Lung cancer mortality among aircraft manufacturing workers with long-term, low-level, hexavalent chromium exposure.

Hexavalent chromium (CrVI) is known to cause lung cancer among workers exposed to high concentrations in certain historical industries. It is also a toxic air contaminant considered to pose a potentially significant cancer risk at comparatively low concentrations in urban air. However, very limited data currently exist to quantify risk at low-concentration occupational or environmental exposures. This study reconstructs individual-level exposures using a job-exposure matrix (JEM) and examines mortality among 3,723 CrVI-exposed aircraft manufacturing workers, including 440 women with long-term low-level CrVI exposures and long-term follow-up. The JEM used Bayesian methods with industrial hygiene data to calculate cumulative worker exposures from 1960 to 1998. A retrospective cohort mortality study was also conducted to calculate standardized mortality ratios (SMRs) by population demographics and to conduct an internally referenced dose-response analysis. CrVI-exposed painters, electroplaters, and aircraft assembly workers, with 1 to 37 years of exposure (median: 8 years) had mean and median cumulative exposures of 16 µg/m-yrs and 2.9 µg/m-yrs, respectively. Based on 1,758 observed deaths, mortality from cancer overall (SMR 1.24; 95% CI 1.13-1.36), smoking-related cancers (SMR 1.31; 95% CI 1.15-1.49), and lung cancer (SMR 1.39; 95% CI 1.17-1.63) were significantly elevated and more highly elevated among women (lung cancer SMR 2.61; 95% CI:1.66-3.92). Internal analyses revealed no dose-response relationship between cumulative exposure and lung cancer mortality. Data available for 12% of CrVI-exposed workers showed smoking prevalence higher than general population norms, especially for women. The absence of a dose-response relationship with cumulative exposure suggests that elevated cancer risks are primarily smoking-related in this cohort, and possibly as a consequence, any increased risk associated with CrVI exposure is not observable. Although an association between lung cancer risk and CrVI exposure was not found, this study provides significant new observations in the low exposure range, and among women, which may be useful for quantitative risk assessment.