ISSN 0017-8748 doi: 10.1111/head.12725 VC 2015 American Headache Society Published by Wiley Periodicals, Inc. Review Articles

Migraine Triggers and : A Narrative Review and Synthesis

Jonathan M. Borkum, PhD

Background.—Blau theorized that triggers are exposures that in higher amounts would damage the . The recent discovery that the TRPA1 ion channel transduces oxidative stress and triggers neurogenic inflammation sug- gests that oxidative stress may be the common denominator underlying migraine triggers. Objective.—The aim of this review is to present and discuss the available literature on the capacity of common migraine triggers to generate oxidative stress in the brain. Methods.—A Medline search was conducted crossing the terms “oxidative stress” and “brain” with “alcohol,” “dehydration,” “water deprivation,” “monosodium glutamate,” “aspartame,” “tyramine,” “phenylethylamine,” “dietary nitrates,” “nitrosamines,” “noise,” “weather,” “air pollutants,” “hypoglycemia,” “hypoxia,” “infection,” “estrogen,” “cir- cadian,” “ deprivation,” “information processing,” “psychosocial stress,” or “nitroglycerin and tolerance.” “Fla- vonoids” was crossed with “prooxidant.” The reference lists of the resulting articles were examined for further relevant studies. The focus was on empirical studies, in vitro and of animals, of individual triggers, indicating whether and/or by what mechanism they can generate oxidative stress. Results.—In all cases except pericranial , common migraine triggers are capable of generating oxidative stress. Depending on the trigger, mechanisms include a high rate of energy production by the mitochondria, toxicity or altered membrane properties of the mitochondria, calcium overload and , neuroinflammation and activation of micro- glia, and activation of neuronal nicotinamide adenine dinucleotide phosphate (NADPH) oxidase. For some triggers, oxi- dants also arise as a byproduct of monoamine oxidase or cytochrome P450 processing, or from uncoupling of nitric oxide synthase. Conclusions.—Oxidative stress is a plausible unifying principle behind the types of migraine triggers encountered in clini- cal practice. The possible implications for prevention and for understanding the nature of the migraine attack are discussed.

Key words: TRPA1, migraine, triggers, oxidative stress,

(Headache 2016;56:12-35)

In clinical experience and patient report, acute odors), dietary (eg, alcohol, nitrates), and pharma- migraine attacks are precipitated by a wide range cological (especially nitroglycerin). On a practical of factors1,2 – behavioral (eg, stress, mental over- level, identifying trigger factors may help in treat- work, irregular sleep), environmental (eg, noise, ment3 by preventing some migraine attacks, impart- ing a sense of control to the patient, and providing advanced warning that allows for early use of acute From the Department of Psychology, University of Maine, medication.4–6 Orono, ME, USA and Health Psych Maine, Waterville, ME, USA. On a theoretical level, triggers raise intriguing questions about the nature of . Clinically, Address all correspondence to J.M. Borkum, Department of Psychology, University of Maine, 301 Little Hall, Orono, ME 04469-5782, USA. [email protected]. Conflicts of Interest: None.

Accepted for publication September 3, 2015. Sources of Financial Support: None.

12 Headache 13 triggers seem to summate, with a migraine resulting of the TRPA1 channel raises the possibility that when a threshold is breached.7,8 In the laboratory, migraine triggers might have in common that they isolated triggers tend to have less effect than expo- induce oxidative stress in the brain, as suggested by sure to several triggers in succession.9 This implies Benemei et al.18 that triggers are converging on a common pathway, Despite this conceptual advance, there has not but the mechanism by which, for example, mental yet been a comprehensive consideration of the effort, hot dogs, diesel fumes, psychosocial stress, types of migraine triggers found in clinical practice and aged cheese interact calls for explanation. in relation to oxidative stress. In what follows, we One approach to the theory of triggers arises will briefly review the main sources of oxidative from the pathophysiology of migraine. Triggers are stress in the brain and then consider how these understood by their capacity to constrict or dilate sources are affected by the various migraine trig- blood vessels, directly irritate peripheral nocicep- gers, drawing on in vitro and animal data. tors, or activate a hyperexcitable cortex, either Oxidative Stress.—By losing an electron or a directly, or by interacting with modulatory seroto- hydrogen atom, a molecule is oxidized and becomes nergic and noradrenergic brainstem nuclei.10–15 capable of oxidizing the molecules it encounters. A second approach, more teleological, was Examples of oxidants are free radicals such as per- given by Blau,16 who noted that migraine triggers oxynitrite (ONOO•), the peroxyl radical (ROO•), tend to be exposures that in higher amounts would the hydroxyl radical (OH•), and the superoxide • 2 damage the brain – hypoxia, hypoglycemia, alcohol, anion ( O2 ). There are also nonradical oxidants, and environmental extremes (eg, of heat or cold) those that do not have charges, such as hydrogen being prime examples. This is consistent with peroxide (H2O2) and hypochlorous acid (HOCl; Loder’s observation that the high prevalence of bleach). migraines implies that they, or the genes that Certain oxidants (eg, H2O2), in controlled underlie them, confer an evolutionary advantage.17 amounts, are physiologic and have important signal- How this would extend to noise or mental effort, ing functions within the .24,25 However, when however, is less clear. Moreover, the physiological oxidants are produced in excess or when the antiox- pathways by which potential harm generates a idant defenses that regulate them are compromised, migraine require explication. the result is oxidative stress, a condition in which As noted by others,18 the recent discovery of biomolecules such as DNA, membrane lipids, the TRPA1 ion channel in nociceptors suggests one , and structural proteins can be damaged possible explanation. In animals, calcitonin gene- through oxidation to an extent that exceeds repair related (CGRP) can be released from dural capacity.26 Oxidative stress is thought to contribute afferents, promoting neurogenic inflammation, pain to a wide range of diseases.25 sensitivity, and behavioral evidence of a migraine, Sources of Oxidants.—There are three broad by agonists of the TRPA1 channel.18 Thus, Dussor endogenous sources of oxidants in the brain: (1) et al note that a number of irritants well known to Mitochondria. Based on animal data, between 0.1% be migraine triggers (eg, formaldehyde19) are in and 4% of electrons ordinarily leak from mitochon- fact activators of the TRPA1 channel.20 drial complexes I, III, and possibly II, generating Particularly important for our purposes is that superoxide anions.27–29 Impediments to the smooth the TRPA1 channel is specifically activated by oxi- flow of electrons through the respiratory chain can dative and nitrosative stress.21 A mechanism for increase their leakage and reaction with oxygen to sensing such conditions would surely have adaptive form superoxide. Thus, damage to the structure of significance, as the brain is uniquely susceptible to mitochondria, accumulated of mitochon- oxidative damage.22 Long-term exposure to oxida- drial DNA, and mitochondrial toxins such as rote- tive stress has been suspected of playing a causal none and cyanide, lead to increased formation of role in a range of brain pathologies.23 The function superoxide.30,31 (2) NADPH oxidase. Oxidative 14 January 2016 stress plays a key role in host defense, particularly rocytes41 and platelets42 in migraine, as has a lower the bactericidal function of white blood cells.32 The activity of glutathione peroxidase in erythrocytes,41 superoxide radical is produced by the NADPH oxi- and lower total capacity and nonoxi- dase (NOX) enzymes and converted into hydrogen dized thiol concentration in serum.43 peroxide in macrophages and hypochlorous acid in Migraine Triggers.—Thus, oxidants arise endoge- neutrophils.32 In the brain, one form of NOX nously in the brain from a number of sources. In , NOX2, is found on , , excess amounts, they are potentially harmful. We and ,33 where it is activated by bacteria, have seen that the brain has ion channels that can toxins, and other threats.32 (3) Other enzymes. For detect oxidative stress, triggering neurogenic infl- certain enzymes, oxidants are a byproduct of the ammation. It is plausible that migraine triggers reactions they catalyze. A key example in the brain have in common a capacity to increase oxidative is monoamine oxidase (MAO), whose role is to stress. In what follows, we will consider this possi- metabolize monoamines such as dopamine, norepi- bility for a range of migraine triggers. nephrine, and serotonin, but which produces hydro- Here, we will use the term “trigger” to mean an gen peroxide in the course of the reaction.34 external or internal exposure that increases the prob- Certain cytochrome P450 enzyme systems that ability of a subsequent migraine attack. Because the metabolize xenobiotics and drugs also produce oxi- association is probabilistic and additive, “potentiating dants as a byproduct.35 factor” is a more accurate description. However, we Antioxidant Defenses.—Oxidants are deactivated, will retain the term “trigger” due to its simplicity that is, reduced, by certain enzymes and by nonen- and nearly universal use in the migraine literature. zymatic antioxidants. Major examples of antioxi- dant enzymes are superoxide dismutase, catalase, METHODS glutathione peroxidase, and peroxyredoxins. Exam- A Medline search was conducted by the author ples of nonenzymatic antioxidants are glutathione crossing the terms “oxidative stress” and “brain” and ubiquinone (coenzyme Q10), both of them with “alcohol,” “dehydration,” “water deprivation,” endogenously generated and, solely from the diet, “monosodium glutamate,” “aspartame,” “tyramine,” ascorbic acid, alpha-tocopherol, and beta caro- “phenylethylamine,” “dietary nitrates,” “noise,” “wea- tene.36 In particular, glutathione is the main nonen- ther,” “hypoglycemia,” “hypoxia,” “infection,” zymatic antioxidant found in cells.22,37 It is easily “estrogen,” “circadian,” “information processing,” oxidized, protecting protein thiol (R-SH) groups “psychosocial stress,” or “nitroglycerin and tolerance.” from oxidation. It is regenerated in cells by the glu- The reference lists of the resulting articles were exam- tathione reductase enzyme. For oxidative damage ined for further relevant studies. The search was that has already occurred, the body has a number restricted to English language articles published of compensatory and repair processes. For example, between 1990 and 2014. Studies were included in the an enzyme, apurinic endonuclease (Ape1), activates review if they examined the exposure as a single inter- the base excision repair pathway, correcting com- vention and reported empirical data on whether and/or mon types of oxidative damage to DNA.38 by what mechanisms the exposure induced oxidative Migraines and Oxidative Stress.—Migraine may stress. In vivo studies were selected when available. be associated with increased vulnerability to oxida- Studies were preferred if they reported on exposures tive stress. Some migraineurs, particularly those plausible for daily life (eg, studies of acoustic trauma with hyperintensities on MRI and/or from blast injuries were excluded for “noise”). In seven who have a family history of migraine, show cases, no relevant articles were found and the searches decreased activity of catalase in the serum, meas- were modified as follows: “tyramine and oxidative and ured interictally,39,40 and thus may have difficulty brain,” “information processing and oxidative stress,” detoxifying hydrogen peroxide. Lower activity of “nitroglycerin and tolerance and oxidative stress,” superoxide dismutase has been found in the eryth- “monosodium glutamate and ,” “sleep Headache 15 deprivation and oxidative stress and brain,” “air pol- Alcohol is metabolized in the body in part by lutants and oxidative stress and brain,” and cytochrome P450-2E1 (CYP2E1) and in part by “nitrosamines and oxidative stress and brain.” For fla- alcohol dehydrogenase. Superoxide and hydrogen vonoids, the search “flavonoids and prooxidant” was peroxide are byproducts of CYP2E1 function- used to exclude the very large number of studies ing.35,49 Some of this enzyme is located within the examining the antioxidant effects of flavonoids. mitochondria, where oxidative damage could pro- Because noninvasive measures of oxidative stress in a mote further oxidant production through the leak- living brain are not yet available, the studies were age of electrons from complexes I and III.50 generally of in vitro or animal data. Moreover, CYP2E1 is inducible – exposure to a substrate such as ethanol prevents its proteosomal RESULTS degradation, causing an increase in the amount of The number of search results vs retained stud- the enzyme within the cell.35,50 ies were: dehydration (4/1), water deprivation (5/1), For both CYP2E1 and alcohol dehydrogenase, aspartame (7/4), phenylethylamine (7/2), nitros- a metabolite of alcohol is acetaldehyde, a toxin that amines (7/1), noise (44/1), hypoglycemia (52/2), psy- activates inducible nitric oxide synthase, xanthine chosocial stress (12/1) and, for the modified oxidase, and neuronal NADPH, causing further searches, tyramine (27/5), monosodium glutamate production of oxidants.49 In addition, high doses of (49/1), air pollutants (42/2), sleep deprivation (42/2), alcohol activate microglia, markedly increasing the and information processing (38/1). For hypoxia (897 production of .51 results), alcohol (296), infection (244), estrogen The fluidity of mitochondrial membranes is (135), flavonoids (112), and nitroglycerin (72), affected acutely by ethanol, leading to altered numerous studies were available reporting the same enzyme function, leakage of electrons, and increased mechanisms. For these triggers, between 1 and 4 oxidant production.52 This in turn is thought to oxi- illustrative papers were retained. dize the membranes further, causing increased oxi- dant production in a vicious cycle.53 Dietary Triggers Moreover, as a diuretic, alcohol induces a com- In 1674, Thomas Willis cited “errata in victu,” pensatory release of arginine vasopressin.54,55 As errors in diet, as eliciting factors in cephalalgia (Willis, will be discussed further when we consider water 1674, cited in Living, 1873).44 In modern times, nutri- deprivation, vasopressin increases superoxide re- tional factors consistently turn up in migraineurs’ lease by the vasculature.56 reports, although the proportion of patients with prom- Alcohol is often cited by patients as among the inent alimentary triggers may be small.1,4,45 Several most prominent dietary triggers.45,57 It is plausible dietary components in particular have been suspected: that this reflects the numerous ways in which alco- Alcohol.—Long-term exposure to high doses of hol generates potentially toxic levels of oxidative alcohol, as seen in chronic alcoholics, leads to loss stress in the brain. of volume, diffuse white matter atro- Water Deprivation.—Water deprivation is sup- phy, and mild .46 Even adolescents and ported as a migraine trigger by retrospective report58 young adults with alcohol use disorders show im- and case studies.59,60 In mice, dehydration increases pairments in executive functioning that resemble the concentration of arginine vasopressin in the premature aging, and studies suggest plasma, through increased production by the hypo- loss of volume of prefrontal cortex and the hippo- thalamus and because plasma volume is lower.56 campus.47 In animal studies, a single day of alcohol Vasopressin, in turn, releases endothelin which, acting consumption, analogous to a binge, is sufficient to through the endothelin A receptor, raises the produc- cause neural degeneration and reactive .48 tion of superoxide anion by the vasculature. The The brain damage in alcohol is thought to be medi- resulting oxidative stress is marked enough to impair ated by oxidative stress from several sources. cerebrovascular reactivity and autoregulation.56 16 January 2016

Monosodium Glutamate.—Some support for a sweeter than sucrose, amounts too small to contrib- general monosodium glutamate (MSG) syndrome, ute calories can be used to sweeten foods and bev- including headache, muscle tightness, numbness, erages.67 Some68,69 but not all70 double-blind, weakness, and flushing, is found in two double-blind, placebo-controlled trials have found depression and placebo-controlled trials.61,62 Both trials used high subtle cognitive impairments with high intake over doses (up to 5 g) on an empty stomach in people 1–2 weeks, implying an effect on the brain. -identified as sensitive to MSG. In both cases, One avenue by which aspartame might have the presence of symptoms was dose-related but the neural effects is through the methyl ester linkage, reproducibility of individual symptoms was low.63 which is metabolized by the body into methanol. MSG that reaches general circulation appears The methanol, in turn, is metabolized by the alco- to raise brain levels of glutamate.64 There, it stimu- hol dehydrogenase system into formaldehyde and lates ionotropic and metabotropic glutamate recep- then to formate.67 Methanol, formaldehyde, and tors on neurons, causing calcium influx. This formate are all , as is well known from increases the neurons’ rates of oxidative phospho- the blindness (from retinal and optic damage) rylation to provide the energy for restoring ion in people consuming methanol during Prohibition.71 homeostasis, which in turn raises oxidant produc- Formate inhibits mitochondrial complex III, causing tion. There can also be uptake of calcium by the release of superoxide, peroxyl, and hydroxyl radi- mitochondria, causing transient opening of perme- cals.72 Methanol freely passes through the blood- ability transition pores and consequent further brain barrier. release of superoxide anion.64 Further, a portion of the methanol is metabo- When glutamatergic stimulation is at normal lized through an alternative, microsomal oxidizing physiological concentrations, the resulting oxidative pathway that creates as a byproduct, free radicals.72 reactions likely play a role in synaptic plasticity, Moreover, in the liver and presumably the brain, learning, and memory, and the collateral DNA enzymatic detoxification of methanol consumes glu- damage is readily repaired.65 At supraphysiologic tathione, directly impairing antioxidant defenses.72,73 concentrations, however, such as when the gluta- Thus, in rats fed high doses of aspartame there are mate comes from an exogenous source, the exces- signs of oxidative stress and a compensatory increase sive signaling and calcium influx leads to in antioxidant enzymes in the cortex and other brain excitotoxicity: damage, , and/or regions.22,74 from oxidative stress.64 Another source of neural effects may be excito- In most cases, it seems doubtful that MSG toxicity: Aspartame is metabolized into phenylala- added to food significantly raises the levels in gen- nine (50%) and aspartic acid (40%), as well as eral circulation. In Western diets, approximately methanol (10%).75 Brain levels of aspartic acid 10 g of glutamate is consumed daily as a normal increase acutely after aspartame ingestion.67 Aspar- constituent of protein and added MSG (0.6 – 2.3 g/ tic acid is a precursor of glutamate and also a direct day) is within the range of dietary variation. More- agonist at NMDA receptors.67 This presumably over, of ingested glutamate, 95% is used by the causes enhanced glutamatergic signaling and excito- enterocytes lining the intestines as a source of toxicity.67 A third mechanism, seen over a 2-week energy and for synthesis of glutathione.63 However, period, is activation of microglia with resulting pro- normal dietary levels might not be benign to sensi- duction of nitric oxide and other inflammatory tive individuals. Foods naturally high in free gluta- mediators.76 The nitric oxide, in turn, would lead to mate, such as citrus and processed meats, figure high oxidative stress via the peroxynitrite radical. in the list of reported migraine triggers.66 Consistent with this evidence of oxidative stress Aspartame.—Aspartame consists of two amino and impact on the brain, some77,78 but not all79 acids, phenylalanine and aspartic acid, bound double-blind, placebo-controlled trials have sug- together as a methyl ester. Because it is 180 times gested a migraine triggering effect of aspartame. Headache 17

Tyramine.—Tyramine is found in red wine and trigger foods as chocolate, wine, and some aged cheese. It first came under suspicion as a cheeses.98 Also, phenylethylamine is a minor migraine trigger from the reported by metabolite of phenylalanine, of which aspartame is people who ingested these foods while taking a significant source.22 Phenylethylamine functions MAO inhibitors, and from evidence that male as an excitatory neuromodulator, and with acute migraineurs may have constitutively lower platelet administration it potentiates dopamine release and MAO activity.80 Migraine without aura is associ- increases motor behavior.98 It is also a norepineph- ated with higher plasma levels of octopamine and rine and dopamine reuptake inhibitor.99 synephrine – metabolites of tyramine – suggesting With long-term administration or at high doses, that baseline levels of tyramine may be elevated.81 however, it inhibits complex 1 of the mitochondria Tyramine has been supported as a migraine trigger in the dopamine-containing cells of the substantia in several double-blind, placebo-controlled trials, nigra, leading to release of hydroxyl radicals, oxida- nearly all from Hanington’s laboratories,82,83 but tive stress-mediated damage, and Parkinsonian there have been negative results as well.84 Only symptoms in laboratory animals.98,100 In fact, at about 5% of migraineurs are thought to be sensi- high doses, b-phenylethylamine behaves similarly to tive to tyramine.85 rotenone and MPTP, used as neurotoxic models of In the brain, tyramine is endogenously pro- Parkinson disease.98 duced, has receptors (trace amine-associated recep- Nonetheless, b-phenylethylamine has not been tors),86 and seems to function by modulating the established with certainty as a migraine trigger. effects of dopamine.87 In particular, tyramine is Other constituents, particularly the flavonoids in indirectly sympathomimetic and both facilitates chocolate and wine may be the true culprit.101,102 dopamine release and interferes with its reuptake.88 And the flavonoid content of chocolate and wine is Tyramine is degraded by MAO, types A and B, generally considered anti-inflammatory. However, found on the outside surface of the outer mem- there are circumstances in which flavonoids may brane of mitochondria. A byproduct of this reaction also generate oxidative stress, as discussed in the is hydrogen peroxide87,89 in amounts far exceeding next section. the background production by mitochondrial respi- Flavonoids.—The case for flavonoids as a ration.90 Although MAO metabolizes other mono- migraine trigger is circumstantial: They are found amines as well, such as dopamine and serotonin, at in fairly high quantity in the likely trigger foods red least in vitro the amount of hydrogen peroxide gen- wine and citrus, they inhibit an enzyme (phenolsul- erated from tyramine is far greater than from dopa- photransferase) that degrades catecholamines, and mine or serotonin.91 Tyramine, in fact, seems to be one class in particular, isoflavones, has weak estro- a source of oxidative damage to mitochondrial genic properties.4 DNA.92 Although pure flavonoids, without a glucose A caveat is that under normal circumstances, moiety, are clearly antioxidants in vitro, the func- tyramine does not cross the blood-brain barrier.87 tioning of flavonoids in the body, where they are Thus, if tyramine sensitivity is centrally mediated, it subject to extensive metabolic processing, is not would require a disruption of the blood-brain bar- known with certainty.103 Indeed, it may be through rier, such as from an inflammatory process, in die- pro-oxidant properties that flavonoids are able to tary migraineurs. In fact, there is some indication induce cell-cycle arrest and apoptosis in cancer that the blood-brain barrier is compromised interic- cells.104,105 Their antioxidant effects in vivo may be tally in migraine93–95 although the evidence is indirect, by inciting increased expression of antioxi- inconsistent and of uncertain interpretation.96,97 dant and detoxifying enzymes.24 Phenylethylamine.—b-phenylethylamine, a bio- What does seem clear is that flavonoids func- genic amine derived from phenylalanine, is a sus- tion as pro-oxidants under three circumstances: pected active ingredient in such possible migraine high concentration,103 oxidation by intracellular 18 January 2016 enzymes such as myeloperoxidase,106 and when in ing increased levels of oxidative DNA damage and contact with transition metals such as iron and cop- 4-hydroxynonenal).121 Because nitrosamines seem per.107 Ordinarily, iron and copper are bound to to cross the blood-brain barrier122 it is possible this proteins (eg, hemoglobin, ferritin for iron, cerulo- same effect occurs in vivo. plasmin for copper) that prevent this effect but free iron or copper may exist in the body after injury or Environmental in disease states.103 Noise.—Exposure to noise seems to increase the In fact, increased nonheme iron has been found probability of a migraine the following day.123 It is in the periaqueductal gray,108 red nucleus, putamen, cited by patients as a trigger in retrospective sur- and globus pallidus109,110 in migraine. In theory, oxi- veys45 and has been supported as a headache trig- dative damage from repeated activation, hyperoxia, ger in a controlled laboratory study.124 and perhaps nociceptive input itself may cause the In young mice, moderate levels of noise (80 dB buildup of iron and underlie the transition from epi- SPL, 2 hours per day for 6 weeks) leads to deficits sodic to chronic migraine.111 In the event that some in spatial learning and avoidance conditioning. The of this increased iron is free, that is, not sequestered extent of these deficits is correlated with increased in ferritin or by microglia, then it could catalyze fla- oxidative stress in the hippocampus, auditory cor- vonoids locally into pro-oxidant derivatives. tex, and inferior colliculus.125 How moderate levels Nitrates.—Dietary nitrites, found in cured and of noise exposure produce such downstream effects processed meats, first emerged as a migraine trigger is not clear, although presumably excitotoxicity is a in a 1972 single-blind case study,112 later confirmed reasonable hypothesis. in a small double-blind, placebo-controlled trial.113 Weather and Pollution.—Certain broad weather There has been only occasional confirmation since patterns may increase or decrease the probability then, for example,114 perhaps because the level of of a migraine.123 Among specific meteorological nitrate added to meats has declined markedly since variables, a wind speed greater than 38 km/hour 1972.115 (24 miles/hour) has received particular support.126 The mitochondria, and certain enzymes (eg, The explanation has not been fully worked out but xanthine oxidase) and proteins (eg, myoglobin, an wind speed may be a proxy for the amount of dust oxygen-binding molecule in the muscles) are all the air has picked up (particulate matter with diam- capable of transforming nitrite into nitric oxide. eter >2.5 lm).127 Air pollution, and particulate These pathways function mostly when oxygen levels matter specifically, increases the probability of hos- are low, and serve to dilate blood vessels, slow pitalization for headache.128,129 metabolism, and prevent injury by free radicals.116 Epidemiological evidence links air pollution to Thus, dietary and endogenous nitrite plays a key neurotoxicity. For example, in Valcamonica, Italy, physiological role and may function as an antioxi- where there is a high level of ferromanganese pol- dant. However, under certain conditions xanthine lution from local industry, ostensibly healthy ado- oxidase produces superoxide,117 which transforms lescents show and subtle deficits in motor nitric oxide into the peroxynitrite radical, a source dexterity and coordination, whose intensity is corre- of oxidative stress.118 lated with the degree of exposure.130 In Alternatively, nitrates – which are abundant highly polluted areas of Mexico City, auditory also in vegetables and thus the brainstem nuclei in children show aggregated alpha – might be innocent. Some other component of (the involved in Parkinson’s dis- processed meats might be to blame, such as nitros- ease and dementia with Lewy bodies) and markers amines, a reaction product of nitrates with protein, of oxidative damage to proteins and DNA.131 which have been linked to brain cancer in animal The mechanism linking exposure to pollutants and some119 but not all120 epidemiological studies. with neurotoxicity is not certain. However, the In vitro, nitrosamines cause oxidative stress, includ- olfactory nerve consists of axonal projections from Headache 19 the nasal epithelium to the olfactory bulb. Particu- thine oxidoreductase enzyme, which ordinarily lates can be transported along the axons, bypassing helps dispose of purines, is cleaved to form xan- the blood-brain barrier, and cross the synaptic cleft thine oxidase.117 Xanthine oxidase may help into the brain parenchyma.132 Based on in vitro evi- increase blood flow to ischemic tissue by producing dence, the pollutants may activate microglial via nitric oxide through the reduction of nitrites and toll-like receptor signaling, inducing an inflamma- nitrates. However, depending on the microenviron- tory response and oxidative stress.133 Thus, the neu- ment, xanthine oxidase also produces superoxide.117 rotoxicity of inhaled particulate matter (collected Nitric oxide, of course, is usually produced by from a public park in Tuxedo, NY, and concen- the nitric oxide synthases, using L-arginine and trated) is exacerbated in apolipoprotein E null rats, molecular oxygen as substrates and NADPH and which are systemically prone to oxidative stress.134 tetrahydrobiopterin as cofactors. When the supply of oxygen is insufficient, however, the nitric oxide Physiological synthases become “uncoupled,” producing superox- Hypoglycemia.—Hypoglycemia is regarded as a ide instead of nitric oxide.147 migraine trigger based on case reports, for exam- Infection.—A sick day not due to headache may ple,135 an experimental study,136 the ability of fast- increase the duration of a subsequent headache,123 ing to produce a headache,137–139 and triggering of suggesting a common pathway. In mice, systemic migraines by glucose tolerance testing.140 Migraine exposure to an inflammatory stimulus, bacterial cell may occur only after blood glucose levels are renor- wall lipopolysaccharides, can cause induction of malized.135 In the glucose tolerance test, migraines NADPH oxidase lasting for 10 months or longer.148 typically occur 1 to 4 hours after glucose ingestion. Estrogen.—The relationship between estrogen Extreme hypoglycemia leads to coma and and migraines is complex. That women have an death, while repeated severe episodes in type I dia- approximately threefold greater risk of migraines, betic patients are associated with of the beginning at menarche, suggests that female sex hor- cortex, especially the occipital and parietal lobes, mones contribute to migraines.149,150 In an animal hippocampus, and basal ganglia.141,142 Animal stud- model of induced neurogenic inflammation, estradiol ies suggest that very severe hypoglycemia in the intensifies vasodilation, CGRP expression, and brain is, initially, an excitotoxic state due to exces- migraine-like behavior.151 However, for immediate sive glutamate release, leading to mitochondrial triggering, the case is strongest for estrogen with- dysfunction and oxidative stress.143 Far higher lev- drawal, that is, the declining estrogen level seen els of oxidative stress follow the reintroduction of perimenstrually and at the time of ovulation.150 glucose, which activates nitric oxide synthase and The phenolic structure of 17b-estradiol confers neuronal NADPH oxidase and causes free radical- to it antioxidant properties in vitro, particularly for mediated death of neurons.144 NADPH oxidase scavenging hydrogen peroxide152 but it is not clear requires a steady supply of NADPH, which cannot that this is relevant in vivo.153 Markers of oxidative be regenerated until glucose is restored.144 stress in the plasma tend to peak during the late Hypoxia.—Hypoxia is a suspected migraine trig- follicular and early luteal phases of the menstrual ger based on the similarity of migraine to acute cycle,154,155 perhaps because of the rapid prolifera- mountain sickness, increased risk of migraine from tion of follicular or endometrial tissue,154 and to living at high altitudes, and a tendency found in a fall during menstruation. controlled trial.145 In the central , however, estro- In animal models, intermittent hypoxia causes gen facilitates a number of protective processes. It oxidative stress via mitochondrial dysfunction (the limits oxidative damage by increasing the expres- mitochondria cannot function efficiently without an sion of Cu, Zn-superoxide dismutase,156 and helps adequate supply of substrates) and via NADPH reverse extant damage by upregulating repair en- oxidase.146 In addition, in hypoxic cells, the xan- zymes thioredoxin157 and Ape1.38 Estradiol also 20 January 2016 inhibits apoptosis by increasing the expression of Sleep deprivation is associated with a depletion Bcl-238,158 and reduces neuroinflammation.159 The of reduced glutathione in the brain.172 As we have reduced activation of microglia, in turn, should seen, glutathione is the body’s main nonenzymatic lower brain oxidative stress. antioxidant and is utilized by the antioxidant enzyme In vitro, estradiol prevents glutamate toxicity glutathione peroxidase, which detoxifies the hydroxyl by downregulating the expression of AMPA- and radical. In rats, cognitive deficits from sleep depriva- 173 NMDA-type glutamate receptors.160 In animals, tion are prevented by antioxidants, further tying estradiol reduces infarct size following middle cere- the harm of sleep deprivation to oxidative stress. 174 bral artery occlusion.161 Mental Overwork.—Living noted that among On the whole, then, estrogen appears to be neu- migraine triggers, “and perhaps more certainly roprotective, with antioxidant, anti-inflammatory, influential than any of them” were “mental emotion antiexcitotoxic, and antiapoptotic actions in vivo.It and excessive brain-work.” is not known, however, whether the short-term phys- Whether triggering by excessive mental work iological withdrawal of estrogen during the men- can be explained by oxidative stress is not clear strual cycle is enough to reduce this protection and because little is known about whether information render cells vulnerable to oxidative stress. processing influences redox state. However, the -like structures of the retina can be damaged by high intensity light exposure (phototoxicity)175,176 Behavioral as can the cochlea and upstream cortical structures Sleep Cycle.—John R. Graham included “sleeping by excessive noise (acoustic trauma);125,177 this dam- late” among his “errors of living for migraineurs to age is specifically through oxidative mechanisms. heed,”162 and “sleeping late,” “oversleeping,” “sleep Similarly, painful stimulation of the limbs in rats disturbance,” and “change in sleeping habits or increases lipid peroxidation in sensorimotor cortex sleep” figure prominently among the triggers nomi- for days afterwards.178 Excessive release of gluta- nated by patients.1,2,45 Of course, the maintenance of mate (excitotoxicity) and the catabolism of dopa- stable biological rhythms is helpful for managing mine179 are sources of oxidative stress. other conditions that involve the brain, such as bipo- Moreover, information processing by the cortex 163,164 lar disorder and dementia. is associated with augmented power and with syn- Disruptions in the circadian rhythm by exposing chronization across cortical regions in the high fre- animals to shifts in the light-dark cycle interfere quency bands of the electroencephalogram (EEG), with learning and with the retention and retrieval of particularly the high gamma range (60–200 Hz).180 165 memories. Hippocampal functions are most read- Lower gamma band activity (30–60 Hz, especially ily affected but with increasing circadian disruption, around 40 Hz) is thought to reflect perceptual bind- cognition that depends on the prefrontal cortex or ing – the integration of stimulus features into a coher- 166 dorsal is affected as well. Anatomically, ent percept or gestalt, integration across different 167,168 there is evidence for reduced . sensory modalities, and the maintenance of items in Corticosterone levels may be directly elevated or the working memory.181 Higher gamma band activity is circadian disruption may potentiate the corticoster- thought to reflect such higher functions as selective 169 one response to laboratory stresses. attention and motor ,181 and perhaps infor- Approximately, 10% of all gene mation processing in general.180 Gamma band activ- in the mouse shows circadian oscillation, suggesting ity may be more intense, with stronger amplitude and regulation by clock genes.170 In particular, the tran- decreased inhibition, in migraineurs, at least in scription of major antioxidant enzymes, including response to visual stimulation.182 superoxide dismutase, catalase, glutathione peroxi- Physically, the correlates of power augmenta- dase, peroxiredoxins, and sestrins, appears to be tion at higher EEG frequencies are higher neuronal controlled by circadian clock genes.171 firing rates and greater utilization of glucose and Headache 21 oxygen.180,183 In particular, gamma oscillations seem ies in which primates must repeatedly establish, to reflect the very rapid, synchronized firing of a cer- challenge, or defend their place in the status hierar- tain class of GABAergic, inhibitory interneurons, chy. This type of stress is not benign but associated parvalbumin-expressing basket cells. Because this fir- with impairments in neurogenesis and dendritic ing is extremely rapid, it entails enormous energy branching and growth in the CA3 region of the hip- demands by ion pumps to restore the neuron to the pocampus.193,194 These types of changes are gener- resting state.184 The basket cells are replete with ally associated with oxidative stress.33 mitochondria that appear to be functioning at full Daily hassles are supported as a migraine trig- capacity during gamma oscillations.184 ger in prospective diary195 and single-subject196 A very high metabolic rate may cause oxidative studies. Both the amount and temporal patterning stress. We have seen that ordinarily, between 0.1% of stress may be key.197 and 4% of electrons leak from mitochondrial com- plexes I and III, generating superoxide anions.29 Pharmacological Thus, the amount of superoxide produced appears Nitroglycerin.—Nitroglycerin, a nitric oxide to be a linear function of metabolic rate. Moreover, releaser, is known to cause an immediate headache with high firing rates, there is potential for transient in susceptible individuals and, in migraineurs, a overload of calcium ions in the neuron, increasing migraine after about 5 hours.198 Nitric oxide is prob- oxidant production further.185 Thus, fast-spiking ably an antioxidant in vivo.199 Nitroglycerin, how- interneurons may be key sources of oxidative stress, ever, also causes the production of superoxide by the contributing to aging and brain pathology.185 enzyme and by vascular Stress.—In patient reports, stress is generally NADPH oxidase.200,201 Moreover, nitric oxide reacts among the most important migraine triggers.45,186 with superoxide approximately three times faster Even this may be an underestimate, as stress may than does the detoxifying enzyme superoxide dismu- have a 3- or 4-day delayed effect,187 making its trig- tase (reaction rate 6.7 3 109 M21s21 vs 2 3 109 gering less observable. It has been confirmed as a M21s21),33,118 forming the highly reactive peroxyni- trigger in laboratory studies.188 Although there are trite radical.202 a large number of major life stresses, a common The main findings in the Results section are unifying characteristic is social loss. summarized in the Table. In rats, being raised in isolation leads to cognitive (impaired sensorimotor gating) and behavioral defi- DISCUSSION cits, as well as a depletion of glutathione in the stria- Thus, migraine triggers encountered in clinical tum and cortex.189 Psychosocial stress seems to practice are capable of generating oxidative stress. activate the same cellular defense processes as infec- Mechanisms include a high rate of energy production tion or toxins, perhaps because stress is a warning of by the mitochondria, toxicity or altered membrane impending injury. In prefrontal cortex and the properties of the mitochondria, calcium overload and nucleus accumbens, psychosocial stress in rats causes excitotoxicity, neuroinflammation and activation of a tremendous increase in NOX2 expression and microglia, and activation of neuronal NADPH oxi- markers of oxidative stress.33,190 Of note, this effect dase. For some triggers, oxidants also arise as a appears to trigger the signaling pathway for hypoxia, byproduct of MAO or cytochrome P450 processing, that is, by hypoxia inducible factor-1.191 Thus, the or from uncoupling of nitric oxide synthase. In turn, idea that social stresses are encoded, not by novel oxidative stress, transduced by the TRPA1 ion chan- structures but by novel ways of activating evolutio- nel on C fibers, can initiate CGRP release and the narily preexisting structures192 appears true on a neurogenic inflammation of migraine. molecular as well as a neural systems level. As a response to oxidative stress, migraines Daily Hassles.—An analog to the minor but could confer a survival advantage, as suggested by repeated stresses of daily life are perhaps the stud- Loder.17 The brain is uniquely vulnerable to 22 January 2016 oxidative damage because of its high metabolic rate, A second weakness is that triggers have multi- the polyunsaturated (easily oxidized) nature of neu- ple effects, of which only one, oxidative stress, has ronal cell membranes, and the presence of iron and been discussed. Thus, for example, hypoglycemia copper ions, which can catalyze the formation of free increases plasma free fatty acids, changes neuronal radicals. Antioxidant resources and cellular repair ion homeostasis, and alters levels of dopamine and capacity, requiring energy, are limited in the brain.22 serotonin, all of which could play a role in Further, apoptosis as a defense mechanism – simply migraine.135 Many triggers can activate the sympa- deleting through programed death cells that have thetic nervous system which, in large amounts, become too badly damaged203 – is not a good option would threaten homeostasis.211 Yet the capacity of in the brain, as neurons generally cannot be replaced. oxidants to elicit neurogenic inflammation through We have seen that migraineurs may be particularly the TRPA1 ion channel, and the relevance of oxi- susceptible to oxidative stress because of lower antioxi- dative stress to neuronal functioning and survival, dant defenses. Moreover, oxidative stress may help make this a tempting unifying principle. account for the relationship between cortical spreading The broad parsimony of the theory is also a depression (CSD) and the subsequent attack in key weakness: In common with other theories of migraine with aura. Pro-oxidants potentiate and migraine triggers, it does not allow us to deduce antioxidants prevent CSD,204 suggesting a common vul- why a particular combination of exposures would nerability. Moreover, CSD itself causes oxidative lead to a migraine in a particular person at a partic- stress205 and in theory might thereby trigger a migraine. ular time. The gap between theory and moment-to- Consistent with this, CSD also rapidly induces certain moment experience is large. A strength of the antioxidant defenses – glutathione-S-transferase 5, apo- theory is that it indicates the types of variables that lipoprotein E, and major protein precursor.206 would need to be measured to bridge this gap: the Further, in animal studies, by inhibiting the level and sources of brain oxidative stress in enzyme synthase and the glutamate response to trigger exposure. transporter on astrocytes, reactive oxygen species Some triggers do not seem well explained by may increase the concentration of glutamate, con- this model – in particular the triggering by pericra- tributing to an excitable brain.207 nial pain such as from eye strain or myofascial Nonetheless, the approach taken here is not problems at the upper neck.123 It is possible that without weakness. Nearly all of the migraine triggers neck pain is actually a premonitory symptom and we have discussed are from traditional case studies not a trigger.212 Alternatively, however, there is in and retrospective reports by migraineurs, subject to vitro evidence that capsaicin and H2O2 applied to misattribution, illusory correlation, and recall dorsal root ganglion neurons, have a multiplicative bias.123,208 The determination of causality is notori- effect in causing CGRP release.205 Thus, we might ously difficult in naturalistic, uncontrolled observa- speculate that TRPV1 channels, sensitive to inflam- tion.209 It is possible that some putative migraine matory pain, and TRPA1 channels, sensitive to oxi- triggers are in fact nothing more than bystanders. dative stress, interact.213 In that event, pericranial Yet we have seen that physiologically, these putative pain might function to lower the threshold of oxida- triggers are far from innocent; all conform to Blau’s tive stress needed to trigger a migraine. dictum that in higher amounts they would damage A further weakness in the theory is that oxida- the brain. Moreover, if the true proximal trigger is tive stress in the brain increases with age214 while oxidative stress then individual triggers may be diffi- the prevalence of migraines declines.215,216 Presum- cult to validate in the laboratory as the dose, combi- ably, one would need to posit that the ability to nation with other triggers, state of antioxidant detect oxidative stress, like perceptual acuity, defenses, and state of pain thresholds (which may be declines with age, or that the mechanisms that raised by the novelty of being in a study) would all transduce oxidative stress become desensitized as moderate the result.210 basal levels rise. There is evidence, for example, Headache 23 that parthenolide, a component of feverfew and a The suggestion here is that this is a class partial agonist of TRPA1, desensitizes the ion chan- effect,43,205 and that other antioxidants will similarly nel over time.217 be effective. Alternatively, the migraine response itself Consistent with this, regular moderate aerobic might become impaired. The vasodilatory compo- , which seems to increase antioxidant nent may be curtailed by endothelial dysfunction reserves in the brain,234 may also attenuate the and atherosclerosis.216 Brainstem cir- intensity or frequency of migraines.235 cuits, and the in particular, have Similarly, frontalis muscle tension biofeedback, been implicated in migraine,218 and perhaps the empirically supported for the prevention of decline of neurotransmission in these circuits migraines,236 has been shown to reduce serum per- accounts for the improvement in migraines.219 In a oxides and raise levels of nitric oxide and superoxide case-control study, 64% of migraineurs recalled an dismutase in chronic migraine.237 The mechanism is improvement or resolution of migraines following not clear, but as a relaxation technique, it might onset of Parkinson disease.220 lower stress-related activity of NADPH oxidase. Moreover, because trigger escape and avoid- (For migraine preventive medications, the case ance can never be perfect, we must ask why migrai- is less clear: Flunarizine,238,239 propranolol,240 and neurs do not show progressive brain damage over topiramate241 appear to be antioxidants, either time. But perhaps they do. Chong et al report that directly or by raising levels of antioxidant enzymes, age-related cortical thinning is accelerated in while valproate242 and amitriptyline243 appear to be migraineurs.221 That this may not be an effect of pro-oxidants.) the attacks themselves is suggested by the inde- On a practical level, the idea that triggers func- pendence of the decline from disease duration and tion via oxidative stress raises the possibility that migraine frequency. However, the physiological antioxidants, taken at the time of trigger exposure, basis of cortical thinning is poorly understood. A could function as “acute preventives” or “preemptive second possibility is that migraines are neuroprotec- .”232,244 There is some preliminary evidence tive, effectively countering oxidative stress, as dis- forthisforginkgo245 and the combination of ginger cussed briefly below. and feverfew.246 Indeed sumatriptan, in some Conversely, viewing migraine triggers from the respects a late-stage preemptive, scavenges superox- perspective of oxidative stress has a number of ide and hydroxyl radicals in vitro247 and may reduce advantages. It is physiologically plausible via the levels of peroxynitrite in vivo.248 However, such a TRPA1 ion channel. It brings unity to a wide vari- role for antioxidants has not been established and ety of behavioral, environmental, dietary, physio- requires further research. logical, and pharmacological exposures. It provides Conceptually, triggers as sources of oxidative an explanation for how triggers can summate. stress adds to the view of migraines as a defensive Moreover, this perspective on triggers links to response designed to limit further exposure to the the further simplification that a number of agents that triggers. The acute environmental intolerances of reduce oxidative stress seem to function as migraine photophobia, phonophobia, and osmophobia, the preventives, including vitamin E,222 ginkgo,223 melato- aversion to exertion and movement, and the tempo- nin,224 butterbur,225 feverfew,226,227 coenzyme Q10,228 rary decline in mental status all fit with an adaptive and possibly alpha lipoic acid,229 and vitamin C purpose of migraines.249 with230 or without231 pine bark extract. The efficacy Further afield, it raises the possibility that of riboflavin, too, might be indirectly due to reduced migraines might actively counter oxidative stress on a oxidative stress:232 Riboflavin is a coenzyme for gluta- physiological level. A number of candidate processes thione reductase, an antioxidant enzyme, and in its can be adduced. In various tissues of the body, reduced form can directly scavenge oxygen CGRP reduces the expression of NADPH oxidase,250 radicals.233 increases the activity of antioxidant enzymes,251 24 January 2016

Table.—Type and Strength of Evidence Linking Migraine Triggers to Oxidative Stress

Strength of Trigger Sources of Oxidative Stress Type of Evidence Evidence

Dietary Alcohol NOX2, microglial activation, In vitro, animal 111 CYP2E1, mitochondrial tox- human (indirect) icity, vasculature Water deprivation Vasculature Animal 11 Monosodium glutamate Excitotoxicity Animal 11 Aspartame Excitotoxicity, microglial acti- In vitro, animal 111 vation, mitochondrial toxic- ity, microsomes, depletion of glutathione Tyramine MAO In vitro 1 b-Phenylethylamine Mitochondrial toxicity In vitro, animal 1 Flavonoids Direct pro-oxidants under In vitro 1 some circumstances Nitrates Xanthine oxidase In vitro 1 Environmental Noise Unknown (? excitotoxicity) Animal 11 Weather and pollution Microglial activation In vitro, animal 11 Physiological Hypoglycemia NOX2, excitotoxicity Animal 1 Hypoxia NOX2, mitochondrial dysfunc- Animal 11 tion, xanthine oxidase, nitric oxide synthase Infection NOX2 Animal 111 Estrogen withdrawal Loss of antioxidant and antiex- In vitro, animal 1 citotoxic properties Behavioral Sleep cycle Decrease in antioxidant Animal 11 enzymes, depletion of glutathione Mental overwork Mitochondria (oxidative Animal 1 phosphorylation) Stress NOX2 Animal 111 Daily hassles Unknown Animal 11 Pharmacological Nitroglycerin Indirect pro-oxidant under Animal 11 some circumstances

“Strength of Evidence” is based on the number of pathways by which a trigger generates oxidative stress and their likelihood of being relevant in daily life: (111) 5 Multiple well-established mechanisms demonstrated in vivo at plausible levels of expo- sure; (11) 5 A single strong source of oxidative stress demonstrated in vivo, but with uncertain mechanism and/or a level of exposure higher than would be encountered in daily life; (1) 5 In vitro or limited in vivo data. CYP2E1 5 cytochrome P450-2E1; MAO 5 monoamine oxidase; NOX2 5 NADPH oxidase-2. decreases oxidative stress,251 and prevents apoptosis trophic.254 Further, serotonin, released into the blood under oxidizing conditions by stimulating growth- by platelets255 and perhaps throughout the cortex by oriented intracellular signaling pathways.252 It is the dorsal raphe nucleus during a migraine,256 may tempting to suspect a similarly protective role for have neurotrophic properties,257,258 and cause the CGRP in the brain. Similarly albumin, released into release of other neural growth factors.259 Platelet acti- the dura by protein extravasation in migraines, is the vation, which takes place in migraine260 as it does in main antioxidant in plasma.253 In the brain, albumin tissue injury, is neuroprotective and neurorestorative causes astrocytes to release oleic acid, which is neuro- in an in vivo model of cerebral ischemia.261 Headache 25

Nonetheless, a neuroprotective role of migraines 11. Ho TW, Edvinsson L, Goadsby PJ. CGRP and remains highly speculative at this point. its receptors provide new insights into migraine pathophysiology. Nat Rev Neurol. 2010;6:573-582. 12. Lambert GA, Zagami AS. The mode of action of CONCLUSION migraine triggers: A hypothesis. Headache. 2009; Regardless, we have seen that underlying each 49:253-275. of the traditional migraine triggers is their propensity 13. Chakravarty A. How triggers trigger acute migraine to generate oxidative stress. This suggests that the attacks: A hypothesis. Med Hypotheses. 2010;74: recent discovery of the TRPA1 ion channel and the 750-753. indications of higher oxidative stress and/or lower 14. Martin VT, Behbehani MM. Toward a rational antioxidant defenses in migraineurs allow us to posit understanding of migraine trigger factors. Med a physiologically informed principle uniting seem- Clin North Am. 2001;85:911-941. ingly disparate migraine triggers and suggest a mech- 15. Monteiro JMP, Dahlof€ CGH. Single use of sub- anism for an evolutionary advantage of migraines. stances. In: Olesen J, Tfelt-Hansen P, Welch KMA, eds. The Headaches. 2nd ed. Philadelphia: Lippincott Williams & Wilkins; 2000:861-869. 16. Blau JN. Migraine: Theories of pathogenesis. Lancet. 1992;339:1202-1207. REFERENCES 17. Loder E. What is the evolutionary advantage of 1. Kelman L. The triggers or precipitants of the acute migraine? Cephalalgia. 2002;22:624-632. migraine attack. Cephalalgia. 2007;27:394-402. 18. Benemei S, Fusi C, Trevisan G, Geppetti P. The 2. Robbins L. Precipitating factors in migraine: A TRPA1 channel in migraine mechanism and retrospective review of 494 patients. Headache. treatment. Br J Pharmacol. 2014;171:2552-2567. 1994;34:214-216. 19. Wantke F, Focke M, Hemmer W, et al. Exposure 3. Dowson AJ, Sender J, Lipscombe S, et al. Estab- to formaldehyde and phenol during an anatomy lishing principles for migraine management in dissecting course: Sensitizing potency of formalde- primary care. Int J Clin Pract. 2003;57:493-507. hyde in medical students. Allergy. 2000;55:84-87. 4. Borkum JM. Chronic Headaches: Biology, Psy- 20. Dussor G, Yan J, Xie JY, Ossipov MH, Dodick chology, and Behavioral Treatment. Mahwah, NJ: DW, Porreca F. Targeting TRP channels for Lawrence Erlbaum Associates; 2007. novel migraine therapeutics. ACS Chem Neuro- 5. Graham JR. Migraine, quo vadis? Headache. sci. 2014;5:1085-1096. 1988;28:681-688. 21. Kozai D, Ogawa N, Mori Y. Redox regulation of 6. Peroutka SJ. What turns on a migraine? A sys- transient receptor potential channels. Antioxid tematic review of migraine precipitating factors. Redox Signal. 2014;21:971-986. Curr Pain Headache Rep. 2014;18:454. 22. Abdel-Salam OME, Salem NA, Hussein JS. Effect 7. Scopp AL. Headache triggers: Theory, research, of aspartame on oxidative stress and monoamine and clinical application – Part I. Headache Quar- levels in lipopolysaccharide-treated terly. 1992;3:32-38. mice. Neurotox Res. 2012;21:245-255. 8. Spierings ELH, Donoghue S, Mian A, Wober€ C. 23. Cahill-Smith S, Li J-M. Oxidative stress, redox Sufficiency and necessity in migraine: How do we signaling and endothelial dysfunction in - figure out if triggers are absolute or partial and, related neurodegenerative diseases: A role of if partial, additive or potentiating? Curr Pain NADPH oxidase 2. Br J Clin Pharmacol. 2014; Headache Rep. 2014;18:455. 78:441-453. 9. Lai C, Dean P, Ziegler DK, Hassanein RS. Clinical 24. Halliwell B. The antioxidant paradox: Less par- and electrophysiological responses to dietary chal- adoxical now? Br J Clin Pharmacol. 2012;75: lenge in migraineurs. Headache. 1989;29:180-186. 637-644. 10. Burstein R, Jakubowski M. Unitary hypothesis 25. Halliwell B, Gutteridge JMC. Free Radicals in for multiple triggers of the pain and strain of Biology and Medicine. 3rd ed. New York: Oxford migraine. J Comp Neurol. 2005;493:9-14. University Press; 1999. 26 January 2016

26. Sies H, Cardenas E. Oxidative stress: Damage to dants, antioxidants and disease mechanisms. intact cells and organs. Philos Trans R Soc Lond Redox Biol. 2013;1:244-257. B Biol Sci. 1985;311:617-631. 37. Meisenberg G, Simmons WH. Principles of Medi- 27. Grivennikova VG, Kareyeva AV, Vinogradov cal Biochemistry. 3rd ed. Philadelphia: Elsevier AD. What are the sources of hydrogen peroxide Saunders; 2012. production by heart mitochondria? Biochim Bio- 38. Dietrich AK, Humphreys GI, Nardulli AM. 17b- phys Acta. 2010;1797:939-944. estradiol increases expression of the oxidative 28. Herrero A, Barja G. ADP-regulation of mito- stress response and DNA repair protein apurinic chondrial free radical production is different with endonuclease (Ape1) in the of Complex I- or Complex II-linked substrates: female mice following hypoxia. J Steroid Biochem Implications for the exercise paradox and brain Mol Biol. 2013;138:410-420. hypermetabolism. J Bioenerg Biomembr. 1997;29: 39. Aytac¸ B, Cos¸kun O,€ Alioglu B, et al. Decreased 241-249. antioxidant status in migraine patients with brain 29. Moran M, Moreno-Lastres D, Marın-Buera L, white matter hyperintensities. Neurol Sci. 2014; Arenas J, Martın MA, Ugalde C. Mitochondrial 35:1925-1929. respiratory chain dysfunction: Implications in 40. Shukla R, Barthwal MK, Srivastava N, et al. neurodegeneration. Free Radic Biol Med. 2012; Neutrophil-free radical generation and enzymatic 53:595-609. antioxidants in migraine patients. Cephalalgia. 30. Indo HP, Davidson M, Yen H-C, et al. Evi- 2004;24:37-43. dence of ROS generation by mitochondria in 41. Bolayir E, Celik K, Kugu N, Yilmaz A, Topaktas cells with impaired electron transport chain and S, Bakir S. Intraerythrocyte antioxidant enzyme mitochondrial DNA damage. . activities in migraine and tension-type headaches. 2007;7:106-118. J Chin Med Assoc. 2004;67:263-267. 31. Marzetti E, Csiszar A, Dutta D, Balagopal G, 42. Shimomura T, Kowa H, Nahano T, et al. Platelet Calvani R, Leeuwenburgh C. Role of mitochon- superoxide dismutase in migraine and tension- drial dysfunction and altered in cardio- type headache. Cephalalgia. 1994;14:215-218. vascular aging and disease: From mechanisms to 43. Alp R, Selek S¸, Alp SI,_ Tas¸kin A, Koc¸yigit A. therapeutics. Am J Physiol Heart Circ Physiol. Oxidative and antioxidative balance in patients of 2013;305:H459-H476. migraine. Eur Rev Med Pharmacol Sci. 2010;14: 32. Jiang F, Zhang Y, Dusting G. NADPH oxidase- 877-882. mediated redox signaling: Roles in cellular stress 44. Living E. On Megrim, Sick-Headache, and response, stress tolerance, and tissue repair. Phar- Allied Disorders: A Contribution to the Pathol- macol Rev. 2011;63:218-242. ogy of Nerve-Storms. London: J. and A. Church- 33. Schiavone S, Jaquet V, Trabace L, Krause K-H. ill; 1873. Severe life stress and oxidative stress in the brain: 45. Camboim Rockett F, Castro K, Rossoni de From animal models to human pathology. Anti- Oliveira V, da Silveira Perla A, Fagundes Chaves oxid Redox Signal. 2013;18:1475-1490. ML, Schweigert Perry ID. Perceived migraine 34. Conrad M, Schick J, Angeli JPF. Glutathione triggers: Do dietary factors play a role? Nutr and thioredoxin dependent systems in neurodege- Hosp. 2012;27:483-489. nerative disease: What can be learned from 46. Mochizuki H, Masaki T, Matsushita S, et al. Cog- reverse genetics in mice. Neurochem Int. 2013;62: nitive impairment and diffuse white matter atro- 738-749. phy in alcoholics. Clin Neurophysiol. 2005;116: 35. Carolina V-OA, Julio M, Rafael C-C, Oscar P-G, 223-228. Javier E-AJ. CYP2E1 induction leads to oxida- 47. Hermens DF, Lagopoulos J, Tobias-Webb J, tive stress and cytotoxicity in glutathione- et al. Pathways to alcohol-induced brain impair- depleted cerebellar granule neurons. Toxicol In ment in young people: A review. Cortex. 2013;49: Vitro. 2014;28:1206-1214. 3-17. 36. Kalyanaraman B. Teaching the basics of redox 48. Hayes DM, Deeny MA, Shaner CA, Nixon K. biology to medical and graduate students: Oxi- Determining the threshold for alcohol-induced Headache 27

brain damage: New evidence with gliosis markers. 61. Geha RS, Beiser A, Ren C, et al. Multicenter, Alcohol Clin Exp Res. 2013;37:425-434. double-blind, placebo-controlled, multiple-chall- 49. Haorah J, Ramirez SH, Floreani N, Gorantla S, enge evaluation of reported reactions to mono- Morsey B, Persidsky Y. Mechanism of alcohol- sodium glutamate. J Allergy Clin Immunol. 2000; induced oxidative stress and neuronal injury. Free 106:973-980. Radic Biol Med. 2008;45:1542-1550. 62. Yang WH, Drouin MA, Herbert M, Mao Y, 50. Aubert J, Begriche K, Knockaert L, Robin MA, Karsh J. The monosodium glutamate symptom Fromenty B. Increased expression of cytochrome complex: Assessment in a double-blind, placebo- P450 2E1 in nonalcoholic fatty liver disease: controlled, randomized study. J Allergy Clin Mechanisms and pathophysiological role. Clin Immunol. 1997;99:757-762. Res Hepatol Gastroenterol. 2011;35:630-637. 63. Beyreuther K, Biesalski HK, Fernstrom JD, et al. 51. Qin L, Crews FT. NADPH oxidase and reactive Consensus meeting: Monosodium glutamate – an oxygen species contribute to alcohol-induced update. Eur J Clin Nutr. 2007;61:304-313. microglial activation and neurodegeneration. 64. Dief AE, Kamha ES, Baraka AM, Elshorbagy J Neuroinflammation. 2012;9:5. doi:10.1186/1742- AK. Monosodium glutamate neurotoxicity 2094-9-5. increases beta amyloid in the rat hippocampus: A 52. Haorah J, Floreani NA, Knipe B, Persidsky Y. potential role for cyclic AMP protein kinase. Stabilization of superoxide dismutase by acetyl-l- Neurotoxicology. 2014;42:76-82. carnitine in endothelium during 65. Yang J-L, Tadokoro T, Keijzers G, Mattson MP, alcohol exposure: Novel protective approach. Bohr VA. Neurons efficiently repair glutamate- Free Radic Biol Med. 2011;51:1601-1609. induced oxidative DNA damage by a process 53. Reddy VD, Padmavathi P, Kavitha G, Saradamma involving CREB-mediated up-regulation of apur- B, Varadacharyulu N. Alcohol-induced oxida- inic endonuclease 1. J Biol Chem. 2010;285: tive/nitrosative stress alters brain mitochondrial 28191-28199. membrane properties. MolCellBiochem.2013; 66. Finkel AG, Yerry JA, Mann JD. Dietary consid- 375:39-47. erations in migraine management: Does a consist- 54. da Silva AL, Ruginsk SG, Uchoa ET, et al. ent diet improve migraine? Curr Pain Headache Time-course of neuroendocrine changes and its Rep. 2013;17:373. correlation with hypertension induced by ethanol 67. Rycerz K, Jaworska-Adamu JE. Effects of aspar- consumption. Alcohol Alcohol. 2013;48:495-504. tame metabolites on astrocytes and neurons. 55. Taivainen H, Laitinen K, Tahtel€ a€ R, Kiianmaa K, Folia Neuropathol. 2013;51:10-17. Valim€ aki€ MJ. Role of plasma vasopressin in changes 68. Lindseth GN, Coolahan SE, Petros TV, Lindseth of water balance accompanying acute alcohol intoxi- PD. Neurobehavioral effects of aspartame con- cation. Alcohol Clin Exp Res. 1995;19:759-762. sumption. Res Nurs Health. 2014;37:185-193. 56. Faraco G, Wijasa TS, Park L, Moore J, Anrather J, 69. Walton RG, Hudak R, Green-Waite RJ. Adverse Iadecola C. Water deprivation induces neurovascu- reactions to aspartame: Double-blind challenge in lar and cognitive dysfunction through vasopressin- patients from a vulnerable population. Biol Psy- induced oxidative stress. J Cereb Blood Flow chiatry. 1993;34:13-17. Metab. 2014;34:852-860. 70. Spiers PA, Sabounjian L, Reiner A, Myers DK, 57. Rains JC, Penzien DB. Precipitants of episodic WurtmanJ,SchomerDL.Aspartame:Neuropsy- migraine: Behavioral, environmental, hormonal, chologic and neurophysiologic evaluation of and dietary factors. Headache. 1996;36:274-275. acute and chronic effects. Am J Clin Nutr. 1998; 58. Blau JN. Water deprivation: A new migraine pre- 68:531-537. cipitant. Headache. 2005;45:757-759. 71. Tephly TR. The toxicity of methanol. Life Sci. 59. Bhatra MS, Gupta R, Srivastava S. Migraine 1991;48:1031-1041. associated with water deprivation and progressive 72. Ashok I, Sheeladevi R. Biochemical responses myopia. Cephalalgia. 2006;26:758-760. and mitochondrial mediated activation of apopto- 60. Martins IP, Gouveia RG. More on water and sis on long-term effect of aspartame in rat brain. migraine. Cephalalgia. 2007;27:372-374. Redox Biol. 2014;2:820-831. 28 January 2016

73. Abhilash M, Paul MVS, Varghese MV, Hair RH. 86. D’Andrea G, Nordera GP, Perini F, Allais G, Long-term consumption of aspartame and brain Granella F. Biochemistry of in antioxidant defense status. Drug Chem Toxicol. primary headaches: Focus on anomalies of tyro- 2013;36:135-140. sine metabolism. Neurol Sci. 2007;28:s94-s96. 74. Iyyaswamy A, Rathinasamy S. Effect of chronic 87. Schmidt N, Ferger B. The biogenic trace amine exposure to aspartame on oxidative stress in tyramine induces a pronounced hydroxyl radical brain discrete regions of albino rats. J Biosci. production via a monoamine oxidase dependent 2012;37:679-688. mechanism: An in vivo microdialysis study in 75. Abhilash M, Alex M, Mathews VV, Nair RH. mouse striatum. Brain Res. 2004;1012:101-107. Chronic effect of aspartame on ionic homeostasis 88. Fairbrother IS, Arbuthnott GW, Kelly JS, Butcher and monoamine in the rat SP. In vivo mechanisms underlying dopamine brain. Int J Toxicol. 2014;33:332-341. release from rat nigrostriatal terminals: II. Studies 76. Abdel-Salam OME, Salem NA, El-Shamarka using potassium and tyramine. JNeurochem.1990; MES, Hussein JS, Ahmed NAS, El-Nagar MES. 6:1844-1851. Studies on the effects of aspartame on memory 89. De Marchi U, Mancon M, Battaglia V, Ceccon S, and oxidative stress in brain of mice. Eur Rev Cardellini P, Toninello A. Influence of reactive Med Pharmacol Sci. 2012;16:2092-2101. oxygen species production by monoamine oxidase 77. Koehler SM, Glaros A. The effect of aspartame activity on aluminum-induced mitochondrial per- on migraine headache. Headache. 1988;28:10-13. meability transition. Cell Mol Life Sci. 2004;61: 78. Van Den Eeden SK, Koepsell TD, Longstreth 2664-2671. WT Jr, van Belle G, Daling JR, McKnight B. 90. Cohen G, Kesler N. Monoamine oxidase and Aspartame ingestion and headaches: A ran- mitochondrial respiration. J Neurochem. 1999;73: domized crossover trial. . 1994;44: 2310-2315. 1787-1793. 91. Grancara S, Battaglia V, Martinis P, et al. Mito- 79. Schiffman SS, Buckley CE III, Sampson HA, chondrial oxidative stress induced by Ca21 and et al. Aspartame and susceptibility to headache. monoamines: Different behaviour of liver and N Engl J Med. 1987;317:1181-1185. brain mitochondria in undergoing permeability 80. Peatfield R, Littlewood JT, Glover V, Sandler M, transition. Amino Acids. 2012;42:751-759. Rose FC. Pressor sensitivity to tyramine in 92. Hauptmann N, Grimsby J, Shih JC, Cadenas E. patients with headache: Relationship to platelet The metabolism of tyramine by monoamine oxi- monoamine oxidase and to dietary provocation. dase A/B causes oxidative damage to mitochondrial J Neurol Neurosurg . 1983;46:827-831. DNA. Arch Biochem Biophys. 1996;335:295-304. 81. D’Andrea G, Terrazzino S, Leon A, et al. Ele- 93. Gonc¸alves FM, Martins-Oliveira A, Lacchini R, vated levels of circulating trace amines in primary et al. Matrix metalloproteinase (MMP)-2 gene headaches. Neurology. 2004;62:1701-1705. polymorphisms affect circulating MMP-2 levels in 82. Kohlenberg RJ. Tyramine sensitivity in dietary patients with migraine with aura. Gene. 2013;512: migraine: A critical review. Headache. 1982;22:30- 35-40. 34. 94. Imamura K, Takeshima T, Fusayasu E, Nakashima 83. Vaughan TR. The role of food in the pathogene- K. Increased plasma matrix metalloproteinase-9 sis of migraine headache. Clin Rev Allergy. 1994; levels in migraineurs. Headache. 2008;48:135-139. 12:167-180. 95. Leira R, Sobrino T, Rodriguez-Yanez~ M, Blanco 84. Jansen SC, van Dusseldorp M, Bottema KC, M, Arias S, Castillo J. MMP-9 immunoreactivity Dubois AEJ. Intolerance to dietary biogenic amines: in acute migraine. Headache. 2007;47:698-702. Areview.Ann Allergy Asthma Immunol. 2003;91: 96. Ashina M, Tvedskov JF, Lipka K, Bilello J, 233-241. Penkowa M, Olesen J. Matrix metalloproteinases 85. Hanington E, Harper AM. The role of tyramine during and outside of migraine attacks without in the aetiology of migraine and related studies aura. Cephalalgia. 2010;30:303-310. on the cerebral and extracerebral circulations. 97. Martins-Oliveira A, Gonc¸alves FM, Speciali JG, Headache.1968;8:84-97. et al. Specific matrix metalloproteinase 9 (MMP- Headache 29

9) haplotype affect the circulating MMP-9 levels brain nuclei in migraine: A population-based in women with migraine. J Neuroimmunol. 2012; magnetic resonance imaging study. Cephalalgia. 252:89-94. 2009;29:351-359. 98. Borah A, Paul R, Mazumder MK, Bhattacharjee 110. Tepper SJ, Lowe MJ, Beall E, et al. Iron deposi- N. Contribution of b-phenylethylamine a compo- tion in pain-regulatory nuclei in episodic migraine nent of chocolate and wine, to dopaminergic neu- and chronic daily headache by MRI. Headache. rodegeneration: Implications for the pathogenesis 2012;52:236-243. of Parkinson’s disease. Neurosci Bull. 2013;29: 111. Welch KMA. Iron in the migraine brain; A resil- 655-660. ient hypothesis. Cephalalgia. 2009;29:283-285. 99. Janssen PA, Leysen JE, Megens AAHP, 112. Henderson WR, Raskin NH. “Hot dog” head- Awouters FHL. Does phenylethylamine act as an ache: Individual susceptibility to nitrite. Lancet. endogenous in some patients? Int J 1972;2:1162-1163. Neuropsychopharmacol. 1999;2:229-240. 113. Cornwell N, Clarke L. Dietary modification 100. Mazumder MK, Paul R, Borah A. b-phenylethyl- in patients with migraine and tension-type amine – a phenylalanine derivative in brain – headache. Cephalalgia. 1991;11(Suppl. 11): contributes to oxidative stress by inhibiting mito- 143-144. chondrial complexes and DT-diaphorase: An in 114. Hansen JM, Hauge AW, Ashina M, Olesen J. silico study. CNS Neurosci Ther. 2013;19:596-602. Trigger factors for familial . 101. Krymchantowski AV, Jevoux CC. Wine and Cephalalgia. 2011;31:1274-1281. headache. Headache. 2014;54:967-975. 115. Sindelar JJ, Milkowski AL. Human safety contro- 102. Panconesi A. Alcohol and migraine: Trigger factor, versies surrounding nitrate and nitrite in the diet. consumption, mechanisms. A review. JHeadache Nitric Oxide. 2012;26:259-266. Pain. 2008;9:19-27. 116. Lundberg JO, Gladwin MT, Ahluwalia A, et al. 103. Prochazkov aD,Bou sova I, Wilhelmova N. Anti- Nitrate and nitrite in biology, nutrition and thera- oxidant and prooxidant properties of flavonoids. peutics. Nat Chem Biol. 2009;5:865-869. Fitoterapia. 2011;82:513-523. 117. Lavie L. Oxidative stress in obstructive sleep 104. Hadi SM, Bhat SH, Azmi AS, Hanif S, Shamim apnea and intermittent hypoxia – Revisited – The U, Ullah MF. Oxidative breakage of cellular bad ugly and good: Implications to the heart and DNA by plant polyphenols: A putative mecha- brain. Sleep Med Rev. 2014;20:27–45. nism for anticancer properties. Semin Cancer 118. Mateo AO, de Artinano~ MAA. Nitric oxide reac- Biol. 2007;17:370-376. tivity and mechanisms involved in its biological 105. Sharma V, Joseph C, Ghosh S, Agarwal A, effects. Pharmacol Res. 2000;42:421-427. Mishra MK, Sen E. Kaempferol induces apopto- 119. Huncharek M, Kupelnick B, Wheeler L. Dietary sis in glioblastoma cells through oxidative stress. cured meat and the risk of adult glioma: A meta- Mol Cancer Ther. 2007;6:2544-2553. analysis of nine observational studies. J Environ 106. Galati G, Sabzevari O, Wilson JX, O’Brien PJ. Pathol Toxicol Oncol. 2003;22:129-137. Prooxidant activity and cellular effects of the 120. Michaud DS, Holick CN, Batchelor TT, phenoxyl radicals of dietary flavonoids and other Giovannucci E, Hunter DJ. Prospective study of polyphenolics. Toxicology. 2002;177:91-104. meat intake and dietary nitrates, nitrites, and 107. Sakihama Y, Cohen MF, Grace SC, Yamasaki H. nitrosamines and risk of adult glioma. Am J Clin Plant phenolic antioxidant and prooxidant activ- Nutr. 2009;90:570-577. ities: Phenolics-induced oxidative damage mediated 121. de la Monte SM, Tong M. Mechanisms of by metals in plants. Toxicology. 2002;177:67-80. nitrosamine-mediated neurodegeneration: Poten- 108. Welch KM, Nagesh V, Aurora SK, Gelman N. tial relevance to sporadic Alzheimer’s disease. Periaqueductal gray matter dysfunction in mig- J Alzheimers Dis. 2009;17:817-825. raine: Cause or the burden of illness? Headache. 122. Cooper SF, Lemoyne C, Gauvreau D. Identifica- 2001;41:629-637. tion and quantitation of N-nitrosamines in 109. Kruit MC, Launer LJ, Overbosch J, van Buchem human postmortem organs. J Anal Toxicol. 1987; MA, Ferrari MD. Iron accumulation in deep 11:12-18. 30 January 2016

123. Wober€ C, Brannath W, Schmidt K, et al. Pro- 136. Blau JN, Cumings, JN. Method of precipitating spective analysis of factors related to migraine and preventing migraine attacks. Br Med J. 1966; attacks: The PAMINA study. Cephalalgia. 2007; 2:1242-1243. 27:304-314. 137. Critchley M, Ferguson FR. Migraine. Lancet. 124. Martin PR, Todd J, Reece J. Effects of noise 1933;1:123-126, 182-187. and a stressor on head pain. Headache. 2005;45: 138. Dalton K. Food intake prior to a migraine attack 1353-1364. – Study of 2313 spontaneous attacks. Headache. 125. Cheng L, Wang S-H, Chen Q-C, Liao X-M. Mod- 1975;15:188-193. erate noise induced cognition impairment of mice 139. Mosek A, Korczyn AD. Yom Kippur headache. and its underlying mechanisms. Physiol Behav. Neurology. 1995;45:1953-1955. 2011;104:981-988. 140. Hockaday JM, Williamson DH, Whitty CWM. 126. Cooke LJ, Rose MS, Becker WJ. Chinook Blood-glucose levels and fatty acid metabolism in winds and migraine headache. Neurology. 2000; migraine related to fasting. Lancet. 1971;1:1153. 54:302-307. 141. Languren G, Montiel T, Julio-Amilpas A, Massieu 127. Li R, Widinmyer C, Hannigan MP. Contrast and L. Neuronal damage and cognitive impairment correlations between coarse and fine particulate associated with hypoglycemia: An integrated view. matter in the United States. Sci Total Environ. Neurochem Int. 2013;63:331-343. 2013;456-457:346-358. 142. Perros P, Deary I, Sellar R, Best JJ, Frier BM. 128. Dales RE, Cakmak S, Vidal CB. Air pollution Brain abnormalities demonstrated by magnetic and hospitalization for headache in Chile. Am J resonance imaging in adult IDDM patients with Epidemiol. 2009;170:1057-1066. and without a history of recurrent severe hypo- 129. Szyszkowich M, Stieb DM, Rowe BH. Air pollu- glycemia. Diabetes Care. 1997;20:1013-1018. tion and daily ED visits for migraine and head- 143. Cardoso S, Santos RX, Correia SC, et al. Insulin- ache in Edmonton, Canada. Am J Emerg Med. induced recurrent hypoglycemia exacerbates dia- 2009;27:391-396. betic brain mitochondrial dysfunction and oxida- 130. Lucchini RG, Guazzetti S, Zoni S, et al. , tive imbalance. Neurobiol Dis. 2012;49:1-12. olfactory and motor changes in Italian adoles- 144. Suh SW, Gum ET, Hamby AM, Chan PH, Swanson cents exposed to historical ferro-manganese emis- RA.Hypoglycemicneuronal death is triggered by sion. Neurotoxicology. 2012;33:687-696. glucose reperfusion and activation of neuronal 131. Calderon-Garcidue nas~ L, D’Angiulli A, Kulesza NADPH oxidase. J Clin Invest. 2007;117:910-918. RJ, et al. Air pollution is associated with brain- 145. Schoonman GG, Sandor PS, Agosti RM, et al. stem auditory nuclei pathology and delayed Normobaric hypoxia and nitroglycerin as trigger brainstem auditory evoked potentials. Int J Dev factors for migraine. Cephalalgia. 2006;26:816-819. Neurosci. 2011;29:365-375. 146. Magalhaes J, Ascensao A, Soares JM, et al. 132. Lucchini RG, Dorman DC, Elder A, Veronesi B. Acute and severe hypobaric hypoxia increases oxi- Neurological impacts from inhalation of pollu- dative stress and impairs mitochondrial function in tants and the nose-brain connection. Neurotoxi- mouse skeletal muscle. JApplPhysiol.2005;99: cology. 2012;33:838-841. 1247-1253. 133. Sama P, Long TC, Hester S, et al. The cellular 147. Li H, Forstermann U. Uncoupling of endothelial and genomic response of an immortalized micro- NO synthase in atherosclerosis and vascular dis- glia cell line (BV2) to concentrated ambient par- ease. Curr Opin Pharmacol. 2013;13:161-167. ticulate matter. Inhal Toxicol. 2007;19:1079-1087. 148. Qin L, Wu X, Block ML, et al. Systemic LPS 134. Veronesi B, Makwana O, Pooler M, Chen L. causes chronic neuroinflammation and progres- Effects of subchronic exposure to concentrated sive neurodegeneration. Glia. 2007;55:453-462. ambient particles. VII. Degeneration of dopami- 149. Chai NC, Peterlin BL, Calhoun AH. Migraine nergic neurons in Apo E -/- mice. Inhal Toxicol. and estrogen. Curr Opin Neurol. 2014;27:315-324. 2005;17:235-241. 150. MacGregor EA. Oestrogen and attacks of 135. Jacome DE. Hypoglycemia rebound migraine. migraine with and without aura. Lancet Neurol. Headache. 2001;41:895-898. 2004;3:354-361. Headache 31

151. Vermeer LMM, Gregory E, Winter MK, 162. Spierings ELH. Migraine prevention and errors McCarson KE, Berman NEJ. Behavioral effects in living: Dr. Graham’s lessons for patients and and mechanisms of migraine pathogenesis follow- physicians. Headache. 2002;42:152-153. ing estradiol exposure in a multibehavioral model 163. Cudney LE, Sassi RB, Behr GA, et al. Altera- of migraine in rat. Exp Neurol. 2015;263:8-16. tions in circadian rhythms are associated with 152. Behl, C. Oestrogen as a neuroprotective hor- increased lipid peroxidation in females with bipo- mone. Nat Rev Neurosci. 2002;3:433-442. lar disorder. Int J Neuropsychopharmacol. 2014; 153. Santanam N, Shern-Brewer R, McClatchey R, 17:715-722. et al. Estradiol as an antioxidant: Incompatible 164. Videnovic A, Lazar AS, Barker RA, Overeem S. with its physiological concentrations and function. ‘The clocks that time us’ – circadian rhythms in J Lipid Res. 1998;39:2111-2118. neurodegenerative disorders. Nat Rev Neurol. 154. Cornelli U, Belcaro G, Cesarone MR, Finco A. 2014;10:683-693. Analysis of oxidative stress during the menstrual 165. Zelinski EL, Deibel SH, McDonald RJ. The trou- cycle. Reprod Biol Endocrinol. 2013;11:74. doi: ble with circadian clock dysfunction: Multiple 10.1186/1477-7827-11-74. deleterious effects on the brain and body. Neuro- 155. Schisterman EF, Gaskins AJ, Mumford SL, et al. sci Biobehav Rev. 2014;40:80-101. Influence of endogenous reproductive hormones 166. Zelinski EL, Tyndall AV, Hong NS, McDonald

on F2-isoprostane levels in premenopausal women: RJ. Persistent impairments in hippocampal, dor- The BioCycle Study. Am J Epidemiol. 2010;172: sal striatal, and prefrontal cortical function fol- 430-439. lowing repeated photoperiod shifts in rats. Exp 156. Rao AK, Dietrich AK, Ziegler YS, Nardulli AM. Brain Res. 2013;224:125-139. 17beta-estradiol-mediated increase in Cu/Zn 167. Alkadhi K, Zagaar M, Alhaider I, Salim S, Aleisa superoxide dismutase expression in the brain: A A. Neurobiological consequences of sleep depriva- mechanism to protect neurons from ischemia. tion. Curr Neuropharmacol. 2013;11:231-249. J Steroid Biochem Mol Biol. 2011;127:382-389. 168. Gibson EM, Wang C, Tjho S, Khattar N, 157. Lee SY, Andoh T, Murphy DL, Chiueh CC. Kriegsfeld LJ. Experimental ‘jet-lag’ inhibits adult 17beta-estradiol activates ICI 182,780-sensitive neurogenesis and produces long-term cognitive estrogen receptors and cyclic GMP-dependent thi- deficits in female hamsters. PLoS One. 2010;5: oredoxin expression for . FASEB e15267. J. 2003;17:947-948. 169. Wright RL, Lightner EN, Harman JS, Meijer 158. Nilsen J, Diaz Brinton R. Mechanism of OC, Conrad CD. Attenuating corticosterone lev- estrogen-mediated neuroprotection: Regulation els on the day of memory assessment prevents of mitochondrial calcium and Bcl-2 expression. chronic stress-induced impairments in spatial Proc Natl Acad Sci USA. 2003;100:2842-2847. memory. Eur J Neurosci. 2006;24:595-605. 159. Brown CM, Mulcahey TA, Filipek NC, Wise PM. 170. Panda S, Antoch MP, Miller BH, et al. Coordi- Production of proinflammatory cytokines and nated transcription of key pathways in the mouse chemokines during neuroinflammation: Novel by the circadian clock. Cell. 2002;109:307-320. roles for estrogen receptors alpha and beta. Endo- 171. Kondratov RV, Vykhovanets O, Kondratova AA, crinology. 2010;151:4916-4925. Antoch MP. Antioxidant N-acetyl-L-cysteine amel- 160. Numakawa Y, Matsumoto T, Yokomaku D, iorates symptoms of premature aging associated et al. 17b-estradiol protects cortical neurons with the deficiency of the circadian clock protein against oxidative stress-induced BMAL1. Aging. 2009;1:979-987. through reduction in the activity of mitogen- 172. D’Almeida V, Lobo LL, Hipolide DC, de Oliveira activated protein kinase and in the accumulation AC, Nobrega JN, Tufilk S. Sleep deprivation indu- of intracellular calcium. Endocrinology. 2007; ces brain region-specific decreases in glutathione 148:627-637. levels. NeuroReport. 1998;9:2853-2856. 161. Suzuki S, Brown CM, Wise PM. Neuroprotective 173. Alzoubi KH, Khabour OF, Rashid BA, Damaj effects of estrogens following ischemic . IM, Salah HA. The neuroprotective effect of Front Neuroendocrinol. 2009;30:201-211. vitamin E on chronic sleep deprivation-induced 32 January 2016

memory impairment: The role of oxidative stress. networks. J Cereb Blood Flow Metab. 2014;34: Behav Brain Res. 2012;226:205-210. 1270-1282. 174. Living E. Observations on megrim or sick-head- 186. Wober€ C, Holzhammer J, Zeitlhofer J, Wessely P, ache. Br Med J. 1872;1:364-366. Wober-Bing€ ol€ C. Trigger factors of migraine and 175. Saenz-de-Viteri M, Heras-Mulero H, Fernandez- tension-type headache: Experience and knowledge Robredo P, et al. Oxidative stress and histologi- of the patients. J Headache Pain. 2006;7:188-195. cal changes in a model of retinal phototoxicity in 187. Martin PR, Teoh H-J. Effects of visual stimuli and rabbits. Oxid Med Cell Longev. 2014;2014:637137. astressoronheadpain.Headache. 1999;39:705-715. 176. Siu TL, Morley JW, Coroneo MT. Toxicology of 188. Haynes SN, Gannon LR, Bank J, Shelton D, the retina: Advances in understanding the Goodwin J. Cephalic blood flow correlates of defence mechanisms and pathogenesis of drug- induced headaches. JBehavMed.1990;13: and light-induced retinopathy. Clin Experiment 467-480. Ophthalmol. 2008;36:176-185. 189. Moller€ M, Du Preez JL, Emsley R, Harvey BH. 177. Fetoni AR, De Bartolo P, Eramo SLM, et al. Isolation rearing-induced deficits in sensorimotor Noise-induced hearing loss (NHL) as a target of gating and social interaction in rats are related to oxidative stress-mediated damage: Cochlear and cortico-striatal oxidative stress, and reversed by cortical responses after an increase in antioxidant sub-chronic clozapine administration. Eur Neuro- defense. J Neurosci. 2013;33:4011-4023. psychopharmacol. 2011;21:471-483. 178. Rokyta R, Holocek V, Pekarkova I, et al. Free 190. Schiavone S, Sorce S, Dubois-Dauphin M, et al. radicals after painful stimulation are influenced Involvement of NOX2 in the development of by antioxidants and analgesics. Neuro Endocrinol behavioral and pathologic alterations in isolated Lett. 2003;24:304-309. rats. Biol Psychiatry. 2009;66:384-392. 179. Meiser J, Weindl D, Hiller K. Complexity of 191. Nayernia Z, Jaquet V, Krause K-H. New insights dopamine metabolism. Cell Commun Signal. on NOX enzymes in the . 2013;11:34. Antioxid Redox Signal. 2014;20:2815-2837. 180. Crone NE, Korzeniewska A, Franaszczuk PJ. 192. Eisenberger NI, Lieberman MD. Why rejection Cortical gamma responses: Searching high and hurts: A common neural alarm system for physical low. Int J Psychophysiol. 2011;79:9-15. and social pain. Trends Cogn Sci. 2004;8:294-300. 181. Uhlhaas PJ, Pipa G, Neuenschwander S, Wibral 193. Kozorovitskiy Y, Gould E. hierarchy M, Singer W. A new look at gamma? High- (>60 influences adult neurogenesis in the dentate Hz) c-band activity in cortical networks: Func- gyrus. J Neurosci. 2004;24:6755-6759. tion, mechanisms and impairment. Prog Biophys 194. Magarinos~ A, McEwen BS. Stress-induced atro- Mol Biol. 2011;105:14-28. phy of apical dendrites of hippocampal CA3c 182. Coppola G, Ambrosini A, Di Clemente L, et al. neurons: Comparison of stressors. . Interictal abnormalities of gamma band activity 1995;69:83-88. in visual evoked responses in migraine: An indi- 195. Holm JE, Lokken C, Myers TC. Migraine and cation of thalamocortical dysrhythmia? Cephalal- stress: A daily examination of temporal relation- gia. 2007;27:1360-1367. ships in women migraineurs. Headache. 1997;37: 183. Niessing J, Ebisch B, Schmidt KE, Niessing M, 553-558. Singer W, Galuske RAW. Hemodynamic signals 196. Kohler T, Haimerl C. Daily stress as a trigger of correlates tightly with synchronized gamma oscil- migraine attacks: Results of thirteen single- lations. Science. 2005;309:948-951. subject studies. J Consult Clin Psychol. 1990;58: 184. Lord L-D, Expert P, Huckins JF, Turkheimer FE. 870-872. Cerebral energy metabolism and the brain’s func- 197. Houle TT, Butschek RA, Turner DP, Smitherman tional network architecture: An integrative review. TA, Rains JC, Penzien DB. Stress and sleep dura- J Cereb Blood Flow Metab. 2013;33:1347-1354. tion predict headache severity in chronic headache 185. Kann O, Papageorgiou IE, Draguhn A. Highly sufferers. Pain. 2012;153:2432-2440. energized inhibitory interneurons are a central 198. Thomsen LL, Olesen JJ. Nitric oxide theory of element for information processing in cortical migraine. Clin Neurosci. 1998;5:28-33. Headache 33

199. Wink DA, Hines HB, Cheng RYS, et al. Nitric 213. Akopian AN. Regulation of nociceptive transmis- oxide and redox mechanisms in the immune sion at the periphery via TRPA1-TRPV1 interac- response. J Leukoc Biol. 2011;89:873-891. tions. Curr Pharm Biotechnol. 2011;12:89-94. 200. Munzel€ T, Daiber A, Gori T. More answers to 214. Salminen LE, Paul RH. Oxidative stress and the still unresolved question of nitrate tolerance. genetic markers of suboptimal antioxidant defense Eur Heart J. 2013;34:2666-2673. in the : A theoretical review. Rev Neu- 201. Otto A, Fontaine J, Tschirhart E, Fontaine D, rosci. 2014;25:805-819. Berkenboom G. Rosuvastatin treatment protects 215. Fernandez-de-las-Pe nas~ C, Palacios-Cena~ D, against nitrate-induced oxidative stress in eNOS Salom-Moreno J, et al. Has the prevalence of knockout mice: Implication of the NAD(P)H oxi- migraine changed over the last decade (2003-2012)? dase pathway. Br J Pharmacol. 2006;148:544-552. A Spanish population-based survey. PLoS One. 202. Doing X-X, Wang Y, Qin Z-H. Molecular mech- 2014;9:e110530. doi:10.1371/journal.pone.0110530. anisms of excitotoxicity and their relevance to 216. Freitag FG. Why do migraines often decrease as pathogenesis of neurodegenerative diseases. Acta we age? Curr Pain Headache Rep. 2013;17:366. Pharmacol Sin. 2009;30:379-387. 217. Materazzi S, Benemei S, Fusi C, et al. Partheno- 203. Skulachev VP. Mitochondrial physiology and lide inhibits nociception and neurogenic vasodila- pathology: Concepts of programmed death of tion in the trigeminovascular system by targeting organelles, cells and organisms. Mol Aspects Med. the TRPA1 channel. Pain. 2013;154:2750-2758. 1999;20:139-184. 218. Barbanti P, Fofi L, Aurilia C, Egeo G. Dopami- 204. Grinberg YY, van Drongelen W, Kraig RP. Insu- nergic symptoms in migraine. Neurol Sci. 2013; lin-like growth factor-1 lowers spreading depres- 34(Suppl. 1):s67-s70. sion susceptibility and reduces oxidative stress. 219. Van Hilten JJ. The migraine-dopamine link: Do J Neurochem. 2012;122:221-229. migraine and Parkinson’s disease coexist? Clin 205. Shatillo A, Koroleva K, Giniatullina R, et al. Neurol Neurosurg. 1992;94(Suppl):s168-s170. Cortical spreading depression induces oxidative 220. Barbanti P, Fabbrini G, Vanacore N, et al. Dopa- stress in the trigeminal nociceptive system. Neu- mine and migraine: Does Parkinson’s disease mod- roscience. 2013;253:341-349. ify migraine course? Cephalalgia. 2000;20:720-723. 206. Choudhuri R, Cui L, Yong C, et al. Cortical 221. Chong CD, Dodick DW, Schlagger BL, Schwedt spreading depression and gene regulation: Rele- TJ. Atypical age-related cortical thinning in epi- vance to migraine. Ann Neurol. 2002;51:499-506. sodic migraine. Cephalalgia. 2014;34:1115-1124. 207. Waldbaum S, Patel M. Mitochondrial dysfunction 222. Ziaei S, Kazemnejad A, Sedighi A. The effect of and oxidative stress: A contributing link to vitamin E on the treatment of menstrual acquired ? J Bioenerg Biomebr. 2010;42: migraine. Med Sci Monit. 2009;15:CR16-19. 449-455. 223. D’Andrea G, Bussone G, Allais G, et al. Efficacy 208. Turner DP, Smitherman TA, Martin VT, Penzien of ginkgolide B in the prophylaxis of migraine DB, Houle TT. Causality and headache triggers. with aura. Neurol Sci. 2009;30(Suppl. 1):s121-s124. Headache. 2013;53:628-635. 224. Alstadhaug KB, Odeh F, Salvesen R, Bekkelund SI. 209. Houle TT, Turner DP. Natural experimentation Prophylaxis of migraine with melatonin: A random- is a challenging method for identifying headache ized controlled trial. Neurology. 2010;75:1527-1532. triggers. Headache. 2013;53:636-643. 225. Lipton RB, Gobel H, Einhaupl KM, et al. Peta- 210. Goadsby PJ, Silberstein SD. Migraine triggers: sites hybridus root (butterbur) is an effective pre- Harnessing the messages of clinical practice. Neu- ventive treatment for migraine. Neurology. 2004; rology. 2013;80:424-425. 63:2240-2244. 211. Hufnagl KN, Peroutka SJ. Glucose regulation in 226. Murphy JJ, Heptinstall S, Mitchell JRA. headache: Implications for dietary management. Randomized double-blind placebo-controlled trial Expert Rev Neurother. 2002;2:311-317. of Feverfew in migraine prevention. Lancet. 1988; 212. Giffin NJ, Ruggiero L, Lipton RB, et al. Premon- 2:189-192. itory symptoms in migraine: An electronic diary 227. Palevitch D, Earon G, Carasso R. Feverfew study. Neurology. 2003;60:935-940. (Tanacetum parthenium) as a prophylactic treat- 34 January 2016

ment for migraine: A double-blind placebo-con- 240. Khaper N, Rigatto C, Seneviratne C, Li T, Singal trolled study. Phytother Res. 1997;11:508-511. PK. Chronic treatment with propranolol induces 228. Sandor PS, Di Clemente L, Coppola G, et al. antioxidant changes and protects against Efficacy of coenzyme Q10 in migraine prophy- ischemia-reperfusion injury. J Mol Cell Cardiol. laxis: A randomized controlled trial. Neurology. 1997;29:3335-3344. 2005;64:713-715. 241. Cardenas-Rodr ıguez N, Coballase-Urrutia E, 229. Magis D, Ambrosini A, Sandor P, Jacquy J, Huerta-Gertrudis B, et al. Antioxidant activity of Laloux P, Schoenen J. A randomized double- topiramate: An antiepileptic agent. Neurol Sci. blind placebo-controlled trial of thioctic acid in 2013;34:741-747. migraine prophylaxis. Headache. 2007;47:52-57. 242. Cardenas-Rodr ıguez N, Coballase-Urrutia E, 230. Chayasirisobhon S. Efficacy of pinus radiata bark Rivera-Espinosa L, et al. Modulation of antioxidant extract and vitamin C combination product as a pro- enzymatic activities by certain antiepileptic drugs phylactic therapy for recalcitrant migraine and long- (valproic acid, oxcarbazepine, and topiramate): Evi- term results. Acta Neurol Taiwan. 2013;22:13-21. dence in humans and experimental models. Oxid 231. Bali L, Callaway E. Vitamin C and migraine: A Med Cell Longev. 2013;2013:598493. Available at: case report. N Engl J Med. 1978;299:364. http://dx.doi.org/10.1155/2013/598493. 232. But€ un€ A, Naziroglu M, Demirci S, C¸elik O, Uguz 243. Moreno-Fernandez AM, Cordero MD, Garrido- AC. Riboflavin and vitamin E increase brain cal- Maraver J, et al. Oral treatment with amitripty- cium and antioxidants, and microsomal calcium line induces coenzyme Q deficiency and oxidative ATP-ase values in rat headache models induced by stress in psychiatric patients. J Psychiatr Res. glyceryl trinitrate. J Membr Biol. 2015;248:205-213. 2012;46:341-345. 233. Ashoori M, Saedisomeolia A. Riboflavin (vitamin 244. Pavlovic JM, Buse DC, Sollars M, Haut S, Lipton

B2) and oxidative stress: A review. Br J Nutr. RB. Trigger factors and premonitory features of 2014;111:1985-1991. migraine attacks: Summary of studies. Headache. 234. Camiletti-Moiron D, Aparicio VA, Aranda P, 2014;54:1670-1679. Radak Z. Does exercise reduce brain oxidative 245. Allais G, D’Andrea G, Maggio M, Benedetto C. stress? A systematic review. Scand J Med Sci The efficacy of ginkgolide B in the acute treat- Sports. 2013;23:e202-e212. ment of migraine aura: An open preliminary trial. 235. Overath CH, Darabaneanu S, Evers MC, et al. Neurol Sci. 2013;34(Suppl. 1):s161-s163. Does an aerobic endurance programme have an 246. Cady RK, Goldstein J, Nett R, Mitchell R, influence on information processing in migrain- Beach ME, Browning R. A double-blind placebo- eurs? J Headache Pain. 2014;15:11. controlled pilot study of sublingual feverfew and 236. Campbell JK, Penzien DB, Wall EM. Evidence- ginger (LipiGesicTM M) in the treatment of based guidelines for migraine headache: Behav- migraine. Headache. 2011;51:1078-1086. ioral and physical treatments. 2000. Available 247. Ikeda Y, Jimbo H, Shimazu M, Satoh K. Suma- at: http://tools.aan.com/professionals/practice/pdfs/ triptan scavenges superoxide, hydroxyl, and nitric g10089.pdf. oxide radicals: In vitro electron spin resonance 237. Ciancarelli I, Tozzi-Ciancarelli MG, Spacca G, Di study. Headache. 2002;42:888-892. Massimo C, Carolei A. Relationship between bio- 248. Read SJ, Manning P, McNeil CJ, Hunter AJ, feedback and oxidative stress in patients with Parsons AA. Effects of sumatriptan on nitric chronic migraine. Cephalalgia. 2007;27:1136-1141. oxide and superoxide balance during glyceryl tri- 238. Ciancarelli I, Tozzi-Ciancarelli MG, Di Massimo nitrate infusion in the rat. Implications for anti- C, Marini C, Carolei A. Flunarizine effects on migraine mechanisms. Brain Res. 1999;847:1-8. oxidative stress in migraine patients. Cephalalgia. 249. Cortelli P, Montagna P. Migraine as visceral 2004;24:528-532. pain. Neurol Sci. 2009;30(Suppl. 1):s19-s22. 239. Kubo K, Yoshitake I, Kumada Y, Shuto K, 250. Zhou Z, Hu C-P, Wang C-J, Li T-T, Peng J, Li Nakamizo N. Radical scavenging action of flunar- Y-J. Calcitonin gene-related peptide inhibits izine in rat brain in vitro. Arch Int Pharmacodyn angiotensin II-induced endothelial progenitor Ther. 1984;272:283-295. cells senescence through up-regulation of Headache 35

klotho expression. Atherosclerosis. 2010;213:92- 256. Raskin NH, Hosobuchi Y, Lamb S. Headache 101. may arise from perturbation of brain. Headache. 251. Fei S, Wei S, Jie-Ming M, Xian W. Calcitonin gene- 1987;27:416-420. related peptide therapy suppresses reactive oxygen 257. Azmitia EC. Serotonin neurons, , species in the pancreas and prevents mice from auto- and homeostasis of neural tissue. Neuropsycho- immune diabetes. Acta Physiol Sin. 2003;55:625-632. pharmacology. 1999;21(Suppl. 1):s33-s45. 252. Schaeffer C, Vandroux D, Thomassin L, Athias 258. Azmitia EC. Serotonin and brain: Evolution, neu- P, Rochette L, Connat J-L. Calcitonin gene- roplasticity, and homeostasis. Int Rev Neurobiol. related peptide partly protects cultured smooth 2007;77:31-56. muscle cells from apoptosis induced by an oxida- 259. Niciu MJ, Ionescu DF, Mathews DC, Richards tive stress via activation of ERK1/2 MAPK. Bio- EM, Zarate CA Jr. Second messenger/signal chim Biophys Acta. 2003;1643:65-73. transduction pathways in major mood disorders: 253. Halliwell B, Gutteridge JM. The antioxidants of Moving from membrane to mechanism of action, human extracellular fluids. Arch Biochem Bio- part I: Major depressive disorder. CNS Spectr. phys. 1990;280:1-8. 2013;18:231-241. 254. Medina JM, Tabernero A. -synthesized 260. Danese F, Montagnana M, Lippi G. Platelets and oleic acid behaves as a neurotrophic factor for migraine. Thromb Res. 2014;134:17-22. neurons. J Physiol Paris. 2002;96:265-271. 261. Hayon Y, Dashevsky O, Shai E, Varon D, Leker 255. Gawel M, Burkitt M, Rose FC. The platelet RR. Platelet microparticles promote neural stem release reaction during migraine attacks. Head- cell proliferation, survival, and differentiation. ache. 1979;19:323-327. J Mol Neurosci. 2012;47:659-665.