Dihexa
Angiotensin IV analog
- Injection
- Oral
- Nasal spray
- Suppository
- Gel or cream
Oral — no established human dose.
No provider we've verified carries Dihexa yet
Dihexa is a small synthetic peptidomimetic — about 504 daltons, two amino-acid-like residues bracketed by fatty-acid caps — designed at Washington State University in the early 2010s. The starting point was angiotensin IV, a fragment of the same hormone system every blood-pressure drug in the cardiology formulary works on. Angiotensin IV is procognitive in rodents but useless as a drug: it cannot cross into the brain, and plasma peptidases destroy it in seconds. Dihexa was built to fix both problems, and in rats it did — orally bioavailable, brain-permeable.
The proposed mechanism is what made it famous, because it is not the usual one. Most peptides sold for cognition modulate a system that already exists — nudge BDNF, lean on a neurotransmitter, provide trophic support. Dihexa was claimed to build new connections: bind hepatocyte growth factor, dimerise with the body’s own HGF, and potentiate the c-Met receptor at picomolar concentrations, driving the cascades that assemble new excitatory synapses. In hippocampal culture the reported effect was roughly a doubling of functional synapse number.
Two of the papers that established that mechanism were retracted in 2025. The compound is still sold on the strength of them.
That is the fact this page exists to state plainly. Kawas 2012 and Benoist 2014 were formally withdrawn after Notices of Concern in 2021; McCoy 2013, the memory and dendritic-spine paper, carries a Notice of Concern and stands. The independent neuroscience putting c-Met at excitatory synapses is untouched — it is the biochemistry tying dihexa to it that is gone. One independent replication survives: a 2021 study in Alzheimer’s mice that saw an effect and traced it to a PI3K-dependent pathway without re-establishing the binding. For the paper-by-paper record, read the deep-dive on PeptideWellness.
Who it's for (or who should avoid)
What it's aimed at
There is no patient group, because there is no human evidence. What the rodent work suggests, and where:
- Cognitively impaired animals — dihexa improved Morris water maze performance in aged rats with induced deficits, and in transgenic Alzheimer’s mice
- Animals with normal baseline cognition — no improvement, which is the more interesting half of the finding: the pathway appears to engage under injury rather than during ordinary function
- Nerve injury models — one rat sciatic-nerve study reported sensory improvement, in combination with stem cells and G-CSF rather than alone
- Ototoxicity protection — a zebrafish study reported protection of lateral-line hair cells, well outside the cognitive frame
None of that is a use. It is a description of what happened to animals in a small number of papers, two of which are no longer in the literature.
Who should avoid it
- Anyone with a current or past cancer diagnosis, or a family history that puts them at elevated risk — c-Met is a pathway oncology works to shut off, and no carcinogenicity study of dihexa exists
- Anyone pregnant, breastfeeding or planning a pregnancy — no reproductive data of any kind
- Anyone taking it for healthy cognitive enhancement, which is the most common reason people buy it and the one thing the animal data specifically failed to show
- Anyone who would be self-dosing a research chemical, which is the only form it comes in: no established human dose, no route, no schedule, no monitoring, and no one to tell if something goes wrong
Route, dose and course
There is nothing to report here, and that is the answer rather than a gap in this page.
No human dose has been established. The rodent studies used systemic doses from picomolar to a few milligrams per kilogram, some oral and some by direct intracerebroventricular injection — a route that involves a needle in the brain and does not translate to a capsule. Oral bioavailability was demonstrated in rats, which is a property of rat physiology until somebody measures it in a person.
What that means in practice:
- No approved product — nothing to dispense, because nothing is approved anywhere
- No compounding route — not on the 503A bulks list, not on the 503B nominated list, so no pharmacy is sanctioned to prepare it
- No prescriber — there is no indication to write for and no dose to write
- What is sold instead — vials and capsules labelled for research use, with no verified potency, sterility or identity, and no batch anyone has tested
The usual arc
Nobody knows, and this is the rare page where that is the whole of the honest answer. There is no human timeline to describe because there has never been a human study — no onset, no plateau, no duration, no discontinuation data. Anyone reporting an arc is reporting their own experience with an unverified product, and the animal work that generated the interest specifically found no effect in subjects whose cognition was already normal.
No provider we've verified carries Dihexa yet
That is the honest answer, and it is not the same as "you can't have it". None of the licensed telehealth providers we track publishes Dihexa, so there is no price to quote and no intake to send you to. Whether that changes depends on federal action rather than on us — which is why the next box exists rather than a button that goes nowhere.
Tell me when someone carries it
Questions we get about Dihexa
Has dihexa ever been tested in humans?
No. Not a Phase 1, not an open-label series, nothing. Everything written about its cognitive effects comes from rodent and cell-culture work, most of it from the lab that designed the molecule. The compound that did reach human trials is a successor from the same group — ATH-1017, developed by Athira Pharma — and it is a different molecule with its own record. Claims about what dihexa does in people are extrapolation, not data.
What happened with the retractions?
The two papers that established the HGF/c-Met mechanism for dihexa specifically — Kawas 2012 and Benoist 2014 — were formally retracted in 2025, after Notices of Concern were issued in 2021. A third, McCoy 2013, which reported the memory and dendritic-spine findings, carries a Notice of Concern and has not been retracted. This does not prove dihexa fails to activate c-Met; it means the published biochemistry that showed it has been withdrawn. What survives is one independent replication in Alzheimer's mice from a Chinese group in 2021, which observed an effect and traced it to a PI3K-dependent pathway without re-establishing the binding chemistry.
Can I get dihexa legally?
Not for human use. It is not approved anywhere, and it is not on the FDA 503A bulks list or the 503B nominated substances list, so no compounding pharmacy is sanctioned to prepare it. It is sold online as a research chemical, and that label is not a permission — it marks a product that has not been made, tested or released for people to take.
See what licensed providers actually carry →What is the safety concern people raise about c-Met?
The pathway dihexa is proposed to potentiate is one that oncology spends considerable effort trying to shut off, because c-Met activation promotes tumour growth and invasion. That is a theoretical concern rather than an observed harm — but the reason it stays theoretical is that no carcinogenicity study of dihexa exists, and no human has been followed on it. An absence of reported harm in a compound nobody has formally studied in people is not a safety record.
Why did the rodent results look so strong?
Two reasons worth separating. The potency was real in the assays reported — picomolar activity and roughly a doubling of functional synapse formation in hippocampal culture. And the effect appeared in impaired animals but not in animals with normal baseline cognition, which fits a pathway that engages under injury rather than during ordinary function. That pattern makes the preclinical story coherent. It does not make it a human result, and two of the papers behind it are no longer in the literature.
We publish no medical reviewer by design. Every claim on this page traces to a source we hold on file, and the reviewed write-up with those citations lives on our hub. Read the full research on PeptideWellness ↗
Fuel your protocol with
the latest research + reviews
Independent updates on peptide therapy, breaking news from compounding pharmacies, and exclusive guides — straight to your inbox.
