The Case Against Monosodium Glutamate—Why MSG is Dangerous (as are Other Sources of Free Glutamate) and How the Dangers Have Been Covered Up

A little over a week ago, I tried doing online searches to try to find out if monosodium glutamate (MSG) was safe or not for my blog post “Too Much of Any Amino Acid is Probably Bad for You, But Monosodium Glutamate Isn't Any Worse Than That.” I overwhelmingly found articles that were reassuring about the safety of MSG and explained why research results that sounded bad about MSG weren’t good evidence for problems from MSG. I have to thank Adrienne Samuels for her comment on my post pointing me to her work documenting not only the dangers from free amino acids such as the glutamate that MSG breaks into, but also the orchestrated, decades-long industry campaign to whitewash glutamate that was behind the difficult I had in getting a clear picture of the dangers of glutamate from online search. This work is summarized in her article “The Toxicity/Safety of Processed Free Glutamic Acid (MSG): A Study in the Suppression of Information,” which I will refer to in this post simply as “Adrienne’s paper” and quote from extensively. Let me say upfront that I find her paper very credible. The amount of traffic my blog gets from Google search gives me some hope that this can be a useful resource to others who, like me, try to search online to see whether MSG (and glutamate more generally) is safe.

One of my mistakes (which I believe was fostered by the industry campaign to whitewash glutamate) when I wrote “Too Much of Any Amino Acid is Probably Bad for You, But Monosodium Glutamate Isn't Any Worse Than That” was not understanding the key distinction between the biological effects of amino acids inside a protein, or during the controlled disassembly of that protein during ordinary protein digestion and the biological effects of amino acids on their own, called “free amino acids.” This is particularly true because certain amino acids, when on their own, are used by our bodies as signaling molecules, or are closely related chemically to signaling molecules. Glutamate (also called “glutamic acid”) in particular is used as a signaling molecule in the nervous system. So it can overstimulate the nervous system in a harmful way. As Adrienne puts it:

… MSG is simply processed free glutamic acid, or processed free glutamic acid combined with sodium (depending on how it is defined), and that glutamic acid is a neurotransmitter that causes nerves to fire; and when present in excess quantities, causes nerves to fire until they die.

Another free amino acid that acts similarly to glutamate is aspartic acid, the key chemical component of the nonsugar sweetener aspartame. Aspartame is on my bad list in “Which Nonsugar Sweeteners are OK? An Insulin-Index Perspective” for reasons related to its neurotoxicity.

Another mistake I made was to believe to readily the claim that glutamate doesn’t get past the blood-brain barrier. Adrienne points out that the blood-brain barrier is far from foolproof. Just as some people have leaky guts, some have leaky blood-brain barriers:

Today scientists know that MSG kills brain cells and causes neuroendocrine disorders in laboratory animals; and that it causes adverse reactions in humans. Scientists know that the blood brain barrier, once thought to prevent glutamate that comes from exogenous sources (eating included) from entering the brain, is not fully developed until puberty; is easily damaged by such conditions as high fever, a blow to the head, and the normal course of aging; and, in the area of the circumventricular organs, is leaky at best at any stage of life. Scientists know that a diverse number of disease conditions such as ALS, Alzheimer's disease, seizures, and stroke are associated with the glutamate cascade (Blaylock, 1994).

The blood-brain barrier is unavoidably leaky for the circumventricular organs because these parts of the brain need to be able to put hormones into the bloodstream. So it is especially easy for glutamate to get to the parts of the brain that are heavily involved in the hormonal regulation of the rest of the body.

The Dangers of Free Glutamate. With the “It’s a natural part of many proteins so it can’t be dangerous” and the “It can’t get past the blood-brain barrier” defenses of glutamate dismantled, what are the dangers of glutamate that evidence points to? Adrienne writes:

I found two sorts of studies: 1) those sponsored by the glutamate industry, which invariably concluded that MSG is safe, and 2) those done by independent neuroscientists and other researchers who found that MSG kills brain cells, causes neuroendocrine disorders, learning disabilities, and a variety of disorders such as tachycardia and seizures.

Some of the key research showing dangers of glutamate goes back to John Olney, but has been well-replicated. Adrienne writes:

… John W. Olney, M.D. reported that laboratory animals suffered brain lesions and neuroendocrine disorders after being exposed to monosodium glutamate (Olney, 1969). Scientists studying retinal degeneration in mice treated with free glutamic acid had noted that these mice became grotesquely obese. Olney, who speculated that the obesity might be a sign of damage to the hypothalamus (the area of the brain that regulates a number of endocrine functions, including weight control), found that infant laboratory animals given free glutamic acid suffered brain damage immediately, and assorted neuroendocrine disorders later in life. Pharmaceutical grade L-glutamic acid was often used to produce these disorders until neuroscientists observed that monosodium glutamate, an inexpensive food additive, could be substituted for laboratory-grade free glutamic acid in these studies and produce the same effects.

Given both the less-than-fully-formed blood-brain barrier of infants and the general sensitivity of infants to toxins, the use of MSG and other sources of free glutamate in baby food is particularly scary. Here is Adrienne’s account of that:

In the years that followed, neuroscientists replicated the work of Olney, and Olney spoke out repeatedly about the toxic potential of glutamic acid freed from protein prior to ingestion. In 1972, for example, Olney testified before the Senate Select Committee on Nutrition and Human Needs that ingestion of MSG places humans at risk, with the greatest risk being for the very young; and that a National Academy of Science panel organized to determine whether MSG ought to be banned from baby food had produced an "industry arranged whitewash" by a group of scientists with almost no experience in neuropathology (Gillette, 1972). In the early 1970s, manufacturers of baby food voluntarily removed the monosodium glutamate from their products, but replaced the monosodium glutamate with [other sources of free glutamate] such as autolyzed yeast and hydrolyzed vegetable protein.

Reading Adrienne’s paper and looking at the Wikipedia article “Monosodium glutamate,” I was surprised to see that food labeling laws don’t require that free glutamate (glutamic acid) content be labeled: it can be listed as “natural flavor.” Most other sources of free glutamate do need to be labeled: “… hydrolyzed vegetable proteinautolyzed yeasthydrolyzed yeastyeast extractsoy extracts, and protein isolate, which must be specifically labeled.” However, the fact that these are all sources of free glutamate may or may not be clear on the label. According to the FDA website, the only time the word “glutamate” is required if glutamate is added in the form of monosodium glutamate. But free glutamate is the issue, not exactly how it got into the food.

Over and above the harm from the left-handed glutamate that is a neurotransmitter, MSG is likely to contain right-handed glutamate and other contaminants. Here, from Adrienne:

… in addition to the L-glutamic acid found in unprocessed, unfermented, unadulterated free glutamic acid, processed free glutamic acid invariably contains D-glutamic acid and brings with it pyroglutamic acid and other contaminants--some of which, depending on procedures used for processing and the protein source, are carcinogenic.

The Industry Campaign to Whitewash Glutamate. Adrienne’s account of the industry campaign of misinformation about glutamate is easy to believe because it is such a distressingly familiar story in other areas. Gary Taubes details the campaign of misinformation about sugar in his book The Case Against Sugar. Keith Woodford details the campaign to discredit the evidence of the dangers of non-A2 milk in his book Devil in the Milk, which I distill in my post “Exorcising the Devil in the Milk.” Those who want to know how a campaign of disinformation works should read Adrienne’s paper in full. One key element is coopted individuals in and out of government referring to other coopted individuals as authorities. But those trying to present glutamate as safe have had a surprising strength not only in getting their story out, but also in being able to prevent or delay the publication of contrary views.

Adrienne very usefully explains strategies that were used to appear to be doing science proving glutamate is safe when proving nothing of the kind. Here is Adrienne’s rundown of those strategies (I omit #11, using student samples or samples of people who say they are sensitive to MSG, which I do not consider a serious flaw):

  1. Use variables and methods known to minimize or be irrelevant to identification of the toxic effects of glutamic acid; then conclude that glutamic acid never produces adverse effects. Studies have focused on the relationship between "objective" parameters such as blood pressure and body temperature and ingestion of MSG. Unless MSG sensitive people are studied, one can not legitimately draw conclusions about the relationship of the variables being studied (no matter how objective they are) to people who are sensitive to MSG. Often, these studies are used to allegedly "prove" that people who are not sensitive to MSG are not sensitive to MSG.

  2. Limit the recorded adverse effects to a few generally mild and transitory reactions occurring simultaneously, such as those first reported in 1968 by Kwok and dubbed "Chinese- restaurant syndrome" (CRS): "...numbness at the back of the neck, gradually radiating to both arms and the back, general weakness and palpitation." Industry researchers do not consider migraine headache, asthma, tachycardia, arrhythmia, depression, anxiety attacks or other obviously debilitating and/or life-threatening reactions reported since 1968.

  3. Make no attempt during a study to prevent subjects from ingesting food to which they might be allergic or sensitive.

  4. Record reactions as reactions to monosodium glutamate or placebo material only if they occur 2 hours or less following ingestion of test or placebo material, even though many symptoms are commonly expressed much later, and reactions may persist for much longer periods.

  5. Fail to report all data.

  6. Draw conclusions that do not follow from the results of the study. The IGTC researchers have concluded, for example, that because approximately one third of their subjects reacted adversely to placebos containing MSG and/or aspartame, they have "proved" that reactions to MSG-containing test material are not reactions to MSG.

  7. Use test material that will minimize the effect of any stated amount of glutamic acid test material in producing adverse reactions. One gram monosodium glutamate encased in capsules, and therefore guaranteeing slow release, will cause less effect than 1g monosodium glutamate sprinkled on food; and 1g monosodium glutamate modified with sucrose will cause less effect than otherwise because sucrose is known to slow monosodium glutamate uptake (Stegink, 1986).

  8. Continue subjects on medications that might block the effects of MSG.

  9. Using placebos to which MSG-sensitive people would react (placebos containing MSG, aspartame, carageenan or enzymes, for example), test potential subjects for sensitivity to those placebos, and eliminate any subjects who react to placebos. Researchers can be fairly certain that those who do not react to their reactive placebos will not react to monosodium glutamate test material.

  10. Advertise for, and presumably use, "well subjects" – people who had never experienced any of the symptoms with which reactions to MSG are associated. (If 50 per cent of the population were sensitive to MSG, but research design precluded inclusion of that 50 per cent who were sensitive, a study claiming to assess the number of people sensitive to MSG would be invalid.)

  11. Use placebos virtually guaranteed to produce as many reactions as might be produced following ingestion of the monosodium glutamate test material. Using toxic material in both test material and placebo, researchers argue that the reactions to MSG-containing test material are not reactions to MSG because subjects also react to placebos, which are assumed to be inert. However, the use of toxic material in placebos, particularly when it is identical or similar to the MSG in the test material, makes it virtually inevitable that there will be approximately as many reactions to placebos as there are reactions to MSG test material.

I find the problem of using placebos that might cause reactions a particularly interesting and serious flaw in these research designs.

Adrienne’s Appendix B is entitled “Excerpt from the prepared statement of John W. Olney, M.D. pertaining to adverse reactions to monosodium glutamate presented before the Federation of American Societies for Experimental Biology April, 1993.” This is a particularly telling account of bad research used to try to whitewash glutamate. Here is the beginning of that excerpt—John Olney’s words:

...When I reported in 1969-70 that glutamate destroys neurons in the hypothalamus when administered either subcutaneously or orally to immature mice (9-12), a U.S. Senate Nutrition Committee was investigating infant nutrition and asked me to comment on the fact that glutamate was being added to baby foods....Under pressure from the Senate committee, FDA arranged for a special ‘scientific' committee to evaluate the safety of glutamate for babies. The committee investigated the matter and concluded that glutamate was safe, but the committee was then investigated and most of its members were found to have close financial ties with the food industry....Of particular note, the committee Chairman, Lloyd J. Filer, was found to be receiving moneys from both the baby food industry and the glutamate industry while he chaired this committee.

Here is John Olney’s critique of the monkey research done to show that glutamate was safe, with my labels added in bold italics inside square brackets:

[Give the monkeys so much glutamate they vomit it up. Then neglect to mention that in the published paper.] … they tube-fed very large doses of glutamate to infant monkeys, which led me to suspect that their infant monkeys probably vomited (large doses of glutamate are known to induce vomiting in monkeys). This raised a crucial issue; if their infant monkeys vomited, they obviously lost dose control and this would render their data unreliable for establishing the safety of glutamate. I questioned Dr. Reynolds on this in public at a scientific meeting a few months before their Science paper appeared in print. In front of a large audience, she admitted that their monkeys vomited. When their Science paper appeared in print (33), I was surprised to read the following description: 'Each infant was maintained in an incubator with handling and cuddling at intervals for a 6 hour period. No unusual behavior was exhibited by the infants.' No mention was made at all of vomiting. Therefore, I wrote a letter to Science pointing out that by the author's own acknowledgment at a public meeting, these infants had vomited. The letter was accepted for publication in Science and was sent to Dr. Reynolds for her response. To my astonishment, in a letter signed by Reynolds, she responded with a denial that they had encountered problems with vomiting or with dose control. Therefore, I withdrew my letter and this exchange was never published.

[Knowingly autopsy the wrong part of the monkeys’ brains.] ....In the following year, I invited Reynolds et al. to send a member of their group to my laboratory to learn how to find glutamate damage in monkey brain. In May 1972, a member of their group (Dr. N. Lemkey-Johnston) did visit my laboratory and reviewed microscopic slides with me and she told me she was convinced that glutamate neuropathology was present in the hypothalamus of my monkeys. She also thanked me for pointing out specifically where to look in the hypothalamus to find these lesions. Two years later, when Reynolds et al. published their second paper (34), they stated that they had treated a few additional monkeys with glutamate and had serially sectioned the hypothalamus to provide definitive evidence of no damage. To my amazement, the illustration they showed was once again from the wrong region of the brain....

[Use monkeys who are likely to be much less susceptible to glutamate problems than humans.] ....In summary, the record shows that FDA for two decades has been assuring the public that glutamate is safe, based almost exclusively on certain industry-generated monkey data which appear upon close scrutiny to be seriously flawed, unreliable and spurious. However, even if these data were not flawed, unreliable and spurious, it is obvious from industry's own finding, shown in Fig. 1 above, that the pharmacokinetics of glutamate absorption and/or metabolism are so disparate between monkeys and man that monkeys, despite their phylogenetic closeness to humans, must be regarded as a singularly inappropriate animal model for evaluating oral glutamate safety. The same oral dose of glutamate that causes a dramatic increase in blood glutamate concentrations in humans, causes no increase at all in monkeys. Therefore, it is difficult to understand why so much money and effort was expended on oral glutamate monkey studies, unless the goal was to amass an unchallengeable mountain of negative evidence that could serve as basis for fostering the misleading impression, and fueling the spurious argument, that if monkeys are resistant to glutamate-induced brain damage, other primates, including humans, must be similarly resistant.

How to Avoid Free Glutamate

MSG and other sources of free glutamate are mainly found in Chinese food and processed food. The post “Is MSG Bad for You” on Real Mom Nutrition, which goes too easy on MSG overall, does have useful information on where MSG is found. It says MSG is especially prevalent in these foods (I am copying the list, with its original bullets):

  • Chinese food

  • canned soup

  • frozen foods

  • fast food

  • packaged snacks

It also gives the list of other sources of free glutamate that I mentioned once above (copying the list with its original bullets):

  • hydrolyzed vegetable protein

  • autolyzed yeast

  • hydrolyzed yeast

  • yeast extract

  • soy extracts

  • protein isolate

The dangers of free glutamate are another good example of the dangers of processed food, which I have warned of in this blog. See for example:

Of course, simply trying to eat low on the insulin index to avoid weight gain would steer one away from a large fraction of processed foods—though it wouldn’t be enough to steer one away from all processed foods. See “Forget Calorie Counting; It's the Insulin Index, Stupid.”

A Note on Umami. Adrienne is suspicious of the idea that umami or the savory taste is a fifth taste along with sweet, salty, bitter and sour. Although I agree with Adrienne’s view that Kikunae Ikeda’s declaration of umami as the fifth taste was intimately bound up with the discovery of monosodium glutamate, that doesn’t mean that it is wrong that this is a fifth taste. Nevertheless, just as using sugar or nonsugar sweeteners to overstimulate our sweet sensors is a bad idea, using MSG or other sources of free glutamates to overstimulate our umami sensors is a bad idea. Tickling the umami sensors in our taste buds with glutamate that arrives bound into proteins is likely to be safer, though overdoing protein is also a problem. Of the three “macronutrients”—carbohydrates, fat and protein, all of which are necessary for human life—I think protein is the most overrated.

What I wrote about umami in needs to be modified to take into account the difference between free amino acids and amino acids bound into proteins, but is otherwise still reasonable.

Umami is not all from glutamate. As Elizabeth Dunn writes in “From MSG Scare to MVP Status: How We Learned to Love Umami”:

Though the term is Japanese, umami is a global phenomenon. The same savory magic in pork and oysters runs through anchovies, seaweed and mushrooms, not to mention breast milk and amniotic fluid. In addition to glutamate, two other molecules, inosinate and guanylate, emit umami. Aging, caramelizing, drying and fermenting intensify it. Garum, the fermented fish sauce ancient Romans adored, teemed with umami, as do oyster sauce in China, miso in Japan, Worcestershire sauce in England and Maggi seasoning the world over. 

(You can learn a lot more about creating umami in the well-done Anime series “Food Wars,” which many of my family marathoned when we were together around Christmas. I recommend you do the version with subtitles. Few Anime shows have good English dubbing.)

Overall, I am willing to bet that it is safer to get the umami taste from some a reasonably balanced combination of glutamic acid, inosinate and guanylate than from an unusually large amount of glutamic acid. But glutamic acid in normal amounts seems to be part of the story of naturally occurring umami.

I worry that “Aging, caramelizing, drying and fermenting” might be ways of enhancing the amount of free amino acids. So intense umami from aging, caramelizing, drying and fermenting might have bad health effects. But umami from whole foods that are not processed in those ways nor more industrial ways should be safer.

For annotated links to other posts on diet and health, see: