Inferring consequences from these effect metrics is challenging but essential. Typical for human effects studies and increasingly common for ecotoxicological studies, the NOEC can be adjusted in a conservative manner to estimate a reference dose (RfD), reference concentration (RfC), or acceptable daily intake (ADI). A common set of adjustments is the following for chronic human exposure:
where UF1 = uncertainty adjustment accounting for variation in natural sensitivity within human populations, UF2 = uncertainty adjustment if extrapolation was performed from animal data to effects to humans, UF3 = uncertainty adjustment if the NOEC comes from a subchronic test data set, UF4 = uncertainty adjustment if the LOEC is used in the calculation instead of an NOEC, and MF = an additional adjustment based on professional judgment. The value for any of these factors can range from 1 to 10 depending on the associated level of uncertainty. This or a similar equation is used to estimate an RfD in the following manner. The literature is searched for all relevant NOEC/LOEC data for the toxicant of interest. The study with the lowest relevant LOEC is identified and the associated NOEC used in the calculation (UF4 = 1). If only the LOEC is available, then the LOEC is applied instead with an UF4 = 10. In the case of chronic human exposure, the RfD is used to estimate the dose level thought to be below that which will cause an adverse effect during chronic exposure. Several types of RfDs are relevant to environmental exposures including short term, subchronic, chronic, or developmental RfDs. The RfD or RfC values may also be developed for different routes of exposure.
With sufficient knowledge, dose- or concentration-effect models can also be applied to estimation of RfD or RfC values. The benchmark dose (BMD) approach uses regression model predictions for a specified effect level (benchmark response) instead of an NOEC or LOEC to estimate the RfD or RfC. Often the lower 95% confidence limit for the estimated BMD (BMDL) is used instead of the NOEC in the above equation to estimate a BMD-based RfD. The UF and MF values can be the same or lower than those used for the NOEC-based approach. Taking the mysid shrimp data as an example, the BMD10 could be used to estimate a certain RfC. The BMD10 is predicted to be 7.2% effluent. This is the predicted lower 95% confidence interval of the LC10 (10.9% effluent) generated from the log normal (probit) model. As another example, the Environmental Protection Agency (EPA) applied such a BMD approach to determine a human chronic oral methylmercury exposure RfD. Using information from several epidemiological studies, the BMD associated with the lowest 5% of methyl-mercury-exposed children (BMD5) was chosen as the basis for calculation of the RfD. The primary advantage of this BMD-based approach is that it avoids many of the shortcomings described earlier for the hypothesis test-related effect metrics.
Regardless of how it is calculated, an RfD is used with information about exposure (e.g., inhalation rates, ingestion rates, bioavailability, and exposure duration) to calculate a maximum allowable concentration (MAC, maximum permitted concentration in a particular source such as food, air, drinking water, or soil) or level of concern (LOC, the concentration in the relevant medium above which an adverse effect could manifest).
Protection of human health is facilitated with a set of RfD or RfC values for various exposure scenarios such as acute, prolonged, lifetime, or developmental exposure. Relative to ecotoxicological testing, calculations associated with estimating 'safe' or acceptable exposures are not as straightforward, requiring consideration of consequences at different stages of life cycles of many species and several levels of biological organization. Partial and complete life cycle tests have emerged to address this requirement. A series of tests are conducted at each major life stage of a species, quantifying important effects notionally linked to an individual organism's fitness. The lowest effect metric for the various tests in such a complete life cycle test is used to generate regulatory limits or goals. The cost and difficulty of performing a complete life cycle test has given rise to a less inclusive set of tests (partial life cycle tests) that assess only the life stages thought to be most sensitive. Often these are the early life stages, leading to a battery of tests called early life stage tests. The emphasis during the interpretation of partial or complete life cycle tests is on protection of the individual; however, the EPA stresses the importance of considering population protection for most nonendan-gered or nonthreatened species existing in ecological communities. That the conventional interpretation of life cycle test-generated effect metrics does not directly address population or community level consequences of exposure is seen as a significant shortcoming in this approach as currently practiced in ecotoxicology. Fortunately, resolving this incongruity between metrics generated with current ecotoxicity tests and prediction of population- and community-level consequences is currently a very active area of research.
Was this article helpful?