Following a request from the European Commission, the Panel on Food Additives and Nutrient Sources added to Food (ANS) of the European Food Safety Authority (EFSA) was asked to deliver a scientific opinion re-evaluating the safety of vegetable carbon (E 153) when used as a food colouring substance.
Vegetable carbon (also called ‘vegetable black’) is a form of finely divided carbon produced by steam activation of carbonized raw material of vegetable origin. Vegetable carbon is used both as a food colouring and as a medicinal substance, i.e. as an intestinal adsorptive drug or an antidote. This opinion relates only to the safety evaluation of vegetable carbon as a food colouring, derived from plant materials.
Vegetable carbon is authorised as a food additive in the EU and previously evaluated by the EU Scientific Committee for Food (SCF) in 1977 and 1983 (SCF 1977, 1984) and by the Joint FAO/WHO Expert Committee on Food Additives (JECFA) in 1970, 1977 and 1987 (JECFA 1971, 1978, 1987). Neither JECFA nor the SCF established an acceptable daily intake (ADI) for vegetable carbon, but the SCF concluded that vegetable carbon could be used in food.
The EINECS number 215-609-9 in the EC specifications for vegetable carbon (E 153) (Directive 2008/128/EC) corresponds to carbon black. In order to avoid confusion with carbon blacks manufactured from petrochemical sources, the Panel supported the proposal from Industry to replace the existing EINECS number for vegetable carbon (that is actually for carbon black) with the EINECS number for carbon (231-153-3).
Specifications have been defined in the EU legislation (Commission Directive 2008/128/EC) and by JECFA (JECFA, 2006). The purity is specified as ≥ 95%, calculated on an anhydrous and ash-free basis. Specifications include a limit for the presence of polycyclic aromatic hydrocarbons (PAHs) measured in comparison to the intensity of the fluorescence of a defined cyclohexane extract. With the proposed method it is not possible to get a quantitative measurement of the level of PAHs that may be present in a vegetable carbon.
Two vegetable carbon products with different particle sizes can be obtained from the manufacturing process and both products are used as food additives. For both products, the industry provided particle size distribution data obtained by laser diffraction. The standard vegetable carbon has a relative colouring power of 25 and the particle size distribution of the two commercial available preparations are characteristically 10% < 2 μm, 50% < 5 μm, and 90% < 55 μm or 10% < 1 μm, 50% < 4 μm, and 90% < 10 μm for the finer product, with an absence of particles below 275 nm (NATCOL, 2011a, 2011b). The Panel concluded that the presence of nanoparticles in vegetable carbon products currently on the market can be excluded. However the current description of the particle size in the specifications does not preclude production of vegetable carbon of a lower particle size.
There were no available data on the toxicokinetics of vegetable carbon. However, for carbon black nanoparticles from hydrocarbon sources, there was one study on gastrointestinal absorption in mice receiving 7 mg 7Be-labelled furnace black nanoparticles (27 nm diameter) by oral administration, showing that these particles were not absorbed. Since the particle size distribution of the standard vegetable carbon used as a food additive is characteristically 10% < 2 μm, 50% < 5 μm, and 90% < 55 μm or 10% < 1 μm, 50% < 4 μm, and 90% < 10 μm for the finer product, again with an absence of particles below 275 nm, the Panel considered that vegetable carbon microparticles may also be assumed to be essentially non-absorbed following oral administration.
There were no data available on the genotoxicity, chronic toxicity, carcinogenicity, and reproductive and developmental toxicity of vegetable carbon. The Panel took into consideration data from studies involving carbon blacks of hydrocarbon origin, considering whether read across would be possible.
The Panel considered that the carcinogenic and genotoxic effects observed after exposure to carbon black extracts can most likely be attributed to the carcinogenic PAHs adsorbed to carbon black and that these impurities are derived from the source material used to produce carbon black. Given that the source material of vegetable carbon is different from the source material used to produce carbon black, the Panel considered that vegetable carbon is unlikely to contain carcinogenic PAHs, or, if present, the carcinogenic PAHs level would be very low. Nevertheless, to further control the possible presence of these carcinogenic and mutagenic impurities, the Panel considered that it may be appropriate to introduce in the specifications for vegetable carbon a requirement for residual carcinogenic PAHs expressed as benzo[a]pyrene using a validated analytical method of appropriate sensitivity (e.g. with a limit of detection of 0.1 µg/kg).
Most in vitro mutagenicity studies of furnace carbon black, including several Ames tests, mouse lymphoma assays and mouse embryo morphological cell-transformation assays, provided negative results. Isolated positive findings with carbon black in vivo are considered secondary to treatment‑related chronic inflammation and related oxidative stress. Overall, the available data do not indicate a genotoxic hazard for carbon black. The positive results obtained in genotoxicity tests in vitro with some carbon black solvent extracts point to the possible presence of genotoxic polycyclic aromatic compounds absorbed onto carbon particles. The Panel noted that the organic compounds extracted by solvents are tightly bound to carbon particles, and may have limited bioavailability in physiological condition. Overall, the Panel considered that for vegetable carbon containing less than 1.0 µg/kg of residual carcinogenic PAHs expressed as benzo[a]pyrene, using a validated analytical method of appropriate sensitivity, the levels of genotoxic polycyclic aromatic compounds possibly present in vegetable carbon would be low enough to be of no genotoxic concern.
The Panel noted that the carcinogenicity of carbon blacks of hydrocarbon origin has been related to the PAH content of these substances and that their PAH content was much higher than the PAH content of vegetable carbon (E 153). Studies done with carbon blacks without PAHs did not show carcinogenic potential and the Panel considered that these data could be used for the evaluation of vegetable carbon. The Panel noted that in a long term toxicity and carcinogenicity study of carbon black without PAHs in rats and mice no adverse effects were observed at levels of carbon black in the diet that ranged from 9 to 18% in mice and rats (estimated to be equivalent to 4500‑9000 mg/kg bw/day for rats and 13 500‑27 000 mg/kg bw/day for mice).
The Panel noted that the available toxicological database on vegetable carbon is sparse and considered therefore that it could not derive an ADI.
When considering the maximum reported use levels, estimates reported for the UK adult population give a dietary exposure to vegetable carbon of 3.7 mg/kg bw/day at the mean and of 28.1 mg/kg bw/day for high level (97.5th percentile) consumers. The main contributors (> 10%) to the total anticipated mean exposure to vegetable carbon were confectionery (41%) and desserts including flavoured milk products (30%). When considering the maximum reported use levels, the dietary exposure of European children, ranged from 3 to 29.7 mg/kg bw/day at the mean, and from 15.3 to 79.1 mg/kg bw/day at the 95th/97.5th percentile. The main contributors to the total anticipated mean exposure to vegetable carbon (> 10% in all countries, these contributions differed per country) were desserts including flavoured milk products (33.7‑95.1%). Confectionery accounted for 11% to 58% of the total exposure in five countries.
The Panel decided to estimate the margin of exposure for benzo[a]pyrene in vegetable carbon using the BMDL10 of 0.07 mg/kg bw/day derived by the EFSA Scientific Panel on Contaminants in Food Chain (EFSA, 2008b). The Panel noted that the CONTAM Panel had estimated margins of exposure of 17 900 and 10 800 using the average and high level dietary exposures of benzo[a]pyrene across Europe. Using the BMDL10 values derived by the CONTAM Panel for benzo[a]pyrene, the margins of exposure for benzo[a]pyrene for mean and high level adult consumers of vegetable carbon would range from 25 to 189 million if benzo[a]pyrene were at the limit of detection (0.1 mg/kg) and from 2.5 to 18.9 million if benzo[a]pyrene were at the limit indicated in the OIV method (1 µg/kg) which was not exceeded by the PAHs content in any of the three commercial batches of vegetable carbon (E 153), as reported by NATCOL. The corresponding values for children would be 8.9 to 24 million and 0.9 to 2.4 million respectively. The Panel noted that these were considerably higher than the margins of exposure estimated from the dietary benzo[a]pyrene exposure.
The Panel noted that the BMDL10 values derived by the CONTAM Panel are based on cancer incidences observed in animals fed with coal tar containing known amounts of benzo[a]pyrene. The Panel considered that the bioavailability of carcinogenic PAHs present in coal tar would be far higher than for PAHs tightly absorbed on carbon black, which are released to a limited extent by organic fluids, this add a further margin of safety to the above estimates, which assume total bioavailability of PAHs absorbed on vegetable carbon. The Panel considered that vegetable carbon is not of concern with respect to carcinogenicity, provided the material of commerce contains less than 1.0 µg/kg of residual carcinogenic PAHs expressed as benzo[a]pyrene µg/kg, using a validated analytical method of appropriate sensitivity.
- the lack of absorption of vegetable carbon,
- the consideration that vegetable carbon is not of concern with respect to genotoxicity and carcinogenicity, provided the material of commerce contains less than 1.0 µg/kg of residual carcinogenic PAHs expressed as benzo[a]pyrene, using a validated analytical method of appropriate sensitivity,
- the history of safe use in medicine showing the absence of toxicologically relevant effects upon exposure to vegetable carbon or comparable carbon preparations for pharmaceutical use at levels 18 to 300 times higher than the mean estimated dietary exposure to vegetable carbon resulting from its use as a food colour, and
- the fact that the margins of exposure for PAHs resulting from the use of vegetable carbon as a food colour are much greater than those estimated to PAHs from the diet,
the Panel concluded that vegetable carbon (E 153) at the reported uses and use levels is not of safety concern.
In order to control their possible presence, the Panel noted that the EC specifications for vegetable carbon may need to be amended to include a requirement for residual carcinogenic PAHs expressed as benzo[a]pyrene using a validated analytical method of appropriate sensitivity (e.g. with a limit of detection of 0.1 µg/kg).
The Panel noted that the anticipated exposure to aluminium based on the dietary exposure vegetable carbon would contribute up to 22% of the olerable eekly ntake (TWI) of 1 mg/kg bw/week in high child consumers. The Panel also noted that the presence of aluminium as a contaminant of the colour could add to the daily intake of aluminium for which a TWI of 1 mg aluminium/kg bw/week has been established and that therefore a maximum level of aluminium in the specifications for vegetable carbon may be required.
The Panel noted that the JECFA specification for lead in vegetable carbon (E 153) is 2 mg/kg, whereas the EC specification is 10 mg/kg.
The Panel also noted that the current description of the particle size of vegetable carbon in the specifications does not preclude production of vegetable carbon of a lower particle size. In consequence, the Panel considered that should the particle size distribution of vegetable carbon change appreciably to include a significant content of particles below 275 nm, its use as a food additive would require re-evaluation. The Panel concluded that the EC specifications for vegetable carbon may need to be amended to include a restriction of the particle size (< 100 nm) in order to exclude the presence of nanoparticles.