Can PDRN outperform retinol on crow's feet? The 2026 split-face trial (Ye et al.)

Appraised by the Skingenetix research team. Medically reviewed by Dr Esther Bodde, Cosmetic & Medical Physician. Last reviewed 22 September 2026.

Verdict: in the only controlled study of PDRN applied to intact skin, a 0.1% PDRN eye cream reduced crow's-feet wrinkle area by up to 23% in 28 days, more than twice the improvement of a 0.1% retinol cream on the other side of the same face. It is well measured and promising, and small: one trial, 31 women, four weeks, and no placebo.

Key facts

  • Design: randomised, double-blind, split-face: each woman used both creams, one on each side
  • Participants: 31 Chinese women aged 35 to 55 (average 47.5), all with self-reported sensitive skin
  • What was applied: an eye cream with 0.1% of a medium-length salmon PDRN as the only active, twice a day
  • Duration: 28 days, measured at day 0, 14 and 28
  • Comparator: 0.1% retinol in an identical cream base; there was no placebo
  • Main measures: VISIA imaging (wrinkles), Antera 3D (eye bags, tear troughs), Cutometer (firmness, elasticity), ultrasound (dermal thickness and density)
  • Result: crow's-feet wrinkle area and number fell about 20–23% with PDRN, against 6–7% with retinol; firmness and elasticity gains about 1.8 times retinol's; PDRN changes visible by day 14
  • Side effects: none on either side
  • Funding and conflicts: no specific funding declared; four authors are affiliated with commercial skincare biotechnology companies, not declared as a conflict
  • Our evidence grade: A for the clinical results (randomised, blinded, instrument-measured); the laboratory findings are graded separately
  • Source: Ye R et al., PLOS ONE, 2026 · PubMed 42430369 · free full text

What the researchers did

Thirty-one women with sensitive skin used two eye creams for 28 days, one on each side of the face, assigned at random. Neither they nor the assessors knew which side had which. One cream contained 0.1% PDRN, the other 0.1% retinol, in exactly the same base, so any difference between the sides came from the active ingredient.

The skin was measured with four separate instruments at the start, at day 14 and at day 28. Alongside the clinical trial, the team tested the PDRN in skin cells, in samples of human skin damaged by UV light, and, in one volunteer, for how far it penetrates the skin.

What they found

On the PDRN side, crow's-feet wrinkle area and wrinkle count fell by about 20 to 23% over 28 days, against about 6 to 7% on the retinol side. The authors report roughly twice retinol's improvement for wrinkles under the eyes, eye-bag volume and tear-trough depth, and about 1.8 times the gain in firmness and elasticity. The PDRN changes were already visible at day 14. No one had a reaction on either side.

In the laboratory arm, the same cream raised nine types of collagen, elastin and fibrillin in UV-damaged human skin samples.

How good is this evidence?

The split-face design is a strength: every woman is her own comparison, which removes differences in age, diet and sun exposure between groups. Blinding, randomisation and four objective instruments put it well above a typical cosmetic study.

The limits are real. It had 31 participants and lasted 28 days. There was no placebo side, so the study shows PDRN beat retinol but not how much better it is than doing nothing, and 0.1% retinol is a slow-acting comparator over four weeks. Exact statistics are shown only in the figures. The authors themselves note that the ultrasound changes may partly reflect short-term hydration, and the penetration test used a single volunteer. Four authors work for commercial skincare biotechnology companies, which the paper does not declare as a conflict.

What it does not show

It tested one specific medium-length PDRN at 0.1%, in an eye cream, around the eyes. It does not test our products, other PDRN sources or lengths, other parts of the face, or results beyond four weeks. Most other PDRN research uses injections or skin opened by microneedling, and those results do not apply to a cream on intact skin.

How it compares with our serum

The PDRN 1% Renewal Serum contains 1% salmon PDRN. The trial used 0.1% of a specific medium-length PDRN. No study has compared PDRN concentrations, so a higher percentage should not be read as a bigger effect, and the trial's results are not our serum's results.

Where this study fits

This is the first controlled trial of topical PDRN on intact skin, and the main reason PDRN has become a credible skincare ingredient rather than only an injectable treatment. A 2026 systematic review of PDRN trials found mostly injected or procedure-based studies and was completed before this one was published (Alhussain et al., 2026). For what PDRN is and how the rest of the evidence stacks up, see what PDRN is. For a comparable appraisal of a peptide trial, see our appraisal of the Argireline trial.

Reference

Ye R, Wang Q, Du L, Li L, Hu F. Topical medium-length PDRN enhances dermal extracellular matrix repair in photodamaged skin via PI3K-Akt/TGF-β-regulated pathways. PLOS ONE. 2026;21(7):e0350905. doi:10.1371/journal.pone.0350905. View on PubMed