This article was downloaded by: [McGill University Library] On: 04 February 2015, At: 04:09 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Food Additives & Contaminants Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tfac19
Safety evaluation of substances consumed as technical ingredients (food additives) E. Poulsen
a
a
National Food Agency of Denmark , M⊘rkh⊘j Bygade 19, S⊘dborg, DK‐2860, Denmark Published online: 10 Jan 2009.
To cite this article: E. Poulsen (1991) Safety evaluation of substances consumed as technical ingredients (food additives), Food Additives & Contaminants, 8:2, 125-133, DOI: 10.1080/02652039109373963 To link to this article: http://dx.doi.org/10.1080/02652039109373963
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sublicensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
FOOD ADDITIVES AND CONTAMINANTS, 1991, VOL. 8, NO. 2, 125-134
Symposium Paper Safety evaluation of substances consumed as technical ingredients (food additives) E. POULSEN National Food Agency of Denmark, Mørkhøj Bygade 19, DK-2860, Sødborg, Denmark
Downloaded by [McGill University Library] at 04:09 04 February 2015
(Received 10 July 1990; revised 19 October 1990) The different types of acceptable daily intakes (ADIs) are described as used by the FAO/WHO Expert Committee on Food Additives (JECFA) and the EEC Scientific Committee for Food (SCF). The allocation is discussed of a full ADI or a temporary ADI, and examples are given for the establishment (or withdrawal) of these ADIs. The flavours cinnamyl anthranillate and the solvent 2-nitropropane (both withdrawn), the sweeteners cyclamate and saccharin and the antioxidant BHA (all three changed) but not abolished. For BHA and saccharin the ADI was retained by both committees in spite of some evidence of carcinogenicity to experimental animals. ADI—'not specified' is specially discussed and it is recommended that numerical ADIs are used whenever possible. With an ADI—'not specified' it should be stated which use (and intake) levels are toxicologically acceptable. Some compounds evaluated by the two committees are discussed, e.g. the colours: Allura red AC, erythrosine, canthaxanthin and the caramels; three anti-oxidants: BHA, BHT and the gallates; the sweeteners: polyols, aspartame, saccharin and cyclamates. Four recommendations are made: (1) a numerical basis be given for the levels allocated an ADI—'not specified' or 'acceptable'; (2) lowering of the conventional safety factor be considered when the effects found are trivial—higher safety factors be considered when the toxic effects are serious or even irreversible; (3) ADIs should, whenever possible, be based on a combination of human and animal data; (4) ADIs might be allocated to compounds indicating animal carcinogenicity, if the compound is non-genotoxic, the mechanism clearly secondary and/or species-specific. Keywords: food additives, safety evaluation
Types of ADI A full ADI will be established when the required toxicological information is available on a food additive. The necessary information is discussed in recent guidelines and reviews (CEC 1980, WHO 1987). The required information varies according to the nature of the compound and the biological effects found in the studies. There have been cases where, for example, JECFA has found the information of such a nature as to recommend not just no ADI, but that 'the compound should not be used' (as a food additive or solvent). Previously established ADIs have been withdrawn or made temporary (and often lowered). The next ADI, the temporary ADI, indicates that the information is inadequate, or that new information raises doubt, which can only be resolved after further study. Such study is necessary, and will generally be forthcoming. In most cases where the ADI is changed to temporary, the compound will continue in use as an approved additive, maybe with restrictions in use levels and/or number of allowed food groups. Often these changes in the ADI come about 0265-203X/90 $3.00 © 1991 Taylor & Francis Ltd.
Downloaded by [McGill University Library] at 04:09 04 February 2015
126
E. Poulsen
because results of new animal studies give indications of carcinogenicity of the additive. Some illustrative examples are the sweeteners cyclamates and saccharin, the antioxidant BHA, the flavouring material cinnamyl anthranilate and the solvent 2nitropropane. Cyclamates, cinnamyl anthranilate and 2-nitropropane have had their acceptance withdrawn (FAO/WHO 1990), but for cyclamates an ADI has now been established by JECFA (FAO/WHO 1982) while SCF has established a temporary ADI (CEC 1985). For saccharin the ADI has been changed to a TADI by both committees, but not abolished (CEC 1985, FAO/WHO 1984). For BHA the ADI was also temporary from both committees (CEC 1990b, FAO/WHO 1987) but JECFA has very recently established a full ADI (FAO/WHO 1989). For both these compounds the IARC monographs evaluate that there is sufficient evidence for carcinogenicity in experimental animals, and the compounds are classified in IARC class 2B, possible carcinogens for humans (IARC 1987). I think it is rather remarkable that the two expert committees have maintained a temporary (or full) ADI for these two compounds. This means that they are approved food additives in most countries, although the appearance of these food additives on national and international lists of carcinogens will undoubtedly raise discussions in some countries. It is my judgement—at least for the SCF—that an animal carcinogen with clear genotoxicity will not be allocated an ADI, while an ADI might very well be established for a compound which is believed to be a non-genotoxic carcinogen or a promoter. It is also my judgement in these last cases that the figure for the ADI would be calculated from a no-effect level (NEL) by the use of a safety factor, maybe larger than the usual 100. After discussion of the types: No-ADI, temporary ADI and full ADI, this paper will discuss other types of ADI: (1) 'not-specified' ADI and (2) acceptable (for use as). I should make it clear that I am not particularly fond of these types of ADI. I think that, when applied, they leave too much to the manufacturer in a rather uncontrolled way. I know, of course, that most manufacturers are completely safety-conscious and knowledgeable about good manufacturing practice (GMP), but they will be without any indications from toxicologists on the level of use for which the compound is found to be toxicologically acceptable. Furthermore where does this leave the consumers and the authorities? This will be especially important in the future when GMP will cover many different countries in the internal market of the EEC. The definition of ADI 'not specified' (WHO 1987) does take all these points into account. Both JECFA and SCF are allocating ADIs 'not specified' and must, therefore, assume 'that the total daily intake of the substance, arising from its use at the levels necessary to achieve the desired effect (as a food additive) and from its acceptable background in food, does not represent a hazard to health'. I know all the arguments against setting a numerical value and have abided by them in both JECFA and SCF. I feel, however, that there ought to be some relation between the highest level which can safely be fed to animals without adverse effects plus other toxicological information on the substance and the maximum level of the additive to which we
Downloaded by [McGill University Library] at 04:09 04 February 2015
Safety evaluation of food additives
127
recommend consumers might be exposed. It also seems somewhat artificial to have the borderline between a numerical value and a 'not-specified' ADI determined by a general rule that you should not feed the animals more than 5% in the diet. If it is below this figure you can fix a numerical ADI, if it is above, the ADI will have to be 'not-specified'. I think the experience which will be obtained when evaluating novel-food and novel-food ingredients will help us to devise a better term than ADI 'not specified'. Most of my discussion concerning ADI 'not-specified' will also be applicable to the term acceptable. JECFA has used this term for solvents (FAO/WHO 1986) and a very specific use of a hydrogen peroxide preservation system for milk (FAO/WHO 1990). In these cases there is a close connection with use conditions (or such a connection can be made legally). The SCF has used the term 'acceptable' for some anti-oxidants which are also vitamins, and for which the committee did not want to set ADIs (CEC 1990b). I cannot say that I was particularly happy with that decision, but I am part of it. I am also part of the SCF decision to use the term 'acceptable' for certain sweeteners, e.g. the different polyols. Some practical examples of ADIs follow, where in most cases nominal figures have been set. I have chosen to use some examples from recent work of the SCF, partly because I am most familiar with them and also because they raise some of the issues I wish to discuss. At the conclusion of the paper I shall try to highlight some of the issues. Table 1 shows some of the ADIs for some colouring matters of a very different nature. On Allura red AC there is a wealth of studies. A NEL of approximately 700 (695) mg/kg body weight in a rat long-term study justified the ADI of 7 mg/kg body weight per day. It may be of interest to note that JECFA in 1980 (FAO/WHO 1980) established a temporary ADI of 7 mg based on the same data. The temporary ADI came from unsolved statistical problems with a mouse study. When these problems were resolved, JECFA made the ADI a full ADI with same numerical value (FAO/WHO 1981). For SCF (CEC 1983, SCF 1985) some questions arose regarding the metabolism studies and the possibility of forming the carcinogen /7-cresidine in vivo. The SCF was reassured by the non-carcinogenicity in long-term studies. It was accepted that the metabolism studies were less than totally adequate but the SCF stuck to the pronouncement that it would be desirable if these details were cleared
Table 1. Recent evaluations by the SCF of colouring matters, 1987-88. Compound Allura red AC Caramels Plain caramel Caustic sulphite caramel Ammonia caramel Ammonia sulphite caramel Canthaxanthin Erythrosine
Evaluation ADI: 0-7 mg/kgb.wt. Acceptable Temporarily acceptable (in alcoholic beverages) Temporary ADI: 0-200 mg/kgb.wt. ADI: 0-200 mg/kgb.wt. Temporary ADI: 0-0-05 mg/kgb.wt. ADI: 0-0-1 mg/kgb.wt.
Data from: CEC 1983, SCF 1985, 1987a, b.
Downloaded by [McGill University Library] at 04:09 04 February 2015
128
E. Poulsen
up. Some behavioural effects seen at higher doses in rat studies did not preclude the ADI at a lower level of exposure. One can see from these and the following examples that the committees in question are flexible in their interpretation of available toxicological information, and do not stick to inflexible rules about the relation between effects, safety factors and numerical ADI values. The caramels (CEC 1983, SCF 1987a) have given the SCF much evaluation work, especially in regard to specifications and identity of products used in toxicological studies. In addition the lymphocyte-depressing effect found in rat studies of ammonia caramel has been complicating the evaluation. The industry has worked intensely on these issues and satisfactory caramel food additives are now available. The factor responsible for the depressing effects on rat lymphocytes, 2acetyl-4(5)-tetrahydroxybutyl imidazole, THI (in animals on a low-pyridoxine diet) can now be kept below a level where a potential effect in ordinary human consumers can be expected. This level of THI should be controlled by the specification of the food additive. To set the temporary ADI for ammonia caramel a 90-day study in the rat was included, and the safety factor was 100. The crucial factor was felt to be the control of THI to below 25 ppm, and in due course down to 10 ppm. For canthaxanthin something very unusual happened as the ADI was lowered considerably (from 25 mg to 0-05 mg) (SCF 1987a). This figure was based on human experience with a non-food use (to obtain artificial browning). This much higher intake had shown effects in humans. Consumption of canthaxanthin can produce crystalline deposits in the retina, and 30 mg per person appeared to be a NEL. If a safety factor of 10 was applied to this level a T-ADI of 0-05 mg/kg body weight could be established. Erythrosine has been evaluated several times because its effect on the thyroid gland and the original ADI was made temporary {and halved) in 1983 (CEC 1983). When new studies from the US became available the SCF discussed the significance of these long-term studies in the rat and short-term studies in man (SCF 1987b). The SCF considered that the oncogenic and possible carcinogenic effects of erythrosine were likely to be secondary to its effect on thyroid and pituitary function. An ADI of 0-1 mg/kg body weight was established based on the NEL for effects of erythrosine on thyroid and pituitary function in humans. On the antioxidants the evaluation is outlined in table 2. The two most controversial compounds were BHA and BHT. BHA has been shown to produce hyperplasia and/or tumours specifically in the forestomach of rats, mice and hamsters. Based on all the available studies in these species and in animals without a forestomach, genotoxicity studies, and on studies on the mechanism of the lesion in the forestomach, the SCF concluded that the production of rodent forestomach tumours was not a manifestation of genotoxicity. In addition to being an effect with a threshold, the preceding hyperplasia may not be of relevance for man. Further reassurance was—as mentioned—to be found in the results of studies in species without a forestomach. In setting the ADI the Committee took into account the NEL in a 90-day study for the production of hyperplasia in the rat forestomach. The SCF, therefore, established a temporary ADI 0-5 mg/kg body weight (CEC 1990b). For BHT the SCF reviewed many studies (CEC 1990b), including a long-term Danish study in rats (Olsen et al. 1986), which showed an increased incidence of hepatocellular adenomas and carcinomas at 100 and 250 mg/kg body weight, while
Safety evaluation of food additives
129
Table 2. Recent evaluations by the SCF of antioxidants (December 1987).
Downloaded by [McGill University Library] at 04:09 04 February 2015
Compound L-Ascorbic Acid Na L-ascorbic acid Ca L-ascorbic acid Ascorbyl palmitate Tocopherol extr. a-tocopherol 7-tocopherol 5-tocopherol Gallates Propyl gallate Octyl gallate Dodecyl gallate BHA (butylated hydroxyanisole) BHT (butylated hydroxytoluene)
Evaluation
Acceptable Acceptable Acceptable
Group ADI: 0-0-5 mg/kgb.wt. Temporary ADI: 0-0-5 mg/kgb.wt. ADI: 0-0-05 mg/kgb.wt.
Data from CEC 1990b.
25 mg/kg appeared to be a NEL. The tumours appeared very late, and the survival of the control animals was markedly poorer than those fed BHT. Previous ordinary long-term studies in rats and mice had been negative. A later Japanese mouse study (Inai etal. 1988), however, had results similar to the Danish rat study. The SCF's view that the recent Danish study suggests a threshold for carcinogenesis (related to promotion?) was reinforced by the results obtained from mutagenicity studies, with an overall lack of evidence of genotoxic effect in in vivo systems. (IARC has considered the evidence as 'limited evidence' for carcinogenicity in experimental animals) (IARC 1986).) In some experiments BHT showed adverse effects on the thyroid. In haematological studies some, but not all, species tested show haemorrhaging effects and/or a reduction in the prothrombin index. Taking all these effects (especially haemorrhaging) into account the NEL was evaluated to be 5 mg/kg body weight and the SCF established an ADI of 0-0-05 mg/kg body weight. This value is at variance with JECFA (FAO/WHO 1987). This committee based its evaluation on the NEL in a recent reproduction study in the rat: 25 mg/kg body weight. (This happens to be identical to the NEL in the long-term rat (Danish) study.) A temporary ADI of 0-125 mg/kg body weight (SF 200) was the JECFA evaluation. In view of the carcinogenic response in the (prolonged) rat and mouse studies I think that a higher SF on the 25 mg/kg NEL might have been more appropriate, and would have avoided the concentration on the haemorrhagic effect, which anyway seems to be reversible. On the other antioxidants in table 2 I have commented elsewhere on the compounds with vitamin effect which were evaluated by the SCF to be 'acceptable' (CEC 1990b). For the gallates the SCF maintained the usual group evaluation based on the study in propyl gallate (CEC 1990b). In its 30th report (FAO/WHO 1987) JECFA departed from their previous evaluation and maintained only an ADI for propyl gallate. The final group of additives I shall comment on are the sweeteners; some of them are outlined in table 3. The SCF have considered both 'bulk' sweeteners and
130
E. Poulsen
Downloaded by [McGill University Library] at 04:09 04 February 2015
Table 3. Recent evaluations by the SCF of sweeteners (1985 and 1987) I. Compound
Evaluation
Isomaltt Lactitolt Maltitolt (and maltitol based products) Mannitolf Sorbitolf Xylitolf
Acceptable Acceptable Acceptable Acceptable Acceptable Acceptable
tLaxation may be observed at high intakes. Consumption of the order of 20 g/person per day of polyols is unlikely to cause undesirable laxative symptoms. Data from SCF 1987a, CEC 1990a.
intense sweeteners (CEC 1985, 1990a). The bulk sweeteners of the polyol types are dealt with in table 3. The laxative effect was especially considered. This was also done in response to comments received by the Commission. The SCF has included a footnote on lactitol and isomalt in the 1987 report: 'Laxation may be observed at high doses. Consumption of the order of 20 g/person/day of polyols is unlikely to cause undesirable laxative symptoms. The level for individual polyols ingested singly is higher in many cases'. In the report it is accepted that the laxation is caused by osmotic pressure and not by a gastroenteric sickness. The SCF feels that the term 'osmotic diarrhoea' describes the effect most accurately. The ADIs for the intense sweeteners are shown in table 4. I shall not comment on acesulfame K, neohesperidine dihydrochalcone and thaumatin, but I shall comment on the other compounds. The ADI for aspartame is not so much under discussion as the fact that the amino acids which it yields on hydrolysis (aspartate and phenylalanine) and most especially the latter, are neuroactive compounds. The ADI has had to be crosschecked with the effect of phenylalanine on blood (and brain) levels, especially in persons who are heterozygotic in relation to phenylketonuria inheritance. Some other effects on the CNS in humans have been postulated in people exceeding the ADI based on animal studies. I think the ADI has withstood the test, but it seems that the safety margin may not be as large as is often the case for food Table 4. Recent evaluations by the SCF of sweeteners (1985 and 1987) II. Compound Aspartame DKP Neohesperidine Dihydrochalcone Thaumatin Cyclamate Saccharin (Na, K and Ca salts) Acesulfame K
Evaluation ADI: 0-40 mg/kgb.wt. ADI: 0-7-5 mg/kgb.wt. ADI: 0-5 mg/kgb.wt. Acceptable Temporary ADI: 0-11 mg/kgb.wt. (as cyclamic acid) Temporary ADI: 0-2-5 mg/kgb.wt. ADI: 0-9 mg/kgb.wt.
Data from SCF 1987a, CEC 1990a.
Downloaded by [McGill University Library] at 04:09 04 February 2015
Safety evaluation of food additives
131
additives, when the ADI is based on animal studies. The real test is the effects seen in people after large-scale release of the substance. For saccharin and cyclamate I have already made some comments. For saccharin I think it has been justified to base the TADI on the NEL in a long-term study with specific carcinogenicity as the end-point. Furthermore I think the importance of the cation (AM.Ca.K) needs to be cleared up, and then we may eventually see a full ADI for saccharin. For cyclamates I think the carcinogenicity issue is nearly settled, though there is still a problem about promotion and potential co-carcinogenicity (NRC/NAS 1985). In general I think it is accepted that no carcinogenic risk for man arises from the use of cyclamate as a food additive. The present ADI established by JECFA (FAO/WHO 1982) and temporarily for SCF (CEC 1985) is based on the NEL in relation to testicular atrophy in the rat. It is calculated to cover the formation of the active compound cyclohexylamine in people who are converters of cyclamate to cyclohexylamine. This compound is also of interest because it is an indirect-acting sympathomimetic agent similar to tyramine, but many times less potent. In the USA this issue is still being discussed in re-establishing cyclamate as an additive, while I tend to agree with Bopp et al. (1988) who say that hypertension does not appear to develop in animals given cyclohexylamine chronically, or in animals and humans ingesting high does of cyclamate. Conclusions What can we extract from my examples? (1) Maybe the terms 'ADI not specified' and 'acceptable' should be discussed further. There should, in my view, be a more firm numerical basis than reference to GMP. The toxicological experts who establish the ADIs will not always have the information available on variations in use levels (intakes) which are covered by the term GMP. I think the use levels proposed should be spelled out for the toxicologists to compare with toxicological data available. Then much later misunderstanding could be avoided. It would not be a problem to evaluate the data from, e.g., feeding studies with levels at 5-10% of the compound in the diet in welldesigned tests, and these data should be combined with appropriate biochemical data (absorption, biotransformation, excretion and kinetics). The experienced toxicologist would then be able to judge when an appropriate and reasonable (low) safety factor would suffice to reach a toxicologically determined level. (2) The conventional approach with a safety factor of 100 applied to a NEL should as a general rule be maintained, as it has worked well. However, the effects observed in the level above the NEL may be trivial: slight depression in weight or weight gain, borderline inhibition of a non-essential enzyme activity, slight increase in the relative organ weight in one organ system without any changes in the histological picture of the tissues, etc. In such cases a lowering of the safety factor should be considered (25-50). If, on the other hand, the effect is a serious, even irreversible toxic effect at the effect level above the NEL, then the safety factor for a full ADI should be increased. (3) ADIs can also be based on a combination of animal and human data. In such cases the human data can be used to check the ADI based on animal data (aspartame) or it can be used to establish a NEL. The safety factor can then be
Downloaded by [McGill University Library] at 04:09 04 February 2015
132
E. Poulsen
decreased, but only if the human data—which will normally be shortterm—indicate that the effect is seen at lower level than the animal data (ammonia caramel). In cases where effects in humans are observed at low (very low) levels the ADI should be based on those levels (canthaxanthin). (4) ADIs can be based on NEL in animal carcinogenicity studies provided the material is clearly a non-genotoxic carcinogen. The effect ought normally to be species-specific to the particular species in question. Effects in other species should be absent, and/or well-conducted epidemiological studies in man should not indicate carcinogenicity (saccharin). For such compounds; and compounds with promotor activity, ADIs can be established, but probably using higher safety factors. There should be discussions on whether ADIs should be established for compounds showing low-potency carcinogenicity (after very high doses) in more than one animal species. This would mean the application of much higher safety factors (1000-5000 or more). It would probably not be acceptable (and practicable) for direct food additives, but for indirect additives it might be appropriate. In such cases the term 'tolerable daily intake' could be more appropriate. (5) The effects seen and used in the evaluation may tend to be more subtle biochemical changes, and not the more traditional parameters such as histological changes, neurotoxicity or teratological effects. In these cases application of lower safety factors to establish ADIs might be appropriate. If the animal species used has been shown to handle a compound qualitatively (biotransformation) and quantitatively (kinetics) like man, a lower safety factor may also be appropriate. (6) It has been suggested that the dose of the food additive which in humans gives a steady-state blood level might be useful in setting the numerical value for the ADI, and in these cases with a lower safety factor than 100, e.g. 5-10. It is assumed that the usual data packet from animal studies does not indicate otherwise. JECFA has used this approach in 1986 (FAO/WHO 1987) when setting a provisional tolerable weekly intake (PTWI) for lead in infants and children. It was concluded that an intake of 3-4 fig of lead per kg body weight would not increase the blood level of lead; with a safety factor of 6-8 the PTWI was set at 25 /ig/kg body weight. I hope that I have given some input to discussion aimed at arriving at improved ways of establishing still more relevant ADIs for food additives in the future. References BOPP, B. A., SONDERS, C , and KESTERSON, J. W., 1988, Toxicological aspects of cyclamate and
cyclohexylamine. Critical Reviews of Toxicology, 16, 213-289. CEC 1980, Reports of the Scientific Committee for Food, 10th Series (Luxembourg: CEC). CEC 1983, Reports of the Scientific Committee for Food, 14th Series (Luxembourg: CEC). CEC 1986, Reports of the Scientific Committee for Food, 16th Series (Luxembourg: CEC). CEC 1990a, Reports of the Scientific Committee for Food, 21st Series (Luxembourg: CEC). CEC 1990b, Reports of the Scientific Committee for Food, 22nd Series (Luxembourg: CEC). FAO/WHO Joint Expert Committee on Food Additives, 1980, Report 24, WHO Technical Report Series 653 (Geneva: WHO). FAO/WHO Joint Expert Committee on Food Additives, 1981, Report 25, WHO Technical Report Series 669 (Geneva: WHO). FAO/WHO Joint Expert Committee on Food Additives, 1982, Report 26, WHO Technical Report Series 683 (Geneva: WHO). FAO/WHO Joint Expert Committee on Food Additives, 1984, Report 28, WHO Technical Report Series 710 (Geneva: WHO).
Safety evaluation of food additives
133
FAO/WHO Joint Expert Committee on Food Additives, 1986, Report 29, WHO Technical Report Series 733 (Geneva: WHO). FAO/WHO Joint Expert Committee on Food Additives, 1987, Report 30, WHO Technical Report Series 751 (Geneva: WHO). FAO/WHO Joint Expert Committee on Food Additives, 1989, Report 33, WHO Technical Report Series 776 (Geneva: WHO). FAO/WHO Joint Expert Committee on Food Additives, 1990, Report 35, Meeting 29 May-7 June 1989 (Geneva: WHO) (In press). IARC, 1986, IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 40, (Lyon: International Agency for Research on Cancer.). IRAC, 1987, IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Supplement 7 (Lyon: International Agency for Research on Cancer).
Downloaded by [McGill University Library] at 04:09 04 February 2015
INAI, K., KOBUKE, T., NAMBU, S., TAKEMOTO, T., Kou, E., NISHINA, H., FUJIHARA, M., YONEHARA, S., SUEHIRO, S., TSUYA, T., HORIUCHI, K., and TOKUOKA, S., 1988, Hepatocellular
tumorigenicity of butulated hydroxytoluene administered orally to B6C3F1 mice. Japanese Journal of Cancer Research (GANN), 79, 49-58. NRC/NAS, 1985, Evaluation of Cyclamate for Carcinogenicity (Washington, DC: National Academy Press). OLSEN, P., MEYER, O., BILLE, N., and WÜRTZEN, G., 1986, Carcinogenicity study on butylated
hydroxytoluene (BHT) in Wistar Rats exposed in utero. Food Chemistry and Toxicology, 24, 1-12. SCF, 1985, Minutes of the Meeting No. 52, Brussels. SCF, 1987a, Minutes of the Meeting No. 59, Brussels. SCF, 1987b, Minutes of the Meeting No. 60, Brussels. WHO (World Health Organization) 1987, Principles for the Safety Assessment of Food Additives and Contaminants in Food. Environmental Health Criteria No 70 (Geneva: WHO).
Food Additives & Contaminants Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tfac19
Safety evaluation of substances consumed as technical ingredients (food additives) E. Poulsen
a
a
National Food Agency of Denmark , M⊘rkh⊘j Bygade 19, S⊘dborg, DK‐2860, Denmark Published online: 10 Jan 2009.
To cite this article: E. Poulsen (1991) Safety evaluation of substances consumed as technical ingredients (food additives), Food Additives & Contaminants, 8:2, 125-133, DOI: 10.1080/02652039109373963 To link to this article: http://dx.doi.org/10.1080/02652039109373963
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sublicensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
FOOD ADDITIVES AND CONTAMINANTS, 1991, VOL. 8, NO. 2, 125-134
Symposium Paper Safety evaluation of substances consumed as technical ingredients (food additives) E. POULSEN National Food Agency of Denmark, Mørkhøj Bygade 19, DK-2860, Sødborg, Denmark
Downloaded by [McGill University Library] at 04:09 04 February 2015
(Received 10 July 1990; revised 19 October 1990) The different types of acceptable daily intakes (ADIs) are described as used by the FAO/WHO Expert Committee on Food Additives (JECFA) and the EEC Scientific Committee for Food (SCF). The allocation is discussed of a full ADI or a temporary ADI, and examples are given for the establishment (or withdrawal) of these ADIs. The flavours cinnamyl anthranillate and the solvent 2-nitropropane (both withdrawn), the sweeteners cyclamate and saccharin and the antioxidant BHA (all three changed) but not abolished. For BHA and saccharin the ADI was retained by both committees in spite of some evidence of carcinogenicity to experimental animals. ADI—'not specified' is specially discussed and it is recommended that numerical ADIs are used whenever possible. With an ADI—'not specified' it should be stated which use (and intake) levels are toxicologically acceptable. Some compounds evaluated by the two committees are discussed, e.g. the colours: Allura red AC, erythrosine, canthaxanthin and the caramels; three anti-oxidants: BHA, BHT and the gallates; the sweeteners: polyols, aspartame, saccharin and cyclamates. Four recommendations are made: (1) a numerical basis be given for the levels allocated an ADI—'not specified' or 'acceptable'; (2) lowering of the conventional safety factor be considered when the effects found are trivial—higher safety factors be considered when the toxic effects are serious or even irreversible; (3) ADIs should, whenever possible, be based on a combination of human and animal data; (4) ADIs might be allocated to compounds indicating animal carcinogenicity, if the compound is non-genotoxic, the mechanism clearly secondary and/or species-specific. Keywords: food additives, safety evaluation
Types of ADI A full ADI will be established when the required toxicological information is available on a food additive. The necessary information is discussed in recent guidelines and reviews (CEC 1980, WHO 1987). The required information varies according to the nature of the compound and the biological effects found in the studies. There have been cases where, for example, JECFA has found the information of such a nature as to recommend not just no ADI, but that 'the compound should not be used' (as a food additive or solvent). Previously established ADIs have been withdrawn or made temporary (and often lowered). The next ADI, the temporary ADI, indicates that the information is inadequate, or that new information raises doubt, which can only be resolved after further study. Such study is necessary, and will generally be forthcoming. In most cases where the ADI is changed to temporary, the compound will continue in use as an approved additive, maybe with restrictions in use levels and/or number of allowed food groups. Often these changes in the ADI come about 0265-203X/90 $3.00 © 1991 Taylor & Francis Ltd.
Downloaded by [McGill University Library] at 04:09 04 February 2015
126
E. Poulsen
because results of new animal studies give indications of carcinogenicity of the additive. Some illustrative examples are the sweeteners cyclamates and saccharin, the antioxidant BHA, the flavouring material cinnamyl anthranilate and the solvent 2nitropropane. Cyclamates, cinnamyl anthranilate and 2-nitropropane have had their acceptance withdrawn (FAO/WHO 1990), but for cyclamates an ADI has now been established by JECFA (FAO/WHO 1982) while SCF has established a temporary ADI (CEC 1985). For saccharin the ADI has been changed to a TADI by both committees, but not abolished (CEC 1985, FAO/WHO 1984). For BHA the ADI was also temporary from both committees (CEC 1990b, FAO/WHO 1987) but JECFA has very recently established a full ADI (FAO/WHO 1989). For both these compounds the IARC monographs evaluate that there is sufficient evidence for carcinogenicity in experimental animals, and the compounds are classified in IARC class 2B, possible carcinogens for humans (IARC 1987). I think it is rather remarkable that the two expert committees have maintained a temporary (or full) ADI for these two compounds. This means that they are approved food additives in most countries, although the appearance of these food additives on national and international lists of carcinogens will undoubtedly raise discussions in some countries. It is my judgement—at least for the SCF—that an animal carcinogen with clear genotoxicity will not be allocated an ADI, while an ADI might very well be established for a compound which is believed to be a non-genotoxic carcinogen or a promoter. It is also my judgement in these last cases that the figure for the ADI would be calculated from a no-effect level (NEL) by the use of a safety factor, maybe larger than the usual 100. After discussion of the types: No-ADI, temporary ADI and full ADI, this paper will discuss other types of ADI: (1) 'not-specified' ADI and (2) acceptable (for use as). I should make it clear that I am not particularly fond of these types of ADI. I think that, when applied, they leave too much to the manufacturer in a rather uncontrolled way. I know, of course, that most manufacturers are completely safety-conscious and knowledgeable about good manufacturing practice (GMP), but they will be without any indications from toxicologists on the level of use for which the compound is found to be toxicologically acceptable. Furthermore where does this leave the consumers and the authorities? This will be especially important in the future when GMP will cover many different countries in the internal market of the EEC. The definition of ADI 'not specified' (WHO 1987) does take all these points into account. Both JECFA and SCF are allocating ADIs 'not specified' and must, therefore, assume 'that the total daily intake of the substance, arising from its use at the levels necessary to achieve the desired effect (as a food additive) and from its acceptable background in food, does not represent a hazard to health'. I know all the arguments against setting a numerical value and have abided by them in both JECFA and SCF. I feel, however, that there ought to be some relation between the highest level which can safely be fed to animals without adverse effects plus other toxicological information on the substance and the maximum level of the additive to which we
Downloaded by [McGill University Library] at 04:09 04 February 2015
Safety evaluation of food additives
127
recommend consumers might be exposed. It also seems somewhat artificial to have the borderline between a numerical value and a 'not-specified' ADI determined by a general rule that you should not feed the animals more than 5% in the diet. If it is below this figure you can fix a numerical ADI, if it is above, the ADI will have to be 'not-specified'. I think the experience which will be obtained when evaluating novel-food and novel-food ingredients will help us to devise a better term than ADI 'not specified'. Most of my discussion concerning ADI 'not-specified' will also be applicable to the term acceptable. JECFA has used this term for solvents (FAO/WHO 1986) and a very specific use of a hydrogen peroxide preservation system for milk (FAO/WHO 1990). In these cases there is a close connection with use conditions (or such a connection can be made legally). The SCF has used the term 'acceptable' for some anti-oxidants which are also vitamins, and for which the committee did not want to set ADIs (CEC 1990b). I cannot say that I was particularly happy with that decision, but I am part of it. I am also part of the SCF decision to use the term 'acceptable' for certain sweeteners, e.g. the different polyols. Some practical examples of ADIs follow, where in most cases nominal figures have been set. I have chosen to use some examples from recent work of the SCF, partly because I am most familiar with them and also because they raise some of the issues I wish to discuss. At the conclusion of the paper I shall try to highlight some of the issues. Table 1 shows some of the ADIs for some colouring matters of a very different nature. On Allura red AC there is a wealth of studies. A NEL of approximately 700 (695) mg/kg body weight in a rat long-term study justified the ADI of 7 mg/kg body weight per day. It may be of interest to note that JECFA in 1980 (FAO/WHO 1980) established a temporary ADI of 7 mg based on the same data. The temporary ADI came from unsolved statistical problems with a mouse study. When these problems were resolved, JECFA made the ADI a full ADI with same numerical value (FAO/WHO 1981). For SCF (CEC 1983, SCF 1985) some questions arose regarding the metabolism studies and the possibility of forming the carcinogen /7-cresidine in vivo. The SCF was reassured by the non-carcinogenicity in long-term studies. It was accepted that the metabolism studies were less than totally adequate but the SCF stuck to the pronouncement that it would be desirable if these details were cleared
Table 1. Recent evaluations by the SCF of colouring matters, 1987-88. Compound Allura red AC Caramels Plain caramel Caustic sulphite caramel Ammonia caramel Ammonia sulphite caramel Canthaxanthin Erythrosine
Evaluation ADI: 0-7 mg/kgb.wt. Acceptable Temporarily acceptable (in alcoholic beverages) Temporary ADI: 0-200 mg/kgb.wt. ADI: 0-200 mg/kgb.wt. Temporary ADI: 0-0-05 mg/kgb.wt. ADI: 0-0-1 mg/kgb.wt.
Data from: CEC 1983, SCF 1985, 1987a, b.
Downloaded by [McGill University Library] at 04:09 04 February 2015
128
E. Poulsen
up. Some behavioural effects seen at higher doses in rat studies did not preclude the ADI at a lower level of exposure. One can see from these and the following examples that the committees in question are flexible in their interpretation of available toxicological information, and do not stick to inflexible rules about the relation between effects, safety factors and numerical ADI values. The caramels (CEC 1983, SCF 1987a) have given the SCF much evaluation work, especially in regard to specifications and identity of products used in toxicological studies. In addition the lymphocyte-depressing effect found in rat studies of ammonia caramel has been complicating the evaluation. The industry has worked intensely on these issues and satisfactory caramel food additives are now available. The factor responsible for the depressing effects on rat lymphocytes, 2acetyl-4(5)-tetrahydroxybutyl imidazole, THI (in animals on a low-pyridoxine diet) can now be kept below a level where a potential effect in ordinary human consumers can be expected. This level of THI should be controlled by the specification of the food additive. To set the temporary ADI for ammonia caramel a 90-day study in the rat was included, and the safety factor was 100. The crucial factor was felt to be the control of THI to below 25 ppm, and in due course down to 10 ppm. For canthaxanthin something very unusual happened as the ADI was lowered considerably (from 25 mg to 0-05 mg) (SCF 1987a). This figure was based on human experience with a non-food use (to obtain artificial browning). This much higher intake had shown effects in humans. Consumption of canthaxanthin can produce crystalline deposits in the retina, and 30 mg per person appeared to be a NEL. If a safety factor of 10 was applied to this level a T-ADI of 0-05 mg/kg body weight could be established. Erythrosine has been evaluated several times because its effect on the thyroid gland and the original ADI was made temporary {and halved) in 1983 (CEC 1983). When new studies from the US became available the SCF discussed the significance of these long-term studies in the rat and short-term studies in man (SCF 1987b). The SCF considered that the oncogenic and possible carcinogenic effects of erythrosine were likely to be secondary to its effect on thyroid and pituitary function. An ADI of 0-1 mg/kg body weight was established based on the NEL for effects of erythrosine on thyroid and pituitary function in humans. On the antioxidants the evaluation is outlined in table 2. The two most controversial compounds were BHA and BHT. BHA has been shown to produce hyperplasia and/or tumours specifically in the forestomach of rats, mice and hamsters. Based on all the available studies in these species and in animals without a forestomach, genotoxicity studies, and on studies on the mechanism of the lesion in the forestomach, the SCF concluded that the production of rodent forestomach tumours was not a manifestation of genotoxicity. In addition to being an effect with a threshold, the preceding hyperplasia may not be of relevance for man. Further reassurance was—as mentioned—to be found in the results of studies in species without a forestomach. In setting the ADI the Committee took into account the NEL in a 90-day study for the production of hyperplasia in the rat forestomach. The SCF, therefore, established a temporary ADI 0-5 mg/kg body weight (CEC 1990b). For BHT the SCF reviewed many studies (CEC 1990b), including a long-term Danish study in rats (Olsen et al. 1986), which showed an increased incidence of hepatocellular adenomas and carcinomas at 100 and 250 mg/kg body weight, while
Safety evaluation of food additives
129
Table 2. Recent evaluations by the SCF of antioxidants (December 1987).
Downloaded by [McGill University Library] at 04:09 04 February 2015
Compound L-Ascorbic Acid Na L-ascorbic acid Ca L-ascorbic acid Ascorbyl palmitate Tocopherol extr. a-tocopherol 7-tocopherol 5-tocopherol Gallates Propyl gallate Octyl gallate Dodecyl gallate BHA (butylated hydroxyanisole) BHT (butylated hydroxytoluene)
Evaluation
Acceptable Acceptable Acceptable
Group ADI: 0-0-5 mg/kgb.wt. Temporary ADI: 0-0-5 mg/kgb.wt. ADI: 0-0-05 mg/kgb.wt.
Data from CEC 1990b.
25 mg/kg appeared to be a NEL. The tumours appeared very late, and the survival of the control animals was markedly poorer than those fed BHT. Previous ordinary long-term studies in rats and mice had been negative. A later Japanese mouse study (Inai etal. 1988), however, had results similar to the Danish rat study. The SCF's view that the recent Danish study suggests a threshold for carcinogenesis (related to promotion?) was reinforced by the results obtained from mutagenicity studies, with an overall lack of evidence of genotoxic effect in in vivo systems. (IARC has considered the evidence as 'limited evidence' for carcinogenicity in experimental animals) (IARC 1986).) In some experiments BHT showed adverse effects on the thyroid. In haematological studies some, but not all, species tested show haemorrhaging effects and/or a reduction in the prothrombin index. Taking all these effects (especially haemorrhaging) into account the NEL was evaluated to be 5 mg/kg body weight and the SCF established an ADI of 0-0-05 mg/kg body weight. This value is at variance with JECFA (FAO/WHO 1987). This committee based its evaluation on the NEL in a recent reproduction study in the rat: 25 mg/kg body weight. (This happens to be identical to the NEL in the long-term rat (Danish) study.) A temporary ADI of 0-125 mg/kg body weight (SF 200) was the JECFA evaluation. In view of the carcinogenic response in the (prolonged) rat and mouse studies I think that a higher SF on the 25 mg/kg NEL might have been more appropriate, and would have avoided the concentration on the haemorrhagic effect, which anyway seems to be reversible. On the other antioxidants in table 2 I have commented elsewhere on the compounds with vitamin effect which were evaluated by the SCF to be 'acceptable' (CEC 1990b). For the gallates the SCF maintained the usual group evaluation based on the study in propyl gallate (CEC 1990b). In its 30th report (FAO/WHO 1987) JECFA departed from their previous evaluation and maintained only an ADI for propyl gallate. The final group of additives I shall comment on are the sweeteners; some of them are outlined in table 3. The SCF have considered both 'bulk' sweeteners and
130
E. Poulsen
Downloaded by [McGill University Library] at 04:09 04 February 2015
Table 3. Recent evaluations by the SCF of sweeteners (1985 and 1987) I. Compound
Evaluation
Isomaltt Lactitolt Maltitolt (and maltitol based products) Mannitolf Sorbitolf Xylitolf
Acceptable Acceptable Acceptable Acceptable Acceptable Acceptable
tLaxation may be observed at high intakes. Consumption of the order of 20 g/person per day of polyols is unlikely to cause undesirable laxative symptoms. Data from SCF 1987a, CEC 1990a.
intense sweeteners (CEC 1985, 1990a). The bulk sweeteners of the polyol types are dealt with in table 3. The laxative effect was especially considered. This was also done in response to comments received by the Commission. The SCF has included a footnote on lactitol and isomalt in the 1987 report: 'Laxation may be observed at high doses. Consumption of the order of 20 g/person/day of polyols is unlikely to cause undesirable laxative symptoms. The level for individual polyols ingested singly is higher in many cases'. In the report it is accepted that the laxation is caused by osmotic pressure and not by a gastroenteric sickness. The SCF feels that the term 'osmotic diarrhoea' describes the effect most accurately. The ADIs for the intense sweeteners are shown in table 4. I shall not comment on acesulfame K, neohesperidine dihydrochalcone and thaumatin, but I shall comment on the other compounds. The ADI for aspartame is not so much under discussion as the fact that the amino acids which it yields on hydrolysis (aspartate and phenylalanine) and most especially the latter, are neuroactive compounds. The ADI has had to be crosschecked with the effect of phenylalanine on blood (and brain) levels, especially in persons who are heterozygotic in relation to phenylketonuria inheritance. Some other effects on the CNS in humans have been postulated in people exceeding the ADI based on animal studies. I think the ADI has withstood the test, but it seems that the safety margin may not be as large as is often the case for food Table 4. Recent evaluations by the SCF of sweeteners (1985 and 1987) II. Compound Aspartame DKP Neohesperidine Dihydrochalcone Thaumatin Cyclamate Saccharin (Na, K and Ca salts) Acesulfame K
Evaluation ADI: 0-40 mg/kgb.wt. ADI: 0-7-5 mg/kgb.wt. ADI: 0-5 mg/kgb.wt. Acceptable Temporary ADI: 0-11 mg/kgb.wt. (as cyclamic acid) Temporary ADI: 0-2-5 mg/kgb.wt. ADI: 0-9 mg/kgb.wt.
Data from SCF 1987a, CEC 1990a.
Downloaded by [McGill University Library] at 04:09 04 February 2015
Safety evaluation of food additives
131
additives, when the ADI is based on animal studies. The real test is the effects seen in people after large-scale release of the substance. For saccharin and cyclamate I have already made some comments. For saccharin I think it has been justified to base the TADI on the NEL in a long-term study with specific carcinogenicity as the end-point. Furthermore I think the importance of the cation (AM.Ca.K) needs to be cleared up, and then we may eventually see a full ADI for saccharin. For cyclamates I think the carcinogenicity issue is nearly settled, though there is still a problem about promotion and potential co-carcinogenicity (NRC/NAS 1985). In general I think it is accepted that no carcinogenic risk for man arises from the use of cyclamate as a food additive. The present ADI established by JECFA (FAO/WHO 1982) and temporarily for SCF (CEC 1985) is based on the NEL in relation to testicular atrophy in the rat. It is calculated to cover the formation of the active compound cyclohexylamine in people who are converters of cyclamate to cyclohexylamine. This compound is also of interest because it is an indirect-acting sympathomimetic agent similar to tyramine, but many times less potent. In the USA this issue is still being discussed in re-establishing cyclamate as an additive, while I tend to agree with Bopp et al. (1988) who say that hypertension does not appear to develop in animals given cyclohexylamine chronically, or in animals and humans ingesting high does of cyclamate. Conclusions What can we extract from my examples? (1) Maybe the terms 'ADI not specified' and 'acceptable' should be discussed further. There should, in my view, be a more firm numerical basis than reference to GMP. The toxicological experts who establish the ADIs will not always have the information available on variations in use levels (intakes) which are covered by the term GMP. I think the use levels proposed should be spelled out for the toxicologists to compare with toxicological data available. Then much later misunderstanding could be avoided. It would not be a problem to evaluate the data from, e.g., feeding studies with levels at 5-10% of the compound in the diet in welldesigned tests, and these data should be combined with appropriate biochemical data (absorption, biotransformation, excretion and kinetics). The experienced toxicologist would then be able to judge when an appropriate and reasonable (low) safety factor would suffice to reach a toxicologically determined level. (2) The conventional approach with a safety factor of 100 applied to a NEL should as a general rule be maintained, as it has worked well. However, the effects observed in the level above the NEL may be trivial: slight depression in weight or weight gain, borderline inhibition of a non-essential enzyme activity, slight increase in the relative organ weight in one organ system without any changes in the histological picture of the tissues, etc. In such cases a lowering of the safety factor should be considered (25-50). If, on the other hand, the effect is a serious, even irreversible toxic effect at the effect level above the NEL, then the safety factor for a full ADI should be increased. (3) ADIs can also be based on a combination of animal and human data. In such cases the human data can be used to check the ADI based on animal data (aspartame) or it can be used to establish a NEL. The safety factor can then be
Downloaded by [McGill University Library] at 04:09 04 February 2015
132
E. Poulsen
decreased, but only if the human data—which will normally be shortterm—indicate that the effect is seen at lower level than the animal data (ammonia caramel). In cases where effects in humans are observed at low (very low) levels the ADI should be based on those levels (canthaxanthin). (4) ADIs can be based on NEL in animal carcinogenicity studies provided the material is clearly a non-genotoxic carcinogen. The effect ought normally to be species-specific to the particular species in question. Effects in other species should be absent, and/or well-conducted epidemiological studies in man should not indicate carcinogenicity (saccharin). For such compounds; and compounds with promotor activity, ADIs can be established, but probably using higher safety factors. There should be discussions on whether ADIs should be established for compounds showing low-potency carcinogenicity (after very high doses) in more than one animal species. This would mean the application of much higher safety factors (1000-5000 or more). It would probably not be acceptable (and practicable) for direct food additives, but for indirect additives it might be appropriate. In such cases the term 'tolerable daily intake' could be more appropriate. (5) The effects seen and used in the evaluation may tend to be more subtle biochemical changes, and not the more traditional parameters such as histological changes, neurotoxicity or teratological effects. In these cases application of lower safety factors to establish ADIs might be appropriate. If the animal species used has been shown to handle a compound qualitatively (biotransformation) and quantitatively (kinetics) like man, a lower safety factor may also be appropriate. (6) It has been suggested that the dose of the food additive which in humans gives a steady-state blood level might be useful in setting the numerical value for the ADI, and in these cases with a lower safety factor than 100, e.g. 5-10. It is assumed that the usual data packet from animal studies does not indicate otherwise. JECFA has used this approach in 1986 (FAO/WHO 1987) when setting a provisional tolerable weekly intake (PTWI) for lead in infants and children. It was concluded that an intake of 3-4 fig of lead per kg body weight would not increase the blood level of lead; with a safety factor of 6-8 the PTWI was set at 25 /ig/kg body weight. I hope that I have given some input to discussion aimed at arriving at improved ways of establishing still more relevant ADIs for food additives in the future. References BOPP, B. A., SONDERS, C , and KESTERSON, J. W., 1988, Toxicological aspects of cyclamate and
cyclohexylamine. Critical Reviews of Toxicology, 16, 213-289. CEC 1980, Reports of the Scientific Committee for Food, 10th Series (Luxembourg: CEC). CEC 1983, Reports of the Scientific Committee for Food, 14th Series (Luxembourg: CEC). CEC 1986, Reports of the Scientific Committee for Food, 16th Series (Luxembourg: CEC). CEC 1990a, Reports of the Scientific Committee for Food, 21st Series (Luxembourg: CEC). CEC 1990b, Reports of the Scientific Committee for Food, 22nd Series (Luxembourg: CEC). FAO/WHO Joint Expert Committee on Food Additives, 1980, Report 24, WHO Technical Report Series 653 (Geneva: WHO). FAO/WHO Joint Expert Committee on Food Additives, 1981, Report 25, WHO Technical Report Series 669 (Geneva: WHO). FAO/WHO Joint Expert Committee on Food Additives, 1982, Report 26, WHO Technical Report Series 683 (Geneva: WHO). FAO/WHO Joint Expert Committee on Food Additives, 1984, Report 28, WHO Technical Report Series 710 (Geneva: WHO).
Safety evaluation of food additives
133
FAO/WHO Joint Expert Committee on Food Additives, 1986, Report 29, WHO Technical Report Series 733 (Geneva: WHO). FAO/WHO Joint Expert Committee on Food Additives, 1987, Report 30, WHO Technical Report Series 751 (Geneva: WHO). FAO/WHO Joint Expert Committee on Food Additives, 1989, Report 33, WHO Technical Report Series 776 (Geneva: WHO). FAO/WHO Joint Expert Committee on Food Additives, 1990, Report 35, Meeting 29 May-7 June 1989 (Geneva: WHO) (In press). IARC, 1986, IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 40, (Lyon: International Agency for Research on Cancer.). IRAC, 1987, IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Supplement 7 (Lyon: International Agency for Research on Cancer).
Downloaded by [McGill University Library] at 04:09 04 February 2015
INAI, K., KOBUKE, T., NAMBU, S., TAKEMOTO, T., Kou, E., NISHINA, H., FUJIHARA, M., YONEHARA, S., SUEHIRO, S., TSUYA, T., HORIUCHI, K., and TOKUOKA, S., 1988, Hepatocellular
tumorigenicity of butulated hydroxytoluene administered orally to B6C3F1 mice. Japanese Journal of Cancer Research (GANN), 79, 49-58. NRC/NAS, 1985, Evaluation of Cyclamate for Carcinogenicity (Washington, DC: National Academy Press). OLSEN, P., MEYER, O., BILLE, N., and WÜRTZEN, G., 1986, Carcinogenicity study on butylated
hydroxytoluene (BHT) in Wistar Rats exposed in utero. Food Chemistry and Toxicology, 24, 1-12. SCF, 1985, Minutes of the Meeting No. 52, Brussels. SCF, 1987a, Minutes of the Meeting No. 59, Brussels. SCF, 1987b, Minutes of the Meeting No. 60, Brussels. WHO (World Health Organization) 1987, Principles for the Safety Assessment of Food Additives and Contaminants in Food. Environmental Health Criteria No 70 (Geneva: WHO).
