Following a request from the European Commission, the Panel on Food Additives and Nutrient Sources added to Food (ANS) was asked to deliver a scientific opinion on dimethyl dicarbonate (DMDC) (E 242) when used as a food additive.
The Panel was not provided with a newly submitted dossier and based its evaluation on previous evaluations, additional literature that has become available since then and the data available following an EFSA call for data. The Panel noted that not all original studies on which previous evaluations were based were available to the Panel.
DMDC (E 242) is an authorised food additive in the EU for treatment of various beverages in accordance with Annex II to Regulation (EC) No 1333/2008. Specifications have been defined in the EU according to the Commission Regulation (EU) No 231/2012.
The use of DMDC for treatment of beverages has been evaluated by the SCF (1992, 1997, 2001) and by JECFA (1991a,b). Both Committees found the treatment of non-alcoholic beverages toxicologically acceptable with a treatment level (as ingoing amount) of 250 mg/L for non-alcoholic beverages and 200 mg/L for wine. No acceptable daily intake (ADI) was allocated.
DMDC is a very reactive substance and is used to inactivate microorganisms. In aqueous solution, DMDC rapidly (half-life between 15 and 20 minutes at 20°C) and fully decomposes into carbon dioxide and methanol, after which no antimicrobial effect remains. With no residue of DMDC remaining in the treated beverage ready for consumption, there is no exposure of the consumers to the substance as such. Therefore, the risk resulting from the use or DMDC as a food additive was assessed considering the main identified reaction products in treated beverages (dimethyl carbonate (DMC), methyl ethyl carbonate (MEC) and methyl carbamate (MC)) and the products of hydrolysis of DMDC (methanol and carbon dioxide). Biological and toxicological data were available for MC, MEC and DMC. Detailed analysis of the biological and toxicological data on methanol has been provided by the Panel on its opinion on the re-evaluation of aspartame as a food additive (EFSA ANS Panel, 2013a). Several DMDC-treated beverages have been tested as models for the evaluation of the toxicity of potential reaction products, which may result from the degradation of DMDC in treated-beverages.
Exposure assessment to the identified reaction products of DMDC from its use as a food additive was carried out based on: (1) Maximum permitted levels (MPLs) set down in the EU legislation (defined as the regulatory maximum level exposure assessment scenario) for DMDC; and (2) the reported use levels of DMDC (defined as the refined exposure assessment scenario).
Fruit juices, beer and white wine to which DMDC was added at a concentration of 4000 mg/L did not induce any adverse effects in short-term or subchronic toxicity studies in rats and dogs. The available data (in vitro and in vivo) did not report a genotoxic potential and no reproductive or developmental toxic effects were reported in rats drinking orange juice treated with DMDC at 4000 mg/L.
The Panel noted that MC has low acute toxicity with a LD50 > 2 000 mg/kg bw. Toxic hepatitis was observed in a 13-week study in female Fisher rat at 500 mg MC/kg bw/day, which was not seen in a further study with Wistar rats or mice. The Panel noted that MC was administered orally by gavage in F344 rats, whereas Wistar rats and mice were given MC in their drinking water; this may have resulted in peak MC concentrations in the liver of F344 rats. The Panel also noted that DNA binding was not detected in F344 rat liver. Finally, the increased incidence in hepatocellular adenomas and carcinomas (combined), which was reported in the 103-week National Toxicology Program (NTP) study with MC, was observed only in females in one strain of rat (F344) and neither in another strain (Wistar) nor in mice, and observed only at the high dose (200 mg MC/kg bw/day). The Panel considered this finding a consequence of the bioaccumulation of MC, which leads to inflammation and hyperplasia of the liver, which was found at a higher incidence in females than in males. Moreover, hepatocellular adenomas and carcinomas combined in the female control group of the NTP study were not observed in contrast to their historical controls, in which incidence varied from 0/100 to 12/100 in female F344 rats. The carcinogenic effect of MC was only observed in one sex and in one strain of rats and at a high dose of exposure (around 200 000 times higher than the highest calculated human exposure using a conservative estimate). There was no genotoxicity concern for MC. No data on the toxic effects of MC on reproduction and development were available. Exposure estimates of MC from the use of DMDC as a food additive ranged from 0.01–0.05 µg/kg bw/day at the mean for the elderly using the non-brand-loyal scenario and from zero to 1.2 µg/kg bw/day at the high level (95th percentile) for toddlers using the regulatory maximum level exposure assessment scenario. The Panel noted that the exposure to MC remains below the threshold of toxicological concern (TTC) of 30 µg/kg bw/day for this Cramer class I substance.
The Panel noted that MEC has also a low acute toxic potential (LD50 > 15 000 mg/kg bw). A NOAEL of 1 094 mg/kg bw/day, the highest dose tested, was identified from a 90-day toxicity study in rats. MEC did not show developmental toxicity up to 1250 mg/kg bw/day when given to rats. No genotoxicity, long-term toxicity, carcinogenicity and reproductive toxicity studies were available. The Panel determined that there were no structural alerts for genotoxic effects using the OECD QSAR Toolbox for MEC. Exposure estimates of MEC from the use of DMDC as a food additive ranged from 0.001 to 0.08 µg/kg bw/day at the mean level for infants and from 13 to 59 µg/kg bw/day at the high level (95th percentile) for the elderly. The Panel noted that exposure to MEC remains below the TTC of 30 µg/kg bw/day for this Cramer class I substance at the mean but was exceeded only for high-level consumers in the adults and the elderly population.
The Panel noted that DMC had a low acute toxicity (LD50 > 10 000 mg/kg bw in mice and rats) and a NOAEL of 890 mg/kg bw/day, the highest dose tested, was identified from a 90-day toxicity study in rats. No genotoxicity, long-term toxicity, carcinogenicity and reproductive and developmental toxicity studies were available. The Panel determined that there were no structural alerts for the genotoxic effects using the OECD QSAR Toolbox for this compound. Exposure estimates of DMC from the use of DMDC as a food additive ranged from 1 µg/kg bw/day at the mean for the elderly (irrespective of the scenario) to 24 µg/kg bw/day at the high level (95th percentile) for toddlers (regulatory maximum level exposure assessment scenario and brand-loyal scenario). The Panel noted that exposure to MC remains below the TTC of 30 µg/kg bw/day for this Cramer class I substance.
When non-alcoholic beverages and wine are treated with the maximum permitted level of 250 mg DMDC/L, the theoretical resulting concentration of methanol could be 120 mg methanol. The Panel evaluated the risks associated with methanol and its subsequent metabolism to formaldehyde in the opinion on aspartame (EFSA ANS Panel, 2013a). The Panel estimated that the exposure to methanol from all sources (basal endogenous pathway and endogenously metabolised pectin, natural food occurrence through the diet and also methanol resulting from the use of aspartame as a food additive) in five population groups would range from 8.4 to 18.9 mg/kg bw/day at the mean and from 15.1 to 35.1 mg/kg bw/day for high-level consumers (EFSA ANS Panel, 2013a). Throughout the scenarios and population groups, methanol exposure assessment ranged at the mean from 0.1 to 0.2 mg/kg bw/day for the elderly, to <0.1–5.8 mg/kg bw/day at the high level (95th percentile) for toddlers (MPL and brand-loyal scenarios). Accordingly, the Panel considered that the additional methanol arising from DMDC at the currently permitted uses and use levels, as a food additive, do not constitute a significant additional risk above the risk from the natural occurrence in foods and the endogenously produced methanol.
Overall, the Panel considered that:
- once in solution, DMDC quickly and fully hydrolyses and/or reacts, with different constituents of the treated beverages, and therefore, with no residue of DMDC remaining in treated beverage ready for consumption. The risk resulting from DMDC use as a food additive can therefore be assessed by considering the exposure to methanol (a product of its hydrolysis) and to the main identified reaction products (DMC, MEC and MC);
- there is limited information available on the identity and toxicity of the possible reaction products, other than DMC, MEC and MC, that might be formed after interactions of DMDC with other food components and/or food additives or flavourings that may be present simultaneously in a beverage;
- no relevant and significant new data were identified since the last evaluation (SCF, 1992), with the exception of genotoxicity data;
- the available toxicity database for the main identified reaction products formed in beverages treated with DMDC (DMC, MEC and MC) is limited but did not raise any concern for genotoxicity;
- the main identified reaction products formed in beverages treated with DMDC (DMC, MEC and MC) belongs to the Cramer class I, for which the TTC is 30 µg/kg bw/day;
- the exposure to the main identified reaction products from the use of DMDC as a food additive is very low and remains below the TTC even when using a conservative approach for the exposure calculation. The only exception being exposure to MEC for high-level consumers in the adults (< 0.01–54 µg/kg bw/day) and the elderly (13–59 µg/kg bw/day);
- the exposure to methanol arising from the use of DMDC at the currently permitted uses and use levels as a food additive does not constitute a significant additional risk greater than the risk from the natural occurrence in foods and endogenously generated methanol.
The Panel concluded that:
- it was not possible to derive an ADI from the available toxicological database;
- there is no indication for a safety concern from the use of DMDC (E 242) as a food additive at its currently reported uses and use levels;
- a new assessment would be warranted in case of any change in the conditions of use.
The Panel recommended that:
- the maximum limits for the impurities of toxic elements (lead, mercury and arsenic) in the EC specification for DMDC (E 242) should be revised in order to ensure that DMDC (E 242) as a food additive will not be a significant source of exposure to those toxic elements in food;
- in order to reduce uncertainties, more information should be generated on the nature and quantity of the reaction products resulting from interaction of DMDC with the components of beverages where its use is permitted (e.g. red wines, tea, cider and perry) and with food additives which could be used in the same beverage;
- in the absence of these data, and in order to limit the formation of any reaction product, other food additives should be added in the beverage only after complete degradation of DMDC.