A comedo is the scientific name for a blackhead. An ingredient that causes comedones is comedogenic. The comedogenic scale is a five point scale that tells you how comedogenic that ingredient is. It is also complete nonsense. Let’s have a look at why.
I don’t support the use of animal testing to develop cosmetics. My objection is not a moral one. People are going to use cosmetics and if the only way to make sure they are safe is to test them on animals then I’d say we have to test them on animals. But I think that animal testing simply isn’t a good way of generating the data you need to decide if something is safe or not.
As it happens despite the large amount of debate it has generated, the cosmetic industry has never been a big user of animal testing. And what it has done has to a large extent been done to comply with regulations rather than something that the industry itself wanted to do.
And what animal testing has been done has never proved to be particularly useful to the cosmetic scientist. In fact some of it has been positively misleading. This blog post is the story of how one particular animal test has proved to be rather worse than a waste of time. This is the story of the rabbit ear model for comedogenicity.
Scientific papers are, for good reason, written in neutral non-emotive language. In plain english the rabbit ear model for comedogenicity sounds less impressive. You try out a chemical behind a rabbit’s ear – and if it causes a blackhead there you assume the same will happen in humans. That at any rate was the idea. The origins of the rabbit ear model go back to the 1950s when it was developed to assess the risks posed to workers by handling certain industrial chemicals. For that purpose the test proved reasonably successful.
After that it drifted into use in the cosmetic industry and in 1979 Albert Kligman, a very well known and respected dermatologist, published a protocol explicitly intended for assessment of cosmetic ingredients. In it he created a scale of comedogenicity from 0 to 5. (0 was good, 5 was bad). Papers started to appear using this method to generate data and drawing conclusions about just how likely to mess up your pores particular ingredients might be. One for example revealed that while lanolin itself was fine, all its derivatives were really bad.
It was a seductive idea. You could run all your ingredients through a test and screen out all the ones that would give you problems. Everyone would be happy. (Except the rabbits. While the test itself was probably not too distressing they were killed when it was completed.)
But the problems started as soon as the results started going around. Experienced formulators noticed quickly that what the data was predicting bore little relationship to their personal experience. I am reporting what I was told here – even I am not old enough to actually remember this. And things got worse when it turned out that there was little agreement between different papers studying the same ingredients. Albert Kligman was, and is, a very respected name and his prestige kept the model on the road for maybe a bit longer than it would otherwise have done. But it lost credibility very quickly even so.
The rabbit ear model was also used for predicting acne – naturally enough since blackheads and acne pustules have a strong family resemblance to blackheads. A new disease name was coined: Acne cosmetica. This was a form of acne supposedly caused by using highly comedogenic cosmetic formulations. The rabbit ear model lasted a bit longer in this context but it was still discredited in time. Even Kligmann himself came round to the idea that the rabbit ear test could produce very misleading results and was noble enough to say so in a paper published in 1996. By 2007 Howard Maibach, someone whose reputation is not too far behind that of Kligmann himself, could dismiss it.
I have to say that even if the rabbit ear model had been more reliable, I think it was doomed from the start. Cosmetic formulations are not simply a sum of their ingredients. It is obvious that you can’t tell how a cake will taste from knowing a lot about flour, butter and sugar. The same is true of a skin cream or a lipstick. The Comedogenic Scale was never going to be of much use.
The trouble is that the papers are still out there and if you don’t know the context it is very easy to be misled. And it is such a nice idea as well. So to this day there are still people using the data generated from the rabbit ear model to judge how comedogenic formulations are. If you do a bit of googling you will soon find them. At least one website has a complete chart of comedogenic ratings supposedly to help you choose the right products.
Bad ideas really are a lot like Star Trek characters. They just don’t stay dead.
Albert Kligman original 1979 protocol
Rabbit Ear Comedogenic Scale
Kligmann renounces rabbit ear model
Maibach dismisses it.
Thanks to Wikipedia for the image and thanks to Wendy for asking the question that triggered of my interest in the comedogenic scale.