Tuesday, November 12, 2019

The Bias Against E-Cigarettes in Medical Research: An Example from the American Heart Association Annual Meeting

Later this week, the American Heart Association will hold its annual scientific conference in Philadelphia. In anticipation of the conference, the American Heart Association issued a press release yesterday announcing the results of a new study on e-cigarettes that will presented at the conference. The headline of the press release is:

"E-cigarettes take serious toll on heart health, not safer than traditional cigarettes."

One of the two papers that is being presented which purportedly support the conclusion that e-cigarettes are no safer than tobacco cigarettes in terms of heart health is a study of cholesterol levels in e-cigarette users compared to current smokers and nonsmokers.

The study is described as follows:

"Researchers evaluated healthy adults (ages 21-45) without existing cardiovascular disease and taking no daily medications in the Cardiovascular Injury due to Tobacco Use (CITU) Study. The study's 476 participants included 94 non-smokers; 45 e-cig smokers; 52 e-cig and t-cig smokers; and 285 t-cig smokers. Analysis was adjusted for age, race, sex, and non-smokers, sole e-cig or t-cig use, or combination e-cig and t-cig use."

The findings were as follows:
  • "Total cholesterol was lower and the bad cholesterol, LDL, was higher in sole e-cigarette users compared to nonsmokers.
  • Good cholesterol, HDL, was lower in dual smokers."
 And the study conclusion was as follows:

"Although primary care providers and patients may think that the use of e-cigarettes by cigarette smokers makes heart health sense, our study shows e-cigarette use is also related to differences in cholesterol levels. The best option is to use FDA-approved methods to aid in smoking cessation, along with behavioral counseling."

The Rest of the Story

This study found a cross-sectional association between e-cigarette use and riskier lipoprotein (cholesterol) profiles. Based on that cross-sectional correlation, the investigators conclude that e-cigarette use leads to higher "bad" cholesterol levels and therefore, switching from smoking to vaping does not improve cardiovascular health.

But let's examine this more closely.

In a cross-sectional study, you have to be very careful in extrapolating from correlation to causation because this type of study design is very susceptible to confounding -- that is, a third variable that is associated with both smoking/vaping status and cholesterol levels and makes it look like they are related but the relationship is actually driven by this third variable.

In this example, there is a very strong potential confounder: diet. It is very likely that smokers and former smokers have significantly less healthy diets than nonsmokers and therefore, worse cholesterol profiles. There may be differences in physical activity as well, which would lead to a finding of worse cholesterol profiles in current and former smokers than in nonsmokers.

In fact, studies have shown (example) that when you compare cholesterol levels in former smokers and nonsmokers, the former smokers have worse cholesterol profiles. Since most vapers are former smokers, it is almost certainly the case that vapers have worse cholesterol profiles than nonsmokers.

Thus, this paper provides no evidence of any effect of vaping on cholesterol levels. That relationship is completely confounded by diet and physical activity, and it has already been demonstrated in the literature that former smokers have worse cholesterol profiles than nonsmokers, even before e-cigarettes were ever invented!

This is why I think that the conclusion of this paper (that e-cigarette use leads to an unhealthy cholesterol profile and that vaping is therefore just as dangerous as smoking in terms of heart health) is unwarranted and indicates a bias against e-cigarettes.

Even if vaping did increase LDL cholesterol and lower HDL cholesterol, this wouldn't indicate that vaping is more dangerous than smoking. There are many other impacts of smoking on cardiovascular health that would need to be examined. And most importantly, one would need to understand the long-term effects of vaping before concluding that it is more dangerous than smoking.

The worst part of this story, however, isn't the evident bias against e-cigarettes in the research. The worst part of the story is the recommendation that smokers not switch to e-cigarettes in order to quit smoking. That is really bad advice. The reality is that 90% of smokers who try to quit using nicotine replacement therapy will fail. So advising vapers to try nicotine patches is tantamount to telling 90% of them to go back to smoking. Moreover, the best scientific evidence available - a randomized clinical trial - showed that e-cigarettes are twice as effective as the nicotine patch for smoking cessation.

The rest of the story is that the bias against e-cigarettes is leading to unfounded scientific conclusions, deceptive communications to the public about the health effects of e-cigarettes, and misguided medical advice.

No comments: