Retinaldehyde: doubling the concentration made no difference on most measures

A 2018 randomised, double-blind trial in Korea. Forty Korean women with photoaged skin applied either 0.1% or 0.05% retinaldehyde cream for three months. Assessment was by Antera 3D imaging.

Both arms improved.

• Texture: 13.7% improvement at 0.1%, 12.6% at 0.05%
• Transepidermal water loss (TEWL): down 14.5% at 0.1%, down 17.9% at 0.05%
• Hydration: up 10.2% at 0.1%, up 6.0% at 0.05%

But there was no statistically significant difference between the groups. Doubling the concentration did not separate the measurements. The TEWL number is actually larger in the lower-concentration arm — and since the difference is not significant, that must not be read as 'lower is better' either.

One measure did differ: melanin index improved by 6.5%, and only in the 0.1% arm.

The most important thing about this trial is what it does not contain: a vehicle arm. All forty applied retinaldehyde. So the trial cannot show 'retinaldehyde works.' What it shows is 'whichever of 0.1% and 0.05% you pick, most measures came out the same.'

For anyone trying to avoid raising the concentration and gaining only irritation, that is a practically useful result.

One more from the same family: the cream used in a 2015 melasma trial also contained retinaldehyde 0.05% — but alongside niacinamide 4%, arbutin 3% and bisabolol 1%. That trial cannot apportion the result between them; the bisabolol page records the same limitation.

Another 2018 randomised controlled trial. Fifty-five women were split into two arms for eight weeks.

• Cream arm: 0.1% retinaldehyde cream daily
• Peel arm: glycolic acid peel sessions

The results:

• Crow's-feet depth: cream −7.61%, peel −4.34%. Between groups, P = .3049 — not a significant difference.
• Texture: cream −5.61%, peel +3.54 (worse). Between groups, P = .0252 — a significant difference.
• Adverse events: fewer and milder in the cream arm.

On wrinkles the two approaches were indistinguishable; on texture the cream won. The peel arm's texture figure was positive, meaning it moved in the worse direction, and that is written as it stands.

The limitation has to be stated plainly. The cream used was not retinaldehyde alone. It also contained a glycosaminoglycan-degradation inhibitor (GGO) and Pre-tocopheryl. So this is a result for that cream, not for retinaldehyde.

And comparing a daily cream with a course of peel sessions means participants necessarily knew which they were getting. Blinding is not possible in this design.

Still worth recording: a cream applied daily for eight weeks was indistinguishable from a course of peels on wrinkles. And as with the item above, there is no untreated arm here either.

In Korea, the ingredients accepted for 'functional cosmetic — helps improve wrinkles' without submitting efficacy data are listed in Annex 4 of the Regulation on the Review of Functional Cosmetics. There are four:

• Retinol
• Retinyl palmitate
• Adenosine
• Polyethoxylated retinamide

Retinaldehyde is not on that list.

To be precise about what that means: it does not mean the ingredient does not work. The list names ingredients for which efficacy data need not be filed separately. Using an ingredient outside the list for a functional claim means going through individual review.

The practical implication for a reader: if a product containing retinaldehyde carries a 'wrinkle-improvement functional' label, that label is either owed to retinol or adenosine also present in the formula, or the product went through individual review. The ingredient list usually tells you which.