Abstract

OBJECTIVES: Animal inhalation studies and theoretical models suggest that the pattern of formation of benzene metabolites changes as exposure to benzene increases. To determine if this occurs in humans, benzene metabolites in urine samples collected as part of a cross sectional study of occupationally exposed workers in Shanghai, China were measured. METHODS: With organic vapour monitoring badges, 38 subjects were monitored during their full workshift for inhalation exposure to benzene. The benzene urinary metabolites phenol, catechol, hydroquinone, and muconic acid were measured with an isotope dilution gas chromatography mass spectroscopy assay and strongly correlated with concentrations of benzene air. For the subgroup of workers (n = 27) with urinary phenol > 50 ng/g creatinine (above which phenol is considered to be a specific indicator of exposure to benzene), concentrations of each of the four metabolites were calculated as a ratio of the sum of the concentrations of all four metabolites (total metabolites) and were compared in workers exposed to > 25 ppm v < or = 25 ppm. RESULTS: The median, 8 hour time weighted average exposure to benzene was 25 ppm. Relative to the lower exposed workers, the ratio of phenol and catechol to total metabolites increased by 6.0% (p = 0.04) and 22.2% (p = 0.007), respectively, in the more highly exposed workers. By contrast, the ratio of hydroquinone and muconic acid to total metabolites decreased by 18.8% (p = 0.04) and 26.7% (p = 0.006), respectively. Similar patterns were found when metabolite ratios were analysed as a function of internal benzene dose (defined as total urinary benzene metabolites), although catechol showed a more complex, quadratic relation with increasing dose. CONCLUSIONS: These results, which are consistent with previous animal studies, show that the relative production of benzene metabolites is a function of exposure level. If the toxic benzene metabolites are assumed to be derived from hydroquinone, ring opened products, or both, these results suggests that the risk for adverse health outcomes due to exposure to benzene may have a supralinear relation with external dose, and that linear extrapolation of the toxic effects of benzene in highly exposed workers to lower levels of exposure may underestimate risk.