Epitalon: the monograph

Epitalon is a synthetic chain of four amino acids, Ala-Glu-Asp-Gly, with the molecular formula C14H22N4O9 and a molecular weight of 390.35 ⁷. It is also written epithalon, and in the more recent literature it is usually called the AEDG peptide. All three names describe the same molecule.

It was not discovered as a drug candidate and then tested. It was carved out of something older. In the 1970s a Soviet research group began studying Epithalamin, a peptide preparation extracted from the pineal gland of cattle, as a geroprotector. Epitalon is the short synthetic sequence identified from that work ^38,35,17. That lineage explains almost everything unusual about the evidence, including the fact that the famous longevity numbers were produced by the extract rather than by the peptide.

In the United States epitalon is not a medicine. No FDA-approved product contains it, FDA lists it among bulk substances that may present significant safety risks in compounding, and in July 2026 FDA proposed that it not be added to the list of substances pharmacists may compound with ^5,1,2. It reaches American buyers as a research chemical in a vial.

This monograph grades every claim by the species it was measured in and by who measured it. That second column matters more here than for any other compound in this fleet, and the section below explains why.

The fact chips on this page are sourced individually. The short version: a four-amino-acid peptide, no approved product anywhere in the United States, no established human dose, no published human pharmacokinetics, no registered trial, no randomized trial of the peptide itself, and a literature two thirds of which comes from the laboratory that created it.

Epitalon is not FDA approved. It is sold for research use; no human efficacy has been established, and FDA has proposed not to permit it in compounded drugs.

Each part of that is separately checkable. A Drugs@FDA query for epitalon as an active ingredient returns no matches, and the same query for the spelling epithalon also returns no matches ⁵. There is no approved product, no approved indication and no approved strength.

FDA lists Epitalon on its page of bulk drug substances that may present significant safety risks in compounding, in the table of substances nominated but withdrawn ¹. The agency's stated concern is specific and worth quoting rather than paraphrasing: compounded drugs containing epitalon "may pose risk for immunogenicity for certain routes of administration due to the potential for aggregation and peptide-related impurities", FDA "has not identified safety-related information regarding epitalon for the proposed route of administration", and therefore "the agency lacks sufficient information to know whether the drug would cause harm if administered to humans" ¹.

Epitalon appears in no category of FDA's interim 503A bulk-substance category list dated May 14, 2026 ⁴. It is not in the group of substances that may be compounded under an interim policy while FDA evaluates them. It is outside that framework entirely.

On July 23 and 24, 2026, FDA brought seven peptide substances to the Pharmacy Compounding Advisory Committee. Epitalon was on the July 24 morning agenda. FDA's Points to Consider, items 11 and 12, read: "FDA is proposing that Epitalon (free base) NOT be included on the 503A Bulks List" and "FDA is proposing that Epitalon acetate NOT be included on the 503A Bulks List" ².

Two things about that proposal are commonly misread. The first is what inclusion would have meant. The 503A Bulks List does not confer approval. It governs what a licensed pharmacist may compound with, inside the statutory exemptions from new drug approval, from labeling with adequate directions for use, and from current good manufacturing practice requirements ². Epitalon would not have become an approved medicine if the answer had gone the other way.

The second is what the briefing document actually contains. It is eight pages: an agenda, the four evaluation criteria from the 2019 final rule, the nominations, and the points to consider ². Under the epitalon section, the sub-heading "FDA Evaluation" has no text beneath it. The substantive review is referred to as linked background material and is not resolvable from the document itself. So this site quotes the proposal, which is written down, and does not attribute reasons to FDA that the document does not state ².

Both nominations to add epitalon had already been withdrawn by the nominators, LDT Health Solutions on behalf of the International Peptide Society, and Wells Pharmacy Network. FDA elected to present the substance to the committee anyway ². That is an unusual thing for an agency to do and it says something about how the agency views the substance.

FDA's own caveat is on page one: the agency "does not intend to issue a final determination on the issues at hand until input from the advisory committee process has been considered and all reviews have been finalized", and the final determination "may be affected by issues not discussed at the advisory committee meeting" ². The rulemaking is pending. No vote tally for epitalon was posted on FDA's meeting page when this page was built, so this site states the proposal and makes no claim about how the committee voted ³.

Here is the fact that reorganizes everything else on this page. Of the 167 records a four-term Europe PMC query returns for this peptide, 111 carry Vladimir Khavinson or Vladimir Anisimov as an author, or name the St. Petersburg Institute of Bioregulation and Gerontology in the affiliation. That is 66.5 percent. Khavinson alone is a named author on 100 of the 167, which is 59.9 percent of everything written about the compound ^44,41.

This is stated as provenance, not as an accusation. Khavinson's institute did the original work, published it, and has kept publishing it for four decades. That is what a research program looks like. The problem is not that the group exists; it is that when a reader is told epitalon is backed by decades of research, the decades and the research are very largely one group's.

The definition used here is deliberately generous to the network and strict about independence. A paper counts as independent only if no author from that network appears on it at all. By that rule the independent literature is real but small, and it is entirely in cells and animals: a Tbilisi State University group working on chromatin in lymphocyte cultures, a Shanxi Medical University group on mouse oocytes, a Gyeongsang National University group on bovine oocytes, and, most substantially, Brunel University London on telomere length in human cell lines in 2025 ^32,35,36,10.

No laboratory outside that network has published a study in which epitalon was given to a human being ⁴⁴. Every human record, including the mortality-reduction work that sells the compound, carries an author from the group that created it ^13,14,18.

The honest way to hold this is not suspicion. It is that a single-source literature has not yet been tested the way medicine tests things: by someone with no stake in the answer trying to get the same result and reporting it either way.

Epithalamin is a peptide complex extracted from the pineal gland. Epitalon is a synthetic tetrapeptide. The literature says so plainly: one 2007 abstract distinguishes "Epithalamin, a complex of peptides isolated from the pineal gland" from "Epitalon, synthetic tetrapeptide" in a single sentence ¹⁷.

This matters because the human evidence splits cleanly along that line, and it splits the wrong way for the marketing. The 266-person study titled "Peptides of pineal gland and thymus prolong human life" tested Thymalin and Epithalamin. It does not mention epitalon ¹³. The 15-year follow-up in elderly coronary patients tested epithalamin courses ¹⁴. The circadian melatonin work in elderly subjects tested epithalamin ¹⁵.

Run the same publication-type query over both literatures and the split becomes a number. Five records in the epithalamin corpus carry a Randomized Controlled Trial tag. Zero records in the epitalon corpus do ^43,42.

The conflation is not only a marketing habit; it is baked into reference data. PubChem's synonym list for the tetrapeptide CID 219042 includes both Epithalamin and Epithalamine ⁷. If the chemical databases blur it, a vendor page will certainly blur it, usually in the direction of borrowing the extract's mortality figures for the peptide in the vial.

One more source-level oddity belongs here rather than in a footnote: the 15-year follow-up paper's title says the geroprotector is from the pituitary gland while its own abstract says pineal gland ¹⁴. Both strings are in the same indexed record. It is quoted here as published.

The evidence table below has 32 rows. Each carries the species it was measured in, the grade of the design, and a provenance value: originating group, collaboration, or independent. Five rows record things that do not exist and are printed at the same weight as the findings.

Read at a glance, the file looks like this. The mechanism work is in cells and it is genuinely interesting. The lifespan and tumor work is in mice, rats and flies and is mixed, including outright null results on mean lifespan. The human work is thin, uncontrolled and almost entirely about a different substance. And 21 of the 27 non-absence rows come from the originating network or a collaboration with it.

That is not a dismissal. It is a map. A compound can be worth studying and still have no evidence that supports what is written on the label of the vial someone is about to inject.

This is the claim epitalon is sold on, and it is the claim that most rewards being split into grades.

In vitro, in human cells, the answer is yes, and it now has independent support. The original 2003 experiment added epithalon to telomerase-negative human fetal fibroblast culture and reported induction of the telomerase catalytic subunit, telomerase enzyme activity and telomere elongation ⁸. A 2004 follow-up in fibroblasts from a 24-week fetus reported telomeres relengthening in senescent cells and 10 extra divisions past the control's limit of 34 passages, which is where the phrase overcoming the Hayflick limit comes from ⁹. Both were done by the originating group.

Twenty-two years later, an independent laboratory at Brunel University London reproduced dose-dependent telomere lengthening in human cell lines, quantitatively, with hTERT expression and telomerase activity in normal cells and alternative lengthening of telomeres in cancer cells ¹⁰. This is the most important recent development in the file and it is worth saying without hedging: the in vitro telomerase claim has now been replicated by people with no stake in it.

Three details from that same paper belong next to it, because they are in the paper and not on any vendor page. At its lowest concentration the peptide SHORTENED telomeres in one cancer line. The authors name their main limitation as being a two-dimensional in vitro study and call for in vivo animal work. And the article carries a published correction replacing three figures, issued about two months after publication; it is a correction, not a retraction ^10,11.

The lengthening seen in the cancer lines ran through alternative lengthening of telomeres, a mechanism cancer cells use to keep dividing ¹⁰. The authors read the normal-cell result as reassuring. A reader deciding whether to inject a telomerase activator should see the cancer-cell result too, stated neutrally: it happened, in a dish, in two breast cancer lines.

In people, the answer is that nobody has measured it. There is no study in which epitalon was given to a person and their telomeres were then measured ⁴¹. The study usually cited for the human claim treated blood lymphocytes from 11 men in a dish. Telomere length changed significantly in 7 of them: it increased in 5, by 41, 55, 156, 18 and 76 percent, and DECREASED in 2, by 37 and 15 percent ¹². The authors describe the pattern as a tendency toward normalization, rising in donors below the group mean and falling in donors above it ¹². That is a different claim from lengthening, and it is the claim the source actually makes.

The mortality numbers are real publications. They are also, all of them, about Epithalamin.

The largest is a clinical assessment of 266 elderly people over 6 to 8 years, with the bioregulators given in the first 2 to 3 years. It reports mortality reductions of 1.6 to 1.8 fold in the Epithalamin group, 2.0 to 2.1 fold with Thymalin, 2.5 fold with both, and 4.1 fold in a subgroup given both annually for six years ¹³. The abstract states no allocation method, no confidence intervals and no p values, and the paper is authored by the developers of both preparations at the institute that created them ^13,44.

The strongest single study is the 15-year follow-up: 39 elderly coronary patients given six courses of epithalamin over three years alongside basic therapy, against 40 patients on basic therapy alone, with significantly lower mortality reported in the treated group ¹⁴. It describes itself as a randomized comparative study. That is a genuine randomized signal and it deserves to be called one. It is also 79 people, in one center, from the same research network, about the extract rather than the peptide.

What these studies can show: that a research group treating elderly patients with pineal peptide preparations over years reported better survival and better cardiovascular indices than their comparison groups. That is a real observation and it is why the compound is interesting at all.

What they cannot show: that epitalon does this. None of them tested it. They cannot show a dose, because the abstracts do not state one. They cannot separate the peptide effect from the effect of being in a long-running clinical program with regular follow-up. And they have not been replicated by anyone outside the network that ran them ⁴⁴.

One finding inside this group is worth carrying forward because it undercuts a marketing line rather than supporting it. In healthy elderly subjects, epithalamin raised night-time melatonin only in people whose pineal function was already low, and melatonin TENDED TO FALL in people whose pineal function was normal ¹⁵. The reported effect is a normalization, not an increase.

The rodent literature is the largest positive body of evidence for this compound, and it is more interesting when the nulls are left in.

In female SHR mice treated monthly from 3 months of age until natural death, epitalon did not change mean lifespan at all ²¹. What it did change was narrower: the lifespan of the last 10 percent of survivors rose 13.3 percent, maximum lifespan rose 12.3 percent, chromosome aberrations fell 17.1 percent, and leukemia incidence fell 6-fold with no change in total tumor incidence ²¹. That is a real result. It is not life extension.

Two rat studies point the same way with a twist. Under a standard day and night cycle the peptide did not change lifespan. Effects appeared only in rats whose lives had already been shortened by constant or northern natural illumination ^25,26. Read plainly, the geroprotector rescued a stressor rather than extending a normal life.

The strongest positive lifespan result, 13.5 percent longer mean lifetime in HER-2/neu transgenic mice, sits on a published dose of 1 mg subcutaneously ²². Every neighbouring study in the same literature uses 0.1 to 1.0 micrograms. That is roughly a thousandfold gap, and it most likely reflects a units error in the source. It is reproduced here as published rather than quietly corrected, and it is flagged wherever the number appears.

The tumor work is more consistent than the lifespan work. Fewer and smaller mammary tumors in transgenic mice with a 3.7-fold fall in HER-2/neu mRNA ²³; colon tumors per rat down from 4.1 to 2.7, though carcinomas still developed in 90 to 100 percent of animals ²⁴; no metastases in treated C3H/He mice against 3 of 9 tumor-bearing controls ²⁹; 17.9 to 30.1 percent fewer chromosome aberrations across three mouse strains where melatonin did nothing ³⁰.

One control result from the same experiments deserves its own line. The comparison peptide Vilon significantly INCREASED mammary cancer incidence in the mice that received it ²³. Short peptides are not benign as a class, and the same laboratory that reported the good result reported that one.

The nearest thing to human-relevant physiology is a primate study: evening melatonin synthesis stimulated and the cortisol rhythm normalized in senescent rhesus monkeys, with no animal count in the abstract ²⁸. And the oldest positive result is in fruit flies, where lifespan rose 11 to 16 percent with no dose-response across five orders of magnitude, which the authors themselves remark on ²⁷.

Strip out the extract studies and the cell-culture studies, and the list of research in which epitalon itself entered a human body is one paper long.

It is a 2002 report on retinitis pigmentosa, and it is the only record in the whole epitalon corpus carrying a Clinical Trial publication tag ^18,43. Its abstract reports "a positive clinical effect in 90% of the cases" and does not state how many patients there were, what the control was, how they were allocated, or what was measured. It reports a Campbell rat experiment and the patient result in the same four sentences.

Everything else described as human is cells from humans, treated in a dish. Lymphocytes from people aged 76 to 80 showing chromatin decondensation ³¹. Lymphocytes from 11 men showing bidirectional telomere changes ¹². Cultures from tuberculosis and breast cancer patients showing partial protection against genome instability ^34,33. Lymphocytes from people aged 75 to 88, in the Tbilisi work, showing the same chromatin effect the originating group reports ³². These are legitimate experiments. They are not evidence about what happens to a person.

And the absences are as informative as the presences. Zero randomized trials of epitalon. Zero meta-analyses. Zero systematic reviews ⁴³. Zero registered studies on ClinicalTrials.gov ⁶. Zero studies from any laboratory outside the originating network in which epitalon was given to a person ⁴⁴.

Five claims carry most of the marketing for this compound. Each traces to a real source, and each source says something narrower than the claim.

"Activates telomerase" traces to a 2003 experiment in telomerase-negative human fetal fibroblast culture, since independently reproduced in human cell lines ^8,10. True, in a dish.

"Lengthens telomeres in humans" traces to blood lymphocytes from 11 men treated in a dish, where telomere length rose in 5 donors and fell in 2 ¹². The source calls it a tendency toward normalization.

"Reduces mortality" and "proven to extend human lifespan" trace to cohort studies of Epithalamin, the pineal extract, run and reported by the developers of both preparations ^13,14. Those studies did not test epitalon.

"Restores melatonin" traces to work in which the extract raised melatonin only in people whose pineal function was already low, and lowered it in people whose function was normal ¹⁵, plus a primate study with no stated animal count ²⁸.

"No side effects" traces to sentences inside efficacy papers written by the originating group, not to any safety study ^17,29.

"Extends lifespan in animals" traces to a literature whose best-designed mouse study found no change in mean lifespan ²¹.

There is no human pharmacokinetic profile of epitalon to summarize. Across 167 archived records, not one reports a human half-life, bioavailability figure, Cmax or AUC for any route ⁴¹. A 2026 review preprint states the same absence independently, and is cited here as corroboration of a gap rather than as evidence of an effect ⁴⁰.

This is not a small gap. Without a pharmacokinetic study nobody can say what fraction of an injected dose survives, how long it persists, or whether two people using the same stated dose reach remotely similar exposure. Every one of the chips below is an honest blank.

Epitalon reaches American buyers as a lyophilized powder in a sealed vial, reconstituted with bacteriostatic water and injected subcutaneously. There is no oral form with any evidence behind it, no nasal form in the literature, and no approved form of any kind.

The vial has no product label in the regulatory sense: no assigned strength, no expiry set by any authority, no directions for use, and no manufacturing standard behind the number printed on it ^5,1. When a reconstitution calculator on any site tells a buyer their vial contains a given number of milligrams, that figure came from the seller and from nowhere else.

It is worth noting how ordinary this is even in research. The independent 2025 laboratory bought its epitalon from a research-chemical vendor and recorded dissolving 10 mg in 4 mL of bacteriostatic water to reach 2.5 mg/mL ¹⁰. That is the same arithmetic, on the same kind of material, in a university laboratory.

There is no established human dose of epitalon. No dose-finding study has ever been published, no approved product exists from which a strength could be taken, and the doses that appear in the literature disagree by up to four orders of magnitude ⁴¹.

The rodent studies cluster at 0.1 to 1.0 micrograms per animal, given on five consecutive days each month or five days a week ^21,24,29. One outlier states 1 mg per mouse ²². The cell-culture work runs 0.1 to 1.0 micrograms per mL, or 0.1 mM ^10,35. The human reports state no dose at all: the retinitis pigmentosa paper gives none, and the epithalamin cohort studies describe courses without stating milligrams ^18,13,14.

The study-dose explorer on this site shows every regimen with the study that used it, its species, its design and its provenance. It computes nothing that looks like a recommendation, because there is nothing to recommend from.

Nobody knows, and that is the accurate answer rather than a cautious one.

No dedicated human safety study of epitalon exists. No toxicology study exists. No adverse-event surveillance exists. No long-term human data at any dose exists ⁴¹.

The reassuring statements that circulate are real sentences from real papers, and they are not safety findings. "Having no side effects" appears inside a 2007 efficacy abstract from the originating group ¹⁷. "Long-term exposure to Epitalon in small doses did not show any toxic effect" appears inside a 6.5-month mouse carcinogenesis study whose design measured tumors, not toxicity ²⁹.

Against them sit three things. FDA states it "lacks sufficient information to know whether the drug would cause harm if administered to humans" and flags immunogenicity risk from aggregation and peptide-related impurities in compounded preparations ¹. The independent 2025 work found telomere lengthening in breast cancer cell lines through alternative lengthening of telomeres, the mechanism cancer cells use to keep dividing ¹⁰. And in the originating group's own mouse experiments, the comparison peptide Vilon significantly increased cancer incidence, which is a direct demonstration that short peptides are not safe as a class ²³.

Add the absence of any pharmacokinetic study ⁴¹ and the position is simple: the exposure is unmeasured, the safety is unstudied, and the product has no quality standard.

There is no systematic adverse-event record for epitalon, because no study was designed to collect one and no surveillance system covers a research chemical.

The literature contains assertions of absence rather than counts of events: no side effects in one abstract, no toxic effect in another, both from within efficacy papers by the originating network ^17,29.

An absence of reported side effects in a literature that never looked for them is not a safety profile. It is a blank page. FDA describes the same blank page in its own words when it says it lacks sufficient information to know whether the drug would cause harm in humans ¹.

No interaction study of epitalon with any drug has been published. With no pharmacokinetic data at all, there is not even a mechanism-level basis for predicting one ⁴¹.

The populations never studied are almost everyone: no data in pregnancy or breastfeeding, none in children or adolescents, none in kidney or liver impairment, none in people taking any specific medication, and none in anyone with a current or past cancer diagnosis, which the alternative-lengthening finding in cancer cell lines makes a live question rather than a formality ¹⁰.

The one signal the literature does offer about combinations is indirect and cautionary: in the same mouse experiments, the related short peptide Vilon made cancer outcomes significantly worse ²³. Peptide effects in this family are not interchangeable and not uniformly benign.

The melatonin work adds a second reason for caution about stacking. In elderly people the extract raised night melatonin in those with low pineal function and lowered it in those with normal function ¹⁵. A compound whose direction of effect depends on baseline is a poor candidate for casual combination with sleep or hormone products.

It comes from research-chemical suppliers. There is no approved manufacturer, no compounding pathway currently open to it, and no product-level quality standard anywhere in the chain.

The compounding route is the part most often misunderstood. Epitalon sits on FDA's safety-risks page under substances nominated but withdrawn ¹. Both nominations to allow it in 503A compounding were withdrawn by the nominators, and FDA presented it to the advisory committee anyway with a proposal not to include it ². It appears in no category of the interim list ⁴.

What a buyer therefore receives is a sealed vial of powder whose mass, purity and identity rest entirely on the seller's word. FDA's specific technical concern about this class of preparation is aggregation and peptide-related impurities, which are exactly the failure modes a certificate of analysis from the seller is least able to rule out ¹.

For calibration, note what a university laboratory does: the independent 2025 replication also bought from a research-chemical vendor, and its methods section reports no certificate of analysis, purity figure or independent identity confirmation ¹⁰.

No stability data for epitalon has been published: no shelf life for the lyophilized powder, no in-use stability for a reconstituted vial, and no compatibility data for any diluent ⁴¹.

The general handling described in the research literature is the ordinary handling of a lyophilized peptide: powder kept cold and dry, reconstituted with bacteriostatic water immediately before use, and refrigerated after reconstitution ¹⁰. That is laboratory practice, not a storage instruction from any authority, and no expiry attaches to it.

Because FDA's stated concern for this class is aggregation and peptide-related impurities ¹, storage matters more here than for a compound with a real label, and there is less to go on.

What is unresolved about epitalon is structural rather than incidental, and the full evidence-limits block follows below. In short: 66.5 percent of the literature comes from the network that created the compound ⁴⁴; no laboratory outside it has ever given epitalon to a person ⁴⁴; the mortality evidence belongs to a different substance ¹³; and there is no human pharmacokinetic study ⁴¹, no registered trial ⁶, no randomized trial of the peptide ⁴³, no dose-finding study ⁴¹ and no safety or toxicology study of any kind ¹.

The questions below are the ones this compound actually generates, answered at the grade the evidence supports.

Every source cited on this page is listed below with its type and year, and every URL was checked for an HTTP 200 response on 2026-08-14. The machine-readable citation manifest ships alongside the page.