Dihexa: what the research shows

  • By Reed Calder
  • Last reviewed Nov 11, 2026
  • 21 studies and official documents checked

In short

Dihexa is a lab-made, peptide-like compound built from a fragment of angiotensin IV and studied for memory. As dihexa, it has been tested in rats, mice and zebrafish, not in people. Its injected phosphate prodrug, fosgonimeton, was tested in people, and in the largest trial it did not beat placebo in Alzheimer’s disease. A 2014 paper that showed how dihexa works was retracted in 2025 for falsified data, and the first rat paper carries a notice of concern. It isn’t FDA-approved.

EvidenceCellAnimalHuman

Dihexa is sold online as capsules, injections and creams for memory and focus. It is often called a nootropic peptide. This page covers what it is, what the animal studies found, what the prodrug trials in people showed, why the early evidence is in doubt, and where it stands legally.

Names and forms you’ll see

The names on dihexa products
Name What it is Studied in people?
Dihexa, PNB-0408, N-hexanoic-Tyr-Ile-(6) aminohexanoic amide A lab-made, peptide-like molecule built from a fragment of angiotensin IV. Formula C27H44N4O5, mass 505 (PubChem) No published study found
Dihexa acetate The salt form the FDA names on its compounding safety-risk page (FDA, 2026) No: the FDA has no human exposure data
Fosgonimeton (ATH-1017, NDX-1017) A phosphate prodrug of dihexa: the same molecule with a phosphate group added to its tyrosine (PubChem). A prodrug is a form the body is meant to turn into the active compound Yes, injected, in Alzheimer’s trials
“Dihexa peptide” A sales label for dihexa No published study found

CAS numbers

A CAS number is a substance’s chemical registry number. Some “research” vials list it in place of a name. A different number usually means a different form, such as a salt, or a different substance. The FDA’s registry record for dihexa lists no CAS number, so both numbers below come from PubChem.

Dihexa CAS numbers
Form CAS number Registry record
Dihexa (free base) 1401708-83-5 PubChem
Fosgonimeton (phosphate prodrug of dihexa) 2093305-05-4 PubChem

What dihexa is

Angiotensin IV is a natural fragment of six amino acids that breaks down fast and doesn’t cross into the brain. Researchers at Washington State University modified a short piece of it to make a molecule that lasts longer and passes the blood-brain barrier, and named it dihexa (McCoy et al., 2013).

The idea is that dihexa binds hepatocyte growth factor (HGF) and strengthens its signal through a receptor called c-Met, which helps nerve cells build connections. The 2014 paper that reported this was retracted in 2025 (Benoist et al., 2014; retraction notice). Later papers from the company that developed the prodrug also describe HGF activity (Johnston et al., 2023), but they come from the company itself. So the how-it-works story has no independent paper behind it that we could find.

What has been studied

In people

  • No published study of dihexa in people. ClinicalTrials.gov has no study for it (searched November 11, 2026). The FDA says it has found no human exposure data on drug products containing dihexa acetate (FDA, 2026).
  • The prodrug was tested in people. The company formed to develop dihexa renamed itself Athira Pharma and developed fosgonimeton, an injected prodrug (ADDF, 2021). In a first trial of 88 people (healthy young men, healthy older adults and 11 people with Alzheimer’s), it was safe and well tolerated over single doses of 2 to 90 mg and 9 days of 20 to 80 mg a day in healthy older adults, with 40 mg a day for 9 days in the 11 people with Alzheimer’s (Hua et al., 2022).
  • In the big Alzheimer’s trial it didn’t work. LIFT-AD gave 40 mg a day by injection under the skin, or placebo, for 26 weeks to 287 people with mild to moderate Alzheimer’s who weren’t taking a cholinesterase-inhibitor drug. There was no difference on the combined thinking and daily-function score (change of −0.08, p = 0.70), and none on either part alone (Porsteinsson et al., 2025). A follow-on open-label study was stopped after that result (ClinicalTrials.gov).
  • A smaller trial ended early. A trial in people with Parkinson’s disease dementia or dementia with Lewy bodies enrolled 28 and ended early because of “study design limitations,” the registry says (ClinicalTrials.gov).
  • Not capsules, creams or sprays. The prodrug trials used injections in Alzheimer’s patients, and the company funded them. No study we found tested dihexa as a capsule, cream or nasal spray in people.

In animals

  • Rats, 2013 (the first paper). Dihexa reversed a drug-induced loss of maze learning when given into the brain (0.1 or 1 nanomole), by injection into the belly (0.25 and 0.5 mg/kg worked, 0.05 did not) or by mouth (2 mg/kg fully reversed it, 1.25 mg/kg partly). In 24-month-old rats, 2 mg/kg by mouth daily improved maze learning, with 6 rats in the group (McCoy et al., 2013). Two authors are founders and shareholders of M3 Biotechnology, the company formed to develop dihexa. In September 2021 the journal added a notice of concern to this paper (notice of concern). Retraction Watch reported that the notice cites possible image manipulation (Retraction Watch, 2021). We found no later notice that settles it.
  • Rats, 2014 (the mechanism paper). Reported that dihexa binds HGF. Retracted in 2025: the journal’s notice says a Washington State University investigation found falsified or fabricated data in two figures and in a later correction, and held two of the authors solely responsible (Benoist et al., 2014; retraction notice).
  • Mice with Alzheimer’s-like disease. Over 3 months at 1.44 or 2.88 mg/kg a day, treated mice found the hidden platform faster and had more nerve cells and less brain inflammation than untreated mice. There were 6 mice per group, and the methods describe the route both as by mouth and as injection (Sun et al., 2021).
  • Rats, a negative result. In rats given a toxin that mimics Huntington’s disease, dihexa (named PNB-0408 in the paper) did not protect against the memory, movement or weight problems (Wells et al., 2024).
  • Other uses. After a cut leg nerve in rats, stem cells combined with dihexa (1 mg per rat, by injection into the belly or leg muscle on days 0, 4, 7 and 10) improved walking scores (Weiss et al., 2021). In zebrafish larvae, dihexa protected ear hair cells from antibiotic damage (Uribe et al., 2015).
  • Overview. A review of 32 animal studies of angiotensin IV and related compounds, dihexa among them, found that 8 of 9 studies in memory-deficit models saw benefits, mostly with short-term dosing and best when given into the brain (Ho et al., 2018). The Alzheimer’s Drug Discovery Foundation’s 2021 report adds that dihexa did not improve memory in rats with normal memory (ADDF, 2021).

EvidenceCellAnimalHuman

The usual claims, checked

Common dihexa claims vs the studies
Claim Verdict What the studies found
“Seven orders of magnitude more potent than BDNF” Untested The figure comes from a 2012 university news release about tests on living nerve cells in a dish (news release, 2012). We found no BDNF comparison in the paper text (McCoy et al., 2013). It is not a dose comparison in animals or people
Orally bioavailable Untested The paper reports blood levels only after injections, and a computer model predicted oral absorption. Oral doses did improve rat maze scores (McCoy et al., 2013). The original group calls it orally active (Wright and Harding, 2015), but we found no measured oral absorption in any species
Crosses the blood-brain barrier Untested In rats, radioactive dihexa given into the blood was found in several brain regions 30 minutes later (McCoy et al., 2013). Nobody has shown it in people
Lasts a long time Untested In rats the half-life was about 12.7 days by vein and 8.8 days by injection into the belly (McCoy et al., 2013). There is no human number
Improves memory or treats Alzheimer’s Contradicted Better maze scores in rats and mice with induced memory loss (Sun et al., 2021). No benefit in a Huntington’s-like rat model (Wells et al., 2024). The prodrug didn’t beat placebo in 287 people (Porsteinsson et al., 2025)
Boosts focus and memory in healthy people Untested No study in healthy people. In rats with normal memory it didn’t improve memory (ADDF, 2021)
Helps depression, heart health or Parkinson’s disease Untested We found no study of dihexa for depression or heart health. For Parkinson’s disease there are review articles by the original group, and the prodrug trial in dementia with Parkinson’s or Lewy bodies ended early (Wright et al., 2015; ClinicalTrials.gov)
Has no toxicity Untested The line comes from a patent, not a published study. No one has studied long-term safety, including tumor effects, in animals or people (ADDF, 2021)

Verdict key. Supported: controlled data in people, or the official record, back it. Mixed: studies disagree, or the support is small or limited to one use. Untested: we found no study in people. Contradicted: the studies or record we found show no effect or the opposite.

What the studies used

Methods of the main dihexa and prodrug studies
Study Species How given Length
McCoy 2013 Rats, including 24-month-old rats Into the brain (0.1 or 1 nanomole), injection into the belly (0.05 to 0.5 mg/kg) or by mouth (1.25 or 2 mg/kg) Hours to 8 days of maze testing
Sun 2021 Mice with Alzheimer’s-like disease Daily, 1.44 or 2.88 mg/kg; route given as both by mouth and injection 3 months
Weiss 2021 Rats after nerve repair 1 mg per rat, by injection into the belly or leg muscle on days 0, 4, 7 and 10, with stem cells 16 weeks
Uribe 2015 Zebrafish larvae In the water Hours
Hua 2022 (prodrug) People Daily injection under the skin Single doses of 2 to 90 mg, or 20 to 80 mg a day for 9 days
LIFT-AD (prodrug) People with Alzheimer’s Daily injection under the skin, 40 mg 26 weeks

These are the studies’ methods, not a dose for anyone to use. Animal doses don’t convert to people, and the prodrug trial doses were for an injected prodrug, not for capsules or creams.

Side effects and safety

  • There are no human safety data for dihexa. The FDA says it lacks important information on whether it would cause harm in people (FDA, 2026).
  • What the prodrug trials show. In the first trial the prodrug was safe and well tolerated (Hua et al., 2022). In LIFT-AD, serious side effects were similar to placebo (4.2% vs 6.9%). More people on the drug stopped because of side effects (14.2% vs 4.6%), mostly from injection-site reactions (Porsteinsson et al., 2025). The company paid for both trials and its employees were authors.
  • The cancer question. HGF and c-Met also drive cell growth. The Alzheimer’s Drug Discovery Foundation says there are theoretical concerns that c-Met activation could promote tumors, and that no study has examined long-term safety, including tumor effects (ADDF, 2021).
  • It stays a long time. A half-life of about 12.7 days in rats means effects, and side effects, could linger. Nobody has measured it in people (McCoy et al., 2013).
  • Doubt about the early data. A retraction for falsified data and a notice of concern don’t make every result wrong, but the early dihexa results need confirming by others. We found few other animal studies, and one was negative.
  • Product quality. We found no independent test of capsules, creams or sprays sold as dihexa, and no study of those forms.
  • Unstudied. Long-term use, use with other medicines, use in pregnancy and use in children.

Side effects people report from online products aren’t covered here; they come from users, not studies.

Legal and approval status

  • Not FDA-approved. We found no approved drug that contains dihexa or fosgonimeton (Drugs@FDA; searched November 11, 2026).
  • On the FDA’s compounding safety-risk page. The FDA lists dihexa acetate among bulk substances that may present significant safety risks if compounded, and says it has no human exposure data (FDA, 2026).
  • No registered trial of dihexa itself. The prodrug was tested in several registered trials (dihexa search; fosgonimeton search).
  • Not in the October 2026 HHS policy. That temporary policy covers six other peptides (Reuters, 2026).
  • Sport. The World Anti-Doping Agency’s 2026 list doesn’t name it, but its S0 class prohibits any substance with no government approval for human therapeutic use, including drugs in development (WADA, 2026).
  • “Research use only” products aren’t approved for use in people.

What to look for if you research it

We don’t name or link sellers. These checks come from the registries:

  • The right name and number. Dihexa is CAS 1401708-83-5, formula C27H44N4O5, mass 504.7 (PubChem). The acetate salt is a different substance with a different mass.
  • Not the prodrug. Fosgonimeton has a phosphate group and a mass of 584.6 (PubChem). Results for it don’t apply to dihexa, and the reverse.
  • The form you are asked to take. The animal studies used injection, gavage or a bath, and the prodrug trials used injection. We found no study of a cream or nasal spray.
  • A report that names a method, such as mass spectrometry, and gives the amount per capsule or vial.

Our guide to what to look for explains what each test on a lab report proves and which routes are legal; how to read a peptide study explains how to judge the studies sellers cite.

What’s still unknown

  • Whether dihexa does anything for memory in people.
  • Whether any outside lab gets the same results as the original lab.
  • How much of an oral dose reaches the blood in any species.
  • Whether a capsule, cream or spray works at all.
  • Long-term safety, including tumor risk.
  • What is really in the products sold online.

Retracted and flagged studies

A retracted study has been pulled by its journal, usually for serious errors or misconduct, and its results shouldn’t be relied on. A withdrawn paper was pulled, often before final publication. An expression of concern means the journal has flagged a problem it is still looking into.

We check every study this page cites against PubMed and the Retraction Watch database, and search both for flagged studies that mention Dihexa. We recheck every month. Last checked Nov 11, 2026.

Flagged studies that mention Dihexa

Retracted, withdrawn and flagged studies
Study What it was about Status
Benoist et al., 2014 The procognitive and synaptogenic effects of angiotensin IV-derived peptides are dependent on activation of the hepatocyte growth factor/c-met system. Retracted (2025) · notice
We cite it above, with this warning.
McCoy et al., 2013 Evaluation of metabolically stabilized angiotensin IV analogs as procognitive/antidementia agents. Expression of concern (2021) · notice
We cite it above, with this warning.

Studies this page relies on

Status of every study cited on this page
Study Published in Status
Wright et al., 2015 Prog Neurobiol No retraction or concern
Wright and Harding, 2015 J Alzheimers Dis No retraction or concern
Uribe et al., 2015 Front Cell Neurosci No retraction or concern
Ho and Nation, 2018 Neurosci Biobehav Rev No retraction or concern
Weiss et al., 2021 Ann Med Surg (Lond) No retraction or concern
Sun et al., 2021 Brain Sci No retraction or concern
Hua et al., 2022 J Alzheimers Dis No retraction or concern
Johnston et al., 2023 Neurotherapeutics No retraction or concern
Wells et al., 2024 J Huntingtons Dis No retraction or concern
Porsteinsson et al., 2025 J Alzheimers Dis Rep No retraction or concern

Bring your questions to other researchers

Our forum opens once 100 researchers join the waitlist. Members post studies, ask what they actually found, and answer with sources.

Questions

What is dihexa?

A lab-made, peptide-like molecule built from a fragment of angiotensin IV and studied for memory in animals (McCoy et al., 2013). It isn’t approved for any use.

Is dihexa a peptide?

It is peptide-like. It is a very short chain: two amino acids (tyrosine and isoleucine) with a six-carbon cap on one end and a six-carbon tail on the other, not a long peptide chain (PubChem).

What does dihexa do?

In rats and mice with induced memory loss, it improved maze scores (McCoy et al., 2013; Sun et al., 2021). The idea that it works through HGF comes from a paper that was retracted (retraction notice). Nobody has shown it works in people.

Has dihexa been tested in humans?

Dihexa itself, no. Its injected prodrug, fosgonimeton, was tested in people with Alzheimer’s and did not beat placebo in the largest trial (Porsteinsson et al., 2025).

Is dihexa safe?

We don’t know. The FDA lacks human data on dihexa acetate (FDA, 2026). The prodrug was generally well tolerated, with more stops because of injection-site reactions (Porsteinsson et al., 2025). Long-term and tumor risks are unstudied (ADDF, 2021).

Is dihexa a stimulant?

We found no study that tested it as a stimulant. The papers describe a different idea, strengthening growth-factor signaling, and that idea is in doubt after a retraction (retraction notice).

Was a dihexa paper retracted?

Yes. In 2025 the journal retracted a 2014 paper that reported how dihexa works. Its notice says a university investigation found falsified or fabricated data (retraction notice). The first 2013 rat paper has a notice of concern (notice of concern).

Is dihexa orally bioavailable?

Not shown. A computer model predicted it, and oral doses improved rat maze scores, but the paper reports blood levels only after injections (McCoy et al., 2013).

What is the half-life of dihexa?

In rats about 12.7 days by vein and 8.8 days by injection into the belly (McCoy et al., 2013). No human value exists.

Does dihexa help Parkinson’s disease or anxiety?

We found no study of dihexa for anxiety. For Parkinson’s there are reviews by the original group, and the prodrug trial in Parkinson’s or Lewy body dementia ended early (Wright et al., 2015; ClinicalTrials.gov).

What is the dihexa dosage?

No dose comes from a study of dihexa in people. In rats, 0.25 to 2 mg/kg, injected into the belly or by mouth, reversed a drug-induced memory deficit (McCoy et al., 2013), and mice got 1.44 or 2.88 mg/kg a day for 3 months (Sun et al., 2021). The prodrug trials gave 40 mg a day by injection (Hua et al., 2022; Porsteinsson et al., 2025). Animal doses don’t convert to people. Dose charts online aren’t evidence.

Is dihexa legal?

It isn’t FDA-approved for any use, and the FDA lists dihexa acetate on its compounding safety-risk page (FDA, 2026). Our guide to what to look for covers the routes people use.

Where can I buy dihexa?

Sellers offer it, but it has no approved use and we found no independent test of the products. We don’t link sellers; see what to look for.

Sources

  1. McCoy AT, et al. Evaluation of metabolically stabilized angiotensin IV analogs as procognitive/antidementia agents. J Pharmacol Exp Ther. 2013;344(1):141-54. PubMed 23055539. Notice of concern, J Pharmacol Exp Ther, 2021;378:313: PubMed 34551989
  2. Benoist CC, et al. The procognitive and synaptogenic effects of angiotensin IV-derived peptides are dependent on activation of the hepatocyte growth factor/c-Met system. J Pharmacol Exp Ther. 2014;351(2):390-402 (retracted). PubMed 25187433. Retraction notice, J Pharmacol Exp Ther, 2025;392:103567: PubMed 40312093 (free text)
  3. Sun X, et al. AngIV-analog dihexa rescues cognitive impairment and recovers memory in the APP/PS1 mouse via the PI3K/AKT signaling pathway. Brain Sci. 2021;11. PubMed 34827486
  4. Wells RG, et al. Effects of an angiotensin IV analog on 3-nitropropionic acid-induced Huntington’s disease-like symptoms in rats. J Huntingtons Dis. 2024;13:55-66. PubMed 38489193
  5. Weiss JB, et al. Stem cell, granulocyte-colony stimulating factor and/or dihexa to promote limb function recovery in a rat sciatic nerve damage-repair model. Ann Med Surg. 2021;71:102917. PubMed 34703584
  6. Uribe PM, et al. Hepatocyte growth factor mimetic protects lateral line hair cells from aminoglycoside exposure. Front Cell Neurosci. 2015;9:3. PubMed 25674052
  7. Ho JK, Nation DA. Cognitive benefits of angiotensin IV and angiotensin-(1-7): a systematic review of experimental studies. Neurosci Biobehav Rev. 2018;92:209-25. PubMed 29733881
  8. Wright JW, Harding JW. The brain hepatocyte growth factor/c-Met receptor system: a new target for the treatment of Alzheimer’s disease. J Alzheimers Dis. 2015;45:985-1000. PubMed 25649658
  9. Wright JW, Kawas LH, Harding JW. The development of small molecule angiotensin IV analogs to treat Alzheimer's and Parkinson's diseases. Prog Neurobiol. 2015;125:26-46. PubMed 25455861
  10. Hua X, et al. Safety, tolerability, pharmacokinetics, and pharmacodynamics of the positive modulator of HGF/MET, fosgonimeton, in healthy volunteers and subjects with Alzheimer's disease. J Alzheimers Dis. 2022;86:1399-1413. PubMed 35180125
  11. Porsteinsson AP, et al. Fosgonimeton in mild-to-moderate Alzheimer's disease. J Alzheimers Dis Rep. 2025;9. PubMed 41393340
  12. Johnston JL, et al. Fosgonimeton, a novel positive modulator of the HGF/MET system, promotes neurotrophic and procognitive effects in models of dementia. Neurotherapeutics. 2023;20:431-51. PubMed 36538176
  13. Alzheimer’s Drug Discovery Foundation. Cognitive Vitality report: dihexa. Last updated August 13, 2021. ADDF report
  14. Retraction Watch. Four papers by Athira CEO earn expressions of concern. September 24, 2021. Retraction Watch
  15. Washington State University news release. Prospective Alzheimer’s drug builds new brain cell connections, improves cognitive function of rats. October 11, 2012. Release via ScienceDaily
  16. National Center for Biotechnology Information. PubChem compound records CID 129010512 (dihexa) and CID 156596375 (fosgonimeton). PubChem
  17. ClinicalTrials.gov search for dihexa and PNB-0408, November 11, 2026: no registered studies. Search
  18. ClinicalTrials.gov search for fosgonimeton (ATH-1017), November 11, 2026: trials NCT03298672, NCT04491006, NCT04488419, NCT04886063, NCT04831281 and NCT05511558. Search
  19. U.S. Food and Drug Administration. Drugs@FDA database; no product with dihexa or fosgonimeton as an active ingredient, searched November 11, 2026. Drugs@FDA
  20. U.S. Food and Drug Administration. Safety risks associated with certain bulk drug substances nominated for use in compounding. Content current as of April 22, 2026. FDA page
  21. World Anti-Doping Agency. 2026 Prohibited List (S0, non-approved substances). WADA document
  22. Reuters. U.S. health department to allow temporary sale of some unapproved peptides. October 28, 2026. Article
  23. National Library of Medicine. PubChem compound record: Dihexa (free base) (CID 129010512). PubChem
  24. National Library of Medicine. PubChem compound record: Fosgonimeton (phosphate prodrug of dihexa) (CID 156596375). PubChem

Last reviewed Nov 11, 2026. Approval and compounding status change often; we recheck this page when they do. Nothing here is medical advice; see our medical disclaimer. Found a mistake? Email editors@peptideresearchstudy.com.