Dihexa
N-hexanoic-Tyr-Ile-(6) aminohexanoic amide
Dihexa (N-hexanoic-Tyr-Ile-(6) aminohexanoic amide) is a cognitive & mood research compound. Angiotensin-IV analog that potentiates HGF/c-Met signaling to drive synaptogenesis — potent nootropic/neuro-repair.
Dihexa quick facts
| Reported research dosing (Oral) | 5mg-20mg |
| Route | Oral |
| Cycle length | 4-8 Weeks |
| Frequency | 1x Daily · 5 On 2 Off or Daily |
| Half-life | Not well characterized (oral-active, hours) |
| Forms | Oral, Nasal |
| Evidence level | Animal |
| Other forms available | Nasal — dosed differently |
Reported orders-of-magnitude stronger than BDNF at building synapses. Powerful and early — respect it.
How Dihexa works
Angiotensin-IV analog that potentiates HGF/c-Met signaling to drive synaptogenesis — potent nootropic/neuro-repair.
Proposed benefits
Synaptogenesis for memory and cognition (potent HGF/c-Met pathway).
Where to get Dihexa
Dihexa is sold in 2 forms. They are not interchangeable — the dose and the route differ, so pick the one this page describes unless you know why you want another.
The evidence for Dihexa
Graded by what exists behind each claim.
Human clinical evidence
- No human trials: development stopped at preclinical, so the ceiling on any claim here is a rodent model — and these transfer poorly.
📊 Correlative data
- Small community use, mostly for cognitive recovery. The safety record is community reports only, and for a compound this potent at promoting synapse formation that gap is worth taking seriously rather than waving through.
- Beyond that the record is self-reported: community dosing logs are real information about tolerability and almost none about efficacy.
🧪 How the mechanism reads
- An angiotensin IV analog that potentiates hepatocyte growth factor signaling through the c-Met receptor. Rodent work reports it is orders of magnitude more potent than BDNF at promoting synaptogenesis, with cognitive restoration in lesion models.
- c-Met is also an oncogene, amplified in several cancers. A compound whose mechanism is potent growth-factor receptor activation carries exactly the proliferation question that BPC-157 and the IGF analogs do, and here it is aimed at the brain.
Why an empty tier is not a verdict → · What community dosing logs are worth →
How to read these tiers: they say how much human evidence exists, not how well something works — and ✗ flags harm, never a disappointing trial. How the evidence tiers work →
What Dihexa actually does
The sequence is the page, and once you read it properly the compound stops being a peptide. Angiotensin IV is the hexapeptide Val-Tyr-Ile-His-Pro-Phe, the 3–8 fragment of angiotensin II. Structure–activity work reduced the active core to the N-terminal end, and McCoy 2013 started from the tripeptide Nle-Tyr-Ile — norleucine substituted at position 1 — then applied N- and C-terminal modifications designed to increase hydrophobicity and decrease hydrogen bonding. The product was N-hexanoic-Tyr-Ile-(6) aminohexanoic amide: dihexa.
Count what is left. Of the original six residues, exactly two survive: Tyr-Ile. The hexanoyl group capping the amino end and the 6-aminohexanoic amide capping the carboxyl end are not amino acids at all — they are fatty-acid-like spacers. And the reason they are there is the interesting part. A free N-terminal amine and a free C-terminal carboxylate are two charges, and they are also two handles: aminopeptidases grab the first, carboxypeptidases the second. Capping both deletes the charges and deletes the handles in one move. That is why McCoy 2013 can describe the result as orally active and blood-brain-barrier permeant — properties no charged tripeptide has.
Which answers the question this site asks of every short peptide, and answers it against the peptide framing. A two-residue stretch cannot carry sequence-specific recognition information in the way a real protein–protein interface does; Tyr-Ile is a tyrosine side-chain hydroxyl and an aliphatic isoleucine side chain, and there are not enough contact points there to encode selectivity by backbone recognition. Dihexa was never designed as a peptide ligand. It is a small lipophilic molecule that happens to contain a dipeptide, and every property that makes it usable came from removing peptide character rather than refining it.
The target is not the receptor. It is the receptor’s ligand, and that distinction matters. Benoist 2014 reports that dihexa binds with high affinity to hepatocyte growth factor (HGF) and thereby activates c-Met signaling. So this is a ligand-directed potentiator: it needs endogenous HGF to be present to do anything, which is a materially different pharmacology from a receptor agonist and predicts that the effect size depends on the user’s own HGF rather than on the dose alone.
And there is an older target the field split away from. Angiotensin IV’s classical action is inhibition of insulin-regulated aminopeptidase (IRAP), and a whole medicinal chemistry program stepwise transformed the hexapeptide into drug-like Ang IV peptidomimetic IRAP inhibitors Hallberg 2020. Dihexa came off the other branch of the same tree and landed on HGF/c-Met Benoist 2014. Two lineages, one parent hexapeptide, two different molecular targets — and a page that calls dihexa an “angiotensin IV analog” without saying which branch it is on has said almost nothing. c-Met is a proto-oncogene, and that is the fact the rest of this page keeps returning to.
Cell, rodent, human — and where it stops
Step one, chemistry. Hexapeptide to tripeptide to capped dipeptide, with hydrophobicity deliberately increased and hydrogen bonding deliberately reduced McCoy 2013.
Step two, rodents, oral, in two deficit models. McCoy 2013 reports marked synaptogenic activity and procognitive effects in aged rats and in the scopolamine model of cholinergic impairment, with the compound given orally and described as blood-brain-barrier permeant. Note what both models have in common: they are impaired animals. Nothing in this literature tests a healthy young one.
Step three, the mechanism nailed down two independent ways. This is the strongest part of dihexa’s file and it deserves to be stated properly. Benoist 2014 delivered an HGF antagonist intracerebroventricularly and it blocked dihexa’s procognitive effect in the Morris water maze; separately, shRNA knockdown of c-Met inhibited spinogenesis and synaptogenesis. A pharmacological block and a genetic knockdown pointing at the same pathway is better mechanistic evidence than most compounds in this catalog have.
Step four, a different organism and a non-neuronal cell. Uribe 2015 used the larval zebrafish lateral line and found dihexa protected hair cells from neomycin and gentamicin toxicity, with optimal protection at 1 µM. That is the only concentration anchor in the entire literature, and it is a bath concentration in water around a fish larva — not a plasma level in anything.
Step five, in humans: nothing at all. No trial, no case series, no phase 1, no published human pharmacokinetics, no human safety data. The program’s own stated destination was the development of small-molecule angiotensin IV analogs to treat Alzheimer’s and Parkinson’s diseases Wright 2015, and it did not get there.
The obstacle, stated exactly, and it is unusually stark. The mechanism is well supported and the translation step is entirely empty — but the specific gap is worse than “no human data”. There is no published rodent milligram-per-kilogram dose in the abstract record for the behavioral results, which means the usual body-surface-area conversion cannot be performed at all. The 5–20 mg doses this site records therefore have no animal dose to be scaled from. That is a different and sharper problem than a compound whose human dose is too low or too high: this one has no anchor in either direction, and the only concentration anyone has published is 1 µM in a fish tank Uribe 2015.
Dihexa pharmacokinetics — how much of it actually gets in
Route: oral, and the molecule was engineered specifically so that this would work. McCoy 2013 describes dihexa as orally active and blood-brain-barrier permeant, and the structural reason is in the name: the N-hexanoyl cap and the 6-aminohexanoic amide cap remove the ionizable termini that would otherwise hold a Tyr-Ile dipeptide out of a lipid membrane. Charge is what blocks membrane crossing; the design deleted the charge.
What degrades it, reasoned from the structure because nothing else is published. Both exopeptidase routes are shut — there is no free amine for an aminopeptidase and no free carboxyl for a carboxypeptidase. What remains chemically vulnerable is the single internal amide bond between the two residues, an endopeptidase or amidase substrate, and the two aliphatic hexanoyl chains, which are the obvious sites for hepatic oxidation and β-oxidation. No cytochrome assignment, no metabolite identification and no clearance route has been published for dihexa in any species, so the paragraph above is chemistry rather than data, and it is offered as such. The practical consequence is that interaction risk with anything else cannot be reasoned about at all.
The half-life question, answered honestly. There is no published half-life for dihexa in any species — not in rodents, not in fish, certainly not in a person. What can be said is what the structure implies: a molecule with both peptide termini capped does not have the seconds-to-minutes serum half-life of the tripeptide it came from, and the oral activity demonstrated in rodents McCoy 2013 sets a floor — a compound cleared in minutes does not produce a behavioral effect after oral gavage. Somewhere between those two statements is the real number, and nobody has measured it.
The oral barrier and the injectable comparator. First-pass hepatic metabolism and gastrointestinal stability are the barriers the capping chemistry was built to defeat, and by the developers’ account it defeated them. There is no subcutaneous or intranasal dihexa with published pharmacokinetics in any species, so an injection or a nasal spray of this compound has no exposure profile anyone can quote — and unlike a peptide, where injection exists because the oral route fails, here the oral route is the one with the data.
The one number, and what it is not. 1 µM was the optimal protective concentration in the zebrafish hair-cell assay Uribe 2015. That is a concentration in the water around a larva with a permeable surface, and converting it to an oral human dose requires a volume of distribution, a bioavailability and a clearance — none of which exist. Anyone quoting a dihexa dose is quoting a number with no published derivation.
What would have to be true, and how you would know it was not
Four predictions. The second is the one that argues most people will get nothing, and the third is a positive control that does not involve the brain at all.
1. If there is a cognitive effect, it needs an instrument and a written baseline. Run a MoCA and a Trail Making Test at the same hour on three separate days before starting, to establish a personal range rather than a single number, then repeat the same three-session protocol at 8 and 12 weeks. Both tests have practice effects, which is precisely why the baseline has to be repeated rather than taken once — a single before-and-after pair on a cognitive test will show improvement whether or not anything happened.
2. The prediction that cuts against it: a healthy adult should see nothing. Every positive behavioral result in this literature came from an animal with an induced or age-related deficit — aged rats and the scopolamine model McCoy 2013 — and a synaptogenic agent restoring an impaired network is a different claim from one enhancing an intact one. The falsifiable version: in someone with a normal baseline MoCA, the 12-week score should sit inside the three-session baseline range. If it does not, that is a genuinely new observation, because no published experiment has ever tested this compound in an unimpaired subject.
3. A positive control that has nothing to do with cognition: wound healing should speed up. HGF/c-Met is an epithelial growth-and-repair pathway, and Uribe 2015 demonstrated dihexa acting on non-neuronal hair cells in a whole animal. So if the compound is reaching a systemic concentration that engages c-Met, the effect should be visible somewhere other than the brain. Photograph and time a small standard skin injury — the same kind of abrasion, the same body region — before and during use. No peripheral effect at all is evidence that systemic exposure is low, which would also explain a null cognitive result.
4. The liver is the organ this growth factor is named after. Hepatocyte growth factor was discovered as the driver of liver regeneration, and dihexa works by binding it Benoist 2014. Run a Comprehensive Metabolic Panel (CMP) at baseline and at 12 weeks. The mechanistic prediction is not injury — it is that nothing should move, because a potentiator of a repair pathway is not a hepatotoxin. An unexplained change in that panel on a c-Met potentiator is worth taking seriously precisely because the mechanism does not predict it.
What nobody has tested yet
Nobody has ever run a proliferation or carcinogenicity study on a c-Met potentiator taken chronically, and this is the experiment that should exist before any other. c-Met is a proto-oncogene, dihexa’s mechanism is to raise signaling through it Benoist 2014, and the published record contains no repeat-dose toxicology, no tumor incidence data and no organ histology from a long-dosing study in any species. The design is standard: a rodent chronic-dosing cohort with liver, lung, gastric and skin histology and Ki-67 staining. It has not been done, or it has not been published, and from outside those are the same thing.
Nobody has published a single pharmacokinetic parameter. No half-life, no bioavailability, no volume of distribution, no brain concentration — despite blood-brain-barrier permeability being one of the compound’s two headline claims McCoy 2013. A brain-to-plasma ratio in six rodents would substantiate the claim everybody repeats, and it is not in the literature.
Nobody has measured whether dihexa changes HGF signaling in a tissue you can actually sample. The mechanism is ligand binding Benoist 2014, so free versus bound HGF in plasma, or phospho-c-Met in a skin biopsy, are both accessible read-outs. Neither has been reported, which means there is no biomarker to confirm target engagement in any species, let alone a person.
And nobody has tested whether the effect requires a deficit to act on. Aged and scopolamine-treated animals McCoy 2013 are damaged networks; the HGF antagonist and c-Met knockdown experiments Benoist 2014 establish the pathway but not the requirement for prior impairment. A single study comparing young intact animals with aged ones on the same dose would answer the question that decides whether this compound is a repair agent or an enhancer, and it has never been run.
Dihexa — its own safety story, not its class's
There is no human safety data for dihexa of any kind. No trial, no case series, no phase 1, no published adverse-event record. Everything below is mechanism and structure, and it is labeled as such because presenting an absence of reports as an absence of risk would be the specific dishonesty this page exists to avoid.
The proto-oncogene problem is this compound’s own, and it is sharper than the generic growth-factor caution. c-Met is a receptor tyrosine kinase that is amplified or overexpressed in several human cancers, and dihexa’s demonstrated mechanism is to bind its ligand and increase signaling through it Benoist 2014. The usual mitigating argument — that a peptide will not reach the tissue in question — does not apply here, because the molecule was deliberately engineered to be orally active and to cross membranes McCoy 2013. What normally limits exposure has been removed on purpose.
And the effect is demonstrably not brain-selective. Uribe 2015 showed dihexa acting on epithelial hair cells in a zebrafish lateral line — a non-neuronal cell type in a peripheral organ, at 1 µM. Whatever selectivity the compound has, it is not tissue selectivity, and the c-Met pathway is present in essentially every epithelium.
Identity, and a specific analytical problem this molecule has. There is no approved product and no pharmacopoeial monograph, so purity rests on a vendor certificate. And the usual peptide certificate is the wrong instrument: dihexa is not a peptide in the analytical sense — it is a Tyr-Ile dipeptide carrying two aliphatic acyl caps, so an HPLC method validated against peptide standards is not obviously fit for it, and a stated purity figure says nothing about which acylation product is in the vial. The N-hexanoyl group can in principle sit on the tyrosine side-chain hydroxyl instead of the terminal amine, and those two products have identical mass.
The honest summary. This is a compound with unusually good mechanistic evidence, unusually good oral and brain delivery, a target that is a named oncogene, and zero human data. Each of the first three would ordinarily be a reason for optimism; together with the fourth they describe a molecule whose exposure is unbounded and whose long-term consequence is entirely unmeasured.
Sources read for this page
- McCoy AT. Evaluation of metabolically stabilized angiotensin IV analogs as procognitive/antidementia agents. J Pharmacol Exp Ther 2013 · PMID 23055539
- Benoist CC. 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 · PMID 25187433
- Uribe PM. Hepatocyte growth factor mimetic protects lateral line hair cells from aminoglycoside exposure. Front Cell Neurosci 2015 · PMID 25674052
- Hallberg M, Larhed M. From Angiotensin IV to Small Peptidemimetics Inhibiting Insulin-Regulated Aminopeptidase. Front Pharmacol 2020 · PMID 33178027
- Wright JW. The development of small molecule angiotensin IV analogs to treat Alzheimer's and Parkinson's diseases. Prog Neurobiol 2015 · PMID 25455861
Dihexa — safety, predicted from mechanism
Predicted from mechanism, not from a human safety trial. How that reasoning works →
What the mechanism predicts
Derived from the molecule, not a trial.
- This class is broad, but the predicted problems cluster by mechanism rather than by molecule. Cholinergics (racetams, and anything raising acetylcholine) predict headache — the classic one, from choline demand outrunning supply. Dopaminergics and eugeroics predict tolerance, sleep disruption and a flat mood on the days off. Anything glutamatergic or AMPA-facing carries a theoretical excitotoxicity concern at high doses.
- The pattern worth internalizing: anything that borrows performance from tomorrow eventually presents the bill. Sleep is the most common currency it gets paid in.
What has actually been reported
- Headache is the most reported effect across the racetam family and usually responds to added choline.
- Irritability, blunted affect and a rebound low on cessation are commonly reported with the stimulant-adjacent members.
- Most of this class has little or no controlled human safety data at the doses actually used.
How to reduce the risk
Same mechanism as the prediction.
- Take a choline source with any racetam. The headache is the mechanism running out of substrate, and it is largely preventable rather than something to push through.
- Dose in the morning. Almost everything in this class has a longer functional tail than its half-life suggests, and sleep is the first thing you lose.
- Use them for something, not as a habit. The compounds that carry tolerance genuinely reward intermittent use aimed at a task, and genuinely punish daily use aimed at feeling normal.
- One at a time, and long enough to judge it. This is the class where people stack five and cannot tell you which one is doing anything — and the effects are subjective, so attribution is already hard enough.
- If you need it to feel normal, stop. That is the line where a tool has become a dependency, and it is the one worth watching for.
What it does to your bloodwork
A fact about the assay.
- No routine marker tracks these. Sleep is the assay — if it is degrading, the compound is costing more than it is producing, and that shows up before anything else does.
Don't run this if
- A seizure history — several of these lower the threshold at least theoretically, and it is not worth establishing empirically.
- Bipolar disorder, for the dopaminergic members especially.
- Alongside prescribed psychiatric medication without knowing exactly how the mechanisms overlap.
The honest unknown
- Chronic use is essentially uncharacterized. The specific unmeasured thing is what daily cholinergic or dopaminergic pressure does to baseline function over years — not whether a few weeks is tolerable.
Not medical advice. If you take prescription medication or have a diagnosed condition, check this with a pharmacist or doctor.
With food
Absorption is better with a meal, and for anything fat-soluble the fat is doing the work rather than the food generally. This is also the version that is easiest to actually remember, which matters more than it sounds.
From half-life and route, not a dosing trial.
Dihexa — interference & stacking
Predicted from mechanism, not from an interaction study. How mechanism-predicted claims are made →
What Dihexa moves on your bloodwork
Expected direction, not a measured one.
- Comprehensive Metabolic Panel (CMP) — ◆ worth watching
Most of this class has no predicted marker movement at all, and saying so is more useful than listing markers that will not move.
What to do: A baseline liver panel is reasonable for anything taken daily and long-term. Beyond that there is nothing specific to chase.
Everything on this page, in an order
This one is free and stays free. What Skool adds is the rest of the shelf — 278 compounds and 371 supplements with the protocol, the stack order and the bloodwork to run beside it.
Join Skool — $10/mo →Bloodwork to run alongside Dihexa
Baseline first, then again at 8–12 weeks.
| Marker | What it’s watching for |
|---|---|
| TSH (Thyroid-Stimulating Hormone) | Thyroid disease imitates every cognitive complaint there is |
| Vitamin B12 | Deficiency causes fog long before it causes anemia |
| Methylmalonic Acid (MMA) | Catches the deficiency a normal B12 hides |
| Ferritin | Low iron flattens cognition at levels most labs call fine |
| Vitamin D (25-Hydroxy) | Commonly low, cheap to correct, associated with mood |
The Brain Fog & Cognition panel covers these in one order — 12 markers, $233.06 with the discount applied.
Check results you already have → · All 103 markers A–Z
Dihexa — frequently asked questions
What is Dihexa?
Dihexa (N-hexanoic-Tyr-Ile-(6) aminohexanoic amide) is a cognitive & mood research compound. Angiotensin-IV analog that potentiates HGF/c-Met signaling to drive synaptogenesis — potent nootropic/neuro-repair.
What dosing does the research reference for Dihexa?
In the research literature, Dihexa is referenced in the 5mg-20mg range, 1x Daily · 5 On 2 Off or Daily. For research use only — not a recommendation for human use.
What is the half-life of Dihexa?
Dihexa has an approximate half-life of Not well characterized (oral-active, hours), which is part of what determines how often it's dosed.
What forms does Dihexa come in?
Dihexa is available as: Oral, Nasal.
What's the evidence behind Dihexa?
Current evidence level: Animal. Dihexa is offered for research purposes only and is not an approved medicine.
Dihexa inside a finished plan
One arm of 1 Protocol Blueprint, free to read in full.
What Dihexa is used for
Dihexa appears under 1 goal in the goal router.
Related Cognitive & Mood compounds
Where this goes next
Dihexa is the bdnf arm of this plan. The page above is the free breakdown of one compound; the plan it belongs to — the dosing, the order to correct things in, the week-by-week schedule and what to retest — is a lesson inside Skool.