Following a request from the European Commission, the Panel on Contaminants in the Food Chain (CONTAM Panel) was asked to deliver a scientific opinion on tetrabromobisphenol A (TBBPA) and its derivatives in food.
TBBPA and its derivatives are a widely used group of flame retardants. TBBPA is primarily used as a reactive flame retardant, covalently bound to epoxy and polycarbonate resins. It is also used as an additive flame retardant in the manufacture of acrylonitrile butadiene styrene (ABS) resins, high impact polystyrene (HIPS) and phenolic resins. TBBPA derivatives may be used as either reactive or additive intermediates in polymer manufacture. When used as additive flame retardants, TBBPA and/or its derivatives might leach from the products into the environment.
TBBPA is susceptible to photolysis, and reductive debromination has been observed under experimental conditions.
A call for data on brominated flame retardants (BFRs) including TBBPA was issued by EFSA in December 2009. EFSA collected and evaluated the results reported from the analysis of 652 food samples, reported by four European countries (Ireland, Norway, Spain and the UK). These results covered the period from 2003 to 2010. The dominant food category was “Fish and other seafood (including amphibians, reptiles, snails and insects)” (n=465), followed by “Meat and meat products (including edible offal)” (n=49), “Milk and dairy products” (n=40), “Animal and vegetable fats and oils” (n=41) and “Eggs and egg products” (n=27). Less than 10 samples were reported for the remaining food categories. All analytical results on TBBPA were reported as less than the limit of quantification (LOQ) (in general ≤ 1 ng/g wet weight). Therefore a meaningful exposure assessment for the general population is not possible. In order to provide some indication of whether there could be a possible health concern with respect to dietary exposure to TBBPA, the CONTAM Panel made a worst case intake estimate for two specific groups of the population, i.e. adult high fish consumers and high cow’s milk consumers (toddlers), by substituting the concentration levels of TBBPA in fish and in cow’s milk, all reported as not quantified, by the maximum LOQ reported for those food categories of 1 ng/g and 0.65 wet weight, respectively. The resulting “upper bound” intake estimate was 2.6 and 55.7 ng/kg body weight (b.w.) per day, respectively.
Data on levels of TBBPA in human milk are scarce. For 3 months old breast-fed infants with average human milk consumption (800 mL per day) concentrations of TBBPA in human milk (ranging from 0.06 to 37.3 ng/g fat) result in daily exposures of 0.28 to 171 ng/kg b.w. For infants with high human milk consumption (1 200 mL per day) the respective daily exposures range from 0.41 to 257 ng/kg b.w.
The limited toxicokinetics data suggest that oral bioavailability of TBBPA in rats is about 70 %. After absorption, TBBPA is distributed in different tissues and rapidly excreted via the bile in faeces. Glucuronide or sulphate conjugates of TBBPA were identified in the bile. Tribromobisphenol A has been identified in faeces, suggesting that debromination of TBBPA can occur in mammals. The plasma half-life in rats was estimated to be about half a day. In humans, the half-life of TBBPA-glucuronide in plasma appeared to be between 48 and 72 h.
Toxicological studies with TBBPA have been carried out using different experimental designs with single or repeated administration during gestation, postnatally or in adulthood. The main target for TBBPA toxicity is thyroid hormone homeostasis.
The limited studies available do not indicate reproductive or teratogenic effects of TBBPA.
The available in vitro data indicate that TBBPA is not genotoxic. There are no long-term toxicity/carcinogenicity studies for TBBPA. However, based on the weight of evidence (absence of genotoxicity in vitro, no indications for proliferative changes or cytotoxicity in studies with up to 90 days repeated administration, no immunosuppression, except possibly at high doses), the CONTAM Panel concluded that there are no indications that TBBPA might be carcinogenic.
The CONTAM Panel identified a lower confidence limit for a benchmark response of 10 % (BMDL10) of 16 mg/kg b.w. reported for changes in thyroid hormone levels (decrease in circulating T4) as the critical reference point. The CONTAM Panel concluded, however, that due to the limitations and uncertainties in the current database, the derivation of a health based guidance value for TBBPA was not appropriate. Therefore, the Panel used a margin of exposure (MOE) approach for the risk characterisation of TBBPA.
Comparison of the “upper bound” dietary exposure estimate of 2.6 ng/kg b.w. per day for the specific group of adult high fish consumers with the BMDL10 of 16 mg/kg b.w. resulted in an MOE of 6 × 106. In the case of high cow’s milk consumers (i.e. toddlers), it resulted in an MOE of 3 × 105.
Usually an MOE of 100 is sufficient to cover uncertainties and variability with respect to kinetic and dynamic differences between animal species and humans and within the human population, and to conclude that there is no health concern. In the case of TBBPA, the CONTAM Panel noted that an additional factor would be needed to cover deficiencies in the database. The MOEs of 6 × 106 and 3 × 105 reported for the “upper bound” exposure scenario for adult high fish consumers and high cow’s milk consumers (i.e. toddlers) are, however, sufficiently large that the CONTAM Panel concluded that the current dietary exposure to TBBPA for these specific population groups with potential high exposure does not raise a health concern.
The limited data for other food groups, all reported as less than the LOQs, did not facilitate a hypothetical exposure assessment. However, since the reported LOQs for these food groups are in general below those reported for milk and fish, and given the large MOEs for adult high fish consumers and high cow’s milk consumers (i.e. toddlers), the CONTAM Panel concluded that it is unlikely that current dietary exposure of the general population to TBBPA raises a health concern.
For breast-fed infants with average or high human milk consumption, MOEs ranging from 6 × 107 to 9 × 104 and 4 × 107 to 6 × 104 have been estimated, respectively. Because of these large MOEs, the CONTAM Panel concluded that exposure via human milk also does not raise a health concern.
Dust in homes, classrooms and cars can be an additional source of exposure to TBBPA, particularly for children. Considering the 95th percentile TBBPA concentration in dust of 460 ng/g, the exposure based on a typical or high end exposure scenario would be 1.2 or 4.6 ng/kg b.w. per day, respectively. The CONTAM Panel concluded that the ‘typical’ exposure scenario provided the most realistic estimate of exposure to TBBPA from dust. Comparing the exposure resulting from this scenario (1.2 ng/kg b.w.) with the BMDL10 of 16 mg/kg b.w. results in an MOE of about 1.3 × 107. This MOE also indicates that exposure of children to TBBPA from dust does not raise a health concern.
Based on the large MOEs derived for both dietary exposure and exposure through dust, the CONTAM Panel concluded that it is unlikely that combined exposure through food and dust would result in a health concern.
For TBBPA derivatives no occurrence data had been submitted to EFSA and no information on their toxicity was identified.
The CONTAM Panel recognised that there is a need for information on production rates, use, chemical characteristics, occurrence in food, especially for food categories other than fish, and toxicity of TBBPA derivatives.