GHK-Cu vs retinol vs vitamin C: where the three headline numbers come from

The 55.8 percent printed on the copper peptide box never happened to anyone's face. It is the gap between two serums in a 39-person trial: the peptide serum cut wrinkle volume 24 percent from baseline over 8 weeks, and the plain control serum on the other side of the same faces managed 15 percent with no peptide in it ¹. The paper's own word for 55.8 is relative, and that word carries the whole meaning ¹. Retinol and vitamin C carry numbers built on different scales again, so lining the three up as a ranking compares three different kinds of measurement. So this page takes each headline number and reports where it came from, who published it, who paid for it, and what it says when you open it.

The three headline numbers, side by side with their sources

Each row is the number people repeat, followed by what the source says. Nothing here is an adjective.

Of the three compounds compared here, only GHK-Cu is a peptide. KPV sits beside it below because our report scores it and it turns up in skin and inflammation contexts, not because it belongs on your face. The cross-compound roster lives on best peptides for skin, GHK-Cu is combined with epithalon and SS-31 on the triple anti-ageing stack, and GHK-Cu with KPV is the KLOW stack.

GHK-CuThe regeneration peptide

KPVThe gut anti-inflammatory

Browse the full reference library →

Disclosure: some outbound links on PeptidesDNA, including on our provider comparison pages, are affiliate links. If you buy through one we may earn a fee at no extra cost to you. It does not change what this page says about the evidence, and we do not sell any compound discussed here.

Where does the 55.8% GHK-Cu figure come from?

It comes from a 2016 trial in the Journal of Aging Science: a randomised, double-blind, split-face study run at Dermatest GmbH in 40 women aged 40 to 65, of whom 39 were analysed, in which every participant applied the GHK-Cu serum in its nano-lipid carrier to the right side of the face and the comparator to the left for 8 weeks ¹. Wrinkle depth and volume were measured with the PRIMOS stripe-projection system on one single large wrinkle chosen per person inside a 30 by 40 mm area near the corner of the eye ¹. The side assignment was fixed rather than randomised, which is a real limitation: the two sides of a face are not identical, and nothing in the design corrects for that. The sentence everybody is quoting, verbatim from the paper, is this one.

Compared to Control there was a 55.8% relative reduction wrinkle volume with the NRFS serum (p<0.01) and a 32.8% decrease in wrinkle depth (p=0.0123).Badenhorst T, Svirskis D, Merrilees M, Bolke L, Wu Z. J Aging Sci 2016;4:166

NRFS is the paper's label for the GHK-Cu serum in its nano-lipid carrier, so that sentence is about the copper peptide product. Relative reduction is the paper's own term and it carries the whole meaning. In the group that produced the 55.8 figure, the GHK-Cu serum cut wrinkle volume 24.1 percent from baseline (plus or minus 8.6) and the plain control serum on the other side of the same faces cut it 15.0 percent (plus or minus 5.2). The 55.8 percent is the gap between those two results, which the paper computes from absolute wrinkle volume: a 0.4 mm3 difference between the two sides of the face, as Table 3 reports it ¹.

Two more details sit in the same tables. The copper-peptide serum's advantage in wrinkle depth over the commercial comparator did not reach statistical significance, coming in at 23.4 percent more than the comparator with p=0.0577 ¹. And the paper gives three different p-values for its own headline: p<0.01 in the prose, p<0.001 in the abstract, and <.0001 in Table 3, all for the same comparison ¹.

Who published that paper, who paid for it, and why both change its weight

The study's own disclosure reads that it was funded by Snowberry New Zealand Ltd in collaboration with Callaghan Innovation, grant ENDU1101, and states that Snowberry had no involvement in the collection or analysis of the data ¹. Snowberry sells GHK-Cu skincare. One co-author, Liane Bolke, is at Dermatest GmbH, the contract institute that ran the trial ¹. None of that makes a result wrong, and this page would not be applying its own test if it disclosed the funding line for the tretinoin, vitamin C and hexapeptide papers and stayed quiet about the one paper the whole comparison is built on.

The venue is the second half. The Journal of Aging Science was published by OMICS. A US federal court in the District of Nevada entered a $50.1 million judgment against OMICS Group Inc., iMedPub LLC, Conference Series LLC and Srinubabu Gedela on 29 March 2019, announced by the FTC on 3 April 2019; the court found that many articles were published with minimal or no peer review, contrary to what the companies told authors, and the order bars them from misrepresenting that their journals engage in peer review ⁴. That is a finding about the publisher's claims, not a finding about this paper, and nobody can conclude from it that this particular trial went unreviewed. What can be said is that the one venue standing behind the most repeated number in copper peptide skincare is a venue a federal court found made false claims about the thing a venue exists to do.

The paper is also not indexed in PubMed, and it reaches the public mostly through a 2018 review in the International Journal of Molecular Sciences written by Loren Pickart, who discovered GHK, and Anna Margolina. Both authors are at R&D Skin Biology in Bellevue, Washington, which sells copper-peptide skincare; the paper reports no external funding and lists no competing interests ⁸.

Two other numbers that travel from that review are worth correcting at the source. GHK does not affect 31.2 percent of human genes in the sense people repeat: the review counts human genes whose activity moved by 50 percent or more in a cell-line database, and the authors picked the 50 percent cutoff themselves because it gave the best correlation with their own biological data ⁸. And the claim that GHK activates 84 DNA repair genes appears in the source as GHK having affected 84 genes associated with DNA repair and other processes ⁸. The mechanism itself, stated properly, is on how GHK-Cu works.

How to check a skincare number in about two minutes

Every correction on this page came from four questions, and none of them needs a subscription. New to peptides in general rather than to this one number? Start at peptides for beginners.

1. 1Find the paper behind the numberA number with no paper is a marketing claim. Brands that have a study usually name it somewhere. If searching the exact figure returns only other brands repeating it, you have your answer already.
2. 2Paste the title into PubMedPubMed is the US National Library of Medicine's free index of biomedical literature. Absence does not make a paper wrong, and the 2016 copper peptide trial discussed above is readable by its DOI while not being indexed, but presence tells you a journal cleared a basic bar and the record can be found by anyone.
3. 3Read the author affiliations and the funding lineThey sit at the top and the bottom of every paper, before and after the science. An author employed by the company that makes the tested ingredient does not make a result wrong. It is still the single most useful thing to know before you read one.
4. 4Ask whether the number is absolute or relativeAn absolute number is what changed in the people who used the product. A relative number is the gap between two groups or two sides, and it is always the larger of the two. A 24.1 percent change from baseline and a 55.8 percent relative reduction can be the same result stated two ways.

Retinol's number is a shift on a grading scale

The 2025 meta-analysis pooled 8 randomised trials and 1,361 patients, with follow-up from 16 weeks to 2 years, and reported improvement as mean differences on clinical grading scales: 0.412 for fine wrinkles (95% CI 0.233 to 0.590) and 0.245 for coarse wrinkles (95% CI 0.119 to 0.370), both P<0.001 ². A grading scale means a dermatologist scored each face against a reference set of photographs, so the result is a shift in score, not a percentage off anything. The 95% CI range after each figure is the band the true effect is most likely to sit in, and because neither band crosses zero, the improvement is unlikely to be chance. It is checkable, it is pooled, and the two authors are hospital doctors who declared no competing interests ².

The second fact about that number matters as much as the first. It tested prescription tretinoin. Over-the-counter retinol is a precursor, meaning skin has to convert it into retinoic acid, the active form, before anything happens, so the evidence for the prescription drug does not transfer intact to a shop-bought serum. Retinol has the strongest checkable trial record of the three, attached to a molecule most people comparing these bottles are not applying.

What comes with it: retinoid irritation is common and expected, meaning redness, peeling, dryness and stinging, especially in the first weeks, and retinoids increase sensitivity to sunlight, so daily sunscreen is part of using one. Topical retinoids are generally avoided in pregnancy and while trying to conceive, and that decision belongs with a doctor before starting rather than after.

Vitamin C's randomised-trial record is 139 volunteers

The prospective randomised trials that a 2023 systematic review found in Medline, CENTRAL and Scopus up to May 2022 come to 7 publications and 139 volunteers in total, and roughly half of that work is about melasma rather than wrinkles ³. The review reports that treated skin appeared smoother and less wrinkled, which the authors say was supported by biopsy data, and that pigmentation lightened significantly, alongside the line most often quoted against it: "Hydration improved equally in the vitamin C and placebo-treated sites." The authors conclude that long-term use may be needed to achieve noticeable changes and that more studies are needed to establish the ideal concentration ³. Small studies show a real signal on pigmentation and skin texture. The record is far too small to settle a strength, a frequency or a dose response.

The 2025 trial in which a peptide beat retinol

A randomised, double-blind trial published in 2025 compared a cyclized hexapeptide-9 serum against a retinol serum and a plain vehicle, meaning the same serum base with no active ingredient in it, over 56 days, with 96 people enrolled and 91 finishing across the three arms, and the peptide came out ahead ⁷. Both serums were 0.002 percent ⁷, and that strength decides what the result means. It is a fair contest between two serums at matched concentration, and nothing like the 0.3 to 1 percent retinol sold on a pharmacy shelf. All eight authors are employed by the two companies behind the tested peptide, per the paper's own affiliation list, and the paper lists no competing interests and no specific funding ⁷. The peptide tested was not GHK-Cu.

Does GHK-Cu actually work? The evidence that exists, and what it was for

The most convincing GHK-Cu trial in the indexed literature is a wound trial, and in that trial the gel outperformed its placebo. A 1994 multicentre, randomised, evaluator-blinded, placebo-controlled study of GHK-Cu gel in diabetic foot ulcers reported median plantar ulcer closure of 98.5 percent against 60.8 percent for the same gel without the peptide (p<0.05), closure roughly three times faster, and ulcer infection in 7 percent of the treated group against 34 percent (p<0.05) ⁵. In ulcers larger than 100 square millimetres the gap widened: median closure 89.2 percent against minus 10.3 percent for the gel without the peptide, meaning those ulcers were on average getting bigger ⁵. What was tested was a 0.5 percent GHK-Cu gel developed as an investigational prescription product, applied inside an aggressive standardised protocol with sharp debridement at entry, pressure-relieving footwear and diabetes education, and the paper states that treatment had to begin immediately after debridement to get the effect ⁵.

So the copper peptide's strongest randomised evidence is about a clinically supervised wound gel that was never approved, not about a bottle on a shelf. That is a more interesting verdict than saying there is no evidence, and it is not a verdict about wrinkles. Which peptides have real randomised evidence for wound and soft-tissue repair, compound by compound, is the subject of best peptides for healing.

One randomised, blinded facial trial of GHK-Cu is indexed in PubMed, and it was run on skin recovering from carbon dioxide laser resurfacing rather than on intact ageing skin, with 13 patients completing it ⁶. Every patient improved significantly in wrinkles and overall skin quality, and computer analysis of standardised photographs and blinded evaluators found no difference between the group using GHK-Cu and the group that did not, while the patients' own questionnaire favoured GHK-Cu at P=.04 ⁶. Both halves belong in any honest summary. Reporting only the half that suits the seller is the same error as reporting only the half that suits the sceptic.

What does not exist is an independently funded randomised trial testing GHK-Cu for wrinkle reduction on intact ageing facial skin with a result anyone can check. An independent review from a university pharmacy school with no copper-peptide product to sell states both sides of that plainly, saying that on cell-culture evidence GHK "can be considered as an anti-wrinkle ingredient" while noting "there is a surprising absence of clinical studies using them" ⁹. The full study record for the compound sits on the GHK-Cu monograph. The claimed benefits are assessed one by one on GHK-Cu benefits, which starts from the fact that your own circulating GHK runs at roughly 200 ng/mL in your twenties and about 80 ng/mL by sixty.

That timeline, and how to photograph your own face so a comparison means something, is the subject of GHK-Cu before and after. For the condition behind all three bottles, skin aging starts from the measured baseline: type I collagen production drops roughly 1 percent per year after age 25, which is the slope any of these compounds is being asked to bend.

In the video below, Dr. Ashley Froese revisits her own position on copper peptides in public.

Dr. Admits: I Was WRONG AGAIN About GHK-CuDr. Ashley Froese · plays on YouTube

If you already own a copper peptide serum, none of this says throw it out. It says the number on the box is a gap between two serums rather than a change on a face, that the serum in its own trial did move wrinkle volume about 24 percent in 8 weeks, and that if you want a topical with a checkable trial record for wrinkles specifically, that is prescription tretinoin, which is a conversation with a clinician rather than a purchase.

Can you use copper peptides with retinol or vitamin C?

No published trial has tested these layering routines on people. What follows is formulation chemistry and what has been measured about it, not trial evidence about faces.

The common warning that acidic vitamin C destroys copper peptides is not supported as it is usually stated. A preformulation study from the School of Pharmacy at the University of Auckland found GHK-Cu stable in water and in buffers across pH 4.5 to 7.4 for at least two weeks at 60 degrees C, and found it vulnerable to breakdown under basic and oxidising conditions and "to a lesser extent acidic stress" ¹⁰. Acid was the weakest of the three stressors tested, not the main threat, and one identified breakdown product is histidine, one of the three amino acids the peptide is built from ¹⁰.

Size is also not the barrier people think it is. GHK, the copper-free tripeptide, weighs 340.38 daltons, and the copper complex sold in serums weighs about 403 daltons, both under the 500-dalton rule of thumb for passing through skin. What limits it is that GHK-Cu strongly prefers water to oil: measured distribution coefficients between an oily phase and a water phase run from log D minus 2.38 to minus 2.49 across pH 4.5 to 7.4, which the authors describe as "the highly hydrophilic nature of GHK-Cu" ¹⁰. A water-loving ingredient has to be carried through an oily barrier, which is why the delivery vehicle on the label decides more than the percentage does. An independent review reaches a more favourable read on how well it crosses in practice, calling GHK and its fat-modified version "effective and relatively skin permeable" ⁹, so penetration is contested rather than settled.

What is approved, and what is only sold

Prescription tretinoin is an approved medicine for photodamaged skin. Over-the-counter retinol, topical vitamin C and GHK-Cu are sold as cosmetic ingredients, and none of them is an FDA-approved drug for wrinkles. The US legal picture for peptides more broadly, including the injectable forms, is on are peptides legal in the US.

One regulatory note for readers outside the US. Commission Regulation (EU) 2024/996 caps retinol at 0.3 percent retinol equivalent in leave-on and rinse-off products and at 0.05 percent in body lotion. Non-compliant products could not be placed on the Union market from 1 November 2025 and may not be made available from 1 May 2027, so stock manufactured before the cut-off can still legally sit on a shelf today ¹¹. PeptidesDNA cannot sell its report into the EU, the EEA or the UK. What European readers can legally buy, and from whom, is covered on peptide pharmacy Europe.

What genetics can and cannot tell you about three bottles on a shelf

No trial of GHK-Cu, retinol or vitamin C has sorted the people who responded from the people who did not by genotype. That means no gene test predicts how a face will respond to any of the three, and a company offering you one is ahead of its own evidence. What genetics supports here is background rather than prediction. MC1R sets which pigment type skin makes, and therefore how much damage a given dose of ultraviolet light does, and ultraviolet exposure is the largest changeable driver of the collagen loss all three of these compounds are sold against. The action that follows is sun protection, not a different serum.

Knowing your MC1R status could reasonably change how seriously you take daily sunscreen. It should not change which of these three bottles you buy.