Risk assessment of mixtures and standard setting: working towards practical compromises.

Basic concepts of 'dose/concentration additivity' and 'response addition/independence' may be applied to evaluate chemical mixtures in human toxicology, as well as in ecotoxicology. In the case of compounds that cause the same toxicological effect by the same mechanism, 'dose addition' is a more plausible form of joint action than 'response addition'. Data on the effects of halogenated hydrocarbons on the kidney, the effects of organic solvents on the central nervous system, and the effects of organophosphates on cholinesterase, are the basis of this assumption. For such compounds, response addition will generally underestimate risk. None the less, both dose addition and response addition are 'non-interactive' forms of joint action. As such, neither additivity assumption will accurately predict risk for compounds that exhibit toxicological interactions regardless of the primary mechanism(s) of toxicity. More often, the additivity approach overestimates the risk of a combination of chemicals. From a public health perspective, such results over-protect the public; hence this approach can be used for standard setting. The introduction of a special safety factor of 10 for the standard setting for mixtures in addition to those normally used for deriving acceptable daily intakes, reference doses or minimal risk levels is not supported by data. Instead, each exposure scenario should be considered on a case-by-case basis. Furthermore, considerations of multiple endpoints, including multiple organs, multiple effects, multiple mechanisms and potential interactions between such mechanisms, are very desirable for overall toxicity and risk assessment of chemical mixtures. In conclusion, additivity could be used to estimate potential risks for a combination of chemicals: only if scientific data support independence of effects can response independence be used as an alternative for standard setting.