New, Made To Order, Meta-Analysis About Indoor Tanning And Skin-Cancer
Today the press-release from a new meta-analysis about the connection between indoor tanning and (non-melanoma) skin-cancer made waves in media around the world.
Here are just some of the first campaign day’s headlines from around the world …
This certainly looks like a well-coordinated and successful PR-effort to demonise indoor tanning, just in time for the discussion about age-limitation for indoor tanning in Canada.
The meta-study, authored by Dr. Eleni Linos, is very similar to the one made in Europe a couple of months ago by the old “melanoma-marketing” team from IARC. It uses more or less the same old reports (most of them paid for by the inventors and sponsors of the sun-scare) and follows exactly the same model.
The problem with some of the reports that show the strongest correlation between sunbed use and non-melanoma skin-cancers is (in addition to their bias) that they include data mainly from tanning in tanning beds in private homes or with equipment used for medical skin-therapy in dermatologist clinics. The radiation from medical sessions is normally much more intensive than the radiation from commercial tanning beds and the risk many times higher. Tanning beds for home use are obviously totally uncontrolled and the risk of burning therefore higher.
The meta-analysis is a perfect tool for manipulation
The meta-analysis is a perfect tool for somebody who wants to manipulate facts. An unscrupulous researcher just have to include all the studies that support his or her case and exclude most of those which don’t. This is exactly the method used by both Eleni Linos and Boniol in their identical strive to fulfil their paymasters wishes. It makes it easy to imagine the same guiding instructions from the PR-wizards of melanoma marketing behind the two latest meta-analysis creating headlines about perceived dangers of indoor tanning.
Another hallmark of manipulative research is its possibility to create bombastic headlines. Very often you will find that the frightening figures used by the spin-doctors of public relation are fetched from one extreme study. Sometimes, like in the case of the Mayo Clinic’s report on melanoma among young women in Olmsted county, the subject creating the headlines doesn’t even have to be a part of the study itself.
This is of course super manipulative and it is working so well because no one, especially no journalist, really bothers to read the report itself, only the press-release.
Affiliations are not fully disclosed
As always (it seems) when it comes to articles published in BMJ, the list of affiliations for this study shows only “innocent” medical establishments. (In the case of the European study by Mathieu Boniol et al, it showed no affiliations by the authors but they refused to provide the non-disclose forms, which they are obliged to deliver upon request).
However, a little bit of Google search shows that the main author of the new meta-analysis, Dr. Eleni Linos, this year received the Career Development Award from Dermatology Foundation.
The main sponsor of Dermatology Foundation is Galderma, a company established in 1981 by L’Oreal and L’Oreal’s second largest shareholder, Nestle. Galderma made almost 2 billion USD in 2011 by selling medical remedies through dermatologists to all the “skin-cancer survivors” created by the early detection campaigns (also sponsored by L’Oreal’s daughter companies).
The early detection campaigns, like Melanoma Days, create the statistics that, together with expensive PR-activities, fuel the public perception of a skin cancer epidemic. Even if 93 out of 100 cases of diagnosed melanoma turn out to be only benign skin-lesions, the statistics are still the cornerstone in L’Oreal’s “melanoma-marketing” with the purpose of increasing their sales of sun-protection cosmetics and medical skin care products.
Dermatology Foundation was also the sponsor of the infamous Mayo Clinic report from April 2012 by Jerry D. Brewer. In that report, the Mayo Clinic managed to put the blame on indoor tanning for a totally imaginary increase of (diagnosed) melanoma among young women in a very small part of the USA even when the indoor tanning habits among the melanoma cases were not at all included in the study.
The problem with meta-analyses based upon case/control studies
The new meta-analysis is based upon case-control studies and the report describe the limitation this means like this:
This study is limited by the fact that it included only observational and mostly case-control studies. However, a randomised trial of indoor tanning is not realistic, and most of the included studies controlled for multiple potential confounders, making this the best level of evidence possible. Another potential limitation of this meta-analysis is the broad time period spanned by the data. Included studies collected data from the 1970s to the 2010s, which is important because indoor tanning devices have changed over time from high ultraviolet B output to predominately ultraviolet A output.
Putting all the complicated and scientific sounding statistical jargon aside, the general problem with this kind of case-control studies is that they include human beings. And as we all know, no person is exactly like another and therefore almost impossible to compare.
A recent case-control study in Europe, presents 13 risk-factors as possible contributors to non-melanoma and melanoma skin-cancers. The thing is that none of those 13 risk-factors are mutually exclusive.
For example, a person that has experienced sunburns as a child may also be a smoker as an adult and also a frequent traveller on sunny vacations (sunburn, smoking and sunny vacations were three of the 13 risk-factors in the large European study).
Even if we were to limit the choice for each factor to two, 13 factors will give 8,192 combinations (2 to the power of 13). Taking into account that the people in the control group share the same possible combination of risk-factors, the total number of combined factors that should be compared in order to get a relevant result becomes 16,384. Obviously, and given this proof of how different people in the case-control studies are, such studies don’t prove anything (or rather, can be tailor-made to prove whatever).
Limiting the number of risk-factors included in a case-control study will of course not help because all other kinds of behaviour still are present and are influencing the results even if they are not included in the study.
In reality indoor tanning is “statistically insignificant”
One interesting and relevant detail in the large European case-control study was that it showed the influence from indoor tanning to be extremely small (“statistically insignificant”). Only 6 persons from 1011 with non-melanoma skin-cancers were habitual indoor tanners. This should be compared to 13 habitual indoor tanners from 1550 persons in the control group. This study was, of course, not included in Eleni Linos’ meta-analysis. Neither did it get even a fraction of the media attention.