhibition with antibodies and decoy receptors, in trials of 251 and 217 patients with posted results. Lean body mass rose, significantly and repeatedly. Six-minute walk distance, gait speed and physical performance did not move once. One programme was terminated explicitly for lack of efficacy. Muscle you cannot use is the most consistent finding in this field, and it was established by people with every incentive to find otherwise.

Browse the growth hormone peptides category, or our complete peptide guide with 118 compounds (verified August 2026). Compare shops through independent lab tests and community-verified shop reviews.

This article is for educational and informational purposes only and does not constitute medical advice. Peptides discussed may be investigational compounds not approved by the FDA (or equivalent regulators in your jurisdiction) for human use. Always consult a qualified healthcare provider before using any peptide or research compound. Peptigrity is an independent review platform and does not sell, endorse, or recommend specific products or vendors.