Factors Influencing the Relationship Between Benzo(a)pyrene and Other Particulate Polycyclic Aromatic Hydrocarbons at a Soderberg Aluminum Smelter

Abstract

P-596 Introduction: Occupational and environmental exposure to polycyclic aromatic hydrocarbons (PAH) occurs as complex mixtures. Exposure to PAH mixtures, such as in primary aluminum smelting, increases cancer risks, but the role of individual PAHs is unknown. Benzo(a)pyrene (BaP) has routinely been used as a marker of PAH exposure, although a few other PAHs seem to be more potent carcinogens in animal studies. The PAH mixtures vary both between and within industries, which complicates the comparison of dose-response relationships across studies even when the same marker of exposure is used. Understanding the factors influencing the relationship between BaP and other PAHs within a mixture may allow for retroactively assessing exposures to these other components and will facilitate comparisons across studies and industries. We investigated the factors that influence the relationship between BaP and other PAHs within an aluminum smelter. Methods: We used 431 personal air measurements analyzed for BaP and 8 other particulate PAHs that were collected between 1978 and 2001 within the aluminum smelter potrooms. Using linear regression we examined differences in the log-transformed PAH/BaP ratio due to anode paste composition (4 types), potroom building, season, and job. The variable potroom building (3 locations) represented differences in technology, process conditions, and building properties between locations. The predicted PAH/BaP ratios were multiplied by BaP relative potency factors to provide an estimate of the relative toxicity of the mixture. Results: For three of the four anode paste types BaP was very strongly correlated with individual particulate PAHs (r> 0.93); however, 70% of the measurements were collected for one job, pot operator. In linear regression, between 23% and 89% of the variability in the log-transformed PAH/BaP ratio was explained by season, anode paste type, potroom building, and job, depending on the PAH. The relative toxicity of the mixture varied more across job (2.1–3.5) than across anode paste composition (1.8–2.8) or potroom building (2.3–2.5). BaP and benzo(ghi)pyrene accounted for 72–82% of the mixture's relative toxicity for the most common anode paste. Although dibenzo(a,h)anthracene is more potent, it accounted for only 1–2% of the mixture's carcinogenicity. Conclusions: Seasonal, job, and location differences in the relationship between BaP and other PAHs have not been previously reported. Assuming that the PAH mixture is constant for a given time period would not account for important job and location differences in the PAH mixture. As such, using a single PAH, such as BaP, would likely introduce exposure misclassification in epidemiologic studies.