The Comedogenic Scale: Assumptions, Flaws and Discrepancies

Those with acne prone skin tend to rely on the comedogenic scale to select skincare products with “noncomedogenic” ingredients, in hopes of minimizing “acne cosmetica” or acne caused by the use of cosmetics. While the comedogenic scale may help to an extent, this method of product selection places a lot of formulas on the chopping block that might not actually be comedogenic. It’s also not always beneficial to only purchase products that claim noncomedogenic, since there is not an FDA definition or regulation for this claim. Comedogenicity is the likelihood an ingredient will cause blackheads and clog pores. Ingredients are ranked on a 0 to 5 point scale with 0 being noncomedogenic and 5 being severely comedogenic. The comedogenic scale is an attempt to quantify the comedogenic potential of ingredients and thereby a product as a whole. In theory if you are concerned about blemishes and breakouts from your beauty products, you look for products containing only ingredients that rate on the scale a 2 or lower. However, the comedogenic scale makes a lot of assumptions and the testing models used are inherently flawed.

Rabbit Ear Assay (REA)

In order to understand the flaws in the comedogenic scale it is important to understand how the tests are carried out. Ingredients were originally evaluated for their comedogenic potential by performing a rabbit ear assay (REA). The substance was placed on the inner ear canal of a rabbit at a concentration of 10% up to 100% and left for two weeks at which point scientists would look at the follicles and their change in dilation.¹

REA Flaws

This model, aside from its questionably ethical use of animals, presents a number of undesired variables. The rabbit ear has a more hypersensitive response than human skin. So if an ingredient is noncomedogenic using the REA it is even less likely to be comedogenic for human skin meaning the model is great for determining absolute negatives. However, this model leads to false positives as an ingredient that is comedogenic with REA may not be comedogenic on human skin. Scientists used subjective grading to evaluate follicle dilation, which is different from microcomedone formation. Follicles can dilate due to irritation from an ingredient and not from compaction or formation of a comedone. Studies were not standardized in their ingredient concentration, application duration, and visual grading leading to differences in comedogenic ranking of substances. This model is only useful for giving absolute negatives or identifying ingredients that have zero comedogenic potential.

Cyanoacrylate Follicular Biopsy

The newer model relies on corneosurfametry and is called cyanoacrylate follicular biopsies. In the newer human model, the substance at a concentration of 10% up to 100% is placed on the upper back of subjects and then occluded with a bandage for 48 to 72 hours. This process is repeated for weeks to months. Then cyanoacrylate glue is placed on the skin and glass slides are pressed down and withdrawn to remove the outer layers of the stratum corneum along with any sebum plugs. With the use of a microscope, scientists calculate the number of follicles and microcomedones per square centimeter to determine an ingredient’s comedogenic potential.²

Cyanoacrylate Follicular Biopsy Flaws

While the human method may be more appropriate to rank an ingredient on the comedogenicity scale than REA, there are still major flaws. Consumers don’t typically occlude a product on their skin; occlusion encourages further product absorption and penetration and therefore leads to a higher chance of comedone formation.³ The subjects selected for these studies have “prominent follicular orifices” or large pores and are more prone to comedone formation. The ingredients are either undiluted or highly concentrated and left on the skin for extended periods of time increasing chances of comedone formation. The physiology of the skin on the back is very different from that on the face. Also the sample size for these studies is relatively small, up to 10 subjects. This small sample size does not provide great statistical power to extrapolate to the general population. A product that causes acne for one person may not have any negative effects on another since there are other variables to consider such as climate, age, skin microbiome, skin type, and hygiene habits.

What is the application?

These models have low external validity meaning the results do not translate well outside the context of the study itself. These models don’t reflect normal use conditions. There are comedogenic ranking discrepancies and new studies have shown that once thought comedogenic materials may not be comedogenic. One such ingredient that was villainized is petrolatum which has proven to be noncomedogenic contrary to original belief.⁴ The comedogenic scale can act as a tentative guide but there are a number of questions one should ask before deeming a product comedogenic. One such question is what is the application? There is a significant difference between a rinse-off and leave-on product. Both models reflect leave-on product application with the human model going a step further and increasing penetration via occlusion. If a product is rinsed off and has a low contact time on the skin then an ingredient’s comedogenic potential in a product significantly drops.

What is the ingredient’s use level?

The comedogenic scale leads one to assume that a finished cosmetic product containing a comedogenic ingredient is therefore comedogenic itself; however, there is a significant dose-response relationship that needs to be considered. One study showed that known comedogenic substances proved to be noncomedogenic at lower concentrations.⁵ For example, the ester isopropyl isostearate has a comedogenic ranking of 4 when applied at 100% which drops down to a 2-3 when applied at 50% and further drops down to a 1-2 when applied at 5%. The use level of a substance plays an important role in its propensity to cause blemishes and breakouts. Along the same line, the vehicle and product type can influence the comedogenic potential of ingredients.⁶ An ingredient in a powder product or in an evaporating vehicle like alcohol or water may be noncomedogenic while the same ingredient in a vehicle such as an oil may become comedogenic. The formula as a whole and its solvents play an essential role in comedogenicity.

What is the grade or purity?

Surprisingly, ingredients with the same INCI may differ in their comedone causing potential. Ingredients such as jojoba oil and beeswax have shown variability in their comedogenicity suggesting contaminants or impurities. Some ingredients can be standardized for various compounds such as safflower seed oil can be standardized for linoleic acid, a fatty acid with low comedogenic potential, or standardized for oleic acid, a fatty acid with high comedogenic potential.⁷ The source can drive an oil or ingredient to be more or less comedogenic.⁸ Some ingredients have a generic INCI but can differ in their composition breakdown. Emulsifying Wax NF can differ in its concentration of longer-chain alcohols and therefore differ in its propensity to cause comedones.

A relevant example

A number of oil cleansers on the market contain comedogenic emollients such as isopropyl palmitate or ethylhexyl palmitate. These wetting agents are great silicone replacements since they are able to reduce the surface tension and improve spreadability, slip, and solubilization. While these substances have proven comedogenic, in cleansing oils they are highly effective at removing sebum, SPF products, and waterproof makeup that can clog pores. Seems quite contradictory until you view the formula as a whole. There are surfactants in the cleansing oil that essentially nullify the comedogenic effects of the emollients by enhancing rinse-off when water is applied. The comedogenic emollients don’t have sufficient contact time to cause comedones and blemishes. These emollients are also diluted to various degrees by other ingredients in the product further reducing their comedogenic potential.