Quantitative approaches to human risk assessment for noncancer health effects.

The estimation of risk for health effects due to chemical exposure is important to the development of standards for regulating the manufacture, use and release of chemicals into the environment. The quantitative data used to develop risk estimates usually come from laboratory animal studies employing relatively high dose levels. Thus, both interpolation from high to low dose levels and extrapolation from laboratory animals to humans are required. The approach most widely used for noncancer end points is to determine the no-observed-adverse-effect level (NOAEL) for the critical effect and then apply uncertainty factors (UFs) to account for scientific uncertainties in the total data base, such as response variability within and between species, the lack of chronic exposure data, the lack of a NOAEL, etc. The resulting value is a reference dose (RfD), i.e., the dose at or below which there is unlikely to be any excess risk. One difficulty with the NOAEL/UF approach is that it does not provide a basis for estimating risk at doses above the RfD; thus, if the exposure assessment indicates that human exposure is above the RfD, there is no way to judge the risk of that exposure. Alternative quantitative approaches for risk estimation of noncancer end points have been proposed and include both statistical and biologically-based dose-response modeling. Research is currently underway to further develop and explore the application of such approaches.