Ipamorelin: There Are Two Human Trials, and Nobody Has Ever Seen the Second One

Most compounds covered here share a problem. The research is thin, or it belongs to a different molecule, or it was all done in rats by one laboratory.

Ipamorelin is not that. A real drug company took it into real human trials, at real doses, in hospitals. The evidence was generated. Then most of it was never shown to anybody.

The compound itself is a chain of five amino acids, made in a lab, built at Novo Nordisk in Denmark. Its foundational characterisation was published in 1998, in the paper that named it the first selective growth hormone secretagogue. (PMID 9849822) It is a designed molecule rather than a fragment of anything found in the body, which already sets it apart from most of this catalogue. It is sold as a ghrelin mimetic, meaning it copies a natural hunger hormone to trigger a growth hormone release, though nobody has ever published how tightly it actually binds to the receptor it supposedly works through.

One distinction the marketing blurs, because people stack these together. Ipamorelin is not the same kind of thing as CJC-1295, sermorelin or tesamorelin. Those act on one receptor, ipamorelin acts on a different one. Both routes end at growth hormone. They are not interchangeable.

The selectivity claim, and the animal it was measured in

Every vendor page says the same thing. Ipamorelin raises growth hormone cleanly, without the cortisol and prolactin that come with older peptides. It is the whole reason anyone picks this one.

That claim comes from one paper, published in 1998, and the only affiliation on it is Novo Nordisk, the company that owned the molecule. The paper is real and says what it is quoted as saying. Its scope is much narrower than the quoting suggests. Its own summary contains the line that governs everything: "The specificity for GH release was studied in swine."

Every measurement of cortisol and prolactin in that study was made in pigs.

There is a detail inside the result that decides how much it is worth. Prolactin did not rise in response to anything in that experiment. Not ipamorelin, and not the older peptides it was compared against, which were GHRP-6 and GHRP-2.

That matters, because one of those comparators has been measured in people. Given intravenously to healthy adults, GHRP-2 raises prolactin, along with ACTH and cortisol. (PMID 9285939) The pig assay recorded no prolactin rise for a compound that demonstrably does raise prolactin in humans.

Picture a metal detector that stays silent for every person walking through the gate, including the one carrying a crowbar. You have not learned the crowd is unarmed. You have learned the detector is not plugged in.

A test that comes back clean for the compounds already known to be dirty cannot certify anything clean.

Nobody has replicated it, and in humans the number is zero. No cortisol or prolactin measurement for ipamorelin has ever been published, at any dose, by any route.

The most repeated fact about this compound is a pig finding, never replicated, from the company that owned it, sold everywhere as a human one.

The trials

Both were run by Helsinn, both in patients whose bowels would not restart after surgery.

The one that was published. NCT00672074. 117 patients, ipamorelin into a vein twice a day for up to a week. The target was how long it took someone to manage a solid meal. Ipamorelin got them there in 25.3 hours against 32.6 on placebo, a gap well within what chance produces. The authors' own conclusion was that there were "no significant differences between ipamorelin and placebo." (PMID 25331030) Nothing else they measured separated either.

On safety, the paper and the FDA read the same data differently. Overall side effects were slightly lower on ipamorelin, which is where "well tolerated" comes from. Reviewing the trial in 2024, FDA noted more low potassium, more insomnia and more high blood sugar on ipamorelin, and two patients in that group died after surgery for colon cancer. FDA says plainly it is unclear whether those deaths had anything to do with the drug, and that matters. Two deaths among very sick post-surgical patients is not a drug signal. What it also is not is the frictionless safety record this is sold on.

The one that was not published. NCT01280344. 320 patients, 45 hospitals, doses up to twice as high. It finished in May 2014.

Three times the size of the published trial, and it has never reported anything. No results in the registry, no journal article, nothing. FDA's own 2024 review mentions it as a thing that exists and reports no outcome from it.

An absence in the research usually means nobody looked. This one is different. Somebody looked, at 320 people, for three years, and the answer has sat somewhere private ever since. The hole in the record is trial-shaped.

Be precise about what is missing, because the registry entry is not blank. It lists all four arms and their doses: 0.03 mg/kg twice daily, 0.06 mg/kg twice daily, 0.06 mg/kg three times daily, and placebo. The doses are public. What has never been described is what happened.

The dose, and the route nobody has studied

Every human dose of ipamorelin ever published went into a vein. No human pharmacokinetic data for injecting it under the skin has ever been published. Animal work by that route does exist, including the mouse study further down this page.

The only human pharmacology study infused it into a vein in healthy men. Growth hormone came out as a single burst peaking around 40 minutes, then faded. (PMID 10496658) Every source quoting a half life of "about two hours" is quoting that study, and almost none mention it was a vein. What an injection under the skin does to that curve has never been published, and growth hormone works in pulses, so the shape of the signal is not a small detail.

The community convention is 200 to 300 mcg under the skin, one to three times a day. Put that next to what humans were actually given.

1

2

3

4

1,000

10,000

mcg per day · log scale

derived from researchcommunity convention

1. 1The community dose, under the skin · 200 to 900 mcg per day200 to 300 mcg per injection, one to three times daily. No published origin, and the route has never been studied in a person.
2. 2Lowest rung of the only human escalation · 210 mcg per dayThe smallest of five doses in the only human escalation study. A starting dose picked to open the study safely, not because it did anything.
3. 3The published trial dose · 4,200 mcg per dayAbout 4.2 mg a day into a vein. This is the dose in the trial that was published, and that trial missed its target.
4. 4The highest daily dose in any registered trial · 12,600 mcg per dayAbout 12.6 mg a day into a vein, in the 320 patient trial. The largest daily total on record, though a 1999 volunteer study gave a larger single dose. Nobody outside the sponsor knows what happened at this one.

Per day, the community dose is 5 to 21 times smaller than the published trial. That is the opposite of BPC-157, where the community number sits above everything the research supports. These are small doses, which is a statement about exposure and not an argument that they work.

So where does 200 to 300 mcg come from? I could not establish it, and I am not going to invent a lineage. Following the citations people offer leads nowhere. One vendor guide credits a journal that has published zero ipamorelin papers, and a university study that does not appear to exist. A 2026 review prints the dose a thousand times too small, crediting a paper published before ipamorelin was first described anywhere. (PMID 42123471)

What nobody knows

Whether it does anything for the reasons people take it. There has never been a human trial for growth hormone levels, body composition, sleep, injury recovery or ageing. And nobody knows what happens at the doses in the 320 patient trial.

What is in the vial. The published finding here is not weak product, it is a different molecule. Two independent lab groups testing illicit material found Gly-ipamorelin, ipamorelin with an extra amino acid stuck on the front, instead of ipamorelin. One found it in black market products and confirmed it against a custom synthesised reference standard. (PMID 29864719) The other tested products seized by Danish customs and reported the same modification in every preparation it analysed, though it does not say how many that was. (PMID 30136411) Neither was measuring vial contents against the label, so this does not tell you what fraction of vials are affected. No published study has measured what Gly-ipamorelin does in a person.

You can check for this one yourself, provided you check the right number. That extra amino acid adds 57.05, so against ipamorelin's 711.9 the altered version has a molecular weight near 768.9.

Those are neutral masses. A mass spectrum reports an ion, so what you want on the certificate is m/z 769.4 where real ipamorelin gives 712.4. Many published methods key on the doubly charged ion instead, in which case the pair is 385.2 against 356.7. A stated molecular weight in the high 760s, or a measured peak at either of those figures, has told on the product.

Whether IGF-1 goes up in humans. This should bother anyone reading for longevity, because IGF-1 is the lasting downstream signal and the thing growth hormone is actually wanted for. No study has ever reported it, at any dose, by any route.

Whether it is safe over time. FDA searched nine databases and reported that it did not identify, in the publicly available literature, any acute toxicity, repeat dose toxicity, genotoxicity, developmental and reproductive toxicity, or carcinogenicity studies of ipamorelin. Its conclusion was that the nonclinical work that does exist is "too limited in scope and duration to inform safety considerations." That is a narrower statement than no animal data existing, because some does. None of it is toxicology.

Whether it raises cancer risk. There is no ipamorelin data in either direction. FDA reasons only by comparison to approved growth hormone drugs, and phrases it as a double negative: there is not enough data to conclude ipamorelin "would not raise safety concerns similar to those associated with approved products that stimulate GH release." That is not a finding of harm. It is a statement that nobody has ruled it out.

One animal finding cuts against the marketing. In mice given twice daily injections under the skin, ipamorelin increased fat pad weight relative to body weight and raised food intake. Growth hormone did the opposite, but only in growth hormone deficient mice; in mice with intact growth hormone it did not change fat at all. Body fat was measured at two weeks, with dosing continued to nine. (PMID 11162489)

Regulatory status

WADA. Named explicitly, not swept in by a category clause, under S2.2.4, growth hormone secretagogues. Prohibited at all times, in and out of competition, and it is a non-Specified substance, which is the more serious tier and carries a four year default sanction. If you are tested, this is disqualifying.

FDA. People still say ipamorelin is Category 2, the tier for significant safety risks. The truth needs two sentences rather than one, because there are two compounding pathways.

On 27 September 2024 FDA moved ipamorelin acetate out of Category 2 under the 503A interim policy, and the reason matters: the two nominators, Wells Pharmacy Network and LDT Health Solutions, withdrew their nominations. Nothing was resolved on the merits. Under the separate 503B policy, which covers outsourcing facilities, ipamorelin acetate remains in Category 2 to this day, added 29 September 2023.

FDA put the question to its Pharmacy Compounding Advisory Committee anyway, on 29 October 2024. There were three votes, not one. First a procedural 12 to 1 to consider the free base and the acetate together, then a vote on each: 0 in favour, 12 against, 1 abstaining for ipamorelin, and the same 0 to 12 with 1 abstaining for ipamorelin acetate. Worth stating plainly because the procedural 12 is a yes count, and it is easy to see a "12" and read it the wrong way round.

It is not part of any approved drug anywhere.

My take

The usual complaint about compounds in this category does not apply here, and that is what makes ipamorelin worth writing about.

Nobody ignored this molecule. Novo Nordisk designed it. Helsinn ran two controlled trials across more than sixty hospitals, with real placebo groups and real patients. That is more genuine human evidence than BPC-157 and TB-500 have between them.

And the outcome is that one trial missed, and the bigger one has never spoken. Somebody knows whether those higher doses did anything, and has known since 2014. That information exists. It is simply not available to anyone deciding whether to inject this.

Set against that, the way the compound is sold is almost a separate subject. A different dose, roughly ten times smaller. A different route, with no human data at all. A different purpose, never tested.

What I would want you to take away is not that ipamorelin does nothing. Nobody has shown that either. It is that the confident version of this compound, the clean and selective one that raises growth hormone and nothing else, is not a summary of the evidence. It is a summary of one paper, about pigs, with the species left out.

If you run ipamorelin, or your clinic prescribes it, I would like to know whether you have ever seen a human trial cited for it, and which one. There are only two. One is negative, and the other has never reported.

---

This content is for educational and informational purposes only and is not medical advice. Peptides discussed are research compounds and may not be approved for human use. Nothing here should be used to diagnose, treat, cure, or prevent any disease. Always consult a qualified healthcare professional before starting any peptide, supplement, or protocol. Individual responses vary. Do not self-administer compounds without proper medical supervision.