Methods for Health Risk Estimation and Uncertainty Analyses For Mixtures of Disinfection By-Products (DBPs)

摘要

The human health risk assessment of drinking water disinfection by-products (DBPs) requires a complex analysis because the exposure is a multiple route, ubiquitous, daily exposure to a highly variable complex mixture of chemicals at low concentrations. Evidence in the epidemiologic literature suggests that exposure to the complex mixture may cause cancer, reproductive or developmental effects. Thus, a complete risk assessment should include all factors that may contribute to potential health risk, including consideration of both measured and unmeasured DBPs. A response addition approach was used to illustrate DBP mixtures risk estimation using a two-stage Monte Carlo simulation. A Quantitative Structure Activity (QSAR) model was used to estimate toxicity for the unmeasured DBPs. Distributions of risks were produced for cancer and reproductive and developmental effects, reflecting population variance in tap water consumption and uncertainty in the DBP concentrations and toxicity estimates. Uncertain factors in the analysis were examined. It was concluded that three assumptions were not substantial sources of uncertainty: the measured and unmeasured DBPs have equal risk (per their concentrations); the proportion of unmeasured DBPs associated with a health endpoint can be determined using QSAR; and the mechanism of action for chloroform is not active at environmental exposure levels. However, factors in the analysis with significant uncertainty that warrant additional research include incremental risk values for the individual DBPs, estimates of the concentration of total organic halogen comprised of unmeasured DBPs, use of epidemiologic data to estimate risks, and application of other risk assessment models for comparison with the response addition approach. Future risk assessments should consider DBP mixtures risk as a multi-dimensional, cumulative risk problem. Potential health risk increases with increases in the number of DBPs considered, in the duration of exposure, and with the inclusion of additional exposure routes (i.e., oral, dermal and inhalation).