Tyramine

In brief

What tyramine is, and where it comes from

Take the amino acid tyrosine, one of the twenty building blocks of protein, and remove the carboxyl group, the acidic -COOH that makes it an amino acid at all. What remains is a benzene ring carrying a hydroxyl group on one side and a short two-carbon arm ending in an amino group on the other. That is tyramine: C₈H₁₁NO, 137.18 grams per mole, formally 4-(2-aminoethyl)phenol ¹. The reaction that makes it is a decarboxylation, and in food it is done by bacteria carrying an enzyme called tyrosine decarboxylase ².

Your own cells make small amounts of it too, which is why it is classed as a trace amine, a term for a group of compounds, tyramine among them with beta-phenylethylamine, tryptamine and octopamine, that are present in mammalian nervous tissue at levels far below those of dopamine, norepinephrine or serotonin ^1,10. The word trace refers to the amounts and not to any judgment about importance: the term was coined to reflect how low the tissue levels are in mammals, which is also why the same molecules are not called trace amines in insects ¹. In invertebrates the same molecules are not trace at all; they run the fight-or-flight machinery that adrenaline runs in us ^1,18.

Food is where tyramine becomes quantitatively serious. Three conditions favor it: protein, so that free tyrosine is released as that protein is broken down; bacteria carrying the decarboxylase; and time. Aging, fermenting and spoiling are three names for the same chemistry on different timetables ^1,14. This is why the risky list has always been a list of old or fermented foods, and why freshness, not food type, is the better rule of thumb ³¹.

The gene for the enzyme belongs to the strain, not the species, which is the part most often misunderstood. In a 2026 challenge study, marinated rainbow trout inoculated with decarboxylase-positive strains of Latilactobacillus fuchuensis reached up to 67 mg of tyramine per kilogram, while decarboxylase-negative strains of the identical species produced only small amounts in the same fillets under the same conditions ³². In that study the strain mattered only where the food let it: in unmarinated fillets tyramine was lower and did not differ between strain types at all ³². So the capacity belongs to the strain, and whether it is used belongs to the food. In cheese-derived model systems published in 2026, the clustering of tyramine, putrescine and cadaverine was driven mainly by which cheese broth the organisms were grown in rather than by their taxonomy or their phenotype in culture, with buffering capacity and the supply of peptides and precursor amino acids the decisive variables ³³. That is laboratory model work on cheese-derived media, not whole cheeses, and it says the label tells you less than either the strain or the food does.

How it acts: a passenger, not a key

The usual picture of a drug is a key in a lock: a molecule arrives at a receptor on the outside of a cell and switches it on. Tyramine does not work that way on blood vessels, and the difference is the whole of its pharmacology. It is carried into the sympathetic nerve ending by the norepinephrine transporter, the protein whose ordinary job is to recover released norepinephrine from the synapse and put it back. Inside, tyramine displaces the stored norepinephrine, which spills out and acts on the vessel wall ^4,6. The pressure that rises is driven by your own transmitter, released at the wrong time and in the wrong amount.

Two classical experiments fix this. In 1958 Burn and Rand published the experiments in reserpine-treated animals that defined the class; the record carries no abstract, so the detail behind the textbook version, that emptying the stores with reserpine abolishes the response to the indirectly acting amines while the response to injected norepinephrine survives, is taken here from the modern account of the mechanism ^5,7. Forty years later the transporter half was nailed down in guinea-pig heart: the release of norepinephrine by an indirectly acting amine was blocked by desipramine, a norepinephrine uptake inhibitor, with a half-maximal concentration near 1 micromolar. In that experiment the amine doing the releasing was dopamine, with tyramine run alongside it for comparison, so the number is a measurement on dopamine and not on tyramine. The same paper recorded something clinically useful: the tyramine effect faded on continued exposure as the stores ran down, while dopamine's did not, because dopamine can be built back up into norepinephrine ⁴.

The same indirect mechanism shows up far from blood vessels. In mouse nerve-muscle preparations published in 2026, tyramine reduced both spontaneous and nerve-evoked acetylcholine release, and the effect was abolished by the alpha-1 blocker doxazosin but not by an alpha-2 blocker, which is the signature of an action carried out by catecholamines that tyramine had released rather than by tyramine itself ³⁴. That study also detected endogenous tyramine and octopamine coming out of the isolated preparation, which is a reminder that the molecule is not purely dietary ³⁴. This is animal work and should be read as mechanism, not as a human finding.

None of this would matter if tyramine reached the circulation, because it does not. Monoamine oxidase in the lining of the small intestine, and then in the liver, deaminates it on the way through; the oxidative deamination of tyramine is specifically named as one of the few phase I reactions in the gut mucosa that is quantitatively important ³. The enzyme that does the work is the A form. Selectively inhibiting MAO-A, or inhibiting both forms, produces the hypertensive reaction; selectively inhibiting MAO-B does not ^6,7. The human evidence came in 1978. After doses of the selective MAO-B inhibitor deprenyl, the drug now called selegiline, large enough to inhibit the platelet enzyme completely, four healthy and six Parkinson's disease volunteers had no pressor reaction to oral tyramine in amounts considerably greater than a normal diet supplies, and normal human intestinal mucosa was shown to contain predominantly the deprenyl-insensitive A form, which presumably did the destroying ⁴⁰. Ten volunteers in an uncontrolled series is thin evidence, and it is still the human basis of the distinction, which itself dates from Johnston's 1968 report of the inhibitor now called clorgyline and of the two activities he named A and B ⁴¹. In anesthetized rats, the pressor response to intravenous tyramine was potentiated by the MAO-A inhibitor clorgyline and by moclobemide, and not by selegiline ⁴².

So the honest one-sentence mechanism is this: tyramine has no appreciable effect in a person with an intact gut enzyme, and in a person without one it produces a surge of their own norepinephrine. Everything written about tyramine in the clinic follows from the state of that enzyme ^7,35.

The receptor, found in 2001, and the trial it failed

For most of the twentieth century tyramine was a molecule with an effect and no address. Then in 2001 two laboratories, working independently, cloned a G protein-coupled receptor that responds to it. Borowsky and colleagues found fifteen related receptors by degenerate polymerase chain reaction, the same copying method described above run with primers that match an approximate rather than an exact sequence, from human and rodent tissue and showed that two of them bind or are activated by trace amines ⁸. Bunzow and colleagues characterized a rat receptor that raises cyclic AMP when exposed to p-tyramine, beta-phenylethylamine, tryptamine and octopamine, and found that amphetamine, MDMA, LSD, various ergolines and the 3-methylated metabolites of the catecholamines are agonists at it too ⁹. In 2009 the international pharmacology nomenclature committee settled the name: trace amine-associated receptor 1, TAAR1 ¹⁰.

Humans carry six functional receptors in this family, TAAR1, TAAR2, TAAR5, TAAR6, TAAR8 and TAAR9; TAAR3 is a working gene in some people and a pseudogene in others, and TAAR4 is a pseudogene in all of us while remaining functional in rodents ¹⁰. Some fish express over a hundred ¹. All of them except TAAR1 live in the olfactory epithelium, where they detect the amines of predators, spoiled food and pheromones ^1,10. TAAR1 is the odd one out: it is not an olfactory receptor, and it is found in the brain and the periphery ¹.

What it does there looks more like a brake than a signal. In mouse brain slices, tyramine acting through TAAR1 slows the firing of dopamine neurons in the ventral tegmental area by opening an inwardly rectifying potassium current, and both that effect and the effect of a selective antagonist vanish in mice lacking the receptor ³⁶. Trace amines acting through the receptor also inhibit uptake at the dopamine, norepinephrine and serotonin transporters and induce efflux, pushing transmitter back out through the transporter, the reverse of its normal direction, and that too is absent in knockout tissue ³⁷. The reviewers' summary is that TAAR1 acts as a rheostat on dopaminergic, glutamatergic and serotonergic transmission ¹. The receptor's structure was solved in 2023, nine complexes with endogenous amines, clinical drugs and synthetic compounds, revealing an amine-recognition pocket built around a conserved aspartate and a pair of toggle switches that open on activation ³⁸.

A brake on dopamine is exactly what an antipsychotic is supposed to be, and without blocking the D2 receptor it might avoid that drug class's movement and metabolic side effects. Ulotaront, a TAAR1 agonist that is also a partial serotonin 5-HT1A agonist, was taken into schizophrenia trials on that logic. Phase 2 looked excellent: a four-week placebo-controlled trial with an effect size of 0.45 on the Positive and Negative Syndrome Scale, fewer adverse events than placebo, 67 percent completion of a six-month open-label extension, and a breakthrough therapy designation from the United States regulator ^19,43. That account was written by the sponsor's own scientists, which is on the face of the paper ¹⁹.

Phase 3 did not reproduce it. A living systematic review pooling four studies and 1291 adults with acute schizophrenia found a standardized mean difference against placebo of 0.15, with a confidence interval from -0.05 to 0.34, meaning the result is compatible with no benefit at all; the related agonist ralmitaront was inferior to risperidone ⁴⁴. The reviewers singled out a large placebo response in the phase 3 trials as a problem for interpretation ^44,45. A 2025 dose-response meta-analysis of three randomized trials in 1144 people squeezed a small signal out of the 100 mg dose, a standardized mean difference of -0.23 with an interval from -0.43 to -0.02, while anxiety-related adverse events were significantly commoner than placebo at the 50 and 75 mg doses, with a risk ratio at 75 mg of 2.06 (95 percent confidence interval 1.11 to 3.80) ⁴⁶. The flattest statement in the literature comes from a 2025 pharmacology paper: the phase 3 trials showed ulotaront to have the same efficacy as placebo ²⁰. Not every review has caught up; a 2026 overview still lists TAAR1 agonists among drugs that substantially improve negative symptoms, which the pooled data do not support ⁴⁷.

The mechanism, oddly, held up better than the drug. An open-label phase 1b study in 22 patients found that fourteen days of added ulotaront lowered striatal dopamine synthesis capacity measured by positron emission tomography, and that the fall correlated with improvement in positive symptoms; mice lacking TAAR1 had higher synthesis capacity than wild-type ⁴⁸. The same group's mouse imaging shows TAAR1 agonists and D2 blockers producing distinct striatal signatures while reaching similar behavior ⁴⁹. So the receptor does regulate dopamine in people. Whether regulating it this way helps anyone is a separate question, and the trials answered it in the negative. Development has continued in other directions. A 2026 review of phase 2 and phase 3 industry-sponsored development for unipolar mood disorders lists ulotaront, under its compound number SEP-363856, among the agents in active United States trials of the previous three years ⁵⁰, and a 2025 review records its continuing development in major depressive disorder and generalized anxiety disorder ⁴³. It has also gone into metabolism, where small sponsor-run studies report favorable movement in insulin and weight measures that have not been confirmed in anything larger ^51,52. Only two TAAR1 agonists, ulotaront and ralmitaront, have reached clinical trials in schizophrenia to date ⁵³.

What the 2023 structures were for is already visible. In 2026 a group docked 65 million molecules against the active state of the receptor, tested 55 of the highest-ranked and found fourteen agonists, with potencies from mid-nanomolar to micromolar; after chemical optimization, three compounds with high brain exposure normalized the loss of prepulse inhibition caused by amphetamine in mice, a startle measure used as a laboratory stand-in for the sensory gating that fails in psychosis, and did so without catalepsy, the frozen immobility that dopamine-blocking drugs produce; one compound matched ulotaront's potency and two exceeded it, and the predicted binding poses were confirmed by cryo-electron microscopy, which images frozen protein complexes directly ⁵⁴. That is cell pharmacology and mouse behavior, not a treatment, and the lesson of ulotaront is that this is the easier half.

• The tyramine pressor testTyramine is given, by mouth or by vein, in rising doses until the systolic blood pressure rises by 30 mmHg; that dose is called TYR30, and the ratio of the drug-free TYR30 to the on-treatment TYR30 is the tyramine sensitivity factor ^28,89.Regulatory pharmacology: deciding whether a new drug that touches monoamine oxidase needs a tyramine diet attached to its labelReproducible enough to rank drugs cleanly. In one controlled comparison the geometric mean sensitivity factors were 1.52 for placebo, 2.15 and 2.74 for safinamide at 100 and 350 mg, 3.12 for selegiline and 9.98 for phenelzine ⁸⁹. A factor of 2 or more is treated as clinically meaningful ⁸⁸.It is done in healthy volunteers, usually fasting, which exaggerates absorption and says nothing about a sick older patient ⁸⁹. Its reading depends on uptake into the nerve ending, so a drug that blocks that uptake can mask the very thing being measured ⁹². Body mass index shifts the threshold, which is why within-person comparison is used ³⁹.
• Liquid chromatography of foodsAmines are separated by high-performance liquid chromatography and quantified, usually with diode-array or mass-spectrometric detection; this is how every number on a tyramine diet sheet was produced ^14,29.Food surveillance, and refining what patients are told to avoidGood for the sample in the tube. A 2025 survey of 125 retail ripened cheeses detected tyramine in 28.0 percent, spanning 5.62 to 646 mg/kg ¹⁴.It measures the sample, not the category. Values vary by orders of magnitude between samples of the same named food, so the tables in diet guides are imprecise by nature and should not be read as a property of the food ²².
• Detecting the decarboxylase gene in food bacteriaThe polymerase chain reaction, a laboratory method that copies one chosen stretch of DNA until there is enough of it to detect, is aimed at the tyrosine decarboxylase gene to find which strains in a food can make tyramine, since the capacity belongs to the strain and not to the species ^2,32.Choosing starter cultures and screening spoilage organismsDecisive when it works: in rainbow trout, decarboxylase-positive strains took marinated fillets to 67 mg/kg while decarboxylase-negative strains of the same species produced little ³².The widely used primers miss many producing strains outright and return false negatives, so a clean result does not mean a safe strain ².

What is actually in the food, measured

The Toronto assay series of 1989 to 1999 is where most of the figures on the diet sheets a patient is handed still come from, and this page leans on them ^{29,31,55,56,57}. They are not the whole measured base: the 2018 reappraisal draws on over 150 original papers and reviews hundreds of more recent estimations in cheeses alone ²¹. It is worth knowing how thin the base behind any single number is, and how wide the values are.

The 1989 series assayed over a hundred of the foods that had accumulated on restriction lists through rumor and case report. Only four categories emerged as requiring absolute restriction: aged cheeses, concentrated yeast extracts of the Marmite type, sauerkraut, and broad bean pods. Alcoholic drinks in moderation, including the Chianti that had become folklore, appeared safe ⁵⁵. The 1996 series extended this to 51 further foods and found the high values in chicken liver aged nine days, at 63.84 mg per 30 g, in sauerkraut at 7.75 mg per 250 g, and in soy sauce at 0.941 mg per milliliter; the air-dried sausage figure is printed in that abstract as 7.56 grams per 30 g, which is a unit error for milligrams on any reading consistent with the paper's own 6 mg threshold ²⁹. The 1999 series put two persistent worries to rest and raised one: no significant tyramine was found in any pizza, including double pepperoni and double cheese, while soy products varied enormously, with clinically significant levels in tofu stored for a week and a high value in one soy sauce ⁵⁷. The overripe banana of the old diet sheets was in the 1996 series, with banana peel, and did not come out among the foods holding 6 mg or more in a serving; that series concluded that only the high-value foods should continue to be absolutely restricted and that the rest were safe or safe in moderation ²⁹. The 1996 diet built on these assays treats any food it does not name as unrestricted ³¹.

A 2025 survey of 125 ripened cheeses from retail in Poland is the best recent measurement. Tyramine was detected in 28.0 percent of samples, spanning 5.62 to 646 milligrams per kilogram, with the high values accompanying Enterococcus faecium and Enterococcus faecalis in the same cheese. Histamine, the only one of these amines with a legal limit in European food law, exceeded 100 mg/kg in 11.2 percent of the cheeses ¹⁴. Note the shape of that result: most ripened cheeses had no measurable tyramine, and a minority had a great deal. That is the pattern that makes category-level advice almost useless and freshness-level advice sound.

The food industry has been working in the other direction, which is the main reason the old numbers no longer describe the modern shelf. Starter cultures and hygiene regulations have cut tyramine in most foods substantially, to the point where even matured cheeses are usually safe in ordinary portions ²¹. Current work is explicitly aimed at lowering it further: of three yeast strains screened in stout beer, one, Pichia kluyveri NW5, degraded 41.75 percent of the tyramine, with the total biogenic amine reduction after mixed fermentation reaching 48.81 percent; no one drank the result, as the study measured the beer and not any person ²⁵, and a Lactiplantibacillus plantarum strain reported a 79.94 percent fall in detectable tyramine under fermentation conditions, with the authors themselves warning that without controls matched for pH, the measure of acidity, this cannot be attributed to enzymatic degradation rather than to chemistry ²⁶. On the other side, the strains that do make it have now been identified in soy sauce and soybean paste, along with the uncomfortable finding that the standard genetic test for the decarboxylase misses many of them ².

The whole history, including the reversals

Tyramine's history is unusually clean in one respect and unusually messy in another. The cheese reaction was identified correctly, quickly, and has held up for sixty years. Almost everything built on top of it, the long restriction lists, the migraine theory, the therapeutic promise of its receptor, has had to be walked back.

The entries below are marked for what they were. The reversals are not failures of care; each was a reasonable reading of the evidence available, overturned by better evidence.

The cheese reaction

In October 1963 Barry Blackwell published a short report of hypertensive crises in patients taking monoamine oxidase inhibitors, with cheese among the suspected precipitants ¹¹. Three weeks earlier, Asatoor, Levi and Milne had put cheese through a chromatograph and named tyramine ¹². The two halves fitted: a drug that removes the enzyme destroying dietary tyramine, and a food containing a great deal of it. The confirmation then came food by food. Blackwell and Marley reproduced the cheese interaction in rats and cats in 1964 ⁶⁵; Blackwell and Mabbitt related the measured tyramine content of cheeses to the hypertensive crises in 1965, with tyrosine decarboxylase and streptococci among that paper's indexed terms ⁶⁶; and the effects of yeast extract after monoamine oxidase inhibition were reported twice more in the Lancet the same year ^67,68. Blackwell and Marley also published the fuller animal work, on cheese and separately on yeast extracts, and Blackwell with colleagues published the clinical survey of these interactions in 1967 ^30,69,70. By 1970 the whole subject had a monograph ⁷².

What the reaction looks like is a sudden, severe headache, often occipital and throbbing, with a blood pressure far above the patient's usual, sometimes with palpitation, sweating, pallor, nausea and vomiting. The danger is not the symptom but the pressure. The clinical survey is indexed with cerebral hemorrhage and subarachnoid hemorrhage among its outcomes, and a later review describes how fear of intracerebral hemorrhage and death drove clinicians away from the entire drug class ^30,77. Honest accounting requires a caveat here: the 1967 paper has no abstract in the database and no number of deaths can be quoted from it. The clearest modern statement is that cheese was the only food associated with documented fatalities resulting from hypertension ²¹.

The mechanism is already described above and is worth restating in the clinical frame. An irreversible inhibitor of MAO-A removes the gut wall's capacity to destroy tyramine ^3,7. Tyramine therefore reaches the circulation and the sympathetic nerve endings, is carried inside by the norepinephrine transporter, and dumps the stored norepinephrine ^4,6. The pressure that results is a surge of the patient's own transmitter. Because this depends on how much is stored and how fast it is released, the response is steep above a threshold and varies widely between people ²².

Food is not the only route in. The same interaction is described with indirectly acting sympathomimetic drugs, which is why decongestants carry a warning; a 2026 review of 219 publications lists hypertensive urgency or emergency reported with these drugs in combination with amphetamines, cocaine, khat and alcoholic drinks containing significant tyramine, while noting that such drinks are now uncommon ⁸³. One rare theoretical route has been tested once: a patient on transdermal selegiline who required intravenous feeding supplying tyrosine, the amino acid tyramine is made from, had no pressor reaction, which the authors correctly present as a single case rather than as evidence of safety ⁸⁴.

The treatment of the reaction itself has come in for a reassessment that cuts against intuition. A 2018 appraisal argues that the harms of treating hypertensive urgency hastily and aggressively, with the overshoot and hypoperfusion that can follow, have been underweighted relative to the harm of the pressure itself, and that the pressor response is usually self-limiting as the nerve endings empty ^4,21. That is one author's strongly argued position rather than a guideline, and it should be read as such.

The diet, and how far it has been relaxed

The original advice was to avoid a long list of foods. Over sixty years that list has been cut roughly three ways: by measuring the foods and finding most of them innocent, by changing which drugs are used, and by the food industry independently lowering tyramine in almost everything.

The measurement cuts came first. By 1996 the list had been reduced to aged cheese, aged or cured meats such as air-dried sausage, any potentially spoiled meat, poultry or fish, broad bean pods, concentrated yeast extract, sauerkraut, soy sauce and soybean condiments, and tap beer, with wine and bottled or canned beer considered safe in moderation ³¹. Broad bean pods are the one item on that list that is not about tyramine. The reaction reported with them is attributed to their levodopa content, first described in a 1964 Lancet report of headache and hypertension after broad beans in a patient taking an inhibitor, a record indexed under dihydroxyphenylalanine, the chemical name of levodopa, and carrying no abstract ⁸⁵; the current prescriber's guide discusses levodopa and L-tryptophan separately from the amines for the same reason ²². Pizza and moderate Chianti, both items of long-standing folklore, were cleared by assay ^55,57. The threshold the whole exercise is built on is around 6 mg of tyramine in a meal for a mild reaction, with 10 to 25 mg associated in case reports with severe headache and hemorrhage, and a working recommendation to stay below 5 mg ¹³. For contrast, the pressor-test literature reckons a high-tyramine meal at up to 40 mg ²⁸.

The drug cuts are the more important ones, because they remove the problem rather than manage it. Selegiline is selective for MAO-B at Parkinson's doses and loses that selectivity at the higher doses needed for depression, which is what the transdermal patch was designed around: deliver antidepressant concentrations to the brain while leaving MAO-A in the small intestine largely alone ¹⁵. The tests bear this out. After nine days at 6 mg per 24 hours the tyramine sensitivity factor was 1.85 plus or minus 0.10, similar to 10 mg a day of oral selegiline and more than twenty times lower than tranylcypromine ¹⁶; at the higher 12 mg/24 h patch, extended treatment produced a larger increase in that factor, which decreased significantly when the tyramine capsules were taken with a meal, so the exemption belongs to the 6 mg patch and not to the higher doses ¹⁶; a single 24-hour application left tyramine doses up to 200 mg without apparent increase in sensitivity ⁷⁸; and twelve healthy men who ate meals supplying roughly 400 mg of tyramine as aged cheeses after thirteen days on the patch showed no clinically significant cardiovascular change ⁷⁹. Twelve healthy men is a small sample run by the sponsor, which is the correct caveat to attach.

Reversibility is the other escape. Moclobemide binds MAO-A reversibly, so a rising tide of tyramine can compete it off the enzyme. In healthy subjects on 450 or 600 mg twice daily, intravenous tyramine sensitivity rose only three- to fourfold, and 50 mg of oral tyramine, far above a normal meal, did not raise systolic pressure by 30 mmHg; 100 mg was compatible with the lower dose but not the higher ⁸⁶. The drug review's conclusion is the practical one: dietary tyramine need not be avoided with moclobemide ¹⁷. That safety is not transferable to combinations. Giving moclobemide with the reversible MAO-B inhibitor lazabemide, blocking both isoforms at once, potentiated oral tyramine 13.5-fold, and only 51 plus or minus 20 mg was then needed to produce the response ⁸⁷.

The same test now gates new drugs that touch the enzyme incidentally. A single 600 mg dose of the antibiotic linezolid gave a tyramine sensitivity factor of 1.8, essentially the same as moclobemide's 2.1 in the same study ³⁹; its successor tedizolid gave a ratio of 1.33, below the factor of 2 regarded as clinically relevant ⁸⁸. Safinamide, a Parkinson's drug, gave 2.15 at 100 mg and 2.74 at 350 mg against placebo's 1.52 ⁸⁹, and, in a separate study at 300 mg once daily in volunteers preselected as tyramine-sensitive, no systolic rise of 30 mmHg occurred with tyramine up to 200 mg ⁹⁰. None of these carries a tyramine diet.

The last cut is the one nobody planned. The foods themselves have changed. A reappraisal of over 150 papers on food tyramine concluded that production methods and hygiene standards have reduced tyramine so far that even matured cheeses are usually safe in healthy-sized portions ²¹, and the current prescriber's guide, co-authored by the man who described the reaction in 1963, concludes that minimizing a small number of risky foods is all that is required and that many patients may hardly need to change their diet at all ²². That guide also makes the methodological point that matters most: tables of tyramine values are imprecise by nature, because levels vary widely within a single food category, so judgment rather than a list is what the prescriber needs ²². Against this background the class was, as of a 2013 review, prescribed far below what guidelines recommend, with the rate then still falling ⁹¹; no more recent prescribing count is cited here, and the 2026 review of these drugs in patients who use psychoactive substances notes only that they remain rarely prescribed in that group ⁸³.

The tyramine pressor test

The test is simple to describe. A person is screened to find the dose of tyramine, by mouth or intravenously, that raises their systolic blood pressure by at least 30 mmHg on consecutive measurements; this is called TYR30. They are then put on the drug to steady state and the ladder is climbed again. The ratio of the two thresholds is the tyramine sensitivity factor ^28,89. In the oral version, doses typically rise through 20, 40 and 80 milligrams, or higher in people found insensitive at screening ^28,90.

What it is for is regulatory rather than clinical: it decides whether a new drug that inhibits monoamine oxidase, deliberately or as a side activity, must carry a dietary warning. A ratio of 2 or more is generally treated as clinically relevant ⁸⁸. It is the reason tedizolid, safinamide and the transdermal patch do not carry a tyramine diet, and the reason their labels can say so with numbers behind them ^16,88,89.

Its strengths are real. It gives a single number on a common scale, so that drugs tested years apart in different places can be ranked, and within-person comparison removes a good deal of noise. Its weaknesses are equally real and worth naming. It is done in healthy volunteers, usually fasting, which exaggerates absorption relative to the same tyramine arriving in a meal, and an older patient with cardiovascular disease is not in the study ^79,89. Body mass index significantly affects the threshold dose, which is why the within-person design is used ³⁹. And the test reads out through the norepinephrine transporter, so a drug that blocks that transporter can flatten the result for reasons that have nothing to do with monoamine oxidase, which is the objection raised in a 2002 letter about drugs with dual serotonin and norepinephrine reuptake inhibition ⁹². That objection now has human numbers behind it. In a randomized, placebo-controlled study of 29 healthy men published in 2025, attenuation of the pressor response to intravenous tyramine was used as the read-out of norepinephrine reuptake inhibition: it separated from baseline at 40 mg of levomilnacipran, the lowest dose effective in depression, and only at 120 mg of duloxetine ⁹³. The test is therefore measuring transporter blockade as readily as it measures monoamine oxidase inhibition, which is exactly the confound the letter raised ⁹².

The deeper limitation is that the test measures the pressor response in a healthy adult who swallowed pure tyramine on an empty stomach. What a patient needs to know is what happens when they eat an unknown quantity of tyramine inside a food, with fat and protein slowing absorption, on a drug they have taken for months. The test is the best available proxy for that, and it is not that.

Tyramine and migraine: a claim that did not survive testing

In May 1967 the British Medical Journal carried a preliminary report titled Tyramine headache, indexed as a controlled clinical trial using lactose as the comparator ⁷¹. The author's surname is spelled Hannington in the database and Hanington in most citations since. The record contains no abstract and no numbers, so what can be said is that the claim was made, in that form, in that year. The timing is not an accident: the cheese reaction had just been explained, migraine patients had long blamed cheese and chocolate, and a mechanism appeared to be lying ready to hand.

The controlled challenges came next, and they did not confirm it. In a double-blind placebo-controlled study of 25 migraine patients, divided into those with migraine alone, those with migraine and epilepsy, and those who believed food precipitated their attacks, headache occurred in 12 of 50 patient sessions. Ten of those were in the diet-attributing group, and within that group headache followed tyramine alone in only two patients, with two patients getting headache after placebo alone and three after both capsules. The authors found no significant relationship between tyramine ingestion and the occurrence of headache, while reporting activation of the electroencephalogram, a recording of the brain's electrical activity made through the scalp, after tyramine in 11 of 15 patients in two of the three groups ⁷³. Two controlled trials were also run in children, though the database record carries no abstract and no numbers for them ⁷⁴.

A crossover study in eight patients with classical migraine went further and looked at mechanism. Between 50 and 75 percent had an attack within 24 hours of a pressor test, but there was no difference in the incidence of attacks between tyramine and dopamine injections, and neither disulfiram nor propranolol altered tyramine pressor sensitivity. The authors concluded that increased adrenergic activity was responsible and that it is unlikely tyramine plays any specific role in migraine except through the adrenergic system ⁷⁶. Eight patients is very small, but the conclusion is the one the rest of the literature keeps reaching. Separately, 27 healthy volunteers given 25 mg of tyramine, 25 mg of histamine or 5 mg of phenylethylamine in apple juice under randomized double-blind placebo-controlled conditions showed no statistically significant effect of tyramine or histamine; phenylethylamine produced headache, dizziness and discomfort in some of them ⁷⁵.

In 2003 the whole challenge literature was audited against explicit quality criteria. Of thirteen oral challenge studies, five with positive results and eight with negative, three were ineligible and six of the remaining ten were judged inconclusive. All four studies that survived as methodologically conclusive reported negative results, and two of them were specifically on tyramine and migraine. The reviewers' conclusion was that the scientific literature shows no relation between the oral ingestion of biogenic amines and food intolerance reactions, and that there is therefore no scientific basis for dietary recommendations about them ⁸⁰. That is as flat a negative as a careful review gets, and the literature it covered ends in 2001.

Where this leaves the dietary trigger idea in 2026 is worth stating precisely, because the honest answer is not simply that patients are wrong. A meta-analysis published in 2026 pooled patient-reported trigger rates from seven studies and found chocolate reported by 13.9 percent of patients (95 percent confidence interval 7.0 to 21.0), caffeine by 12.0 percent (6.0 to 19.0) and cheese by 8.0 percent (3.0 to 12.0), with heterogeneity above 95 percent and all data self-reported and observational. Its recommendation is individualized trigger identification using headache diaries rather than routine avoidance of these foods ⁸². So: a minority of people with migraine do report cheese as a trigger; controlled challenge with tyramine does not reproduce attacks; and the likeliest readings are that the trigger is something else in the food, that the association runs backward, with premonitory craving in the hours before an attack mistaken for its cause, or that it is a real effect in a small subgroup too small for these studies to find. None of those has been demonstrated, and saying so is better than choosing one.

Tyramine among the amines: histamine intolerance and fermented food

Tyramine rarely travels alone. The same decarboxylating bacteria working on the same breaking-down protein produce histamine from histidine, putrescine from ornithine, cadaverine from lysine and tryptamine from tryptophan, which is why food chemists measure them as a set ^1,14. In the Polish cheese survey, cadaverine was found in 28.0 percent of samples and putrescine in 20.8 percent, histamine exceeded 100 mg/kg in 11.2 percent, and tyramine was found in 28.0 percent, with different amines tracking different organisms: putrescine and cadaverine with Hafnia alvei, tyramine with the enterococci ¹⁴.

Histamine intolerance is the clinical idea built on this. The proposal is that some people degrade ingested histamine poorly, chiefly through low activity of diamine oxidase in the gut, so that amine-rich food provokes diarrhea, headache, flushing, hives, palpitation and low blood pressure that can be mistaken for allergy ⁹⁴. Tyramine appears on low-histamine diet sheets because it travels with histamine in the same foods, and sometimes with the further claim that it competes for the same disposal enzymes. That second claim should be handled carefully. The enzymatic-competition argument in the plant-food literature concerns putrescine, which competes with histamine at diamine oxidase, and spermidine, which the enzyme handles with lower affinity; tyramine is tracked in these reviews but is not the molecule that argument is about ⁹⁵.

How solid is the condition itself? Its own reviewers are candid. The founding review concedes that because the symptoms are so varied the existence of the condition has been underestimated and that further studies based on double-blind placebo-controlled provocation are needed ⁹⁴. A 2020 state-of-the-art review states that more scientific evidence is still required to define, diagnose and clinically manage it ⁹⁶. And the 2003 audit of the challenge literature, which covered histamine and red wine as well as tyramine, found no relation between oral ingestion of biogenic amines and intolerance reactions in the studies that were methodologically conclusive ⁸⁰. The reasonable position is that histamine poisoning from spoiled fish is beyond dispute, that interindividual susceptibility to ingested amines is plausible and has been proposed as the explanation for puzzling features of such outbreaks ⁹⁶, and that histamine intolerance as a chronic diagnosis in an individual patient currently rests on symptom response to diet rather than on a validated test.

Practically, the advice that follows is the same advice that the tyramine literature arrives at from the other direction, and it is about freshness rather than food type. Amines accumulate with storage and with spoilage, so the same named food is a different food at day one and day eight ^32,57,95. That is why a week-old tofu differs from a fresh one, and why a nine-day-old chicken liver differs from a fresh one by orders of magnitude ^29,57.

Tyrotoxicon: the poison that was not tyramine

Before tyramine there was tyrotoxicon, and the two are often run together in modern retellings in a way the historical record does not support. In the nineteenth century, ice cream and cheese caused outbreaks of foodborne illness often enough to need an explanation, and the explanations tried included toxic vanilla, galvanism in the ice cream freezers, and extreme indigestion. In the late 1880s Victor C. Vaughan's proposal won wide acceptance: the illness was caused by a ptomaine, a putrefaction alkaloid, which he named tyrotoxicon, from the Greek for cheese poison ⁶⁰. The term entered clinical use, and a note in the British Medical Journal in 1899 is headed simply Tyrotoxicon Poisoning ⁶¹.

The ptomaine theory did not survive bacteriology. The historical account of this episode traces the emergence, impact and decline of the whole idea, and makes the point that historians had neglected how much the ptomaine theory shaped both scientific thinking and public health in the late nineteenth century before it collapsed ⁶⁰. What actually caused most of those outbreaks, as bacteriology established, was infection, not a preformed alkaloid.

The reason the story belongs on this page is the shape of the mistake rather than any chemical identity. Tyrotoxicon and tyramine share a prefix, a food and a plausible chemistry: both would be products of protein breakdown in cheese. One turned out to be a real molecule with a real and lethal interaction; the other turned out to be an explanatory category that dissolved. Neither of the sources here identifies tyrotoxicon with tyramine, and the honest statement is that tyrotoxicon was never characterized well enough for the question to be answerable.

The other half of the story: tyramine in invertebrates

In us, tyramine is a trace. In an insect, a worm or a crustacean, it is a transmitter in its own right, and the direct biosynthetic precursor of octopamine, which is the invertebrate equivalent of norepinephrine. Both are decarboxylation products of tyrosine, both act through their own G protein-coupled receptors, and together they run the behaviors adrenaline runs in us: arousal, flight, the mobilization of energy ¹⁸. The ability to use trace amines as signals arose at least twice in evolution, with entirely distinct receptor families in invertebrates and vertebrates, which is why insect receptors and human TAAR1 are not the same protein doing the same job ¹.

The relationship between the two amines is not simply precursor and product. In isolated larval nervous systems of Drosophila published in 2026, bath-applied octopamine promoted forward crawling and suppressed backward crawling, while tyramine, conceived as its inhibitory counterpart, promoted collisions and overlap between motor programs, activity-dependent inhibition, and bouts of silence. Applying both revealed a state-dependent effect: a forward bias appeared only if the network was not already forward-dominant ²³. This is fictive locomotion in a dish, not behavior in an intact animal, but it shows two amines from one pathway pulling a network in different directions.

The reason this matters beyond entomology is insecticides. Because the tyramine and octopamine systems are, as far as is known, restricted to invertebrates, their receptors have long been regarded as attractive targets for compounds that would act on pests and not on us ¹⁸. The insect type 1 tyramine receptor has been characterized in structure, signaling and behavior for exactly this purpose, and its history contains a cautionary note of its own: it was initially classified as an octopamine receptor that tyramine could also activate, before being recognized as a true tyramine receptor ⁹⁷. Compounds with insecticidal properties acting at it are under development ⁹⁷.

One recent negative result is worth recording, because it is the kind that rarely gets published. Tyramine had been proposed as a virulence factor in European foulbrood, a bacterial disease of honeybee brood. Knocking out the putative tyrosine decarboxylase gene in a virulent strain of Melissococcus plutonius abolished tyramine production entirely and did not reduce the organism's virulence at all ²⁴. The gene does what it was thought to do; the molecule does not do what it was thought to do.

The same enzyme, working on a drug

The bacterial enzyme at the center of this page is not confined to the cheese cave. Tyrosine decarboxylase is carried by organisms living in the human gut, and in 2019 it was shown to act on a drug. Enterococcus faecalis in the human colon decarboxylates levodopa, the mainstay treatment for Parkinson's disease, turning it into dopamine before the drug can be absorbed and used; a second organism, Eggerthella lenta, then removes a hydroxyl group to make m-tyramine. A drug that inhibits the human decarboxylase, given alongside levodopa precisely to stop this happening in the body's own tissues, does not stop the bacterial version, and the investigators found a separate compound that did inhibit it in Parkinson's patients' gut communities and raised levodopa availability in mice ⁸¹.

Two things in that are worth keeping straight. The amine produced is m-tyramine, the meta isomer, a different molecule from the para-tyramine this page is about, although both are made by the same family of bacterial enzyme. And the clinical consequence described is not a pressor reaction but a loss of drug: the gut flora eating a Parkinson's medication before the patient gets it. The work was done in human gut communities in the laboratory and in mice, not as a treatment trial ⁸¹.

The reason it belongs here is what it says about the enzyme. The decarboxylation that makes tyramine in a cheese is not a quirk of dairy microbiology; it is a common bacterial capacity, carried by organisms that also live in us, acting on whatever aromatic amino acid or drug happens to be in reach. The microbiota both produce and degrade trace amines, which means the quantity of tyramine a person is exposed to is not simply the quantity on the plate ¹.

What was believed, and is not

Each of the following was standard teaching, and each was a reasonable reading of the evidence at the time. They are listed because the corrections are the most useful thing on this page.

Cheese and ice cream poisoning is caused by a ptomaine called tyrotoxicon. Widely accepted in the late 1880s and dissolved by bacteriology within a generation. Tyrotoxicon was never characterized well enough to be identified with any modern molecule, and it was not shown to be tyramine ^60,61.

Tyramine raises blood pressure by stimulating adrenergic receptors. It does not stimulate them at all in any important way. It is carried into the sympathetic nerve ending by the norepinephrine transporter and displaces the transmitter already stored there, which is why reserpine abolishes its effect and desipramine blocks it ^4,5,6.

A long list of foods must be avoided on these drugs. Measurement cut the list to a handful. Over a hundred contested foods were assayed in 1989 and only aged cheeses, concentrated yeast extracts, sauerkraut and broad bean pods required absolute restriction; Chianti and pizza, two items of long folklore, were cleared ^55,57.

All monoamine oxidase inhibitors require a tyramine diet. Only those that inhibit MAO-A, irreversibly, in the gut. Selective MAO-B inhibition does not produce the reaction ^6,7, transdermal selegiline at 6 mg/24 h bypasses the gut enzyme and needs no diet ^15,16,79, and reversible MAO-A inhibition is competed off by rising tyramine ^17,86.

Tyramine causes migraine. Proposed in 1967 ⁷¹ and not confirmed by controlled challenge. A double-blind study found no significant relation ⁷³, a crossover study concluded tyramine plays no specific role beyond the adrenergic system ⁷⁶, 25 mg produced no significant effect in healthy volunteers ⁷⁵, and the systematic audit found that every methodologically conclusive challenge study was negative ⁸⁰.

The tyramine data in the old tables describe today's food. They do not. Modern production methods and hygiene standards have cut tyramine levels substantially in most foods, including many cheeses ²¹. In 2025, 72 percent of 125 retail ripened cheeses had no detectable tyramine at all ¹⁴.

TAAR1 agonism is a validated treatment for schizophrenia. A phase 2 effect size of 0.45 became a pooled phase 3 standardized mean difference of 0.15 with a confidence interval crossing zero, and a 2025 pharmacology paper states plainly that the phase 3 trials showed the drug to have the same efficacy as placebo ^19,20,44. Some 2026 reviews have not registered this ⁴⁷.

Tyramine is a virulence factor in European foulbrood. Deleting the gene abolished production of the amine and left the organism exactly as virulent ²⁴.

What is still unknown

What tyramine does in a healthy person is genuinely unsettled. It is present, it is made and degraded by our own microbiota as well as by our tissues, and it has a receptor in the brain that responds to it ^1,36. Whether endogenous tyramine does anything physiologically important, as opposed to being a metabolic by-product with a receptor that evolved for other ligands, has not been established. The receptor's own discoverers framed this as the field's central question and it has not been answered ⁹⁸.

Why TAAR1 agonism worked in phase 2 and not in phase 3 is unresolved. The reviewers point to an unusually large placebo response in the later trials as a candidate explanation, which would mean the drug effect is real but small ^44,45. The alternative, that the phase 2 result was chance or optimism, is equally compatible with the data. A 2025 dose-response analysis found a small signal at 100 mg and recommended that higher doses be tested, so the dose question is open too ⁴⁶. Meanwhile the receptor demonstrably does regulate presynaptic dopamine in people ⁴⁸, which makes the clinical failure harder, not easier, to explain.

Whether the modern monoamine oxidase inhibitor diet could be relaxed further is a live question with no trial behind it. The argument that even matured cheeses are now usually safe in ordinary portions rests on assay data and inference rather than on anyone being fed cheese on phenelzine in a controlled study ^21,22. No one is likely to run that study. The practical consequence is that the current advice sits on an uneasy base: measurements that are solid and a translation to the individual patient that is not.

Who reacts and who does not is unexplained. The pressor response varies widely between people for reasons beyond body mass index, which accounts for some of it ^22,39. Whether the variation is in gut enzyme activity, in transporter function, in sympathetic stores or in vascular responsiveness is not known, and there is no test that identifies a sensitive patient in advance.

Whether a chronic sensitivity to dietary amines exists as a clinical entity remains open. The reviewers of histamine intolerance say so themselves, calling for the double-blind placebo-controlled provocation studies that would settle it ^94,96, and the challenge literature that exists is negative in its conclusive studies ⁸⁰. Something explains why some people reliably react to fermented food, and the amine hypothesis has not been demonstrated.

The food-safety question is narrower and more tractable. There is no legal limit on tyramine in European Union food law, where histamine is the only biogenic amine that carries one, under Regulation 2073/2005 ¹⁴; this page makes no claim about other jurisdictions. A quarter or more of ripened cheeses carry measurable tyramine and a few carry a great deal ¹⁴. Whether that matters to anyone not taking an inhibitor is unknown. The only long exposure experiment is in mice, where large daily doses for ten to twelve weeks produced no rise in blood pressure, a fall in nonfasting glucose, a small rise in heart rate at the higher dose, and increased hydrogen peroxide production in heart preparations, a combination that is reassuring and interesting and not human ²⁷.

Where it connects

Sources

98 sources, numbered as they are cited. Every one was checked against PubMed or its publisher before it was cited here; the note under each says what it shows and what it does not.

1. 1
   Gainetdinov RR, Hoener MC, Berry MD. Trace Amines and Their Receptors.Pharmacol Rev · 2018 · 70(3):549-620doi:10.1124/pr.117.015305 · PMID 29941461

   The most complete review of the field. Establishes that trace amines are abundant in common foodstuffs and are produced and degraded by the constitutive microbiota, that the ability to use them arose at least twice in evolution, and that TAAR1 acts as a rheostat on dopaminergic, glutamatergic and serotonergic transmission.

2. 2
   Zhang X, Li S, Liu H, et al. Isolation and Characterizations of Histamine- and Tyramine-Producing Strains Isolated from Fermented Soybean Food: Soy Sauce and Soybean Paste.Foods · 2025 · 14(14):2407doi:10.3390/foods14142407 · PMID 40724227

   Identifies the strains that actually make the amine in soy sauce and soybean paste, and shows that the usual polymerase chain reaction primers for the tyrosine decarboxylase gene miss many producing strains, giving false negatives. A laboratory finding about methods, not about human harm.

3. 3
   George CF. Drug metabolism by the gastrointestinal mucosa.Clin Pharmacokinet · 1981 · 6(4):259-74doi:10.2165/00003088-198106040-00002 · PMID 6113909

   Review of first-pass metabolism in the gut wall, which names the oxidative deamination of tyramine as one of the few phase I reactions there that matters quantitatively.

4. 4
   Habuchi Y, Tanaka H, Nishio M, et al. Dopamine stimulation of cardiac beta-adrenoceptors: the involvement of sympathetic amine transporters and the effect of SKF38393.Br J Pharmacol · 1997 · 122(8):1669-78doi:10.1038/sj.bjp.0701574 · PMID 9422813

   Guinea-pig heart. Shows that the release of norepinephrine by an indirectly acting amine runs through a desipramine-sensitive transporter, and that the tyramine effect fades on continued exposure while dopamine's does not. Animal work.

5. 5
   Burn JH, Rand MJ. The action of sympathomimetic amines in animals treated with reserpine.J Physiol · 1958 · 144(2):314-36doi:10.1113/jphysiol.1958.sp006104 · PMID 13611696

   The classical demonstration that emptying the nerve endings of their stored amine abolishes the action of the indirectly acting sympathomimetics. No abstract in the record; cited for the experiment and the date.

6. 6
   Finberg JPM. Update on the pharmacology of selective inhibitors of MAO-A and MAO-B: focus on modulation of CNS monoamine neurotransmitter release.Pharmacol Ther · 2014 · 143(2):133-52doi:10.1016/j.pharmthera.2014.02.010 · PMID 24607445

   Review. States plainly that potentiation of indirectly acting sympathomimetic amines is caused by MAO-A inhibitors and not MAO-B inhibitors, and that reversible MAO-A inhibitors cause minimal potentiation.

7. 7
   Finberg JPM, Gillman K. Selective inhibitors of monoamine oxidase type B and the "cheese effect".Int Rev Neurobiol · 2011 · 100:169-90doi:10.1016/B978-0-12-386467-3.00009-1 · PMID 21971008

   The reference account of the mechanism: inhibiting MAO-A, or both isoforms, produces the cheese effect; inhibiting MAO-B alone does not. Also surveys food contents and pressor testing.

8. 8
   Borowsky B, Adham N, Jones KA, et al. Trace amines: identification of a family of mammalian G protein-coupled receptors.Proc Natl Acad Sci U S A · 2001 · 98(16):8966-71doi:10.1073/pnas.151105198 · PMID 11459929

   One of the two 2001 cloning papers. Fifteen receptors identified by degenerate PCR from human and rodent tissue; two bind or are activated by trace amines; three of the four human receptors of the family are present in the amygdala.

9. 9
   Bunzow JR, Sonders MS, Arttamangkul S, et al. Amphetamine, 3,4-methylenedioxymethamphetamine, lysergic acid diethylamide, and metabolites of the catecholamine neurotransmitters are agonists of a rat trace amine receptor.Mol Pharmacol · 2001 · 60(6):1181-8doi:10.1124/mol.60.6.1181 · PMID 11723224

   The other 2001 paper. A rat receptor that raises cyclic AMP when exposed to p-tyramine, beta-phenylethylamine, tryptamine and octopamine, and also to amphetamines, ergolines and 3-methylated catecholamine metabolites. Transfected cells.

10. 10
    Maguire JJ, Parker WAE, Foord SM, et al. International Union of Pharmacology. LXXII. Recommendations for trace amine receptor nomenclature.Pharmacol Rev · 2009 · 61(1):1-8doi:10.1124/pr.109.001107 · PMID 19325074

    The official naming. Also states the standard attribution of the "beer, wine, and cheese effect" to the sympathomimetic action of dietary trace amines, and that humans carry six functional receptors of the family with TAAR3 a pseudogene in some people and TAAR4 a pseudogene in all.

11. 11
    Blackwell B. Hypertensive crisis due to monoamine-oxidase inhibitors.Lancet · 1963 · 2(7313):849-50doi:10.1016/s0140-6736(63)92743-0 · PMID 14056007

    The letter that opened the case. The PubMed record carries no abstract, so it is cited for the report, the date and its indexed subject headings (cheese, headache, hypertension, phenelzine, tranylcypromine), not for any number.

12. 12
    Asatoor AM, Levi AJ, Milne MD. Tranylcypromine and cheese.Lancet · 1963 · 2(7310):733-4doi:10.1016/s0140-6736(63)90368-4 · PMID 14057166

    The chromatographic work conventionally credited with naming tyramine as the agent in cheese. No abstract in the record; cited for the finding and the date.

13. 13
    McCabe BJ. Dietary tyramine and other pressor amines in MAOI regimens: a review.J Am Diet Assoc · 1986 · 86(8):1059-64PMID 3525654

    The source of the numbers quoted ever since: about 6 mg of tyramine may produce a mild crisis and 10 to 25 mg severe headache with intracranial hemorrhage, with a recommendation to keep intake below 5 mg. These are the review's estimates from case reports and amine assays, not from a dose-finding study. No DOI in the record.

14. 14
    Pawul-Gruba M, Denis E, Kiljanek T, Osek J. Prevalence of Biogenic Amines and Their Relation to the Bacterial Content in Ripened Cheeses on the Retail Market in Poland.Foods · 2025 · 14(14):2478doi:10.3390/foods14142478 · PMID 40724300

    The best recent food survey: 125 ripened cheeses from retail, analyzed by high-performance liquid chromatography with diode-array detection. Tyramine was found in 28.0 percent of samples at 5.62 to 646 mg/kg, the high values accompanying Enterococcus faecium and Enterococcus faecalis. One country, one year, retail sampling.

15. 15
    Frampton JE, Plosker GL. Selegiline transdermal system: in the treatment of major depressive disorder.Drugs · 2007 · 67(2):257-65; discussion 266-7doi:10.2165/00003495-200767020-00006 · PMID 17284087

    Drug profile. States the design logic: deliver antidepressant concentrations to the brain without substantially inhibiting monoamine oxidase A in the small intestine, so that at the 6 mg/24 h target dose tyramine restrictions are not needed.

16. 16
    Azzaro AJ, Vandenberg CM, Blob LF, et al. Tyramine pressor sensitivity during treatment with the selegiline transdermal system 6 mg/24 h in healthy subjects.J Clin Pharmacol · 2006 · 46(8):933-44doi:10.1177/0091270006289852 · PMID 16855078

    The tyramine sensitivity factor was 1.85 plus or minus 0.10 after nine days of the 6 mg/24 h patch, similar to 10 mg a day of oral selegiline and more than twenty times lower than tranylcypromine. Healthy volunteers, industry authors.

17. 17
    Fulton B, Benfield P. Moclobemide. An update of its pharmacological properties and therapeutic use.Drugs · 1996 · 52(3):450-74doi:10.2165/00003495-199652030-00013 · PMID 8875133

    Drug review. States that moclobemide only weakly potentiates the pressor response to tyramine and that dietary tyramine need not be avoided with it, in contrast to the older drugs.

18. 18
    Roeder T. Tyramine and octopamine: ruling behavior and metabolism.Annu Rev Entomol · 2005 · 50:447-77doi:10.1146/annurev.ento.50.071803.130404 · PMID 15355245

    The standard account of the invertebrate side: octopamine and tyramine are the invertebrate counterparts of the vertebrate adrenergic transmitters, tyramine is the biological precursor of octopamine, and both are independent transmitters acting through their own G protein-coupled receptors.

19. 19
    Achtyes ED, Hopkins SC, Dedic N, et al. Ulotaront: review of preliminary evidence for the efficacy and safety of a TAAR1 agonist in schizophrenia.Eur Arch Psychiatry Clin Neurosci · 2023 · 273(7):1543-1556doi:10.1007/s00406-023-01580-3 · PMID 37165101

    The optimistic phase 2 account, written by the sponsor's own scientists: a four-week trial with an effect size of 0.45 on the Positive and Negative Syndrome Scale and fewer adverse events than placebo. A commercial conflict of interest is on the face of the authorship.

20. 20
    Huey JK, Shi X, Schutzer WE, Janowsky A, Abbas AI. Trace amine-associated receptor 1 agonists differentially regulate dopamine transporter function.Mol Pharmacol · 2025 · 107(9):100064doi:10.1016/j.molpha.2025.100064 · PMID 40840012

    Cell and rodent synaptosome work. Cited here for its plain statement that the phase 3 trials showed ulotaront to have the same efficacy as placebo, and for the finding that different TAAR1 agonists act differently on the dopamine transporter.

21. 21
    Gillman PK. A reassessment of the safety profile of monoamine oxidase inhibitors: elucidating tired old tyramine myths.J Neural Transm (Vienna) · 2018 · 125(11):1707-1717doi:10.1007/s00702-018-1932-y · PMID 30255284

    Appraises over 150 original papers on food tyramine. Argues that starter cultures and hygiene regulations have cut tyramine in modern foods so far that even matured cheeses are usually safe in ordinary portions, that cheese was the only food associated with documented fatalities, and that aggressive treatment of the blood pressure rise has its own harms. A strongly argued single-author review, not a systematic one.

22. 22
    Van den Eynde V, Gillman PK, Blackwell BB. The Prescriber's Guide to the MAOI Diet-Thinking Through Tyramine Troubles.Psychopharmacol Bull · 2022 · 52(2):73-116doi:10.64719/pb.4434 · PMID 35721816

    The current practical guide, co-authored by the man who described the reaction in 1963. Makes the key methodological point that tables of tyramine values are imprecise because levels vary widely within a food category, and that the pressor response varies widely between people. The authors disclose consultancy for companies developing these drugs.

23. 23
    Smith WV, Pulver SR. Octopamine and Tyramine Modulate Motor Program Competition in Drosophila larvae.J Neurophysiol · 2026doi:10.1152/jn.00564.2025 · PMID 42789988

    Isolated larval nervous systems. Octopamine promoted forward crawling and suppressed backward crawling; tyramine, its supposed inhibitory counterpart, promoted collisions between motor programs and bouts of silence. Fictive locomotion in a dish, not behavior in an intact animal.

24. 24
    Okamoto M, Nakamura K, Uegaki R, et al. Tyramine is not essential for European foulbrood pathogenesis in a virulent Melissococcus plutonius strain.J Vet Med Sci · 2026 · 88(9):1300-1306doi:10.1292/jvms.26-0187 · PMID 42438008

    A clean negative result. Knocking out the tyrosine decarboxylase gene abolished tyramine production in a virulent bee pathogen without reducing its virulence, which disposes of tyramine as the proposed virulence factor in that disease.

25. 25
    Hong K, Li C, Ai J, et al. Biogenic amines degradation ability of Saccharomyces cerevisiae I45 and Pichia sp. NW5 & LB60 and their application in beer fermentation.Food Res Int · 2025 · 202:115726doi:10.1016/j.foodres.2025.115726 · PMID 39967102

    Yeast strains selected to lower amines in stout beer; one degraded 41.75 percent of the tyramine. A fermentation experiment, with no human testing of the resulting beer.

26. 26
    Wang LH, Xu H, Chen S, et al. Combined Genome-Wide and Phenotypic Profiling of Lactiplantibacillus plantarum XHQ-007: Genome-Guided Insights into Tyramine Reduction, Safety Characteristics, and Probiotic Potential.Foods · 2026 · 15(11):1977doi:10.3390/foods15111977 · PMID 42279763

    Reports a 79.94 percent fall in detectable tyramine under fermentation conditions, and the authors themselves state that without pH-matched controls this cannot be attributed to enzymatic degradation. An honest negative caveat worth repeating.

27. 27
    Carpéné C, Schaak S, Guilbeau-Frugier C, Mercader J, Mialet-Perez J. High intake of dietary tyramine does not deteriorate glucose handling and does not cause adverse cardiovascular effects in mice.J Physiol Biochem · 2016 · 72(3):539-53doi:10.1007/s13105-015-0456-2 · PMID 26634369

    Mice given roughly 300 or 1100 micromoles per kilogram of body weight per day for ten to twelve weeks. Blood pressure did not rise, nonfasting glucose fell, the higher dose raised heart rate slightly, and hydrogen peroxide production rose in heart preparations. Mice with intact monoamine oxidase; it says nothing about a person on an inhibitor.

28. 28
    Burch D, Asgharnejad M, Gerson W, Fielding RM, Azzaro AJ. Lack of tyramine pressor response effect with oral CX157: A specific reversible MAOI.Clin Pharmacol Drug Dev · 2013 · 3(1):4-12doi:10.1002/cpdd.61 · PMID 27128224

    Describes the method of the pressor test precisely and states the working assumption that a high-tyramine meal contains up to 40 mg. No tyramine reaction occurred at oral doses up to 80 mg on the drug. Twelve subjects, ten on active drug.

29. 29
    Walker SE, Shulman KI, Tailor SA, Gardner D. Tyramine content of previously restricted foods in monoamine oxidase inhibitor diets.J Clin Psychopharmacol · 1996 · 16(5):383-8doi:10.1097/00004714-199610000-00007 · PMID 8889911

    Fifty-one foods by liquid chromatography. Chicken liver aged nine days held 63.84 mg per 30 g, sauerkraut 7.75 mg per 250 g and soy sauce 0.941 mg per mL. The air-dried sausage figure is printed in the abstract as 7.56 g per 30 g, which is a unit error for milligrams.

30. 30
    Blackwell B, Marley E, Price J, Taylor D. Hypertensive interactions between monoamine oxidase inhibitors and foodstuffs.Br J Psychiatry · 1967 · 113(497):349-65doi:10.1192/bjp.113.497.349 · PMID 6034391

    The clinical survey. No abstract in the record; its indexing lists cerebral hemorrhage and subarachnoid hemorrhage among the outcomes, which is the strongest statement the database entry supports. It gives no count of deaths that can be quoted.

31. 31
    Gardner DM, Shulman KI, Walker SE, Tailor SA. The making of a user friendly MAOI diet.J Clin Psychiatry · 1996 · 57(3):99-104PMID 8617704

    The list most modern diets descend from: avoid aged cheese, aged or cured meats, potentially spoiled meat, poultry or fish, broad bean pods, concentrated yeast extract, sauerkraut, soy sauce and soybean condiments, and tap beer; wine and bottled or canned beer in moderation are considered safe. No DOI in the record.

32. 32
    Johansson P, Säde E, Adhikari L, Björkroth J, Jääskeläinen E. Effects of Latilactobacillus fuchuensis and marination on spoilage and tyramine formation in rainbow trout packaged under modified atmosphere.Food Microbiol · 2026 · 139:105160doi:10.1016/j.fm.2026.105160 · PMID 42215236

    Challenge study in fish fillets at 3 degrees Celsius for 12 days. Marinated fillets inoculated with decarboxylase-positive strains reached up to 67 mg/kg of tyramine, while decarboxylase-negative strains of the same species produced little. Shows that the risk is a property of the strain, not of the species.

33. 33
    Pasculli L, Farooq M, Natrella G, et al. Biogenic amine production and degradation profiles of selected cheese isolates in cheese-derived model ecosystems.Food Res Int · 2026 · 242(Pt 4):120148doi:10.1016/j.foodres.2026.120148 · PMID 42637469

    Cheese isolates grown in broths made from long-ripened, medium-ripened and surface-ripened cheeses. Clustering of tyramine, putrescine and cadaverine was driven mainly by the broth rather than by microbial taxonomy or in vitro phenotype, with buffering capacity and the supply of peptides and precursor amino acids decisive. Model systems, not whole cheeses.

34. 34
    Nevsky ES, Fedorov NS, Mikshina PV, et al. Tyramine modulates synaptic transmission in vertebrate neuromuscular junction by releasing catecholamines and activating alpha-adrenergic receptors.Eur J Pharmacol · 2026 · 1037:179425doi:10.1016/j.ejphar.2026.179425 · PMID 42874907

    Mouse nerve-muscle preparation. Tyramine reduced both spontaneous and evoked acetylcholine release; the effect was abolished by an alpha-1 blocker and not by an alpha-2 blocker, consistent with indirect action through released catecholamines. The authors also detected endogenous release of tyramine and octopamine from the preparation. Animal work.

35. 35
    Finberg JPM, Rabey JM. Inhibitors of MAO-A and MAO-B in Psychiatry and Neurology.Front Pharmacol · 2016 · 7:340doi:10.3389/fphar.2016.00340 · PMID 27803666

    Review. Notes that several reversible MAO-A inhibitors free of the cheese effect have been described but only moclobemide is in clinical use, and that modifying selegiline's pharmacokinetics by the transdermal route produced a new drug form.

36. 36
    Bradaia A, Trube G, Stalder H, et al. The selective antagonist EPPTB reveals TAAR1-mediated regulatory mechanisms in dopaminergic neurons of the mesolimbic system.Proc Natl Acad Sci U S A · 2009 · 106(47):20081-6doi:10.1073/pnas.0906522106 · PMID 19892733

    Mouse brain slices. p-Tyramine slows the firing of ventral tegmental dopamine neurons through TAAR1 by opening an inwardly rectifying potassium current; the effects vanish in TAAR1 knockout mice. Animal work.

37. 37
    Xie Z, Miller GM. Beta-phenylethylamine alters monoamine transporter function via trace amine-associated receptor 1: implication for modulatory roles of trace amines in brain.J Pharmacol Exp Ther · 2008 · 325(2):617-28doi:10.1124/jpet.107.134247 · PMID 18182557

    Transfected cells and rodent and monkey synaptosomes. A trace amine acting through TAAR1 inhibits uptake and induces efflux of dopamine, norepinephrine and serotonin, and the effect is absent in TAAR1 knockout mouse tissue. Cell and animal work, not human.

38. 38
    Shang P, Rong N, Jiang JJ, et al. Structural and signaling mechanisms of TAAR1 enabled preferential agonist design.Cell · 2023 · 186(24):5347-5362.e24doi:10.1016/j.cell.2023.10.014 · PMID 37963465

    Nine structures of the receptor bound to its G proteins with endogenous amines, clinical drugs and synthetic compounds, showing the primary amine recognition pocket built around a conserved acidic aspartate and a twin toggle switch for activation.

39. 39
    Cantarini MV, Painter CJ, Gilmore EM, et al. Effect of oral linezolid on the pressor response to intravenous tyramine.Br J Clin Pharmacol · 2004 · 58(5):470-5doi:10.1111/j.1365-2125.2004.02186.x · PMID 15521893

    Twelve healthy men, placebo-controlled crossover. The tyramine sensitivity factor was 1.8 (90% CI 1.6 to 2.0) for a single 600 mg dose of the antibiotic linezolid and 2.1 (1.8 to 2.4) for moclobemide as the positive control. Body mass index significantly affected the threshold dose.

40. 40
    Elsworth JD, Glover V, Reynolds GP, et al. Deprenyl administration in man: a selective monoamine oxidase B inhibitor without the 'cheese effect'.Psychopharmacology (Berl) · 1978 · 57(1):33-8doi:10.1007/BF00426954 · PMID 96466

    The human demonstration behind the whole A-versus-B clinical rule. Four normal and six Parkinson's volunteers, given deprenyl in doses sufficient for complete inhibition of the platelet enzyme, had no pressor reaction to oral tyramine in amounts considerably greater than a normal diet supplies; normal human intestinal mucosa was shown to contain predominantly the deprenyl-insensitive A form. Ten volunteers, uncontrolled.

41. 41
    Johnston JP. Some observations upon a new inhibitor of monoamine oxidase in brain tissue.Biochem Pharmacol · 1968 · 17(7):1285-97doi:10.1016/0006-2952(68)90066-x · PMID 5659776

    The report of the inhibitor now called clorgyline and of the two enzyme activities named A and B. No abstract in the record; cited for the observation and the date.

42. 42
    Humphrey SJ, Curry JT, Turman CN, Stryd RP. Cardiovascular sympathomimetic amine interactions in rats treated with monoamine oxidase inhibitors and the novel oxazolidinone antibiotic linezolid.J Cardiovasc Pharmacol · 2001 · 37(5):548-63doi:10.1097/00005344-200105000-00007 · PMID 11336106

    Rat work. Gives linezolid's competitive inhibition constant for human MAO-A as 55 micromolar, and shows the tyramine pressor response potentiated by clorgyline and moclobemide but not by selegiline. Animal data.

43. 43
    Alnefeesi Y, Murtazina RZ, Gainetdinov RR. Trace Amine-associated Receptors (TAARs): Candidate Targets in the Treatment of Bipolar Disorders.Actas Esp Psiquiatr · 2025 · 53(5):1140-1153doi:10.62641/aep.v53i5.1916 · PMID 41117250

    Review. Records the six functional human receptors, the breakthrough therapy designation given to ulotaront in schizophrenia, and its continuing development in major depressive disorder and generalized anxiety disorder.

44. 44
    Siafis S, Chiocchia V, Macleod MR, et al. Trace amine-associated receptor 1 (TAAR1) agonism for psychosis: a living systematic review and meta-analysis of human and non-human data.Wellcome Open Res · 2024 · 9:182doi:10.12688/wellcomeopenres.21302.1 · PMID 39036710

    The pooled human result: across four studies and 1291 adults with acute schizophrenia the standardized mean difference against placebo was 0.15 (95% CI -0.05 to 0.34), and ralmitaront was inferior to risperidone. Notes the large placebo response in the ulotaront phase 3 trials.

45. 45
    Smith KA, Boyce N, Chevance A, et al. Triangulating evidence from the GALENOS living systematic review on trace amine-associated receptor 1 (TAAR1) agonists in psychosis.Br J Psychiatry · 2024 · 226(3):162-170doi:10.1192/bjp.2024.237 · PMID 39710623

    The expert appraisal of the same evidence, including people with lived experience. Confirms the animal-to-human gap: animal models suggested an antipsychotic effect that the human trials did not deliver.

46. 46
    Hsu YC, Hung TY, Chen YB, et al. Trajectory of efficacy and safety across ulotaront dose levels in schizophrenia: a systematic review and dose-response meta-analysis.Int J Neuropsychopharmacol · 2025 · 28(9)doi:10.1093/ijnp/pyaf059 · PMID 40795331

    Three randomized trials, 1144 participants. The best dose, 100 mg, gave a standardized mean difference of -0.23 (95% CI -0.43 to -0.02) on total symptoms; anxiety-related adverse events were commoner than placebo at 75 mg (risk ratio 2.06, 95% CI 1.11 to 3.80).

47. 47
    Werner FM, Coveñas R. Clinical Effects of Recently Developed Antipsychotic Drugs in Schizophrenia.Cent Nerv Syst Agents Med Chem · 2026doi:10.2174/0118715249384914251122074130 · PMID 41588897

    A 2026 review that still lists ulotaront among promising new antipsychotics and states that TAAR1 agonists substantially improve negative symptoms. It is cited here as an example of how unevenly the phase 3 result has been absorbed; the meta-analyses are the better guide.

48. 48
    Howes OD, Vano LJ, Petty AE, et al. TAAR1 Regulates Presynaptic Dopamine Function: Evidence From Preclinical Studies and a Phase 1b Trial in Patients With Schizophrenia.Biol Psychiatry · 2025 · 100(2):199-207doi:10.1016/j.biopsych.2025.10.013 · PMID 41130556

    Open-label phase 1 trial in 22 patients plus mouse work. Fourteen days of added ulotaront lowered striatal dopamine synthesis capacity, and the fall correlated with improvement in positive symptoms (r = 0.5). Open-label, no placebo group, and sponsor co-authors.

49. 49
    Huang S, Yang SM, Bellier L, et al. Antipsychotic-like effects of TAAR1 agonists and D2 receptor blockers are mediated through distinct modulation of striatal neuronal dynamics.Mol Psychiatry · 2026 · 31(11):6559-6571doi:10.1038/s41380-026-03724-y · PMID 42481744

    Calcium imaging in freely behaving mice. TAAR1 agonists and dopamine D2 blockers both reverse a drug-induced psychosis-like state but do so through different striatal signatures. Animal work; several authors are employed by the sponsor.

50. 50
    King A, Hardy M, Rhee TG, Wilkinson ST. Next-Generation Pharmacotherapies for Unipolar Mood Disorders: A Leading Article on Phase II/III Drug Development Across Diverse Mechanisms.CNS Drugs · 2026 · 40(11):1429-1444doi:10.1007/s40263-026-01330-8 · PMID 42645703

    Review of phase 2 and phase 3 industry-sponsored United States trials active within the past three years. SEP-363856, which is ulotaront, is among the compounds listed, which is the currency statement for its continuing development in mood disorders.

51. 51
    Milanovic S, Lew R, Xiong K, et al. An open-label study on ulotaront's effects on insulin-glucose regulation in schizophrenia patients with metabolic syndrome and prediabetes: Part I.Diabetes Obes Metab · 2026 · 28(1):103-111doi:10.1111/dom.70136 · PMID 41126393

    Phase 1b, open-label, small, terminated early, and conducted by the sponsor. Insulin response after a solid meal fell with nominal significance (p = 0.021); the glucose and C-peptide changes were trends.

52. 52
    Lenham RK, Elgammal F, Harding SV, Berry MD. TAAR1 as a Potential Alternate Molecular Target to GLP-1 for Novel Anti-diabetic, Anti-obesity Medications.Handb Exp Pharmacol · 2026doi:10.1007/164_2026_821 · PMID 42432129

    Review of the receptor's metabolic side: preclinical and early clinical evidence that activating it enhances glucose-stimulated insulin secretion. A proposal, not a demonstrated treatment.

53. 53
    Zhang W, Wu W, Gao J, et al. Medicinal chemistry advances of TAAR1 agonists: SAR, structural innovation, and mechanistic insights.Eur J Med Chem · 2026 · 316:119058doi:10.1016/j.ejmech.2026.119058 · PMID 42302664

    Review. Records that only two TAAR1 agonists, ulotaront and ralmitaront, have reached clinical trials in schizophrenia to date.

54. 54
    Huang S, Liu H, Wu Y, et al. Structure based discovery of antipsychotic-like TAAR1 agonists.Nat Commun · 2026 · 17(1)doi:10.1038/s41467-026-77484-2 · PMID 42894001

    Sixty-five million molecules docked against the active state of the receptor, 55 tested, fourteen agonists found with potencies from mid-nanomolar to micromolar; three optimized compounds with high brain exposure normalized amphetamine-disrupted prepulse inhibition in mice without catalepsy, and the docked poses were confirmed by cryo-electron microscopy. Cell pharmacology and mouse behavior.

55. 55
    Shulman KI, Walker SE, MacKenzie S, Knowles S. Dietary restriction, tyramine, and the use of monoamine oxidase inhibitors.J Clin Psychopharmacol · 1989 · 9(6):397-402PMID 2592589

    Over 100 controversial foods assayed. Only aged cheeses, concentrated yeast extracts, sauerkraut and broad bean pods emerged as requiring absolute restriction; alcoholic drinks in moderation, including Chianti, appeared safe. No DOI in the record.

56. 56
    Shulman KI, Tailor SA, Walker SE, Gardner DM. Tap (draft) beer and monoamine oxidase inhibitor dietary restrictions.Can J Psychiatry · 1997 · 42(3):310-2doi:10.1177/070674379704200311 · PMID 9114949

    A single documented case of hypertensive crisis after a modest amount of tap beer in a patient taking phenelzine, which with an earlier similar case is the basis for restricting draft beer. Two case reports, not a trial.

57. 57
    Shulman KI, Walker SE. Refining the MAOI diet: tyramine content of pizzas and soy products.J Clin Psychiatry · 1999 · 60(3):191-3PMID 10192596

    High-pressure liquid chromatography of pizzas and soy products against a safety threshold of 6 mg. No significant tyramine in any pizza, including double pepperoni and double cheese; marked variability in soy products, with clinically significant levels in tofu stored a week. No DOI in the record.

58. 58
    Ueda Y, Fujii Y, Omae S, Aoyanagi N, Nishimura K. [Investigation of Nonvolatile Amines in Agricultural Products].Shokuhin Eiseigaku Zasshi · 2025 · 66(6):141-145doi:10.3358/shokueishi.66.141 · PMID 41443853

    One hundred and two samples of 64 agricultural products in Japan. Tyramine was present in 29 products, peaking at 132 micrograms per gram in hanabiratake mushroom, and the authors state that patients taking monoamine oxidase inhibitors may need to exercise caution with the high-tyramine products. In Japanese with an English abstract; the bracketed title is the database's translation.

59. 59
    Rampazzo G, Depau G, Pagliuca G, et al. A Rapid LC-MS/MS Method for Quantification of Biogenic Amines in Meat: Validation and Application for Food Safety Monitoring.Methods Protoc · 2025 · 8(5):106doi:10.3390/mps8050106 · PMID 40981224

    A method paper applied to ten commercial meat samples, with a limit of quantification of 10 micrograms per gram. Putrescine, tyramine and spermidine were the amines most often detected; histamine and tyramine were below toxicological thresholds for healthy individuals, but one sample showed tyramine at a level the authors call potentially concerning for a consumer treated with a monoamine oxidase inhibitor. Ten samples.

60. 60
    Geist E. When ice cream was poisonous: adulteration, ptomaines, and bacteriology in the United States, 1850-1910.Bull Hist Med · 2012 · 86(3):333-60doi:10.1353/bhm.2012.0053 · PMID 23241909

    The historical account of tyrotoxicon: Victor C. Vaughan's proposed cheese and ice cream poison, its wide acceptance in the late 1880s, and the decline of the whole ptomaine idea. The article does not identify tyrotoxicon with tyramine.

61. 61
    Tyrotoxicon Poisoning.Br Med J · 1899 · 2(2027):1300PMID 20758714

    A contemporary note showing that tyrotoxicon was treated as a working clinical diagnosis at the end of the nineteenth century. No abstract, no listed author and no DOI in the record; cited as a primary document of its period.

62. 62
    Dale HH, Dixon WE. The action of pressor amines produced by putrefaction.J Physiol · 1909 · 39(1):25-44doi:10.1113/jphysiol.1909.sp001325 · PMID 16992970

    The first pharmacology of the amines of putrefaction, the class tyramine belongs to. No abstract in the record; cited for the work and the date.

63. 63
    Barger G, Dale HH. Chemical structure and sympathomimetic action of amines.J Physiol · 1910 · 41(1-2):19-59doi:10.1113/jphysiol.1910.sp001392 · PMID 16993040

    The structure-activity work that produced the word sympathomimetic. No abstract in the record; cited for the work and the date.

64. 64
    Hare ML. Tyramine oxidase: A new enzyme system in liver.Biochem J · 1928 · 22(4):968-79doi:10.1042/bj0220968 · PMID 16744124

    The paper that named the enzyme after this molecule: tyramine oxidase, later renamed monoamine oxidase. Written from the Biochemical Laboratory, Cambridge. No abstract in the record; cited for the finding, the name and the date.

65. 65
    Blackwell B, Marley E. Interaction between cheese and monoamine-oxidase inhibitors in rats and cats.Lancet · 1964 · 1(7332):530-1doi:10.1016/s0140-6736(64)92914-9 · PMID 14100188

    The animal reproduction of the cheese interaction, a year after the clinical report. No abstract in the record; its indexed terms include cheese, hypertension, phenelzine, tranylcypromine and tyramine. Cited for the work and the date.

66. 66
    Blackwell B, Mabbitt LA. Tyramine in cheese related to hypertensive crises after monoamine-oxidase inhibition.Lancet · 1965 · 1(7392):938-40doi:10.1016/s0140-6736(65)91257-2 · PMID 14275713

    The paper that tied measured tyramine in cheeses to the crises; tyrosine decarboxylase and Streptococcus are among the indexed terms, which is the bacterial half of the story. No abstract in the record; cited for the work and the date.

67. 67
    Blackwell B, Marley E, Mabbitt LA. Effects of yeast extract after monoamine-oxidase inhibition.Lancet · 1965 · 1(7392):940-3doi:10.1016/s0140-6736(65)91258-4 · PMID 14275831

    The second food, confirmed. No abstract in the record; cited for the work and the date.

68. 68
    Blackwell B, Marley E, Taylor D. Effects of yeast extract after monoamine-oxidase inhibition.Lancet · 1965 · 1(7396):1166doi:10.1016/s0140-6736(65)91994-x · PMID 14287333

    The follow-up letter four weeks later. No abstract in the record; cited for the work and the date.

69. 69
    Blackwell B, Marley E. Interactions of cheese and of its constituents with monoamine oxidase inhibitors.Br J Pharmacol Chemother · 1966 · 26(1):120-41doi:10.1111/j.1476-5381.1966.tb01816.x · PMID 5950679

    The animal experiments behind the clinical claim, in cats and rats. No abstract in the record; cited for the existence and design of the work, not for numbers.

70. 70
    Blackwell B, Marley E. Interactions of yeast extracts and their constituents with monoamine oxidase inhibitors.Br J Pharmacol Chemother · 1966 · 26(1):142-61doi:10.1111/j.1476-5381.1966.tb01817.x · PMID 5919499

    The companion paper on yeast extracts, the second food to be implicated. No abstract in the record.

71. 71
    Hannington E. Preliminary report on tyramine headache.Br Med J · 1967 · 2(5551):550-1doi:10.1136/bmj.2.5551.550 · PMID 5337268

    The origin of the migraine claim, indexed as a controlled clinical trial with lactose as the comparator. The record carries no abstract and no numbers, so it is cited for the claim and the date. PubMed spells the author Hannington; the literature usually writes Hanington.

72. 72
    Marley E, Blackwell B. Interactions of monoamine oxidase inhibitors, amines, and foodstuffs.Adv Pharmacol Chemother · 1970 · 8:185-349doi:10.1016/s1054-3589(08)60597-9 · PMID 4947578

    The 165-page monograph that consolidated the field. No abstract in the record.

73. 73
    Moffett A, Swash M, Scott DF. Effect of tyramine in migraine: a double-blind study.J Neurol Neurosurg Psychiatry · 1972 · 35(4):496-9doi:10.1136/jnnp.35.4.496 · PMID 4559027

    Twenty-five patients, placebo-controlled. Headache occurred in 12 of 50 sessions, 10 of them in the group who believed food triggered their attacks, and within that group headache followed tyramine alone in only two patients. The authors found no significant relation between tyramine and headache, while reporting electroencephalographic changes.

74. 74
    Forsythe WI, Redmond A. Two controlled trials of tyramine in children with migraine.Dev Med Child Neurol · 1974 · 16(6):794-99doi:10.1111/j.1469-8749.1974.tb03400.x · PMID 4613591

    Two controlled trials in children. The PubMed record carries no abstract, so this is cited for the existence and design of the trials only, not for their numbers.

75. 75
    Lüthy J, Schlatter C. [Biogenic amines in food: effects of histamine, tyramine and phenylethylamine in the human].Z Lebensm Unters Forsch · 1983 · 177(6):439-43doi:10.1007/BF01409672 · PMID 6364621

    Twenty-seven healthy volunteers given 25 mg of histamine, 25 mg of tyramine or 5 mg of phenylethylamine in apple juice, randomized, double-blind, placebo-controlled. Neither histamine nor tyramine produced a statistically significant effect; phenylethylamine did in some. Healthy volunteers, not people who report reacting to these foods. In German.

76. 76
    Ghose K, Carroll JD. Mechanism of tyramine-induced migraine: similarity with dopamine and interactions with disulfiram and propranolol in migraine patients.Neuropsychobiology · 1984 · 12(2-3):122-6doi:10.1159/000118123 · PMID 6527752

    Eight patients with classical migraine, double-blind crossover. Between 50 and 75 percent had an attack within 24 hours of a pressor test, with no difference between tyramine and dopamine. The authors concluded it is unlikely that tyramine plays any specific role except through the adrenergic system. Eight patients is a very small sample.

77. 77
    Brown C, Taniguchi G, Yip K. The monoamine oxidase inhibitor-tyramine interaction.J Clin Pharmacol · 1989 · 29(6):529-32doi:10.1002/j.1552-4604.1989.tb03376.x · PMID 2666453

    Short review of how the interaction was worked out and how fear of intracerebral hemorrhage and death drove clinicians away from the drug class.

78. 78
    Barrett JS, Hochadel TJ, Morales RJ, et al. Pressor response to tyramine after single 24-hour application of a selegiline transdermal system in healthy males.J Clin Pharmacol · 1997 · 37(3):238-47doi:10.1002/j.1552-4604.1997.tb04786.x · PMID 9089426

    Fifteen healthy men in three groups. Tyramine doses up to 600 mg at baseline and up to 200 mg on treatment produced no apparent increase in sensitivity after a single 24-hour patch application. Single-blind, very small, single dose.

79. 79
    Blob LF, Sharoky M, Campbell BJ, et al. Effects of a tyramine-enriched meal on blood pressure response in healthy male volunteers treated with selegiline transdermal system 6 mg/24 hour.CNS Spectr · 2007 · 12(1):25-34doi:10.1017/s1092852900020496 · PMID 17192761

    Twelve healthy adult men ate meals supplying roughly 400 mg of tyramine as aged cheeses after thirteen days on the patch, with no clinically significant change in cardiovascular vital signs. Twelve healthy men is a small sample, and the sponsor conducted the study.

80. 80
    Jansen SC, van Dusseldorp M, Bottema KC, Dubois AEJ. Intolerance to dietary biogenic amines: a review.Ann Allergy Asthma Immunol · 2003 · 91(3):233-40; quiz 241-2, 296doi:10.1016/S1081-1206(10)63523-5 · PMID 14533654

    The most rigorous appraisal of the challenge literature. Of thirteen oral challenge studies, three were ineligible and six of the remaining ten inconclusive; all four methodologically conclusive studies were negative, two of them specifically on tyramine and migraine. The literature reviewed ends in 2001.

81. 81
    Maini Rekdal V, Bess EN, Bisanz JE, Turnbaugh PJ, Balskus EP. Discovery and inhibition of an interspecies gut bacterial pathway for Levodopa metabolism.Science · 2019 · 364(6445):eaau6323doi:10.1126/science.aau6323 · PMID 31196984

    The same bacterial enzyme family, tyrosine decarboxylase from Enterococcus faecalis, acting on a drug instead of on food protein. Note that the amine produced downstream is m-tyramine, the meta isomer, not the para-tyramine of this page. Human gut communities in vitro plus mouse work.

82. 82
    Alzahrani H. Assessment of Chocolate, Cheese, and Caffeine as Migraine Triggers: A Systematic Review and Meta-Analysis of Patient-Reported Trigger Rates.J Nutr Metab · 2026 · 2026:4693275doi:10.1155/jnme/4693275 · PMID 42662243

    Pooled patient-reported trigger rates from seven studies: chocolate 13.9 percent (95% CI 7.0 to 21.0), cheese 8.0 percent (3.0 to 12.0), caffeine 12.0 percent (6.0 to 19.0). These are what patients report, not what challenge studies confirm, and heterogeneity exceeded 95 percent.

83. 83
    Rached G, Campana A, Fiani D, et al. Safety and Efficacy of Monoamine Oxidase Inhibitors in Patients Who Use Psychoactive Substances: Potential Drug Interactions and Substance Use Disorder Treatment Data.CNS Drugs · 2026 · 40(3):359-417doi:10.1007/s40263-025-01256-7 · PMID 41546846

    Narrative review of 219 publications spanning 1955 to 2025. Lists hypertensive urgency or emergency reported with these drugs in combination with alcoholic drinks containing significant tyramine, amphetamines, cocaine, khat and others, and notes such tyramine-rich drinks are now uncommon. Case reports and forum posts are included among the sources.

84. 84
    Bharani T, Mogensen KM, Rosen JH, Gura KM, Robinson MK. Concomitant use of monoamine oxidase inhibitor and tyrosine in parenteral nutrition.Eur J Clin Nutr · 2024 · 78(3):274-276doi:10.1038/s41430-023-01375-8 · PMID 38057413

    A single case. Intravenous feeding supplying tyrosine was given with transdermal selegiline without a pressor reaction. One patient is not evidence of safety, and the authors say so.

85. 85
    Hodge JV, Nye ER, Emerson GW. Monoamine-oxidase inhibitors, broad beans, and hypertension.Lancet · 1964 · 1(7342):1108doi:10.1016/s0140-6736(64)91316-9 · PMID 14132623

    The origin of the broad bean restriction, which is a levodopa story and not a tyramine one: the record is indexed under dihydroxyphenylalanine, headache and hypertension. No abstract in the record; cited for the report and the date.

86. 86
    Dingemanse J, Wood N, Guentert T, et al. Clinical pharmacology of moclobemide during chronic administration of high doses to healthy subjects.Psychopharmacology (Berl) · 1998 · 140(2):164-72doi:10.1007/s002130050754 · PMID 9860107

    Twenty-four healthy subjects on 450 or 600 mg twice daily for eight days. Intravenous tyramine pressor sensitivity rose three- to fourfold; 50 mg of oral tyramine did not raise systolic pressure by 30 mmHg, and 100 mg was compatible with the lower dose but not the higher.

87. 87
    Dingemanse J, Hussain Y, Korn A. Tyramine pharmacodynamics during combined administration of lazabemide and moclobemide.Int J Clin Pharmacol Ther · 1996 · 34(4):172-7PMID 8861736

    Blocking both isoforms reversibly at once potentiated oral tyramine 13.5 plus or minus 6.9 fold, and only 51 plus or minus 20 mg of tyramine was then needed to raise systolic pressure by 30 mmHg. Thirteen healthy subjects in total. No DOI in the record.

88. 88
    Flanagan S, Bartizal K, Minassian SL, Fang E, Prokocimer P. In vitro, in vivo, and clinical studies of tedizolid to assess the potential for peripheral or central monoamine oxidase interactions.Antimicrob Agents Chemother · 2013 · 57(7):3060-6doi:10.1128/AAC.00431-13 · PMID 23612197

    Tedizolid inhibited human MAO-A with a 50 percent inhibitory concentration of 8.7 micromolar and MAO-B at 5.7, against 46.0 and 2.1 for linezolid. In people the tyramine sensitivity ratio was 1.33, below the ratio of 2 regarded as clinically relevant. Sponsor-run.

89. 89
    Marquet A, Kupas K, Johne A, et al. The effect of safinamide, a novel drug for Parkinson's disease, on pressor response to oral tyramine: a randomized, double-blind, clinical trial.Clin Pharmacol Ther · 2012 · 92(4):450-7doi:10.1038/clpt.2012.128 · PMID 22948897

    The cleanest side-by-side ranking of tyramine sensitivity factors in healthy volunteers: placebo 1.52, safinamide 100 mg 2.15, safinamide 350 mg 2.74, selegiline 3.12, phenelzine 9.98. Healthy volunteers under fasting conditions, which exaggerates absorption.

90. 90
    Di Stefano AFD, Rusca A. Pressor response to oral tyramine during co-administration with safinamide in healthy volunteers.Naunyn Schmiedebergs Arch Pharmacol · 2011 · 384(6):505-15doi:10.1007/s00210-011-0674-2 · PMID 21850574

    Twenty volunteers preselected as tyramine-sensitive. No systolic rise of 30 mmHg or more occurred with tyramine up to 200 mg during safinamide 300 mg once daily.

91. 91
    Shulman KI, Herrmann N, Walker SE. Current place of monoamine oxidase inhibitors in the treatment of depression.CNS Drugs · 2013 · 27(10):789-97doi:10.1007/s40263-013-0097-3 · PMID 23934742

    Review. States that despite guideline recommendations the prescription rate of these drugs is far below what would be expected and is falling, and that with attention to diet and interactions the class can be used safely.

92. 92
    Chalon S, Bieck PR, Goldstein DJ, et al. The tyramine pressor test may have limited sensitivity, especially in the presence of dual serotonin/norepinephrine uptake inhibition.Neuropsychopharmacology · 2002 · 26(5):698-9; author reply 700-1doi:10.1016/S0893-133X(01)00409-2 · PMID 11927196

    A letter, not a study. Raises the point that the test's result depends on uptake into the nerve ending, so a drug that blocks that uptake can mask what the test is meant to measure.

93. 93
    Nikolitch K, Phillips JL, Daniels S, Blier P. Levomilnacipran, but Not Duloxetine, Inhibits Serotonin and Norepinephrine Reuptake Throughout Its Therapeutic Range.J Clin Psychiatry · 2025 · 86(3)doi:10.4088/JCP.25m15867 · PMID 40875503

    Randomized, placebo-controlled, 29 healthy men (10 on levomilnacipran, 9 on duloxetine, 10 on placebo; ClinicalTrials.gov NCT03249311). Attenuation of the tyramine pressor response was the read-out of norepinephrine reuptake inhibition: it separated from baseline at 40 mg of levomilnacipran and only at 120 mg of duloxetine. Cited here for what it shows about the test rather than about the drugs.

94. 94
    Maintz L, Novak N. Histamine and histamine intolerance.Am J Clin Nutr · 2007 · 85(5):1185-96doi:10.1093/ajcn/85.5.1185 · PMID 17490952

    The review that defined the modern concept. Concedes that because the symptoms are so varied the condition has been underestimated and that double-blind placebo-controlled provocation studies are still needed.

95. 95
    Sánchez-Pérez S, Comas-Basté O, Rabell-González J, et al. Biogenic Amines in Plant-Origin Foods: Are They Frequently Underestimated in Low-Histamine Diets?Foods · 2018 · 7(12):205doi:10.3390/foods7120205 · PMID 30558197

    Review of amine content in non-fermented plant foods. Only eggplant, spinach, tomato and avocado carried significant amounts of histamine, with wide variability. The enzymatic competition argument it makes concerns putrescine and spermidine at diamine oxidase, not tyramine.

96. 96
    Comas-Basté O, Sánchez-Pérez S, Veciana-Nogués MT, Latorre-Moratalla M, Vidal-Carou MDC. Histamine Intolerance: The Current State of the Art.Biomolecules · 2020 · 10(8):1181doi:10.3390/biom10081181 · PMID 32824107

    Current review. States that more scientific evidence is still required to define, diagnose and manage the condition, and introduces the idea of individual susceptibility to ingested amines.

97. 97
    Finetti L, Roeder T, Calò G, Bernacchia G. The Insect Type 1 Tyramine Receptors: From Structure to Behavior.Insects · 2021 · 12(4):315doi:10.3390/insects12040315 · PMID 33915977

    Review of the insect tyramine receptor, including that it was first misclassified as an octopamine receptor, and of insecticidal compounds that act on it.

98. 98
    Grandy DK. Trace amine-associated receptor 1-Family archetype or iconoclast?Pharmacol Ther · 2007 · 116(3):355-90doi:10.1016/j.pharmthera.2007.06.007 · PMID 17888514

    Review by one of the 2001 discoverers. Sets out the structure-activity profile of the receptor and states the central problem: heterologously expressed TAAR1 behaves as a trace amine receptor, but the physiology of each family member in a living animal is unsettled.