This just in: moisterizers cause cancer. So say many recent news articles about a Rutgers dermatological study.

The popular press is getting a few things right about this study. Most seem to emphasize that the study was in mice, not people, and that as such the results must be viewed as preliminary. For example, Business Week quotes the principal researcher who did the study as saying "We don't know what the implications are for humans." The LA Times' includes quotes from dermatologists unaffiliated with the study:

"The components in moisturizers are tested. There's no evidence for this being a problem in humans."

The Times also explicitly raised the possibility that the current wave of moisturizer-cancer articles might be a "another needless public health scare." So far, so good.

But of course, there's more. The studies author Allan Conney is quoted as saying

"I think it raises a red flag indicating that there's a need to determine whether or not these products could cause this problem in people."

I went looking for "red flags" in the original study. Here's one: The researchers treated mice with 100 mg of moisturizer per day. It doesn't sound like much, but mice weigh about 20 grams, meaning that the treatment was about 0.5% of body weight. Extrapolated to a 70 kg human, this would be 350 g of moisturizer every day. That's a whole bottle of moisturizer! I doubt I will ever be a daily user of that much moisturizer.

Doing the calculation on a surface area basis puts the extrapolated human dosage to a somewhat more reasonable ~23 g per 70 kg human per day. But it's not clear to me why surface area normalization is better than body mass normalization. Especially since the the original study says "mouse skin is much thinner and more permeable than human skin." Oops!

Another detail: "During the course of our studies, we developed a Custom Blend cream that was not tumorigenic..." Oh great! The researchers already solved the problem -- and patented it! -- but no one happened to mention that in the popular reports.

Here's another interesting tidbit: "[O]ur [previous] studies suggested the possible utility of caffeine and caffeine sodium benzoate as topical agents for inhibiting sunlight-induced skin cancer in humans..." Why were no journalists around to write articles about these older results when they came ou: "Caffeine cures skin cancer!" or "A link between caffeine and a cure for cancer". Maybe journalistic overreaching to find popular import for research findings is acceptable only when it leads to scary thoughts instead of hopeful ones.

All in all, the current study showed that some moisturizers cause an increased rate of skin cancer in mouse models. I guess the popular press mostly got that part right. Here's what they missed out on:
1. The dosage of moisturizer was very high, far higher than people are ever likely to use regularly.
2. The researchers apparently understood the tumorigenic effect of the moisturizers enough to design a moisturizer which did not cause an increased rate of tumors.
3. New, caffeine-based prophylactics for skin cancer are being developed. Caffeine!

No calculation is required to arrive at points 2 and 3. They are stated explicitly in the plain text of the original study. That no articles mention these points suggests either (i) journalists don't habitually read the study they're writing about, or (ii) if they do, they ignore anything that doesn't fit into a "it's a new public health crisis" template for their article.

UPDATE: I edited the post slightly to remove a few extraneous sentences and make my central point clearer.