Monday, November 07, 2016

Investigators Botch Interpretation of New E-Cigarette Study Results; American Academy of Pediatrics Misrepresents the Findings

A new study published online this morning in the journal Pediatrics purports to show that flavored e-cigarettes serve as a gateway to youth smoking by increasing nonsmoking youths' intentions to smoke and decreasing smoking youths' intentions to quit.

The study reports results from the 2014 National Youth Tobacco Survey. In a cross-sectional analysis, the investigators examined the relationship between flavored e-cigarette use and (1) intentions to smoke among nonsmokers; (2) intentions to quit among smokers; and (3) perceptions of the danger of tobacco products among both smokers and nonsmokers.

The study concludes that the use of flavored e-cigarettes by youth is a gateway to future smoking: "Consistent with previous findings about e-cigarette use, our findings suggest that youth use of flavored e-cigarettes might serve as a gateway for future cigarette use."

A press release from the American Academy of Pediatrics (the publisher of Pediatrics) reports the conclusion of the study as follows: "Flavored E-Cigarette Use May Increase Teens' Taste for Smoking."

According to this press release: "New research shows the use of electronic cigarettes with flavors such as gummy bear and bubble gum among U.S. middle- and high-school students may serve as a gateway for future smoking." The press release also states that: "The study ... suggests use of these products increases young people's intentions to begin smoking... ."

The Rest of the Story

This study makes one of the cardinal mistakes of causal inference: assuming that correlation in a cross-sectional study equals causation. The fact that there is an association between the use of flavored e-cigarettes and higher susceptibility to smoking does not demonstrate that the use of e-cigarettes increases susceptibility to smoking. Because the data are cross-sectional, it could alternatively be the case that higher susceptibility to smoking causes youth to experiment with e-cigarettes.

In other words, the direction of this relationship could go either way: from e-cigarettes to an intention to smoke or from an intention to smoke to use of e-cigarettes.

Despite the fact that the relationship could be in either direction and despite the fact that the paper acknowledges that "the data are cross-sectional; thus, we were unable to establish causal inferences," the paper nevertheless concludes that e-cigarettes may be a gateway to smoking. Moreover, the press release clearly presents the findings as suggesting that e-cigarettes cause smoking initiation among youth.

This is a great example of the widespread bias against e-cigarettes that has taken hold in the tobacco control movement. Instead of presenting the study as showing equivocal results, the investigators and the American Academy of Pediatrics have both "chosen sides." This is not science; it is biased interpretation and presentation of science.

Why should we even be spending money on conducting this research if investigators and organizations are going to draw pre-determined conclusions regardless of whether those conclusions are supported by the data? Why not just save the time and money and jump right to the conclusion that e-cigarettes are a gateway to smoking without the need for actual scientific evidence? After all, we're drawing that conclusion without the presence of scientific evidence anyway.

No comments: