I’m currently in Las Vegas anxiously waiting for The Amazing Meeting to start. Believe it or not, I’ll even be on a panel! While I’m gone, I’ll probably manage to do a new post or two, but, in the meantime, while I’m away communing with fellow skeptics at TAM7, I’ll be reposting some Classic Insolence from the month of July in years past. (After all, if you haven’t been following this blog at least a year, it’ll be new to you. And if you have I hope you enjoy it again.) This particular post first appeared in July 2007.
“What do you think about second hand smoke?” he asked me. I sensed ulterior motives behind the question, but I wasn’t sure. I suspected that he was just looking for an argument.
“It’s bad,” I joked.
“Some have told me that the studies don’t show any health problems from second hand smoke,” he replied.
“I’m sure ‘some’ have,” I retorted somewhat sarcastically.
“No, really, is there any evidence,” he replied. “I’m open-minded about this topic.”
Somehow I doubted this, but I figured, what the heck, and did a little reviewing. It makes for some interesting reading.
The question of whether second hand smoke (SHS) causes lung cancer or other health problems in people exposed to it on a chronic basis is a tough one to examine epidemiologically because, as with most studies of exposures, it’s difficult to control for other factors. It’s not like studying smoking as a cause for cancer, where it’s relatively easy to examine the exposure. More importantly, whatever the effect of SHS may be, we can, based on the simple principle of dose-response, expect that it would produce considerably less profund of an effect than smoking itself, given the obvious difference between inhaling someone else’s smoke at a bar and inhaling 10-30 cigarettes a day over the course of many years. Moreover, it should be remembered that the science behind whether SHS is a health risk is a distinct from the policy question of what we should do about it as far as indoor smoking bans go. Finally, when millions of people are exposed, small increased risks can lead to many, many additional cases of smoking-caused disease.
Before I get into the studies, as a background, I’d point out that the tobacco industry managed to try to deny and obfuscate for decades over the very real and now indisputable risk of health problems from smoking. How on earth did they manage to do this? One has to remember that most people who smoke will not get cancer. As hard as it is for many lay people to believe, it’s true. A person who smokes two packs a day smoker for 40-50 years will have approximately a 20% chance of dying of lung cancer. True to a dose-response curve, that means people who smoke less (which is most smokers these days) will have even less of a chance of dying of cancer. That means that there are a lot of smokers who never develop lung cancer and a lot of smokers who never develop cardiovascular disease. Indeed, I once said that it’s impossible to tell whether any individual smoker will or won’t get lung cancer; I can only quote the odds. In any case, if it was possible for tobacco companies to obfuscate such a powerful and now indisputable epidemiological risk for so long, imagine how much easier it is to sow doubt over the considerably smaller elevation in risk due to SHS.
I realize that “skeptics” of the health dangers of SHS will immediately start spinning conspiracy theories, but an excellent source of information on the topic is the most recent Surgeon General’s report, released about a year ago, entitled The Health Consequences of Involuntary Exposure to Tobacco Smoke: A Report of the Surgeon General. I had thought of plowing through the literature myself without a guide until the hundreds of studies out there convinced me that it was such a massive undertaking that it would be hard to justify it for just one blog article. The report lists many, many studies that support the contention that SHS causes health problems and critically compares them to the fewer studies that don’t support such a link. I realize that to those who don’t accept the science indicating an increased health risk from passive smoke, invoking any government document will be viewed in much the same way as HIV/AIDS denialists view the NIH statement on the link between HIV and AIDS, but keep in mind that this document was a product of the Bush administration, hardly a champion of regulating human exposure to environmental toxins.
Also, such “skeptics” often forget or ignore a key element in making a connection, namely biological plausibility. We already know that smoking causes cancer and heart disease; it is thus plausible that exposure to secondhand smoke would be likely to cause cancer as well, just at a decreased rate because of its much lower dose. This is not like the arguments over mercury in thimerosal in vaccines, where not only is it not biologically plausible that mercury causes autism but the epidemiological evidence does not support such a link. Moreover, “skeptics” also often will point to single studies as “proof” that there is no connection between SHS and human disease. The problem is, for a complex epidemiological problem like whether SHS causes cancer or heart disease, no single study will ever definitively prove or disprove a link. The totality of evidence has to be examined critically. Many “skeptics” that SHS is a danger will cherry pick studies that didn’t find a link, but usually those studies fail to find a link because of inadequate sample size or statistical power.
Because the issue at hand is smoke exposure in the workplace and in indoor spaces, I’m not going to focus so much on the well-known and well-documented harmful effects of smoking on children’s respiratory health. The evidence for this is well-documented in Chapter 6 of the Surgeon General’s report. Suffice it to say that children exposed to secondhand smoke are at increased risk for acute respiratory infections, more severe asthma, and a variety of other health conditions.
In adults, numerous studies support the existence of approximately a 25% elevated risk of lung cancer from those exposed to secondhand smoke chronically. The types of epidemiologic studies generally include studies of secondhand smoke exposure in the home, which usually take the form of studies looking at cancer rates in nonsmoking spouses of smokers. Criticisms of these studies generally takes the form of criticizing how well they control for other factors, given that spouses living together generally share more than just exposure to smoke. Even so, the better-controlled studies here consistently show elevated risk of lung cancer in nonsmoking spouses of smokers. There are at least 52 spousal smoking studies, and meta-analysis produces estimates of relative risk ranging from 1.15 to 1.43; i.e., a 15-43% elevation in lung cancer risk. These are summarized in Chapter 7 of the Surgeon General’s report.
The second type fo study tends to look at smoking in the workplace. As of the Surgeon General’s report, there were at least 25 such studies, of which seven were from the U.S., one from Canada, seven from Europe, and ten from Asia. The studies are fairly heterogeneous, but their conclusions are remarkably consistent for such a diverse bunch of epidemiological studies. Not all studies reached statistical significance, namely because of small numbers and inadequate statistical power, but they still came to similar estimates, with relative risks ranging from 1.12 to 1.32, with a pooled estimate of 1.22, or a 22% increased risk of lung cancer. More interestingly, there are studies that support not just a correlation between exposure to SHS and lung cancer but a dose-response effect (Hackshaw et al, 1997; Lubin 1999; Zhong et al, 2000; Brennan et al, 2004). When these studies are looked at, neither confounding factors, publication bias, nor misclassification errors account for the observed increased relative risk.
A number of studies have also looked at the risk of cardiovascular disease due to SHS. As summarized in Chapter 8 of the Surgeon General’s report, they consistently show a roughly 30% increase in risk of cardiovascular events. This data is summarized well in Chapter 8 of the Surgeon General’s report. Epidemiological links are supported also by a wealth of preclinical evidence showing the effect of SHS on vascular tone, endothelial cell function, platelet aggregation, and a number of other markers of vascular injury.
One study that is frequently cited by “skeptics” who don’t accept that SHS can cause health problems is J. E. Enstrom and G. C. Kabat Br. Med. J. 326, 1057; 2003. This study failed to find a causal relation between SHS and tobacco-related mortality. There’s just one problem. The study was crap, and the BMJ was forced to reveal that the authors had accepted money from tobacco companies and that the study had several fatal flaws. In an editorial in Nature recently, this fiasco was recounted:
But top scientists at the cancer society say they repeatedly warned Enstrom of possible deficiencies in his analysis — particularly a 25-year gap in which exposure to second-hand smoke could not be verified. The society also says that when it gave Enstrom computerized records of study subjects, it was not aware that he was receiving funding from the tobacco industry. Later tobacco-related lawsuits revealed he had received money from industry funneled through an organization called the Center for Indoor Air Research. And court records show Enstrom previously did consulting and research for attorneys defending the tobacco companies R. J. Reynolds and Philip Morris.
Perhaps my favorite paper of all is one that I’m saving for last because if I started with it I might have been accused of “poisoning the well.” This one, which appeared in JAMA in 1998, is revealing. It examined review articles, and, well, here’s the abstract:
Why Review Articles on the Health Effects of Passive Smoking Reach Different Conclusions
Deborah E. Barnes, MPH; Lisa A. Bero, PhD
Objective.– To determine whether the conclusions of review articles on the health effects of passive smoking are associated with article quality, the affiliations of their authors, or other article characteristics.
Data Sources.– Review articles published from 1980 to 1995 were identified through electronic searches of MEDLINE and EMBASE and from a database of symposium proceedings on passive smoking.
Article Selection.– An article was included if its stated or implied purpose was to review the scientific evidence that passive smoking is associated with 1 or more health outcomes. Articles were excluded if they did not focus specifically on the health effects of passive smoking or if they were not written in English.
Data Extraction.– Review article quality was evaluated by 2 independent assessors who were trained, followed a written protocol, had no disclosed conflicts of interest, and were blinded to all study hypotheses and identifying characteristics of articles. Article conclusions were categorized by the 2 assessors and by one of the authors. Author affiliation was classified as either tobacco industry affiliated or not, based on whether the authors were known to have received funding from or participated in activities sponsored by the tobacco industry. Other article characteristics were classified by one of the authors using predefined criteria.
Data Synthesis.– A total of 106 reviews were identified. Overall, 37% (39/106) of reviews concluded that passive smoking is not harmful to health; 74% (29/39) of these were written by authors with tobacco industry affiliations. In multiple logistic regression analyses controlling for article quality, peer review status, article topic, and year of publication, the only factor associated with concluding that passive smoking is not harmful was whether an author was affiliated with the tobacco industry (odds ratio, 88.4; 95% confidence interval, 16.4-476.5; P<.001).
Conclusions.– The conclusions of review articles are strongly associated with the affiliations of their authors. Authors of review articles should disclose potential financial conflicts of interest, and readers of review articles should consider authors’ affiliations when deciding how to judge an article’s conclusions.
Indeed. Lest you think that this isn’t still going on, it’s instructive to take a look at a couple of more recent articles that demonstrate the tobacco industry’s favored techniques for denying a link between SHS and health problems: namely, challenging causation of adverse health effects of passive smoking by citing limitations of epidemiologic research, raising methodological and statistical issues, and disputing biological plausibility.
The bottom line is that, although the effect of SHS is relatively small in epidemiological terms because it is a relative risk that is well under 2, namely a roughly 25% increased risk of lung cancer and heart disease, plus lower and less clear associations to a number of other cancers and diseases, it is real. It consistently shows up in multiple studies done using multiple different methods in many different countries. Moreover, even though it is relatively small, because of the sheer numbers of people exposed to SHS, it is most definitely significant. Indeed, in the case of heart disease, which is the most common cause of death, even an increase of risk due to SHS of only a few percent would be significant, although such a small increase would be almost impossible to detect in epidemiological studies. Consequently, a strong argument can be made for indoor smoking bans in the workplace. We may not be able to do much aside from education and the encouragement of smoking cessation to prevent the exposure of children and nonsmoking spouses to the dangers of SHS. Arguments on a public policy basis and a cost-benefit basis about whether banning indoor smoking makes sense, but arguments that the science doesn’t show that SHS can cause lung cancer and cardiovascular disease and adults and a variety of health problems in children are nonstarters. The science doesn’t support them (although ideologues will try to take advantage of controversies over the exact magnitude of the health risks posed by SHS). Indeed, even Penn Jillette has admitted that he was wrong about SHS in an episode of Bullsh*t! in which it was claimed that SHS was not a danger, although he was full of, well, Bullsh*t! when he stated that a “lot of this was very new.”
No, it wasn’t. The data at the time that original episode aired was quite strong. The newer data only confirm the older data available then.
That being said, there is a risk of going too far in legislating smoking bans. Certainly indoor smoking bans at workplaces and in restaurants and bars are defensible on a scientific and public health basis. However, outdoor smoking bans, such as I’ve heard about on many California beaches and a recent law passed by the Beverly Hills City Council to ban smoking in all outdoor eating areas, are just plain stupid and wrong-headed, at least if the rationale is to prevent SHS-caused health problems in the population. This is particularly true since such bans apply to sidewalk cafes, where, in my experience, the exhaust fumes from passing traffic tend to overwhelm any smoke that comes from cigarettes. It doesn’t help the reasonable and scientifically supported cause of indoor workplace smoking bans to overstep and impose bans in cases where the science doesn’t support it.
And here are some studies that I picked almost at random. If anyone doesn’t think they’re enough, you can review the Surgeon General’s report for literally hundreds more, and I could post many more as well.
- de Waard F, Kemmeren JM, van Ginkel LA, Stolker AAM. Urinary cotinine and lung cancer risk in a female cohort. British Journal of Cancer 1995; 72(3):784-7.
- Johnson KC, Hu J, Mao Y, the Canadian Cancer Registries Epidemiology Research Group. Lifetime residential and workplace exposure to environmental tobacco smoke and lung cancer in never-smoking women, Canada 1994-97. International Journal of Cancer 2001;93(6):902-6.
- Zhong L, Goldberg MS, Gao YT, Jin F. A case-control study of lung cancer and environmental tobacco smoke among nonsmoking women living in Shanghai, China. Cancer Causes Control. 1999 Dec;10(6):607-16.
- Eisner MD, Wang Y, Haight TJ, Balmes J, Hammond SK, Tager IB. Secondhand smoke exposure, pulmonary function, and cardiovascular mortality. Ann Epidemiol. 2007 May;17(5):364-73.
- Venn A, Britton J. Exposure to secondhand smoke and biomarkers of cardiovascular disease risk in never-smoking adults. Circulation. 2007 Feb 27;115(8):990-5.
- Hill SE, Blakely T, Kawachi I, Woodward A. Mortality among lifelong nonsmokers exposed to secondhand smoke at home: cohort data and sensitivity analyses. Am J Epidemiol. 2007 Mar 1;165(5):530-40.
- Barnoya J, Glantz SA. Cardiovascular effects of secondhand smoke: nearly as large as smoking. Circulation. 2005 May 24;111(20):2684-98.
- Raupach T, Schafer K, Konstantinides S, Andreas S. Secondhand smoke as an acute threat for the cardiovascular system: a change in paradigm. Eur Heart J. 2006 Feb;27(4):386-92.
- Brennan P, Buffler PA, Reynolds P, Wu AH, Wichmann HE, Agudo A, Pershagen G, Jockel KH, Benhamou S, Greenberg RS, Merletti F, Winck C, Fontham ET, Kreuzer M, Darby SC, Forastiere F, Simonato L, Boffetta P. Secondhand smoke exposure in adulthood and risk of lung cancer among never smokers: a pooled analysis of two large studies. Int J Cancer. 2004 Mar;109(1):125-31.