The relationship between the occupational exposure limits (OEL) and the lethal dose 50 (LD50) values of rats or mice for metals and metallic compounds was statistically analyzed by a stepwise multivariate regression method. The OEL values were predicted from LD50 values and metallic compensation coefficients (MCC), which were developed as the regression coefficients of dummy variables that represented the metallic element contained in the substance of interest. The value of the MCC indicated the extent of the adverse health effects of the metal in the substance. Smaller values of the MCC were assigned to metals that would have the more severe adverse health effects, such as carcinogenesis, while larger values were given to the less toxic metals. The Health Index (HI) based on the OEL values was proposed as a convenient measure of the toxicity of industrial products. The prediction method could be applied to toxicity risk assessments by using the HI when a designer of consumer products wants to use substances for which OEL values have not been determined. Two case studies were conducted to estimate the potential toxicity of materials used in solders and in rechargeable batteries.

No abstract

To ensure selection of appropriate materials for industrial use in terms of toxicity, a procedure to estimate a health index (HI) that can be used for prioritizing less hazardous components of consumer products is described. The HI is based on the occupational exposure limits (OELs) of organic substances that compose the products. To calculate the HI of a product, it is often necessary to predict OEL values for the substances for which an OEL value has not yet been promulgated. We developed a method to estimate the OEL values from median lethal dose (LD50) data of rodent. A good correlation between known OEL values and the LD50 data was obtained by a multivariate regression analysis by introducing organic compensation coefficients, which were caluculated as 10 to the Bth power where B is the regression coefficient of dummy variables denoting the characteristics of the organic compounds such as functional groups, molecular structures, and so on. We believe that the use of the present method should be limited to predicting unknown OEL values for the HI and used for material prioritization. It should not be extended to determinations of regulatory OELs.