Follistatin and Follistatin 344: Evidence, Isoforms and the Injection Problem
Follistatin is the most biologically interesting entry on the muscle-growth peptide lists and the one where the gap between the cited evidence and the sold product is widest. The trial data people quote comes from gene therapy delivered by viral vector into muscle. The product people buy is a recombinant protein in a vial. This guide separates the two, walks through what the human studies actually reported including the published objection to them, and explains why the most informative result in the field is a trial that succeeded on its primary endpoint and was cancelled anyway.
What Follistatin Is
Follistatin is a glycoprotein that works by binding things and taking them out of circulation. Its principal targets are ligands in the TGF-beta superfamily: activin A above all, plus myostatin (GDF-8) and GDF-11. By binding these ligands directly, follistatin sterically blocks them from engaging their cell-surface receptors, so the signal never arrives.
Myostatin is a negative regulator of skeletal muscle mass. Remove it and muscle grows, which is why myostatin-null cattle and the rare human cases exist as textbook photographs. Follistatin removes myostatin along with activin and GDF-11 at once, which makes it a broader intervention than a myostatin-specific antibody and, in rodent work, a more potent one.
That breadth cuts both ways. Activin signalling is not confined to muscle; it participates in reproduction, inflammatory regulation and tissue repair. A local injection into one muscle keeps the effect where it is wanted. Systemic exposure does not, and nobody has published what happens when a healthy adult suppresses activin signalling body-wide for weeks. For related mechanisms, see our peptides for muscle growth overview.
FST344, FS-315 and FS-288: What the Numbers Mean
Three numbers circulate and they refer to different things, which is a large part of why the marketing is so confused.
| Name | What it refers to | Behaviour |
|---|---|---|
| FST344 / FS-344 | A transcript of the follistatin gene, encoding a precursor | Processed to the circulating mature form; the construct used in gene therapy |
| FS-315 | Mature circulating isoform, 315 amino acids | Weak cell-surface binding, distributes systemically |
| FS-288 | Shorter mature isoform, 288 amino acids | Binds heparan sulfate on cell membranes; acts locally |
A vial labelled follistatin 344 is borrowing the name of a gene transcript to describe a protein. It is not wrong exactly, since FST344 is where the circulating form comes from, but the number is doing rhetorical work: it points at the gene therapy trials without claiming to be one.
Note also the size. This is a glycoprotein of roughly 300 amino acids, an order of magnitude larger than the short peptides most vendors handle. Correct folding and glycosylation are not optional for activity, and neither is something a buyer can verify. The storage and handling rules that apply to a 15-amino-acid peptide are not sufficient for a protein this size.
The Gene Therapy Trials
The human evidence people cite comes from a Phase 1/2a program at Nationwide Children’s Hospital using AAV1-FS344, an adeno-associated viral vector carrying the follistatin transcript under a CMV promoter, injected directly into the quadriceps. Two populations were studied: Becker muscular dystrophy and sporadic inclusion body myositis.
In the Becker cohort, two participants improved on the six-minute walk test, by 58 metres and 125 metres. In the inclusion body myositis cohort, four of six treated subjects increased six-minute walk distance by 58 to 153 metres while two changed minimally, by 5 to 23 metres; muscle biopsies were reported to show decreased fibrosis and improved regeneration. Safety was the strongest finding: the vector was well tolerated at the doses given.
Then the important part, which is missing from every vendor page that cites this work. A commentary in Molecular Therapy specifically disputed the functional-outcome claims in the inclusion body myositis cohort as unfounded. The objections are the ones you would expect: an open-label design with no randomised control group, six subjects, and a six-minute walk test whose variability in slowly progressive neuromuscular disease can easily produce swings of the size reported. The distances were measured. Whether they represent drug effect rather than practice effect, motivation or noise is genuinely contested in the literature, and citing the trial without citing the rebuttal is selective.
ACE-083: The Result That Should Change Your Expectations
Acceleron developed ACE-083, a locally acting follistatin-based fusion protein designed to be injected into a specific muscle and stay there. It went into Phase 2 in facioscapulohumeral muscular dystrophy and Charcot-Marie-Tooth disease, with Fast Track designation in FSHD.
It worked, on the endpoint it was measuring. ACE-083 produced a robust, statistically significant increase in mean total muscle volume, its primary endpoint. It then failed to convert that into statistically significant improvement on functional tests, and development was discontinued in both FSHD and Charcot-Marie-Tooth disease.
This is the single most useful data point for anyone considering follistatin for physique or performance. A properly resourced pharmaceutical program, using a purpose-engineered follistatin construct delivered directly into the target muscle at controlled doses, reliably made muscles bigger and could not show that the bigger muscles did more. Larger cross-sectional area produced by removing a growth brake is not automatically functional tissue. That result comes from a diseased-muscle population and does not transfer directly to healthy trained adults, but it should temper the assumption that volume gain equals capability gain.
Why Injecting the Protein Is a Different Intervention
Every piece of positive human data on follistatin involves sustained local exposure. The gene therapy converts muscle cells into a factory producing the protein continuously for months. ACE-083 was a fusion protein engineered for local retention and injected into the target muscle. Neither resembles a subcutaneous shot of recombinant protein into abdominal fat.
Circulating follistatin clears quickly. Figures on the order of an hour to ninety minutes are commonly cited for the protein in plasma, and the entire reason fusion constructs and Fc-fusions exist in this space is to extend a half-life that is otherwise too short to be useful as a drug. A daily subcutaneous injection produces a brief systemic pulse; the trials produced months of continuous local concentration. Expecting the second result from the first is the same error described in our kisspeptin dosage guide, where trial results from continuous infusion get quoted alongside twice-weekly injection protocols.
There is also no published human pharmacokinetic data for subcutaneous recombinant follistatin, no dose-ranging study, and no immunogenicity data. Repeated injection of a large recombinant human protein carries a real risk of anti-drug antibodies, and in this case those antibodies would be directed at an endogenous regulatory protein. That is a theoretical concern rather than a documented one, which is precisely the problem: nobody has looked.
On Dosage
This guide does not provide a follistatin dosing protocol, and the reason is not caution for its own sake. There is nothing to base one on. No approved product, no published human dose-ranging study of the injected protein, no subcutaneous pharmacokinetics, no exposure-response relationship.
The figures that circulate, typically around 100 micrograms per day for one to two weeks, trace back to vendor product pages and forum convention. They are not derived from any trial, and the gene therapy work cannot supply them because a viral vector dose in vector genomes does not convert into micrograms of injected protein. If you see a dosing chart for follistatin presented with the same confidence as one for a studied compound like tesamorelin, that confidence is manufactured.
Legal and Anti-Doping Status
No injectable follistatin product is approved by the FDA for any indication. The gene therapy form exists only inside registered clinical trials at specific research institutions. Material sold as follistatin 344 is research-use-only product outside the approved supply chain, with no verified identity or purity.
For competitive athletes the position is simple: agents inhibiting myostatin function are prohibited at all times under the World Anti-Doping Code, and gene doping is prohibited separately. Follistatin is not a grey area in tested sport. See our FDA-approved peptides list for what has actually cleared regulatory review.
The Bottom Line
Follistatin is real biology with a genuine mechanism and a small amount of contested human data, all of it generated by delivery methods that are not available to buy. The gene therapy results are interesting and disputed in print. The one program that ran a controlled functional endpoint increased muscle volume and could not increase function.
What is sold as follistatin 344 is a large recombinant protein with no human dosing data, a short circulating half-life, unverifiable folding, and a plausible immunogenicity question nobody has studied. The name on the vial points at trials whose design it does not share. For muscle-focused options with actual human dose data behind them, compare IGF-1 LR3 dosing and MK-677, neither of which is uncomplicated but both of which have something to reason from.
Frequently Asked Questions About Follistatin
FST344 is one of the transcripts of the human follistatin gene. It encodes a precursor that, after the signal peptide is removed, yields the 315-amino-acid circulating form of the protein. The name matters because it is the construct used in the AAV1-FS344 gene therapy trials, and vendors selling injectable vials borrow that number to imply a connection to those trials. What the vials contain is a recombinant protein, not a gene therapy vector, and the two produce completely different exposure.
They are the two main isoforms of the mature protein, numbered by amino acid count. FS-315 is the predominant circulating form; it binds cell surfaces weakly, so it distributes systemically. FS-288 binds tightly to heparan sulfate proteoglycans on cell membranes and therefore acts locally where it is produced. That difference is not trivia. It is the reason local production inside muscle and a systemic injection are not interchangeable, and the reason a drug designer choosing between them is choosing between two different pharmacologies.
Yes, though describing it as a myostatin inhibitor undersells the promiscuity. Follistatin binds and neutralises several TGF-beta superfamily ligands, primarily activin A, along with myostatin (GDF-8) and GDF-11, by sterically preventing them from engaging their receptors. Activin binding is arguably its primary role. That breadth is what makes it more potent than a myostatin-specific antibody in animal models, and also what makes systemic exposure harder to reason about, since activin signalling is involved in reproduction, inflammation and wound repair, not just muscle.
A Phase 1/2a program at Nationwide Children’s Hospital delivered AAV1-FS344 by direct intramuscular injection into the quadriceps in Becker muscular dystrophy and sporadic inclusion body myositis. In the Becker cohort, two participants improved on the six-minute walk test by 58 and 125 metres. In the inclusion body myositis cohort, four of six subjects gained 58 to 153 metres on the same test while two were minimally changed at 5 to 23 metres, with reported reductions in fibrosis and improved regeneration on biopsy. The vector was described as exceptionally safe at the doses used.
Yes, and anyone citing the trial should know it. A commentary published in Molecular Therapy under the title describing the functional-outcome claims in inclusion body myositis as unfounded challenged the interpretation of those six-minute walk results. The core objections concern open-label design without a randomised control, very small numbers, and the natural variability of the six-minute walk test in a slowly progressive disease. The measurements happened; whether they demonstrate a treatment effect is what is contested.
ACE-083 was a locally acting follistatin-based fusion protein injected directly into target muscles, developed by Acceleron and taken into Phase 2 in facioscapulohumeral muscular dystrophy and Charcot-Marie-Tooth disease. It hit its primary endpoint: a robust, statistically significant increase in mean total muscle volume. Those volume gains did not translate into statistically significant improvement on functional tests, and development was discontinued in both indications. That is the most instructive result in the entire follistatin literature, because it separates the outcome people want from the outcome the drug delivers.
Short, and this is the crux of the problem with the injectable market. Circulating follistatin is cleared rapidly, with figures on the order of an hour to an hour and a half commonly cited for the protein; engineered fusion constructs are built specifically to extend that, which is why they exist at all. Set that against gene therapy, where a viral vector turns muscle cells into a local factory producing the protein continuously for months. Sustained local production and an injection that clears within hours are not the same intervention, whatever the label says.
No, and there is no honest way to construct one. No injectable follistatin product is approved for any indication, no human dose-ranging study of the injected protein has been published, and no pharmacokinetic data exists for the subcutaneous route in humans. Circulating protocols quoting figures in the range of 100 micrograms daily for a week or two are derived from vendor copy and forum convention, not from published exposure or response data. Anyone quoting you a dose with confidence is quoting a number with nothing behind it.
Yes. Agents that inhibit myostatin function are prohibited at all times under the World Anti-Doping Code, and follistatin falls squarely in that description regardless of whether it is delivered as protein or gene therapy. Gene doping provisions cover the vector-based route separately. Tested athletes should treat this as an unambiguous prohibition, not a grey area.
Different mechanisms and very different evidence positions. IGF-1 LR3 is a modified growth factor with a long half-life and at least a coherent pharmacological rationale for injection, discussed in our IGF-1 LR3 guide. MK-677 is an orally active ghrelin receptor agonist with substantial human data on growth hormone and IGF-1 elevation. Follistatin acts by removing a brake on muscle growth rather than pressing an accelerator, which is mechanistically attractive, but the only human evidence sits in gene therapy and locally injected fusion proteins, and the one program that reached a properly powered functional endpoint missed it.