#61 Particulate air pollution and lung cancer: A case-control study

PURPOSE: To evaluate the association between particulate air pollution and lung cancer in a retrospective study in Connecticut, Utah and southern Idaho. METHODS: We collected detailed residence histories from 1,474 pathologically confirmed lung cancer cases and 1,811 population controls selected with randomized recruitment. Analysis was limited to the 3,073 participants who lived only in Connecticut, Utah and Idaho during the period 5–20 years before lung cancer diagnosis for cases or interview for controls. Using geographical information system technology to interpolate total suspended particle (TSP) exposure from fixed-site monitoring data, we assigned a TSP exposure level to each residence and averaged individual exposures over the 15-year period from 5 to 20 years before the diagnosis/interview. Odds ratios (OR) associated with average TSP were estimated with logistic regression after adjusting for age, gender and decade-specific pack-years of cigarette smoking and taking into account selection probabilities used in randomized recruitment. RESULTS: The OR for a 10 μg/m3 increase in TSP was 2.4 (95% CI = 2.0, 3.0) in Connecticut and 1.3 (95% CI = 1.1, 1.5) in Utah and Idaho. ORs were similar for smokers and nonsmokers. The OR associated with the top tertile for TSP and heavy smoking compared with nonsmokers in the lowest tertile was 57.4 (95% CI = 26.7, 123.4) in Connecticut and 75.7 (95% CI = 37.6, 152.5) in Utah/Idaho. CONCLUSION: Long-term exposure to higher levels of TSP may increase the risk for lung cancer. The higher observed OR in Connecticut despite lower average TSP exposures may be due to differences in the physical and chemical composition of TSP between Connecticut and Utah/Idaho, or to differences in the extent of exposure misclassification between those two regions. Alternatively, these differences might be an accurate reflection of the shape of the dose-response curve and our use of a measure of relative rather than absolute risk.