intraspecies factor).2 EPA then makes its decision with regard to the FQPA mandate, which requires it to apply up to an additional 10-fold factor to take into account the potential for increased sensitivity for fetuses and children. The statute allows EPA to apply a factor other than 10 (i.e., lesser or greater) if reliable data are available to show that this different factor is protective of infants and children. The cumulative effect of this approach to determining safe levels of exposure to pesticides is a potential 1,000-fold margin of safety between the NOAEL in animals and allowable exposures in humans. It has long been EPA’s practice to adjust the interspecies and intraspecies uncertainty factors if justified by scientific evidence showing that a different factor would provide a more scientifically sound or “accurate” extrapolation from the animal test results.

In response to FQPA, several pesticide manufacturers conducted and submitted to EPA intentional oral dosing studies involving humans for purposes of determining a NOAEL that might justify the reduction or elimination of the interspecies safety factor for certain pesticides in the widely used OP and carbamate classes. The submission of these studies has generated substantial controversy. Although it is not unusual or controversial for EPA to rely on human-derived data in its risk assessments, such data are typically derived from case reports, observational studies, or epidemiological studies that do not involve intentional dosing of humans.

In part, the pesticide studies involving humans are controversial because they were conducted by economically interested third parties, whose motivation was to justify reducing the interspecies uncertainty factor, thereby increasing the acceptable or safe human exposure level and possibly permitting the continuation of certain pesticide uses that might otherwise have been precluded under FQPA’s new safety standards. Some scientists and environmental and other public interest groups challenged the ethical and scientific validity of the studies, contending among other things that people should not be put at risk for the purpose of reducing

2  

Application of additional uncertainty factors in deriving the RfD may be necessary (1) to account for the lack of chronic data if deriving a traditional, chronic RfD (i.e., the subchronic-to-chronic factor), (2) to extrapolate from a LOAEL (lowest observed adverse effect level) to an estimated NOAEL, if no appropriate NOAEL can be identified in the toxicity database (the LOAEL-to-NOAEL factor), or (3) to account for the absence of key data in the toxicity database for a given chemical (the database factor). The default values for the inter- and intraspecies uncertainty factors are 10; those for the other three generally range from 3 to 10. Of course, EPA has the discretion to modify any of these default uncertainty factors if justified by the available scientific evidence.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement