April 03, 2009
Cosmetics Database reliability questioned
1000fragrances expresses some concerns about the reliability of the Skin Deep cosmetics Database. This reiterates some concerns I had when I was looking at the database a few months ago. For those of you unfamiliar with the database, it allows companies to enter their product ingredients and then evaluates the safety of the product, assigning a hazard score of 0-10. Consumers concerned about the safety of products can look them up and obtain an assessment of their relative safety. It sounds like a good idea, but it appears to have a faulty evaluation mechanism. Unless the user is careful in evaluating the data, and understands the limitations of the evaluation, it may be useless. They do state what they call a “data gap” that assesses what data is unknown. In the case cited by 1000fragrances the data gap for the “good” product is 81%, which suggests to me that they are admitting they don’t really know whether it is safe or not. The “bad” product has a data gap of 64%.
The difference seems to be that the ingredients in the “good” product are listed as essential oils, while the ingredients in the “bad” product are listed by their constituent chemical names. And the “bad” product contains a number of ingredients that are indeed synthetic or hazardous in some way, in addition to the essential oil constituent chemicals that are listed. So indeed the “bad” product may be bad. 1000fragrances makes the point that the “good” product is also potentially hazardous, even though it gets a “good” score.
One reason the “bad” product scores badly is because one of the ingredients is “fragrance”, which is allowable as an ingredient by the FDA, but is considered as “bad” by the Environmental Working Group because it could contain anything.
Given that the Compact for Safe Cosmetics and EWG seem to be prepared to jettison the small businesses who have been the innovators in the Safe Cosmetics industry by blindly supporting the regulation of the industry through the FDA Globalization Act of 2008/9, perhaps it is time to reconsider whether participation in the Cosmetics Database is a good idea. It certainly products some strange results.