Summary of Retrospective Asbestos and Welding Fume Exposure Estimates for a Nuclear Naval Shipyard and Their Correlation with Radiation Exposure Estimates

Abstract

In support of a nested case-control study at a U.S. naval shipyard, the results of the reconstruction of historical exposures were summarized, and an analysis was undertaken to determine the impact of historical exposures to potential chemical confounders. The nested case-control study (N = 4388) primarily assessed the relationship between lung cancer and external ionizing radiation. Chemical confounders considered important were asbestos and welding fume (as iron oxide fume), and the chromium and nickel content of welding fume. Exposures to the potential confounders were estimated by an expert panel based on a set of quantitatively defined categories of exposure. Distributions of the estimated exposures and trends in exposures over time were examined for the study population. Scatter plots and Spearman rank correlation coefficients were used to assess the degree of association between the estimates of exposure to asbestos, welding fume, and ionizing radiation. Correlation coefficients were calculated separately for 0-, 15-, 20-, and 25-year time-lagged cumulative exposures, total radiation dose (which included medical X-ray dose) and occupational radiation dose. Exposed workers’ estimated cumulative exposures to asbestos ranged from 0.01 fiber-days/cm ³ to just under 20,000 fiber-days/cm ³ , with a median of 29.0 fiber-days/cm ³ . Estimated cumulative exposures to welding fume ranged from 0.16 mg-days/m ³ to just over 30,000 mg-days/m ³ , with a median of 603 mg-days/m ³ . Spearman correlation coefficients between cumulative radiation dose and cumulative asbestos exposures ranged from 0.09 (occupational dose) to 0.47 (total radiation dose), and those between radiation and welding fume from 0.14 to 0.47. The estimates of relative risk for ionizing radiation and lung cancer were unchanged when lowest and highest estimates of asbestos and welding fume were considered. These results suggest a fairly large proportion of study population workers were exposed to asbestos and welding fume, that the absolute level of confounding exposure did not affect the risk estimates, and that weak relationships existed between monitored lifetime cumulative occupational radiation dose and asbestos or welding fume.

Keywords:

• confounding
• exposure assessment
• lung cancer

ACKNOWLEDGMENTS

The authors would like to thank Jeffrey Brann, James Braun, Mark Duval, and Christopher Moore at PNS; Sarah Locke, Susan Viet, Mary Engle, Karin Davis, Bin Chen, and Tim McAdams at Westat; and Lynette Hartle, Greg Kinnes, James Couch, Misty Hein, Sharon Silver, and Robert Daniels at NIOSH. Funding for this study was provided through an agreement between the U.S. Department of Energy (DOE) and the U.S. Department of Health and Human Services (DHHS).

The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the National Institute for Occupational Safety and Health.

Notes

^A Percent of workers whose cumulative exposure exceeded their respective cumulative ACGIH threshold limit value (TLV) or REL.

^B Workers who had at least one work history segment with an exposure to the hazard above background.

^C The NIOSH recommended exposure limit (REL, 15 μg/m³) is used for nickel due to its considerably lower value compared with the TLV.

^D Two workers had no exposure assigned to them due to an incomplete work history, hence N = 4386.

^E Values not calculated due to heavily censored data distribution.

^A From all sources, including diagnostic medical X-rays.

^B Excluding diagnostic medical X-rays.

^C Workers exposed above background to both asbestos and welding fume, no lag.

^D All correlation coefficients were statistically significant (p < 0.05).

^E Sample sizes vary because not all workers in the population had occupational or medical X-ray exposure, and because lags reduced sample sizes. In all cases, the equivalent lags were applied to both variables being correlated (e.g., radiation, welding fume, and asbestos).

^A Model with all potential confounders including sex, radiation monitoring status, socioeconomic status, and exposures listed above modeled log-linearly or linearly. RR and the corresponding 95% CI for these potential confounders (from the log-linear model) are very similar whether occupational dose is modeled log-linearly, linearly, or categorically.