Glutamate Safety in the Food Supply

The Safety Evaluation of Monosodium Glutamate1

Ronald Walker2 and John R. Lupien*
School of Biological Sciences, University of Surrey, Guildford GU2 5XH, Surrey, UK and
*Food and Nutrition Division, FAO, 00100 Roma, Italy

ABSTRACT L-Glutamic acid and its ammonium, calcium, monosodium and potassium salts were evaluated by
the Joint FAO/WHO Expert Committee on Food Additives (JECFA) in 1988. The Committee noted that intestinal and
hepatic metabolism results in elevation of levels in systemic circulation only after extremely high doses given by
gavage (30mg/kg body weight). Ingestion of monosodium glutamate (MSG) was not associated with elevated
levels in maternal milk, and glutamate did not readily pass the placental barrier. Human infants metabolized
glutamate similarly to adults. Conventional toxicity studies using dietary administration of MSG in several species
did not reveal any specific toxic or carcinogenic effects nor were there any adverse outcomes in reproduction and
teratology studies. Attention was paid to central nervous system lesions produced in several species after
parenteral administration of MSG or as a consequence of very high doses by gavage. Comparative studies

Downloaded from jn.nutrition.org by guest on October 13, 2016

indicated that the neonatal mouse was most sensitive to neuronal injury; older animals and other species (including
primates) were less so. Blood levels of glutamate associated with lesions of the hypothalamus in the neonatal
mouse were not approached in humans even after bolus doses of 10 g MSG in drinking water. Because human
studies failed to confirm an involvement of MSG in Chinese Restaurant Syndrome or other idiosyncratic
intolerance, the JECFA allocated an acceptable daily intake (ADI) not specified to glutamic acid and its salts. No
additional risk to infants was indicated. The Scientific Committee for Food (SCF) of the European Commission
reached a similar evaluation in 1991. The conclusions of a subsequent review by the Federation of American
Societies for Experimental Biology (FASEB) and the Federal Drug Administration (FDA) did not discount the
existence of a sensitive subpopulation but otherwise concurred with the safety evaluation of JECFA and the
SCF. J. Nutr. 130: 1049S1052S, 2000.

KEY WORDS: humans safety food safety food additives monosodium glutamate toxicity

The Joint FAO/WHO Expert Committee on Food Addi- lished; its primary aims were to protect the health of the
tives (JECFA)3 was established in the mid-1950s by the Food consumer and facilitate international trade in food. It was
and Agriculture Organization of the United Nations (FAO) decided that JECFA would provide expert advice to Codex on
and the World Health Organization (WHO) to assess the matters relating to food additives. Additionally, JECFA pro-
safety of chemical additives in food on an international basis. vides advice directly to FAO and WHO member states, and
Its brief has been broadened subsequently to include contam- requests for assessment may come directly from them.
inants and veterinary drug residues. In the early 1960s, the Members of JECFA are independent scientists, drawn
Codex Alimentarius Commission (CAC), an international mainly from government or academic research institutes, who
intergovernmental body that sets food standards, was estab- serve in their individual expert capacity and not as represen-
tatives of their governments or institutions. Members are
assisted by Temporary Advisers (WHO) or Consultants
1
Presented at the International Symposium on Glutamate, October 1214, (FAO), also appointed in their personal capacity. In relation
1998 at the Clinical Center for Rare Diseases Aldo e Cele Dacco, Mario Negri to food additives, the goals are to establish safe levels of intake
Institute for Pharmacological Research, Bergamo, Italy. The symposium was and to develop specifications for identity and purity.
sponsored jointly by the Baylor College of Medicine, the Center for Nutrition at the
University of Pittsburgh School of Medicine, the Monell Chemical Senses Center, Through mid-1998, 51 meetings of JECFA have been held.
the International Union of Food Science and Technology, and the Center for The reports are published in the WHO Technical Report
Human Nutrition; financial support was provided by the International Glutamate Series and the toxicological evaluations, which form the basis
Technical Committee. The proceedings of the symposium are published as a
supplement to The Journal of Nutrition. Editors for the symposium publication of the safety assessment, are published in the WHO Food
were John D. Fernstrom, the University of Pittsburgh School of Medicine, and Additives Series; specifications are published in the FAO Food
Silvio Garattini, the Mario Negri Institute for Pharmacological Research. and Nutrition Paper Series.
2
To whom correspondence should be addressed.
3
Abbreviations used: ADI, acceptable daily intake; CAC, Codex Alimentarius The safety evaluation of monosodium glutamate (MSG) by
Commission; CNS, central nervous system; FASEB, Federation of American JECFA was conducted along with the group of related com-
Societies for Experimental Biology; FDA, Food and Drug Administration; JECFA, pounds, i.e., L-glutamic acid and its ammonium, calcium,
Joint FAO/WHO Expert Committee on Food Additives; LD50, dose that is lethal to
50% of subjects; MSG, monosodium glutamate; SCF, Scientific Committee for monosodium and monopotassium salts. These substances were
Food. first evaluated at the fourteenth and seventeenth meetings in

0022-3166/00 $3.00 2000 American Society for Nutritional Sciences.

1049S
1050S SUPPLEMENT

1971 and 1974, respectively (FAO/WHO 1971 and1974). At clinical indices, mortality or general behavior. Reproduction
that time, an Acceptable Daily Intake (ADI) of 0 120 mg/kg and teratology studies using the oral route of administration
body weight was allocated, encompassing the L-glutamic acid have been uneventful even when the dams were fed glutamate
equivalents of the salts; this was considered additional to the at high doses, indicating that the fetus and suckling neonate
intake from all nonadditive dietary sources. In the absence of was not exposed to toxic levels from the maternal diet through
human infant data at that time, and in view of the observation transplacental transfer. This latter observation is in accord
that neonatal rodents appeared to be more sensitive than with reports that glutamate levels in fetal blood do not rise in
adults to the neurologic effects of high blood levels of gluta- parallel with maternal levels. For example, in rats, although
mate, it was stated that the ADI did not apply to infants 12 single oral doses of 8000 mg/kg given to pregnant females late
wk of age. A more recent and comprehensive safety evaluation in gestation caused plasma levels to rise from 100 to 1650
was conducted in 1987 (Joint FAO/WHO Expert Committee nmol/mL, no significant increases were observed in plasma
on Food Additives 1988); the basis of that evaluation will be levels of the fetuses. Similarly, in pregnant rhesus monkeys,
discussed below. The Scientific Committee for Food of the the infusion of 1 g MSG/h led to a 10- to 20-fold increase in
Commission of the European Communities (SCF) also re- maternal plasma levels but no changes in fetal plasma levels.
viewed the data in 1991 and reached conclusions similar to In rats and monkeys, oral ingestion of these large doses of
those of the JECFA (SCF 1991). Subsequently, the Federation MSG did not lead to detectable increases in glutamic acid
of American Societies for Experimental Biology (FASEB) levels in maternal milk.
conducted a review of reported adverse reactions to MSG and The toxicologic picture arising from conventional studies
reported in 1995 (FASEB 1995). This report and the response therefore seemed quite reassuring. Nevertheless, two other
of U.S. Food and Drug Administration (FDA) will also be major issues had to be addressed in relation to high intakes of
mentioned briefly. MSG, namely, 1) potential neurotoxicity, especially to the
infant, and 2) the putative role of MSG in Chinese Restau-

Downloaded from jn.nutrition.org by guest on October 13, 2016

The safety evaluation of the JECFA rant Syndrome (e.g., flushing, tightness of the chest or diffi-
culty in breathing ) after consumption of Chinese foods.
The JECFA reviewed the then available data on metabo- In relation to the neurotoxicity, the Committee considered
lism and pharmacokinetics of MSG, together with relevant reports of 59 studies conducted in mice (40), rats (12), ham-
experimental toxicologic data and results of studies in humans. sters, guinea pigs, chicks, ducks, rabbits, dogs and primates
The Committee noted that, after ingestion, transamination to (21). Lesions (focal necrosis) in the arcuate nucleus of the
alanine occurs during intestinal absorption, leading to the hypothalamus were observed reproducibly in rodents and
formation of -ketoglutarate; glutamine, -aminobutyrate and rabbits after parenteral administration of glutamate (intrave-
glutathione are other quantitatively minor but physiologically nously or subcutaneously) or after very high bolus doses by
important metabolites. Excess glutamate, after deamination, gavage. These neural lesions were observable within hours of
may be utilized in gluconeogenesis. The available data indi- administration. The mouse appeared to be the most sensitive
cated that, under normal conditions, mammals have the met- species, and there were significant differences with age and
abolic capacity to handle large oral doses although the more maturity; the neonate was particularly sensitive. Notably, most
readily available nature of free MSG compared with the slow of the studies in primates were negative with regard to hypo-
release during protein digestion must be borne in mind. As a thalamic lesions; these were reported in only 2 of 21 studies,
consequence of the ready metabolism, concentrations of glu- both conducted in the same laboratory (Olney and Sharpe
tamate in portal blood show only a small rise after adminis- 1969, Olney et al. 1972)
tration of MSG unless very large bolus doses are administered The oral gavage doses required to produce the lesions were
by gavage. Further metabolism occurs in the liver, and sys- on the order of 1000 mg/kg body weight as a bolus dose, and
temic blood levels rise only when such large bolus doses are in only one study were lesions seen after voluntary ingestion
given as to overwhelm this hepatic metabolism or if MSG is of MSG. In that case, weanling mice were deprived of food and
given by irrelevant parenteral routes. Gavage doses in excess of water overnight, then given solutions containing 5 or 10%
30 mg/kg body weight are required to produce detectable MSG as the sole drinking fluid. No such lesions were seen
elevations of blood levels, and the same dose of MSG admin- when MSG was given at 10% of the diet even though plasma
istered in food produces lower peak plasma levels than if glutamate levels were doubled, nor after administration at high
administered in aqueous solution. Foods rich in available car- concentrations in drinking water ad libitum.
bohydrate were most effective in blunting the peak plasma In relation to the question of the relevance of this neuro-
levels. In fact, only slight rises in plasma glutamate were toxicity for humans receiving MSG in the diet, biochemical
observed after a dose of 150 mg MSG/kg body weight in and human studies have been crucial. The transport rate of
human adults. Infants, including premature babies, could also glutamate from blood to brain in mature animals is much lower
metabolize similar doses given in infant formulae (Tung and than that for neutral or basic amino acids, and normal plasma
Tung 1980). levels of glutamate are nearly four times the Michaelis-Menten
The conventional toxicologic database available for review constant (Km) of the transport system to the brain i.e. the
by the JECFA was very extensive, including acute, subchronic concentration associated with half maximal velocity. This
and chronic toxicity studies in rats, mice and dogs, together implies that this transport system is virtually saturated under
with studies on reproductive toxicity and teratology. physiologic conditions. However, the blood-brain barrier may
Glutamate has a very low acute toxicity under normal be less effective in the neonatal mouse, which raises the
circumstances; the oral dose that is lethal to 50% of subjects possibility that this is the reason for the exquisite sensitivity of
(LD50) in rats and mice is 15,000 18,000 mg/kg body this model. However, the question of the comparability of the
weight, respectively. Subchronic and chronic toxicity studies mouse and the human infant remains an issue, in that the level
of up to 2 y duration in mice and rats, including a reproductive of brain development in the two species is quite dissimilar at
phase, did not reveal any specific adverse effects at dietary parturition.
levels of up to 4%. A 2-y study in dogs at dietary levels of 10% The threshold blood levels associated with neuronal dam-
also did not reveal any effects on weight gain, organ weights, age in the mouse (the most sensitive species) are 100 130
 SAFETY EVALUATION OF MSG 1051S

mol/dL in neonates rising to 380 mol/dL in weanlings and The Scientific Committee for Food of the Commission of
630 mol/dL in adult mice. In humans, plasma levels of this the European Communities
magnitude have not been recorded even after bolus doses of
150 mg/kg body weight (10 g for an adult). Additionally, the The SCF (1991) conducted a safety evaluation similar to
studies in infants previously mentioned have confirmed that that of the JECFA and reached the same conclusion, i.e., that
the human infant can metabolize glutamate as effectively as MSG could be allocated an ADI not specified, and this is the
adults. It is thus concluded that blood levels of glutamate current situation in the European Union.
aspartate do not rise significantly even after abuse doses of up
to 10 g, and infants are no more at risk than adults. Similarly, FASEB and the FDA
the comparisons of maternal and fetal blood levels after high
doses indicate that the fetus is not at greater risk. Intake levels Because of the FDAs concern over continuing reports of
associated with the use of MSG as a food additive and natural adverse reactions to MSG and other glutamate-containing
levels of glutamic acid in foods, therefore, do not raise toxi- ingredients, and in light of the expanding knowledge on the
cologic concerns even at high peak levels of intake because the role of glutamate in brain function, the FDA contracted with
mechanism of toxicity appears to be related to the peak plasma FASEB to conduct a review with the following objectives:
level achieved rather than the area under the curve. A puta- 1. To determine whether MSG and hydrolyzed protein
tive mechanism for the neuronal damage is that high levels of products, as used in the American food supply, contrib-
glutamate at the target site lead to continuous excitation of ute to the presentation of a complex of symptoms
the glutaminergic neurons, depleting ATP and leading to cell (initially described as Chinese Restaurant syndrome)
death. Such a situation is difficult to achieve with oral admin- after oral ingestion of levels up to or beyond 5 g per
istration in food. Furthermore, the JECFA noted that the oral eating occasion . . . and or the elicitation of other reac-
ED50 for production of hypothalamic lesions in the neonatal tions, including more serious adverse reactions . . . re-

Downloaded from jn.nutrition.org by guest on October 13, 2016

mouse is 500 mg MSG/kg body weight by gavage, whereas ported to occur following ingestion of 25100 mg per
the largest palatable dose for humans is 60 mg/kg body eating occasion.
weight with higher doses causing nausea; thus, voluntary in- 2. To determine whether MSG and hydrolyzed protein
gestion would not exceed this level. products . . . have the potential to contribute to brain
lesions in neonatal or adult nonhuman primates and
Idiosyncratic intolerance (Chinese Restaurant Syndrome) whether there is any risk to humans ingesting dietary
MSG.
With regard to the second issue, reports of the so-called 3. To assess whether hormones are released from the
Chinese Restaurant Syndrome were linked to the use of pituitary of nonhuman primates following ingestion of
MSG in Chinese cuisine and suggested that there may be MSG or hydrolyzed protein products and whether any
idiosyncratic intolerance in some individuals. Most of the comparable risk to humans ingesting these substances
reports of these subjective symptoms were anecdotal, although exists.
in some investigative studies, MSG was also claimed to pro- 4. To define the metabolic basis that might underlie any
voke these symptoms. However, more extensive studies in adverse reactions to MSG and hydrolyzed protein prod-
human volunteers were reviewed, and these failed to demon- ucts.
strate that MSG was the causal agent in provoking the full The FASEB report was submitted to the FDA in July 1995
range of symptoms. Properly conducted and controlled double- (FASEB 1995). In this report, the term MSG symptom com-
blind crossover studies have failed to establish a relationship plex is used instead of Chinese Restaurant Syndrome be-
between Chinese Restaurant Syndrome and ingestion of cause the latter was considered pejorative and characterized
MSG, even in individuals claiming to suffer from the syn- the symptoms as an acute, temporary and self-limiting com-
drome. Some food symptom surveys were considered techni- plex including the following: 1) a burning sensation of the
cally flawed because of inappropriate questionnaire design. back of the neck, forearms and chest; 2) facial pressure or
In its conclusion on this matter, the JECFA stated con- tightness; 3) chest pain; 4) headache; 5) nausea; 6) upper body
trolled double-blind crossover trials have failed to demonstrate tingling and weakness; 7) palpitation; 8) numbness in the back
an unequivocal relationship between Chinese Restaurant of the neck, arms and back; 9) bronchospasm (in asthmatics
Syndrome and consumption of MSG. MSG has not been only); and 10) drowsiness.
shown to provoke bronchoconstriction in asthmatics. In passing, it is interesting to note the term MSG symptom
complex was used when the terms of reference clearly in-
JECFA safety evaluation cluded protein hydrolysates and other natural sources of glu-
tamic acid.
The overall safety evaluation led the JECFA to conclude The report concluded that, although there was no scientif-
that the total dietary intake of glutamates arising from their ically verifiable evidence of adverse effects in most individuals
use at levels necessary to achieve the desired technological exposed to high levels of MSG, there is sufficient documen-
effect and from their acceptable background in food do not tation to indicate that there is a subgroup of presumably
represent a hazard to health. For that reason, the establish- healthy individuals that responds, generally within 1 h of
ment of an ADI expressed in numerical form was not deemed exposure, with manifestations of the MSG symptom complex
necessary and an ADI not specified was allocated to L- when exposed to an oral (bolus) dose of MSG of 3 g in the
glutamic acid and the monosodium, potassium, calcium and absence of food. Although the FDA appears to have accepted
ammonium salts. this conclusion of the existence of the MSG symptom complex
The JECFA also noted the evidence that it was not neces- (Hattan, 1996), it was pointed out that the key data relate to
sary to treat pregnant women and infants as special cases; single-dose challenges in capsules or simple solutions and are
however, they did retain the previously expressed position that limited in their ability to predict adverse reactions resulting
food additives, in general, should not be used in infant foods to from the use of MSG in food. This is an important caveat
be consumed before 12 wk of age. because available carbohydrate in foods appears to modulate
1052S SUPPLEMENT

the pharmacokinetics. The Hattan memorandum also indi- Finally, the FDA interpreted the findings of the FASEB
cates that the FDA did not consider the evidence regarding Report to be generally consistent with the safety assessments of
sensitivity of asthmatics to MSG compelling and questioned other authoritative organizations (presumably including the
the inclusion of bronchoconstriction in the MSG symptom JECFA and SCF) that have affirmed the safety of MSG at
complex in the absence of confirmatory data in a well-con- levels normally consumed by the general population, and
trolled study. The reasons are outlined and relate to limita- concurred with the conclusion that there is no evidence link-
tions in the key study (Allen et al. 1987), and a call was made ing current MSG food use to any serious, long-term medical
for further work in this area. problems in the general population.
The FASEB report concludes that there is no evidence to
support a role for dietary MSG or other forms of free glutamate LITERATURE CITED
in causing or exacerbating serious, long-term medical problems
Allen, D. H., Delohery, J. & Baker, G. (1987). Monosodium L-Glutamate-
resulting from degenerative nerve cell damage. The FDA induced asthma. J. Allergy Clin. Immunol. 80: 530 537.
accepted the conclusion that serious neurotoxicologic effects Carlson, H. E., Miglietta, J. T., Roginsky, M. S. & Stegink, L. D. (1989) Stim-
from MSG are limited to animals given very large doses by ulation of pituitary hormone secretion by neurotransmitter amino acids in
humans. Metabolism 38: 1179 1182.
parenteral, pharmacologic or other nondietary conditions of FAO/WHO (1971) Evaluation of food additives: specifications for the identity
use or administration. and purity of food additives and their toxicological evaluation; some extrac-
With regard to the potential disruption of the neuroendo- tion solvents and certain other substances; and a review of the technological
efficiency of some antimicrobial agents. 14th Report of the Joint FAO/WHO
crine axis, the FASEB Expert Panel gave particular consider- Expert Committee on Food Additives. FAO Nutrition Meetings Report Series
ation to the potential of dietary MSG to affect adversely the no. 48, WHO Technical Report Series no. 462.
structure and function of areas of the brain not protected by FAO/WHO (1974) Toxicological evaluation of certain food additives with a
the blood-brain barrier. The Panel focused their evaluation on review of general principles and of specifications. 17th Report of the Joint
FAO/WHO Expert Committee on Food Additives. FAO Nutrition Meetings
a study conducted by Carlson et al. (1989) of the stimulation

Downloaded from jn.nutrition.org by guest on October 13, 2016

Report Series no. 53, WHO Technical Report Series no. 539.
of pituitary hormone secretion by neurotransmitter amino FASEB (1995) Analysis of Adverse Reactions to Monosodium Glutamate
acids, which showed that a dose of 10 g glutamic acid in saline (MSG), Report. Life Sciences Research Office, Federation of American Soci-
eties for Experimental Biology, Washington, DC.
caused a twofold increase in peak serum concentrations of Fernstrom, J. D., Cameron, J. L., Fernstrom, M. H., Maconaha, Weltzin, T. E. &
prolactin and cortisol over baseline values. However, a subse- Kaye, W. H. (1996) Short term neuroendocrine effects of a large oral dose
quent study (Fernstrom et al. 1996) using equivalent (phar- of monosodium glutamate in fasting male subjects. J. Clin. Endocrinol.
Metab. 81: 184 191.
macologic) doses of MSG rather than the free acid failed to Hattan, G. G. (1996) Evaluation of the Federation of American Societies for
demonstrate any effect on plasma prolactin, luteinizing Experimental Biology (FASEB) July 1995 report: Analysis of Adverse Reac-
hormone, follicle-stimulating hormone, testosterone, growth tions to Monosodium Glutamate (MSG). Memorandum from Director of Health
Effects Evaluation to Dr. Lawrence Lin, HFS-206, dated August 30.
hormone, cortisol, thyroid-stimulating hormone or thyroid Joint FAO/WHO Expert Committee on Food Additives (1988) L-glutamic acid
hormones despite an 11-fold increase in plasma glutamate and its ammonium, calcium, monosodium and potassium salts. In: Toxico-
levels (cf. Carlson et al. 1989). The FDA concurred with the logical Evaluation of Certain Food Additives and Contaminants. New York,
Cambridge University Press, pp. 97161.
conclusion (from animal studies) that large doses of glutamate Olney, J. W. & Sharpe, L. G. (1969) Brain lesions in an infant rhesus monkey
can influence hormonal function but concluded further that it treated with monosodium glutamate. Science (Washington, DC) 166: 386
did not believe that there was evidence to indicate that MSG 388.
as ordinarily consumed in foods disrupts the neuroendocrine Olney, J. W., Sharpe, L. G. & Fergin, R. D. (1972) Glutamate-induced brain
damage in infant primates. J. Neuropathol. Exp. Neurol. 31: 464 488.
axis in humans. SCF (1991) Reports of the Scientific Committee for Food on a First Series of
It has been contended in some quarters that glutamate in Food Additives of Various Technological Functions, Commission of the Eu-
commercial products such as MSG or hydrolyzed protein, is ropean Communities, Reports of the Scientific Committee for Food, 25th
Series. Brussels, Belgium.
different in some way from naturally occurring glutamate. The Tung, T. C. & Tung, T. S. (1980) Serum free amino acid levels after oral
FASEB Panel rejected this contention. glutamate intake in infant and adult humans. Nutr. Rep. Int. 22: 431 443.