Most studies on the association between lung cancer and air pollution have investigated mortality. There have been few studies of lung cancer incidence. We used data from the ongoing Netherlands Cohort Study on Diet and Cancer for 114,378 subjects with follow-up from September 1986 to December 1997. Exposure to black smoke, nitrogen dioxide (NO2), sulfur dioxide (SO2), and particulate matter < or =2.5 microm (PM2.5) and traffic intensity variables (intensity on nearest road, intensity in a 100 m buffer, and an indicator variable for living close to a major road) were estimated at the home address. We conducted Cox proportional hazard analyses in the full cohort adjusting for age, sex, smoking status, and area-level socioeconomic status. We also carried out case-cohort analyses using more potential confounders on a subset of study participants for whom complete information from the baseline questionnaire had been processed. Adjusted analyses included 1940 cases for the full cohort and 1295 cases for the case-cohort analysis. Relative risks (RRs) for the overall air pollution concentrations were slightly below unity, and for the traffic variables RRs were slightly elevated. Risk was elevated among people who never smoked cigarettes (40,114 participants; 252 cases), with RRs of 1.47 (95% confidence interval = 1.01-2.16) for overall black smoke concentration, 1.11 (0.88-1.41) for traffic intensity on nearest road, and 1.55 (0.98-2.43) for living near a major road. We found evidence for an association of exposure to black smoke and traffic with lung cancer incidence in people who had never smoked. No associations were found for the full cohort, or for other categories of smoking.
