Abstract
Rationale: Industrial facilities emit known lung carcinogens into air, but the association of these agents with lung cancer risk at environmental levels is unknown. Objectives: We sought to investigate industrial emissions and lung cancer risk. Methods: We used a U.S. regulatory database to estimate airborne exposure to known and probable human carcinogens (n = 31) emitted from industrial sources (1987-1995) for 442,986 participants in the NIH-AARP Diet and Health Study. We estimated inverse distance- and wind-weighted average exposures within 1, 2, 5, and 10 km of the enrollment residence. Using Cox proportional hazards models adjusted for smoking and other confounders, we evaluated lung cancer risk overall and by major histologic subtype (adenocarcinoma, squamous cell carcinoma, small cell carcinoma) for levels (tertiles and medians) of exposure to each agent. Measurements and Main Results: Among agents with prior evidence of lung carcinogenicity, overall risk was elevated for cobalt (5-km hazard ratio (HR) for tertile(3) (T(3)) = 1.19; 95% confidence interval [CI] = 1.10-1.29; p-trend ⩽ 0.0001; 10-km HR(T3) = 1.15; 95% CI = 1.09-1.21; p-trend ⩽ 0.0001) and beryllium (5-km HR(T3) = 1.20; 95% CI = 0.94-1.55; p-trend = 0.15; 10-km HR(T3) = 1.15; 95% CI = 1.01-1.31; p-trend = 0.02). We also observed associations with benzene and nickel. For agents without prior evidence, styrene was associated with risk at 1 km (HR(T3) = 1.22; 95% CI = 1.00-1.48; p-trend = 0.03). Diethyl sulfate, chromium, and lead were also associated with risk. Associations for cobalt, benzene, nickel, and diethyl sulfate were most apparent for squamous cell carcinoma. Conclusions: Our novel findings show that relatively high air emissions of numerous carcinogenic industrial agents near the home were associated with lung cancer risk unexplained by smoking. These and the stronger associations for squamous cell carcinoma highlight the potential role of industrial exposures in lung cancer development.