Dihexa turns up in nootropic catalogs as a memory compound. The research behind it is narrower than that label suggests. It comes mostly from one laboratory, it is entirely preclinical, and its central mechanism paper no longer stands. This page reads the studies one at a time and sits under the longevity goal, where cognitive claims are filed.
What dihexa is
Angiotensin IV is a small peptide made in the body, and analogs of it had long been studied in animals as possible memory drugs. A laboratory working on them set out to make a version that is metabolically stable and can reach the brain [10].
The result was N-hexanoic-Tyr-Ile-(6) aminohexanoic amide, named dihexa, derived from an earlier analog called Nle1-angiotensin IV [1]. Its developers describe it as orally active and able to cross the blood-brain barrier. Strictly, it is a modified peptide fragment rather than a full peptide.
What the animal studies found
The same laboratory reported that dihexa improved memory and formed new connections between neurons in animal models of Alzheimer’s disease [1]. Its 2013 discovery paper, the source of those rat results, has carried an expression of concern from its journal since September 2021 [8], so this page does not lean on its figures.
| Study | Model | Result |
|---|---|---|
| 2013, discovery paper | Rats | Memory gains; under expression of concern |
| 2014, mechanism paper | Cells and rats | Retracted in 2025 |
| 2015, hair cell study | Zebrafish larvae | Protected against antibiotic damage at 1 μM |
| 2021, Alzheimer's model | APP/PS1 mice | Water maze performance restored |
| 2021, nerve repair | Rats, 6 to 8 per group | Motor gains when combined with stem cells |
| 2024, Huntington's model | Rats (n = 40) | No protection |
A 2021 study from a separate group in China tested dihexa in APP/PS1 mice, a strain bred to develop Alzheimer’s-like changes [2]. It reported restored learning in the water maze, more neurons, and lower levels of the inflammatory signals IL-1β and TNF-α. Dose groups were used, but the abstract does not give the amounts.
A 2015 zebrafish study looked at hearing-type hair cells [3]. Dihexa at 1 μM gave the best protection against two antibiotics known to damage them. A 2021 rat nerve-repair study gave dihexa at 2 to 4 mg/kg alongside other treatments [4]. Motor function improved significantly with stem cells plus dihexa, or stem cells plus a growth factor, injected into the calf muscle.
The independent result that did not go its way is the most recent. In 2024, a team randomized 40 male rats to three groups in a model of Huntington’s disease induced by a mitochondrial toxin [5]. Dihexa, under its code name PNB-0408, did not protect them from the motor and memory deficits the toxin caused.
A 2018 systematic review of angiotensin IV research gives the wider picture [6]. In models of memory impairment, 8 of 9 studies found angiotensin IV and its analogs, dihexa among them, improved performance. Every study was in animals, and the compounds worked best given directly into the brain’s fluid spaces near the time of learning.
The retracted mechanism paper
The usual account of how dihexa works is that it boosts hepatocyte growth factor and that factor’s receptor, c-Met, to build new synapses. A 2014 paper was titled as showing exactly that, that the memory and synapse effects of angiotensin IV-derived peptides depend on this growth factor system [7].
The growth factor account also appears in the 2015 zebrafish paper, where a growth factor blocker weakened dihexa’s protection [3], and in two 2015 reviews from the same laboratory [1] [10]. Those are not retracted. The paper that set out to establish the mechanism is.
Side effects and the cancer question
No human has taken dihexa in a registered trial, so there is no record of side effects in people. ClinicalTrials.gov lists no study under the name dihexa or its code name. A 2026 review of peptides in orthopedics lists dihexa among neuroactive peptides and notes that clinical trials are lacking across the category [11].
The cancer question comes from the pathway itself. The c-Met receptor that dihexa was said to boost is a recognized cancer drug target. A 2021 review describes MET as druggable across several human cancers, with the inhibitor capmatinib approved for a type of metastatic lung cancer [12]. No published study has tested what long-term dihexa exposure does to tumor growth, in animals or people.
Where it stands with the FDA
Dihexa appears on neither FDA bulk drug substance list: not in the 503A categories, updated May 14, 2026 [13], and not in the 503B categories, updated March 21, 2025 [14]. A pharmacy therefore has no lawful route to compound it, and no human dose has been established. Anything sold as dihexa sits outside the prescription system.
For how those lists decide what a pharmacy may prepare, see are peptides legal. For another brain-targeted peptide with some human data behind it, see semax, and for one with only laboratory work, KPV.