Update Hazard Assessment of Bisphenol A : Abstract

We closely investigated the information on bisphenol A (BPA) regarding its potential as a human health hazard from 2005 onwards, and updated the hazard assessment of BPA. Since reproductive toxicity in the next generation was one of concerns with regard to the human health hazard induced by BPA, a two-generation reproductive study using mice under OECD GLP was carried out. However, no toxic effects on the reproductive potential of the next generation except for a slight prolongation of gestational length of F1/F2 at 300 mg/kg bw, and a no-observed-adverse-effect level (NOAEL) of 50 mg/kg bw have been noted. Regarding the general toxicity of BPA, multinucleated giant hepatocytes, centrilobular hepatocytomegaly, centrilobular hepatocyte hypertrophy and nephropathy were observed in mice. Considering that the NOAEL of centrilobular hepatocyte hypertrophy (3 mg/kg bw) found in mice was the lowest, this finding was determined as the endpoint of general toxicity upon oral administration of BPA.

The carcinogenicity of BPA by oral administration has already been determined to be negative from bioassays. With regard to skin irritation, skin sensitization, skin photo-irritability, and photo-sensitization due to BPA, it was believed that there is almost no need for concern because these were found to be negative in animal testings at a practical dose level.

Regarding the developmental neurotoxicity of BPA, a GLP-compliant rat developmental neurotoxicity study on BPA under the OECD testing guideline 426 and testing guideline 870.6300 of the U.S. EPA OPPTS was performed. However, evaluating the developmental neurotoxicity of BPA in our institute was not carried out because the validity of the testing protocol on chemical compounds with estrogenic activities has not yet been proven, although the protocol is valid to detect known developmental neurotoxicants. Moreover, the RISS determined that any influence on the brain function as well as behavior of children exposed to BPA in-utero or via breast milk cannot currently be evaluated because reports related to sexually differentiation of the brain, sexual behavior, social behavior, brain neurotransmitters, receptor expression, etc. in experimental animals caused by prenatal or neonatal exposure of BPA were all too uncertain to conclude as being adverse to humans.

Accordingly, NOAEL for the hazard assessment of BPA was determined to be 3 mg/kg bw/day, with the uncertainty factor of 25 (= species difference: 2.5 x individual difference: 10), although NOAEL and uncertainty factor were determined to be 5 mg/kg bw/day and 100, respectively, in our previous publication (Nakanishi et al., 2005). In the present assessment, the uncertainty factor related to species difference with regard to the extrapolation of animal data to human has been determined to be 2.5 because BPA has been shown to be detoxified mainly by glucuronide-conjugation in the liver, but it is rapidly metabolized and excreted in humans in comparison to rodents.

According to the BPA exposure estimate in Japanese individuals (Miyamoto and Kotake, 2006), exposure was highest in 1 to 6 year-olds, with an estimated 95% tile value of 3.9 μg/kg bw/day (men) and 4.1 μg/kg bw/day (women). In addition, the 95% tile value of BPA intake estimated from the amount of BPA excreted in 24-hour urine in adults was 0.037 to 0.064 μg/kg bw/ day (men) and 0.043 to 0.075 μg/kg bw/day (women). Using the 95% tile value of these exposure estimates and the NOAEL (3 mg/kg bw/day) from the animal testings, the Margin of Exposure (MOE) became 730 to 770 in 1 to 6 year-olds, and 40,000 to 81,000 in adults.

These values were much larger than the MOE (25) estimated to cause health hazards in humans mentioned above or the conventional and conservative MOE (100), and thus the risk of BPA with regard to human health was believed to be very small.