Air pollution and lung cancer: an investigation of county-level residual incidence rates with smoking eliminated

Smoking is the leading cause of lung cancer and this can make it difficult to focus public attention on other preventable causes, like air pollution. To provide easily understandable quantitative information on the relative importance of smoking and air pollution as contributors to lung cancer, we used a two-stage regression method to simulate what would happen to county-level lung cancer rates if smoking were completely eliminated. Residual lung cancer rates would vary substantially among counties after eliminating smoking; the objective of this paper was to identify air pollutants that drive high lung cancer rates post smoking elimination. We used multilevel negative binomial regression to model 2019 lung cancer incidence data, provided by the Surveillance Epidemiology and End Results (SEER) Program, at the county level (n = 1,078), to estimate incidence rates with observed smoking prevalence (lagged 20 years) and with smoking eliminated. In a second step, age-standardized county rates were dichotomized into those above and below the top 20th percentile; logistic regression was used to model the effects of specific air pollutants on the odds of having a high residual lung cancer rate after smoking was eliminated. Using 1999 U.S. EPA data on mean concentrations of carcinogenic air pollutants (lagged 20 years) we evaluated the contributions of: PM2.5, diesel exhaust particulate, benzene, polycyclic organic materials (POM), and radon (chosen a priori) to high county lung cancer rates. Consistent with previous literature, mean PM2.5 concentrations and radon were important contributors to high lung cancer rates after accounting for smoking. Additional lung cancer determinants were mean concentrations of diesel exhaust particulate, benzene, and POM. The contribution of PM2.5 to lung cancer appeared to be largely explained by the 3 more specific pollutants diesel exhaust, benzene, and POM. The study suggests there are geographic inequities in focusing cancer prevention only on tobacco control because places with high concentrations of carcinogenic air pollutants would still have disproportionately high lung cancer rates, even if smoking could be completely eliminated. These results can be used to identify counties where control of carcinogenic air pollutants might have a particularly important impact on cancer prevention.