Dihexa

The 2013 report, published online in late 2012 in the Journal of Pharmacology and Experimental Therapeutics, tested dihexa in two kinds of forgetful rats. One group had memory loss induced by scopolamine, a drug that blocks a key memory chemical. The other group was simply old. In both, dihexa improved performance in the Morris water maze, the standard pool-and-platform test of spatial memory.

The authors also reported marked synaptogenic activity in hippocampal neurons, meaning the growth of new synapses, the connections between neurons. Memory is stored largely in the pattern and strength of those connections, so a pill that builds them is the kind of result that gets attention well beyond a pharmacology journal.

Two details deserve a second look. The memory tests were short laboratory tasks, and the models were a drug-induced deficit and normal aging, not the plaque-and-tangle disease that defines Alzheimer’s in people. Those are standard first steps in drug development. The step that usually comes next, a careful human safety study, has never been published for dihexa.

In 2021, the journal added an expression of concern to this paper, a formal flag that questions have been raised about the work. That notice does not by itself mean the findings were wrong. It does mean the founding study of dihexa sits under a question mark that readers should know about.¹

Hepatocyte growth factor, HGF, was named for its effect on liver cells, but it acts across the body, brain included. It works through a receptor called MET, sometimes written c-Met. When MET is switched on, cells grow, move, change shape and resist injury. In neurons, the same signal encourages the branching and synapse building that learning depends on.

In 2014, the Washington State group published the explanation that made dihexa’s reputation. They reported that dihexa binds tightly to HGF itself and, in the presence of amounts of HGF too small to act on their own, switches on MET. In effect, dihexa was said to work like an amplifier, letting a faint growth signal be heard.

The paper went further. Dihexa and its parent compound built new dendritic spines and synapses in hippocampal neurons, much as HGF does. Blocking HGF or silencing MET stopped those effects. Most persuasive of all, an HGF blocker delivered into the brain erased dihexa’s memory benefit in the water maze.

A 2015 review by two of the authors, who also listed an affiliation with M3 Biotechnology, described the HGF/MET system as an overlooked target for Alzheimer’s disease and dihexa as a first-in-class compound. That review restated the HGF/MET mechanism, the explanation that would later be retracted, a version still widely repeated about dihexa online.⁷²

In 2021, the Journal of Pharmacology and Experimental Therapeutics published expressions of concern on both the 2013 paper that introduced dihexa and the 2014 paper that explained it. In April 2025, the journal retracted the 2014 mechanism paper. The retraction notice lists the original Washington State authors.

A retraction does not delete a paper. It stays readable, but it is formally flagged as no longer reliable. The notice indexed on PubMed does not state the reasons in an abstract, so the honest reading is narrow. The study that tied dihexa to HGF and MET can no longer be relied on, and the 2013 founding paper still carries a formal flag.

This does not prove that dihexa does nothing, or that HGF plays no part. Work by another group, a Chinese team studying Alzheimer’s mice in 2021, reported effects involving PI3K/AKT signalling, without testing whether HGF/MET sat upstream of it. What the retraction does is remove the strongest piece of evidence for how dihexa works, at exactly the point where online sellers lean on it most.⁷⁸¹⁵

A 2018 systematic review from the University of Southern California pulled together the animal research on angiotensin IV and related peptides. Of 450 articles, 32 met their standards. In animals with an induced memory deficit, 8 of 9 studies found that angiotensin IV or an analogue, including dihexa, improved spatial working memory or avoidance learning.

The same review noted a detail that matters for anyone reading forum dosing threads. These peptides looked most effective when injected directly into the fluid spaces of the brain, close to the time the animal was learning or being tested. That is a long way from a capsule taken by a person on an ordinary Tuesday.

In 2021, researchers at China Pharmaceutical University gave dihexa, which they describe as orally active, to APP/PS1 mice, a common Alzheimer’s model. It restored learning in the water maze, increased neurons and a synapse marker, calmed the brain’s immune cells and lowered the inflammatory signals IL-1β and TNF-α. Blocking the PI3K/AKT pathway reversed the anti-inflammatory effects, pointing to that route.

Not every test has gone well. In 2024, a team at Whitworth University split 40 rats into three groups: no toxin, a toxin that mimics Huntington’s disease, and the toxin plus dihexa. Over five weeks, dihexa did not protect the rats from the stunted weight gain, memory problems or movement problems the toxin caused.⁴⁵⁶

The idea behind dihexa, that boosting HGF and MET could rebuild synapses in Alzheimer’s disease, has been tested in people, with a different molecule. That drug is fosgonimeton, a small-molecule positive modulator of the HGF/MET system developed by Athira Pharma.

Its phase 1 trial, published in 2022, gave fosgonimeton or placebo to 88 people: healthy young men, healthy older adults and 11 people with Alzheimer’s disease. It was well tolerated, and brain-wave recordings suggested it reached its target. In the small Alzheimer’s group, a measure of how quickly the brain responds to a stimulus moved toward normal.

Then came the real test. LIFT-AD, a randomized, placebo-controlled phase 2/3 trial, based its primary analysis on 287 people with mild-to-moderate Alzheimer’s disease who were not taking cholinesterase inhibitors, and tracked safety in 549 dosed participants. At 26 weeks, fosgonimeton did not achieve its primary goal, a combined score of thinking and daily function, or its secondary goals. The differences leaned slightly toward the drug but were not statistically meaningful.

Fosgonimeton is not dihexa, and nothing here shows that one came from the other. But it is the closest thing to a human answer for the pathway dihexa was said to use. Pushing HGF/MET in a carefully run trial of people with Alzheimer’s did not produce a statistically significant improvement in thinking or daily function at six months. That is a sobering backdrop for a compound with no human data at all.⁹¹⁰

The short answer is nothing. FDA’s April 2026 compounding safety page states that FDA has not identified any human exposure data on drug products containing dihexa, by any route. There is no published trial, no case series and no pharmacology study in people.

That puts every claim about dihexa’s effects in humans, from sharper focus to Alzheimer’s prevention, in the category of extrapolation from rodents. Some of those rodent results are interesting. One of the most cited mechanism papers is retracted, and the founding paper carries an expression of concern.

The closest human evidence is indirect, from fosgonimeton, and that trial missed its goals. If someone tells you dihexa has been shown to work in people, ask for the study. There is not one to show you.¹²⁸¹¹⁰

With no human exposure data, dihexa has no known side-effect profile in people. FDA states that it lacks important information on any safety issues dihexa acetate raises, including whether it would cause harm in humans. So a lack of reported side effects reflects a lack of study.

The biggest theoretical concern comes from the very pathway dihexa was designed to boost. HGF and MET drive cell growth, invasion, movement and new blood vessel formation in tumours, and MET is a major target for cancer drugs, several of which are already on the market. Whether switching that system on for months could encourage a cancer is unknown.

The angiotensin connection raises a second, practical question. Dihexa comes from the same hormone family that ACE inhibitors and ARBs act on, and no one has studied how they interact. The fosgonimeton trial adds one data point for drugs aimed at this pathway. More people stopped the active drug because of side effects than stopped placebo, mostly from injection-site reactions.

Finally, there is the product itself. With no approved or compounded source, anything sold as dihexa comes from unregulated suppliers, so neither its identity nor its purity can be assumed.¹²¹¹¹⁰

Most people who search for dihexa are worried about memory, their own or a parent’s. A clinical work-up may include blood tests for conditions that can contribute to memory slips, though no blood test by itself explains them. Vitamin B12 deficiency can cause fog and forgetfulness, and replacing it may help, though recovery is not always complete. Homocysteine, an amino acid that rises when B vitamins run low, is linked with memory problems and vascular risk.

HbA1c, your average blood sugar over about three months, matters because high blood sugar damages the small vessels that feed the brain over years. hs-CRP measures inflammation anywhere in the body; it does not measure the brain inflammation that dihexa’s mouse research focused on.

This is how Vitals Vault approaches brain health. A blood panel at Quest Diagnostics comes first, and a licensed clinician reviews your results and history. New or worsening memory changes deserve a proper medical work-up, which no peptide replaces.⁵

Dihexa is not FDA-approved for any use. FDA’s compounding safety page lists dihexa acetate, noting no human exposure data, and it is not one of the six peptides covered by FDA’s October 29, 2026 enforcement discretion for 503A pharmacies.

That leaves no legitimate US prescription route. Vials sold online as research use only are not made for people.¹²¹³

Dihexa is a clever piece of chemistry with a striking set of rat and mouse results, a retracted mechanism, a founding paper under an expression of concern and no human data. The largest human trial cited here of its proposed pathway, using a different drug, did not significantly improve outcomes in Alzheimer’s disease.

If memory is the worry, the most useful next step is to find what is already measurable and fixable. And when anyone recommends dihexa, the question worth asking is simple. Which human study are we relying on?⁸¹⁰¹²