Is MSG Actually Bad For You? | The Everyday Biologist

MSG might be the single most maligned ingredient in the grocery store. It’s been blamed for headaches, obesity, liver damage, kidney damage, infertility, and a made-up “syndrome” that was originally attached to one specific cuisine. At the same time, it’s been sitting on FDA’s generally-recognized-as-safe list since 1958, and it’s in half the savory food you eat without a second thought. So which is it?

The honest answer is that the picture depends almost entirely on two things nobody puts on the label: how much MSG a study actually used, and how it got into the animal or person eating it. Once you sort studies by those two variables, the controversy gets a lot less confusing.

What MSG actually is

MSG is the sodium salt of glutamic acid, a non-essential amino acid your body already makes and uses constantly for neurotransmission and protein synthesis [3]. When you eat it, it dissociates into sodium and free glutamate, the same molecule that’s naturally present, just bound to protein, in foods like tomatoes, mushrooms, seaweed, and aged cheeses like Parmesan [3,5]. The free form is absorbed faster than the protein-bound form, which is part of why added MSG behaves differently in the body than the glutamate sitting inside a steak.

There’s no chemical difference between glutamate that comes from MSG and glutamate your body produces on its own [3]. The dose and the delivery route are what change the story.

The dose and route problem

This is the part that gets buried in almost every scary headline about MSG. A systematic appraisal of 39 studies used to support MSG’s “ameliorative” and toxicity literature found that only two of them used a dietary route of exposure, meaning MSG mixed into food the way a person would actually eat it [2]. The rest used gavage (a tube directly into the stomach), or subcutaneous or intraperitoneal injection, often in newborn rodents during the first two weeks of life [2]. Doses ranged from 0.08 mg/kg/day all the way up to 16,000 mg/kg/day, and only 17 of the 39 studies used doses at or below the NOAEL used to set the regulatory acceptable daily intake [2].

Why does route matter this much? Glutamate is heavily metabolized in the small intestine before it ever reaches the bloodstream in a normal meal, and food itself slows absorption further through a “first pass effect” [2,4]. A bolus dose delivered straight into the stomach or under the skin skips that buffering system entirely, so blood glutamate spikes in a way that never happens when MSG is eaten with food [1]. Human studies confirm the same pattern: plasma glutamate barely moves when MSG is consumed with a meal, even at fairly high doses, but rises sharply when the same dose is taken in water on an empty stomach [1].

That neonatal detail matters too. A large share of the studies linking MSG to obesity and neurological damage rely on injecting newborn rodents during a developmental window when their blood-brain barrier is still immature, which lets glutamate accumulate in the hypothalamus in a way that wouldn’t happen from a person eating soup [2]. The lesions produced this way are real, but they’re a model for studying excitotoxicity mechanisms, not a preview of what happens when someone orders pho.

What the actual dietary safety data show

The cleanest evidence comes from two unpublished-until-recently GLP-compliant dietary studies in Sprague-Dawley rats, now reported in full [1]. In the 28-day study, rats ate diets delivering roughly 5,100 mg/kg bw/day (males) and 4,800 mg/kg bw/day (females), with no toxicologically significant effects observed [1]. In the 90-day study, the highest dose tested, about 3,170 to 3,620 mg/kg bw/day depending on sex, produced no adverse effects either, meaning the actual no-observed-adverse-effect level (NOAEL) is at least that high [1]. For context, average human MSG intake is estimated around 0.6 g/day in North America and Europe and 1.1 to 2.3 g/day in parts of East Asia, which works out to somewhere in the single digits of mg/kg bw/day for most adults [1,5]. The doses tested in these dietary rat studies are hundreds of times higher than that.

Regulatory bodies including FDA, JECFA, and the European Food Safety Authority have all concluded MSG is safe as a food additive, with JECFA assigning it an acceptable daily intake of “not specified,” the least restrictive category available [1]. EFSA did set a numerical ADI of 30 mg/kg bw/day, but it’s worth knowing where that number came from: it was based on a 1979 preliminary developmental neurotoxicity study that its own original author later re-examined and concluded showed no valid evidence of developmental neurotoxicity at all [1]. No other regulatory authority has used that study to set MSG policy.

So where do all the scary findings come from?

They’re real studies, just not studies that map onto how people eat. A broad narrative review pulling together 27 papers from 2014 to 2024 catalogued MSG-linked findings across nearly every organ system: liver injury, kidney injury, cardiovascular effects, reproductive toxicity, oxidative stress, embryotoxicity [3]. Read the fine print on those studies, though, and the pattern from the reliability analysis above holds up again and again. The reported doses run from 60 mg/kg up to 30,000 mg/kg bw/day, frequently administered by gavage or injection rather than diet, and rarely under standardized testing guidelines [1,3]. A separate independent review reached a similar conclusion: MSG toxicity findings in the literature tend to come from studies with methodological flaws, excessive doses, or non-dietary administration, and offer limited relevance for extrapolating to how humans actually consume it [1].

The maternal and fetal literature has the same shape. A review of MSG and offspring development found 14 eligible animal studies and zero eligible human studies, meaning everything we know about MSG during pregnancy comes from rodents, and most of those studies used subcutaneous injection or gavage at doses that don’t have a clean human equivalent [4]. The one primate study in the mix, monkeys given MSG orally during the last trimester, found no significant effects on pregnancy, birthweight, behavior, or hypothalamic development [4]. That’s thin evidence in both directions, which is really the honest takeaway: there just isn’t good human data on MSG and pregnancy yet, so caution is reasonable, but the animal findings that do exist aren’t strong grounds for alarm either.

What about headaches?

This is the one area where the evidence is mixed rather than just a route-and-dose artifact. A review of human trials found that when MSG was given with a meal, no study found a statistically significant increase in headache incidence [5]. When MSG was dissolved in soda or another beverage and given without food, some studies did find a significant increase, but usually at doses like 2.5 to 5 grams, well above the roughly 0.6 to 2 grams most people eat across an entire day [5]. There’s also a documented sensitive subgroup: a 1995 FASEB-commissioned review found that a small subset of people may respond to about 3 grams of MSG taken on an empty stomach within an hour of exposure [3]. If you suspect MSG triggers headaches for you specifically, a structured elimination diet is a more useful tool than trying to interpret a rat study [5].

The Chinese Restaurant Syndrome problem

It’s worth saying plainly that the origin story here is not great science. The term was coined in 1968 after a letter published in a medical journal, and it was, from the start, attached specifically to Chinese food rather than to MSG as an ingredient, even though MSG shows up constantly in western processed food, seasoning blends, and fast food [5]. Since then, multiple double-blind, placebo-controlled challenge studies have failed to reproduce the syndrome in a consistent, reliable way [5]. Reviewers studying this history have pointed out that the label persisted for decades because of preexisting bias against a specific cuisine, not because the science supported it, and that framing has continued to circulate in print as recently as 2023 [5].

Where this leaves things

MSG is not inert. At high enough acute doses delivered the wrong way, glutamate can produce real physiological effects, and there are legitimate open questions, especially around a possibly sensitive subgroup and the near-total absence of human pregnancy data. But the studies that mimic how people eat, dietary exposure at doses well above typical intake, fail to find harm, while the studies that generate alarming headlines tend to rely on injections, stomach tubes, or bolus doses that no one encounters at dinner. If you enjoy the umami hit MSG gives your cooking, the current dietary evidence doesn’t give you much reason to feel guilty about it.

Common Questions

Is MSG actually dangerous to eat?

The best-conducted studies, dietary rodent studies run under Good Laboratory Practice at doses far above what people actually eat, show no adverse effects. The studies that report toxicity almost always use gavage, injection, or acute high doses in newborn animals, which don't reflect how humans are exposed to MSG through food.

What is Chinese Restaurant Syndrome and is it real?

It's a cluster of symptoms first described in 1969 and blamed on MSG in Chinese food specifically. Double-blind, placebo-controlled trials have failed to reproduce the syndrome, and researchers now generally treat the label as more of a cultural artifact than a real medical phenomenon.

Can MSG give you a headache?

Maybe, in some people, at doses well above typical intake. Studies where MSG was mixed into a meal found no difference in headache rates compared to placebo. Studies where large doses were dissolved in soda, without food, sometimes did find a difference.

Is MSG safe during pregnancy?

There's little human data either way. The animal studies that exist mostly used injection or gavage rather than dietary exposure, at doses that don't map cleanly onto typical human consumption, so it's hard to draw a firm conclusion in either direction.

References