Konstantinos Farsalinos | 18 September 2013
As I recently mentioned in a letter to the editor of Journal of Chromatography A, the way results of research are presented is crucial for the message obtained by the public and can have important implications on decisions implemented by authorities.
This is especially important in the case of e-cigarettes, mostly for 2 reasons:
1. The issue of smoking addiction and cessation is an extremely sensitive health matter, especially when you consider that smoking is one of the most difficult habits to quit and, 2. There is significant controversy in the scientific community and in society about e-cigarettes.
Over the past month, a series of articles was published in a consumer magazine in France, presenting the results of a chemical study on e-cigarettes that was conducted on their behalf. The initial article mentioned that some e-cigarettes emit more toxic chemicals compared to tobacco cigarettes. However, there was no mention on the methodology or the results in detail. That article sparked worldwide media frenzy, with respected news-media reproducing the news with titles such as: “E-cigarettes are as harmful as cigarettes and could cause cancer” (Daily Mail), or “Cancer from e-cigarettes” (Greek newspaper). Obviously, none cared about the true results of the study, it was the headline (and its distortion) that was “selling” to readers.
When the results were released (at first the range of findings was reported and subsequently a table with findings per sample tested), it was evident that the initial statement was not true. Chemicals like aldehydes (formaldehyde, acetaldehyde and acrolein) were indeed found, but the levels were similar to those found by Goniewicz and coworkers. In his study, Goniewicz concluded that the levels of such chemicals were 9-450 times lower in e-cigarettes compared to tobacco. Despite the identical results of the French magazine study, their conclusion was that e-cigarettes may be more dangerous. How did they do that? By comparing the highest emission from e-cigarette with the lowest emission from tobacco cigarette (which was not even tested in their experiment, they used literature data). Of course this is false from a scientific and statistical point of view. When you test a group of products, it is wrong to use a selected case and compare it with another selected case from another group (tobacco). The same mistake was made in the Journal of Chromatography A paper which evaluated the presence of nitrosamines in e-cigarette liquids. They tested 105 liquids, they reported the mean values (12.99 nanograms per ml) but in conclusions they chose the highest nitrosamine-containing liquid and compared it with the levels found 5 years ago in a Ryan e-cigarette sample. That was a completely unjustified and useless comparison. No comparison with the levels present in tobacco smoke was made. In my letter, I emphasized the authors’ mistake, and I calculated that, based on their findings, an e-cigarette user is exposed 76-142 times less nitrosamines from the daily consumption of e-cigarette liquid compared to smoker’s exposure from 1 tobacco cigarette. This is a valid, realistic and useful comparison. Useful for the scientific community and for the public health authorities, who should be properly informed about the results of research and their implication, since their decisions are (and should be) based on scientific data.