Abstract
OBJECTIVES: The study aimed to determine whether the National Cancer Institute's (NCI) recent suggestion of associations between acrylonitrile (AN) exposure and mortality in lung and bladder cancer and pneumonitis is robust to alternative methods of data analysis. MATERIALS AND METHODS: We used the Richardson method to indirectly adjust risk ratios (RRs) in relation to AN exposure for potential confounding by smoking and asbestos. We repeated key analyses omitting workers from Plant 4 to account for possible local, historical shipyard-related asbestos exposures. RESULTS: The adjustment of lung cancer RRs for confounding by both smoking and asbestos and omitting Plant 4 workers yielded mostly decreased RRs and much less evidence of a positive association with cumulative AN exposure. CONCLUSION: Overall, our reanalysis provided little evidence to support NCI's suggestion of associations between AN exposure and mortality in lung and bladder cancer and pneumonitis.