Dihexa Dosage: What Animal Studies Used (and Why Human Dosing Is Unknown)

Key takeaways

• There is no human dose of dihexa, because dihexa has never been tested in a person. FDA says it has found no human exposure data by any route. Every dihexa dosage chart online was written by a seller or a forum.
• The studied doses are rodent doses: in rats, 1.25 to 2 mg/kg by stomach tube, 0.25 to 0.5 mg/kg injected into the abdomen, or a nanomole infused into the brain; in mice, 1.44 or 2.88 mg/kg for three months.
• Converted by body surface area, the 2 mg/kg rat dose corresponds to about 23 mg a day for a 70 kg adult. That is our arithmetic, not a dose; by body weight alone it would be 140 mg.
• Nobody knows how long dihexa lasts in a person. In rats its half-life was 9 to 13 days; the active metabolite of fosgonimeton, a related injectable drug, lasted about 1.5 hours in people. A schedule cannot be built on numbers that far apart.
• This page gives no reconstitution example and no units table: no US pharmacy may compound dihexa, so there is no legal product to measure.

The direct answer: there is no dihexa dosage for humans. Dihexa has never been given to people in a registered trial, and FDA's compounding page says the agency has not identified any human exposure data for it by any route (FDA Category 2). What exists are rat and mouse doses, which this page lays next to the oral, cream and injection charts that sellers circulate, with our conversions labeled as the arithmetic they are. Our dihexa profile covers what dihexa is, the evidence and the retractions; the companion page on whether dihexa is safe or legal covers the risks.

The studied doses

Three things matter in reading that table. The effective doses rose steeply with the route: a nanomole into the brain, half a milligram per kilogram into the abdomen, two milligrams per kilogram by mouth, which is what a drug that is only partly absorbed would show. The experiments were short, 8 days of testing in rats and three months in mice, so nothing here describes months or years of use. And the founding paper, the source of most of these numbers, has carried an expression of concern from its journal since 2021, as our dihexa profile explains.

Our arithmetic, labeled as arithmetic

Animal doses are converted to a first guess at a human-equivalent dose by body surface area rather than weight, because small animals clear drugs faster. The standard factors divide a rat's mg/kg dose by 6.2 and a mouse's by 12.3 (Nair 2016). For a 70 kg adult:

None of those numbers is a dose. Drug developers use surface-area conversion to choose a cautious starting point for a supervised first-in-human safety study, and then divide it by a further safety factor, usually 10 (Nair 2016). The method assumes the drug behaves in people roughly as it does in rodents, which for dihexa nobody has checked. The doses infused into the brain cannot be converted at all; no product delivers dihexa that way.

The half-life problem

A dosing schedule depends on how long a drug lasts, and for dihexa that number is missing in the one species that matters. In rats, the plasma half-life was 12.7 days after 10 mg/kg intravenously and 8.8 days after 20 mg/kg into the abdomen, estimates extrapolated from only 5 days of blood sampling (McCoy 2013). If anything like that held in people, a daily capsule would keep accumulating for weeks.

The only human measurement in the family points the other way. Athira Pharma, which licensed dihexa and abandoned it in 2015, developed fosgonimeton, an injectable drug its 2021 SEC filing describes as a prodrug converted in the blood into dihexa (Athira 2021). In fosgonimeton's phase 1 study, the active metabolite reached its peak about half an hour after an injection under the skin and had a half-life of about 1.5 hours, with an occasional slower tail of about 5 hours (Hua 2022). A half-life of hours and a half-life of days imply completely different schedules, and no one has measured what happens after a dihexa capsule, cream or vial.

The only human doses in the family belong to a different drug

Fosgonimeton is not dihexa: it is a separate, pharmaceutical-grade molecule with a phosphate group attached, injected under the skin. Its studied doses were single injections of 2 to 90 mg in 48 healthy young men, 20 to 80 mg daily for 9 days in 29 healthy older adults, and 40 mg daily in 11 people with Alzheimer's disease, with no maximum tolerated dose reached (Hua 2022). In the 26-week LIFT-AD trial, 40 mg daily did not beat placebo on any primary or secondary measure in 287 people with mild-to-moderate Alzheimer's, and 14.2% stopped because of side effects, mostly injection-site reactions, against 4.6% on placebo (Porsteinsson 2025). Those numbers are not a dihexa dose, and they do not validate any seller's chart; they show what a tested product in this family looked like, and that it did not work for the disease it was tested in.

The charts sellers circulate

These are claims, not doses. Some pages also quote an oral bioavailability percentage; the founding paper reports only a computer prediction of absorption from the human gut, and we found no measured figure in any species.

Why there is no reconstitution example

People searching for a dihexa dosage often want one more thing: how much bacteriostatic water to add to a 10 mg vial and how many units to draw. We are not going to give that, for three reasons.

• No pharmacy may make it. Dihexa is not approved anywhere and is not on the list of substances 503A pharmacies may compound. It came off FDA's Category 2 list on April 15, 2026 only because its nomination was withdrawn, and the advisory committee review is not due until February 2027, as our legal tracker records.
• The arithmetic would describe an unknown. Reconstitution math assumes the vial holds what its label says. A research vial's content, purity and sterility have not been verified by anyone, so a precise volume would be precise about nothing.
• Reconstitution help is what FDA cites. In its 2026 warning letters to research-peptide sellers, FDA pointed to dosing guides, reconstitution calculators and bundled bacteriostatic water as evidence that the vials were drugs for human use.

Our reconstitution calculator exists for medications a pharmacy has dispensed with a labeled concentration, and our peptide dosage chart covers the peptides that have one. Dihexa has neither.

Bottom line

A truthful dihexa dosage chart has an empty human row. Above it sit rat doses of 0.25 to 2 mg/kg for about a week of testing, a mouse study of three months, and a related injectable drug that did not help in Alzheimer's disease. Below it sit seller charts for capsules, creams and vials that no study has tested. Between them is the measurement nobody has made: how much dihexa reaches a human brain, and for how long. Until that exists, and until a pharmacy may legally prepare it, there is no dihexa dose, only a number someone chose.

Frequently asked questions

+−What is the dihexa dosage for humans?

There is none. Dihexa has never been given to people in a registered trial, and FDA says it has found no human exposure data for it by any route. The only studied doses are rat and mouse doses, mainly 2 mg per kilogram a day by stomach tube in rats. Every human dosage chart online was written by a seller or a forum, not derived from a study.

+−How much dihexa did the rat studies use?

The founding 2013 study gave 1.25 or 2 mg per kilogram by stomach tube, 0.05 to 0.5 mg per kilogram injected into the abdomen, or 0.1 to 1 nanomole infused into the brain, daily during 8 days of maze testing, in rats with drug-induced memory loss; 24-month-old rats got 2 mg per kilogram by mouth daily. A 2021 mouse study used 1.44 or 2.88 mg per kilogram for three months.

+−Is oral dihexa better than injecting it?

Nobody knows. In rats, 2 mg per kilogram by mouth and 0.5 mg per kilogram into the abdominal cavity both reversed a memory deficit, but no study has given dihexa under the skin, which is how injectable vials are used, and no human has had either route measured. The oral-absorption figure in the original paper is a computer prediction.

+−How many milligrams are in dihexa pills?

Seller charts in October 2026 list capsules of 5 to 10 mg a day to start, rising to 20 to 45 mg, and one guide quotes 10 to 50 mg a day. No study tested any of these amounts. By our arithmetic, the 2 mg per kilogram rat dose corresponds to about 23 mg a day for a 70 kg adult by body surface area; that is a calculation, not a dose.

+−Is there a dihexa cream or transdermal dose?

No. We found no study in any species that applied dihexa to the skin, so the 5 to 20 mg cream figures on seller charts have no source and nobody knows how much, if any, is absorbed.

+−How long does dihexa stay in the body?

In rats, the plasma half-life was about 12.7 days after an intravenous dose and 8.8 days after an injection into the abdomen, estimates extrapolated from 5 days of blood sampling. In people it has never been measured after a capsule or a vial. In a trial of fosgonimeton, a different drug that its developer says becomes dihexa in the blood, the active metabolite's half-life was about 1.5 hours.

+−How do you reconstitute dihexa?

We do not give reconstitution instructions for dihexa. No US pharmacy may legally prepare it, so a vial has no label anyone stands behind and unverified content, and FDA has treated reconstitution guides and calculators on vendor websites as evidence that such vials are unapproved drugs for people.

+−Can a doctor prescribe dihexa at a proper dose?

Not in October 2026. Dihexa is not FDA-approved, is not on the list of substances pharmacies may compound, and its FDA advisory committee review is not due until February 2027. A prescription cannot make a non-compoundable substance compoundable, so there is no labeled product and no labeled dose.

Sources

1. StudyMcCoy AT et al. Evaluation of metabolically stabilized angiotensin IV analogs as procognitive/antidementia agents. J Pharmacol Exp Ther 2013;344:141-54. Expression of concern published 2021 (J Pharmacol Exp Ther 378:313). (2013)
2. StudySun X et al. AngIV-analog dihexa rescues cognitive impairment and recovers memory in the APP/PS1 mouse via the PI3K/AKT signaling pathway. Brain Sci 2021;11:1487. (2021)
3. StudyMartino KA et al. Hepatocyte growth factor/MET activator rescues working memory deficits after repeated mild traumatic brain injury. Neurotrauma Rep 2025;6:991-8. (2025)
4. ReviewNair AB, Jacob S. A simple practice guide for dose conversion between animals and human. J Basic Clin Pharm 2016;7:27-31 (divide the rat mg/kg dose by 6.2 and the mouse dose by 12.3). (2016)
5. StudyHua X et al. Safety, tolerability, pharmacokinetics, and pharmacodynamics of the positive modulator of HGF/MET, fosgonimeton, in healthy volunteers and subjects with Alzheimer's disease: randomized, placebo-controlled, double-blind, phase I clinical trial. J Alzheimers Dis 2022;86:1399-413. (2022)
6. StudyPorsteinsson AP et al. Fosgonimeton in mild-to-moderate Alzheimer's disease (LIFT-AD). J Alzheimers Dis Rep 2025;9:25424823251405817. (2025)
7. SourceAthira Pharma, Inc. Form 8-K filed October 21, 2021: dihexa development 2011 to 2015 and its discontinuation; ATH-1017 (fosgonimeton) described as a prodrug that metabolizes in plasma into dihexa. (2021)
8. RegulatoryFDA. Certain bulk drug substances for use in compounding that may present significant safety risks: Category 2 list and bulk drug substances nominated but withdrawn (dihexa acetate). Content current as of April 22, 2026. (2026)