Variability in airborne and biological measures of exposure to mercury in the chloralkali industry: Implications for epidemiologic studies

Exposure assessment is a critical component of epidemiologic studies, and m ore sophisticated approaches require that variation in exposure be consider ed. We examined the intra- and interindividual sources of variation in expo sure to mercury vapor as measured in air, blood, and urine among four group s of workers during 1990-1997 at a Swedish chloralkali plant. Consistent wi th the underlying kinetics of mercury in the body, the variability of biolo gical measures was dampened considerably relative to the variation in airbo rne levels. Owing to the effects of intraindividual variation, estimating w orkers' exposures from a few measurements can attenuate measures of effect. To examine such effects on studies relating long-term exposure to a contin uous heath outcome, we evaluated the utility of each exposure measure by co mparing the necessary sample sizes required for accurate estimation of a sl ope coefficient obtained from a regression analysis. No single measure outp erformed the others for all groups of workers. However, when workers were e valuated together, creatinine-corrected urinary mercery better discriminate d workers' exposures than airborne or blood mercury levels. Thus, pilot stu dies should be conducted to examine variability in both air and biomonitori ng data because quantitative information about the relative magnitude of th e intra- and interindividual sources of variation feeds directly into our e fforts to design an optimal sampling strategy when evaluating health risks associated with occupational or environmental contaminants.