Peptide Glow Journal

Peptides for Skin and Hair: The Evidence Gap

We counted the human research for each peptide on the skin side and the hair side. The two lists barely overlap, and products borrow across the gap.

Peptide Glow Journal Editorial · Published 2026-09-06

"Peptides for skin and hair" is one of the fastest-growing searches in this category and one of the least examined claims in it. So we split the question in two. On 2026-09-06 we counted, for each peptide sold into beauty, how many human research records exist on skin terms and how many on hair terms — separately, using the same method for every ingredient. The result is a matrix, and the matrix is nearly empty down its diagonal.

The census

PubMed, 2026-09-06. Each cell is the count of records indexed as human for that ingredient's title and abstract terms, intersected with skin terms (skin, wrinkle, photoaging, dermal, facial) or hair terms (hair, alopecia, follicle, scalp).

Ingredient Human records: skin Human records: hair
Oral collagen peptides 188 17
Copper tripeptide / GHK-Cu 21 2
Acetyl hexapeptide-8 (Argireline) 20 0
Thymosin beta-4 17 6
Palmitoyl pentapeptide-4 (Matrixyl) 9 1
Biotinoyl tripeptide-1 (Procapil) 0 4
Acetyl tetrapeptide-3 (Capixyl) 0 0

Read down the columns and the pattern is immediate. The ingredients that dominate skincare marketing — Argireline, Matrixyl — have essentially nothing on the hair side. The ingredient that dominates hair marketing — Procapil — has nothing on the skin side. These are not competing camps within one field; they are two separate literatures that happen to share the word "peptide".

Why the sets barely overlap

Partly biology: a follicle is a mini-organ with its own cycle, and a wrinkle is a change in dermal matrix and muscle movement. Nothing in principle stops one molecule from acting on both, and copper peptides are proposed to do so.

Mostly, though, it is money. These are not academic research programmes with a question to answer; they are ingredient-development programmes with a product to sell. A cosmetic-actives company developing an anti-wrinkle peptide runs a wrinkle test, because that is the claim its customers will make. A hair-actives company runs an anagen/telogen count. Neither pays to test the other endpoint, so the other endpoint stays empty — and an empty cell in this table usually means nobody looked, not that somebody looked and found nothing.

That distinction cuts both ways, and it is the honest reading of every zero above.

Copper peptides: the widest gap, and the most confident marketing

GHK-Cu is the ingredient most often sold for skin and hair, and it has 21 human skin records against 2 hair records. Worse, the hair side does not contain what the marketing implies. No placebo-controlled human trial isolates GHK-Cu for hair. The one placebo-controlled trial involving a GHK peptide tested a copper-free GHK plus 5-aminolevulinic acid complex in 45 men over six months, which cannot separate the two ingredients; the most-cited laboratory study tested AHK-Cu — a different tripeptide — on follicles in a dish; the animal reports came from companies then developing copper peptides. Our peptides for hair growth page works through all of it.

On the skin side, the single randomized trial indexed on PubMed found no objective benefit in 13 laser-resurfacing patients, and the positive 40-woman trial was funded by the serum's maker. See copper peptides.

So the ingredient with the loudest dual claim has, on close reading, one small negative indexed trial on one side and no isolating trial at all on the other. That is not a reason to conclude it does nothing. It is a reason to notice that the confidence in the marketing has no counterpart in the literature.

Collagen is the exception, and it is not a serum

Oral collagen peptides are the only entry with a genuine record on both sides: 188 human skin records and 17 hair records, with two meta-analyses pooling 19 trials in 1,125 participants and 26 trials in 1,721 participants for skin hydration and elasticity. The hair evidence is thinner and industry-run — the largest placebo-controlled trial we could open followed 114 women on 1 g per day of a fish-collagen peptide for 24 weeks and reported increased hair diameter.

The catch is the one this site keeps returning to: collagen is digested and travels through the blood, while a topical peptide is applied at parts per million and has to cross the stratum corneum. Evidence for the drink is not evidence for the serum. Our collagen peptides page separates them.

Thymosin beta-4, and what "evidence on both sides" can hide

Thymosin beta-4 looks balanced at 17 skin and 6 hair records, and that appearance is misleading in a way the raw counts cannot show. Almost all of its human research is wound healing — venous leg ulcers, in randomized trials measuring closure. None of it is cosmetic, on either side. A count tells you how much has been published, never what it was about, which is why we read the records rather than only counting them. The compound's marketing as part of an injectable "glow" combination is covered on our glow stack page.

What this census cannot tell you

Three limits, stated because a table of numbers invites more confidence than it earns.

Indexing is not existence. Acetyl tetrapeptide-3 returns zero on both sides, yet a 30-volunteer manufacturer trial for hair exists in J Cosmet Sci 2013. Ingredient trade names, journals PubMed does not index, and supplier dossiers all sit outside this method. Every zero in the table means "not indexed under these terms", which is a weaker claim than "does not exist".

Counts are not quality. One well-run randomized trial outweighs ten uncontrolled supplier tests, and the census cannot see the difference. That is why our per-ingredient pages read each study rather than tallying it.

Absence of evidence is not evidence of absence. An untested pairing is untested. It may work; nobody has checked.

What the census does establish is narrower and still useful: when a product claims skin and hair from a single peptide, there is usually research on one of those claims and none on the other, and a reader can now check which is which.

Next: the whole skin category counted and ranked by evidence; the hair side in peptides for hair growth; what not to mix with copper peptides. Our testing notes set out the standard this publication works to.

Sources

  • PubMed census run 2026-09-06 via NCBI E-utilities esearch, humans[mh] filter, ingredient title/abstract terms intersected with skin terms (skin, wrinkle*, photoaging, dermal, facial) and hair terms (hair, alopecia, follicle*, scalp). PubMed.
  • de Miranda RB, Weimer P, Rossi RC. Effects of hydrolyzed collagen supplementation on skin aging: a systematic review and meta-analysis. Int J Dermatol 2021;60(12):1449-61. PubMed 33742704.
  • Pu SY, Huang YL, Pu CM, et al. Effects of Oral Collagen for Skin Anti-Aging: A Systematic Review and Meta-Analysis. Nutrients 2023;15(9). PubMed 37432180.
  • Miller TR, Wagner JD, Baack BR, Eisbach KJ. Effects of topical copper tripeptide complex on CO2 laser-resurfaced skin. Arch Facial Plast Surg 2006;8(4):252-9. PubMed 16847171.
  • Guarnera G, DeRosa A, Camerini R. The effect of thymosin treatment of venous ulcers. Ann N Y Acad Sci 2010;1194:207-12. PubMed 20536470.
  • Trial-level detail for each ingredient, with participant numbers and funders, is set out with its primary citations on our peptides for hair growth, copper peptides, collagen peptides, Argireline and Matrixyl pages.

Published by Peptide Glow Journal Editorial. No product is ranked, scored or affiliate-linked on this page.

Frequently asked questions

Are there peptides that work for both skin and hair?

Almost none has evidence for both. Counted on 2026-09-06, the peptides with published human skin research and the peptides with published human hair research are nearly disjoint sets: Argireline has 20 skin records and 0 hair records, Procapil has 0 skin and 4 hair. The only ingredient with a real record on both sides is oral collagen, at 188 and 17, and that is something you swallow. A product claiming both is usually combining two different peptides, or moving one peptide's evidence somewhere it has never been tested.

Do copper peptides work for hair as well as skin?

The evidence is lopsided. Copper tripeptide returned 21 human records on skin terms and 2 on hair terms, and no placebo-controlled human trial isolates GHK-Cu for hair: the one placebo-controlled trial with a GHK peptide tested a copper-free GHK plus 5-aminolevulinic acid complex in 45 men, so it cannot separate the two ingredients, and the most-cited laboratory study used a different peptide, AHK-Cu, on follicles in a dish. On the skin side the single indexed randomized trial was negative on its objective endpoints. Copper peptides are the most confidently marketed skin-and-hair ingredient and the one with the widest gap between the two evidence bases.

Why do skin and hair need different peptides?

Because the targets are different tissues with different biology and, in practice, because the research programmes were funded by different industries. Skin peptides are developed and tested for wrinkle appearance in cosmetic laboratories; hair peptides are developed and tested for follicle counts and anagen ratios, usually against or alongside minoxidil. Nothing prevents one molecule from doing both, but no one has paid to find out for most of them, so the claim that an ingredient does both is nearly always an extrapolation.

What about collagen for skin and hair?

Oral collagen is the one ingredient with published trials on both sides, and the two sides are not equal. For skin, two meta-analyses pooled 19 trials with 1,125 participants and 26 trials with 1,721, finding improved hydration and elasticity against placebo. For hair, the placebo-controlled trials are few, small and industry-run, the largest being 114 women taking 1 g of a fish-collagen peptide for 24 weeks with an increase in hair diameter. Nothing shows collagen regrows hair in diagnosed hair loss on its own.

Are skin-and-hair peptide supplements worth taking?

This page does not tell anyone what to take. What the census supports is a way to read the label: check whether the specific peptide named has human research on the side of the claim being made, rather than on the other side. For most ingredients marketed for both, the answer is that it has research on one side and nothing on the other.

Does a peptide serum help hair if you put it on your scalp?

No published trial has tested a facial peptide serum on scalp. The hair trials that exist used hair-specific formulations at their own concentrations, and the skin trials measured wrinkles. Applying one to the other is untested, which is a different statement from unsafe, and neither this publication nor anyone else can tell you what would happen.

How was this census done, and what are its limits?

On 2026-09-06 we searched PubMed for each ingredient's title and abstract terms, filtered to human-indexed records, then intersected each with skin terms and with hair terms, counting the results. The limits are real and worth stating: an ingredient can be studied under a name we did not search, trials published in journals PubMed does not index are invisible to it, and a record count is not a quality measure. Acetyl tetrapeptide-3 shows exactly that failure mode, returning zero on both sides while a manufacturer trial for hair does exist.