Ethnicity
Genetically inherited variation in the efficiency of caffeine-catabolizing enzymes may account for differences in migraine prevalence between different ethnic groups.

Aging
The age-related slowdown in caffeine-catabolizing efficiency71 may partly explain the age-related decrease in primary headache prevalence.

Cluster headache
The typical cluster patient is a man who smokes150 and uses a lot of caffeine. Cluster headaches, briefer and more intense than typical migraine headaches, may result from an interaction between a genetic predisposition and (typically) male gender, cigarette smoking, and heavy caffeine intake. Both smoking and caffeine induce caffeine catabolism, and rapid caffeine catabolism induced by smoking and heavy caffeine intake might cause sudden onset of severe caffeine withdrawal headache. That may explain the sudden and intense headache characteristic of cluster headache. Part of the reason few cluster patients are female may be that fewer women are smokers.

Typical locations of migraine pain, sensory disturbances
Why does primary headache hurt in the head? Why not in the foot, or some other part of the body? That may seem like a stupid question; after all, if it hurt in the foot, it would be a foot ache, not a headache. Semantics aside, though, there must be something special about the head that makes migraine hurt there rather than elsewhere.

The head contains the brain, of course, and the prevailing view assumes headache originates within the brain. But the head also contains an array of major sensory apparatus: the retinas, cochleae, vestibular apparatus, semicircular canals, olfactory bulbs, ventral surface of the tongue, and tooth pulp, all densely packed with sensory neurons. During a migraine episode, those numerous sensory neurons emit adenosine, which travels only a short distance before being dismantled or taken back up into cells. In experiments, infused radiolabeled (tagged with radioactivity) adenosine congregates in the sensory apparatus of the head.137 Perhaps migraine originates mainly in the major sensory apparatus of the head. That would readily explain the typical locations of migraine pain as well as the sensory disturbances associated with migraine. For example: the most typical location of migraine pain is behind and around the eyes,138 which is the location of the retinas. Visual aura may originate in the retinas rather than in the visual cortex of the brain. A vestibular (the vestibular apparatus senses balance and motion) disturbance is likely the cause of nausea/vomiting associated with primary headache. The emotional effects of migraine may originate in the olfactory system outside the brain.

This idea is not such a radical departure from the prevailing assumption that migraine originates within the brain, as the major sensory apparatus of the head are, in a real sense, extensions of the brain; the retinas and olfactory bulbs literally extend from the brain on stalks.

Cortical spreading depression, if it occurs during migraine episodes, may be a consequence of migraine rather than the origin of migraine aura. Or a third, underlying factor may cause both cortical spreading depression and migraine symptoms including aura.

Duration of migraine episodes
The stability of receptor-bound adenosine may explain the typically hours-long duration of migraine episodes and caffeine withdrawal headache episodes. Unlike free extracellular adenosine, which is normally dismantled or taken back up within cells within seconds, receptor-bound adenosine is quite stable: the half-life of bound adenosine ex vivo at 24° C (75.2° F) is more than three hours.72

Spontaneous resolution of migraine episodes
Untreated migraine episodes spontaneously resolve themselves. So do untreated caffeine withdrawal headache episodes. The mechanism may be the same in both conditions.

In caffeine withdrawal headache, a nervous system deranged by caffeine eventually adapts to the absence of caffeine by adjusting its adenosine neurochemistry through feedback mechanisms; neurochemistry distorted by caffeine returns to normal. The main short-term adjustment may involve decreased adenosine output.

Other adjustments to a shortage of caffeine may include increased rate of extracellular adenosine reuptake or destruction, decreased adenosine receptor affinity, and decreased adenosine receptor density, along with opposite changes in excitatory neurotransmitter output, reuptake, destruction, receptor affinity, receptor density, or other changes.

One cause or many: the trigger theory
Migraine is irregularly episodic, so there needs to be some explanation for why a particular migraine episode occurs at a particular time. The trigger theory supposes that exposure to various environmental and variable internal factors triggers individual migraine episodes. Many people report that one or more dietary, physical, hormonal, emotional, or environmental factors precipitate their migraines. The most-often reported triggers include stress, alcohol, foods, too much or too little sleep, and weather.

Physicians, drug companies, and headache self-help books have long promoted the trigger theory, advising patients to try to identify personal headache triggers by looking for associations between their headaches and various suspected trigger factors. Patients are urged to keep a "headache diary" in which to record what they eat and when they get a headache, to look for correlations, and to try to avoid headache by avoiding factors they identify as triggers. Typically this advice is accompanied by a list of notorious trigger factors.

One problem with such advice is that patients encouraged to look for associations between their headaches and various factors are bound to mistake apparent associations for real ones, coincidental associations for causal ones. Like most nonscientists, headache patients tend to be poor observers and even poorer record keepers. Patients tend to make the error of selective observation: they notice positive associations, such as headaches after eating rutabagas, but fail to notice, remember, or record non-positive associations, such as headache but no rutabagas, rutabagas but no headache, and no headache nor rutabagas. Those non-positive associations may not seem as remarkable to the patient, but are just as important as the positive associations for the purpose of distinguishing between real and merely apparent associations due to observational bias.

Even if an association the patient notices and remembers is real, it could be coincidental. And even if an association the patient notices is real and non-coincidental, the patient will often jump to the unsound conclusion that a factor associated with one or more headaches must have caused or precipitated those headaches. The patient will likely fail to consider several other logical possibilities. In some cases, for example, the factor identified as a trigger is a symptom of migraine, rather than a cause of migraine symptoms; it's all to easy to mistake effect for cause. Many patients, for example, believe odors trigger their migraines, but it's the other way around: migraine makes them more sensitive to odors. The same goes for strong emotions associated with migraine episodes: anger or crying73 are associated with migraine but don't trigger migraine; rather, migraine causes emotional effects.

One popular theory is that different people are sensitive to different sets of triggers; that each migraine patient has a personal set of triggers, and patients can decide for themselves what their personal triggers are. Migraine may be the only disease in which patients are encouraged to identify for themselves the cause of the disease in their particular cases. This leads to some absurd beliefs and behavior. I've corresponded with a man who avoids venturing outside because he believes electrical fields emanating from outdoor power lines give him migraine headaches. He prefers to remain safe inside his house, where he had his house wiring reconfigured to minimize electrical fields around his head. I've read of a physician who believes using her mobile phone gives her migraines, and for that reason fought the installation of a mobile telephone antennae mast in her neighborhood. Some people, convinced their metal amalgam dental fillings cause their migraines, have had all their fillings removed, to no avail.

One author tried to impose discipline on this free-for-all by proposing that a given factor should be considered a migraine trigger only if it causes headache in non-migrainers.74

Saying anything and everything can set off a migraine episode is tantamount to admitting you don't know what sets off migraine episodes. Like a messy garage, the trigger theory needs cleaning out. The hundreds of accused triggers need to be sorted into categories. Many factors accused of triggering migraine are innocent, others may be guilty. Suspicion should fall first, however, upon factors demonstrated to cause headache.

Food triggers
A patient might insist rutabagas are a personal migraine trigger. But what if the result of a blinded trial involving a large group of migrainous subjects who believe rutabagas trigger their migraines was that headache followed rutabagas no more often than placebo? The simplest and most likely explanation is that rutabagas don't trigger migraines, and that people who believe they do are mistaken.

Some authors, however, favor a more elaborate explanation: they claim multiple trigger factors are required to trigger a migraine episode. Dr. David Buchholz, author of the popular migraine self-help book Heal your headache, proposes the existence of a "migraine threshold." Or maybe it's the rim of a migraine bucket that various triggers pile up in until the whole mess metaphorically overflows. Buchholz attempts to explain why no food has been demonstrated to trigger migraine:

The recognition of certain dietary items as triggers is based on cumulative experience among headache sufferers. Scientific studies have shed little light on dietary triggers for migraine, in part because it's impossible to isolate individual dietary items for study while maintaining a steady-state background of other, uncontrollable triggers and in part because no single trigger always triggers migraine.

The failure of most of the few scientific studies that have looked at migraine dietary triggers to demonstrate a clear effect is a good example of the limitations of science in identifying truth that is otherwise obvious. Try telling a headache sufferer who has had killer headaches after eating chocolate or drinking red wine that it's all in his or her imagination!75

Buchholz is right when he says science doesn't support the belief of many patients that foods trigger their migraines. He loyally sides with his patients and dismisses the science. But the truth is, if foods triggered migraine, trial studies would have identified food triggers as such. Even if, as Buchholz contends, a trigger doesn't always trigger a migraine in subjects sensitive to that trigger, or only triggered migraine in combination with other contributing factors, a food trigger would nevertheless be associated with migraine episodes more often than could be accounted for by random chance. Trial studies would have revealed such non-coincidental associations.

Once a trigger food had been so identified, researchers could have isolated the specific component of that food that triggers migraine. They could have liquefied the trigger food in a blender, then separated the liquid into various fractions by centrifuging, filtration, distillation, evaporation, or other laboratory techniques. Trials and a process of elimination could have identified the particular fraction, and, eventually, the particular chemical component that triggers migraine. Researchers would be motivated to go to all that trouble because identifying a migraine-causing substance in foods would yield valuable insight into the causal mechanism of food triggers. In short: if foods triggered migraine, researchers could take advantage of that fact to learn about migraine pathogenesis.

So, if Buchholz was correct about foods triggering migraine episodes, by now researchers should have demonstrated that certain foods trigger migraine, should have isolated the specific migraine-triggering substances in those foods, and should by now be well on their way to determining the mechanism by which those substances trigger migraine. None of that has happened. No migraine-triggering component has been isolated from a trigger food, for the simple reason that no whole food has been demonstrated to trigger migraine. There's no point in trying to determine which part of a food triggers migraine if the whole food can't be demonstrated to trigger migraine.

Buchholz mistakes the unsound conclusions of his patients for facts. If a patient concludes that rutabagas trigger migraines, credulous physicians and authors take the patient's conclusion at face value, and add rutabagas to the list of migraine triggers. The trouble with that practice is that evidence-based medicine is based on evidence. A patient's testimony that the patient ate rutabagas then got a migraine is evidence. The patient's conclusion that rutabagas must have triggered the migraine, however, is not evidence, because that conclusion, based on fallacious post hoc reasoning, is unsound, and unsound conclusions aren't evidence. Similarly, if the patient reports quitting rutabagas was followed by a reduction in migraine frequency or severity, that's evidence. But if the patient concludes that quitting rutabagas must have caused the subsequent improvement, that's an unsound conclusion the patient has arrived at, not evidence.

When researchers review the available scientific evidence, they find no suspected trigger food has been demonstrated to trigger migraine, and restricted diets haven't been demonstrated to work. So we're left with unsound conclusions reached by migraine patients, but no good evidence or argument to support the belief that foods trigger migraine, or that avoiding certain foods will help migraine. On any restricted diet, some migraine patients will improve, some will get worse, and some will remain the same. Buchholz and many others misinterpret what this means, mistakenly concluding that what works for some people doesn't work for others because everybody's migraine is different and everybody has different triggers. The truth is, if a given factor isn't associated with migraine episodes on average among groups of subjects in blinded trials, we can't soundly conclude that particular factor triggers migraine. And if eliminating a given factor doesn't on average decrease migraine episodes among groups of subjects in blinded trials, we can't soundly conclude that avoiding that factor helps any individual patient avoid migraines.

Many foods stand accused of being migraine triggers. The following is a list I've compiled of some of the foods accused of triggering migraine headaches:

Pumpkin seeds, sunflower seeds, sesame seeds, lima beans, pinto beans, navy beans, garbanzo beans (chick peas), kidney beans, black beans, red beans, pole beans, green beans, English beans, lentils, peas, pea pods, snow peas, sugar peas, snap peas, black-eyed peas, soy, soy flour, soy sauce, tamari, tofu, miso, tempeh, spinach, peppers, artichoke, asparagus, eggplant, chives, potato, coriander, leek, seaweed, carrot, parsnip, tomato, tomato sauce, tomato paste, apple juice, apple cider, pear, strawberries, berry pie filling, canned berries, banana, apricot, peach, nectarine, coconut, canned figs, dried fruits, bruised fruits, dates, mincemeat, raisins, grapes, grape juice, currants, gooseberries, avocado, guacamole, pineapple, mango, papaya, passion fruit, melon, cherry, maraschino cherries, plum, red plum, kiwi, orange, grapefruit, lemon, lime, persimmon, pomegranate, carob, chicory, chicken, chicken liver, liverwurst, pâté, turkey, beef, roast beef, chipped beef, corned beef, ham, deviled ham, pork, pigs feet, bacon, hot dogs, salami, bologna, summer sausage, head cheese, luncheon meats (cold cuts), smoked meats, Slim Jims, beef jerky, broth, canned soup, instant soup, TV dinners, vinegar, pickles, sauerkraut, olives, marinades, catsup, mustard, teriyaki sauce, Worcestershire sauce, steak sauce, barbecue sauce; Blue, Boursault, Brick, Brie, Camembert, Cheddar, Emmenthaler, Gruyère, Limburger, Mozzarella, Parmesan, Provolone, Romano, Roquefort, Stilton, and Swiss cheeses; macaroni and cheese, Welsh rarebit, fondue, cream cheese, cottage cheese, buttermilk, caseinate, yogurt, sour cream, ice cream, whipped cream, cream pie, pudding, milk, whey, corn, popcorn, oats, rye, wheat, barley, alfalfa, breakfast cereal, bran, malt, almonds, cashews, chestnuts, filberts, hazel nuts, hickory nuts, pecans, pine nuts, walnuts, peanuts, yeast, brewer's yeast, Marmite (yeast extract), Vegemite (yeast extract), shellfish, shrimp, deviled crab, anchovy, herring, cod, plaice, mackerel, smoked eel, sardines, caviar, pizza, white bread, fresh bread, sourdough bread, breadsticks, pretzels, crackers, bagels, coffee cake, macaroons, sandwich cookies, animal crackers, eclairs, brownies, fruitcake, raised doughnuts, olive oil, fat, animal fat, fried foods, butter, herbs, spices, salt, seasoned salt, baking powder, caramels, sugar, ginseng, Chinese food, potato chips, Doritos corn chips.

Here's the article abstract of a review article that sums up the current scientific state of affairs regarding food triggers:

Based on a review of the literature the authors discuss the role of nutrition in the precipitation of migraine and tension-type headache (TTH). The available information relies largely on the subjective assessment of the patients. Controlled trials suggest that alcohol and caffeine withdrawal are the most important nutritional precipitating factors of migraine and TTH. In addition, there is some evidence that missing meals is also an important factor. Dehydration seems to deserve more attention. A selective sensitivity to red wine has been shown in some patients, the importance of chocolate has been doubted seriously, and scientific evidence for cheese as a precipitating factor is lacking. Despite a series of experimental studies demonstrating that NO [nitric oxide] donors such as nitroglycerin and parenteral histamine cause headache, the role of histamine, nitrates, and nitrites in food remains unclear. Similarly, other biogenic amines and aspartame have not been proven to precipitate headache. Sodium glutamate causes adverse reactions including headache probably at large doses ingested on an empty stomach. Therefore, patients should be advised that food plays a limited role as a precipitating factor of migraine and TTH. Subjective sensitivity to certain foods should be examined critically, and proven precipitating factors should be avoided. General dietary restrictions have not been proven to be useful.76

Given the total lack of demonstrated evidence that foods have ever caused or triggered a single migraine, why do so many people believe foods cause or trigger their migraines? The main reason may be that nausea often accompanies migraine. When nausea occurs it's only natural to suspect the culprit must have been something you ate. That's probably why foods make up the bulk of suspected migraine triggers.

Another reason foods are so often scapegoated is a false analogy between food allergies and migraine. It's probably no coincidence many of the foods on my big list contain, correspond to, or are botanically related to the nine foods (milk, eggs, wheat, soy, fish, shellfish, peanuts, tree nuts, and sesame seeds) that account for more than 90 percent of food allergies. Allergy, however, is an immune disorder, migraine isn't, and headache isn't among the symptoms of food allergy.

Major pharmaceutical firms that manufacture and market migraine drugs continue to promote the myth that foods trigger migraine. In their brochures and on their websites, they provide a list of trigger foods, recommend that migraine patients record headaches and everything they eat in a diary, and encourage migraine patients to attempt to identify and avoid personal food triggers. Drug companies endorse the food trigger theory and dietary restriction even though science doesn't back them up. Drug companies aren't altruistic; they are for-profit corporations hardly motivated to spend money distributing advice that would decrease demand for their headache medicines. They know promoting the food trigger theory and dietary restriction won't hurt sales.

People who believe foods trigger their migraines have mistaken a coincidental association for a causal one. They happened to eat a given food prior to the onset of headache and then reasoned post hoc ergo propter hoc (before the event, therefore because of it) that the food must have caused the migraine.

I sort putative dietary triggers into foods, food additives (such as preservatives, artificial colorings, and artificial sweeteners), and psychoactive drugs (alcohol and caffeine added to or naturally occurring in foods).

The following is a closer look at a few of the most-often cited food triggers.

Chocolate: Does chocolate trigger headaches? Taken together, the results of trial studies say no:

• In a 1974 double-blind, controlled study, chocolate delivered to the stomach via a nasal tube, to prevent the subjects from tasting the chocolate and thereby defeating the blind, did not increase headache frequency or severity compared to placebo.77

• In a 1991 trial study using subjects who believed chocolate triggered their migraines, chocolate was followed by a migraine episode in five of 12 subjects. Carob placebo was followed by a migraine episode in none of eight subjects.78 Median time between ingestion of chocolate and onset of a migraine episode was 22 hours.

• In a larger 1997 double-blind study using 63 women subjects with chronic migraine, chronic tension-type headache, or both, chocolate was no more likely than carob placebo to provoke headache or migraine.79

Notwithstanding, chocolate contains plenty of theobromine and some theophylline, which are methylxanthines (caffeine-like molecules), and some sources say chocolate also contains caffeine. So it may be the methylxanthines in chocolate can produce a withdrawal headache. That means it's possible ingestion of chocolate can be followed by a methylxanthine withdrawal headache, if not a caffeine withdrawal headache.

I've experimented on myself by abstaining from caffeine for several weeks, then eating 55 grams of Scharffen Berger 62 percent cacao semisweet dark chocolate on each of two consecutive evenings. I got a mild headache approximately 52 hours after the second dose. I got the headache around bedtime, went to sleep, awoke in the middle of the night with the same headache, went back to sleep, and arose in the morning with the same headache. When I say the headache was mild I mean it was mild compared to severe caffeine withdrawal headaches I'd experienced in the past. The chocolate-related headache was nonetheless unpleasant and accompanied by a general feeling of illness. A small amount (one ounce) of regular coffee aborted the headache episode.

Most people who claim chocolate triggers their migraines, however, likely mean eating chocolate is followed within several hours by headache. Can chocolate provoke or precipitate a headache soon after being eaten? Because caffeine induces its own metabolism, it's theoretically possible a large dose of chocolate could induce a caffeine or methylxanthine withdrawal headache within several hours of ingestion.

Cheese, tyramine: Fermented, aged, or ripened cheese, and even cottage cheese are suspected of triggering migraines. Many migraine patients believe tyramine-containing foods, especially cheese, trigger migraine, but in a double-blind, controlled trial, tyramine delivered into the stomach through a nasal tube did not increase headache compared to placebo.80

The following is a brief discussion of several food additives.

MSG: From a 2006 review of the medical literature regarding MSG:

This article reviews the literature from the past 40 years of research related to monosodium glutamate (MSG) and its ability to trigger a migraine headache... MSG has been described as a trigger for...migraine headache..., but there are no consistent data to support this relationship. Although there have been reports of an MSG-sensitive subset of the population, this has not been demonstrated in placebo-controlled trials. ... Despite a widespread belief that MSG can elicit a headache...there are no consistent clinical data to support this claim. Findings from the literature indicate that there is no consistent evidence to suggest that individuals may be uniquely sensitive to MSG. Nurse practitioners should therefore...seek more consistently documented etiologies for symptoms such as headache...

In the absence of clinical data, it is premature to make any conclusions about MSG as a potential trigger for migraine headaches. With no consistent data to suggest that MSG causes any type of headache, much more extensive clinical research would be required to establish a link between MSG and migraine headaches.152

Caffeine and caffeine withdrawal headache weren't controlled for during the above-mentioned MSG trial studies, which means it's possible caffeine caused every headache that occurred in those studies.

Aspartame: Many people believe the artificial sweetener aspartame triggers their migraines,84 but the few placebo-controlled trial studies of the effect of aspartame on headache have had contradictory results.85,86,87 Aspartame has never been demonstrated to cause or trigger a headache or migraine.77

A theory of migraine pathogenesis needn't explain why foods trigger migraine, because there isn't sufficient reason to believe foods trigger migraine. Having said that, it's possible an indirect interaction between foods and caffeine could promote or precipitate headache. Some foods contain substances that alter the rate at which enzymes catabolize (destructively metabolize) caffeine. In that way, certain foods might influence caffeine withdrawal headache. Certain substances in foods induces enzymes and thereby speed caffeine catabolism, so may promote headache by making caffeine withdrawal more likely and therefore more frequent. Conversely, if a chemical substance in food makes caffeine stay intact in the body longer, that substance may protect against headache by making caffeine withdrawal less likely to occur.

Complicating matters, however, is the two-edged sword thing: caffeine both causes and forestalls caffeine withdrawal headache. Chemical substances that impede caffeine catabolism expose the nervous system to higher concentrations of caffeine for longer periods of time. The nervous system adapts to the increased and extended presence of caffeine by becoming more sensitive to adenosine. So although factors that slow caffeine catabolism may forestall caffeine withdrawal in the short run, they may at the same time help prime the nervous system for more severe withdrawal headaches in the longer run.

Some foods contain naturally-occurring chemicals that may influence caffeine withdrawal headache and thereby influence migraine and other primary headaches:

• Cruciferous vegetables (vegetables in the genus Brassica: cabbage, red cabbage, broccoli, broccoli sprouts, turnip, rape, Brussels sprouts, cauliflower, kohlrabi, chard, kale, bok choy, radish, horseradish, daikon, mizuna, tatsoi, collard greens, mustard greens, rutabaga [!], and watercress) contain indole-3-carbinole, a chemical that accelerates caffeine metabolism,88,89 so may promote caffeine withdrawal and thereby promote migraine.
   
  
• Vegetables in the genus Allium, including onions and garlic, may promote migraine by increasing the concentration of extracellular adenosine.
   
  
• High-protein foods, such as cheese, may accelerate caffeine metabolism91 and thereby promote migraine.
   
  
• Seared meat contains chemicals that increase the rate of caffeine metabolism.92
   
  
• Grape juice. A scientific study in which grape juice was accidentally substituted for grapefruit juice serendipitously revealed that grape juice increases the activity of the enzyme CYP1A2.93 That means grape juice likely increases the rate of caffeine metabolism and may thereby promote caffeine withdrawal and therefore migraine. If so, that might explain why so many people report that wine, more than other alcoholic drinks, seems to promote their migraines.

Some foods may decrease migraine frequency:

• Grapefruit may slow caffeine catabolism94 so may make caffeine withdrawal headache less likely to occur, and thereby protect against primary headache. Grapefruit has not, however, been demonstrated to decrease headache frequency. Naringin had been thought to be the active ingredient in grapefruit that slows caffeine catabolism, but a recent study casts doubt on that.94
   
  
• Apiaceous vegetables such as carrots and celery slow caffeine metabolism,94 so may make caffeine withdrawal less likely and thereby decrease migraine frequency. Apiaceous vegetables have not, however, been demonstrated to decrease migraine frequency.

Fava beans (broad beans) deserve special mention, because fava beans can cause serious illness and symptoms including headache and nausea/vomiting in persons with glucose-6-phosphate dehydrogenase (G6PD) deficiency. This condition is called favism, and it's possible symptoms of favism could be mistaken for migraine. A headache caused by favism, however, isn't a primary or migraine headache.

Stress
Many people blame their headaches on stress. What, exactly, is stress? The word stress, in the context of health, has two different meanings. One is the definition given by Hans Selye, the originator of the stress theory: "The nonspecific response of the body to any demand."95 When most people talk of stress, however, they mean anxiety.

Dr. Selye with subject: "And now you will swim, my little friend, and swim and swim and swim."

First we need to dispose of Dr. Selye's stress theory, which, though highly popular and deeply entrenched, is utter nonsense. Dr. Selye's definition of stress, as he readily admitted, is abstract. "The nonspecific reaction of the body to any demand" doesn't exist as a real entity. Diseases are caused by real things, not by abstract concepts. Existing entities are quantifiable; that is, measurable; an existing entity possesses measurable qualities such as length, mass, location, color, and velocity. Stress has no such quality, because it's an abstract concept, not a real thing. Tellingly, there's no unit of stress, because stress can't be measured or quantified; there's no way to determine that Bob has seven "Selyes" of stress, or that Helmut has twice as much stress as Akbar. Stress isn't an existing entity, so can't cause, precipitate, aggravate, promote, or otherwise influence disease. In short, the stress theory is bunk.

Anxiety, on the other hand, is real, insofar as it consists of tangible neurochemistry. Many physicians and headache patients unsoundly conclude that when anxiety, often called by the misnomer stress, is associated with headaches, the anxiety must have caused or promoted the headaches. They overlook several other logical possibilities: the headache may have caused the anxiety, rather than the other way around, or a third, underlying factor may cause both the headaches and the associated anxiety. Caffeine, a drug that causes both headache and anxiety, could be the underlying factor linking the two.

Curiously, both stress and lack of stress are said to cause or trigger headaches including migraine. Migraine episodes are more likely to occur on Saturdays than on other days of the week,96 and many people report getting migraines after the big project at work is completed. When the pressure is finally off and the late nights end, that's when "stress letdown" headaches and migraines happen.

The idea that both stress and lack of stress can cause headaches is like "heads I win, tails you lose;" if I asserted that both bunny rabbits and lack of bunny rabbits cause migraine, there's no test that could prove me wrong. If bunny rabbits are present, I can blame the migraine on the bunny rabbits. If bunny rabbits are absent, I can blame the migraine on the absence of bunny rabbits.

Nevertheless, caffeine appears to readily explain why headache is associated with both the presence and absence of stress. What is popularly referred to as stress may consist largely of caffeine-caused anxiety. It's not surprising that stress is associated with work, as work is associated with caffeine consumption. Caffeine is used to stay awake and alert while working, especially when doing sedentary work. Work-related caffeine intake may cause work-related anxiety, often called stress, as well as headaches associated with work, and, therefore, headache associated with work-related anxiety.

Caffeine may also cause "stress letdown" headaches: headaches associated with a lack of stress. Lack of stress means lack of anxiety, and for many people a state of relaxation reflects a respite from caffeine-caused anxiety thanks to a decrease or delay in caffeine intake. But that same relaxing decrease or delay in caffeine intake also makes caffeine withdrawal headache more likely to occur. Caffeine withdrawal appears to readily explain why more migraine episodes occur on Saturdays than on other days of the week:96 weekends mean a break from the office coffee pot. Conversely, migraine is least likely to happen on Mondays and Tuesdays,96 after the return to the office coffee pot. Other factors associated with weekends and relaxation, including alcohol, exposure to bright sunlight, and sleeping in late, also promote headache. But there's no demonstrated evidence those factors are associated with primary headache absent caffeine use. For that matter, there's no demonstrated example of primary headache absent caffeine use.

Alcohol
In the movie The Philadelphia Story, the character Tracy Lord, played by Katherine Hepburn, drinks too much champagne one night. The next morning she suffers a classic morning hangover, including acute discomfort at sunlight. Her uncle Willie, who also overindulged, winces at noise and jostling.

How does alcohol—ethanol and other alcohols in alcoholic drinks—cause a hangover headache? A popular explanation is that alcohol causes dehydration, and the dehydration causes the headache. That explanation assumes dehydration causes dry mouth accompanying hangover. Dry mouth normally indicates dehydration, but dry mouth can also occur absent dehydration; various drugs and diseases can cause dry mouth absent dehydration. Dry mouth can, for example, be a migraine symptom. During a migraine episode the functioning of the sympathetic nervous system, which controls salivation, is impaired.

Migraine and alcohol hangover have much in common. Both are associated with alcohol intake, weekends, mornings, and late sleeping. Both feature headache, nausea/vomiting, photophobia (excessive sensitivity to light), phonophobia (excessive sensitivity to sound), and dysphoria (depressed mood). And both are often relieved by administered caffeine.

It may be alcohol hangover largely amounts to migraine associated with and promoted by acute alcohol intake. At the same time, it may be alcohol hangover is caffeine withdrawal associated with and promoted by acute alcohol intake. It may be hangover, migraine, and caffeine withdrawal are three names for the same condition. Like migraine, alcohol hangover headache has never been demonstrated to occur absent caffeine withdrawal. It may be people who verifiably use no caffeine are immune to hangover headache.

Caffeine works by blocking adenosine, so it must be caffeine relieves hangover by blocking adenosine. Adenosine is thought to cause the stimulating, sedative, and anesthetic effects of ethanol.97 Adenosine seems the likely endogenous cause of migraine symptoms as well.

Red Wine
Many people report that red wine, more than other alcoholic drinks, gives them migraines. That may be because, in addition to alcohol, red wine contains grape juice. Grape juice increases the rate of caffeine metabolism92 and may thereby promote caffeine withdrawal and therefore migraine.

Sleep
Primary headache has a relationship with sleep, though the nature of that relationship remains poorly understood.

• Many people report that lack of sleep gives them headaches.
• Many people report that sleeping in late gives them headaches.
• Headache is most likely to occur in the morning, upon awakening.
• Many people find their best strategy during a migraine is to go to sleep and sleep through it. Many other people, on the other hand, are awakened by headaches, or can't go to sleep during their headaches.
• Behavior during a migraine episode often resembles preparation for sleep: retiring to a darkened room and lying down.

Adenosine is the neurochemical that gives you that sleepy feeling and causes the behavior associated with preparing for sleep. Adenosine also very likely causes the head pain and other symptoms associated with migraine. Caffeine postpones sleep by blocking adenosine, and exposure to caffeine alters the normal functioning of adenosine. Withdrawal from caffeine has been demonstrated to cause headache.

Lack of sleep may be associated with headache because lack of sleep is associated with intake of caffeine, a potent stimulant drug demonstrated to cause headache.

The delay in morning caffeine intake readily explains headache associated with sleeping in late.

Odors
People who blame their headaches on odors mistake a migraine symptom for a migraine trigger. Migraine often causes amplified or distorted sensitivity to odors, olfactory illusions (such as one thing smelling like something else), and olfactory hallucinations.98

Weather
Many people believe weather influences their migraines. Those who believe so may be mistaken, for several reasons. For one, most people are poor observers and rarely keep complete and accurate records. They tend to make the error of selective observation: when a migraine coincides with a thunderstorm, they notice and remember that, but fail to notice or remember negative associations, such as thunderstorm with no migraine, migraine but no thunderstorm, or no thunderstorm with no migraine. Though those last three permutations may seem less noteworthy than the first, they are just as important for determining whether a perceived association is real or merely an illusion.

Even if the observed association is real, it could be a coincidental association. Coincidence can't be ruled out in any one specific individual subject. The only way to rule out coincidence is to look at a group of subjects to see whether their migraines, as a group, are associated with weather. Anecdotal evidence can't prove weather influences migraine. Only methodical studies of groups of migraine patients can do so.

The following are summaries of the scientific studies of weather and migraine to date (omitting survey studies in which researchers simply asked subjects if weather gives them headaches):

1. Kugler J, Laub M (1978) found no correlation between headache and atmospheric pressure, temperature, humidity, and ionization in four subjects over a five-year period, but did find a significantly high incidence of headache symptoms in "biometeorologic phase 6Z," whatever that is.99
    
   
2. Wilkinson and Woodrow (1979) found no correlation between headache frequency and adverse weather conditions in London.100
    
   
3. Schulman, et al. (1980) showed no correlation between migraine and barometric pressure in Boston.101
    
   
4. Nursall (1981) showed no correlation between migraine and barometric pressure in southern Ontario, but did show headache frequency increased as temperature and humidity increased.102
    
   
5. Cull (1982) found that a sharp rise in barometric pressure reduced the frequency of migraine attacks in Scotland.103
    
   
6. Diamond, et al. (1990) found no correlation between headache frequency and adverse weather conditions in Chicago.104
    
   
7. De Mattels G, et al. (1994) found no correlation between migraine and humidity or temperature in 40 migraine patients, but did find a correlation between geomagnetic activity and migraine frequency.105
    
   
8. Larmande, et al. (1996) found no correlation between weather and 4,421 migraines in France.96
    
   
9. Piorecky, et al. (1996) found that among 13 subjects in the Canadian Rockies migraines occurred on 17.26 percent of days with Chinook winds compared to 14.65 percent of days without Chinook winds.106
    
   
10. Cooke, et al. (2000) found that among 75 subjects in the Canadian Rockies migraines occurred 1.19 times more often on days with Chinook winds, and 1.24 times more often on the day prior to days with Chinook winds, than on non-Chinook days. For Chinook days, the relative risk of migraine increased only on days with winds in excess of 38 kilometers per hour.107
     
    
11. Vaitl D, et al. (2001) found low frequency sferics, pulse-shaped electromagnetic fields originating from atmospheric discharges (lightning), correlated with migraines and headaches in 37 German women from October through December, but not in July and August, when thunderstorm activity had been very intense.108
     
    
12. Walech H, et al. (2002) found a small correlation between weather and headache among 98 patients in Germany, but only during the summer half of the year.109
     
    
13. Prince PB, et al. (2004) found a statistical correlation in 39 (50.6 percent) of 77 subjects between migraine and one or more of three synthetic combinations of multiple functions derived from actual weather conditions. Oddly, subject beliefs about the kind of weather they are sensitive to usually didn't correspond to the kind of weather the study found them sensitive to.110
     
    
14. Villeneuve PJ, et al. (2006) found no correlation between 4,039 emergency room visits for migraine in Ottawa and weather conditions 24 hours before those visits.111
     
    
15. Mukamal KJ, et al. (2009) examined the records of 7,054 headache patients (2,250 cases of migraine and 4,803 cases of tension or unspecified headache) who came to a single Boston hospital emergency department between May 2000 and December 2007, and found that the ambient temperature during the 24 hours before each visit was on average higher than the averaged temperature of the other same days of the week during that same calender month. Hospital visits were also associated with days that had a lower barometric pressure during the preceding 48 to 72 hours than the averaged barometric pressure preceding the other same days of the week during that same calender month. The study found no association, however, between migraine and low barometric pressure.153

The Prince study methodology seems complicated enough to conceal some shenanigans; most lay readers would find it difficult to judge whether the conclusions of the authors are sound. Instead of looking for associations between migraine episodes and specific weather factors such as temperature, humidity, barometric pressure, wind, rain, snow, fog, cloudiness, thunderstorms, etc., the Prince study looks for associations between migraine episodes and synthetic weather conditions derived from mathematical functions combining multiple weather factors:

There were 43 weather variables including lead-time, lag-time, and delta variables included in the data set. To eliminate collinearity of the weather variables, the first step was to perform factor analysis on the weather data alone. Using the resulting rotated factor matrix, a determination was made as to which weather variables were closely related and had the greatest impact on overall weather trends, and three factors from the original 43 variables were generated. The resulting three complex factors from the weather data analysis were then used for subsequent analysis. This analysis also allowed us to determine those variables that carried the most weight in determining the value of the corresponding factor.110

No statistical method can determine that weather causes migraine in any one particular subject. If a subject gets a migraine whenever there's a thunderstorm, that would be remarkable, but it wouldn't necessarily mean the subject is sensitive to thunderstorms; it could be mere coincidence. Nevertheless, in the Prince study, migraine was associated with certain kinds of weather in too many of the subjects for the associations in all those subjects to be purely coincidental. It seems weather may influence migraine. If weather influences migraine, a theory of migraine pathogenesis would have to explain why.

The authors wrote:

In our study we found that there were some patients sensitive to high temperatures with high humidity, and some sensitive to low temperatures with low humidity.

I think a behavioral explanation more likely: weather may indirectly influence headache by influencing habits of caffeine intake and thereby influencing caffeine withdrawal headache. Caffeine is a demonstrated cause of headache, and consumption of caffeinated beverages varies with the seasons. Coffee sales plummet (fig. 3), and fewer people drink coffee (fig. 4), during warm weather. Demand for soda pop, including caffeinated soda pop, spikes during hot spells. Tea consumption varies with the seasons also.

Fig. 3, right: Relative seasonal variation in coffee contract prices
Fig. 4, right: Seasonal prevalence of coffee drinking

Given the weather-driven variation in consumption of specific caffeine-containing drinks, it seems highly likely caffeine intake varies seasonally in some individuals. Though many people may maintain consistent caffeine intake by substituting caffeinated soda, iced coffee, or iced tea for hot coffee or hot tea during hot weather, and vice-versa during cold weather, not everybody will manage smooth transitions.

The effect of weather on consumption of caffeinated drinks may explain why some subjects are sensitive to cold, dry weather, others to hot, humid weather. Some migraine patients get their caffeine mostly from hot drinks such as hot tea and coffee, others mostly from cold drinks such as soda pop and iced tea. Those who get their caffeine mostly from hot drinks may tend to reduce their caffeine intake during hot weather, while those who get their caffeine mostly from cold drinks may tend to reduce their caffeine intake during cold weather. Reduction in caffeine intake tends to promote caffeine withdrawal headache. The Prince study neglects to control for caffeine withdrawal headache, so may have inadvertently counted severe caffeine withdrawal headaches as migraines. Their results may reflect the effect of weather on caffeine withdrawal headache rather than on migraine.

The caffeine explanation has several advantages over the sferics theory. For one, caffeine has been demonstrated to cause headaches, sferics have not. For another, caffeine causes headache by a fairly well-understood mechanism of addiction and withdrawal. The sferics theory, in contrast, invokes an unknown causal mechanism. Additionally, on a practical level, headache patients can do something about caffeine headaches, but there doesn't seem to be much utility in the idea that sferics may somehow influence migraine. What can a headache patient do about electromagnetic pulses from lightning? Wear an aluminum foil hat? Move to a region with a different climate? I haven't seen any evidence moving helps migraine. California has mild weather year round, yet Californians seem to get their fair share of migraines.

The authors of the Prince study suggest patients could consult weather forecasts and thereby anticipate their migraines. That would enable patients to cancel activities for fear of weather-related migraine episodes. Or, the authors suggest, patients forewarned of "headache weather" can attempt to avoid other, compounding trigger factors and thereby avoid inducing a weather-related migraine, a strategy premised on the dubious migraine threshold theory.

Most disturbing is Dr. Prince's suggestion that migraine patients can head off anticipated weather-related headaches by using migraine abortant drugs prophylactically (to prevent migraines), a practice that would inevitably result in patients taking abortant medicines in a futile attempt to prevent migraines that weren't going to happen anyway. Such a wasteful practice would benefit sellers of migraine drugs, but wouldn't be such a good deal for their customers. Besides: according to the Prince authors, most patients whose headaches correlate with a certain type of weather are mistaken about the type of weather their headaches correlate with, so a patient's attempt to predict headaches by consulting weather forecasts will probably be misdirected.

Caffeine withdrawal as migraine trigger
A leading headache researcher told me he believes caffeine is a migraine trigger but cannot cause migraine. I asked him how he would distinguish between a migraine triggered by caffeine, a migraine caused by caffeine, and a caffeine withdrawal headache that resembles migraine. I asked him that question three times, but he never answered it. His silence on this point is understandable, as there is no known way to distinguish between those three possibilities. He isn't alone in his confusion on this point. A different author, Buchholz, writes:

Letdown from the stress of the workweek can be a trigger... But weekend headaches also arise from oversleep on weekends, excessive alcohol intake on Friday and Saturday nights, and caffeine withdrawal (from drinking less coffee on weekends than on weekdays). All these triggers can be addressed.75

From a Medscape (National Libraries of Medicine) web page:

People with migraines may inherit the tendency to be affected by certain migraine triggers, such as fatigue, bright lights, and weather changes. Additional possible triggers include:

• Emotional stress
• Sensitivity to specific chemicals and preservatives in food. ...
• Caffeine. Excessive caffeine consumption or withdrawal from caffeine can cause headaches when the caffeine level abruptly drops. The blood vessels seem to become sensitized to caffeine. When caffeine is not ingested, a headache may occur. Caffeine itself is often helpful in treating acute migraine attacks.112

Like many who write about the relationship between migraine and caffeine withdrawal headache, this author assumes the two are separate and distinct from one another, yet muddles the two.

From a U.S. Food and Drug Administration website:

A change in caffeine intake, either up or down, can also trigger migraine.113

University of Michigan neurologist Dr. Linda Selwa told an interviewer:

"Many of my patients have told me stories about needing caffeine when they feel a headache coming on, or having headaches on the weekends when they haven't had their usual workday amount of caffeine," says Selwa.

"In fact, caffeine is a key ingredient in almost all of the over-the-counter migraine medications. That's because caffeine is useful in stopping a headache once it starts," she continues. "The unfortunate thing is that, in patients who use caffeine chronically, they're much more likely to get a migraine as the caffeine begins to wear off."114

Dr. Seymour Diamond writes:

...caffeine withdrawal...can trigger a severe headache.115

Prominent headache researcher Dr. Richard Lipton says:

Caffeine withdrawal can be a powerful trigger of migraine and caffeine withdrawal is a major cause of weekend headache in people who sleep through their morning cup of coffee. What I recommend to patients is that they limit their use of caffeine on most days to just one cup of coffee a day, but that they use caffeine, if it's effective for them, as a headache comes on, either in combination with a painkiller or in combination with whatever they're taking to relieve their headache.139

Like Lois Lane, who notices the association between Clark Kent and Superman but can't figure out what the relationship is, headache researchers seem to have an inductive blind spot. They need to clarify their view of the interaction between caffeine withdrawal and migraine.

Caffeine as migraine trigger
Caffeine induces its own metabolism by inducing the CYP1A2 enzyme.116,117,118 That may at least partly explain why headache frequency correlates to caffeine intake: the more caffeine a patient uses, the shorter the delay between the latest dose of caffeine and onset of caffeine withdrawal symptoms becomes, so the more frequent the patient's headaches will tend to be. This mechanism appears to readily explain development of chronic headache, including medicine overuse headache, and could also explain why caffeine was found to be a risk factor for chronic daily headache.6

But can a dose of caffeine trigger a migraine episode? A dose of caffeine can be closely followed by a headache. Sometimes that may be simply coincidence: a caffeine withdrawal headache may have been developing at the time the dose of caffeine was taken, and the dose may have come too late to halt and reverse the developing caffeine withdrawal headache.

It may also be possible that a large dose of caffeine—a bigger dose than a person is accustomed to—can provoke a caffeine withdrawal headache when none had been developing. If so, that would make caffeine seem a headache trigger. It's another caffeine paradox: too little caffeine can cause headache, but so can too much. If adenosine causes headache and associated symptoms, one might expect a dose of caffeine, which blocks adenosine, should prevent or at least delay headache, not precipitate it.

There are several possible mechanisms by which a big dose of caffeine could cause a headache.

One possibility is that a large dose of caffeine produces a spike in body caffeine concentration, which induces rapid caffeine catabolism, causing a subsequent rapid decrease in body caffeine concentration. Thus a large dose of caffeine may compress the delay between ingestion of caffeine and onset of caffeine withdrawal symptoms. The result might be a caffeine withdrawal headache closely following acute caffeine ingestion.

Another possibility is that large doses of caffeine have the opposite effect of small doses. The "biphasic" effect of caffeine on movement is seen in experiments with rats: small doses of caffeine increase movement, but large doses decrease movement.146,147 The reason for this biphasic effect is that there are different types of adenosine receptors, named A1, A2a, A2b, and A3. Types A1 and A2a have different and sometimes opposite effects when activated. The various types also have different affinities for adenosine, so that activation of each type depends on adenosine concentration. Adenosine receptors interact with each other by increasing or decreasing the concentration of adenosine outside cells. A2a receptor activation desensitizes A1 receptors.149 It may be small doses of caffeine prevent or reverse headache by blocking type A2a adenosine receptors but large doses promote headache by blocking type A1 adenosine receptors.

Headache medicines
Use of caffeine-containing headache medicines increases the patient's overall caffeine intake, and an increase in caffeine intake induces caffeine metabolism, thereby shortening the delay between caffeine intake and onset of caffeine withdrawal symptoms. The shortened grace period makes caffeine withdrawal more likely to occur during any given period of time, so withdrawal headaches tend to become more frequent. The patient treats increasingly frequent headaches with increasingly frequent doses of caffeine-containing headache medicines. This vicious spiral often leads to daily or near-daily headaches.

Some chronic headache is associated with headache medicines that don't contain caffeine, such as simple (single-ingredient) analgesics and triptans. Neither aspirin by itself nor acetaminophen by itself is addictive, and neither has been demonstrated to cause withdrawal symptoms or headaches, yet simple analgesics are associated with development of medicine overuse headache, which is clearly a withdrawal headache. This association could be due to an indirect interaction between those medicines and dietary caffeine: aspirin and acetaminophen inhibit CYP1A2-mediated caffeine catabolism. When aspirin or acetaminophen exit the body, caffeine catabolism may speed up, which would make caffeine withdrawal, and therefore headache, more likely to occur.

Triptan medicines do not contain caffeine, yet overuse of triptans is associated with chronic headache, and in such cases stopping triptan use usually results in fewer headaches. It may be triptans alter the nervous system by inducing nervous adaptation to triptans, and such adaptation leaves a person vulnerable to frequent headaches. Nevertheless, an association between triptans and chronic headache hasn't been demonstrated absent caffeine use. It may be people who don't use caffeine don't get migraines so don't use triptans.

Smoking
Smoking is associated with primary headache, especially cluster headache. Smoking accelerates caffeine catabolism by inducing the enzyme CYP1A2.119 Smoking may promote primary headache by shortening the delay between caffeine intake and onset of withdrawal symptoms, thereby making caffeine withdrawal headache more likely.

Rhythmic stimuli
Many people report rhythmic stimuli, such as flickering lights, computer screens, drumming, light filtered through Venetian blinds, boldly-patterned wallpaper, etc., can trigger, promote, or aggravate their migraine episodes.

Regular or rhythmic sensory stimulation provokes rhythmic nerve firing and thereby tends to induce convulsions. The nervous system responds to regular or rhythmic stimuli by releasing adenosine, which inhibits nerve firing and thereby suppresses convulsions. In an individual with an abnormal nervous system oversensitive to adenosine, adenosine released in response to rhythmic stimuli may promote or aggravate headache or migraine.

Bright light, exertion
Migraine patients have an abnormal neurochemical reaction to light and to exertion. In normal people, light doesn't change norepinephrine output, exertion increases norepinephrine output. In migrainers, light and exertion decrease norepinephrine output.120

Light striking neurons in the retina causes them to release adenosine triphosphate (ATP),121 which only adds to the problem of excessive adenosine during a migraine attack. ATP is thought to initiate pain by binding to purinoreceptors located on sensory neurons.3 Once released from cells, ATP is rapidly converted into adenosine, which also causes pain. Light-evoked adenosine and ATP release may explain photophobia associated with migraine.

Eyestrain
It's often assumed eyestrain causes headaches associated with eyestrain, but it may be the other way around; it may be eyestrain is a migraine symptom, rather than a headache cause or trigger. Excessive adenosine receptor activation due to caffeine withdrawal may be the underlying cause of both eyestrain and headache.

The retinas, at the back of the eyeballs, are densely packed with sensory neurons. Adenosine emitted from retinal neurons may cause the pain of eyestrain, just as it may cause migraine pain behind and around the eyes, the most common location of migraine pain. Adenosine emitted from the retinas may also cause the sensation of muscle tension or stiffness by causing stretch-, length-, or tension-sensing neurons in sensory fibers embedded within eye muscles or tendons to malfunction and inaccurately report the condition of those muscles.

Neck pain or stiffness
Migraine is sometimes associated with pain or stiffness in the neck, and many people assume the neck pain or stiffness causes the associated migraine episodes. More likely it's the other way around: the migraine causes the neck pain or stiffness. Neck pain or stiffness associated with headache may be due to adenosine, emitted from neurons in the major sensory apparatus of the head, diffusing down into the neck. A stiff neck associated with migraine may feel stiff because excessive adenosine has caused stretch-sensing neurons in sensory fibers within neck muscles to malfunction and inaccurately report the degree to which neck muscles are contracted.

Normally, extracellular adenosine is dismantled within seconds, but adenosine concentration in the head and neck is elevated during migraine episodes.1 It may be that, during a migraine episode, neurons in the major sensory apparatus of the head release so much adenosine that the molecular mechanisms that dismantle adenosine or transport it back into cells are overwhelmed, and adenosine may be able to drift far from its origins. In extreme cases, adenosine might survive intact long enough to reach the neck.

Hypoglycemia, skipping meals
An association between missing a meal and getting a headache may be simply a coincidental association, unless the missed meal means missed caffeine, which could result in caffeine withdrawal headache.

Cause and effect may be the other way around: rather than a missed meal causing a migraine, it may be migraine causes loss of appetite that leads to a missed meal.

Many people believe their headaches or migraines are caused or triggered by hypoglycemia, or low blood sugar, also known as insulin shock. Though hypoglycemia can cause shaking and headache, it's quite rare in non-diabetics. Fasting for 24 hours or longer followed by moderate or high intake of alcohol can lead to hypoglycemia, and thereby to headache, especially in young children, the elderly, or very small adults.

Symptoms of hypoglycemia are quickly relieved by ingestion of sugar. If drinking a sugar solution or fruit juice doesn't relieve a headache within ten or 15 minutes, hypoglycemia didn't cause the headache. Caffeinated beverages shouldn't be used for this test, since that would confound the result: there would be no way to know whether it was the sugar or the caffeine that relieved the headache.

Persons who discover that drinking caffeinated soda pop relieves their headaches can mistakenly assume the sugar rather than the caffeine in the soda does the trick, and so mistakenly conclude that low blood sugar causes their headaches.

Caffeine impairs insulin action and slows glucose disposal,122 so that's another possible mechanism whereby caffeine could interact with food, blood sugar, and diabetes.

Associations between primary headaches and other diseases
Caffeine appears to readily explain associations between migraine and certain other diseases and conditions.

Epilepsy:	Caffeine, which both aggravates epilepsy and causes severe headache and nausea/vomiting, may be the factor responsible for the association between epilepsy and migraine. Epileptics are twice as likely as the average person to have migraine: 24 percent of epileptics have migraine, compared to 12.5 percent of the general population.123 (Because migraine is quite common and epilepsy relatively rare, the association is much weaker in reverse; migrainers are only slightly more likely than the average person to have epilepsy.) Caffeine, which can cause migraine-like headaches, aggravates epilepsy. An overdose of caffeine can cause epilepsy-like symptoms; namely, convulsions. Intravenous caffeine has precipitated epileptic seizure.124 50 percent of epileptics suffer headache following a seizure.125 One possible explanation for headaches following epileptic seizures is that they are caused by large amounts of adenosine released from neurons in an attempt to arrest the seizure. Adenosine is an anti-convulsant. The nervous system releases adenosine in order to inhibit excitatory neurotransmitter release and thereby inhibit the rhythmic nerve firing that causes convulsions. That may also explain why headache often follows electroconvulsive therapy: the nervous system releases large amounts of adenosine to stop the convulsions caused by electric shock. Patients are often given caffeine before receiving electroconvulsive therapy because caffeine not only prolongs the convulsions, which are thought to be therapeutic, but also prevents the post-convulsion headache.
Asthma:	Caffeine may be the factor linking asthma and migraine. Inhaled aerosolized adenosine-5'-monophosphate provokes bronchoconstriction in asthmatics.126 Asthma patients have greater-than-normal amounts of adenosine in their blood and breath.127 Caffeine and another methylxanthine, theophylline, are known to cause bronchodilation, improve lung and airway function, and relieve asthma attacks. Asthma, like migraine, is influenced by female sex steroid hormones: many women experience worsening of asthma symptoms in the week prior to menses.128,129 Together these facts suggest caffeine, by provoking adaptive changes in neurochemical mechanisms, may play a causal role in asthma. It stands to reason that if caffeine causes bronchodilation and relieves asthma symptoms, caffeine withdrawal might have the opposite effect. Asthma may be a variation of caffeine withdrawal syndrome.
Depression:	Migraine and depression have a similar prevalence and average age of onset, and both conditions afflict more women than men. Lifetime prevalence of major depression is three times higher in migrainers than in the general population. Caffeine, which alters the nervous system, may account for the strong association between migraine and depression. Dysphoria (depressed mood) is among the symptoms of caffeine withdrawal. Adenosine, which, overall, inhibits the release of dopamine, norepinephrine, and other excitatory neurotransmitters, may be the endogenous causal factor in depression. If so, caffeine may be the exogenous (originating outside the body) causal factor. Certainly caffeine use is common enough to account for the high prevalence of depression. An indirect interaction between estrogen and caffeine may underlie the associations between female gender and both migraine and depression.
Anxiety/panic disorder:	Caffeine, a potent anxiogenic (anxiety-causing) drug that can induce panic attacks and cause severe headache, may be the factor underlying the very strong association between anxiety/panic disorder and migraine.
Obesity, diabetes:	Obesity is associated with diabetes and also with migraine. Estrogen withdrawal promotes headache and migraine. It may be extraovarian production of estrogen and estradiol in adipose tissue (body fat) promotes headache and migraine by influencing the rate of caffeine catabolism and thereby influencing caffeine withdrawal headache. Another possible explanation for the association between obesity and migraine is that obesity is associated with heavy intake of soda pop, consumption of soda pop is associated with caffeine consumption, and caffeine is a demonstrated cause of severe headache and nausea/vomiting. Repeated large doses of sugar in the form of caffeine-containing soda pop may promote the development of diabetes. Coffee intake is associated with a reduced risk of type 2 diabetes,45 but coffee is not the same thing as caffeine; caffeine is a single chemical substance, coffee a complex mixture of hundreds of different chemical substances, some of which may be beneficial to health. Nor is caffeine-containing soda pop, the major source of caffeine for children, equivalent to coffee.
Irritable bowel syndrome:	Coffee, known to induce symptoms of irritable bowel syndrome, contains caffeine, a demonstrated cause of severe headache, so coffee may be the factor underlying the association between irritable bowel syndrome and migraine.
Restless legs syndrome:	One study found 17.3 percent of migraine patients had restless legs syndrome, compared to 5.6 percent of controls.151 Caffeine, known to cause headaches, alter the nervous system, and induce muscle twitching, may underlie the association between migraine and restless legs syndrome. Like migraine, restless legs syndrome hasn't been demonstrated to occur absent caffeine use.