Mandelic acid: it tied with salicylic acid — and then split by which kind of lesion

A 2020 trial in Rohtak, India.

50 patients with mild-to-moderate facial acne were randomised into two groups.

• 30% salicylic acid peel
• 45% mandelic acid peel

Six sessions at two-week intervals, over 12 weeks. Assessment used the Michaelsson acne score (MAS) and clinical photographs.

Results:

• Overall the two peels were about equally effective. Neither MAS improvement nor percentage decrease in MAS differed significantly between the groups.
• But split by lesion type they diverged — salicylic acid did better on non-inflammatory lesions (comedones) and mandelic acid on inflammatory ones.

The authors write that mandelic acid is emerging as a peeling agent because of antibacterial and anti-inflammatory properties, which makes comparing it with long-established salicylic acid worthwhile. The split above runs with that account.

How to read it.

First, nothing was left untreated. So this trial cannot say 'a mandelic acid peel reduces acne'. It says 'it came out similar overall to a salicylic acid peel' and stops.

Second, 'no difference' is not 'the same'. Splitting 50 people in two and finding no difference can mean there is none, or that there was not the power to find one.

Third, the split by lesion type is a secondary analysis. The primary endpoint was the overall score, and that tied. The broken-out result is better read as a hypothesis for the next trial.

It is still practical information. It means the answer may differ depending on whether comedones or red angry lesions are the main problem. Read the salicylic acid page alongside.

So this entry is 'not established'.

Same trial, same 50 people.

Adverse effects of both agents were noted at each visit.

Result: they were fewer with the mandelic acid peel. The authors conclude that a 45% mandelic acid peel was as effective as a 30% salicylic acid peel, while its safety and tolerability were better.

Why we grade this differently from the entry above.

Up there, the comparator being another acid blocked the conclusion. Here it is an advantage.

Because the question is not 'does a peel reduce acne' but 'which of the two acids is less unpleasant'. For that question, a randomised head-to-head under the same conditions gives a good answer. We said the same on the erbium laser page — the same trial earns different grades depending on the question.

This result does not contradict the account about mandelic acid's large molecule either. But we will keep the distinction: the account did not produce the result; the result happened to fit the account.

So it sits at 'early evidence'. One trial, 50 people, no further.

And this is a kind of information this site writes down relatively rarely. One of the things people actually ask when choosing an ingredient or a procedure is how much it stings and how much it flakes — a question that is often easier to answer than the one about effect, and usually the one left out of the writing.