Chemical Mixtures 



shellfish. This fractional RfD would then replace the RfD in the 

 denominator of the hazard index. The index would be compared to a 

 value of 1.0 to evaluate the potential for concern. However, the uncer- 

 tainties associated with such an approach should be clearly stated. 

 Further research on this problem is clccu-ly needed. 



The margin of exposure (MOE) is an alternative indicator of noncar- 

 cinogenic risk. The MOE is the ratio of the NOAEL to an estimated 

 exposure dose. When the MOE is equal to or greater than the product 

 of the uncertainty factor and the modifying factor used to derive the 

 RfD, the level of regulatory concern is usually low (see U.S. EPA 1987a 

 for details of the derivation of MOE). Concerns about mixed-species 

 diets and exposures from non-fishery sources, as discussed above for 

 hazard indices, also apply to MOE for exposure to contaminated 

 fisheries. 



U.S. EPA (1986d) discussed various models for assessment of the 

 upper limit to risk from chemical mixtures. Because of present data 

 limitations and the complexity of possible contaminant interactions, it 

 is virtually impossible at present to predict synergistic or antagonistic 

 effects of most chemical mixtures. The approach used most frequently 

 for multiple-chemical assessment is the additive-risk (or response-ad- 

 ditive) model. Thus, total upper-limit risk for a chemical mixture is 

 usually estimated as the sum of upper-limit risks for carcinogens or of 

 hazard indices for noncarcinogens. A sum of noncarcinogenic hazard 

 indices should be calculated only for a group of chemicals acting on 

 the same target organ (Stara et al.l983). The numerical estimates 

 obtained using the response-additive model are useful in terms of 

 relative comparisons (e.g., among fishing areas or among fishery 

 species). However, risk estimates for chemical mixtures should be 

 regarded only as very rough measures of absolute risk (U.S. EPA 

 1986d). Because technological Umitations preclude analyzing fishery 

 samples for all potentially toxic chemicals, risk estimates for chemical 

 mixtures should not be interpreted as estimates of total chemical risk 

 associated with ingestion. 



66 



