MENU

SUBSCRIPTIONS

Migraine

By Serge Kreutz
2025

This article advances the hypothesis that in many cases, migraine is a foodborne illness. The common causative microorganisms are a range of bacteria, including Lactobacillus strains used in the fermentation of milk and vegetables and affecting the decay of many foods. These bacteria convert amino acids into biogenic amines. All mammalian organisms use some amines as neurotransmitters and in the immune system, and can synthesize them themselves in necessary quantities. Excess amines, whether produced by an organism itself or entering the body as food, are broken down by amine oxidase enzymes present throughout any mammalian organism. The particularities of these enzymes are genetically encoded, and thus vary from person to person. Some people handle amines entering the body with food better than others. This is in line with migraine running in families. If excess biogenic amines are not broken down in a timely manner, they cause, in extreme cases, food poisoning, even fatal. On a lesser scale, so the hypothesis of this article, poorly managed excess biogenic amines cause migraine.

Foodborne disease

The US FDA, on its consumer pages, classifies 16 conditions as foodborne diseases. [What You Need to Know about Foodborne Illnesses]. Of those, 14 are caused by bacteria and 2 by viruses. Salmonella infection is the most well-known bacterial, and Hepatitis A the best-known viral foodborne disease.

The most dangerous foodborne illness on the FDA list is botulism, caused by Clostridium botulinum bacteria. An outbreak in the US was linked to potatoes baked in aluminum foil [Botulism: the hazardous baked potato]. When Otto Warmbier, sentenced to 15 years of hard labor for allegedly removing a propaganda poster from a hotel floor, was returned by North Korea on June 13, 2017, he was in a coma from a botulism infection and died 4 days later. Payment of a 2 million US dollars North Korean hospital bill was approved by Trump [Reuters].

However, the FDA definition of foodborne illnesses is too narrow. Heavy metal poisoning should also be included, as it is easily avoided by not consuming contaminated food. Celiac disease, an autoimmune condition caused by the consumption of wheat, should also be on the list, and of course obesity, the mother of all foodborne diseases. Also add scombroid poisoning, caused by high levels of histamine (scombroid is a category of fish that includes mackerel, barracuda, and tuna). And don't forget migraine, caused by tyramine which, like histamine, is introduced into food by bacteria that decarboxylate amino acids.

What happens during migraine attacks?

The traditional view of migraine explained it as a vascular condition. During the aura phase, blood vessels around the skull were thought to be contracted (or cramped). This would reduce blood flow to the eyes, causing typical visual effects like flickering or blind spots, or other sensual deficiencies. Then the blood vessels would expand disproportionately and exert pressure on nerves around the skull, thus causing pain.

Turned out old science was wrong. When blood vessels were measured during the aura and headache phases, they weren't constricted or expanded in line with the theory. In 2009, Brain (one of the world's foremost journals of neurology, published by Oxford University Press since 1878) summarized the change of winds in the headline: "The vascular theory of migraine - a great story wrecked by the facts" [Brain, Volume 132, Issue 1, January 2009]

In the early years of the third millennium, a new theory gained traction. Migraine was considered a neurological disorder, like epilepsy, but usually not as severe. The supposed key event is called cortical spreading depression, "a slowly propagated wave of depolarization followed by suppression of brain activity" [Nature]. The Brazilian neurologist Aristides de Azevedo Pacheco Leão discovered the phenomenon in 1944. He opened the skulls of anesthetized rabbits and inserted some electrodes because he wanted to provoke epileptic seizures. What happened instead was a spreading depression of brain activity [Aristides Leão's discovery]. Interestingly, "the cortical spreading depression (CSD) could be blocked by an incision in the cortex" [Scientific Electronic Library Online].

Remember: Frontal lobotomy (the drilling of holes into the foreheads of people) was initiated by the Portuguese neurologist Antonio Caetano de Abreu Freire Egas Moniz in the mid-1930s in order to calm down unruly psychiatric patients. This "solution" earned him a Nobel Prize in 1949. Nowadays, surgical treatment of migraine is also being offered. It involves the severance of various branches of the trigeminal nerve. Like frontal lobotomy, it is irreversible.

So, what is the link between migraine and cortical spreading depression. No interference with cortical spreading depression ever stopped a migraine, and it was never possible to induce a migraine by setting off a cortical spreading depression. All research on cortical spreading depressions was done in rabbits and cats. It wasn't even possible to prove the existence of cortical spreading depression in humans, unless a person had suffered extensive brain injury. Reason: cortical spreading depression propagates on the surface of brains. It does not jump easily over sulci (the furrows in brain folds) from brain gyri (bulged folds) to gyri; but heavily damaged brains do not have divided folds like healthy brains, thus cortical depressions spread more easily. During migraine attacks in humans, cortical spreading depression has never been measured (a fact for which neurologists came up with many excuses). [Persistent questions]

So, how was cortical spreading depression ever linked to migraine? Well, by intuition! By 1958, Peter Milner, an electrical engineer who retrained as a neuroscientist, felt that the time frame and slow intensity variations of migraine auras were similar to the time frame calculated (not measured!) for cortical spreading depressions (2 to 6 mm / minute) for the human brain. [Cortical spreading depression in neurological disorders: migraine ...] Milner based his hypothesis on the observations of a psychologist, Karl Lashley, who described his own migraine auras as slow waves overwhelming his eyesight.

Wow! Just that?

Yeap.

In 2018, the journal Neurological Research published a review by cardiologist Piet Borgdorff under the headline "Arguments against the role of cortical spreading depression in migraine". The abstract stated: "CSD is difficult to evoke in man, and electroencephalography (EEG) readings are not flattened during migraine (as opposed to EEG during CSD). Moreover, in contrast to CSD, migraine can occur bilaterally, and is not accompanied by a disrupted blood' brain barrier, increased cerebral metabolism, or cerebral cell swelling. Calcitonin gene-related peptide, which is thought to be characteristic of migraine pain, is increased in the blood from the external jugular vein during migraine in humans, but not during CSD in cats or rats." Another great story wrecked by the facts.

A less grandiose hypothesis views migraine as an often sporadic inflammation of the trigeminal nerve. This theory has the advantage that there are parameters that can actually be measured. One of them is the neuropeptide calcitonin gene-related peptide which, as indicated in the quote above, is elevated in the blood of the jugular vein during migraine headache.

As a scientific explanation for migraine, cortical spreading depression has probably undergone brain death.

Medical marketing

Cortical spreading depression, however, is still very much alive in the marketing efforts of neurologists. YouTube is full of clips that have been uploaded by headache web sites. Neurologists want you to hire their services... elaborate diagnostics, expensive consultations which they bill to your insurance company. Roughly 20 percent of Americans suffer from migraine [Source: The Prevalence and Impact of Migraine and Severe Headache in the United States: Figures and Trends From Government Health Studies]. 65 million potential patients. If only half shows up for consultations with neurologists, that's going to be a bonanza.

Neurologists will prescribe neurological medications, typically expensive. Thus, big pharma is on the bandwagon of the neurologists.

Nutritional triggers

The vascular, neurological, and inflammatory theories of migraine focus on what happens during migraine episodes. An entirely different perspective concentrates on what causes migraine episodes. Patients who are cured because they can avoid the causes of migraine attacks won't care much about what happens during them. Anyway, they no longer are afflicted.

Biogenic amines, mostly tyramine and histamine, have long been considered migraine triggers. In high enough dosages, tyramine and histamine in food make not just migraine sufferers sick, but any human.

Lists of prohibited and allowed foods for migraineurs are plentiful on non-academic websites. Many of these lists are dubious. It's not the specific foods that impact migraine sufferers, but the tyramine and histamine contents. And for any food, these vary from batch to batch, depending on several parameters.

As a drastic example, the histamine content of a tuna sandwich can vary by a factor of 1000 when the same tuna sandwich is bought at the same shop, but a week apart. Maybe their refrigerator broke down, you never know.

For one and the same kind of fruit, the tyramine load can vary by a factor of 100, depending on the condition under which if was harvested and how it was processed. Once the tyramine is in, you can't get it out, unless the whole lot is injected with amine oxidase enzymes that convert an amine group into ammonia, (NH₃). Boiling, waving, or even deep-frying won't remove the tyramine or histamine, but can easily mask a taste that would otherwise betray spoilage.

To be meaningful at all, histamine and tyramine must be quantified. It's NOT that cheese is "prohibited" because it contains tyramine, and fruit and vegetables are "allowed".

Amino acid decarboxylation

The human body uses 20 amino acids to synthesize proteins. 9 are essential, they must come from nutrition. They are: histidine, isoleucine, leucine, lysine, methionine, phenylalanine, threonine, tryptophan, valine. Another 11 may come from food, or are derived by the human organism from essential amino acids. They are: alanine, arginine, asparagine, aspartic acid, cysteine, glutamic acid, glutamine, glycine, proline, serine, tyrosine

Human enzymes or enzymes of bacteria either within the human body or present in food we consume turn amino acids into amines by decarboxylation (the removal of carbondioxide, CO₂, molecules).

The enzymatic derivates of amino acids include:

Histamine (a crucial amine in the immune system) from histidine
Tyramine from tyrosine; several important neurotransmitters, grouped as catecholamines, are also derived from tyramine: dopamine, noradrenaline, adrenaline
Agmatine from arginine
Putrescine, spermidine, spermine from arginine, glutamine, or ornithine
Cadaverine from lysine, glutamine, or ornithine
0-phenylethylamine from phenylethylalanine
Serotonin and tryptamine from tryptophan

Most of these amines have important functions in the human body. Cadaverine, putrescine, spermidine, and spermine, and are needed in cells for protein synthesis and membranes. In human pathology, however, histamine and tyramine are the most relevant. And this article is focused on tyramine.

The following bacteria have been identified as generating tyramine:

Gram-positive: Bacillus thuringiensis. Carnobacterium divergens1, Enterococcus durans, Enterococcus hirae, Enterococcus faecalis, Enterococcus faecium, Lactobacillus brevis, Lactobacillus curvatus, Tetragenococcus halophilus
Gram-negative: Pseudomonas entomophila, Pseudomonas putida, Pseudomonas putida, Gluconacetobacter diazotrophicus, Granulibacter bethesdensis [Source: Tyramine and Phenylethylamine Biosynthesis by Food Bacteria ]

Tyramine

The most comprehensive list of tyramine measurements was compiled by Gaby Andersen, Patrick Marcinek, Nicole Sulzinger, Peter Schieberle, Dietmar Krautwurst and published in February 2019 in the journal Nutrition Reviews [Food sources and biomolecular targets of tyramine ]. Be prepared for some surprises.

Dairy (except cheese)

Buttermilk, 2.2 mg/kg, Souci et al (2016)

Cow’s milk, none detected, Novella-Rodriguez et al (2000)

Cream, 1.7 mg/kg, Souci et al (2016)

Sour cream, 1.4 mg/kg, Souci et al (2016)

Quark, 2.4 mg/kg, Souci et al (2016)

Yogurt, 1.3 mg/kg, Souci et al (2016)
None detected, Mayr & Schieberle (2012)
None detected, Novella-Rodriguez et al (2000)

Out of three tests, two were negative, and the one test that detected tyramine at all only found a very small amount. But mind you: spoiled yogurt can be loaded with tyramine up to dangerous levels. Lactobacillus varieties, responsible for yogurt fermentation, are the champions when it comes to the conversion of tyrosine to tyramine.

Cheese

A wide spectrum of measurements was reported for cheeses. For brie, it ranged from none detected to 260.0 mg/kg in tests conducted by one team, for parmesan 4.0 mg/kg to 290.0 mg/kg, for gauda 20.0 mg/kg to 670.0 mg/kg, for roquefort 27.0 mg/kg to 1100.0 mg/kg. For edam, it was 13.5 mg/kg to 310.0 mg/kg by three teams. For no food item are the measurements as confusing, and unequal, as they are for cheese. Different brands or batches of the same kind of cheese can have more than 50 times as much tyramine as other brands or batches, and you can't know why. It could be a flaw in the culture, or the aging process, or an event during transportation.

Appenzeller, 55.0 mg/kg, Souci et al (2016)

Brie, none detected to 260.0 mg/kg, Souci et al (2016)

Camembert, 37.0 mg/kg, Souci et al (2016)

Cheddar, 350.0 mg/kg, Souci et al (2016)
130.0 mg/kg, Tarjan & Janossy (1978)

Edam, 310.0 mg/kg, Souci et al (2016)
13.5 mg/kg, Lange et al (2002)
25.6 mg/kg, Tarjan & Janossy (1978)

Emmental, 42.0 mg/kg, Souci et al (2016)
128.7 mg/kg, Tarjan & Janossy (1978)

Feta, 152.0 mg/kg to 246.0 mg/kg, Valsamaki et al (2000)

Gorgonzola, 8.0 mg/kg, Lange et al (2002)

Gouda, 20.0 mg/kg to 670.0 mg/kg, Souci et al (2016)

Gruyère, 37.0 mg/kg, Souci et al (2016)

Leerdammer, none detected, Mayr & Schieberle (2012)

Parmesan, 4.0 mg/kg to 290.0 mg/kg, Souci et al (2016)
3.75 mg/kg, Mayr & Schieberle (2012)

Roquefort, 27.0 mg/kg to 1100.0 mg/kg, Souci et al (2016)
152.0 mg/kg, Lange et al (2002)

Tilsit, 32.0 mg/kg, Souci et al (2016)

Meat

Chicken liver, 100.0 mg/kg, Souci et al (2016)
50.0 mg/kg, Tarjan & Janossy (1978)

Ham, cooked, 6.0 mg/kg to 108.0 mg/kg, Saccani et al (2005)

Ham, dry cured, 7.5 mg/kg, Lange et al (2002)
4.0 mg/kg to 171.0 mg/kg, Saccani et al (2005)

Onion sausage, 32.0 mg/kg, Lange et al (2002)

Ox liver, 270.0 mg/kg, Souci et al (2016)

Pork, fresh meat, none detected to 56.0 mg/kg, Saccani et al (2005)

Salami, 77.1 mg/kg, Mayr & Schieberle (2012)
17.0 mg/kg, Lange et al (2002)

Fish

Cod, 2.0 mg/kg, Lange et al (2002)

Fermented fish sauce, 276.0 mg/L to 357.0 mg/L, Kirschbaum et al (2000)

Herring, none detected, Lange et al (2002)

Herring, cured, none detected to 3000.0 mg/kg, Souci et al (2016)

Mackerel, 25.8 mg/kg to 27.4 mg/kg, Shakila et al (2001)

Mackerel in brine, none detected, Shakila et al (2001)

Mackerel salt-dried, 398.4 mg/kg to 413.8 mg/kg, Shakila et al (2001)

Salmon, none detected, Lange et al (2002)

Sardines, 16.2 mg/kg to 11.8 mg/kg, Shakila et al (2001)

Sardines in oil, none detected, Shakila et al (2001)

Sardine, salt-dried, 169.5 mg/kg to 178.1 mg/kg, Shakila et al (2001)

Seer fish, 9.4 mg/kg to 10.7 mg/kg, Shakila et al (2001)

Seer fish, salt dried, 154.2 mg/kg to 154.1 mg/kg, Shakila et al (2001)

Shrimp, 8.8 mg/kg to 12.6 mg/kg, Shakila et al (2001)

Shrimp, salt dried, 693.2 mg/kg to 704.7 mg/kg, Shakila et al (2001)

Tuna, 0.06 mg/kg, Mayr & Schieberle (2012)

Tuna in oil, 0.72 mg/kg, Souci et al (2016)
None detected to 1.2 mg/kg, Shakila et al (2001)

Alcoholic beverages

Beer, Pilsner lager, 1.4 mg/L, Souci et al (2016)
6.85 mg/L, Kalac et al (1997)

Beer, German Vollbier, 1.8 mg/L to 12.0 mg/L, Souci et al (2016)

Beer, alcohol-free, 1.2 mg/L, Souci et al (2016)
6.16 mg/L, Kalac et al (1997)

Wine, port, 0.51 mg/L, Cunha et al (2011)

Wine, red, none detected to 20.0 mg/L, Souci et al (2016)
1.93 mg/L, Mayr & Schieberle (2012)
0.8 mg/L to 2.6 mg/L, Landete et al (2005)
38.8 mg/L, Tarjan & Janossy (1978)
1.40 mg/L, Marcobal et al (2005)
None detected to 0.292 mg/L, Anli et al (2004)
3.1 mg/L, Lüthy & Schlatter (1983)

Wine, white, none detected to 3.0, Souci et al (2016)
110.8 mg/L, Tarjan & Janossy (1978)
1.2 mg/L to 22.7 mg/L, Lüthy & Schlatter (1983)

Fruit

Apple, none detected, Tarjan & Janossy (1978)

Apple juice, 0.1 mg/L, Maxa & Brandes (1993)

Avocado, 23.0 mg/kg, Souci et al (2016)

Banana, 7.0 mg/kg, Souci et al (2016)
0.9 mg/kg, Lavizzari et al (2006)

Currants, none detected, Tarjan & Janossy (1978)

Currant juice, freshly squeezed, 3.26 mg/L, Maxa & Brandes (1993)

Grape, 691.0 mg/kg, Tarjan & Janossy (1978)

Grape juice, 0.04 mg/L, Cunha et al (2011)
0.1 mg/L, Maxa & Brandes (1993)

Grapefruit juice, freshly squeezed, 0.1 mg/L, Maxa & Brandes (1993)

Hazelnut, 1.8 mg/kg, Lavizzari et al (2006)

Orange, 10.0 mg/kg, Souci et al (2016)

Orange juice, 0.21 mg/L, Maxa & Brandes (1993)

Orange juice, freshly squeezed, 0.1 mg/L to 0.49 mg/L, Maxa & Brandes (1993)

Peach, none detected, Tarjan & Janossy (1978)

Pear, none detected, Tarjan & Janossy (1978)

Plum, none detected to 6.0 mg/kg, Souci et al (2016)
None detected, Tarjan & Janossy (1978)

Raspberry, 10.0 mg/kg to 90.0 mg/kg, Souci et al (2016)

Raspberry juice, freshly squeezed, 66.66 mg/L, Maxa & Brandes (1993)

Watermelon, 460.0 mg/kg, Tarjan & Janossy (1978)

Vegetables

Beetroot, 160.0 mg/kg, Tarjan & Janossy (1978)

Cabbage, 670.0 mg/kg, Tarjan & Janossy (1978)

Chinese cabbage, 1.26 mg/kg, Simon-Sarkadi & Holzapfel (1994)

Cabbage, sauerkraut, 20.0 mg/kg, Souci et al (2016)
60.66 mg/kg, Mayr & Schieberle (2012)
6.0 mg/kg, Lange et al (2002)

Cabbage, fermented juice, 37.1 mg/L to 73.0 mg/L, Kirschbaum et al (2000)

Kohlrabi, 930.0 mg/kg, Tarjan & Janossy (1978)

Carrot, 0.001 mg/kg, Sulzinger et al (2016)
119.0 mg/kg, Tarjan & Janossy (1978)

Carrot juice, 0.002 mg/L, Sulzinger et al. (2016)

Cauliflower, 400.0 mg/kg, Tarjan & Janossy (1978)

Cucumber, 250.0 mg/kg, Tarjan & Janossy (1978)

Endive, 1.60 mg/kg, Simon-Sarkadi & Holzapfel (1994)

Green pea, frozen, 8.7 mg/kg, Kalac et al (2002)

Haricot bean, 160.0 mg/kg, Tarjan & Janossy (1978)

Iceberg lettuce, 0.94 mg/kg, Simon-Sarkadi & Holzapfel (1994)

Miso, 24.6 mg/kg to 349.0 mg/kg, Kirschbaum et al (2000)
none detected to 49.8 mg/kg, Yen (1986)

Olives, none detected, Lange et al (2002)

Paprika, 266.0 mg/kg, Tarjan & Janossy (1978)

Potato, 1.14 mg/kg, Sulzinger et al. (2016)
840.0 mg/kg, Tarjan & Janossy (1978)
2.0 mg/kg, Lavizzari et al (2006) French fried potatoes, baked, 1.77 mg/kg, Sulzinger et al. (2016)

French fried potatoes, raw, 0.77 mg/kg, Sulzinger et al. (2016)

Radicchio, 2.73 mg/kg, Simon-Sarkadi & Holzapfel (1994)

Radish, 200.0 mg/kg, Tarjan & Janossy (1978)

Soy sauce, 17.7 mg/L to 172.0 mg/L, Kirschbaum et al (2000)
16.1 mg/L to 1699.0 mg/L, Yen (1986)

Soybeans, 9.05 mg/kg, Bartkiene et al (2015)
none detected, Gloria et al (2005)

Soybean, fermented, 27.8 mg/kg to 416.1 mg/kg, Bartkiene et al (2015)

Sufu, fermented tofu, none detected to 1125.4 mg/kg, Yen (1986)

Spinach, 3.78 mg/kg, Sulzinger et al. (2016)
286.0 mg/kg, Tarjan & Janossy (1978)
2.2 mg/kg, Lavizzari et al (2006)

Spinach purée, frozen, 10.2 mg/kg, Kalac et al (2002)

Spinach, boiled, 2.58 mg/kg, Sulzinger et al. (2016)

Tomato, 4.0 mg/kg, Souci et al (2016)
1.07 mg/kg, Sulzinger et al. (2016)
4.0 mg/kg, Lange et al (2002)
250.0 mg/kg, Tarjan & Janossy (1978)

Tomato ketchup, 33.6 mg/kg, Kalac et al (2002)

Tomato, concentrated purée, 7.23 mg/kg, Sulzinger et al. (2016)
10.4 mg/kg, Kalac et al (2002)

Turnip, none detected, Tarjan & Janossy (1978)

Zucchini, 0.06 mg/kg, Sulzinger et al. (2016)

Others

Chocolate, 3.11 mg/kg, Mayr & Schieberle (2012)
0.3 mg/kg, Lavizzari et al (2006)

Coffee, ground, 1.26 mg/kg to 16.14 mg/kg, Restuccia et al (2015)

Coffee, brewed, 0.25 mg/L to 1.89 mg/L, Restuccia et al (2015)

Privacy policy