Arsenic exposure and bladder cancer: quantitative assessment of studies in human populations to detect risks at low doses.

While exposures to high levels of arsenic in drinking water are associated with excess cancer risk (e.g., skin, bladder, and lung), exposures at lower levels (e.g., <100-200 µg/L) generally are not. Lack of significant associations may result from methodological issues (e.g., inadequate statistical power, exposure misclassification), or a different dose-response relationship at low exposures, possibly associated with a toxicological mode of action that requires a sufficient dose for increased tumor formation. The extent to which bladder cancer risk for low-level arsenic exposure can be statistically measured by epidemiological studies was examined using an updated meta-analysis of bladder cancer risk with data from two new publications. The summary relative risk estimate (SRRE) for all nine studies was elevated slightly, but not significantly (1.07; 95% confidence interval [CI]: 0.95-1.21, p-Heterogeneity [p-H]=0.543). The SRRE among never smokers was 0.85 (95% CI: 0.66-1.08, p-H=0.915), whereas the SRRE was positive and more heterogeneous among ever smokers (1.18; 95% CI: 0.97-1.44, p-H=0.034). The SRRE was statistically significantly lower than relative risks predicted for never smokers in the United States based on linear extrapolation of risks from higher doses in southwest Taiwan to arsenic water exposures >10 µg/L for more than one-third of a lifetime. By contrast, for all study subjects, relative risks predicted for one-half of lifetime exposure to 50 µg/L were just above the upper 95% CI on the SRRE. Thus, results from low-exposure studies, particularly for never smokers, were statistically inconsistent with predicted risk based on high-dose extrapolation. Additional studies that better characterize tobacco use and stratify analyses of arsenic and bladder cancer by smoking status are necessary to further examine risks of arsenic exposure for smokers.