(see Hazard Identification section). A LOAEL of 13.3 mg boron/kg-d, based on decreased fetal body weight and skeletal malformations in Sprague-Dawley rats, and a NOAEL of 9.6 mg boron/kg-d were identified. Heindel et al. (1992) identified a LOAEL of 13.6 boron/kg-d based on developmental effects.

Allen et al. (1996) calculated a benchmark dose for the developmental toxicity data associated with boric acid exposure in rats and determined that fetal weight, averaged within litters, is the most sensitive, relevant end points to use. Combining the data from Price et al. (1996a) and Heindel et al. (1992), a benchmark dose (BMD₅) of 59 mg boric acid/kg-d (10.3 mg boron/kg-d) was calculated for a 5% decrease in fetal weight. An expert scientific panel (consisting of 28 individuals from academia, industry, and government) concluded that the benchmark dose estimated by Allen et al. (1996) was appropriate (Moore 1997). The subcommittee agrees that a benchmark dose (BMD₅) of 10.3 mg boron/kg-d is appropriate for use in this risk assessment.

The default uncertainty factor for the derivation of the RfD for boric acid would be 100, composed of a factor of 10 for interspecies extrapolation and 10 for intraspecies variability (further broken down to a factor of 4 for toxicokinetic considerations and 2.5 for toxicodynamic considerations). It has been suggested that, in the case of boron compounds, the database is sufficient to warrant adjusting down the UF based on similar absorption, distribution, and elimination of boron in rats and humans, and lack of metabolism in either rats or humans. The International Programme on Chemical Safety (IPCS) (1998) used a total uncertainty factor of 25 when calculating their tolerable intake level for boron. An ad hoc panel, which included representatives from the EPA and Health Canada, that reviewed the data on the elimination of boron has suggested that a total UF of 60 be used (Dourson et al. 1998). The subcommittee, however, decided that an uncertainty factor of 30 should be used. Division of the BMD₅ of 10.3 mg boron/kg-d by the UF of 30 yields an RfD of 0.3 mg boron/kg-d for boric acid. The derivation of the oral RfD for boric acid is summarized in Table 8–5. Confidence in the critical studies for boric acid (Heindel et al. 1992; Price et al. 1996a) is high because they are well-conducted studies. There is a large body of literature indicating developmental and reproductive effects within the same order of magnitude, therefore, the confidence in the overall database is high. Therefore, the subcommittee has high confidence in the oral RfD for boric acid high.

In order to derive an oral RfD for zinc borate from the RfDs for zinc compounds and boric acid, the relative contributions of zinc and boron to zinc borate were determined. Boron comprises approximately 11.3% (w/w) of zinc