Irisin
The exercise hormone found in mice in 2012, and the long argument over whether human blood tests for it measure anything
Irisin is the piece of the muscle membrane protein FNDC5 that is clipped off and released into the blood. In mice it turned white fat into a heat-burning, brown-like tissue, strengthened bone and protected memory, which made it famous as an exercise hormone. In people the story is far less settled: most human studies measured it with commercial antibody kits that later proved unreliable, mass spectrometry finds only a few nanograms per millilitre, and no one has ever been given irisin as a treatment.
No established human dose Details ↓
Recommended Dosing
No established human dose
Irisin has never been given to a person in any study read. Every human study on this page measures it in blood or spinal fluid, and the measurement itself is disputed: antibody kits bound other proteins, and mass spectrometry finds only a few nanograms per millilitre.
What is it?
Irisin was described in 2012 by a Harvard group studying PGC1-alpha, a gene switch that exercise turns on in muscle. In mice, more PGC1-alpha meant more of a membrane protein called FNDC5, and a piece of FNDC5 was cut off and released into the blood. The group named that piece irisin. Added to white fat cells, it switched on UCP1, the heat-producing protein of brown fat, and a modest rise in blood irisin made mice burn more energy and handle glucose better without moving or eating differently. The paper reported that exercise raised irisin in mice and in people. Because its title does not use the word irisin, it is listed under References rather than as a card.
Later mouse work widened the claims. Small weekly injections of recombinant irisin made the cortical bone of young male mice thicker and stronger. Irisin was found to bind αV integrins on bone cells and fat cells, and mice lacking FNDC5 were protected from the bone loss that follows removal of the ovaries. In mouse models of Alzheimer's disease, raising brain FNDC5 or irisin restored memory, and mice without the gene lost some of the cognitive benefit of exercise. In a Parkinson's model, irisin delivered by a viral vector reduced the spread of toxic α-synuclein and protected dopamine neurons. All of this is animal and cell work.
The human picture has been contested from the start. Human FNDC5 begins with an unusual start codon (ATA rather than ATG), and in one 2013 study a construct with that start produced only about 1% as much protein as the standard version; the same study found no rise in FNDC5 messenger RNA in human muscle after training. In 2015 a group tested four commercial antibodies used in irisin ELISA kits and found they stuck to many other blood proteins, calling earlier human measurements into question. The original group answered later that year with mass spectrometry, the most specific method available, and found irisin in human plasma at roughly 3.6 ng/mL, rising to about 4.3 ng/mL with interval training. That is far lower than many antibody-based studies had reported.
The argument did not end there. A 2020 study by the sceptical group found that the antibodies still bound non-specific proteins and that even mass spectrometry needed a multi-step preparation that introduced variation it could not control. Mass spectrometry has also found glycosylated irisin in the cerebrospinal fluid of older men, though in that study it could not find irisin in their paired plasma. So the plain position today is that irisin probably exists in human blood at low levels, but its true concentration, and whether it changes meaningfully with exercise, is not reliably known.
This matters for reading everything else. Hundreds of human studies, and the meta-analyses that pool them, link irisin levels to diabetes, osteoporosis, thyroid disease and exercise, almost all using ELISA kits. Those pooled results are listed on this page because they are what the literature contains, but they inherit the measurement problem: a difference in an ELISA reading may not be a difference in irisin. Training trials disagree too. A 26-week randomised trial found no training effect once the age of stored samples was accounted for, and meta-analyses report everything from a fall to a rise.
Irisin has never been given to a person. There is no approved product, no human pharmacokinetic study and no human dose. Anything sold as irisin for injection would be outside any regulated supply, and the mouse doses do not translate into a human amount.
Mechanism of Action
In mice, exercise turns on PGC1-alpha in muscle, which raises FNDC5; the outer part of FNDC5 is cut off and released as irisin, which acts on other tissues through αV integrins. How much of this happens in people is uncertain.
From muscle to blood
Exercise raises PGC1-alpha and FNDC5 in mouse muscle, and irisin is the cleaved, circulating part of FNDC5.
Brown-like fat (mice)
Irisin switched on UCP1 and a brown-fat programme in mouse white fat, raising energy expenditure; recombinant irisin failed to do this in human fat precursor cells.
Integrin signalling in bone and brain
Binding to αVβ5 integrin, helped by extracellular Hsp90α, irisin acts on osteocytes and osteoblasts and, in mouse models, on neurons involved in memory.
Published Research
These are studies of irisin read from their PubMed abstracts: the mass-spectrometry measurements and the papers that challenged antibody tests, a randomised training trial and five meta-analyses of measured levels, and the mouse and cell work behind its reputation. No study here gave irisin to a person. Most human level studies used ELISA kits whose reliability is disputed, which each card notes where it applies.
Mass spectrometry finds irisin in human plasma
Using heavy-isotope labelled reference peptides, the original group identified irisin in human plasma, translated mainly from the unusual ATA start codon. It circulated at about 3.6 ng/mL in sedentary people and about 4.3 ng/mL in people doing aerobic interval training.
Glycosylated irisin in cerebrospinal fluid, not seen in plasma
Mass spectrometry detected the glycosylated form of irisin in the cerebrospinal fluid of men over 80 with various illnesses, at roughly 0.26 to 1.86 ng/mL with large differences between people. The same method did not detect irisin in their paired plasma samples.
Age, sex and irisin in fluid around the brain
In paired cerebrospinal fluid and plasma from 71 healthy people, measured by ELISA and mass spectrometry, irisin in the fluid rose with age and fell with body mass index, and plasma levels were higher in men than women. The authors suggest transport into the brain may be saturable.
26 weeks of training did not raise irisin
In a randomised trial of 102 inactive adults aged 30 to 60, neither endurance nor strength-endurance training three times a week for 26 weeks changed serum irisin more than in controls, though fitness improved. An apparent upward drift was explained by how long the frozen samples had been stored.
Commercial irisin antibodies bind the wrong proteins
Western blots of four polyclonal antibodies used in commercial irisin kits showed strong binding to unrelated proteins in human and animal serum and no band at the expected size, even with recombinant irisin as a control. An FNDC5 signal near 20 kDa was found in human serum by mass spectrometry but was missed by the antibodies.
Still chasing shadows: a warning on irisin assays
Testing transcripts, antibodies and a mass-spectrometry method across species, the authors found more human FNDC5 transcript variety than annotated but none predicting a truncated irisin; antibodies still bound non-specific serum proteins; and the mass-spectrometry workflow introduced variation that could not be controlled. They advise against measuring circulating irisin with the methods available.
Evidence against an effect in humans
Human FNDC5 carries an ATA start codon, and cells given a human construct with it made about 1% of the protein made with ATG. Contracting human muscle cells raised PGC1-alpha but not FNDC5, training did not change FNDC5 in human muscle biopsies, and recombinant irisin did not brown human fat precursor cells.
Does circulating irisin detection really work?
A short commentary arguing that most of the conflicting human results on irisin trace back to weaknesses in the methods used to detect and measure it.
Meta-analysis: chronic training and irisin
Pooling three randomised trials (173 untrained healthy adults), at least eight weeks of training lowered circulating irisin moderately compared with controls; nine non-randomised studies showed essentially no change from baseline. The authors called for more research given the inconsistency.
Meta-analysis: resistance training and irisin
Seven randomised trials (282 adults) using ELISA kits showed a non-significant tendency for irisin to rise after at least eight weeks of resistance training, significant in older adults and with progressive, demanding programmes; the authors ask for studies using mass spectrometry.
Meta-analysis: one endurance session
Across 16 studies of 412 healthy adults, a single session of continuous endurance exercise raised circulating irisin by a small amount, while interval sessions produced no clear change.
Meta-analysis: lower readings in type 2 diabetes
In 23 observational studies (1,745 patients, 1,337 controls), irisin readings were lower in type 2 and gestational diabetes than in controls but about 30% higher in type 1 diabetes, with larger falls in Asian than European patients.
Meta-analysis: lower readings in osteoporosis
Seven studies of 1,018 middle-aged and older people found lower irisin in those with osteoporosis, lowest in postmenopausal women and people with fractures, and a weak positive link with bone mineral density; publication bias was possible.
Low-dose irisin thickened mouse cortical bone
Young male mice given weekly recombinant irisin at a dose too low to brown their fat gained cortical bone mass and bending strength, mainly through more bone formation and fewer bone-resorbing cells; the spongy (trabecular) bone and vertebrae were unchanged.
αV integrins identified as irisin's receptor
Irisin bound αV-class integrins, and blocking them stopped its signalling in bone and fat cells. It raised osteocyte survival and sclerostin production, and mice lacking FNDC5 were protected from the bone loss that follows removal of the ovaries.
FNDC5/irisin and memory in Alzheimer's models
FNDC5/irisin was reduced in hippocampus and cerebrospinal fluid from people with Alzheimer's disease and in mouse models. Lowering it in mouse brain impaired memory, raising it rescued synaptic plasticity and memory, and blocking it blunted the protective effect of exercise.
Mice without irisin lose exercise's cognitive benefit
Mice lacking Fndc5/irisin had worse cognition with exercise, ageing and Alzheimer-type disease, and abnormal new neurons in the dentate gyrus. Raising blood irisin through liver gene delivery increased brain irisin and improved memory and pathology in Alzheimer mouse models.
Irisin in a mouse model of Parkinson's disease
In mice injected with α-synuclein fibrils, irisin delivered through a viral vector reduced pathological α-synuclein, prevented loss of dopamine neurons and striatal dopamine, and lessened motor deficits; in cultured neurons it speeded the breakdown of the toxic protein.
Crystal structure: irisin forms a dimer
The 2.28 Å crystal structure of the FNDC5 ectodomain showed a fibronectin III-like domain followed by a flexible tail, with two molecules joined through a continuous β-sheet not seen before in this protein family; glycosylation did not affect dimer formation.
How irisin signals through its integrin
Muscle released extracellular Hsp90α with exercise, which activated integrin αVβ5 and allowed irisin to bind it tightly; irisin docked on a surface of the integrin different from the one its known ligands use.
Recombinant irisin and bone-forming cells
Bacterially produced irisin increased growth and maturation of several osteoblast cell types, mainly by shifting them towards aerobic glycolysis; blocking that metabolic shift removed the effect.
No Established Human Dose
Irisin has never been given to a person in any study read. Every human study on this page measures it in blood or spinal fluid, and the measurement itself is disputed: antibody kits bound other proteins, and mass spectrometry finds only a few nanograms per millilitre. The treatment studies are all in mice or in cells, so there is no human dose to report and none can be worked out from them.
What this rests on: PMID 26278051, PMID 25749243, PMID 24191966, PMID 26374841.
Educational information, not medical advice. The doses on this page come from approved labels, published studies and, where no study exists, amounts reported by users. None is a recommendation for you. An unapproved compound has no established safe or effective human dose, and products sold for “research use only” are not made or tested for people. Talk to a doctor before acting on anything on this site, including before you start, stop or change any medicine or dose. Full medical disclaimer.
What a Buyer Needs to Know
What is sold, how it is mixed, stored and taken, and what is known about its safety, quality and legal status. Each part names its source. Where we have none, it says unknown.
What is sold
Unknown: we hold no label and no published test that says what forms or vial sizes Irisin is sold in. A vial's own label is a seller's claim, not a source.
Mixing a vial: a worked example
Unknown: no human dose of Irisin has been established (see No Established Human Dose below), so there is no amount to work out. The reconstitution calculator does the arithmetic for any vial, water and dose.
Storage before and after mixing
Unknown: no label or published stability study on file says how to store Irisin, before or after mixing. The mixing guide sets out the general rules, with their sources: keep a mixed vial in the fridge, and use it promptly, as the bacteriostatic water label says of any drug dissolved in it. The 28 days often quoted is the limit for an opened multiple-dose vial from a manufacturer, an outer limit set for sterility, not a time a mixed solution has been shown to keep. Nobody has published how long Irisin stays intact once mixed.
How it is taken
Unknown: with no established human dose, there is no established way of taking Irisin. What the human studies gave and how is set out under No Established Human Dose.
What to expect, and when
Unknown: no cited half-life and no label or trial duration on this page says how long Irisin lasts or how long it was given. Results reported in the studies are under Published Research.
Side effects, and when to stop or get help
From the Side Effects section: No human trial, case report or pharmacokinetic study of irisin administration was found; In mice, higher weekly doses browned fat while low doses did not, so effects depend on amount in ways unknown in people (PMID 26374841). Full list and cautions ↓
When to stop or get help: unknown; no label or trial on file sets rules for stopping Irisin. Treat a severe or unexpected reaction as a reason to stop and see a doctor.
Source: the Side Effects section below (“No human safety data exist, because irisin has not been given to people”).
Interactions
Unknown: a PubMed search for human interaction studies of Irisin (reviewed 2026-10-28) found none. Unknown is not the same as safe to combine.
Every peptide's record, with its source: the interaction matrix.
Checking what you bought
Unknown: no published independent test of Irisin products is in our lab-results database yet, so how products sold under this name measure up is not known.
To read a certificate of analysis, paste it into the COA decoder. Trust & Verification explains what independent testing can and cannot show.
Legal status
US: Research Chemical (Not a medicine anywhere: no approval and no human treatment trial in the sources read). Sport: Prohibited (S0 catch-all).
Country by country ↓ · Regulatory tracker
Source: this entry's approval status and the WADA Prohibited List, as in Legal Status below.
Side Effects & Contraindications
Reported Side Effects
No human safety data exist, because irisin has not been given to people.
Contraindications & Cautions
No product label exists. These cautions follow from the evidence.
Legal Status
Irisin is a research protein; it is not approved as a medicine anywhere in the sources read.
Frequently Asked Questions
?What is irisin?
A protein hormone cut from FNDC5, a protein in the membrane of muscle cells. It was described in 2012 as a hormone released with exercise that turned white fat brown-like in mice.
?Does exercise raise irisin in people?
Possibly a little. Mass spectrometry found about 3.6 ng/mL in sedentary people and 4.3 ng/mL with interval training, but a 26-week randomised training trial found no change, and meta-analyses disagree on the direction.
?Why is irisin measurement controversial?
Most human studies used commercial antibody kits, and tests showed those antibodies bind many other blood proteins. Mass spectrometry is more specific but still variable, so true human levels are uncertain.
?Has irisin been tested as a treatment?
Only in animals and cells, where it strengthened bone, protected memory in Alzheimer models and protected dopamine neurons in a Parkinson model. No study has given it to people.
?Is there a human dose of irisin?
No. It has never been given to a person, so there is no dose, and the mouse doses cannot be converted into one.
?Can I buy irisin?
No approved irisin product exists. Anything sold as irisin is outside regulated supply, and its contents are unverified.
References
- [pubmed] Jedrychowski MP, et al. "Detection and Quantitation of Circulating Human Irisin by Tandem Mass Spectrometry." Cell Metab, 2015;22(4):734-740. PMID: 26278051.
- [pubmed] Ruan Q, et al. "Detection and quantitation of irisin in human cerebrospinal fluid by tandem mass spectrometry." Peptides, 2018;103:60-64. PMID: 29574076.
- [pubmed] Ruan Q, et al. "The effects of both age and sex on irisin levels in paired plasma and cerebrospinal fluid in healthy humans." Peptides, 2019;113:41-51. PMID: 30716355.
- [clinical-trial] Hecksteden A, et al. "Irisin and exercise training in humans - results from a randomized controlled training trial." BMC Med, 2013;11:235. PMID: 24191966.
- [pubmed] Albrecht E, et al. "Irisin - a myth rather than an exercise-inducible myokine." Sci Rep, 2015;5:8889. PMID: 25749243.
- [pubmed] Albrecht E, et al. "Irisin: Still chasing shadows." Mol Metab, 2020;34:124-135. PMID: 32180552.
- [pubmed] Raschke S, et al. "Evidence against a beneficial effect of irisin in humans." PLoS One, 2013;8(9):e73680. PMID: 24040023.
- [review] Sanchis-Gomar F, et al. "Circulating irisin detection: Does it really work?." Trends Endocrinol Metab, 2015;26(7):335-6. PMID: 26008878.
- [review] Qiu S, et al. "Chronic Exercise Training and Circulating Irisin in Adults: A Meta-Analysis." Sports Med, 2015;45(11):1577-88. PMID: 26392122.
- [review] Cosio PL, et al. "Effect of Chronic Resistance Training on Circulating Irisin: Systematic Review and Meta-Analysis of Randomized Controlled Trials." Int J Environ Res Public Health, 2021;18(5). PMID: 33802329.
- [review] Cosio PL, et al. "Systematic Review and Meta-Analysis of Circulating Irisin Levels Following Endurance Training: Results of Continuous and Interval Training." Biol Res Nurs, 2023;25(3):367-381. PMID: 36426821.
- [review] Du XL, et al. "Lower Circulating Irisin Level in Patients with Diabetes Mellitus: A Systematic Review and Meta-Analysis." Horm Metab Res, 2016;48(10):644-652. PMID: 27300472.
- [review] Zhou K, et al. "Lower circulating irisin in middle-aged and older adults with osteoporosis: a systematic review and meta-analysis." Menopause, 2019;26(11):1302-1310. PMID: 31688577.
- [pubmed] Colaianni G, et al. "The myokine irisin increases cortical bone mass." Proc Natl Acad Sci U S A, 2015;112(39):12157-62. PMID: 26374841.
- [pubmed] Kim H, et al. "Irisin Mediates Effects on Bone and Fat via αV Integrin Receptors." Cell, 2018;175(7):1756-1768.e17. PMID: 30550785.
- [pubmed] Lourenco MV, et al. "Exercise-linked FNDC5/irisin rescues synaptic plasticity and memory defects in Alzheimer's models." Nat Med, 2019;25(1):165-175. PMID: 30617325.
- [pubmed] Islam MR, et al. "Exercise hormone irisin is a critical regulator of cognitive function." Nat Metab, 2021;3(8):1058-1070. PMID: 34417591.
- [pubmed] Kam TI, et al. "Amelioration of pathologic α-synuclein-induced Parkinson's disease by irisin." Proc Natl Acad Sci U S A, 2022;119(36):e2204835119. PMID: 36044549.
- [pubmed] Schumacher MA, et al. "The structure of irisin reveals a novel intersubunit β-sheet fibronectin type III (FNIII) dimer: implications for receptor activation." J Biol Chem, 2013;288(47):33738-33744. PMID: 24114836.
- [pubmed] A M, et al. "Irisin acts through its integrin receptor in a two-step process involving extracellular Hsp90α." Mol Cell, 2023;83(11):1903-1920.e12. PMID: 37267907.
- [pubmed] Zhang D, et al. "The bone anabolic effects of irisin are through preferential stimulation of aerobic glycolysis." Bone, 2018;114:150-160. PMID: 29775761.
- [pubmed] Boström P, et al. "A PGC1-α-dependent myokine that drives brown-fat-like development of white fat and thermogenesis." Nature, 2012;481(7382):463-8. PMID: 22237023.
Sources & Citations
Every study on this page was read from its PubMed record on October 28, 2026. Regulatory status comes from Drugs@FDA (openFDA) and the EMA medicines list, and trial status from ClinicalTrials.gov, read the same day. Grey Peptides sells nothing and links to no vendor.
Medical Disclaimer: This content is for educational and informational purposes only. It is not medical advice, and nothing here is an instruction to obtain or use any compound. Talk to a licensed clinician about your own situation.