Monday, October 06, 2008

New Study Concludes that Secondhand Smoke Exposure May Cause Nicotine Dependence Among Young Neversmokers

A new study published online ahead of print in the journal Addictive Behaviors has concluded that secondhand smoke exposure, especially in a car, may cause nicotine dependence among children who have never themselves smoked (see: Belanger M, et al. Nicotine dependence symptoms among young never-smokers exposed to secondhand smoke. Addictive Behaviors 2008, doi:10.1016/j.addbeh.2008.07.011).

The study examined symptoms of nicotine dependence among a cross-sectional sample of 1,488 never smokers, ages 10-12 years, in a stratified sample of schools in Montreal. Using multiple logistic regression and controlling for smoking among friends and siblings, the paper finds an increased odds of self-reported nicotine dependence symptoms among children with higher levels of secondhand smoke exposure in automobiles.

The authors conclude that secondhand smoke exposure - and in particular, exposure to tobacco smoke in cars - may cause nicotine dependence in young neversmokers: "Overall, these data suggest an interesting and potentially important hypothesis -- that exposure to secondhand smoke may potentiate nicotine dependence symptoms in young never-smokers. ... If replicated, these findings support public health interventions that promote non-smoking in the presence of children, and more specifically uphold policy interventions to restrict non-smoking in motor vehicles in which children are present."

The results of the study have been widely reported in the media as showing that secondhand smoke exposure can cause nicotine dependence among nonsmoking children. For example, the headline about this study in the Calgary Herald read: "Smoking Parents Can Hook Kids on Nicotine."

Other news headlines about the article include:

Second-Hand Smoking Can Lead to Nicotine Addiction in Children (InjuryBoard.com)

Some Children of Smokers Show Symptoms of Nicotine Dependence (National Business Review)

Smoking Parents Likely to Cultivate Nicotine Dependence in Children (HealthJockey.com)

Some Children Exposed to Second-Hand Smoke Show Dependence, Says Study (The News)

Smokers' Children 'Become Addicted to Nicotine Through Passive Smoking' (The Daily Mail)

Smokers’ Children Get Addicted to Nicotine (The Metro - UK)

Second-hand Smoke May Get Kids Hooked on Nicotine (The Leader-Post)

Parental Warning: Second-Hand Smoke May Trigger Nicotine Dependence Symptoms In Kids (ScienceDaily)

Smoking Parents Can Foster Nicotine Dependence in Kids (The MedGuru)

Second-Hand Smoke May Trigger Nicotine Dependence Symptoms In Kids (Medical News Today)

Does Secondhand Smoke Make Kids Addicts? Symptoms Of Withdrawal Seen In Non-Smokers (WISN-TV, ABC-Milwaukee)

Exposure to Secondhand Smoke Increases Risk of Childhood Nicotine Dependence (MedWire News)

The Rest of the Story

When I went to read the actual study, I was expecting research that measured symptoms of nicotine withdrawal in children, such as cravings for nicotine that are initiated when one is in the absence of nicotine but relieved by nicotine or symptoms of nicotine withdrawal that occur when nicotine exposure is ended, such as irritability, inability to concentrate, insomnia, fatigue, or headache.

To my surprise, the study did not actually measure nicotine dependence among young neversmokers. Instead, what the study used was a nicotine dependence assessment tool that was designed specifically for smokers. The measurement tool that was used was not relevant to measuring nicotine dependence symptoms in nonsmokers. In fact, nearly all the questions assumed that the respondent was a smoker and these questions were therefore irrelevant and not applicable to these nonsmoking respondents. Any nonsmoking youth would have had considerable difficulty answering these questions which clearly did not apply to them.

For example, one question was: "Do you find it difficult not to smoke in places where it is not allowed (at a movie theatre, at home if your parents don't know you smoke)?"

What in the world does this mean for a person who is a nonsmoker? Clearly, the question is designed for smokers and it has no place in the assessment of nicotine dependence among neversmokers.

The other questions, while arguably not as strictly irrelevant, were still inappropriate for the assessment of addiction among a nonsmoker. In fact, the questions, if anything, were leading questions that could easily lead a nonsmoker to report feeling addicted to cigarettes or to answering in a way that indicated addiction, even in the absence of any actual addiction.

One question asked: "How physically addicted to smoking cigarettes are you?"

Another asked: "How mentally addicted to smoking cigarettes are you?"

Both of these are leading questions which suggest to the respondent that it is not only possible for them to be addicted but that it is expected that they might be addicted.

The other questions were:
  • "How often do you have cravings to smoke cigarettes?"
  • "How often have you felt like you really need a cigarette?"
  • "When you see other kids your age smoking cigarettes, how easy is it for you not to smoke?" and
  • "[Do you] sometimes have strong cravings for cigarettes where it feels like [you are] in the grip of a force that [you] cannot control?"
Answering "yes" or "sometimes" to any one of these questions yes puts you into the addicted category for the analysis upon which the authors based their conclusions.

Quite clearly, in the context of nonsmokers, what these questions measure is not physical dependence upon nicotine, but the propensity to start smoking. In other words, for a nonsmoker, these questions are a measure of that individual's closeness to starting to smoke. What these questions really ask is not "Are you nicotine dependent?", but instead "How much of an urge are you experiencing to experiment with cigarettes?"

Now, here's the rub:

One would expect that exposure to secondhand smoke (i.e., living with one or more smokers, having a smoking parent, being in a car with a smoker [which likely means having a parent who smokes]) would be highly correlated with the urge to experiment with cigarettes. Thus, the study was set up in such a way to find that secondhand smoke exposure is associated with nicotine dependence.

What the study is actually finding is that the degree of secondhand exposure is associated with the extent of an individual's urge to experiment with cigarettes. This is quite plausible, since higher secondhand exposure usually indicates a higher number of parents/siblings/friends who smoke and/or a heavier intensity of smoking and higher level of dependence on smoking among family members or close friends. Such a finding is not at all a surprise, and does not really add anything new to the smoking literature.

However, this study concludes - instead - that the results indicate that the degree of secondhand smoke exposure is related to the extent of nicotine dependence. This alternative hypothesis is clearly a wild one and is unsupported by the data, which do not actually measure nicotine dependence among nonsmokers.

Before one could purport to have found a relationship between secondhand smoke and nicotine dependence among 10-12 year-old nonsmokers, one would have to validate a measurement instrument to assess nicotine dependence among nonsmokers. The study fails to do that. In fact, it borrows, without adaptation, an instrument that was specifically designed to assess nicotine dependence among adolescents who smoke.

The conclusions of the study and the worldwide headlines are way overblown. In fact, overblown isn't really right because that suggests that the results are exaggerated. It is not a question of exaggeration. The conclusion is simply wrong. The results do not support - at all - the conclusion that these young children were nicotine dependent.

The study reports that the measurement instrument was validated by testing it among a sample of smokers. Apparently, there was no testing conducted to see whether it makes any sense to use this scale among nonsmokers, whether nonsmokers will understand or have any idea how to answer questions that are not designed for them, and whether the ultimate result of the questioning will have any construct validity (i.e., whether it will actually correlate with nicotine dependence).

One could design a set of questions to assess the possibility of nicotine dependence among young nonsmokers, but these questions would not be on the list to try. They are leading questions which are not relevant to nonsmokers and whose use will ensure that nicotine dependence is found among nonsmokers.

In essence, this study was designed in such a way to create the very result it now reports. The study was set up such that it could not fail to find "nicotine dependence" among young nonsmokers. Furthermore, it was set up in such a way that it could not fail to find a correlation between "nicotine dependence" and secondhand smoke exposure.

Normally, I would assume that the intent of the researchers was pure and that setting up the study in this way was simply a bad mistake. However, the statement inserted into the paper about how these results are important in the movement to ban smoking in private cars with children leads me to believe that there is a strong bias inherent in this research and that the investigators are trying to find an association between secondhand smoke exposure and nicotine dependence, whether one exists or not.

This is an example of biased and shoddy science. Unfortunately, this is what tobacco control research is gradually deteriorating into. That the flaw inherent in this study is so blatant and that the results have already been disseminated worldwide in conclusive and striking fashion point to the damage that the degradation of the scientific integrity in tobacco control research is doing, not only to the movement, but to the public's appreciation of the scientific and medical facts about tobacco.

No comments: