One compound with two evidential lives. Plus the most-quoted statistic in the field, and why moving a third of the genome is not the selling point it sounds like.
GHK-Cu is the odd one out in this section. The other healing compounds here have thin human evidence, full stop. This one has a real human clinical record and none at all — and which applies depends entirely on how you put it into your body.
That split is the whole article.
As a cream or serum, GHK-Cu appears on ingredient lists as Copper Tripeptide-1. It is a normal cosmetic ingredient, unrestricted, with decades of use behind it and randomised controlled trials supporting it.
As an injection, it is a research chemical. The FDA put it on the Category 2 bulk drug substances list in 2023, which shuts compounding pharmacies out of it. The stated reason was insufficient human safety data for injected routes. There are no published human trials of injectable GHK-Cu, for safety or for effect. None.
Same molecule. One route has trials and a shelf at the chemist; the other has animal data and forums.
Anyone weighing the two should notice something. The clinical results quoted in favour of the injection were, almost without exception, produced by the cream.
GHK-Cu is glycyl-L-histidyl-L-lysine bound to a copper ion — three amino acids and a metal. Your body makes it and carries it in plasma, saliva and urine.
Loren Pickart pulled it out of human plasma albumin in 1973, after an odd observation. Liver tissue from older donors, bathed in blood from younger donors, started behaving like younger tissue. GHK turned out to be the factor responsible.
Levels fall sharply with age — roughly 200 nanograms per millilitre at 20, down to around 80 by 60. More than half gone. That decline is the entire premise for supplementing it, and it is a good premise. It is not, by itself, evidence that adding it back works.
The copper is not decorative. Several enzymes cannot function without it, and the one that matters most here is lysyl oxidase, which cross-links collagen and elastin into durable networks.
Without enough copper, new collagen stays weak and breaks down easily. You can stimulate all the collagen you like and still end up with poor material. GHK-Cu delivers usable copper straight to tissue in a bound form, rather than as free copper in the blood, which is toxic.
That pairing — build the collagen and deliver what is needed to cross-link it — is a better argument than either half alone.
You will see this everywhere: GHK changes the expression of 31.2% of all human genes by 50% or more.
It comes from the Broad Institute's Connectivity Map, a database tracking how compounds shift gene activity across thousands of profiles. In the cancer analysis, 1,309 active molecules were examined. GHK reversed expression in 70% of the overactive genes in a metastasis-linked colon cancer signature. It also raised 47 DNA repair genes and 14 antioxidant genes, all at tiny concentrations.
Two things are worth saying about it.
First, these findings come from review papers, not original trials. And the reviews are largely written by Pickart, the same researcher who found the compound in 1973 and has published much of the literature since. That is not an accusation. It is a reason to weight a review differently from one written by people with no stake in the answer.
Second, and more interesting: a compound that moves a third of the human genome by half is not obviously reassuring. The claim is that it shifts expression toward healthier patterns — repair up, inflammation and breakdown down. Perhaps. But "changes almost everything" describes a very blunt instrument, and it gets quoted as though breadth were the same as precision.
One mechanism deserves picking out because it is unusually sensible.
MMPs are the enzymes that break collagen down. The naive approach is to block them. GHK-Cu does not. In cell culture it raised MMP1 and MMP2 — and simultaneously raised TIMP1, the natural inhibitor, at every concentration tested.
The result is a higher inhibitor-to-enzyme ratio. In practice that means turnover, not build-up: old damaged collagen is cleared while new collagen goes in. And it was that ratio shift that tracked with more collagen and elastin, not the raw enzyme numbers.
It also promotes decorin, a small proteoglycan that organises collagen fibres. Again the same theme — make the material and arrange it properly.
This is the solid ground.
Badenhorst and colleagues, 2016. Randomised, double-blind, 40 women aged 40 to 65, applied twice daily for eight weeks. Wrinkle volume fell 55.8% and wrinkle depth 32.8% against a control serum, and it beat an established competitor peptide by 31.6% on wrinkle volume. The lab portion confirmed increased collagen and elastin, with the TIMP-to-MMP ratio moving favourably. Small sample, and the delivery used lipid nanocarriers rather than a plain cream.
Finkley and colleagues, 2005. 71 women with mild to advanced photoageing, cream twice daily for 12 weeks. Skin density and thickness up, sagging and fine lines down, no adverse events. Published as a book chapter rather than a journal article, which is a real limitation on how much scrutiny it received.
Abdulghani and colleagues, 1998. The head-to-head everyone cites: collagen production improved in 70% of women on GHK-Cu, against 50% for vitamin C and 40% for retinoic acid. Small study, minor journal. Suggestive rather than settled.
Safety on this route is well characterised. A 2023 review across 12 studies and 512 subjects found transient redness at 4.2% and itching at 2.8%, with no consistent systemic effects.
Fu and colleagues, 2015, 48 rats, ACL reconstruction. Injected GHK-Cu into the joint improved graft healing and reduced knee laxity at six weeks.
By twelve weeks, after treatment stopped, the benefit was gone.
The paper's own word for the effect is transient. That is an unwelcome finding for anyone running a course and expecting the result to stay. It does not prove the same happens in people. But it is the closest thing to a controlled injectable outcome study we have, and it raises a question nobody has answered.
The most universally reported thing about injecting this compound is that it stings. Burning, redness, a raised bump, sometimes for 15 to 20 minutes. Occasionally a blue-green tint at the site.
There is a mechanism. Free copper can briefly separate from the GHK at the injection site, producing mild oxidative activity and a histamine release. It is a property of the chemistry, not a sign of a bad batch.
The blue colour of the solution has the same origin, and is expected. A reconstituted vial that is not blue would be the thing worth questioning.
The copper drives contraindications the other compounds in this section do not have.
Wilson's disease and any copper metabolism disorder rule it out outright. These are conditions where copper builds up, and the last thing needed is more of it. Active infection at the injection site is another, because copper can feed bacterial growth. High-dose zinc is worth a mention too, since zinc and copper compete for absorption.
Then the ones shared with everything else here. Active cancer or tumours, because it builds blood vessels. As with BPC-157, and unlike TB-500, the gene work points the other way, showing reversal of metastasis-linked expression. The mechanism concern stands anyway. A cancer history needs an oncologist first.
Caution with pregnancy and breastfeeding (no data), autoimmune disorders, severe liver disease, kidney impairment, immunosuppressants and blood thinners.
On copper load specifically: at the doses in practical use, toxicity is not a live concern. The estimated LD50 sits around 330 mg/kg, which for a 70 kg adult would be roughly 23,000 mg — orders of magnitude above anything anyone takes.
The topical case is the strongest human evidence for any compound in this section. Randomised, controlled, repeated, with a well-mapped safety profile. If the goal is facial skin, that route is not a compromise. It is the evidenced option.
The injectable case is an extrapolation. It runs from topical human results, from animal wound and tissue work, and from gene analysis written mostly by the compound's discoverer. It may be right. The one controlled animal study of an injected outcome found the benefit faded once dosing stopped, and nobody has checked whether that holds in people.
Knowing which of those two you are relying on is worth more than any protocol.
This article explains published research and the rules around it. It carries no dose, no protocol, and no claim that any compound treats, cures or prevents a condition in anyone. Where the evidence is thin we say so. See our editorial standards.
Research-grade material, sold for laboratory research use. We link to a single supplier rather than a list, so you know exactly where the recommendation goes. Check the certificate of analysis against what a COA can and cannot show before you buy anything.
Commercial link. Nothing on this page is a dose, a protocol, or a claim that this compound treats or prevents any condition.
KLOW is GLOW plus 10 mg of KPV — that part is simple. The more useful fact is that both are, by mass, mostly copper peptide, and that is the ingredient whose injected form has no human trials.
An open study found sleep normalised in six of seven patients. The controlled study by the same investigator found weak effects and said so. The difference between them is a placebo arm.
It does not raise dopamine — it magnifies what other things do to it. That one property explains why caffeine feels stronger on it, and why MAO inhibitors are a hard contraindication.
