Dihexa: Evidence, Retracted Studies and Legal Status (2026)

Quick answers

What is it?: A small lab-made molecule built from a fragment of angiotensin, a hormone best known for controlling blood pressure¹². University scientists designed it to help brain cells form new connections². It is sold online as a memory booster⁵.
Does it work?: Unknown in people: we found no study that gave it to humans³⁶. Rat and mouse studies reported better maze learning, but one independent rat study found no benefit¹⁷⁸. A key paper from its inventors was retracted for falsified data⁴.
Is it safe?: Unknown. FDA found no human safety data and cannot say whether it would harm people³. It is meant to boost a growth signal that many cancers use to grow and spread, so experts flag a possible cancer risk⁹¹⁰.
Is it legal?: It is not an approved medicine, and US pharmacies may not make it¹¹¹²¹³. FDA plans to ask its compounding advisers about it at a meeting before the end of February 2027³⁴. Likely banned in sport¹⁵¹⁶.
What is in the vials?: Unknown. Finnrick, an independent testing service, had published no tests of it as of 7 October 2026¹⁷. It is sold mostly as powder or capsules labeled for research only, and one seller describes the molecule wrongly⁵¹⁸.

What it is

Dihexa is a small, lab-made molecule derived from angiotensin IV, a six-amino-acid fragment of the blood-pressure hormone angiotensin¹². Its inventors kept two of angiotensin IV’s amino acids, tyrosine and isoleucine, and capped both ends so that the molecule resists breakdown, passes from the gut into the blood and reaches the brain in rats¹². They reported that dihexa binds hepatocyte growth factor (HGF), a natural growth signal, and boosts its action at its receptor, called MET, which helps nerve cells form new connections (synapses)¹. Two papers from the same laboratory on this mechanism, from 2012 and 2014, were retracted in 2025⁴¹⁹. An independent 2025 rat study did find that blocking MET canceled dihexa’s effect on memory²⁰, and FDA’s substance registry lists dihexa as an activator of the HGF receptor²¹.

Joseph Harding and Jay Wright of Washington State University began studying angiotensin IV in 1992, and Harding designed dihexa about five years before the university announced it in October 2012². Their company, M3 Biotechnology, was incorporated in 2011 and renamed Athira Pharma in April 2019²²². Athira licensed a dihexa patent from the university but developed a different drug, fosgonimeton, which the blood converts into dihexa²²²³.

Benefits: what the research shows

We found no study that gave dihexa itself to people. FDA found no human exposure data for products containing it by any route³, ClinicalTrials.gov listed no trial under ‘dihexa’ or ‘PNB-0408’ on 7 October 2026⁶, and a 2021 review by the Alzheimer’s Drug Discovery Foundation found no published human studies⁹. Everything known about dihexa itself comes from rats and mice. The only human data we found involve a different molecule. Athira Pharma’s fosgonimeton (ATH-1017) is an injected prodrug, an inactive form that the blood quickly converts into dihexa²¹²³. Its trial papers call the active product ATH-1001, a name FDA’s substance registry lists for dihexa²¹²⁸. Those trials tested an injected prodrug, not the dihexa powders and capsules sold online, so we do not count them as evidence for dihexa. The inventors’ main paper, published in 2013, reported that dihexa reversed the learning problems caused by the drug scopolamine in rats, whether given into the brain, by injection or by mouth, and that old rats given it by mouth learned a water maze better¹. In 2021 the journal published a notice of concern about that paper²⁵. In April 2025 it retracted a 2014 paper from the same group after a Washington State University investigation found falsified or fabricated data, for which Leen Kawas and Joseph Harding were found solely responsible⁴. It also retracted two related papers from 2011 and 2012¹⁹²⁶. A special committee of Athira’s board, which finished its investigation in October 2021, found that Kawas, Athira’s former chief executive, had altered images in her 2011 doctoral dissertation and in at least four papers she co-wrote at the university, published from 2011 to 2014²²²³. A 2021 review noted that dihexa did not improve rats with normal memory, a finding from the 2014 paper that was later retracted⁴⁹. Independent results are mixed. A team in China reported that dihexa restored maze learning and reduced brain inflammation in mice bred to develop Alzheimer’s-like disease⁷. A US team found that dihexa infused into the brain rescued working memory in rats after repeated mild head injuries, and that a MET blocker canceled the effect²⁰. Another team found that dihexa did not protect rats from the movement and memory damage caused by a toxin that mimics Huntington’s disease, in a study of 40 rats⁸. A rat nerve-repair study co-written by Harding reported better limb function when dihexa was given with stem cells into the muscle²⁷.

Key human studies

No completed human studies were found.

Doses studied

Doses below are what studies used or what a label says. They are not advice, and gray-market products may not contain what their label says.

In human studies

None. We found no published human study of dihexa, so we know of no dose tested in people by any route³⁹. Doses of fosgonimeton, the injected prodrug Athira Pharma tested, are doses of a different molecule and do not apply²³.

In animal studies

As reported in the papers, in animals only. Rats given scopolamine: 0.1 or 1 nmol into the brain; 0.05, 0.25 or 0.50 mg/kg by injection into the belly cavity; 1.25 or 2 mg/kg by mouth, each before daily training¹. Old rats: 2 mg/kg a day by mouth¹. Blood-level studies in rats: 10 mg/kg into a vein and 20 mg/kg into the belly cavity¹. Alzheimer’s-model mice: 1.44 or 2.88 mg/kg once a day for three months; the paper describes the route both as by stomach tube and as injection into the belly cavity⁷. Rats after repeated head injury: 1 pmol or 1 nmol a day into the brain²⁰. Rats after nerve repair: 2 to 4 mg/kg, given at the repair site, into the muscle or by injection²⁷. These are animal doses. They are not human doses, and we do not convert them.

What users report Anecdotal · not validated · not a recommendation

One community site lists 5 to 20 mg a day, taken once daily by mouth or on the skin, and says the figures come from vendor copy and forums, not trial data⁵. We found no human study to support any dose³.

Half-life: About 12.7 days in rats after an injection into a vein; we found no measurement in people who take dihexa¹The inventors’ 2013 rat study reported a half-life of 12.68 days after a 10 mg/kg dose into a vein, and 8.83 days after a 20 mg/kg injection into the belly cavity in 4 rats¹. In rat serum in a test tube, dihexa lasted 335.5 minutes, against 1.42 minutes for the angiotensin IV analog it was built from¹. The journal later published a notice of concern about this paper²⁵. In a trial of fosgonimeton, the injected prodrug that the blood converts into dihexa, the active product (ATH-1001) left the blood with a half-life of about 1.5 hours²¹²⁸. We found no study measuring dihexa in people after the oral or skin products sold online.
Routes: By mouth, as powder or capsules (the main gray-market form; no human data found)
On the skin, often dissolved in DMSO (sold this way; no human data found)
Injection into a vein, the belly cavity or the brain (rat and mouse studies only)

Every dose studied for Dihexa, with sources →

Legal status

United States: Dihexa is not approved by FDA for any use. A search of FDA’s approved-drug database (Drugs@FDA, through openFDA) for the ingredient name ‘dihexa’ returned no match on 7 October 2026¹¹. FDA says it ‘has not identified any human exposure data on drug products containing dihexa acetate administered via any route of administration’ and ‘lacks important information regarding any safety issues raised by dihexa acetate, including whether it would cause harm if administered to humans’ (page current 22 April 2026)³. A search of FDA’s adverse event database (FAERS, through openFDA) for ‘dihexa’ returned no reports on 7 October 2026³¹. Reporting is voluntary, so that does not show that no harm has occurred. FDA’s substance registry gives dihexa the identifier UNII 9WYX65A5C2²¹; that is a naming record, not an approval.
US pharmacies: FDA’s 503A list sorts the substances nominated for use by compounding pharmacies into three categories: under evaluation (Category 1), raising significant safety concerns (Category 2) and nominated without adequate support (Category 3)¹². Dihexa is in none of them on the list updated 14 May 2026¹². Dihexa acetate was in Category 2 until its nominator withdrew the nomination. FDA now lists it under ‘Bulk drug substances nominated but withdrawn’, substances ‘previously in category 2’ (page current 22 April 2026)³. A law firm that reported the April 2026 change stressed that removal from Category 2 ‘does not render these bulk drug substances eligible for compounding under section 503A’, and that they stay outside FDA’s interim enforcement discretion policy, ‘which currently extends only to substances in Category 1’¹³. Dihexa was not on the agenda of the Pharmacy Compounding Advisory Committee meeting of 23 July 2026 and 24 July 2026¹³³⁵. FDA says the committee will discuss dihexa acetate, with GHK-Cu, LL-37, melanotan II and PEG-MGF, at a meeting before the end of February 2027, listing them as substances ‘being considered for inclusion on the 503A bulks list’³⁴. Two law firms reported the same plan on 16 April 2026 and 27 July 2026¹³¹⁴. The committee’s advice is not binding, and FDA must still go through formal rulemaking before adding any substance to the list¹³. On 8 October 2026 FDA’s page for the meeting still said its time and place ‘will be scheduled in the coming months’, and a Federal Register search that day found no notice of the meeting, which FDA says it intends to publish³²³⁴.
Other countries: Australia: dihexa is not named in the Poisons Standard in force from 1 October 2026³⁰; how Australian law otherwise treats it was not verified. We found no approval of dihexa as a medicine in any country, which fits FDA’s finding that there are no human exposure data at all³. Its status in the UK, the EU and Canada was not verified.
Sport: Dihexa is not named on WADA’s 2026 Prohibited List, which is in force now, or on the 2027 List, which takes effect 1 January 2027¹⁵¹⁶. In our reading it is banned at all times anyway. Class S2.3 bans hepatocyte growth factor (HGF) and ‘other growth factors or growth factor modulators’ with certain effects on muscle, tendon, blood vessels or regeneration, and FDA’s substance registry lists dihexa as an activator of the HGF receptor¹⁵²¹. If S2.3 did not apply, class S0 would: it bans ‘any pharmacological substance which is not addressed by any of the subsequent sections of the List and with no current approval by any governmental regulatory health authority for human therapeutic use’¹⁵. That is our inference, not a WADA ruling. WADA’s explanatory note on the 2027 List warns that a peptide not named on the List ‘may still be prohibited and may be harmful to health’²⁹.

Dihexa in the legal tracker →

Quality and the gray market

What testing finds

No independent test results exist. As of 7 October 2026, Finnrick, an independent testing service that publishes laboratory results for gray-market peptides, listed no published dihexa tests and no rated vendors¹⁷. Seller claims cannot be checked. One seller’s page guaranteed 99% purity and called dihexa a ‘synthetic heptapeptide’, a chain of seven amino acids, although the molecule contains only two standard amino acids and one non-standard one¹⁸²⁴.

How it is sold

Dihexa is sold online as a ‘nootropic’, mostly as powder or 5 mg capsules labeled for research use only, and also as solutions applied to the skin, often made with DMSO⁵¹⁸. On 7 October 2026 one seller listed 500 mg of powder at $97.99, with a notice that ‘any form of bodily introduction into humans or animals is strictly prohibited by law’¹⁸. The community site cited here for doses promotes that same seller⁵. In 2021 the Alzheimer’s Drug Discovery Foundation reported that dihexa was not commercially available in the US but that some private clinics may administer it⁹. The popular claim that dihexa is ‘seven orders of magnitude’ (10 million times) more powerful than BDNF, a natural brain growth factor, comes from the university’s 2012 news release about tests on nerve cells in a dish². The published paper did not compare the two directly⁹.

Side effects and risks

Unknown human safety. We found no study that gave dihexa itself to people, and FDA says it cannot tell whether dihexa acetate would cause harm if given to humans³.
Possible cancer risk. Dihexa is meant to boost HGF and its receptor MET, a pathway that helps many cancers grow and spread⁹¹⁰. No animal or human study has tested its long-term safety⁹.
Long-lasting, at least in rats. Its half-life in rats was about 12.7 days after a vein injection, and reviewers flag the long half-life alongside the cancer concern¹⁹.
Unreliable evidence. A 2014 paper from its inventors was retracted for falsified or fabricated data, two related papers were retracted too, and the main 2013 paper carries a notice of concern^4192526.
Unknown product content. No independent laboratory had published tests of gray-market dihexa as of 7 October 2026, and one seller’s description of the molecule is wrong¹⁷¹⁸.
Anti-doping. Dihexa is not named on WADA’s Lists, but by our reading it is banned at all times under class S2.3 or S0¹⁵¹⁶.
Legal. US pharmacies have no lawful route under section 503A to compound dihexa¹²¹³.

Questions people ask

Is dihexa FDA approved?

No. No FDA-approved drug contains dihexa, and US pharmacies have no lawful route to compound it¹¹¹²¹³. FDA plans to ask its compounding advisers, at a meeting before the end of February 2027, whether to add it to the list of substances pharmacies may use³⁴.

Has dihexa been tested in humans?

We found no study that gave dihexa itself to people, and FDA found no human exposure data for it³⁶. People in trials of Athira Pharma’s injected drug fosgonimeton were exposed to dihexa, because the blood converts that drug into it²³.

Is dihexa the same as fosgonimeton?

No. Fosgonimeton (ATH-1017) is an inactive prodrug that the blood quickly converts into dihexa, according to its maker²¹²³. Fosgonimeton’s 26-week Alzheimer’s trial did not meet its main goals, and Athira paused the drug²²³³. That result is for an injected prodrug, not for dihexa sold as powder or capsules.

Is dihexa really 10 million times stronger than BDNF?

That figure comes from Washington State University’s 2012 news release about tests on nerve cells in a dish². A 2021 review noted that the published paper did not compare dihexa with BDNF directly⁹, and we found no study testing dihexa itself in people³.

Why were the dihexa studies retracted?

In April 2025 the journal retracted a 2014 paper on how dihexa works, after a Washington State University investigation found falsified or fabricated data⁴. It also retracted two related papers from 2011 and 2012, and the main 2013 dihexa paper carries a notice of concern¹⁹²⁵²⁶.

Can dihexa cause cancer?

Nobody knows. Dihexa is meant to boost the HGF and MET growth pathway, which many cancers use to grow and spread, so experts flag a possible risk⁹¹⁰. A 2021 review by the Alzheimer’s Drug Discovery Foundation found no study of dihexa’s long-term safety in animals or people⁹.

What is still unknown

Whether dihexa does anything in people, for memory or anything else. We found no human study of it.
How much of the inventors’ animal data can be trusted, given one retraction for falsified or fabricated data, two further retractions and a notice of concern on the main 2013 paper.
How long dihexa lasts in people who take it by mouth or on the skin. The 12.68-day figure comes from rats given a vein injection; the 1.5-hour figure comes from people given the injected prodrug fosgonimeton.
What gray-market dihexa products contain. No independent test results were available on 7 October 2026.
When FDA’s compounding advisers will meet on dihexa, and what FDA will then decide. FDA has said only that the meeting will be before the end of February 2027, and had set no date by 8 October 2026.
The dose and route used in the 2024 Huntington’s disease rat study; only its abstract was available to us.
Why the 2011 and 2012 papers were retracted. The PubMed records of those retraction notices carry no text.
Dihexa’s legal status in the UK, the EU and Canada.

Sources

1Evaluation of metabolically stabilized angiotensin IV analogs as procognitive/antidementia agents Journal of Pharmacology and Experimental Therapeutics, Jan 2013. pmc.ncbi.nlm.nih.gov
2Prospective Alzheimer’s drug builds new brain cell connections Washington State University, 11 Oct 2012. archive.news.wsu.edu
3Certain Bulk Drug Substances for Use in Compounding that May Present Significant Safety Risks U.S. Food and Drug Administration, 22 Apr 2026. fda.gov
4Retraction notice to ‘The Procognitive and Synaptogenic Effects of Angiotensin IV-Derived Peptides Are Dependent on Activation of the Hepatocyte Growth Factor/c-Met System’ Journal of Pharmacology and Experimental Therapeutics, 29 Apr 2025. pmc.ncbi.nlm.nih.gov
5Dihexa: How It Works, Mechanism & Evidence (2026) The Peptide Catalog, 11 Jun 2026. thepeptidecatalog.com
6ClinicalTrials.gov search for ‘dihexa’ or ‘PNB-0408’ (no records) ClinicalTrials.gov, U.S. National Library of Medicine, accessed 7 Oct 2026. clinicaltrials.gov
7AngIV-Analog Dihexa Rescues Cognitive Impairment and Recovers Memory in the APP/PS1 Mouse via the PI3K/AKT Signaling Pathway Brain Sciences, 11 Nov 2021. pmc.ncbi.nlm.nih.gov
8Effects of an Angiotensin IV Analog on 3-Nitropropionic Acid-Induced Huntington’s Disease-Like Symptoms in Rats Journal of Huntington’s Disease, 2024. pubmed.ncbi.nlm.nih.gov
9Dihexa (Cognitive Vitality report) Alzheimer’s Drug Discovery Foundation, 13 Aug 2021. alzdiscovery.org
10Known and novel roles of the MET oncogene in cancer: a coherent approach to targeted therapy Nature Reviews Cancer, Jun 2018. pubmed.ncbi.nlm.nih.gov
11Drugs@FDA API search for the active ingredient name ‘dihexa’ (no matches) openFDA, U.S. Food and Drug Administration, accessed 7 Oct 2026. api.fda.gov · data query
12Bulk Drug Substances Nominated for Use in Compounding Under Section 503A of the Federal Food, Drug, and Cosmetic Act (categories list) U.S. Food and Drug Administration, 14 May 2026. fda.gov
13FDA to Remove 12 Popular Peptides from the Category 2 ‘Do Not Compound’ List Frier Levitt, 16 Apr 2026. frierlevitt.com
14Bulk-list bound? PCAC backs majority of peptides in two-day public meeting McDermott Will & Schulte, 27 Jul 2026. mcdermottlaw.com
15World Anti-Doping Code International Standard: Prohibited List 2026 World Anti-Doping Agency, 24 Sep 2025. wada-ama.org
16World Anti-Doping Code International Standard: Prohibited List 2027 World Anti-Doping Agency, 21 Sep 2026. wada-ama.org
17Dihexa: safety testing results and vendor ratings (no published tests) Finnrick, accessed 7 Oct 2026. finnrick.com
18Dihexa - Powder, 500mg (product page) Swiss Chems (vendor), accessed 7 Oct 2026. swisschems.is (seller page, deliberately not linked)
19Retraction notice to ‘Development of Angiotensin IV Analogs as Hepatocyte Growth Factor/Met Modifiers’ Journal of Pharmacology and Experimental Therapeutics, Apr 2025. pubmed.ncbi.nlm.nih.gov
20Hepatocyte Growth Factor/MET Activator Rescues Working Memory Deficits after Repeated Mild Traumatic Brain Injury Neurotrauma Reports, 2025. pmc.ncbi.nlm.nih.gov
21DIHEXA (UNII 9WYX65A5C2): substance record Global Substance Registration System, U.S. Food and Drug Administration, accessed 7 Oct 2026. precision.fda.gov
22Athira Pharma, Inc. annual report (Form 10-K) for the year ended 31 December 2024 Athira Pharma, filed with the U.S. Securities and Exchange Commission, 27 Feb 2025. sec.gov
23Athira Pharma Announces Leadership Changes (press release, exhibit 99.1 to Form 8-K) Athira Pharma, filed with the U.S. Securities and Exchange Commission, 21 Oct 2021. sec.gov
24Dihexa, CID 129010512 PubChem, U.S. National Library of Medicine, accessed 7 Oct 2026. pubchem.ncbi.nlm.nih.gov
25Notice of Concern: McCoy AT, et al. (2013) Evaluation of Metabolically Stabilized Angiotensin IV Analogs as Procognitive/Antidementia Agents Journal of Pharmacology and Experimental Therapeutics, Sep 2021. pubmed.ncbi.nlm.nih.gov
26Retraction notice to ‘Mimics of the Dimerization Domain of Hepatocyte Growth Factor Exhibit Anti-Met and Anticancer Activity’ Journal of Pharmacology and Experimental Therapeutics, Apr 2025. pubmed.ncbi.nlm.nih.gov
27Stem cell, Granulocyte-Colony Stimulating Factor and/or Dihexa to promote limb function recovery in a rat sciatic nerve damage-repair model: Experimental animal studies Annals of Medicine and Surgery, Nov 2021. pubmed.ncbi.nlm.nih.gov
28Safety, Tolerability, Pharmacokinetics, and Pharmacodynamics of the Positive Modulator of HGF/MET, Fosgonimeton, in Healthy Volunteers and Subjects with Alzheimer’s Disease: Randomized, Placebo-Controlled, Double-Blind, Phase I Clinical Trial Journal of Alzheimer’s Disease, 2022. pmc.ncbi.nlm.nih.gov
29Summary of Major Modifications and Explanatory Notes: 2027 Prohibited List World Anti-Doping Agency, 21 Sep 2026. wada-ama.org
30Therapeutic Goods (Poisons Standard—October 2026) Instrument 2026 (F2026L01327) Federal Register of Legislation, Australian Government, 28 Sep 2026. legislation.gov.au
31FDA Adverse Event Reporting System API search for reports naming ‘dihexa’ (no matches) openFDA, U.S. Food and Drug Administration, accessed 7 Oct 2026. api.fda.gov · data query
32Federal Register search: FDA documents mentioning “Pharmacy Compounding Advisory Committee” published since 1 April 2026 Federal Register, accessed 8 Oct 2026. federalregister.gov
33Fosgonimeton in mild-to-moderate Alzheimer’s disease Journal of Alzheimer’s Disease Reports, 12 Dec 2025. pmc.ncbi.nlm.nih.gov
34Meeting of the Pharmacy Compounding Advisory Committee (to be held before the end of February 2027) U.S. Food and Drug Administration, 15 Apr 2026. fda.gov
35July 23-24, 2026: Meeting of the Pharmacy Compounding Advisory Committee U.S. Food and Drug Administration, 6 Aug 2026. fda.gov

Not medical advice. This page reports what studies and regulators say; it does not recommend using any product. How we grade evidence · Report an error.