Peels have already appeared on this site three times — and each time against something else

A 2020 randomised, double-blinded, split-face controlled trial in Malaysia.

The participants are what make this trial valuable. 94.5% of the 36 subjects were Fitzpatrick type IV–V — that is, darker skin. Peels carry more risk of post-inflammatory hyperpigmentation the darker the skin, so data looked at directly in this population is scarce and worth having.

• Each side of the face was randomly assigned — Jessner's solution (JS) on one, 30% salicylic acid (SA) on the other
• Once fortnightly, three sessions in total
• Assessed by lesion counting, the Michaelsson acne score (MAS), photographs and the post-acne hyperpigmentation index (PAHPI)

The results.

Against baseline, all four measures improved significantly — inflammatory lesions, non-inflammatory lesions, MAS and PAHPI (P < .001 each).

But mixed model analysis comparing the two sides against each other found no significant difference.

That the pigmentation index fell is worth noting separately. Pigmentation is the main worry when peeling darker skin, and in this trial post-acne hyperpigmentation improved. But that means marks already present faded, which is a different question from the risk of new pigmentation appearing.

And the shape of this conclusion is familiar here. As the hyaluronic acid filler page sets out, almost every randomised trial in aesthetic medicine is A versus B, ending in 'no difference'. That design cannot rule out that neither worked. The third item below is about that.

This was 36 people, three sessions, six weeks.

Two trials already on this site, gathered here. In both, the peel was the comparator, not the subject.

One — it lost to a cream.

The 2018 trial from the retinaldehyde page. Fifty-five women were split into two arms for eight weeks.

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

Result: the two were indistinguishable on crow's feet (cream −7.61%, peel −4.34%; between groups P = .3049). But the cream won on texture — cream −5.61%, peel +3.54 (worse); between groups P = .0252.

That the peel arm's texture number was positive is written as it stands. At eight weeks it had moved in the worse direction.

Two — adding it helped.

The 2017 trial from the microneedling page, in sixty people with atrophic acne scars.

• Group 1: microneedling only (weeks 0, 6, 12)
• Group 2: microneedling plus 70% glycolic peels at weeks 3, 9 and 15

At week 22, on acne scar scores graded by blinded assessors, group 2's reduction was larger, and group 2 was also better on participant-rated texture.

Why this design is good. Both groups received microneedling equally and only one had the peel added, so any difference belongs to the peel. The kojic acid page has an add-on design of the same shape.

Setting the three trials side by side:

• Peel versus peel — no difference
• Peel versus a daily cream — wrinkles similar, texture to the cream
• Microneedling with or without a peel — the peel side won

Note too that the concentrations and depths differ throughout: 30% salicylic, glycolic (concentration not stated), 70% glycolic. The single word 'peel' does not denote the same thing.

The comparators in the three trials above were another peel solution, a topical cream and microneedling alone.

Not once was there a side that received nothing.

Why that matters. In the first trial all four measures improved significantly. But acne waxes and wanes on its own, and people enter trials mostly when it is at its worst. Simply returning toward their own average brings the score down (regression to the mean).

On top of that, being in a trial makes people cleanse and moisturise more diligently, and six weeks changes the season too.

So 'improved at P < .001' does not tell you the peel's contribution. It means the faces that were peeled were better six weeks later.

Look at the places in this site's procedures layer that did have an inert control and you can see why the worry is practical. On PRP the blinded raters could not tell it from saline; on LED Cochrane gathered 71 trials and concluded no clinically significant effects were shown.

We found no randomised trial comparing a peel against a sham procedure or no treatment.

That said, the microneedling trial in the second item partly sidesteps this problem. Both groups had the same procedure and only one had the peel added, so time and regression to the mean apply equally to both. That the peel won there is the firmest single line on this page.