--- canonical_name: Follistatin alternate_names: FST, Follistatin-344, FS-344, FS-315, FS-288, Activin-Binding Protein canonical_topic: Follistatin for Muscle Growth short_topic_lc: follistatin_muscle creation_date: 2026-0912-1550 creator_ai_fullname: Opus 5 ep_keywords: Myostatin Inhibitors, Activin Inhibitors --- # Follistatin for Muscle Growth
Evidence Review created on 09/12/2026 using [AI4L](https://github.com/forever-healthy/AI4L) / Opus 5 **Also known as:** FST, Follistatin-344, FS-344, FS-315, FS-288, Activin-Binding Protein ## Motivation Follistatin is a protein the body makes on its own. It acts as a sponge for myostatin, the signal that skeletal muscle secretes to limit its own growth, so raising follistatin lifts a brake rather than pressing an accelerator. That simple idea has made it one of the most talked-about targets for building and keeping muscle. Animals lacking a working myostatin brake are strikingly heavy-muscled, and the same build appears in certain breeds of cattle and dogs and in a small number of people. That observation moved follistatin out of the laboratory. It has since been placed into muscle by gene transfer in people with wasting diseases, engineered into an injectable protein, offered by private clinics operating outside conventional regulatory oversight, and sold as unlicensed vials by online vendors. This review sets out what is actually known: how follistatin works, which molecules have been tested in people and what those studies measured, what the safety signals look like, how the protocols used in trials differ from those used in clinics and by individuals, and which questions remain open. **[Benefits](#expected-benefits) - [Risks](#potential-risks--side-effects) - [Protocol](#therapeutic-protocol) - [Conclusion](#conclusion)** ## Recommended Reading High-level overviews of follistatin and myostatin-pathway modulation from expert commentators and the primary literature. * [#370 – AMA #76: Peter evaluates longevity drugs, aspirin for CVD, and strategies to improve muscle mass — proven, promising, fuzzy, noise, or nonsense?](https://peterattiamd.com/ama76/) - Peter Attia An ask-me-anything (AMA) episode spanning aspirin for cardiovascular disease (CVD) and muscle; it rates follistatin gene therapy on an evidence-strength scale and explains why injectable follistatin is impractical. * [The Prospect of Human Age Reversal](https://www.lifeextension.com/magazine/2025/1/prospect-human-age-reversal) - William Faloon Its follistatin section traces the animal findings behind the offshore gene-therapy market, the reported human plasmid results, and the cost and dosing interval quoted for that route. * [Follistatin Gene Therapy Doubles Muscle Mass in Mice](https://www.fightaging.org/archives/2020/05/follistatin-gene-therapy-doubles-muscle-mass-in-mice/) - Reason A sceptical longevity-field commentary on why follistatin gene therapy is already sold through medical tourism despite thin efficacy data, and what delivery problems still block systemic use. * [Follistatin Gene Therapy Improves Ambulation in Becker Muscular Dystrophy](https://pubmed.ncbi.nlm.nih.gov/27858738/) - Al-Zaidy et al., 2015 A narrative review by the investigators: why the FS-344 variant (a follistatin isoform) was chosen over FS-288 to avoid pituitary effects, and what the gene-transfer trial in Becker muscular dystrophy (an inherited muscle-weakening disease) showed. * [Detection of black market follistatin 344](https://pubmed.ncbi.nlm.nih.gov/31758732/) - Reichel et al., 2019 Doping-control laboratory analysis of seventeen grey-market follistatin products, documenting what was actually inside the vials and how the material differs from human follistatin. Two of the six priority platforms carried directly relevant material. No substantive treatment of follistatin or myostatin inhibition was found on chriskresser.com (Chris Kresser) or hubermanlab.com (Andrew Huberman), whose site search returns only broad episodes in which the term appears incidentally, and foundmyfitness.com (Rhonda Patrick) likewise returns only passing mentions. Lifespan.io's matches are news and interview items that mention follistatin gene therapy incidentally, plus a Rejuvenation Roadmap entry, which is a registry record rather than a discussion; none treats the topic in depth, so all are excluded. ## Grokipedia * [Follistatin](https://grokipedia.com/page/Follistatin) A dedicated encyclopaedia entry covering follistatin's structure, its isoforms, its antagonism of transforming growth factor-β family ligands (proteins that govern cell growth and repair), and the gene-transfer trials, with source citations throughout. ## Examine * [Follistatin](https://examine.com/outcomes/follistatin/) Summarises follistatin as an understudied regulator, grades the single 40-participant supplement trial in its database, and flags the reproductive and activin-related effects that sit beyond muscle. ## ConsumerLab No ConsumerLab article on follistatin exists. ConsumerLab tests retail supplements; follistatin is supplied only as an unapproved injectable biologic or gene-transfer product and is not sold as a dietary supplement, so it falls outside the organisation's testing scope. ## Systematic Reviews Pooled analyses covering follistatin as a measurable biological signal and myostatin-pathway blockade as a therapy. * [The effects of resistance training on myostatin and follistatin in adults: A systematic review and meta-analysis](https://pubmed.ncbi.nlm.nih.gov/37328021/) - Khalafi et al., 2023 Pools 36 interventions in 768 adults: resistance training raises circulating follistatin and lowers myostatin at every age studied. * [Circulating biomarkers in older adults with and without sarcopenia: a systematic review and meta-analysis](https://pubmed.ncbi.nlm.nih.gov/42202210/) - Prokopidis et al., 2026 Across 26 studies, follistatin did not differ between older adults with and without sarcopenia (age-related muscle loss), weakening the biomarker rationale for raising it. * [Circulating resistin and follistatin levels in obese and non-obese women with polycystic ovary syndrome: A systematic review and meta-analysis](https://pubmed.ncbi.nlm.nih.gov/33740002/) - Raeisi et al., 2021 Nine studies show follistatin is elevated in polycystic ovary syndrome independent of obesity, tying higher follistatin to reproductive-metabolic dysfunction. * [Myostatin Modulation in Spinal Muscular Atrophy: A Systematic Review of Preclinical and Clinical Evidence](https://pubmed.ncbi.nlm.nih.gov/40565321/) - Gnazzo et al., 2025 Reviews myostatin-pathway blockade as a class, not follistatin itself, in spinal muscular atrophy (an inherited motor-neuron disease): safety is favourable but efficacy unproven. Coverage of the trade-off is uneven. The claimed effect and the metabolic-endocrine risk each have a pooled analysis, but no systematic review or meta-analysis evaluates the safety or efficacy of administered follistatin — whether as gene transfer, engineered protein or injected peptide — because too few controlled trials exist to pool. ## Mechanism of Action Follistatin is a secreted glycoprotein that works as a molecular sponge. It binds and neutralises several members of the transforming growth factor-β family (TGF-β, a group of signalling proteins that govern cell growth and repair), above all myostatin — also called growth differentiation factor 8 (GDF-8) — and activin A. Myostatin is muscle's own brake: muscle cells secrete it, it docks on the activin receptor type IIB (ActRIIB, the receptor shared by myostatin and activin), and the resulting signal suppresses protein synthesis and satellite-cell activity (satellite cells are muscle's own stem cells). Sequestering myostatin releases that brake, enlarging existing fibres and, in animals, adding new ones. Two mechanistic readings compete. The dominant one holds that follistatin beats a pure myostatin antibody because it also blocks activin A, adding anti-inflammatory and anti-fibrotic effects. The opposing reading treats that breadth as the defect: follistatin also binds bone morphogenetic proteins (BMPs, signals that build bone and blood vessels) and interacts with the inhibin axis (the reproductive hormone feedback loop), so wide exposure invites endocrine and vascular consequences a selective blocker avoids. Pharmacologically, native follistatin-315 is a poor drug. Heparan-sulfate binding holds it at cell surfaces near the injection site, the liver clears it within minutes, and being a protein it is destroyed in the gut and broken down by lysosomal proteolysis rather than by cytochrome P450 enzymes (the liver's main drug-metabolising family). Adding an antibody Fc tail and removing heparan-sulfate binding lengthened terminal half-life roughly 100-fold in mice ([Datta-Mannan et al., 2013](https://pubmed.ncbi.nlm.nih.gov/23249626/)). ## Historical Context & Evolution Follistatin was identified in the 1980s in ovarian follicular fluid as a suppressor of follicle-stimulating hormone (FSH, the pituitary signal that drives egg and sperm production). For roughly a decade it was studied purely as a reproductive-endocrine protein, and its name still records that origin. The muscle story began elsewhere. In 1997 myostatin was described as a negative regulator of muscle mass, and the heavily muscled Belgian Blue and Piedmontese cattle were shown to carry loss-of-function mutations in it. Because follistatin was already known to bind activins, it was tested against myostatin and found to neutralise it. Transgenic mice overexpressing follistatin in muscle grew dramatically; crossing them onto a myostatin-null background quadrupled muscle mass, showing follistatin acts on more than myostatin alone ([Lee, 2007](https://pubmed.ncbi.nlm.nih.gov/17726519/)). Translation followed a deliberate path. Gene delivery of the FS-344 splice variant enlarged and strengthened quadriceps in macaque monkeys without organ toxicity ([Kota et al., 2009](https://pubmed.ncbi.nlm.nih.gov/20368179/)), and the same construct entered human trials in Becker muscular dystrophy and sporadic inclusion body myositis (a progressive inflammatory muscle-wasting disease of later life). In parallel, Acceleron Pharma engineered a locally acting follistatin-domain fusion protein, and went on to design, fund and co-author every human trial of it. Opinion has moved in both directions since. The engineered-protein programme was discontinued when muscle volume rose without matching function, while gene transfer moved outside conventional regulation into private clinics. Neither outcome settles the question: the muscle-growth effect is reproducible, and what that muscle does is not. ## Expected Benefits ### High 🟩 🟩 🟩 #### Increased Skeletal Muscle Volume at the Injected Site Injecting a follistatin-domain fusion protein (ACE-083) into one muscle reliably enlarges that muscle. Three placebo-controlled trials — in healthy postmenopausal women, in facioscapulohumeral muscular dystrophy (a dystrophy of face, shoulder and upper-arm muscles) and in Charcot-Marie-Tooth disease (an inherited nerve disorder causing weakness furthest from the trunk) — measured the gain by magnetic resonance imaging; in the two disease trials contractile (non-fat) volume rose too. All three were funded and co-authored by the manufacturer, Acceleron Pharma. The effect is confined to injected muscles; no trial tested whole-body exposure. **Magnitude:** +14.5% ± 4.5% in rectus femoris and +8.9% ± 4.7% in tibialis anterior versus baseline in healthy volunteers ([Glasser et al., 2018](https://pubmed.ncbi.nlm.nih.gov/29486514/)); +16.4% in biceps brachii and +9.5% in tibialis anterior versus placebo ([Statland et al., 2022](https://pubmed.ncbi.nlm.nih.gov/35428982/)); +13.5% in tibialis anterior versus placebo ([Thomas et al., 2022](https://pubmed.ncbi.nlm.nih.gov/35545446/)). Values after ± are standard deviations (SD), a measure of spread. ### Medium 🟩 🟩 No benefit reaches Medium: the only human evidence for follistatin itself, as opposed to an engineered follistatin-domain protein, comes from open-label six-patient gene-transfer series with no randomised or concurrent placebo comparison, which is the uncontrolled evidence class that caps at Low. ### Low 🟩 #### Muscle Fibre Hypertrophy and Reduced Fibrosis after Gene Transfer Quadriceps biopsies after intramuscular AAV1-FS344 (an adeno-associated virus, a harmless gene-carrying virus) showed larger fibres, more normal fibre-size distribution, less endomysial scarring and improved regeneration. Open-label, six patients per trial, no placebo arm. **Magnitude:** Direction only — hypertrophy and reduced fibrosis were clearest at the higher dose in both trials ([Mendell et al., 2015](https://pubmed.ncbi.nlm.nih.gov/25322757/); [Mendell et al., 2017](https://pubmed.ncbi.nlm.nih.gov/28279643/)); neither report gives a quantified effect size for the histological change. #### Improved Walking Distance in Muscle-Wasting Disease Six-minute walk distance improved in most, not all, recipients of follistatin gene transfer. Both trials were open-label with six participants; the myositis trial used age-, sex- and baseline-matched untreated comparators rather than randomisation. **Magnitude:** +58 m and +125 m in two of three low-dose and +108 m and +29 m in two of three high-dose Becker muscular dystrophy patients, with two non-responders ([Mendell et al., 2015](https://pubmed.ncbi.nlm.nih.gov/25322757/)); +56.0 m/year versus −25.8 m/year in matched untreated myositis patients ([Mendell et al., 2017](https://pubmed.ncbi.nlm.nih.gov/28279643/)). #### Muscle Strength ⚠️ Conflicted Three placebo-controlled trials of the follistatin-domain protein disagree: no strength change in healthy volunteers, no consistent gain in facioscapulohumeral dystrophy, but a significant ankle dorsiflexion gain in Charcot-Marie-Tooth disease. Net reading: added muscle volume does not dependably become added strength. **Magnitude:** Direction only — strength rose measurably in one of three trials, in ankle dorsiflexion after tibialis anterior injection ([Thomas et al., 2022](https://pubmed.ncbi.nlm.nih.gov/35545446/)); the other two report null strength results without effect figures ([Glasser et al., 2018](https://pubmed.ncbi.nlm.nih.gov/29486514/); [Statland et al., 2022](https://pubmed.ncbi.nlm.nih.gov/35428982/)). #### Reduced Fat Inside the Treated Muscle ⚠️ Conflicted Fat fraction fell in the tibialis anterior group of one placebo-controlled trial, but not in that trial's biceps group and not in a second trial injecting the same muscle. This is intramuscular fat in one muscle, not body-wide fat loss. Net reading: the effect is not reproducible. **Magnitude:** Direction only — fat fraction decreased in the tibialis anterior group but not the biceps group of the facioscapulohumeral trial ([Statland et al., 2022](https://pubmed.ncbi.nlm.nih.gov/35428982/)), and was not significantly improved in the Charcot-Marie-Tooth trial injecting the same muscle ([Thomas et al., 2022](https://pubmed.ncbi.nlm.nih.gov/35545446/)); neither report gives a stand-alone effect figure for fat fraction. ### Speculative 🟨 #### Whole-Body Lean Mass Gain with Fat Loss Systemic follistatin-288 raised muscle mass and cut fat accumulation in mice ([Gangopadhyay, 2013](https://pubmed.ncbi.nlm.nih.gov/23942549/)). The only human report is an uncontrolled, non-peer-reviewed plasmid series measuring body composition without a comparator, so the basis stays animal work. #### Reduced Biological Age ⭕️ Not Central to Muscle Growth The plasmid programme reports epigenetic-clock age falling several years after one dose. It bears on longevity, not muscle growth. The basis is an uncontrolled company series, released by the seller, using an unvalidated biomarker. #### Reduced Joint Damage and Metabolic Inflammation ⭕️ Not Central to Muscle Growth [Follistatin gene transfer in mice](https://pubmed.ncbi.nlm.nih.gov/32426485/) lowered obesity-driven inflammatory signals and protected knees from injury-induced damage. It bears on joint and metabolic health, not muscle growth. The basis is animal work only. ## Benefit-Modifying Factors * **Pre-existing vector immunity:** Prior exposure to adeno-associated virus leaves neutralising antibodies that destroy the vector before it reaches muscle. Trials screening for this require a titre of 1:100 or lower, and roughly a third to a half of adults screen out. * **Glucokinase regulatory protein (GCKR) variants:** Variation in GCKR, the gene controlling how the liver handles glucose, was the strongest genetic determinant of circulating follistatin in a [genome-wide analysis](https://pubmed.ncbi.nlm.nih.gov/34759311/), so baseline levels and the increment achieved differ between genotypes. * **Baseline myostatin and activin A:** Follistatin works by removing an existing brake. Where baseline myostatin is already low — in trained individuals or myostatin-mutation carriers — less brake remains to remove, and the achievable gain is correspondingly smaller. * **Sex:** Every trial of the follistatin-domain protein in healthy people enrolled postmenopausal women only, so no direct comparison exists. In premenopausal women the activin-follicle-stimulating hormone axis is active, which constrains dosing more than it does in men. * **Pre-existing metabolic disease:** Obesity, fatty liver and insulin resistance already raise endogenous hepatic follistatin. Starting from an elevated level narrows the margin before the concentrations associated with adverse cardiometabolic outcomes are reached. * **Age:** Older muscle has fewer and less responsive satellite cells. Above roughly 70 the volume gain is expected to be smaller and to translate into function less readily, though no trial has tested this directly. * **Concurrent resistance training:** Loading is what converts added tissue into usable strength. Resistance training independently [raises follistatin and lowers myostatin](https://pubmed.ncbi.nlm.nih.gov/37328021/), so the untrained state is where the ceiling on benefit is lowest. ## Potential Risks & Side Effects ### High 🟥 🟥 🟥 #### Injection-Site Reactions Pain, swelling, redness and induration (hardening of the tissue) at the injection site were the most common treatment-emergent adverse event in all three placebo-controlled trials of the locally acting follistatin-domain protein, and were graded mild to moderate. They are the predictable consequence of depositing a protein depot intramuscularly, they resolve without intervention, and no participant discontinued because of them. Repeated three-weekly dosing into the same muscle is what drives cumulative exposure. **Magnitude:** Injection-site pain affected 38 of 42 treated healthy volunteers (90%) and 16 of 16 on placebo (100%), with other injection-site events in roughly 10–20% of treated participants ([Glasser et al., 2018](https://pubmed.ncbi.nlm.nih.gov/29486514/)); the same reactions were the most frequent adverse event in the facioscapulohumeral and Charcot-Marie-Tooth trials ([Statland et al., 2022](https://pubmed.ncbi.nlm.nih.gov/35428982/); [Thomas et al., 2022](https://pubmed.ncbi.nlm.nih.gov/35545446/)). ### Medium 🟥 🟥 #### Higher Circulating Follistatin Tracks with Cardiometabolic Disease In two European cohorts followed for up to 22 years, people with higher plasma follistatin developed more type 2 diabetes, stroke, heart failure and chronic kidney disease, and died sooner. The mechanism proposed is that liver-derived follistatin drives fat-tissue insulin resistance and free fatty acid release. The data are observational and describe endogenous hepatic follistatin, not injected or gene-expressed follistatin, so causal transfer to the intervention is not established — but no study has shown the two are biologically distinct either. **Magnitude:** Hazard ratio per standard deviation increase 1.24 (confidence interval 1.04–1.47) over 19 years and 1.31 (1.09–1.58) over 4 years for type 2 diabetes ([Wu et al., 2021](https://pubmed.ncbi.nlm.nih.gov/34759311/)); 1.05 for death, 1.13 for ischaemic stroke, 1.16 for heart failure and 1.38 for chronic kidney disease ([Pan et al., 2024](https://pubmed.ncbi.nlm.nih.gov/38000997/)). Hazard ratio is the relative rate of an event between groups; confidence interval is the range the true value most likely occupies. ### Low 🟥 #### Vector Immune Response and Liver Injury Adeno-associated virus vectors can provoke fever, flu-like illness and liver enzyme rises, and high systemic doses have caused deaths in other neuromuscular gene-therapy programmes. The follistatin trials reported no adverse effects, but enrolled only 15 patients in total at low intramuscular doses. **Magnitude:** Not quantified in available studies. No controlled trial of follistatin gene transfer has been large enough to estimate an event rate, and the class risk is drawn from other vectors and dose ranges ([Suoranta et al., 2022](https://pubmed.ncbi.nlm.nih.gov/39086961/)); the follistatin series report zero events in six patients each ([Mendell et al., 2015](https://pubmed.ncbi.nlm.nih.gov/25322757/)). #### Misidentified and Adulterated Grey-Market Product A doping-control laboratory analysed seventeen products sold as "follistatin 344". Nearly half contained no follistatin; some contained other growth-promoting peptides instead. Every positive sample carried a bacterial purification tag and protein clumping absent from human follistatin. **Magnitude:** 9 of 17 products (53%) contained follistatin; the remainder contained mechano growth factor, growth hormone releasing peptide 2 or nothing identifiable, and all nine positives contained histidine-tagged FS-344 plus oligomers ([Reichel et al., 2019](https://pubmed.ncbi.nlm.nih.gov/31758732/)). #### Vascular and Mucosal Effects of Broad Pathway Blockade A [trial](https://pubmed.ncbi.nlm.nih.gov/27462804/) of a systemic activin-receptor decoy in boys with Duchenne dystrophy (an inherited muscle-wasting disease) was halted for nosebleeds and dilated surface vessels, attributed to bone morphogenetic protein blockade. Follistatin binds those proteins too, so the signal carries across indirectly; no follistatin trial has reported these events. **Magnitude:** Direction only — the events appeared under repeated systemic dosing of a broad pathway blocker and were judged serious enough to stop that trial early ([Campbell et al., 2017](https://pubmed.ncbi.nlm.nih.gov/27462804/)); the report names the safety concern without giving an incidence rate. ### Speculative 🟨 #### Suppression of Follicle-Stimulating Hormone Follistatin was first identified as a suppressor of follicle-stimulating hormone, and mice overexpressing it develop [reproductive defects](https://pubmed.ncbi.nlm.nih.gov/9440814/). The FS-344 variant was chosen to limit this. No human endocrine outcome data exist. #### Tumour-Promoting Activin Blockade Activin A restrains cell proliferation, and a [review of solid tumours](https://pubmed.ncbi.nlm.nih.gov/27807065/) finds follistatin overexpressed and tracking with worse prognosis. Whether giving follistatin promotes tumours is untested in people. #### Tendon and Connective-Tissue Weakening [Tendons of myostatin-deficient mice](https://pubmed.ncbi.nlm.nih.gov/18162552/) are small, brittle and cell-poor, raising the concern that muscle outgrows its attachments. No human tendon outcome has been measured after any follistatin intervention. ## Risk-Modifying Factors * **Adeno-associated virus serostatus:** Pre-existing neutralising antibodies raise the chance of an immune reaction while lowering the chance of benefit. Protocols require a titre no higher than 1:100 and prior exposure to any vector serotype is disqualifying. * **Baseline glucose handling:** Higher follistatin is [linked to insulin resistance](https://pubmed.ncbi.nlm.nih.gov/34759311/), so raising it from an already-impaired baseline compounds the signal. Active protocols exclude fasting glucose ≥126 mg/dL or haemoglobin A1c ≥6.5% (average blood sugar over three months). * **Baseline creatine kinase:** Creatine kinase is a muscle enzyme that leaks into blood when muscle is damaged. Elevation above the upper limit of normal on two occasions signals underlying myopathy (muscle disease) and disqualifies from gene-transfer protocols. * **Sex and reproductive status:** Premenopausal women face the follicle-stimulating hormone suppression risk that postmenopausal women and men largely do not. Registered protocols enrol only postmenopausal women and require contraception for six months afterwards. * **Cancer history and genetics:** Because follistatin blocks a growth-restraining signal, any cancer history, or a hereditary cancer syndrome, is treated as disqualifying rather than cautionary in current protocols. * **Liver and kidney status:** The liver clears follistatin and is the vector's main off-target organ. Cirrhosis, viral hepatitis, transaminases (liver enzymes released by damaged liver cells) above 1.5 times normal, or kidney filtration below 60 mL/min/1.73 m² magnify risk. * **Age:** Above roughly 75 the combination of reduced hepatic reserve, higher baseline vector seroprevalence and greater comorbidity burden shifts the balance; registered protocols cap enrolment at 75 years. ## Key Interactions & Contraindications * **Immunosuppressants (rapamycin, prednisolone, tacrolimus):** Caution and deliberate use. Gene-transfer protocols give rapamycin for about two months to blunt the vector response, accepting mouth ulcers, raised lipids and infection risk as the trade. * **Corticosteroids (prednisone, dexamethasone):** Caution — chronic use causes muscle wasting that directly opposes the intervention, and confounds any assessment of whether it worked. * **Anticoagulants and antiplatelets (warfarin, apixaban, clopidogrel):** Caution before multi-site intramuscular injection; consequence is intramuscular haematoma. Protocols require washout or exclude platelet counts below 100 × 10⁹/L. * **Statins and other muscle-toxic drugs (simvastatin, daptomycin, colchicine):** Monitor. These raise creatine kinase, making drug-induced myopathy indistinguishable from injection-related muscle injury without a pre-treatment baseline. * **Muscle-targeting interventions (resistance training, blood-flow restriction training, growth hormone secretagogues — which prompt the body's own growth hormone release — such as ipamorelin):** Potentiating for training, additive and unquantified for secretagogues; monitor tendon tolerance, since two hypertrophic stimuli load connective tissue faster than it adapts. * **Anti-vascular endothelial growth factor agents (bevacizumab, ranibizumab):** Absolute contraindication in protocols combining follistatin with a vascular growth factor plasmid, since the two directly oppose each other. * **Androgens and anabolic steroids (testosterone, oxandrolone, selective androgen receptor modulators):** Additive anabolic effect with additive tendon and cardiac strain; monitor, and separate their introduction so attribution stays possible. * **Incretin mimetics (semaglutide, tirzepatide — drugs that copy a gut hormone to curb appetite):** Monitor. These drive lean-mass loss alongside fat loss, so combining them obscures whether follistatin preserved muscle or the other drug removed it. * **Over-the-counter anti-inflammatories (ibuprofen, naproxen, aspirin):** Caution around dosing days; they increase injection-site bleeding and bruising, and high sustained doses blunt the training-induced signalling the added muscle depends on. * **Supplements with additive myostatin-pathway claims (epicatechin, fertilised egg-yolk isolates, sulforaphane):** Monitor. Evidence for each is weak, but stacking them makes any observed change uninterpretable and is a common reason for over-attribution. * **Supplements that raise creatine kinase or mask it (creatine monohydrate, high-dose caffeine pre-workout):** Monitor; they shift the creatine kinase baseline used to detect muscle injury. Separation of at least 72 hours from any blood draw is required. **Populations who should avoid Follistatin:** * Anyone with any history of cancer, or a hereditary cancer syndrome (BRCA1/2, Lynch, Li-Fraumeni — inherited faults in genes that normally restrain tumour growth) without genetic counselling clearance * Pregnant, breastfeeding or premenopausal women, and anyone unwilling to use contraception for six months afterwards * People with diabetes of any type, fasting glucose ≥126 mg/dL, or haemoglobin A1c ≥6.5% * People with chronic kidney disease stage 3–5 (filtration below 60 mL/min/1.73 m²) * People with cirrhosis, chronic viral hepatitis, or transaminases above 1.5 times the upper limit of normal * People with dysferlinopathy (an inherited muscle-wasting disorder), in which myostatin-blockade-driven hypertrophy accelerated degeneration in animal models * People with heart failure, ejection fraction below 50%, prior myocardial infarction or stroke at any interval, or left ventricular wall thickness ≥15 mm * People with prior exposure to any adeno-associated virus gene therapy, or an antibody titre above 1:100 * People with active infection, tuberculosis, hepatitis B or C, or human immunodeficiency virus * People with inflammatory myopathy, systemic connective-tissue disease, or a history of rhabdomyolysis (breakdown of muscle releasing damaging contents into blood) ## Risk Mitigation Strategies * **Vector antibody screening before committing:** An adeno-associated virus serotype 9 antibody titre above 1:100 predicts both wasted dose and immune reaction. Testing first avoids paying for a therapy that cannot work. * **Dose escalation from the lower vector dose:** The [registered protocol](https://clinicaltrials.gov/study/NCT07443826) escalates from 5 × 10¹⁰ to 1 × 10¹¹ vector genomes per kilogram only after a 21-day observation window, limiting exposure to dose-dependent liver and immune toxicity. * **Rapamycin cover for the vector:** About two months of rapamycin from the time of dosing blunts the anti-vector immune response that otherwise destroys transduced fibres and drives fever and transaminase rises. * **Baseline and serial liver panel:** Alanine aminotransferase (a liver enzyme) and bilirubin before dosing, then weekly for eight weeks, catches vector hepatotoxicity while it is still reversible; stopping rules trigger above three times the upper limit. * **Three-monthly glucose-handling checks:** Fasting insulin, glucose and haemoglobin A1c detect the insulin resistance that higher follistatin is associated with, before it becomes diagnosable diabetes. * **Lab-verified product or none:** Nearly half of grey-market vials contain no follistatin. Demanding an independent certificate of analysis and endotoxin result is the only defence against injecting an unknown peptide. * **Progressive loading of the muscle:** Tendon and connective tissue adapt more slowly than muscle. Holding load increases to roughly 5–10% per week for the first three months reduces the mismatch that animal tendon data predict. * **Strength baseline, not just a scan:** Recording leg-press one-repetition maximum and grip strength before treatment is what distinguishes real functional gain from the volume-only result all three controlled trials produced. ## Therapeutic Protocol * **Intramuscular gene transfer, the trial regimen:** Adeno-associated virus serotype 1 carrying FS-344 at [3 × 10¹¹ or 6 × 10¹¹ vector genomes per kilogram per leg](https://pubmed.ncbi.nlm.nih.gov/25322757/), injected bilaterally into quadriceps under image guidance as a single administration. * **Intramuscular gene transfer, the longevity-clinic regimen:** Adeno-associated virus serotype 9 carrying follistatin at [5 × 10¹⁰ or 1 × 10¹¹ vector genomes per kilogram](https://clinicaltrials.gov/study/NCT07443826), distributed bilaterally across quadriceps, gluteus maximus, gastrocnemius and biceps brachii, with rapamycin cover. * **Plasmid gene transfer, the alternative approach:** A [single subcutaneous dose](https://clinicaltrials.gov/study/NCT06411366) of non-integrating FS-344 plasmid, giving transient expression measured against a three-month serum follistatin change; repeat dosing is offered rather than a permanent effect. * **Engineered local protein, the discontinued approach:** ACE-083 at 150–240 mg per muscle, injected into a named muscle every three weeks for up to twelve months. Enlarges only the injected muscle; the programme was stopped. * **Competing approaches presented without default:** Gene transfer, engineered protein and grey-market peptide differ in duration, reversibility and evidence. Gene transfer has the only peer-reviewed efficacy signal; the engineered protein has the only randomised muscle-volume data. * **Who popularised each:** Mendell and colleagues at Nationwide Children's Hospital pioneered AAV1-FS344 gene transfer; Acceleron Pharma developed ACE-083; Minicircle brought plasmid follistatin to private clinics in Roatán, Honduras. * **Best time of day:** No circadian data exist. Trial dosing occurs in daytime clinic hours. Endogenous follistatin rises with fasting and after exercise, so pre-dose sampling is standardised to a fasted morning state. * **Half-life determines the format:** Native follistatin-315 clears within minutes through heparan-sulfate binding and hepatic uptake. Gene transfer and Fc fusion exist precisely because the unmodified protein cannot sustain an exposure. * **Single versus split dosing:** Gene transfer is a single administration by design. The engineered protein was split across muscles and repeated three-weekly. Grey-market peptide protocols use daily microdoses with no pharmacokinetic justification. * **Genetic factors in dose choice:** Glucokinase regulatory protein genotype sets baseline circulating follistatin, and vector antibody titre — partly genetically influenced — determines whether the dose reaches muscle at all. * **Sex-based differences:** No head-to-head data exist. Randomised follistatin-domain protein data come exclusively from postmenopausal women; gene-transfer trials enrolled mostly men. Protocols exclude premenopausal women rather than dose-adjust. * **Age-related considerations:** Protocols cap enrolment at 75 years. Satellite-cell reserve falls with age, so equivalent vector doses are expected to yield smaller gains in the oldest recipients. * **Baseline biomarkers guiding response:** Serum follistatin, myostatin and activin A before dosing define the brake available to release; body composition by scan defines the starting point against which change is judged. * **Pre-existing conditions altering response:** Insulin resistance, fatty liver and obesity raise endogenous follistatin and blunt the relative increment; inflammatory myopathy and advanced muscle loss predicted poor response in the myositis trial. ## Discontinuation & Cycling * **Viral gene transfer cannot be discontinued:** Adeno-associated virus genomes persist as free-floating DNA circles in muscle fibres, which never divide, so expression continues for years. There is no off switch, no antidote and no way to lower the dose afterwards. * **Plasmid gene transfer is self-limiting:** Non-integrating plasmid expression decays over weeks to months as the plasmid is diluted and degraded, which is why repeat administration is offered. This makes it the reversible option by default. * **Injected protein stops when dosing stops:** The engineered local protein was given three-weekly; muscle volume regresses once injections cease, since nothing sustains the exposure. * **No withdrawal syndrome documented:** Neither trial programme reported rebound weakness, symptom flare or hormonal disturbance on stopping. Absence of reports reflects small samples and short follow-up, not established safety. * **No tapering protocol exists:** Tapering is inapplicable to a single-dose gene transfer and was not used for the three-weekly protein. Nothing in the literature describes a step-down schedule. * **Cycling is not established:** No evidence supports cycling to maintain effect. For viral vectors the reverse holds — neutralising antibodies generated by the first dose make re-dosing with the same serotype ineffective. * **Lifelong versus short-term framing:** Viral gene transfer is a one-way, effectively lifelong commitment. Plasmid and protein approaches are short-term and repeatable. This distinction matters more than any other protocol choice. ## Sourcing and Quality * **No licensed product exists anywhere:** No regulator has approved a follistatin product for any indication. Every route of access is investigational, compassionate, offshore or illicit, and that is the governing sourcing fact. * **Grey-market vials frequently contain something else:** Of seventeen products sold as "follistatin 344", fewer than ten contained follistatin; others held mechano growth factor or growth hormone releasing peptide 2 ([Reichel et al., 2019](https://pubmed.ncbi.nlm.nih.gov/31758732/)). * **What to look for in a certificate of analysis:** Identity by mass spectrometry, purity above 98% by chromatography, an endotoxin result below 1 endotoxin unit per milligram, and a batch number matching the vial label rather than a generic template. * **The histidine tag is a red flag:** Every grey-market sample containing follistatin carried a bacterial purification tag and protein clumping absent from human follistatin — a marker of research-grade material never intended for injection. * **Gene-therapy vector quality is unverifiable by the buyer:** Vector genome titre, empty-capsid fraction and residual genetic material from the production cells determine both dose and immune risk, and only good manufacturing practice production documents them. Private clinics rarely release these data. * **Compounding pharmacies are not a route:** Follistatin is a 315-residue glycoprotein requiring mammalian cell expression; no compounding pharmacy can produce it, so any "compounded follistatin" offer is a resale of unverified material. * **Reputable sources are the registered trials:** Nationwide Children's Hospital ran the peer-reviewed gene-transfer programme; the currently registered protocols are run by Minicircle and Unlimited Biotechnology at the Global Alliance for Regenerative Medicine site in Roatán. ## Practical Considerations * **Time to effect:** Muscle volume changes are measured at three to six months in every protocol. The engineered protein showed significant volume gain by six months; gene-transfer trials assessed walking distance at six months and one year. * **Pitfall — equating volume with strength:** The most consistent finding across the three placebo-controlled trials ([Glasser et al., 2018](https://pubmed.ncbi.nlm.nih.gov/29486514/); [Statland et al., 2022](https://pubmed.ncbi.nlm.nih.gov/35428982/); [Thomas et al., 2022](https://pubmed.ncbi.nlm.nih.gov/35545446/)) is bigger muscle without better function. Expecting strength to track the scan is the field's defining error. * **Pitfall — treating grey-market vials as the same intervention:** Injected peptide, plasmid and viral vector differ in everything that matters. Evidence generated with one does not transfer to another, and the peptide has no human pharmacokinetic data at all. * **Pitfall — not training the treated muscle:** Untrained added tissue is metabolically active but functionally inert. None of the clinic protocols mandate a resistance-training programme; the gene-transfer trials included one. * **Pitfall — stacking and losing attribution:** Combining follistatin with androgens, peptides or myostatin-claim supplements makes any result uninterpretable and multiplies the tendon and cardiac load. * **Regulatory status:** Not approved by the Food and Drug Administration or any other regulator. Clinic administration occurs offshore. Follistatin is prohibited in sport under the World Anti-Doping Agency hormone and metabolic modulator category, and gene transfer separately under gene doping. * **Cost and accessibility:** Private plasmid gene therapy has been priced around 25,000 US dollars per administration, viral protocols more, plus travel to Honduras and local residency status. No insurer or health system reimburses any route, so payer incentives bias neither guidelines nor funding. ## Interaction with Foundational Habits * **Sleep:** Indirect and unmeasured. No study has examined follistatin and sleep in either direction. The plausible link runs through growth hormone secretion and muscle protein synthesis during slow-wave sleep, which set the ceiling on what any anabolic signal can achieve, so sleep restriction is expected to blunt rather than block the effect. * **Nutrition:** Direct and bidirectional. Circulating follistatin is liver-derived and driven by the glucagon-to-insulin ratio, so fasting and carbohydrate restriction raise it ([Hansen et al., 2016](https://pubmed.ncbi.nlm.nih.gov/26652766/)). Practically, adequate protein remains the limiting input for building the muscle follistatin permits, and fasted blood draws are needed for meaningful monitoring. * **Exercise:** Direct and potentiating. Resistance training raises follistatin and lowers myostatin in adults of every age ([Khalafi et al., 2023](https://pubmed.ncbi.nlm.nih.gov/37328021/)), acting on the same axis. Because added tissue only becomes strength under load, training is the mechanism that converts the volume gains seen on imaging into anything usable. * **Stress management:** Indirect and opposing. Chronic cortisol elevation drives muscle protein breakdown, working against the intervention, and activin A — one of the ligands follistatin sequesters — is itself an inflammatory mediator released under systemic stress. No study has measured whether stress reduction alters response, so the direction is inferred from mechanism. ## Monitoring Protocol & Defining Success Before any follistatin protocol, a baseline set is drawn: serum follistatin, myostatin and activin A to define the starting signal; fasting glucose, insulin and haemoglobin A1c because higher follistatin tracks with insulin resistance; a liver panel and creatine kinase because the vector targets liver and the injections traumatise muscle; follicle-stimulating hormone and sex hormones to detect activin-axis suppression; kidney filtration; and a body-composition scan with recorded leg-press and grip strength, since volume and function must be tracked separately. Adeno-associated virus antibody titre is measured once, before dosing, as an eligibility test. Afterwards, liver enzymes and blood counts are checked weekly for eight weeks while the vector immune response plays out, then at three, six and twelve months alongside glucose handling, follistatin, hormones and body composition; thereafter every six to twelve months indefinitely, because gene transfer is not reversible. | Biomarker | Optimal Functional Range | Why Measure It? | Context/Notes | |---|---|---|---| | Serum follistatin | No established target; fold-change from the individual's own baseline is what protocols track (healthy adults sit roughly 1–10 ng/mL) | Confirms the product expressed at all | Fasted morning draw; levels rise with fasting and exercise, so conditions are standardised | | Fasting insulin | 2–5 µIU/mL | Earliest sign of the insulin resistance linked to higher follistatin | Conventional labs report up to about 25 µIU/mL as normal, far above the functional target; 12-hour fast; paired with fasting glucose to derive HOMA-IR (homeostatic model assessment of insulin resistance, a calculated index of insulin action) | | Haemoglobin A1c | Below 5.3% | Confirms glucose handling has not drifted | Conventional threshold is 5.7%; protocols exclude above 6.5% | | Fasting glucose | 75–85 mg/dL | Complements insulin for the same signal | Conventional reference tops out at 99 mg/dL, well above the functional target | | Alanine aminotransferase | Below 20 U/L (men), below 17 U/L (women) | Detects vector-related liver injury while reversible | Conventional labs report up to 40–55 U/L as normal; weekly for 8 weeks post-vector, stopping rule at 3× upper limit | | Creatine kinase | 45–200 U/L | Distinguishes injection trauma and myopathy from expected soreness | Strenuous exercise within 72 hours of the draw renders the result uninterpretable | | Follicle-stimulating hormone | Men 1.5–8 IU/L; postmenopausal women 25–135 IU/L | Detects suppression of the activin axis follistatin acts on | IU/L is international units per litre; morning draw; paired with luteinising hormone, testosterone or oestradiol | | Appendicular lean mass index | Men 8.0–9.5 kg/m²; women 6.5–7.5 kg/m² | The actual outcome being purchased | Same scanner, same hydration and fasting state each time; sarcopenia thresholds sit at 7.0 and 5.5 kg/m² | | Estimated glomerular filtration rate | Above 90 mL/min/1.73 m² | Higher follistatin associates with chronic kidney disease | Cystatin C-based estimate is preferable in high-muscle individuals, where creatinine overstates impairment | | High-sensitivity C-reactive protein | Below 0.5 mg/L | Tracks the inflammatory tone that activin blockade shifts | Conventional cardiovascular cut-offs place low risk below 1.0 mg/L and high risk above 3.0 mg/L; testing is deferred for two weeks after injections, which transiently raise it | Qualitative markers worth tracking alongside the laboratory set: * Perceived strength at a fixed working weight, recorded as reps at a constant load rather than by feel * Recovery time between hard training sessions * Joint and tendon comfort during loaded movement, since connective tissue is the predicted lagging element * Visible limb girth asymmetry, which after single-muscle injection is expected rather than alarming * Energy, appetite and sleep quality * Nosebleeds, visible small dilated surface vessels, or unexplained bruising ## Emerging Research * **First trial in healthy ageing adults:** [NCT07443826](https://clinicaltrials.gov/study/NCT07443826) is a recruiting Phase 1/2a study of intramuscular adeno-associated virus serotype 9 follistatin, alone or after a vascular growth factor plasmid, in 12 adults aged 35–75. Primary endpoint is treatment-emergent adverse events through day 90. * **Completed plasmid safety study:** [NCT06411366](https://clinicaltrials.gov/study/NCT06411366) enrolled 43 healthy adults in a single-arm Phase 1 of injectable FS-344 plasmid, with adverse events and three-month serum follistatin change as co-primary endpoints and body composition as secondary. Results circulated as a company preprint, not peer-reviewed. * **Combination gene therapy:** [NCT07285629](https://clinicaltrials.gov/study/NCT07285629) completed in April 2026, an early Phase 1 of klotho plus follistatin gene therapy in 14 healthy adults. Whether combining two longevity-associated genes changes either safety or effect is entirely unreported so far. * **The source of the human efficacy signal:** [NCT01519349](https://clinicaltrials.gov/study/NCT01519349) is the completed Phase 1 that enrolled 15 patients with Becker muscular dystrophy or sporadic inclusion body myositis and produced the only peer-reviewed walking-distance results for follistatin gene transfer. * **Evidence that could weaken the case:** [Prokopidis et al., 2026](https://pubmed.ncbi.nlm.nih.gov/42202210/) found circulating follistatin no different between older adults with and without sarcopenia, and [Wu et al., 2021](https://pubmed.ncbi.nlm.nih.gov/34759311/) tied higher follistatin to incident type 2 diabetes across two cohorts. * **The unresolved volume-versus-function question:** The discontinued engineered-protein programme ([Statland et al., 2022](https://pubmed.ncbi.nlm.nih.gov/35428982/); [Thomas et al., 2022](https://pubmed.ncbi.nlm.nih.gov/35545446/)) showed muscle can be grown without function following. Future trials powered on strength rather than imaging would settle it. * **Delivery and durability:** Whether systemic rather than single-muscle delivery is achievable at tolerable vector doses, and how long expression persists, remain open. [Datta-Mannan et al., 2013](https://pubmed.ncbi.nlm.nih.gov/23249626/) shows protein engineering can solve the exposure problem without a virus. ## Conclusion Follistatin blocks the body's own brake on muscle growth. That much is settled, and it is why the molecule keeps drawing attention from people who take building and keeping muscle seriously. What the human evidence supports is narrower than the enthusiasm around it. Injecting a follistatin-based protein into a single muscle makes that muscle measurably bigger, and this has been shown repeatedly against placebo. Placing the follistatin gene into the thigh muscles of people with wasting diseases improved how far most of them could walk, in studies too small, and run without a comparison group, to settle anything. The consistent disappointment is that bigger muscle has not reliably become stronger muscle, and the company that produced the clearest growth data stopped its programme for exactly that reason. On the other side sit signals that deserve weight. People whose own follistatin runs high develop more diabetes, stroke, heart failure and kidney disease over decades, although that describes the body's own protein rather than an injected one. Blocking a growth-restraining signal for life raises questions about tumours and tendons that no study has answered. Nearly half the material sold online is not follistatin at all. Much of the strongest evidence also comes from parties selling the product: the growth trials were run by the maker, and the clinic results released by the clinic. The gap between what has been shown and what is offered is unusually wide, and the durable choices — viral gene transfer above all — cannot be undone. **[Top](#top) - [Benefits](#expected-benefits) - [Risks](#potential-risks--side-effects) - [Protocol](#therapeutic-protocol)**