Dihexa: The Cognitive Peptide Evidence Reviewed (2026)
By Theo Park · Editor, Privacy & Safety
Updated Jun 2026Dihexa is a synthetic peptide that gets sold online as the most potent "brain-repair" compound ever made, with claims that it builds new connections between neurons. The real picture is far messier: the animal data is thin, the single most important mechanistic study was retracted in 2025 for fabricated data, and the drug built on the same idea just failed a large Alzheimer's trial. This review walks through what dihexa actually is, what the evidence does and does not show, and why the honest grade here is "interesting in theory, unproven in practice."
Dihexa is a synthetic peptide that gets sold online as the most potent "brain-repair" compound ever made, with claims that it builds new connections between neurons. The real picture is far messier: the animal data is thin, the single most important mechanistic study was retracted in 2025 for fabricated data, and the drug built on the same idea just failed a large Alzheimer's trial. This review walks through what dihexa actually is, what the evidence does and does not show, and why the honest grade here is "interesting in theory, unproven in practice."
What Dihexa Is
Dihexa (also written as N-hexanoic-Tyr-Ile-(6) aminohexanoic amide, development code PNB-0408, CAS 1401708-83-5) is a small synthetic peptide. It was designed in a lab at Washington State University by Joseph Harding and John Wright in the late 2000s.
The starting point was angiotensin IV. That's a natural fragment your body makes when it breaks down angiotensin, a hormone best known for controlling blood pressure. Researchers noticed that angiotensin IV could improve memory in animals. The problem: it breaks down fast in the body and does not cross from the blood into the brain. So the Harding lab chemically modified it. They added a fatty chain on one end and a different cap on the other. The result, dihexa, was meant to survive longer, slip across the blood-brain barrier, and be taken by mouth.
On paper, dihexa is not a hormone, not a steroid, and not a growth-hormone peptide. It is a designed molecule meant to switch on a growth signal in the brain.
Where it came from
To understand dihexa you have to understand the renin-angiotensin system. Most people know it as the blood-pressure system, the one that ACE inhibitor drugs target. But the brain has its own local version of this system, and one of its byproducts, angiotensin IV, turned out to do something unexpected in animal experiments: it sharpened memory and learning. Researchers found a binding site in the brain (sometimes called the AT4 site) that angiotensin IV docked onto, and activity there tracked with better performance on memory tasks.
That discovery set off a search for a "drug-like" version. Natural angiotensin IV is a short peptide that enzymes chop up within minutes, and it can't get into the brain when given by mouth or injection into the bloodstream. So the chemistry work focused on three goals: make it last longer in the body, make it cross the blood-brain barrier, and make it work when swallowed. Dihexa was the molecule that, in the developers' hands, hit all three. The fatty hexanoic chain on one end is what gives the molecule its grease-loving character, which is thought to help it slip through the fatty membrane that walls off the brain.
This origin story matters because it explains both the appeal and the limits. The appeal: dihexa came out of legitimate university neuroscience, not a supplement marketing department. The limit: that same university later had to investigate the lab's data, and the conclusions were damning.
The proposed mechanism
The leading theory is that dihexa works through the hepatocyte growth factor (HGF) system. HGF is a natural protein. It binds to a receptor called c-Met (also written MET). When HGF locks onto c-Met, it triggers signals inside the cell, mainly two pathways called PI3K/AKT and MAPK, that help cells survive, grow, and form new connections.
In the brain, that signaling is linked to synaptogenesis, the building of new synapses between neurons. The pitch is that dihexa amplifies HGF's effect at c-Met, so neurons sprout more connections and memory improves.
That mechanism is plausible. But "plausible" is not the same as "proven in humans," and as you'll see below, the single study that nailed down the HGF/c-Met link for dihexa no longer stands.
It's worth being precise about what "synaptogenesis" would and wouldn't mean. Building more synapses is not the same as making someone smarter, healing a specific injury, or reversing a disease. A brain that sprouts connections indiscriminately is not obviously a better brain. Useful cognition depends on the right connections forming in the right places at the right time, pruned and shaped by experience. A drug that globally cranks up connection-building is a blunt instrument. So even if dihexa does what the theory says at the molecular level, translating "more synapses in a dish" into "better thinking in a person" is a long, unproven leap. The field has watched many compounds clear that first hurdle and then do nothing measurable for patients.
The Potency Claim, Examined
The most repeated marketing line is that dihexa is "seven orders of magnitude more potent than BDNF." BDNF (brain-derived neurotrophic factor) is the body's own famous brain-growth protein. Seven orders of magnitude means ten million times.
That number traces back to a lab dish assay measuring how much new neuronal connection formed at a given concentration. Two cautions matter here. First, potency in a dish does not equal benefit in a living brain, and it certainly does not equal benefit in a person. A compound can be wildly "potent" at flipping a molecular switch and still do nothing useful, or something harmful, in a whole organism. Second, the early figures behind dihexa's potency and mechanism came from a research group whose work later drew formal integrity concerns. So treat the ten-million-times headline as a lab curiosity, not a clinical fact.
What the Evidence Actually Shows
Here is the part that matters most. Dihexa has never been tested in a published human trial. Every efficacy claim comes from rodents, and the rodent record itself has serious cracks.
| Evidence type | What exists | Honest grade |
|---|---|---|
| Human clinical trials (dihexa itself) | None published | No evidence |
| Animal cognition studies (aged/impaired rats) | A handful, mostly from one lab | Weak, integrity-flagged |
| Key HGF/c-Met mechanism study | Published 2014, RETRACTED 2025 | Withdrawn |
| Foundational characterization study (2013) | Published, carries a 2021 Notice of Concern | Compromised |
| Related clinical drug (fosgonimeton) | Phase 2/3 trial completed | Failed primary endpoint |
| Long-term human safety data | None | Unknown |
The animal data
The early animal work suggested oral dihexa could improve performance in memory tasks. In the foundational 2013 paper, aged rats given oral dihexa reportedly did better in the Morris water maze (a standard rodent memory test), and the compound was said to reverse a drug-induced memory deficit in younger rats (McCoy et al., 2013, J Pharmacol Exp Ther, PMID 23055539). Earlier work from the same group reported that related angiotensin IV analogs boosted hippocampal synapse formation and spatial memory (Benoist et al., 2011, PMID 21719467).
Two problems sink the confidence you can put on this. The 2013 foundational paper now carries a formal Notice of Concern issued by the journal in 2021 (Notice of Concern, PMID 34551989). And the body of work came overwhelmingly from a single lab, never independently replicated at scale. In science, results that only ever appear from one source, and then get flagged for integrity, are exactly the results you should not bank on.
The retracted mechanism study
The 2014 paper titled "The Procognitive and Synaptogenic Effects of Angiotensin IV-Derived Peptides Are Dependent on Activation of the Hepatocyte Growth Factor/c-Met System" was the keystone. It was the study that supposedly proved dihexa's cognitive effects ran specifically through HGF/c-Met (original article, PMID 25187433).
In April 2025 the journal retracted it (Retraction Notice, PMID 40312093). A Washington State University investigation found that figures in the paper contained falsified and/or fabricated data, including reused images presented as different experiments and digitally altered measurements. Two of the authors were found responsible.
Read that again. The single most-cited evidence that dihexa works the way vendors say it works was withdrawn for fabricated data. That does not automatically mean the HGF/c-Met theory is wrong. It means the strongest "proof" for it is gone, and dihexa's mechanism is now back to being a reasonable hypothesis rather than a demonstrated fact.
The clinical-drug failure
Athira Pharma (formerly M3 Biotechnology) was founded to turn this science into a real drug. The company decided dihexa itself was not drug-like enough, so it developed fosgonimeton (ATH-1017), a prodrug that converts in the body to an active form closely related to dihexa and works through the same HGF/MET system (Johnston et al., 2023, Neurotherapeutics, PMID 36538176).
Fosgonimeton went all the way into a large Phase 2/3 Alzheimer's trial called LIFT-AD (NCT04488419, ClinicalTrials.gov). In September 2024 the company reported that the trial missed its primary endpoint, and the key secondary measures did not reach statistical significance either (Practical Neurology, 2024).
This is the closest thing to a real-world test of the dihexa idea, and it did not pan out for Alzheimer's. The drug was generally safe and well tolerated, but it did not clearly help patients think better. That is sobering context for anyone buying raw dihexa expecting brain repair.
It's fair to note a few caveats in the drug's favor, because honest grading cuts both ways. Some smaller subgroup signals (for example, in patients carrying the APOE4 Alzheimer's risk gene) trended in the drug's direction without reaching statistical significance. Trends like that can be real or can be noise; you don't get to count them as wins until a properly designed study confirms them, and none has. Also, fosgonimeton is a different molecule from dihexa, given by injection as a prodrug, so a failure for fosgonimeton is not a one-to-one verdict on raw dihexa. But it is the same biological hypothesis, taken further into rigorous human testing than dihexa ever was, and it came up empty for its main goal. When the best-funded, most carefully run version of an idea fails, that should lower your expectations for the unregulated powder version, not raise them.
Why no dihexa human trial ever happened
A reasonable question: if dihexa was supposedly ten million times stronger than BDNF, why did the company switch to a different molecule and never run a human dihexa trial? The answer the developers gave was that dihexa lacked the "drug-like" properties needed for a clean clinical program, things like consistent solubility and predictable pharmacokinetics. In plain terms, dihexa was hard to formulate and dose reliably as a medicine. That's a quiet but important admission. The compound now sold freely online as a finished "nootropic" is the same one a funded biotech decided was not good enough to put into people. The version that did go into people was re-engineered, and it failed.
Dihexa vs. Other Cognitive Peptides
People shopping for dihexa usually compare it to other "nootropic" peptides. Here's how the evidence stacks up. Note that "human data" here means published, peer-reviewed human studies, not anecdotes.
| Compound | Proposed mechanism | Human data | Evidence strength |
|---|---|---|---|
| Dihexa | Amplifies HGF/c-Met signaling, synaptogenesis | None | Very weak (key study retracted) |
| Semax | Modulates BDNF, dopamine; ACTH fragment | Some, mostly Russian | Limited |
| Selank | Anxiolytic, GABA/serotonin effects | Some, mostly Russian | Limited |
| Vasopressin (cognitive use) | Memory consolidation via V1a receptors | Mixed, older | Mixed/weak |
| FGL peptide | NCAM mimetic, synaptic plasticity | None human | Preclinical only |
The honest takeaway across this whole category: none of these are proven cognitive enhancers in healthy people. Dihexa sits at the weaker end because its mechanism study was retracted and its clinical-drug cousin failed. For deeper looks at the better-studied options, see our reviews of Semax vs Selank, N-Acetyl Semax Amidate, vasopressin's cognitive research, and the FGL peptide cognitive research.
Safety: What's Known and What Isn't
There is no published human safety data for dihexa specifically. Everything below is either animal data, theory, or extrapolated from the related drug.
The c-Met cancer question
This is the safety issue worth understanding. The c-Met receptor that dihexa is designed to activate is the same receptor that, when overactive, drives many cancers. c-Met signaling is a known cancer-growth pathway. A compound built to boost that pathway across the body raises an obvious theoretical concern about feeding tumor growth.
In fairness, patent documents from the developers reported that short-term animal safety studies did not show tumor formation. But "short-term" and "in rodents" are doing heavy lifting in that sentence. Nobody has studied what happens when a person takes dihexa for months or years. The cancer question is unresolved, not cleared.
Other unknowns
Because dihexa amplifies a growth-and-survival signal, other theoretical concerns include effects on existing abnormal cell growth and unknown interactions with blood-vessel formation. None of this has been characterized in humans. Reported side effects in the online community (headaches, irritability, brain fog the next day) are anecdotal and unverified.
Legal and quality status
Dihexa is not an FDA-approved drug and is not a dietary supplement. It is sold as a "research chemical," which means no regulatory oversight of purity, dose accuracy, or what's actually in the vial. Counterfeit and contaminated research peptides are a documented problem (FDA on counterfeit medicine). If you don't know what's in the vial, you can't reason about safety at all. For how vendor quality is verified, see our peptide vendor quality standards guide and the broader peptide legality map.
The dosing problem
You'll see dihexa "protocols" online quoting specific milligram amounts and routes (oral, transdermal in a cream, sometimes intranasal). Treat those numbers with deep skepticism. There is no human dose-finding study, so any dose figure circulating online was not derived from controlled human research. It was either scaled down from animal experiments using rough conversions or simply passed around the community until it sounded official. Animal-to-human dose conversion is not a clean multiplication; metabolism, absorption, and target sensitivity all differ between species. On top of that, dihexa is notoriously hard to dissolve, which means homemade preparations may not deliver the dose a person thinks they're taking. A confident-looking protocol is not the same as a validated one. In this case, none of them are validated.
Who Dihexa Is (and Isn't) For
Let's be blunt about this, because the marketing isn't.
Dihexa is not for anyone looking for a proven cognitive enhancer. The evidence does not support that claim. It is not for people with a personal or family history of cancer, given the c-Met concern. It is not for anyone who wants a regulated, quality-controlled product, because none exists.
Who might reasonably be interested: people who understand they would be self-experimenting with an unapproved research chemical that has no human data, a retracted key study, and a failed clinical cousin, and who accept those risks with full information. That is a small and well-informed group. For most people asking "will this make me smarter or fix my brain," the honest answer based on current evidence is: there's no good reason to expect it will, and real reasons to be cautious.
If your actual goal is cognitive health, the boring, evidence-backed levers (sleep, exercise, treating high blood pressure, hearing loss correction, social engagement) have far stronger support than any peptide on this list. These aren't glamorous, and nobody sells them in a vial, but the human evidence behind them dwarfs anything dihexa has produced. That contrast is the whole point. The reason boring interventions win is that they've actually been tested in people, repeatedly, and held up.
A note on the "brain repair" framing
The word "repair" does a lot of selling in dihexa marketing. It implies the drug fixes damage, regrows lost tissue, or reverses decline. Nothing in the published record supports that framing for dihexa in humans. Even in the animal work, "improvement on a memory test" is not the same as "repaired brain." And the one human-stage program that tested this exact biology in people with actual brain disease, Alzheimer's, did not show meaningful repair or slowing. When a product's central promise is the one thing the evidence most clearly fails to support, that's not a small marketing exaggeration. It's the core claim being unproven.
The Bottom Line
Dihexa is a genuinely clever piece of chemistry built on a real biological idea: tune up the brain's growth signaling and you might rebuild lost connections. The idea attracted serious money and a real biotech company. But the chain of evidence supporting dihexa has been breaking link by link. The key mechanism study was retracted for fabricated data in 2025. The foundational characterization paper carries a Notice of Concern. The clinical drug built on the same mechanism failed its big Alzheimer's trial. And there has never been a single published human study of dihexa itself.
Grade: weak-to-no evidence of benefit, real unanswered safety questions, zero regulatory standing. Interesting science. Not a proven product.
Frequently Asked Questions
Is dihexa FDA approved?
No. Dihexa is not approved by the FDA for any use, and it is not a dietary supplement. It is sold only as a "research chemical," which means there is no oversight of its purity, dosing, or labeling. A related prodrug, fosgonimeton, was studied in clinical trials but failed its main Alzheimer's endpoint and is also not approved.
Does dihexa actually work for memory or brain repair?
There is no published human evidence that dihexa improves memory or repairs the brain. The animal data is limited, comes mostly from one lab, and is now clouded by a retracted mechanism study and a Notice of Concern on the foundational paper. The most rigorous real-world test of the same biology, the fosgonimeton LIFT-AD trial, did not show a clear cognitive benefit.
Why was the dihexa research retracted?
In April 2025, the journal retracted the 2014 study that linked dihexa's effects to the HGF/c-Met system. A Washington State University investigation found the paper contained falsified or fabricated data, including reused and digitally altered images. The retraction removed the strongest published support for how dihexa was claimed to work.
Is dihexa safe?
There is no human safety data for dihexa. The biggest theoretical concern is that it activates the c-Met receptor, a pathway involved in cancer growth. Short-term animal studies reportedly did not show tumors, but no long-term or human safety studies exist. Because it is an unregulated research chemical, product purity and contamination are also real risks.
How does dihexa compare to other nootropic peptides like semax?
All of these peptides are short on solid human evidence, but dihexa sits at the weaker end. Semax and selank at least have some published (mostly Russian) human studies, while dihexa has none and carries a retracted key paper. None of them are proven to enhance cognition in healthy people.
This article is for educational purposes only and is not medical advice. Dihexa is an unapproved research chemical with no human safety or efficacy data. Talk to a qualified healthcare provider before using any peptide or making health decisions.
On Google
Get our answers in your Google results.
Add Peptide Front as a preferred source and Google will surface our peptide research more often — in Top Stories and AI answers, marked with a preferred badge. One tap, free, undo anytime.
Add us as a preferred sourceOpens Google's source preferences for peptidefront.com. No sign-up with us — it's a Google setting.