Comedogenic ratings explained: where the 0-5 numbers come from

Where the numbers came from

In 1972 Kligman and Mills published a paper called "Acne cosmetica".¹ The concept linked certain cosmetic ingredients to comedones, the blocked follicles of acne.² Animal models were then used to judge the comedogenic potential of raw materials, on the assumption that finished products containing them would behave the same way, and dermatologists were given lists of substances to avoid.²

The animal model was the rabbit ear. James Fulton's group tested cosmetics and ingredients in rabbits in the 1970s³, and in 1984 published results for products and ingredients, describing the rabbit ear assay as "not an ideal animal model but is the best we have".⁴ A 1989 paper extended the work into a long table of ingredients, each graded from 0 to 5.⁵

How the rabbit ear test worked

In the 1989 study, each ingredient was usually diluted to 10% in propylene glycol and spread on the inside of one ear of each rabbit once a day, five days a week, for two weeks; the other ear served as an untreated control. The width of the blocked follicles was then measured and converted into a grade from 0 to 5. Grade 5 matched the large comedones produced by Isopropyl Myristate, which the study used as its standard positive substance.⁵

The study read the grades this way: 0 to 1 not significant, 2 to 3 borderline, 4 to 5 positive. It gave the model's advantage as speed, two weeks in the rabbit ear against perhaps six months on human skin, and its disadvantage as extreme sensitivity: not everything that irritates the rabbit ear will irritate human skin.⁵

Others found the same limits. Mills and Kligman, who developed a test on the upper backs of human volunteers with large follicles, found the rabbit model more sensitive than the human one.⁶ A 1986 paper noted controversy over the assay's reliability because results differed between laboratories, and suggested that the subjective standard grading contributed.⁷

One ingredient, several numbers

The 1989 paper itself shows how much a grade depends on the test. Four fatty acids, Lauric Acid, Myristic Acid, Palmitic Acid and Stearic Acid, each scored lower when dissolved in a quickly evaporating solvent than in sunflower oil:⁵

The same paper's main table, which tested ingredients in propylene glycol, gave stearic acid 2-3, a third answer for one ingredient. The author also showed a colourant's grade rising or falling with the vehicle it was mixed into, called his survey "not at all definitive", and wrote that whether ingredients that score high in rabbits are always comedogenic in people "still remains uncertain".⁵

Why an ingredient's grade does not carry over

A grade describes one raw material, at one concentration, in one vehicle, on rabbit skin. A finished product combines that ingredient with many others, at a concentration the label does not give: EU ingredient lists carry no percentages, and below 1% ingredients may be listed in any order.⁸

When Draelos and DiNardo tested finished cosmetic products directly, with a modified version of the Mills and Kligman human assay on six volunteers, they concluded that finished products using comedogenic ingredients are not necessarily comedogenic. They note that only a limited number of products could be tested.² Adding up grades across a whole ingredient list, or flagging a product because one name once scored 4, applies a number to something it never measured.

"Non-comedogenic" is a claim

A 2018 systematic review of marketing terms found that "non-comedogenic", like "hypoallergenic" and "oil-free", is not regulated by any governing body.⁹ There is no official test a product must pass to carry the word.

In the EU the general rules for claims still apply. Article 20 of the cosmetics regulation forbids labelling that implies a product has characteristics it does not have¹⁰, claims must be backed by adequate and verifiable evidence, and a claim that carries an ingredient's properties over to the finished product needs evidence of its own.¹¹ The company responsible chooses the method.¹² A 2017 paper on claim substantiation describes two ways to support a non-comedogenic claim: a human model descended from Mills and Kligman's test, and the more costly clinical trial, which it calls the gold standard.¹³ The label does not say which was used.

What to look for on the label

• Treat a rating as the result of one old test, at one strength and in one mixture, not as a fixed property of the ingredient.⁵
• Read "non-comedogenic" as the maker's claim about that formula, resting on whatever evidence the maker holds.¹¹
• Look at the whole list, not one name. The skincare ingredient checker shows what each ingredient is there to do, and what an ingredient scanner can and cannot tell you explains why a list is not a verdict.
• Ask a dermatologist if breakouts are a concern. No label or rating table can answer how a product will behave on your skin.