The Smoking Gun?
Rothman, Kenneth J. Editor; Cann, Cristina I. Associate Editor
The Associated Press reported recently that we are beginning to experience a global decline in tobacco smoking. 1 This good news is most welcome. If the trend is permanent, the better health that will result will be partly attributable to the efforts of epidemiologists and other public health professionals who for many years have investigated the afflictions that tobacco smoking brings and who worked to free us from its grip.
Tobacco smoke, unfortunately, is not the only combustion product that threatens human health. The air of many cities and rural valleys around the globe is choked with combustion products from burning wood, vehicle exhaust, power plants, and industry smokestacks. The character of polluted air varies with place and season. Whatever the pollution mix, however, we already know that we are better off without this collective exhaust of our industrial enterprise in our air. Tobacco smoking has had its defenders, however misguided, who argued that smoking tobacco is a choice and its effects are sustained largely by those who make the choice. Even these flawed arguments cannot be raised to defend air pollution. What stops us, then, from clearing the air?
Only the cost, it seems. Everyone prefers cleaner air, but how much should we pay to get there? Questions like that evoke cost-benefit comparisons that examine the cost of cleaner air in terms of lives saved or other health outcomes. For our part, we think that the benefit side of the equation is short-changed when the esthetic component is neglected. If we focus on the health effects, however, we need to consider the results of epidemiologic research and what it has revealed about air pollution.
Much of the air pollution research in epidemiology nowadays involves time-series analyses of daily variation in air pollution concentrations and acute outcomes. 2,3 These studies take advantage of short-term fluctuations in air quality to learn about their relation to health outcomes that have a short induction time. Relatively few confounders can distort this type of analysis, although there remains plenty of areas for debate, 4,5 especially in regard to the active agents in the pollution mix. 6 Long induction times are not suitable for study by time-series analyses, because the effects of day-to-day fluctuations in exposure to pollution on the risk of disease get smoothed out as the induction time increases.
Until now, most studies looking at possible long-induction-time effects of air pollution on lung cancer risk have been based on aggregate data. Getting individual-level data on air pollution exposure over a long induction period is a formidable task. Cohort studies with individual-level data on air pollution have not been undertaken, with the exception of a study by Beeson et al.7 Using historical ambient pollution data based on zip codes of homes and offices, they studied a cohort of more than 6,000 nonsmoking Californians whom they followed for 15 years. Lung cancer was diagnosed in only 36 of their cohort members.
In this issue, Nyberg and colleagues 8 report findings from a case-control study of lung cancer, using air-pollution information obtained on an individual level going back more than 4 decades into the past. 8 They used geographic information systems to assign the individual exposure patterns to cases and controls. Their study goes beyond the work of Beeson et al. in two important ways: (1) because they conducted a case-control study, they were able to include a study base that gave rise to nearly 1,200 cases of lung cancer, considerably more than the 36 cases that Beeson et al. reported in their cohort study; and (2) the case-control approach, coupled with the historical information available in Sweden on air pollution, enabled them to consider individual-level exposures going back nearly half a century.
Nyberg et al. focused their interest on NOx/NO2 as an indication of pollution from road traffic and SO2 as an indicator of pollution from heating. For the latter indicator they found little effect. For exposure to pollution from road traffic, they found that for an induction time of 21–30 years, those with the highest exposure had an estimated 44% increase in risk of developing lung cancer (RR = 1.44; 95% CI = 1.05–1.99) compared with those who experienced the lowest levels, adjusting for age, year, smoking habits, radon exposure, and occupational exposures known to be associated with lung cancer. They also found that the relative risk was greater for never-smokers (RR = 1.68; 95% CI = 0.67–4.19). In contrast, when NO2 exposure was averaged over a longer 30-year period, the effect estimates were much smaller. It thus appears that exposure to air pollution more than 20 years before the diagnosis of lung cancer is more relevant than recent exposure. This study by Nyberg et al. is the first epidemiologic study that could have reported such a finding, as it is the first to have the data needed to do it.
This careful study corroborates a growing body of research evaluating the effect of air pollution on lung cancer risk using a variety of methodologic approaches and exposure metrics. 7,9–12 These studies of necessity encompassed a shorter induction time than the study by Nyberg et al; effect estimates ranged from up to a 90% increase in risk of lung cancer in areas of moderate pollution to more than a 250% increase in heavily polluted areas. 7,11,12 Relative effects, as in the current study by Nyberg et al., have tended to be greatest among nonsmokers, as one might predict if the effects of air pollution and smoking on lung cancer risk were independent. Among these studies effect estimates, as usual, vary considerably amid a variety of potential biases. Disagreement over interpretations is lively in this area of epidemiologic research, as in others; questions can certainly be raised about the exposure models and their source-specificity in the study of Nyberg et al. Indeed, questions can be raised about the ability of all these studies to distinguish the effects of separate sources of air pollution. Nevertheless, the findings of Nyberg and colleagues represent a new plateau on which to rest the debate over air pollution effects on lung cancer. Even if this study is not the smoking gun in the debate over air pollution effects, these individual-level data spanning much of an individual’s life will be influential, making it clearer that there is an increased risk of cancer from inhaling combustion products other than tobacco.
1. Associated Press. Facts about declines in smoking. May 22, 2000, 11:00 EDT.
2. Zmirou D, Schwartz J, Saez M, Zanobetti A, Wojtyniak B, Touloumi G, Spix C, Ponce de León A, Le Moullec Y, Bacharova L, Schouten J, Pönkä A, Katsouyanni K. Time-series analysis of air pollution and cause-specific mortality. Epidemiology 1998; 9: 495–503.
3. Schwartz J. Air pollution and hospital admissions for heart disease in eight US counties. Epidemiology 1999; 10: 17–22.
4. McMichael AJ, Anderson HR, Brunekreef B, Cohen AJ. Inappropriate use of daily mortality analyses to estimate longer-term mortality effects of air pollution. Int J Epidemiol 1998; 27: 450–453.
5. Zeger SL, Thomas D, Dominici F, Samet JM, Schwartz J, Dockery D, Cohen AJ. Exposure measurement error in time-series studies of air pollution: concepts and consequences. Environ Health Perspect 2000; 108: 419–426.
6. Moolgavkar SH, Luebeck EG. A critical review of the evidence on particulate air pollution and mortality. Epidemiology 1996; 7: 420–428.
7. Beeson WL, Abbey DE, Knutsen SF. Long-term concentrations of ambient air pollutants and incident lung cancer in California adults: results from the AHSMOG Study. (Adventist Health Study on Smog.) Environ Health Perspect 1998; 106: 813–823.
8. Nyberg F, Gustavsson P, Järup L, Bellander T, Berglind N, Jakobsson R, Pershagen G. Urban air pollution and lung cancer in Stockholm. Epidemiology 2000; 11: 587–595.
9. Dockery DW, Pope CA III, Xu X, Spengler JD, Ware JH, Fay ME, Ferris BG, Speizer FE. An association between air pollution and mortality in six US cities. N Engl J Med 1993; 329: 1753–1759.
10. Pope CA III, Thun MJ, Namboodiri MM, Dockery DW, Evans JS, Speizer FE, Heath CW Jr. Particulate air pollution as a predictor of mortality in a prospective study of US adults. Am J Respir Crit Care Med 1995; 151: 669–674.
11. Katsouyanni K, Pershagen G. Ambient air pollution exposure and cancer. Cancer Causes Control 1997; 8: 284–291.
12. Katsouyanni K, Trichopoulos D, Kalandidi A, Tomos P, Riboli E. A case-control study of air pollution and tobacco smoking in lung cancer among women in Athens. Prev Med 1991; 20: 271–278.
© 2000 Lippincott Williams & Wilkins, Inc.