Acne is one of the most heavily trialled conditions in dermatology. That makes it an unusually good place to ask what the peptide evidence looks like, because the denominator is enormous and the comparison is not a matter of opinion.
The denominator
PubMed, 2026-09-18, acne in the title or abstract, publication type randomized controlled trial: 1,411 records.
| Intervention | Randomized acne trials |
|---|---|
| Benzoyl peroxide | 207 |
| Adapalene | 152 |
| Salicylic acid | 47 |
| Niacinamide | 11 |
| Any peptide | 3 |
| A peptide product, after reading all three | 1 |
Three of 1,411 records name a peptide. Reading them is the whole exercise, because two are not what the count implies:
- A trial of a peptide-containing cleanser — the one real record, below.
- A mechanistic readout of an isotretinoin trial, measuring how the antimicrobial peptide S100a7a is downregulated in patients on the drug. The peptide is an outcome measure, not a treatment.
- A study of dendritic cells and cytokine-induced killer cells in late-stage non-small cell lung cancer. It matches on vocabulary. It has nothing to do with skin.
A count published without opening the records would have reported three peptide acne trials. There is one. This is the fourth time this site has caught a keyword count inflating a literature by two or three times, and the pattern is consistent enough to be a rule: a raw count is a search result, not a census.
The one trial, read in full
Lee HJ and colleagues, Journal of Cosmetic Dermatology, 2020. Sixty volunteers with IGA grade II–III acne, randomized into groups of thirty, washing twice daily for eight weeks with either a cleanser containing 5-aminolevulinic acid and peptides or a basic cleanser.
| Endpoint, treatment group | Baseline | Week 4 | Week 8 |
|---|---|---|---|
| Inflammatory lesions | 5.9 | 4.5 | 4.1 |
| Non-inflammatory lesions | 11.4 | 8.8 | 7.4 |
Both reported as significant, alongside improvements in Michaelson's acne severity index and IGA, and better investigator assessment and patient satisfaction than the control at week 8. Adverse events were similar between groups.
Three things about it that a reader deserves before the numbers are used for anything.
The absolute change is small because the baseline is small. 5.9 inflammatory lesions is mild by the standards of an acne trial, and 1.8 fewer lesions after eight weeks is the size of the effect being described.
The peptide is not isolated. 5-aminolevulinic acid is in the same bottle. The control was a basic cleanser, not a 5-ALA cleanser without peptides — the one comparison that would have attributed anything to the peptide.
The peptides are not named. The title and abstract say "peptides". Which peptides, at what concentration, is not in the indexed record, so the trial cannot be used as evidence for any specific ingredient on any specific label.
One department, two indications, the same design
Here is the part that does not appear anywhere else. The only randomized trial of a peptide product in acne and the only randomized trial of a peptide product in pattern hair loss come from the same dermatology department, and both pair the peptide with 5-aminolevulinic acid.
Lee WJ is the last author of the 2020 acne paper and the first author of Lee WJ et al., Annals of Dermatology, 2016 — a trial of a complex named ALAVAX, 5-aminolevulinic acid with glycyl-histidyl-lysine peptide, in 45 patients with male pattern hair loss. Both are from the Department of Dermatology, Kyungpook National University School of Medicine, Daegu. Our peptides for hair growth page reads the 2016 trial in detail and reaches the identical structural conclusion for hair: because the peptide is complexed with 5-ALA, nothing in the result can be attributed to the peptide.
Two fields, two indications, one research programme, one unresolvable confound. That is not a criticism of the group, whose papers are clear about what was in the bottle. It is a description of how thin the interventional peptide literature in dermatology actually is: pull on the only randomized thread in two separate conditions and the same lab is holding both ends.
The other peptide literature, and why it is not this one
Search acne against antimicrobial peptides and the picture inverts. 144 records join the two subjects. Cathelicidin or LL-37 appears in 30, defensins in 38.
Then the composition of those 144:
- 44 carry the review publication type.
- 1 is a clinical trial of any design — the isotretinoin readout already counted above.
This is a mechanism literature. The skin manufactures its own antimicrobial peptides, their expression shifts in acne-involved skin, and the subject is reviewed far more often than it is tested. Thirty per cent reviews and one trial is the signature of a field that is interesting and not yet actionable.
The third member of that set is eczema, where the same family runs the other way entirely: our count of the randomized eczema record found LL-37 and beta-defensin expressed at lower levels in atopic skin, and the one antimicrobial peptide trialled on it cleared the bacteria without moving the symptoms.
It is worth setting beside our peptides and rosacea page, where the same molecule family appears with a much sharper evidence trail — cathelicidin measurably overexpressed and abnormally processed in rosacea skin, and inflammatory when the processed forms are injected into mouse skin. Same peptides, different disease, and in rosacea the peptide is implicated in causing the condition rather than proposed as a remedy for it.
The peptides actually on sale have no acne file at all
Searching acne against the ingredient names that appear on cosmetic labels returned seven records on 2026-09-18 — two for GHK or copper peptide terms, five for palmitoyl, Argireline, Matrixyl or acetyl hexapeptide terms.
Not one studies acne. They are a mouse atherosclerosis metabolome study of a plum extract, a review of stearoyl-CoA desaturase inhibitors, a gene-association paper on glycaemic index and acne, a pharmacy compounding-compatibility study and a 2004 practice column on skin care. The two GHK matches are a scar-gel study and that same column.
So the honest summary of the commercial question is short. The peptides with an acne literature are the ones the skin makes. The peptides sold in acne skincare have none. Our copper peptides guide covers what GHK-Cu has been tested for, and acne is not on the list.
What this page is, and is not
This is a document of an absence, and it is written because the question is asked often enough to deserve a straight answer rather than a shrug. The absence is reproducible: every count above is one search anyone can re-run, and each comes with a control count from the same database on the same day so that a zero can be told apart from a broken query.
It is not a treatment page. Acne that is persistent, painful or scarring is a medical problem with treatments carrying hundreds of randomized trials behind them, and that decision belongs with a dermatologist. Nothing here is a recommendation about anyone's skin.
Sources
- Lee HJ, Kim JY, Park KD, Lee WJ. Randomized controlled double-blind study of a cleanser composed of 5-aminolevulinic acid and peptides on mild and moderate acne vulgaris. J Cosmet Dermatol 2020;19(7):1745-50. PubMed 31778021
- Lee WJ, Sim HB, Jang YH, Lee SJ, Kim do W, Yim SH. Efficacy of a Complex of 5-Aminolevulinic Acid and Glycyl-Histidyl-Lysine Peptide on Hair Growth. Ann Dermatol 2016;28(4):438-43. PubMed 27489425
- Downregulation of S100a7a antimicrobial peptide in acne vulgaris patients after isotretinoin therapy. Dermatol Ther 2019. PubMed 31639246
- PubMed record counts and record lists via NCBI E-utilities, run 2026-09-18. Search shapes are given in the text so that every figure can be reproduced.