Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

The metabolic face of migraine — from pathophysiology to treatment


Migraine can be regarded as a conserved, adaptive response that occurs in genetically predisposed individuals with a mismatch between the brain’s energy reserve and workload. Given the high prevalence of migraine, genotypes associated with the condition seem likely to have conferred an evolutionary advantage. Technological advances have enabled the examination of different aspects of cerebral metabolism in patients with migraine, and complementary animal research has highlighted possible metabolic mechanisms in migraine pathophysiology. An increasing amount of evidence — much of it clinical — suggests that migraine is a response to cerebral energy deficiency or oxidative stress levels that exceed antioxidant capacity and that the attack itself helps to restore brain energy homeostasis and reduces harmful oxidative stress levels. Greater understanding of metabolism in migraine offers novel therapeutic opportunities. In this Review, we describe the evidence for abnormalities in energy metabolism and mitochondrial function in migraine, with a focus on clinical data (including neuroimaging, biochemical, genetic and therapeutic studies), and consider the relationship of these abnormalities with the abnormal sensory processing and cerebral hyper-responsivity observed in migraine. We discuss experimental data to consider potential mechanisms by which metabolic abnormalities could generate attacks. Finally, we highlight potential treatments that target cerebral metabolism, such as nutraceuticals, ketone bodies and dietary interventions.

Key points

  • Prevalent triggers of migraine attacks can all be linked to unbalanced cerebral energy metabolism and/or oxidative stress.

  • Magnetic resonance spectroscopy studies have shown that mitochondrial phosphorylation potential and ATP are decreased in the brains of people with migraine between attacks. Glucose (and lipid) metabolism and mitochondrial functions are abnormal in the peripheral blood.

  • Among patients with migraine, various single nucleotide polymorphisms are present in non-coding mitochondrial DNA and nuclear-encoded mitochondrial proteins; common variants associated with migraine are functionally involved in mitochondrial metabolism.

  • Metabolic enhancers, such as riboflavin and coenzyme Q10, and dietary or pharmacological ketogenesis improve migraine but novel, more efficient metabolic strategies are needed.

  • Experimental studies indicate a link between cerebral energy disequilibrium and cortical spreading depression and/or trigeminovascular system activation; calcitonin gene-related peptide and pituitary adenylate cyclase-activating peptide could also help restore energy homeostasis.

  • Migraine can be regarded as a conserved, adaptive response that occurs in individuals with a genetic predisposition and a mismatch between the brain’s energy reserve and workload.


As early as 1935, migraine — which now affects >15% of the population worldwide — was referred to as a ‘hypoglycaemic headache’1. Despite this early connection between migraine and energy metabolism, clinical and basic research in migraine largely focused on the vasculature, neurovasculature and neurotransmission until Willem Amery revived the idea that metabolism is involved in the pathogenesis of migraine in his hypothesis-generating review in 1982 (ref.2). Since then, accumulating evidence — much of it clinical — indicates that migraine is at least partially an energy deficit syndrome with mitochondrial dysfunction. Technological advances (for example, in neuroimaging and genetics) have enabled examination of different aspects of cerebral metabolism in patients with migraine, and complementary animal research has deciphered possible links between metabolic factors and trigeminovascular activation in migraine pathophysiology. Evidence that cortical responsivity and sensory processing are abnormal in patients with migraine between attacks (reviewed elsewhere3) led to the suggestion that a combination of sensory overload and lowered energy reserve ignites the major pain-signalling system of the brain, the trigeminovascular system, leading to the migraine attack4.

In this Review, we describe the abnormalities of energy metabolism observed in migraine with a particular focus on clinical data, including phenotypic, biochemical, genetic and therapeutic studies. We also discuss experimental data to elaborate on the potential role of such metabolic abnormalities in migraine attack generation. Finally, we highlight therapeutic approaches to targeting of cerebral metabolism (antioxidants, nutraceuticals, pharmaceuticals and dietary ketogenesis).

Triggers and metabolic dysfunction

Two systematic reviews5,6 and a study of 1,207 patients with migraine7 have identified that the most common migraine trigger factors are stress or subsequent relaxation, fasting or skipping a meal, sleep changes (too much or too little), ovarian hormone changes (including menstruation and oral contraception), weather changes (including certain winds, hypoxia and high altitude), physical exercise (including sexual activity), alcohol, strong odours (especially perfume or cigarette smoke), intense light (especially bright or blue light) and loud noises. The distinction between trigger factors and premonitory symptoms of migraine attacks is not always easy, as some premonitory symptoms might be misinterpreted8. For example, premonitory photophobia could lead to interpretation of light as a trigger. Similarly, as a result of premonitory craving for sweets, chocolate is frequently mistaken as a trigger. Nevertheless, individual triggers seem to have an additive effect9, leading to an attack only when a threshold has been reached. This observation suggests that trigger factors act on common pathways.

Some triggers — such as skipping a meal or fasting, exercise, dehydration, hypoxia and lack of sleep — have a clear link to metabolism. However, many other triggers, including hormonal changes, also have a potential common metabolic denominator: changes in mitochondrial metabolism and/or oxidative stress10. For instance, intense physical11,12 or psychological stress can increase oxidative stress in the CNS13. In healthy people, one night of sleep deprivation is enough to substantially reduce levels of glutathione, ATP, cysteine and homocysteine14. Intense sensory stimuli, including odours15, perfumes containing phthalates16, blue light17 and loud noises18, can increase oxidative stress. Experimental hypoxia can trigger migraine headaches8,19 (less so migraine aura19), even in most healthy people20, and in line with this observation, migraine prevalence is higher in populations that live at high altitude21.

In animals, alcohol-induced oxidative or nitrosative stress alters mitochondrial membrane properties in the brain22. In rodents, oestrogen and, to a lesser extent, progesterone increase susceptibility to cortical spreading depression (CSD)23, the cause of migraine aura. These hormones also modulate the ability of 5-hydroxytryptophan (the precursor of serotonin) to inhibit CSD24, and influence oxidative metabolism in the rat brain25. Furthermore, oestrogen, which greatly modulates the course of migraine in females, is involved in insulin sensitivity, the regulation of insulin secretion and nutrient homeostasis26. In combination, these observations show that most migraine triggers or aggravating factors have a link to energy metabolism and oxidative stress.

Biochemical studies

A large number of biochemical studies in migraine point towards a variety of different metabolic abnormalities, discussed below, all of which are related to energy homeostasis (Supplementary Tables 1 and 2). A combination of metabolic and endocrinological abnormalities, possibly together with abnormal cerebral responsivity, are likely to determine the migraine attack threshold of an individual. The cumulative number of abnormalities, in combination with unfavourable environmental factors, is likely to determine disease severity6.

Oxidative phosphorylation, ATP and lactate

Magnetic resonance spectroscopy (MRS) enables non-invasive measurement of numerous substances in various tissues. Some of these substances, such as lactate, magnesium and ATP, provide pivotal information about energy metabolism, and this approach has been used in studies of migraine (Supplementary Table 1).

The use of 31P-MRS has shown that mitochondrial oxidative phosphorylation is impaired in the brain of patients with migraine during27 and between migraine attacks28,29,30,31,32,33,34. This impairment is seen as increased levels of ADP, decreased levels of organic phosphate and a decreased phosphorylation potential. Similar patterns have been observed in skeletal muscles28,35,36, suggesting a generalized rather than brain-specific alteration (reviewed in detail elsewhere33,37). Subsequently, a modified 31P-MRS methodology was used to directly quantify brain ATP, which was found to be decreased by 16% between attacks in patients with migraine without aura compared with healthy controls38. The lowest ATP concentrations were detected in the most severely affected patients, a finding that agrees in part with those of other studies showing modest associations between brain hypometabolism and attack frequency31,35,38. Magnesium is often also measured in 31P-MRS studies of neural metabolism because it is a crucial cofactor for ATP production. These measurements have shown that, in line with alterations in oxidative phosphorylation, cytosolic free magnesium is reduced in the occipital lobes of patients with migraine30,31,39.

The more widely available 1H-MRS technique can be used to determine concentrations of lactate, a key cellular metabolite. Variability in methodologies and patient selection criteria in studies of brain lactate levels in patients with migraine mean that strong conclusions cannot be drawn33,37. Elevated levels of brain lactate have been found in patients with migraine with aura40,41 but not in those with migraine without aura42,43,44,45. Occipital baseline lactate levels were increased in patients who had strictly visual aura compared with healthy controls but not in those who had complex neurological auras; lactate levels increased considerably during photic stimulation in the latter group of patients but not in the former group40. An important consideration is that stimulus-induced increases in cortical lactate levels are physiological46 and are explained by the astrocyte-to-neuron lactate shuttle47, the mechanism by which astrocytes provide energy to neurons when they become activated. Hence, the absence of a stimulus-induced increase in lactate levels in patients with migraine could be considered pathological, as it might render them vulnerable to an energy crisis, particularly because neuronal activation is likely to have a higher energy demand in patients with migraine than in healthy individuals because their sensory information processing is abnormal48. A study that combines quantification of lactate in the cortex and electrophysiological testing of brain-evoked responses would be able to clarify this relationship between function and metabolism.

Another useful method for assessing energy dynamics in the brain is 18F-fluorodeoxyglucose (18F-FDG) PET. This procedure involves measuring positron emission from the systemically administered, radiotracer-labelled glucose analogue 18F-FDG that is taken up by metabolically active tissue, although this technique does not allow glucose uptake by neurons to be distinguished from that by non-neuronal cells such as astrocytes. In a study published in 2018, the use of 18F-FDG-PET to compare glucose uptake at rest with that during interictal neuronal activation (visual evoked potentials) showed that glucose uptake during interictal neuronal activation exceeded glucose uptake in visual processing areas in 90% of patients with migraine without aura, but in only 15% of healthy controls49. Given that at least 50% of glucose taken up in the brain goes to the astrocytes, where energy is stored in the CNS, this observation suggests that energy reserves are reduced in patients with migraine, thereby supporting our hypothesis that a mismatch between brain activation and glucose metabolism is a cornerstone of migraine pathophysiology. Further studies are needed to confirm this hypothesis.

Peripheral metabolic abnormalities

Several metabolic abnormalities in peripheral tissues have been described (Supplementary Table 2). These abnormalities include alterations in mitochondrial enzyme function, oxidative stress and glucose metabolism.

Mitochondrial enzyme function

Evidence that generalized metabolic dysfunction is a feature of migraine comes from studies showing that activity of mitochondrial enzymes, such as monoamine oxidase, succinate dehydrogenase, NADH dehydrogenase, cyclooxygenase and citrate synthetase, is reduced in the platelets of patients with migraine with or without aura50,51. Interestingly, these biochemical changes are restricted to enzymes of the respiratory chain that are encoded by mitochondrial DNA (mtDNA), which is more vulnerable than nuclear DNA to oxidative stress and is, like migraine, chiefly maternally inherited. This observation suggests that both inherited and acquired mtDNA abnormalities play a role in migraine pathophysiology (see Genetic studies below). Elevated lactate and pyruvate levels in the plasma provide further evidence for mitochondrial abnormalities in migraine, although these abnormalities are mostly found in patients with migrainous stroke52,53.

Oxidative and nitrosative stress and antioxidant capacity

As discussed above, all common migraine triggers are likely to increase levels of oxidative stress. This link is supported by studies in which markers of increased oxidative or nitrosative stress and/or decreased antioxidant capacity have been directly examined54,55,56,57,58,59,60,61,62,63,64,65,66,67 (Supplementary Table 2). All studies of oxidative stress and/or antioxidant capacity in migraine have demonstrated that at least one marker is abnormal54,55,56,57,58,59,60,61,62,63,64,65,66,67. Of all biomarkers examined, superoxide dismutase activity seems to be the only one that is consistently reduced in patients with migraine, including interictally68. Inconsistent results for other markers could be due to differences in methodology, patient selection criteria (for example, low-frequency versus high-frequency migraine, or migraine with aura versus migraine without aura) and variations related to the migraine cycle — for example, nitrosative stress, oxidative stress67 and nitric oxide68 are elevated during migraine attacks, but not interictally.

Another possible marker of oxidative stress is heavy metals, levels of which can be increased in migraine69. For example, free iron is highly pro-oxidant and accumulates in the brainstem of patients with migraine in proportion to disease duration70.

Glucose metabolism

The human brain is highly dependent on energy sources from the circulation owing to limited glycogen stores, high energy needs and the exclusion of large, energy-dense molecules by the blood–brain barrier, so is particularly vulnerable to their shortages. Hypoglycaemia has been associated with migraine for almost a century1,71,72, and a simple comparison between migraine-associated symptoms — premonitory symptoms in particular — and symptoms of hypoglycaemia73 reveals several similarities. Shared symptoms include dizziness, pale skin, cold hands and feet, binge eating and/or sugar cravings, yawning, nausea, low blood pressure, shaking, cognitive difficulties, tiredness, fatigue, visual dysfunction and slurred speech. All of these symptoms can be caused by an insufficient supply of glucose to the brain and/or by release of catecholamines as a result of sympathetic activation73.

The hypothalamus controls homeostasis and is activated early during the premonitory phase of triggered and spontaneous migraine attacks74,75,76. This activation could represent the underlying physiological correlate of premonitory symptoms or could be part of an adaptive behavioural response77 to a hypoglycaemic or energy-compromised brain that initiates increased yawning (to increase brain levels of oxygen), craving (to restore energy balance), fatigue, sickness and hypersensitivity (all energy-conserving behaviours) and other symptoms.

Circumstantial evidence from early experimental studies suggests that metabolic changes induced by fasting or administration of glucose or insulin can trigger migraine attacks. Although insulin-induced hypoglycaemia elicited an attack in only 2 of 20 patients during an observation time of 2 h (ref.78), a 50 g glucose tolerance test (GTT) after 10 h of fasting initiated an attack within an 8-h test period in 6 of 10 patients with migraine whose attacks were associated with fasting or craving79. Interestingly, the metabolic responses in patients who developed an attack differed substantially from those in patients who did not: in those who developed an attack, free fatty acid and ketone body levels increased substantially before headache onset and increased further during the attack, despite similar food intake by all patients79. No differences in glucose and glycerol levels were apparent between patients who developed attacks and those who did not. These findings suggest that, similar to attacks triggered by nitroglycerin80, attacks triggered by metabolic stress develop after a latency of several hours, which is probably needed for activation of the trigeminovascular system.

Abnormal metabolic responses have been observed in several studies in which the GTT has been used (Supplementary Table 2). Comparison of responses to an intravenous GTT during and outside attacks in patients with migraine with aura81 showed that, during an attack, glucose tolerance was impaired, levels of free fatty acid, ketone bodies, glycerol and cortisol were increased, and the ratio of β-hydroxybutyrate to acetoacetate (both ketone bodies) was increased. However, insulin levels were decreased, which was considered to be an ictal stress response that was accompanied by increased lipolysis and ketogenesis. These increases in lipolysis and ketogenesis can also be interpreted as counter-regulatory responses to a cerebral energy deficit. Given that ketone bodies are an efficient alternative fuel for the brain when glucose availability is low, their elevation would be expected to restore brain energy homeostasis. However, the Western carbohydrate-laden diet means that in most people in Western countries the brain is not keto-adapted so does not have the enzymatic composition and transporters to make use of ketone bodies produced during an energy crisis.

Interictal impairments of glucose tolerance and insulin resistance have been found in various other studies of migraine82,83 (Supplementary Table 2). Insulin is the key regulator of glucose homeostasis, promoting absorption of glucose from the blood into predominantly fat and muscle cells with the help of insulin-sensitive glucose transporters (GLUTs), in particular GLUT4. Insulin also blocks carnitine transporters and, consequently, penetration of free fatty acids into cells. In the endothelial cells of the blood–brain barrier and in astrocytes and oligodendrocytes, insulin-independent GLUT1 is responsible for glucose transport under basal conditions. Multiple studies have provided evidence for an association between migraine and insulin resistance (reviewed elsewhere84), although this association was not seen at all in one study85 and only in women with migraine in another86. One study has shown that β-cell function, and therefore insulin production, is normal in patients with migraine but that the degree of insulin resistance correlates with disease severity87.

Rather than being directly involved in migraine pathogenesis, reduced insulin sensitivity could be part of a temporary adaptive response to ensure that the brain’s energy needs are met. Such a ‘glucose-sparing’ effect is typically observed when glucose availability is low (for example, during fasting or carbohydrate restriction)88,89,90. Some evidence suggests that diabetes mellitus protects against migraine91,92, and this finding supports the hypothesis that insulin resistance is an adaptive response to migraine that increases energy supply to the brain rather than a causal factor. In the long-term, however, chronic insulin resistance might contribute to metabolic diseases, such as the metabolic syndrome that is associated with migraine with aura93 and with chronic migraine in women94. Whether metabolic derangements are risk factors for or consequences of migraine, however, remains unclear. Findings in relation to BMI, for instance, have shown that this measure is associated with frequency of attacks but not with migraine prevalence95.

Elevated interictal cortisol levels have been observed in episodic86,96 and chronic migraine97, as well as during migraine attacks81, although a review of the evidence published in 2017 concluded that the results are mixed overall98. High catecholamine levels have also been associated with early morning migraine99. The observed increases in cortisol and catecholamines were not accompanied by the increase in glucose that would be expected, indicating that prior glucose levels were low but corrected by a hypoglycaemia-induced stress response. The body’s physiological reaction to hypoglycaemia does involve secretion of cortisol, adrenaline and noradrenaline, which protect cells by increasing gluconeogenesis and glycogenolysis100, stimulating protein catabolism and blocking the action of insulin. These mechanisms ensure that prolonged hypoglycaemia is avoided at all costs, even if this requires constant elevation of stress hormone levels. For this reason, and because a migraine attack can take several hours to develop, hypoglycaemia might not be detectable as a trigger101, unless blood glucose levels are monitored over a long time period.

The body’s response to hypoglycaemia also involves the release of glucagon, the antagonist of insulin. In one study, increases in blood glucose levels in response to glucagon injection were less pronounced in patients with migraine than in healthy controls102. This reduced response to glucagon could lead to a deficiency in energy compensation that partly explains migraine attacks that are induced by fasting.

Investigation of other hormones involved in energy homeostasis, such as leptin and adipocytokines, in migraine has produced conflicting data103. Low leptin levels were identified in patients with episodic migraine in one study104 and could exacerbate a cellular energy deficit. However, another study showed that levels of leptin and adiponectin are increased in migraine, and these increases could increase inflammation103,105.

Genetic studies

Genetic studies (Supplementary Table 3) directly and indirectly support the hypothesis that people with migraine have an increased vulnerability to oxidative stress, suboptimal mitochondrial functioning and/or altered metabolism. In contrast to nuclear DNA, mtDNA is particularly sensitive to reactive oxygen species (ROS)106,107. However, whether accumulation of mtDNA damage as a result of oxidative stress over time has a role in migraine chronification and increased vascular risk remains to be determined. As pointed out in a previous review108, epigenetic mechanisms, particularly mitochondrial methylation, could be a new avenue of investigation for exploring the underpinnings of mitochondrial dysfunction in migraine, as the mtDNA epigenetic status of healthy individuals differs from that of individuals with complex neurological disorders.

Coding mitochondrial DNA

Migraine is approximately threefold more prevalent among women than men109, and maternal transmission of the condition is more common than paternal transmission110, suggesting that either an X-linked form of inheritance is involved or that mtDNA has a role. The prevalence of migraine among people with mitochondrial disorders (29–35.5% of patients) is more than double that among the general population111,112 and migraine-like attacks in mitochondrial encephalomyopathy, lactic acidosis and stroke-like episodes (MELAS) are especially severe and prolonged113. Furthermore, the lifetime prevalence of migraine among healthy carriers of the Ala3243Gly mutation in coding mtDNA that causes MELAS is significantly higher than among the general population (58% versus 18%; P < 0.001)114. Among patients with the mitochondrial Ala8344Gly mutation in the MTTK gene, which causes myoclonic epilepsy with ragged red fibres (MERRF), 52% also have migraine115. These findings suggest a clinical association between a monogenic inherited disorder of mtDNA and susceptibility to migraine.

In contrast to these studies of migraine prevalence in mitochondriopathies, most genetic studies in patients with migraine have not identified the classic mutations in coding mtDNA116,117,118,119,120,121,122 (Supplementary Table 3). However, such mutations have been associated with migrainous stroke episodes123.

Non-coding mitochondrial DNA

Mitochondrial function could be impaired in migraine as a result of single nucleotide polymorphisms (SNPs) in the non-coding portion of mtDNA, which could influence mitochondrial metabolism. Specific mitochondrial haplogroups, defined by specific combinations of highly conserved polymorphisms in non-coding mtDNA, are associated with specific metabolic profiles. For example, haplogroup H confers an advantage in oxidative phosphorylation function124,125 and is associated with a poorer response to metabolism-stimulating treatment with riboflavin than that associated with other haplogroups126 (Supplementary Table 3). A high prevalence of specific SNPs in non-coding mtDNA has been seen in patients with migraine and occipital stroke associated with haplogroup U127, as well as in migraine without aura and in cyclic vomiting, a childhood equivalent of migraine128. In a similar patient population, the prevalence of the common Cys16519Thr polymorphism in non-coding mtDNA was greater than in healthy controls, and the difference from controls was even greater for the combination of this polymorphism with the less common Gly3010Ala polymorphism129. A high prevalence of these two polymorphisms has also been associated with chronic fatigue syndrome and depression130. As is the case when studying nuclear DNA, large cohort studies of cases and controls are needed to detect small to moderate associations between mtDNA variants and a complex disease such as migraine; a lack of such studies might explain why some results have not been replicated118.

Nuclear-encoded mitochondrial proteins

Given that most proteins involved in mitochondrial function are encoded by nuclear DNA, the role of genes that encode nuclear-encoded mitochondrial proteins (NEMPs) in migraine susceptibility is a promising line of investigation, yet has received little attention131. A gene-centric association analysis of NEMPs within the genetically isolated Norfolk Island population revealed an association between migraine and three genes that encode NEMPs involved in phosphorylation, fatty acid metabolism and oxidative demethylation132. This finding provides further evidence for a link between mitochondrial function and migraine susceptibility.

Non-mitochondrial genes

Some evidence suggests that genetic predisposition can lead to reduced antioxidant capacity or increased oxidative stress in migraine. For example, a polymorphism (the rs4880 TT (Val/Val) genotype) in the gene that encodes superoxide dismutase 2 (SOD2), a crucial enzyme in the clearance of mitochondrial ROS, has been associated with unilateral cranial autonomic symptoms in patients with migraine with aura133. In paediatric patients with migraine, the C–T genotype and C allele at position 16 of SOD2 and the A–A genotype and A allele at position 21 of CAT (which encodes catalase) were more frequent among patients with migraine with or without aura than among healthy controls134. Both enzymes act in concert to reduce ROS accumulation, and the polymorphisms reported in this study are thought be associated with reduced transcriptional and/or enzyme activity, meaning people with migraine are more vulnerable to oxidative stress.

Other genetic variants that affect oxidative stress have also been identified in migraine. For example, migraine, particularly migraine with aura, has been associated with the Cys677Thr mutation in the gene that encodes methylenetetrahydrofolate reductase (MTHFR)135,136. This mutation diminishes the enzyme’s capacity to remethylate homocysteine to form methionine135,136, and increased homocysteine levels favour vascular pathologies in humans and induce oxidative stress and reduce antioxidant capacity in rats137, although in a Finnish population of patients with migraine with aura, the association between MTHFR and migraine could not be replicated138. Next-generation nuclear DNA sequencing has also revealed an association between cyclic vomiting syndrome, which is considered to be a migraine equivalent, and variants in RYR2, which encodes a stress-induced calcium channel; this association could favour ROS-mediated mitochondrial damage139.

Further genetic findings that indicate alterations in metabolism in migraine include associations of the condition with polymorphisms in insulin-related genes140,141,142,143. Similarly, GLUT1 deficiency syndrome, a genetic condition of impaired glucose transport to the brain, has been linked to hemiplegic migraine and migraine with aura144.

Genome-wide associated loci

Genome-wide association studies (GWAS) have identified 38 gene loci that are associated with migraine145. In a study published in 2016, data from GWAS were integrated with high-resolution spatial gene expression data from normal adult brains to identify specific brain regions and molecular pathways that might be involved in migraine pathophysiology146. Genes associated with mitochondrial function were enriched in migraine-associated loci identified by GWAS, a finding that establishes a genetic link between mitochondrial function and migraine146. Similarly, genes related to mitochondrial function are differentially expressed in adolescent patients with menstrual migraine compared with healthy controls147.

Therapeutic studies

A large number of metabolic treatments have already been investigated in migraine (Supplementary Table 4) and most therapeutic agents used in migraine prevention can influence metabolism and mitochondrial functioning via several possible mechanisms of action (Fig. 1). The fact that some of these treatments are effective in migraine does not prove that migraine is primarily a disorder of brain energetics, but reinforces the idea that migraine is a multifactorial disorder in which the predominant pathophysiology can vary between patients. Nevertheless, the observed benefits do strongly suggest that these metabolic treatments act by improving brain energetics in some patients, although assessment of brain metabolism before and after treatment is needed to enable definitive conclusions to be drawn.

Fig. 1: Cerebral metabolomics that might be involved in migraine pathogenesis and therapeutic targets.

Glucose crosses the blood–brain barrier via insulin-independent glucose transporter 1 (GLUT1) (step 1), deficiency of which can contribute to migraine, as can insufficient glucose, oxygen, water or minerals. GLUT1 is expressed by capillary endothelial cells and astrocytes. Non-oxidative glucose metabolism produces pyruvate (step 2), which is converted to lactate and shuttled to neurons through monocarboxylate transporters (MCTs; mainly MCT1 and MCT4 in astrocytes and MCT2 in neurons). In neurons, this lactate can be used as an energy substrate following its conversion to pyruvate, which can be converted into acetyl-coenzyme A (Acetyl-CoA) and fed in turn into the tricarboxylic acid cycle (TCA) (step 3). Neurons can also take up glucose via neuronal GLUT3 (step 4), which is insulin-dependent and can be influenced by insulin resistance. Ketone bodies (β-hydroxybutyrate (BHB) and acetoacetate (AcAc)) cross the blood–brain barrier via MCT1 transporters (step 5), and penetrate astrocytes via MCT1 or MCT4 (step 6) and neurons via MCT2 (step 7). BHB provides an alternative to glucose as a substrate for oxidative phosphorylation; it is converted to AcAc and, subsequently, acetyl-CoA, which enters the TCA to produce ATP (step 8). BHB also has antioxidant properties and, compared with glucose, its conversion to ATP produces fewer reactive oxygen species (ROS) per oxygen molecule consumed. Increased ROS, nitric oxide, lack of energy substrates or lack of necessary co-enzymes inhibit mitochondrial function and reduce ATP levels (step 9). Antioxidants and co-enzymes, such as riboflavin, other B vitamins, coenzyme Q10 (CoQ10), magnesium, α-lipoic acid (ALA), l-carnitine and the anticonvulsant topiramate, support mitochondrial function and protect against migraine. AQP4, aquaporin 4; ETC, electron transport chain.

Acute treatment

Only two abortive migraine treatments have a proven link to energy metabolism. Corticosteroids that stimulate gluconeogenesis, are amongst the most effective drugs for abortion of prolonged migraine attacks and status migrainosus148. Caffeine (>100 mg), on the other hand, has a beneficial (though small) analgesic effect, at least when used in conjunction with common analgesics149. Besides its suppression of transient receptor potential A1 (TRPA1) activity150, caffeine also stimulates cortisol secretion151,152 and, consequently, gluconeogenesis153. In addition, caffeine increases levels of free fatty acids and decreases insulin responses152, effects that are similar to the metabolic changes that occur during a migraine attack81. This similarity supports the idea that the ictal metabolic abnormalities reflect counter-regulatory effects rather than pathogenic changes. Long-term use of excess caffeine, however, is associated with insulin resistance154 and migraine chronification155, whereas caffeine discontinuation is associated with higher efficacy of acute migraine treatment156.

Prophylactic nutraceuticals

Several nutraceuticals157 have been shown to be beneficial in migraine prevention158, and most of these can be linked to energy metabolism and/or mitochondrial function159. The level of evidence, however, is variable (Table 1), and not all of them are included in international guidelines for migraine prevention. They are almost devoid of adverse effects, in contrast to most classic preventive drugs.

Table 1 Evidence levels for treatment of migraine with nutraceuticals and steroids and classic preventive drugs


Riboflavin has an important role in the metabolism of carbohydrates, proteins and fats and in the recycling of oxidized glutathione, and is a precursor of flavin nucleotides, which are necessary for activity of flavoenzymes that participate in the electron transport chain160,161. Furthermore, riboflavin has neuroprotective properties, as it alleviates oxidative stress, mitochondrial dysfunction, neuroinflammation and glutamate excitotoxicity161,162. Several studies have demonstrated the efficacy of high-dose (200–400 mg daily) riboflavin for migraine prevention in adults and children163,164,165,166, but not of a low dose (50 mg daily)167. In a single-blind, comparative, parallel group study (n = 90), 400 mg riboflavin daily was as effective as 500 mg sodium valproate daily for migraine prevention168. A systematic review published in 2017 showed that high-dose riboflavin (400 mg daily) is well tolerated, inexpensive and effectively reduces migraine headache frequency169. In one study, mtDNA haplogroup influenced the therapeutic response to riboflavin126: most patients who responded had non-H haplogroups, whereas most patients who did not respond had haplogroup H with the best oxidative phosphorylation function.

Coenzyme Q10

Coenzyme Q10 (CoQ10; known as ubiquinone in its oxidized form and ubiquinol in its reduced form) is an essential cofactor of the electron transport chain with strong antioxidant properties170,171. In four placebo-controlled double-blind trials and two open-label studies, CoQ10 treatment (400 mg capsules or 300 mg liquid suspension daily) reduced migraine frequency in adults172,173,174,175. In another randomized controlled trial in children and adolescents with migraine, a 100 mg dose of CoQ10 was not superior to placebo176, but did have beneficial preventive effects in paediatric patients with migraine and low CoQ10 blood levels in an open-label study177.

Alpha-lipoic acid

Alpha-lipoic acid (ALA; also known as thioctic acid) is a water-soluble and fat-soluble antioxidant that can reduce oxidative stress directly by removing reactive species, or indirectly by chelating transition metal ions178,179. Results of a randomized placebo-controlled trial of 600 mg ALA daily for 3 months indicated a trend towards reduction of attack frequency, number of headache days and headache severity180. In an open-label trial of ALA in patients with migraine and insulin resistance, a 50% response was seen in 69% of participants181. In another study, combined treatment with topiramate and ALA for 1 month was more effective for migraine prevention than either drug alone182.

Other B vitamins

In double-blind randomized placebo-controlled trials, a 2 mg daily dose of folic acid (vitamin B9) combined with 25 mg pyridoxine (vitamin B6) and 400 μg cobalamin (vitamin B12) reduced migraine-related disability and the frequency and severity of migraine with aura183,184. In this trial, the greatest clinical effects and reductions in homocysteine levels were seen in carriers of the C allele of the MTHFR Cys677Thr variant and of the A allele of the Ala66Gly variant of the gene that encodes methionine synthase reductase184. The same vitamin B combination with a 1 mg dose of folic acid was not superior to placebo185. According to a systematic review186 in which nine open studies were identified, niacin (nicotinic acid or vitamin B3) might also have a prophylactic effect in migraine, but randomized controlled trials are lacking.


Magnesium acts as a cofactor for as many as 300 enzymes and has a vital role in energy metabolism. Blood levels of magnesium are reduced in migraine31,39,187 and a meta-analysis of randomized controlled trials provides evidence that intravenous magnesium has a modest beneficial effect on acute migraine attacks and that oral magnesium reduces attack frequency and intensity188. A randomized controlled trial published in 2019 showed that 500 mg magnesium oxide daily had a similar preventive effect in migraine as 400 mg sodium valproate daily189.


l-Carnitine transports fatty acids into the mitochondria for lipid oxidation and energy production. Trials of l-carnitine for migraine prevention have been conducted, but results are conflicting. In a large (133 patients) but single-blinded study the effects of daily administration of 500 mg l-carnitine, 500 mg magnesium oxide or both for 3 months were compared with each other and with routine treatment190. Migraine attacks and days decreased significantly in all patient groups, although this decrease was greater among patients who received magnesium oxide alone190. By contrast, in a randomized crossover trial with successive 3-month treatment periods separated by a 4-week washout, the effect of 3 g acetyl l-carnitine daily was not significantly different from that of placebo191. In a randomized controlled trial published in 2019, a combination of 500 mg l-carnitine and 30 mg CoQ10 daily was significantly more effective than placebo in reducing headache severity, frequency and duration during an 8-week treatment period175.

Ketogenic diet and exercise

A ketogenic diet mimics, to some extent, the state of fasting and promotes hepatic production of an alternative to glucose as an energy substrate for the brain. Ketone body transport is not GLUT1-dependent and ketosis has a variety of other effects that are potentially beneficial in migraine pathophysiology, including: increased mitochondrial biogenesis; increased antioxidant capacity; upregulation of GLUT1 and ketone body transporters; increased GABA but inhibition of glutamate transport and, therefore, reduced excitatory synaptic transmission pain and inflammation (reviewed in detail elsewhere192,193). Ketone bodies can also stabilize neuronal excitability by inhibiting ATP-sensitive potassium channels (KATP) that might play a crucial role in migraine attack generation194 (see below). Several case studies have shown that ketosis can protect against migraine195,196,197,198,199,200. In addition, in a 1-month observational study of the ketone diet in 96 patients with migraine as part of a weight-loss programme, attack frequency, attack severity and acute medication use were reduced by up to 80%199. Similarly, in a study of 18 patients with episodic migraine, the same intervention reduced migraine days by 62.5%, and this reduction was accompanied by normalization of interictal deficits in habituation of visual evoked responses200. In a double-blind study published in 2019, a very low-calorie ketogenic diet and a very low-calorie non-ketogenic diet were compared in a 2-month crossover study in a population of 35 overweight individuals with episodic migraine201. The ketogenic diet was superior to the non-ketogenic diet for reducing monthly migraine days, and the 50% response rate was much higher (74.3% of patients on the ketogenic diet and 8.6% of patients on the non-ketogenic diet)201. The potential for supplementation with the ketone body β-hydroxybutyrate without a strict dietary change to prevent migraine is currently being examined in a randomized controlled trial202.

Aerobic exercise is often recommended in migraine management, and a randomized study showed that regular aerobic exercise is associated with a decrease in migraine frequency comparable to that achieved with topiramate, a prophylactic drug with level A evidence for efficacy203. The metabolic effects of aerobic exercise mimic those of ketosis and include upregulation of multiple proteins that are involved in brain energy metabolism, such as enzymes involved in glucose catabolism, ATP synthesis and hydrolysis, and glutamate turnover204. Furthermore, in mice, exercise training increases the number of mitochondria not only in muscle but also in the brain205.

Prophylactic pharmaceuticals

The level of evidence for the classic preventive migraine drugs is higher overall than that for nutraceuticals (Table 1). The precise mechanisms of action of most drugs used in migraine prophylaxis are not known, but many downregulate neuronal reactivity206 and might, therefore, reduce the energy demands of the brain. Certain prophylactic agents can also have direct metabolic effects. For instance, topiramate protects against oxidative stress, inflammation207 and mitochondrial membrane depolarization208, prolongs mitochondrial survival208, slightly increases lipolysis in children209 and increases the sensitivity of adipocytes to insulin in female rats210.

Other drugs used for prevention of migraine treatment have clear metabolic effects. Amitriptyline also reduces markers of oxidative stress and increases antioxidant capacity65. Valproate attenuates nitroglycerin-induced trigeminovascular activation by preserving mitochondrial function in a rat model of migraine211 and increases mitochondrial biogenesis212. Amitriptyline and flunarizine increase serum levels of leptin and insulin and increase BMI after 12 weeks of therapy in patients with migraine213. Gabapentin, atenolol, verapamil, valproate, pizotifen and amitriptyline all increase body weight in a substantial number of patients after 6 months of use214. Beta-blockers decrease the whole-body metabolic rate and body fat215, thereby theoretically leaving more energy for the brain. In addition, beta-blockers reduce rebound headache that often follows stress in patients with migraine216, and this effect could result from regulation of noradrenaline-mediated consumption of energy reserves100 during stress. Together, the findings suggest that the double action of prophylactic drugs — on neurons and metabolism — favours the equilibrium between metabolic needs and available energy that is necessary to maintain cerebral homeostasis.

Metabolism and migraine pathophysiology

Above, we highlight the clinical evidence that suggests a role for energy metabolism and mitochondrial function in migraine pathophysiology. In this section, we examine how such abnormalities could lead to a migraine attack. We consider these changes in relation to three hallmarks of the pathophysiological migraine cascade: hypothalamic and brainstem activation (which is thought to initiate and modulate the attack), CSD (which is responsible for the aura) and trigeminovascular activation (which causes the headache and associated symptoms). For this purpose, we mainly draw upon experimental data obtained in rodents.

Hypothalamic and brainstem activation

The hypothalamus is activated early in migraine attack initiation, during the premonitory phase75,76. What activates the hypothalamus is not known; possibilities include an intrinsic biorhythm, an environmental trigger or stimulation from a CNS centre that is highly connected to the hypothalamus, such as the amygdala79. The hypothalamus can sense a metabolic disequilibrium in the brain, owing to the presence of chemosensitive neurons, and in the periphery, partly because some hypothalamic areas lack a fully functional blood–brain barrier. Chemosensitive neurons, notably those that detect oxygen, form a network that extends from the thalamus to the brainstem217. Therefore, we hypothesize that metabolic changes in the brain and/or in the blood might activate these neuronal systems and ignite a migraine attack.

Involvement of the amygdala in activation of the hypothalamus in migraine, as suggested above, could provide a link between mitochondrial function and the sexual dimorphism of migraine. In the human and mouse basolateral amygdala, mitochondrion-related biological pathways are the most strongly associated with sexual dimorphism and, in females, genes related to mitochondrial function are downregulated whereas genes related to regulation of the circadian rhythm are upregulated218.

Cortical spreading depression

CSD is the pathophysiological cause of migraine aura. Susceptibility to CSD is strongly modulated by metabolic factors. For example, cerebral glucose availability modulates induced CSD219,220. Hypoglycaemia prolongs CSD219 and inhibition of cerebral glycogen reduces the threshold for CSD in vivo220. Hyperglycaemia protects against induction of CSD219. In rats, supply of an energy substrate other than glucose via short-term and long-term treatment with a medium-chain triglyceride-enriched ketogenic diet has a similar protective effect against CSD221.

Hypoxia negatively influences energy metabolism and can trigger CSD222,223. Its effects in mice and rats include inhibition of astroglial mitochondrial respiration, leading to mitochondrial depolarization, production of free radicals, lipid peroxidation and release of calcium ions from intracellular stores224. When hypoxia is preceded by pharmacological mitochondrial inhibition, hypoxia-induced CSD in rat hippocampal slices is greatly facilitated222. This observation suggests a mechanism by which genetic or acquired mitochondrial dysfunction could exacerbate the impact of a metabolic stressor.

Trigeminovascular system

In experimental animals, CSD can activate the trigeminovascular system225. One mechanism of this process is the opening of pannexin 1 large-pore channels in neurons, which are associated with the ligand-gated ion channels P2X7 (ref.226), leading to downstream activation of the inflammatory pathway in astrocytes and — via cytokines and prostanoids — to sensitization of meningeal nociceptors227 (Fig. 2). In addition, CSD and subsequent restoration of ion homeostasis have an extremely high energy demand228. Moreover, in mice, CSD causes tissue hypoxia223,229 and increases expression of mitochondrial uncoupling proteins, which decreases ATP synthesis and increases thermogenesis230. These changes could create a vicious circle in which the metabolic disequilibrium that favours or triggers CSD, and hence trigeminovascular system activation, is perpetuated. CSD also induces oxidative stress in the trigeminal nociceptive system231. For example, hydrogen peroxide production can activate TRPA1 and acid-sensing ion channels (ASICs), thereby promoting release of calcitonin gene-related peptide (CGRP) from meningeal nociceptors, which is known to be pivotal in mediating the headache of the migraine attack232,233,234 (Fig. 2).

Fig. 2: Metabolic face of migraine attack generation and resolution.

In genetically predisposed individuals, migraine triggers can increase oxidative and nitrosative stress and/or reduce cerebral ATP and glycogen levels. These processes ignite a cascade of events that assists restoration of cerebral energy homeostasis but also favours cortical spreading depression (CSD), trigeminovascular activation via activation of transient receptor potential channel A1 (TRPA1), acid-sensing ion channels (ASICs) and the pannexin-1–P2X7 pore complex, which leads to release of calcitonin gene-related peptide (CGRP) and pituitary adenylate cyclase-activating peptide (PACAP) and opening of ATP-sensitive potassium channels (KATP). Together, this molecular cascade can produce the symptoms of migraine. In this way, the factors that are known to participate in generation of migraine attacks might also be involved in its resolution. Dashed lines indicate mechanisms by which metabolic homeostasis may be restored. AcAc, acetoacetate; BHB, β-hydroxybutyrate; FFA, free fatty acid; GLUT1, glucose transporter 1.

Clinically, most patients with migraine never experience an aura. Therefore, triggers of the trigeminovascular system other than CSD must exist. TRP channels, which are expressed in meningeal nociceptive nerve terminals, might contribute to generation of a migraine attack235, as they can be directly activated by various exogenous and endogenous agents that are associated with migraine, and their activation induces the release of CGRP234. TRP channels are also inhibited or desensitized by abortive migraine drugs235. These channels can also sense reactive species indirectly through second messengers or directly via oxidative modification of cysteine residues236, and the subchannel TRPA1 is strongly activated by oxidative, nitrosative and electrophilic stress235,236. In combination, these observations suggest a mechanism by which known migraine trigger factors that increase oxidative stress could lead to migraine pain.

Cerebral energy deficiency caused by inhibition of glycogen use in the brain was recently shown to cause CSD-independent opening of pannexin 1 large pore channels in neurons220, which leads to activation of meningeal nociceptors227. Furthermore, a study published in 2017 showed that metabolic changes can directly modify activity of central trigeminovascular nociceptors — changes in blood glucose levels after injection of insulin, glucagon or leptin were associated with changes in baseline firing of dural responsive nociceptive neurons in the spinal trigeminal nucleus237. Recent evidence indicates that KATP channels link metabolic stress with activation of trigeminovascular nociceptors. These channels are located in cranial arteries and the trigeminal ganglion and are modulated by the intracellular ratio of ATP to ADP and levels of cAMP and cGMP. In a study published in 2019, intravenous infusion of the KATP channel opener levcromakalim induced migraine headache in 16 of 16 patients with migraine, whereas placebo treatment did so in only 1 of these 16 patients194 (Box 1). The role of KATP channels in the metabolic aspects of migraine pathophysiology might not, however, be restricted to cranial vessels and the trigeminovascular system. In slice preparations of rodent cerebrum, ketone bodies238 or a decrease in extracellular glucose levels239 can open KATP channels and decrease the excitability of central neurons. This process is thought to be mediated by pannexin 1 hemichannel-mediated ATP release and activation of adenosine receptors239 (Fig. 2).

High nitric oxide concentrations stimulate CGRP release and activate the trigeminovascular system, and nitric oxide can also modulate mitochondrial activity via various mechanisms240. The mitochondrial effects of nitric oxide could, therefore, explain the headache experienced by patients with migraine upon administration of nitric oxide donors241. Examples of the mitochondrial effects of nitric oxide in animals include inhibition of the mitochondrial respiratory chain in cultured astrocytes242. By contrast, use of 1H-MRS in rats has shown that cortical lactate increases as early as 10 min after intraperitoneal injection of nitroglycerin, suggesting that the widely used nitroglycerin model of migraine involves metabolic, in addition to vascular, changes243.

Finally, limited evidence from animal experiments suggests that the migraine attack itself can affect mitochondrial energy metabolism in the trigeminal ganglion. In a rat model of chronic migraine in which an inflammatory soup is applied to the dura mater, abnormal mitochondrial dynamics and impaired mitochondrial biogenesis were observed in the trigeminal ganglion244,245.

Metabolic functions of CGRP and PACAP

CGRP is known to play a role in spontaneous and triggered migraine headache232,233. Blood levels of CGRP are elevated during attacks246, CGRP triggers attacks in patients with migraine247, and blocking the action of CGRP transiently (with CGRP antagonists) or durably (with monoclonal antibodies) aborts attacks or reduces attack frequency, respectively248. As a consequence, CGRP is overwhelmingly considered to be the ‘villain’ in migraine pathophysiology. However, in CGRP-triggered migraine attacks, only 28% of patients with migraine with aura experience an aura249, and CGRP does not elicit premonitory symptoms114. These observations suggest that CGRP does not trigger a complete migraine attack, but is the physiological correlate of the headache pain and the headache-related behaviour.

Some evidence suggests that CGRP release is a response to oxidative stress or cerebral energy disequilibrium and might be part of an adaptive response, thereby challenging the perception that CGRP is the pathophysiological trigger of migraine. For example, CGRP has antioxidant and anti-inflammatory actions250,251,252,253, supporting the hypothesis that its release mediates an adaptive response to oxidative stress and/or energy deficiency. The idea that CGRP increases endogenous energy availability for the brain is also supported by rodent studies in which CGRP inhibited insulin-stimulated glucose transport254, decreased tolerance to glucose in the GTT without altering plasma insulin levels255, inhibited muscle glycogen synthesis and caused insulin resistance upon activation of skeletal muscle sensory nerves256. In addition, intravenous injections of CGRP in the rat increased plasma glucose concentrations255, an effect that could help to restore energy homeostasis.

Moreover, CGRP is widely distributed in the brainstem and diencephalon, including a network that includes hypothalamic nuclei, the locus coeruleus, the area postrema and the nucleus tractus solitarius, and is involved in energy homeostasis257. Several of these nuclei do not have a functional blood–brain barrier and are therefore accessible to CGRP therapies that are administered systemically.

Pituitary adenylate cyclase-activating peptide (PACAP) is also released during migraine attacks and induces migraine-like headache when administered to patients with migraine. Consequently, PACAP is another promising molecular target for migraine treatment. PACAP is present in the trigemino-parasympathetic viscero-autonomic circuit, where it contributes to headache pain and associated autonomic symptoms, but it also has a role in the hypothalamus, where it modulates circadian rhythms and food anticipatory behaviour (reviewed elsewhere257). In rats, PACAP stimulated glucose production via sympathetic hepatic innervation, indicating the PACAP release could also be an adaptive response to restore energy homeostasis258.


Cerebral energy deficiency and/or increased oxidative stress decrease the threshold for CSD and activate TRP channels and ASICs, thereby stimulating CGRP and PACAP release. These peptides are pivotal in eliciting the migraine headache and associated symptoms, but we hypothesize that they also induce an antioxidant response and various metabolic changes that, together with energy-conserving behavioural changes, decrease oxidative stress levels and increase glucose and ketone body availability for the brain to help restore energy homeostasis (Fig. 2).

Future research

Studies that combine assessments of brain and mitochondrial metabolism with those of sensory processing are necessary to disentangle the disequilibrium between metabolic reserve and brain activity in migraine. In these studies, age-related adaptive increases in glucose uptake in the brainstem and in visual areas must be taken into account259. Similarly, studies to examine the role of specific alterations in mitochondrial function and/or energy metabolism in migraine subgroups are needed. More data are also needed to determine whether therapeutic interventions that improve mitochondrial function lead to changes in sensory processing and cerebral energy availability that correlate with treatment responses.

Despite the need for more data, the findings described in this Review already have several potential therapeutic implications. We suggest a four-step approach to improve mitochondrial function and energy metabolism in migraine (Box 2).


The evidence discussed in this Review indicates that, from a metabolic perspective, migraine is a conserved adaptive response77 that helps to reduce harmful oxidative stress levels and restore brain energy homeostasis, a concept that was proposed by Edward Liveing as early as 1873 (ref.260). Given the high prevalence of migraine and the fact that it is associated with common gene polymorphisms, a migraine-prone nervous system might be, or at least might have been, associated with reproductive or survival advantages77,261 from an evolutionary perspective. Over time, possibly owing to changes in nutrition, our environment might have become inadequate or suboptimal for the conserved adaptive genetic response patterns. Alternatively, migraine could be the price the human species has to pay for having a developed, high-performing and energy-consuming brain.

Commonly reported migraine trigger factors can be linked to energy disequilibrium and oxidative stress, and numerous biochemical and genetic studies point towards a variety of different metabolic abnormalities in migraine. Most preventive migraine treatments can improve metabolic functioning in addition to their effects on brain responsiveness and excitability. Disruption of cerebral metabolic homeostasis is a plausible trigger of the trigeminovascular system and its limbic connections via induction of CSD, stimulation of chemosensitive brainstem neurons or direct activation of TRP channels and/or ASICs that stimulate CGRP and PACAP release by meningeal nociceptive fibres, or modulation of KATP channels, which might be the final common pathway in activation of trigeminal nociceptors. These neuropeptide pathways are the likely culprits for the migraine headache and associated symptoms, but they could also participate in an antioxidant response and various metabolic changes that help restore energy homeostasis.

Looking at the metabolic face of migraine has several potential therapeutic implications (Box 2). Some progress has been made in developing anti-migraine therapies that target metabolism, and novel strategies, such as ketone body supplementation, are being explored. More research is needed on different metabolic subtypes, the association between metabolic phenotypes and genotypes and treatment responses to metabolic agents, interactions between the sensory system and metabolism, and metabolic nutraceutical treatments for migraine.


  1. 1.

    Gray, P. A. & Burtness, H. I. Hypoglycemic headache. Endocrinology 19, 549–560 (1935).

    Google Scholar 

  2. 2.

    Amery, W. K. Brain hypoxia: the turning-point in the genesis of the migraine attack? Cephalalgia 2, 83–109 (1982). This article is the first to have drawn attention to the possible role of brain hypoxia in migraine pathogenesis.

    CAS  PubMed  Google Scholar 

  3. 3.

    Schoenen, J., Ambrosini, A., Sándor, P. S. & Maertens de Noordhout, A. Evoked potentials and transcranial magnetic stimulation in migraine: published data and viewpoint on their pathophysiologic significance. Clin. Neurophysiol. 114, 955–972 (2003).

    PubMed  Google Scholar 

  4. 4.

    Schoenen, J. Pathogenesis of migraine: the biobehavioural and hypoxia theories reconciled. Acta Neurol. Belg. 94, 79–86 (1994). This article presents the hypothesis that the conjunction of abnormal energy metabolism and information processing in the brain may be able to generate migraine attacks.

    CAS  PubMed  Google Scholar 

  5. 5.

    Pavlovic, J. M., Buse, D. C., Sollars, C. M., Haut, S. & Lipton, R. B. Trigger factors and premonitory features of migraine attacks: summary of studies. Headache 54, 1670–1679 (2014).

    PubMed  Google Scholar 

  6. 6.

    Peroutka, S. J. What turns on a migraine? A systematic review of migraine precipitating factors. Curr. Pain Headache Rep. 18, 454 (2014).

    PubMed  Google Scholar 

  7. 7.

    Kelman, L. The triggers or precipitants of the acute migraine attack. Cephalalgia 27, 394–402 (2007).

    CAS  PubMed  Google Scholar 

  8. 8.

    Schoonman, G. G., Evers, D. J., Terwindt, G. M., van Dijk, J. G. & Ferrari, M. D. The prevalence of premonitory symptoms in migraine: a questionnaire study in 461 patients. Cephalalgia 26, 1209–1213 (2006).

    CAS  PubMed  Google Scholar 

  9. 9.

    Spierings, E. L. H., Donoghue, S., Mian, A. & Wöber, C. Sufficiency and necessity in migraine: how do we figure out if triggers are absolute or partial and, if partial, additive or potentiating? Curr. Pain Headache Rep. 18, 455 (2014).

    PubMed  Google Scholar 

  10. 10.

    Borkum, J. M. Migraine triggers and oxidative stress: a narrative review and synthesis. Headache 56, 12–35 (2016).

    PubMed  Google Scholar 

  11. 11.

    Pingitore, A. et al. Exercise and oxidative stress: potential effects of antioxidant dietary strategies in sports. Nutrition 31, 916–922 (2015).

    CAS  PubMed  Google Scholar 

  12. 12.

    Powers, S. K., Radak, Z. & Ji, L. L. Exercise-induced oxidative stress: past, present and future. J. Physiol. 594, 5081–5092 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  13. 13.

    Schiavone, S., Jaquet, V., Trabace, L. & Krause, K.-H. Severe life stress and oxidative stress in the brain: from animal models to human pathology. Antioxid. Redox Signal. 18, 1475–1490 (2013).

    CAS  PubMed  PubMed Central  Google Scholar 

  14. 14.

    Trivedi, M. S., Holger, D., Bui, A. T., Craddock, T. J. A. & Tartar, J. L. Short-term sleep deprivation leads to decreased systemic redox metabolites and altered epigenetic status. PLOS ONE 12, e0181978 (2017).

    PubMed  PubMed Central  Google Scholar 

  15. 15.

    Angelucci, F. L. et al. Physiological effect of olfactory stimuli inhalation in humans: an overview. Int. J. Cosmet. Sci. 36, 117–123 (2014).

    CAS  PubMed  Google Scholar 

  16. 16.

    Franken, C. et al. Phthalate-induced oxidative stress and association with asthma-related airway inflammation in adolescents. Int. J. Hyg. Environ. Health 220, 468–477 (2017).

    CAS  PubMed  Google Scholar 

  17. 17.

    Nakamura, M., Kuse, Y., Tsuruma, K., Shimazawa, M. & Hara, H. The involvement of the oxidative stress in murine blue LED light-induced retinal damage model. Biol. Pharm. Bull. 40, 1219–1225 (2017).

    CAS  PubMed  Google Scholar 

  18. 18.

    Demirel, R. et al. Noise induces oxidative stress in rat. Eur. J. Gen. Med. 6, 20–24 (2009).

    CAS  Google Scholar 

  19. 19.

    Arngrim, N. et al. Migraine induced by hypoxia: an MRI spectroscopy and angiography study. Brain 139, 723–737 (2016). In this study, normobaric hypoxia induced migraine headache, but not aura in patients with migraine.

    PubMed  Google Scholar 

  20. 20.

    Broessner, G. et al. Hypoxia triggers high-altitude headache with migraine features: a prospective trial. Cephalalgia 36, 765–771 (2016).

    PubMed  Google Scholar 

  21. 21.

    Arregui, A. et al. High prevalence of migraine in a high-altitude population. Neurology 41, 1668–1668 (1991).

    CAS  PubMed  Google Scholar 

  22. 22.

    Reddy, V. D., Padmavathi, P., Kavitha, G., Saradamma, B. & Varadacharyulu, N. Alcohol-induced oxidative/nitrosative stress alters brain mitochondrial membrane properties. Mol. Cell. Biochem. 375, 39–47 (2013).

    CAS  PubMed  Google Scholar 

  23. 23.

    Chauvel, V., Schoenen, J. & Multon, S. Influence of ovarian hormones on cortical spreading depression and its suppression by L-kynurenine in rat. PLOS ONE 8, e82279 (2013).

    PubMed  PubMed Central  Google Scholar 

  24. 24.

    Chauvel, V., Multon, S. & Schoenen, J. Estrogen-dependent effects of 5-hydroxytryptophan on cortical spreading depression in rat: modelling the serotonin-ovarian hormone interaction in migraine aura. Cephalalgia 38, 427–436 (2018).

    PubMed  Google Scholar 

  25. 25.

    Irwin, R. W. et al. Progesterone and estrogen regulate oxidative metabolism in brain mitochondria. Endocrinology 149, 3167–3175 (2008).

    CAS  PubMed  PubMed Central  Google Scholar 

  26. 26.

    Mauvais-Jarvis, F., Clegg, D. J. & Hevener, A. L. The role of estrogens in control of energy balance and glucose homeostasis. Endocr. Rev. 34, 309–338 (2013).

    CAS  PubMed  PubMed Central  Google Scholar 

  27. 27.

    Welch, K. M., Levine, S. R., D’Andrea, G., Schultz, L. R. & Helpern, J. A. Preliminary observations on brain energy metabolism in migraine studied by in vivo phosphorus 31 NMR spectroscopy. Neurology 39, 538–541 (1989). A pioneering study that showed that mitochondrial energy metabolism is abnormal in the brain of patients with migraine.

    CAS  PubMed  Google Scholar 

  28. 28.

    Barbiroli, B. et al. Abnormal brain and muscle energy metabolism shown by 31P magnetic resonance spectroscopy in patients affected by migraine with aura. Neurology 42, 1209–1214 (1992).

    CAS  PubMed  Google Scholar 

  29. 29.

    Kim, J. H. et al. Interictal metabolic changes in episodic migraine: a voxel-based FDG-PET study. Cephalalgia 30, 53–61 (2010).

    CAS  PubMed  Google Scholar 

  30. 30.

    Lodi, R. et al. Deficit of brain and skeletal muscle bioenergetics and low brain magnesium in juvenile migraine: an in vivo 31P magnetic resonance spectroscopy interictal study. Pediatr. Res. 42, 866–871 (1997).

    CAS  PubMed  Google Scholar 

  31. 31.

    Lodi, R. et al. Deficient energy metabolism is associated with low free magnesium in the brains of patients with migraine and cluster headache. Brain Res. Bull. 54, 437–441 (2001).

    CAS  PubMed  Google Scholar 

  32. 32.

    Montagna, P. et al. 31P-magnetic resonance spectroscopy in migraine without aura. Neurology 44, 666–669 (1994).

    CAS  PubMed  Google Scholar 

  33. 33.

    Reyngoudt, H., Achten, E. & Paemeleire, K. Magnetic resonance spectroscopy in migraine: what have we learned so far? Cephalalgia 32, 845–859 (2012).

    PubMed  Google Scholar 

  34. 34.

    Schulz, U. G. et al. Association between cortical metabolite levels and clinical manifestations of migrainous aura: an MR-spectroscopy study. Brain J. Neurol. 130, 3102–3110 (2007).

    CAS  Google Scholar 

  35. 35.

    Lodi, R. et al. Quantitative analysis of skeletal muscle bioenergetics and proton efflux in migraine and cluster headache. J. Neurol. Sci. 146, 73–80 (1997).

    CAS  PubMed  Google Scholar 

  36. 36.

    Barbiroli, B. et al. Complicated migraine studied by phosphorus magnetic resonance spectroscopy. Cephalalgia 10, 263–272 (1990).

    CAS  PubMed  Google Scholar 

  37. 37.

    Cevoli, S., Favoni, V. & Cortelli, P. Energy metabolism impairment in migraine. Curr. Med. Chem. (2018). Excellent comprehensive review of energy metabolism in migraine highlighting the pivotal contribution of the Bologna group initiated by the late Pasaquale Montagna.

  38. 38.

    Reyngoudt, H., Paemeleire, K., Descamps, B., De Deene, Y. & Achten, E. 31P-MRS demonstrates a reduction in high-energy phosphates in the occipital lobe of migraine without aura patients. Cephalalgia 31, 1243–1253 (2011). This study showed that absolute brain levels of ATP are reduced interictally in patients with migraine without aura.

    PubMed  Google Scholar 

  39. 39.

    Ramadan, N. M. et al. Low brain magnesium in migraine. Headache 29, 416–419 (1989). The first study to show that magnesium is low in the brain of patients with migraine.

    CAS  PubMed  Google Scholar 

  40. 40.

    Sandor, P. S. et al. MR-spectroscopic imaging during visual stimulation in subgroups of migraine with aura. Cephalalgia 25, 507–518 (2005). This study revealed that resting and stimulation-induced levels of lactate in the occipital cortex differ among phenotypic subtypes of patients with migraine with aura.

    CAS  PubMed  Google Scholar 

  41. 41.

    Watanabe, H., Kuwabara, T., Ohkubo, M., Tsuji, S. & Yuasa, T. Elevation of cerebral lactate detected by localized 1H-magnetic resonance spectroscopy in migraine during the interictal period. Neurology 47, 1093–1095 (1996).

    CAS  PubMed  Google Scholar 

  42. 42.

    Reyngoudt, H. et al. Does visual cortex lactate increase following photic stimulation in migraine without aura patients? A functional (1)H-MRS study. J. Headache Pain 12, 295–302 (2011).

    CAS  PubMed  PubMed Central  Google Scholar 

  43. 43.

    Prescot, A. et al. Excitatory neurotransmitters in brain regions in interictal migraine patients. Mol. Pain 5, 34 (2009).

    PubMed  PubMed Central  Google Scholar 

  44. 44.

    Mohamed, R. E., Aboelsafa, A. A. & Al-Malt, A. M. Interictal alterations of thalamic metabolic concentration ratios in migraine without aura detected by proton magnetic resonance spectroscopy. Egypt. J. Radiol. Nucl. Med. 44, 859–870 (2013).

    Google Scholar 

  45. 45.

    Becerra, L. et al. A ‘complex’ of brain metabolites distinguish altered chemistry in the cingulate cortex of episodic migraine patients. Neuroimage Clin. 11, 588–594 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  46. 46.

    Sappey-Marinier, D. et al. Effect of photic stimulation on human visual cortex lactate and phosphates using 1H and 31P magnetic resonance spectroscopy. J. Cereb. Blood Flow Metab. 12, 584–592 (1992).

    CAS  PubMed  Google Scholar 

  47. 47.

    Magistretti, P. J. & Pellerin, L. Cellular mechanisms of brain energy metabolism and their relevance to functional brain imaging. Phil. Trans. R. Soc. Lond. B 354, 1155–1163 (1999). This is a review of brain energy metabolism, in particular of the “neuron–astrocyte lactate shuttle” by the authors who first described this important phenomenon.

    CAS  Google Scholar 

  48. 48.

    Gantenbein, A. R. et al. Sensory information processing may be neuroenergetically more demanding in migraine patients. Neuroreport 24, 202–205 (2013).

    PubMed  Google Scholar 

  49. 49.

    Lisicki, M. et al. Evidence of an increased neuronal activation-to-resting glucose uptake ratio in the visual cortex of migraine patients: a study comparing 18FDG-PET and visual evoked potentials. J. Headache Pain 19, 49 (2018). This study showed that in patients with migraine there is a mismatch between neuronal activation and glucose uptake in the visual system between attacks.

    PubMed  PubMed Central  Google Scholar 

  50. 50.

    Littlewood, J. et al. Low platelet monoamine oxidase activity in headache: no correlation with phenolsulphotransferase, succinate dehydrogenase, platelet preparation method or smoking. J. Neurol. Neurosurg. Psychiatry 47, 338–343 (1984).

    CAS  PubMed  PubMed Central  Google Scholar 

  51. 51.

    Sangiorgi, S. et al. Abnormal platelet mitochondrial function in patients affected by migraine with and without aura. Cephalalgia 14, 21–23 (1994).

    CAS  PubMed  Google Scholar 

  52. 52.

    Montagna, P. et al. Mitochondrial abnormalities in migraine. Preliminary findings. Headache 28, 477–480 (1988).

    CAS  PubMed  Google Scholar 

  53. 53.

    Okada, H., Araga, S., Takeshima, T. & Nakashima, K. Plasma lactic acid and pyruvic acid levels in migraine and tension-type headache. Headache 38, 39–42 (1998).

    CAS  PubMed  Google Scholar 

  54. 54.

    Alp, R., Selek, S., Alp, S. I., Taşkin, A. & Koçyiğit, A. Oxidative and antioxidative balance in patients of migraine. Eur. Rev. Med. Pharmacol. Sci. 14, 877–882 (2010).

    CAS  PubMed  Google Scholar 

  55. 55.

    Aytaç, B. et al. Decreased antioxidant status in migraine patients with brain white matter hyperintensities. Neurol. Sci. 35, 1925–1929 (2014).

    PubMed  Google Scholar 

  56. 56.

    Bernecker, C. et al. Oxidative stress is associated with migraine and migraine-related metabolic risk in females. Eur. J. Neurol. 18, 1233–1239 (2011).

    CAS  PubMed  Google Scholar 

  57. 57.

    Bolayir, E. et al. Intraerythrocyte antioxidant enzyme activities in migraine and tension-type headaches. J. Chin. Med. Assoc. 67, 263–267 (2004).

    PubMed  Google Scholar 

  58. 58.

    Ciancarelli, I., Tozzi-Ciancarelli, M., Massimo, C. D., Marini, C. & Carolei, A. Urinary nitric oxide metabolites and lipid peroxidation by-products in migraine. Cephalalgia 23, 39–42 (2003).

    CAS  PubMed  Google Scholar 

  59. 59.

    Ciancarelli, I., Tozzi-Ciancarelli, M., Spacca, G., Massimo, C. D. & Carolei, A. Relationship between biofeedback and oxidative stress in patients with chronic migraine. Cephalalgia 27, 1136–1141 (2007).

    CAS  PubMed  Google Scholar 

  60. 60.

    Eren, Y., Dirik, E., Neşelioğlu, S. & Erel, Ö. Oxidative stress and decreased thiol level in patients with migraine: cross-sectional study. Acta Neurol. Belg. 115, 643–649 (2015).

    PubMed  Google Scholar 

  61. 61.

    Geyik, S., Altunısık, E., Neyal, A. M. & Taysi, S. Oxidative stress and DNA damage in patients with migraine. J. Headache Pain 17, 10 (2016).

    PubMed  PubMed Central  Google Scholar 

  62. 62.

    Gumusyayla, S. et al. A novel oxidative stress marker in migraine patients: dynamic thiol-disulphide homeostasis. Neurol. Sci. 37, 1311–1317 (2016).

    PubMed  Google Scholar 

  63. 63.

    Shimomura, T. et al. Platelet superoxide dismutase in migraine and tension-type headache. Cephalalgia 14, 215–218 (1994).

    CAS  PubMed  Google Scholar 

  64. 64.

    Tozzi-Ciancarelli, M. et al. Oxidative stress and platelet responsiveness in migraine. Cephalalgia 17, 580–584 (1997).

    CAS  PubMed  Google Scholar 

  65. 65.

    Tripathi, G. M., Kalita, J. & Misra, U. K. A study of oxidative stress in migraine with special reference to prophylactic therapy. Int. J. Neurosci. 128, 318–324 (2018).

    CAS  PubMed  Google Scholar 

  66. 66.

    Tuncel, D., Tolun, F. I., Gokce, M., İmrek, S. & Ekerbiçer, H. Oxidative stress in migraine with and without aura. Biol. Trace Elem. Res. 126, 92–97 (2008).

    CAS  PubMed  Google Scholar 

  67. 67.

    Yilmaz, G., Sürer, H., Inan, L. E., Coskun, O. & Yücel, D. Increased nitrosative and oxidative stress in platelets of migraine patients. Tohoku J. Exp. Med. 211, 23–30 (2007).

    CAS  PubMed  Google Scholar 

  68. 68.

    Neri, M. et al. A meta-analysis of biomarkers related to oxidative stress and nitric oxide pathway in migraine. Cephalalgia 35, 931–937 (2015).

    PubMed  Google Scholar 

  69. 69.

    Gonullu, H. et al. The levels of trace elements and heavy metals in patients with acute migraine headache. J. Pak. Med. Assoc. 65, 694–697 (2015).

    PubMed  Google Scholar 

  70. 70.

    Welch, K. M., Nagesh, V., Aurora, S. K. & Gelman, N. Periaqueductal gray matter dysfunction in migraine: cause or the burden of illness? Headache 41, 629–637 (2001).

    CAS  PubMed  Google Scholar 

  71. 71.

    Blau, J. N. & Cumings, J. N. Method of precipitating and preventing some migraine attacks. Br. Med. J. 2, 1242–1243 (1966).

    CAS  PubMed  PubMed Central  Google Scholar 

  72. 72.

    Roberts, H. J. Migraine and related vascular headaches due to diabetogenic hyperinsulinism. Observations on pathogenesis and rational treatment in 421 patients. Headache 7, 41–62 (1967).

    CAS  PubMed  Google Scholar 

  73. 73.

    Binder, C. & Bendtson, I. Endocrine emergencies. Hypoglycaemia. Baillieres Clin. Endocrinol. Metab. 6, 23–39 (1992).

    CAS  PubMed  Google Scholar 

  74. 74.

    Denuelle, M., Fabre, N., Payoux, P., Chollet, F. & Geraud, G. Hypothalamic activation in spontaneous migraine attacks. Headache 47, 1418–1426 (2007). This study was the first to demonstrate activation of the hypothalamus during attacks of migraine without aura.

    PubMed  Google Scholar 

  75. 75.

    Maniyar, F. H., Sprenger, T., Monteith, T., Schankin, C. & Goadsby, P. J. Brain activations in the premonitory phase of nitroglycerin-triggered migraine attacks. Brain J. Neurol. 137, 232–241 (2014). This study showed that the hypothalamus is activated during the premonitory phase of migraine attacks.

    Google Scholar 

  76. 76.

    Schulte, L. H. & May, A. The migraine generator revisited: continuous scanning of the migraine cycle over 30 days and three spontaneous attacks. Brain J. Neurol. 139, 1987–1993 (2016). Daily recordings in a patient with migraine confirmed that the hypothalamus is activated hours before the migraine headache starts.

    Google Scholar 

  77. 77.

    Montagna, P., Pierangeli, G. & Cortelli, P. The primary headaches as a reflection of genetic darwinian adaptive behavioral responses. Headache 50, 273–289 (2010).

    PubMed  Google Scholar 

  78. 78.

    Pearce, J. Insulin induced hypoglycaemia in migraine. J. Neurol. Neurosurg. Psychiatry 34, 154–156 (1971).

    CAS  PubMed  PubMed Central  Google Scholar 

  79. 79.

    Hockaday, J. M., Williamson, D. H. & Whitty, C. W. M. Blood-glucose levels and fatty-acid metabolism in migraine related to fasting. Lancet 297, 1153–1156 (1971). An early study of metabolic responses and migraine attack generation after a glucose tolerance test; 6 of 10 patients developed an attack and increases in free fatty acids and ketone bodies.

    Google Scholar 

  80. 80.

    Christiansen, I., Thomsen, L. L., Daugaard, D., Ulrich, V. & Olesen, J. Glyceryl trinitrate induces attacks of migraine without aura in sufferers of migraine with aura. Cephalalgia 19, 660–667 (1999). discussion 626.

    CAS  PubMed  Google Scholar 

  81. 81.

    Shaw, S. W., Johnson, R. H. & Keogh, H. J. Metabolic changes during glucose tolerance tests in migraine attacks. J. Neurol. Sci. 33, 51–59 (1977).

    CAS  PubMed  Google Scholar 

  82. 82.

    Dexter, J. D., Roberts, J. & Byer, J. A. The five hour glucose tolerance test and effect of low sucrose diet in migraine. Headache 18, 91–94 (1978).

    CAS  PubMed  Google Scholar 

  83. 83.

    Cavestro, C. et al. Insulin metabolism is altered in migraineurs: a new pathogenic mechanism for migraine? Headache 47, 1436–1442 (2007).

    PubMed  Google Scholar 

  84. 84.

    Rainero, I., Govone, F., Gai, A., Vacca, A. & Rubino, E. Is migraine primarily a metaboloendocrine disorder? Curr. Pain Headache Rep. 22, 36 (2018).

    PubMed  Google Scholar 

  85. 85.

    Sacco, S. et al. Insulin resistance in migraineurs: results from a case-control study. Cephalalgia 34, 349–356 (2014).

    CAS  PubMed  Google Scholar 

  86. 86.

    Kokavec, A. Effect of sucrose consumption on serum insulin, serum cortisol and insulin sensitivity in migraine: evidence of sex differences. Physiol. Behav. 142, 170–178 (2015).

    CAS  PubMed  Google Scholar 

  87. 87.

    Siva, Z. O. et al. Determinants of glucose metabolism and the role of NPY in the progression of insulin resistance in chronic migraine. Cephalalgia 38, 1773–1781 (2018).

    PubMed  Google Scholar 

  88. 88.

    Brand-Miller, J. C., Griffin, H. J. & Colagiuri, S. The carnivore connection hypothesis: revisited. J. Obes. 2012, 258624 (2012).

    PubMed  Google Scholar 

  89. 89.

    Issad, T. et al. Effects of fasting on tissue glucose utilization in conscious resting rats. Major glucose-sparing effect in working muscles. Biochem. J. 246, 241–244 (1987).

    CAS  PubMed  PubMed Central  Google Scholar 

  90. 90.

    Stepien, M. et al. Increasing protein at the expense of carbohydrate in the diet down-regulates glucose utilization as glucose sparing effect in rats. PLOS ONE 6, e14664 (2011).

    CAS  PubMed  PubMed Central  Google Scholar 

  91. 91.

    Wang, X. et al. Are glucose and insulin metabolism and diabetes associated with migraine? A community-based, case-control study. J. Oral Facial Pain Headache 31, 240–250 (2017).

    PubMed  Google Scholar 

  92. 92.

    Antonazzo, I. C. et al. Diabetes is associated with decreased migraine risk: a nationwide cohort study. Cephalalgia 38, 1759–1764 (2018).

    PubMed  Google Scholar 

  93. 93.

    Streel, S. et al. Screening for the metabolic syndrome in subjects with migraine. Cephalalgia 37, 1180–1188 (2017).

    PubMed  Google Scholar 

  94. 94.

    He, Z. et al. Metabolic syndrome in female migraine patients is associated with medication overuse headache: a clinic-based study in China. Eur. J. Neurol. 22, 1228–1234 (2015).

    CAS  PubMed  Google Scholar 

  95. 95.

    Bigal, M. E., Liberman, J. N. & Lipton, R. B. Obesity and migraine: a population study. Neurology 66, 545–550 (2006). A population study that revealed an association between obesity and chronic migraine.

    PubMed  Google Scholar 

  96. 96.

    Ziegler, D. K., Hassanein, R. S., Kodanaz, A. & Meek, J. C. Circadian rhythms of plasma cortisol in migraine. J. Neurol. Neurosurg. Psychiatry 42, 741–748 (1979).

    CAS  PubMed  PubMed Central  Google Scholar 

  97. 97.

    Peres, M. F. et al. Hypothalamic involvement in chronic migraine. J. Neurol. Neurosurg. Psychiatry 71, 747–751 (2001).

    CAS  PubMed  PubMed Central  Google Scholar 

  98. 98.

    Lippi, G. & Mattiuzzi, C. Cortisol and migraine: a systematic literature review. Agri 29, 95–99 (2017).

    PubMed  Google Scholar 

  99. 99.

    Hsu, L. K. et al. Early morning migraine. Nocturnal plasma levels of catecholamines, tryptophan, glucose, and free fatty acids and sleep encephalographs. Lancet 1, 447–451 (1977).

    CAS  PubMed  Google Scholar 

  100. 100.

    Coggan, J. S. et al. Norepinephrine stimulates glycogenolysis in astrocytes to fuel neurons with lactate. PLOS Comput. Biol. 14, e1006392 (2018).

    PubMed  PubMed Central  Google Scholar 

  101. 101.

    Jacome, D. E. Hypoglycemia rebound migraine. Headache 41, 895–898 (2001).

    CAS  PubMed  Google Scholar 

  102. 102.

    De Silva, K. L., Ron, M. A. & Pearce, J. Blood sugar response to glucagon in migraine. J. Neurol. Neurosurg. Psychiatry 37, 105–107 (1974).

    PubMed  PubMed Central  Google Scholar 

  103. 103.

    Peterlin, B. L., Sacco, S., Bernecker, C. & Scher, A. I. Adipokines and migraine: a systematic review. Headache 56, 622–644 (2016).

    PubMed  PubMed Central  Google Scholar 

  104. 104.

    Guldiken, B., Guldiken, S., Demir, M., Turgut, N. & Tugrul, A. Low leptin levels in migraine: a case control study. Headache 48, 1103–1107 (2008).

    PubMed  Google Scholar 

  105. 105.

    Domínguez, C. et al. Role of adipocytokines in the pathophysiology of migraine: a cross-sectional study. Cephalalgia 38, 904–911 (2018).

    PubMed  Google Scholar 

  106. 106.

    Van Houten, B., Hunter, S. E. & Meyer, J. N. Mitochondrial DNA damage induced autophagy, cell death, and disease. Front. Biosci. Landmark Ed. 21, 42–54 (2016).

    PubMed  PubMed Central  Google Scholar 

  107. 107.

    Yang, J.-L., Weissman, L., Bohr, V. A. & Mattson, M. P. Mitochondrial DNA damage and repair in neurodegenerative disorders. DNA Repair 7, 1110–1120 (2008).

    CAS  PubMed  PubMed Central  Google Scholar 

  108. 108.

    Roos-Araujo, D., Stuart, S., Lea, R. A., Haupt, L. M. & Griffiths, L. R. Epigenetics and migraine; complex mitochondrial interactions contributing to disease susceptibility. Gene 543, 1–7 (2014).

    CAS  PubMed  Google Scholar 

  109. 109.

    MacGregor, E. A. Oestrogen and attacks of migraine with and without aura. Lancet Neurol. 3, 354–361 (2004).

    CAS  PubMed  Google Scholar 

  110. 110.

    Lemos, C. et al. Assessing risk factors for migraine: differences in gender transmission. PLOS ONE 7, e50626 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  111. 111.

    Kraya, T., Deschauer, M., Joshi, P. R., Zierz, S. & Gaul, C. Prevalence of headache in patients with mitochondrial disease: a cross-sectional study. Headache 58, 45–52 (2018).

    PubMed  Google Scholar 

  112. 112.

    Vollono, C., Primiano, G., Della Marca, G., Losurdo, A. & Servidei, S. Migraine in mitochondrial disorders: prevalence and characteristics. Cephalalgia 38, 1093–1106 (2018).

    PubMed  Google Scholar 

  113. 113.

    Montagna, P. et al. MELAS syndrome: characteristic migrainous and epileptic features and maternal transmission. Neurology 38, 751–754 (1988).

    CAS  PubMed  Google Scholar 

  114. 114.

    Guo, S. et al. Prevalence of migraine in persons with the 3243A>G mutation in mitochondrial DNA. Eur. J. Neurol. 23, 175–181 (2016).

    CAS  PubMed  Google Scholar 

  115. 115.

    Altmann, J. et al. Expanded phenotypic spectrum of the m.8344A>G “MERRF” mutation: data from the German mitoNET registry. J. Neurol. 263, 961–972 (2016).

    CAS  PubMed  Google Scholar 

  116. 116.

    Buzzi, M. G. et al. mtDNA A3243G MELAS mutation is not associated with multigenerational female migraine. Neurology 54, 1005–1007 (2000).

    CAS  PubMed  Google Scholar 

  117. 117.

    Cevoli, S. et al. High frequency of migraine-only patients negative for the 3243 A>G tRNALeu mtDNA mutation in two MELAS families. Cephalalgia 30, 919–927 (2010).

    PubMed  Google Scholar 

  118. 118.

    Fachal, L. et al. No evidence of association between common European mitochondrial DNA variants in Alzheimer, Parkinson, and migraine in the Spanish population. Am. J. Med. Genet. B Neuropsychiatr. Genet. 168B, 54–65 (2015).

    PubMed  Google Scholar 

  119. 119.

    Haan, J. et al. Search for mitochondrial DNA mutations in migraine subgroups. Cephalalgia 19, 20–22 (1999).

    CAS  PubMed  Google Scholar 

  120. 120.

    Klopstock, T. et al. Mitochondrial DNA in migraine with aura. Neurology 46, 1735–1738 (1996). A negative search for mtDNA mutations associated with migraine with aura.

    CAS  PubMed  Google Scholar 

  121. 121.

    Rozen, T. D. et al. Study of mitochondrial DNA mutations in patients with migraine with prolonged aura. Headache 44, 674–677 (2004).

    PubMed  Google Scholar 

  122. 122.

    Russell, M. B., Diamant, M. & Nørby, S. Genetic heterogeneity of migraine with and without aura in Danes cannot be explained by mutation in mtDNA nucleotide pair 11084. Acta Neurol. Scand. 96, 171–173 (1997).

    CAS  PubMed  Google Scholar 

  123. 123.

    Ojaimi, J., Katsabanis, S., Bower, S., Quigley, A. & Byrne, E. Mitochondrial DNA in stroke and migraine with aura. Cerebrovasc. Dis. 8, 102–106 (1998).

    CAS  PubMed  Google Scholar 

  124. 124.

    Larsen, S. et al. Increased intrinsic mitochondrial function in humans with mitochondrial haplogroup H. Biochim. Biophys. Acta 1837, 226–231 (2014).

    CAS  PubMed  Google Scholar 

  125. 125.

    Martínez-Redondo, D. et al. Human mitochondrial haplogroup H: the highest VO2max consumer–is it a paradox? Mitochondrion 10, 102–107 (2010).

    PubMed  Google Scholar 

  126. 126.

    Di Lorenzo, C. et al. Mitochondrial DNA haplogroups influence the therapeutic response to riboflavin in migraineurs. Neurology 72, 1588–1594 (2009). A pharmacogenic study that showed that patients with migraine who have non-H mtDNA haplogroups respond better to high-dose riboflavin than those with the H haplogroup.

    PubMed  Google Scholar 

  127. 127.

    Majamaa, K., Finnilä, S., Turkka, J. & Hassinen, I. E. Mitochondrial DNA haplogroup U as a risk factor for occipital stroke in migraine. Lancet 352, 455–456 (1998).

    CAS  PubMed  Google Scholar 

  128. 128.

    Wang, Q. et al. Mitochondrial DNA control region sequence variation in migraine headache and cyclic vomiting syndrome. Am. J. Med. Genet. A. 131, 50–58 (2004).

    PubMed  Google Scholar 

  129. 129.

    Zaki, E. et al. Two common mitochondrial dna polymorphisms are highly associated with migraine headache and cyclic vomiting syndrome. Cephalalgia 29, 719–728 (2009). A study that showed that childhood migraine and cyclic vomiting are associated with a high prevalence of two mtDNA polymorphisms.

    CAS  PubMed  Google Scholar 

  130. 130.

    Boles, R. G. et al. Increased prevalence of two mitochondrial DNA polymorphisms in functional disease: are we describing different parts of an energy-depleted elephant? Mitochondrion 23, 1–6 (2015).

    CAS  PubMed  Google Scholar 

  131. 131.

    Stuart, S. & Griffiths, L. R. A possible role for mitochondrial dysfunction in migraine. Mol. Genet. Genomics 287, 837–844 (2012).

    CAS  PubMed  Google Scholar 

  132. 132.

    Stuart, S. et al. Gene-centric analysis implicates nuclear encoded mitochondrial protein gene variants in migraine susceptibility. Mol. Genet. Genomic Med. 5, 157–163 (2017). The first evidence that polymorphisms in genes that encode nuclear-encoded mitochondrial proteins might play a role in migraine.

    CAS  PubMed  PubMed Central  Google Scholar 

  133. 133.

    Palmirotta, R. et al. Is SOD2 Ala16Val polymorphism associated with migraine with aura phenotype? Antioxid. Redox Signal. 22, 275–279 (2015).

    CAS  PubMed  PubMed Central  Google Scholar 

  134. 134.

    Saygi, S. et al. Superoxide dismutase and catalase genotypes in pediatric migraine patients. J. Child Neurol. 30, 1586–1590 (2015).

    PubMed  Google Scholar 

  135. 135.

    Kowa, H. et al. The homozygous C677T mutation in the methylenetetrahydrofolate reductase gene is a genetic risk factor for migraine. Am. J. Med. Genet. 96, 762–764 (2000).

    CAS  PubMed  Google Scholar 

  136. 136.

    Lea, R. A., Ovcaric, M., Sundholm, J., MacMillan, J. & Griffiths, L. R. The methylenetetrahydrofolate reductase gene variant C677T influences susceptibility to migraine with aura. BMC Med. 2, 3 (2004).

    PubMed  PubMed Central  Google Scholar 

  137. 137.

    Bhattacharjee, N. & Borah, A. Oxidative stress and mitochondrial dysfunction are the underlying events of dopaminergic neurodegeneration in homocysteine rat model of Parkinson’s disease. Neurochem. Int. 101, 48–55 (2016).

    CAS  PubMed  Google Scholar 

  138. 138.

    Kaunisto, M. et al. Testing of variants of the MTHFR and ESR1 genes in 1798 Finnish individuals fails to confirm the association with migraine with aura. Cephalalgia 26, 1462–1472 (2006).

    CAS  PubMed  Google Scholar 

  139. 139.

    Lee, J., Wong, S. A., Li, B. U. K. & Boles, R. G. NextGen nuclear DNA sequencing in cyclic vomiting syndrome reveals a significant association with the stress-induced calcium channel (RYR2). Neurogastroenterol. Motil. 27, 990–996 (2015).

    CAS  PubMed  Google Scholar 

  140. 140.

    Curtain, R., Tajouri, L., Lea, R., MacMillan, J. & Griffiths, L. No mutations detected in the INSR gene in a chromosome 19p13 linked migraine pedigree. Eur. J. Med. Genet. 49, 57–62 (2006).

    PubMed  Google Scholar 

  141. 141.

    Kaunisto, M. A. et al. Chromosome 19p13 loci in Finnish migraine with aura families. Am. J. Med. Genet. B Neuropsychiatr. Genet. 132B, 85–89 (2005).

    PubMed  PubMed Central  Google Scholar 

  142. 142.

    McCarthy, L. C. et al. Single-nucleotide polymorphism alleles in the insulin receptor gene are associated with typical migraine. Genomics 78, 135–149 (2001).

    CAS  PubMed  Google Scholar 

  143. 143.

    Netzer, C. et al. Replication study of the insulin receptor gene in migraine with aura. Genomics 91, 503–507 (2008).

    CAS  PubMed  Google Scholar 

  144. 144.

    Mohammad, S. S., Coman, D. & Calvert, S. Glucose transporter 1 deficiency syndrome and hemiplegic migraines as a dominant presenting clinical feature. J. Paediatr. Child Health 50, 1025–1026 (2014).

    PubMed  Google Scholar 

  145. 145.

    Gormley, P. et al. Meta-analysis of 375,000 individuals identifies 38 susceptibility loci for migraine. Nat. Genet. 48, 856–866 (2016). A large meta-analysis of genome-wide association studies, which showed that a great number of loci are each responsible for a very small proportion of the genetic risk of migraine.

    CAS  PubMed  PubMed Central  Google Scholar 

  146. 146.

    Eising, E. et al. Gene co-expression analysis identifies brain regions and cell types involved in migraine pathophysiology: a GWAS-based study using the Allen Human Brain Atlas. Hum. Genet. 135, 425–439 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  147. 147.

    Hershey, A., Horn, P., Kabbouche, M., O’Brien, H. & Powers, S. Genomic expression patterns in menstrual-related migraine in adolescents. Headache 52, 68–79 (2012).

    PubMed  PubMed Central  Google Scholar 

  148. 148.

    Woldeamanuel, Y., Rapoport, A. & Cowan, R. The place of corticosteroids in migraine attack management: a 65-year systematic review with pooled analysis and critical appraisal. Cephalalgia 35, 996–1024 (2015).

    CAS  PubMed  Google Scholar 

  149. 149.

    Derry, C. J., Derry, S. & Moore, R. A. Caffeine as an analgesic adjuvant for acute pain in adults. Cochrane Database Syst. Rev. 11, CD009281 (2014).

    Google Scholar 

  150. 150.

    Nagatomo, K. & Kubo, Y. Caffeine activates mouse TRPA1 channels but suppresses human TRPA1 channels. Proc. Natl Acad. Sci. USA 105, 17373–17378 (2008).

    CAS  PubMed  Google Scholar 

  151. 151.

    Lovallo, W. R. et al. Caffeine stimulation of cortisol secretion across the waking hours in relation to caffeine intake levels. Psychosom. Med. 67, 734–739 (2005).

    CAS  PubMed  PubMed Central  Google Scholar 

  152. 152.

    Wu, B. H. Dose effects of caffeine ingestion on acute hormonal responses to resistance exercise. J. Sports Med. Phys. Fit. 55, 1242–1251 (2015).

    CAS  Google Scholar 

  153. 153.

    Khani, S. & Tayek, J. A. Cortisol increases gluconeogenesis in humans: its role in the metabolic syndrome. Clin. Sci. 101, 739–747 (2001).

    CAS  PubMed  Google Scholar 

  154. 154.

    Pagano, G. et al. An in vivo and in vitro study of the mechanism of prednisone-induced insulin resistance in healthy subjects. J. Clin. Invest. 72, 1814–1820 (1983).

    CAS  PubMed  PubMed Central  Google Scholar 

  155. 155.

    Bigal, M. E. & Lipton, R. B. Concepts and mechanisms of migraine chronification. Headache 48, 7–15 (2008).

    PubMed  Google Scholar 

  156. 156.

    Lee, M. J., Choi, H. A., Choi, H. & Chung, C.-S. Caffeine discontinuation improves acute migraine treatment: a prospective clinic-based study. J. Headache Pain 17, 71 (2016).

    PubMed  PubMed Central  Google Scholar 

  157. 157.

    Kalra, E. K. Nutraceutical–definition and introduction. AAPS PharmSci 5, 27–28 (2003).

    PubMed Central  Google Scholar 

  158. 158.

    Rajapakse, T. & Pringsheim, T. Nutraceuticals in migraine: a summary of existing guidelines for use. Headache 56, 808–816 (2016).

    PubMed  Google Scholar 

  159. 159.

    Shaik, M. M. & Gan, S. H. Vitamin supplementation as possible prophylactic treatment against migraine with aura and menstrual migraine. Biomed Res. Int. 2015, 469529 (2015).

    PubMed  PubMed Central  Google Scholar 

  160. 160.

    Barile, M., Giancaspero, T. A., Leone, P., Galluccio, M. & Indiveri, C. Riboflavin transport and metabolism in humans. J. Inherit. Metab. Dis. 39, 545–557 (2016).

    CAS  PubMed  Google Scholar 

  161. 161.

    Bütün, A., Nazıroğlu, M., Demirci, S., Çelik, Ö. & Uğuz, A. C. Riboflavin and vitamin E increase brain calcium and antioxidants, and microsomal calcium-ATP-ase values in rat headache models induced by glyceryl trinitrate. J. Membr. Biol. 248, 205–213 (2015).

    PubMed  Google Scholar 

  162. 162.

    Marashly, E. T. & Bohlega, S. A. Riboflavin has neuroprotective potential: focus on Parkinson’s disease and migraine. Front. Neurol. 8, 333 (2017).

    PubMed  PubMed Central  Google Scholar 

  163. 163.

    Boehnke, C. et al. High-dose riboflavin treatment is efficacious in migraine prophylaxis: an open study in a tertiary care centre. Eur. J. Neurol. 11, 475–477 (2004).

    CAS  PubMed  Google Scholar 

  164. 164.

    Condò, M., Posar, A., Arbizzani, A. & Parmeggiani, A. Riboflavin prophylaxis in pediatric and adolescent migraine. J. Headache Pain 10, 361–365 (2009).

    PubMed  PubMed Central  Google Scholar 

  165. 165.

    Gaul, C., Diener, H.-C. & Danesch, U., Migravent® Study Group. Improvement of migraine symptoms with a proprietary supplement containing riboflavin, magnesium and Q10: a randomized, placebo-controlled, double-blind, multicenter trial. J. Headache Pain 16, 516 (2015).

    PubMed  Google Scholar 

  166. 166.

    Schoenen, J., Jacquy, J. & Lenaerts, M. Effectiveness of high-dose riboflavin in migraine prophylaxis. A randomized controlled trial. Neurology 50, 466–470 (1998). A placebo-controlled trial that showed that 400 mg riboflavin daily is effective for migraine prevention.

    CAS  PubMed  Google Scholar 

  167. 167.

    Bruijn, J. et al. Medium-dose riboflavin as a prophylactic agent in children with migraine: a preliminary placebo-controlled, randomised, double-blind, cross-over trial. Cephalalgia 30, 1426–1434 (2010).

    PubMed  Google Scholar 

  168. 168.

    Rahimdel, A., Mellat, A., Zeinali, A., Jafari, E. & Ayatollahi, P. Comparison between intravenous sodium valproate and subcutaneous sumatriptan for treatment of acute migraine attacks; double-blind randomized clinical trial. Iran. J. Med. Sci. 39, 171–177 (2014).

    PubMed  PubMed Central  Google Scholar 

  169. 169.

    Thompson, D. F. & Saluja, H. S. Prophylaxis of migraine headaches with riboflavin: a systematic review. J. Clin. Pharm. Ther. 42, 394–403 (2017).

    CAS  PubMed  Google Scholar 

  170. 170.

    Prangthip, P., Kettawan, A., Posuwan, J., Okuno, M. & Okamoto, T. An improvement of oxidative stress in diabetic rats by ubiquinone-10 and ubiquinol-10 and bioavailability after short- and long-term coenzyme Q10 supplementation. J. Diet. Suppl. 13, 647–659 (2016).

    CAS  PubMed  Google Scholar 

  171. 171.

    Yang, X. et al. Neuroprotection of coenzyme Q10 in neurodegenerative diseases. Curr. Top. Med. Chem. 16, 858–866 (2016).

    CAS  PubMed  Google Scholar 

  172. 172.

    Dahri, M., Hashemilar, M., Asghari-Jafarabadi, M. & Tarighat-Esfanjani, A. Efficacy of coenzyme Q10 for the prevention of migraine in women: a randomized, double-blind, placebo-controlled study. Eur. J. Integr. Med. 16, 8–14 (2017).

    Google Scholar 

  173. 173.

    Dahri, M., Tarighat-Esfanjani, A., Asghari-Jafarabadi, M. & Hashemilar, M. Oral coenzyme Q10 supplementation in patients with migraine: effects on clinical features and inflammatory markers. Nutr. Neurosci. 22, 607–615 (2019).

    CAS  PubMed  Google Scholar 

  174. 174.

    Sándor, P. S. et al. Efficacy of coenzyme Q10 in migraine prophylaxis: a randomized controlled trial. Neurology 64, 713–715 (2005).

    PubMed  Google Scholar 

  175. 175.

    Hajihashemi, P., Askari, G., Khorvash, F., Reza Maracy, M. & Nourian, M. The effects of concurrent coenzyme Q10, l-carnitine supplementation in migraine prophylaxis: a randomized, placebo-controlled, double-blind trial. Cephalalgia 39, 648–654 (2019).

    PubMed  Google Scholar 

  176. 176.

    Slater, S. K. et al. A randomized, double-blinded, placebo-controlled, crossover, add-on study of coenzyme Q10 in the prevention of pediatric and adolescent migraine. Cephalalgia 31, 897–905 (2011).

    PubMed  Google Scholar 

  177. 177.

    Hershey, A. D. et al. Coenzyme Q10 deficiency and response to supplementation in pediatric and adolescent migraine. Headache 47, 73–80 (2007).

    PubMed  Google Scholar 

  178. 178.

    Müller, U. & Krieglstein, J. Prolonged pretreatment with α-lipoic acid protects cultured neurons against hypoxic, glutamate-, or iron-induced injury. J. Cereb. Blood Flow Metab. 15, 624–630 (1995).

    PubMed  Google Scholar 

  179. 179.

    Packer, L., Witt, E. H. & Tritschler, H. J. alpha-Lipoic acid as a biological antioxidant. Free. Radic. Biol. Med. 19, 227–250 (1995).

    CAS  PubMed  Google Scholar 

  180. 180.

    Magis, D. et al. A randomized double-blind placebo-controlled trial of thioctic acid in migraine prophylaxis. Headache 47, 52–57 (2007).

    PubMed  Google Scholar 

  181. 181.

    Cavestro, C. et al. Alpha-lipoic acid shows promise to improve migraine in patients with insulin resistance: a 6-month exploratory study. J. Med. Food 21, 269–273 (2018).

    CAS  PubMed  Google Scholar 

  182. 182.

    Ali, A. M., Awad, T. G. & Al-Adl, N. M. Efficacy of combined topiramate/thioctic acid therapy in migraine prophylaxis. Saudi Pharm. J. 18, 239–243 (2010).

    CAS  PubMed  PubMed Central  Google Scholar 

  183. 183.

    Lea, R., Colson, N., Quinlan, S., Macmillan, J. & Griffiths, L. The effects of vitamin supplementation and MTHFR (C677T) genotype on homocysteine-lowering and migraine disability. Pharmacogenet. Genomics 19, 422–428 (2009).

    CAS  PubMed  Google Scholar 

  184. 184.

    Menon, S. et al. Genotypes of the MTHFR C677T and MTRR A66G genes act independently to reduce migraine disability in response to vitamin supplementation. Pharmacogenet. Genomics 22, 741–749 (2012).

    CAS  PubMed  Google Scholar 

  185. 185.

    Menon, S. et al. The effect of 1 mg folic acid supplementation on clinical outcomes in female migraine with aura patients. J. Headache Pain 17, 60 (2016).

    PubMed  PubMed Central  Google Scholar 

  186. 186.

    Prousky, J. & Seely, D. The treatment of migraines and tension-type headaches with intravenous and oral niacin (nicotinic acid): systematic review of the literature. Nutr. J. 4, 3 (2005).

    PubMed  PubMed Central  Google Scholar 

  187. 187.

    Facchinetti, F., Sances, G., Borella, P., Genazzani, A. R. & Nappi, G. Magnesium prophylaxis of menstrual migraine: effects on intracellular magnesium. Headache 31, 298–301 (1991).

    CAS  PubMed  Google Scholar 

  188. 188.

    Chiu, H.-Y., Yeh, T.-H., Huang, Y.-C. & Chen, P.-Y. Effects of intravenous and oral magnesium on reducing migraine: a meta-analysis of randomized controlled trials. Pain Physician 19, E97–E112 (2016).

    PubMed  Google Scholar 

  189. 189.

    Karimi, N., Razian, A. & Heidari, M. The efficacy of magnesium oxide and sodium valproate in prevention of migraine headache: a randomized, controlled, double-blind, crossover study. Acta Neurol. Belg. (2019).

  190. 190.

    Tarighat Esfanjani, A. et al. The effects of magnesium, L-carnitine, and concurrent magnesium-L-carnitine supplementation in migraine prophylaxis. Biol. Trace Elem. Res. 150, 42–48 (2012).

    PubMed  Google Scholar 

  191. 191.

    Hagen, K. et al. Acetyl-l-carnitine versus placebo for migraine prophylaxis: a randomized, triple-blind, crossover study. Cephalalgia 35, 987–995 (2015).

    PubMed  Google Scholar 

  192. 192.

    Yang, H., Shan, W., Zhu, F., Wu, J. & Wang, Q. Ketone bodies in neurological diseases: focus on neuroprotection and underlying mechanisms. Front. Neurol. 10, 585 (2019).

    PubMed  PubMed Central  Google Scholar 

  193. 193.

    Gross, E. C., Klement, R. J., Schoenen, J., D’Agostino, D. P. & Fischer, D. Potential protective mechanisms of ketone bodies in migraine prevention. Nutrients 11, 811 (2019). A comprehensive review of the various potential mechanisms by which ketone bodies might protect against migraine.

    CAS  PubMed Central  Google Scholar 

  194. 194.

    Al-Karagholi, M. A.-M., Hansen, J. M., Guo, S., Olesen, J. & Ashina, M. Opening of ATP-sensitive potassium channels causes migraine attacks: a new target for the treatment of migraine. Brain J. Neurol. awz199 (2019). A demonstration that a K ATP channel opener induces migraine headache in patients with migraine.

  195. 195.

    Strahlman, R. S. Can ketosis help migraine sufferers? A case report. Headache 46, 182 (2006).

    PubMed  Google Scholar 

  196. 196.

    Di Lorenzo, C. et al. Diet transiently improves migraine in two twin sisters: possible role of ketogenesis? Funct. Neurol. 28, 305–308 (2013). The first clinical indication that a ketogenic diet can improve migraine.

    PubMed  Google Scholar 

  197. 197.

    Maggioni, F., Margoni, M. & Zanchin, G. Ketogenic diet in migraine treatment: a brief but ancient history. Cephalalgia 31, 1150–1151 (2011).

    PubMed  Google Scholar 

  198. 198.

    SCHNABEL, T. G. An experience with a ketogenic dietary in migraine. Ann. Intern. Med. 2, 341 (1928).

    Google Scholar 

  199. 199.

    Di Lorenzo, C. et al. Migraine improvement during short lasting ketogenesis: a proof-of-concept study. Eur. J. Neurol. 22, 170–177 (2015).

    PubMed  Google Scholar 

  200. 200.

    Di Lorenzo, C. et al. Cortical functional correlates of responsiveness to short-lasting preventive intervention with ketogenic diet in migraine: a multimodal evoked potentials study. J. Headache Pain 17, 58 (2016).

    PubMed  PubMed Central  Google Scholar 

  201. 201.

    Di Lorenzo, C. et al. A randomized double-blind, cross-over trial of very low-calorie diet in overweight migraine patients: a possible role for ketones? Nutrients 11, 1742 (2019).

    PubMed Central  Google Scholar 

  202. 202.

    Gross, E. et al. Efficacy and safety of exogenous ketone bodies for preventive treatment of migraine: a study protocol for a single-centred, randomised, placebo-controlled, double-blind crossover trial. Trials 20, 61 (2019).

    PubMed  PubMed Central  Google Scholar 

  203. 203.

    Varkey, E., Cider, A., Carlsson, J. & Linde, M. Exercise as migraine prophylaxis: a randomized study using relaxation and topiramate as controls. Cephalalgia 31, 1428–1438 (2011).

    PubMed  PubMed Central  Google Scholar 

  204. 204.

    Ding, Q., Vaynman, S., Souda, P., Whitelegge, J. P. & Gomez-Pinilla, F. Exercise affects energy metabolism and neural plasticity-related proteins in the hippocampus as revealed by proteomic analysis. Eur. J. Neurosci. 24, 1265–1276 (2006).

    PubMed  Google Scholar 

  205. 205.

    Steiner, J. L., Murphy, E. A., McClellan, J. L., Carmichael, M. D. & Davis, J. M. Exercise training increases mitochondrial biogenesis in the brain. J. Appl. Physiol. (1985) 111, 1066–1071 (2011).

    CAS  Google Scholar 

  206. 206.

    Silberstein, S. D., Latsko, M. & Schoenen, J. in Multidisciplinary Management of Migraine: Pharmacological, Manual, and Other Therapies (eds Fernandez-de-las-Penas, C., Chaitow, L., & Schoenen, J.) 91–102 (Jones & Bartlett Learning, 2013).

  207. 207.

    Motaghinejad, M., Motevalian, M. & Shabab, B. Neuroprotective effects of various doses of topiramate against methylphenidate induced oxidative stress and inflammation in rat isolated hippocampus. Clin. Exp. Pharmacol. Physiol. 43, 360–371 (2016).

    CAS  PubMed  Google Scholar 

  208. 208.

    Kudin, A. P., Debska-Vielhaber, G., Vielhaber, S., Elger, C. E. & Kunz, W. S. The mechanism of neuroprotection by topiramate in an animal model of epilepsy. Epilepsia 45, 1478–1487 (2004).

    CAS  PubMed  Google Scholar 

  209. 209.

    Franzoni, E. et al. Topiramate: effects on serum lipids and lipoproteins levels in children. Eur. J. Neurol. 14, 1334–1337 (2007).

    CAS  PubMed  Google Scholar 

  210. 210.

    Wilkes, J. J., Nelson, E., Osborne, M., Demarest, K. T. & Olefsky, J. M. Topiramate is an insulin-sensitizing compound in vivo with direct effects on adipocytes in female ZDF rats. Am. J. Physiol. Endocrinol. Metab. 288, E617–E624 (2005).

    CAS  PubMed  Google Scholar 

  211. 211.

    Li, R. et al. Valproate attenuates nitroglycerin-induced trigeminovascular activation by preserving mitochondrial function in a rat model of migraine. Med. Sci. Monit. 22, 3229–3237 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  212. 212.

    Sitarz, K. S. et al. Valproic acid triggers increased mitochondrial biogenesis in POLG-deficient fibroblasts. Mol. Genet. Metab. 112, 57–63 (2014).

    CAS  PubMed  PubMed Central  Google Scholar 

  213. 213.

    Berilgen, M. S. et al. Comparison of the effects of amitriptyline and flunarizine on weight gain and serum leptin, C peptide and insulin levels when used as migraine preventive treatment. Cephalalgia 25, 1048–1053 (2005).

    CAS  PubMed  Google Scholar 

  214. 214.

    Maggioni, F., Ruffatti, S., Dainese, F., Mainardi, F. & Zanchin, G. Weight variations in the prophylactic therapy of primary headaches: 6-month follow-up. J. Headache Pain 6, 322–324 (2005).

    PubMed  PubMed Central  Google Scholar 

  215. 215.

    Lamont, L. S. Beta-blockers and their effects on protein metabolism and resting energy expenditure. J. Cardiopulm. Rehabil. 15, 183–185 (1995).

    CAS  PubMed  Google Scholar 

  216. 216.

    Lipton, R. B. et al. Reduction in perceived stress as a migraine trigger: testing the “let-down headache” hypothesis. Neurology 82, 1395–1401 (2014).

    PubMed  PubMed Central  Google Scholar 

  217. 217.

    Neubauer, J. A. & Sunderram, J. Oxygen-sensing neurons in the central nervous system. J. Appl. Physiol. (1985) 96, 367–374 (2004).

    CAS  Google Scholar 

  218. 218.

    Lin, L.-C., Lewis, D. A. & Sibille, E. A human-mouse conserved sex bias in amygdala gene expression related to circadian clock and energy metabolism. Mol. Brain 4, 18 (2011).

    CAS  PubMed  PubMed Central  Google Scholar 

  219. 219.

    Hoffmann, U., Sukhotinsky, I., Eikermann-Haerter, K. & Ayata, C. Glucose modulation of spreading depression susceptibility. J. Cereb. Blood Flow Metab. 33, 191–195 (2013).

    CAS  PubMed  Google Scholar 

  220. 220.

    Kilic, K. et al. Inadequate brain glycogen or sleep increases spreading depression susceptibility. Ann. Neurol. 83, 61–73 (2018). A pivotal study that showed that in rodents, glucose shortage and sleep deprivation increase susceptibility to cortical spreading depression and open pannexin-1 megachannels, enabling activation of the trigeminovascular system.

    CAS  PubMed  Google Scholar 

  221. 221.

    de Almeida Rabello Oliveira, M. et al. Effects of short-term and long-term treatment with medium- and long-chain triglycerides ketogenic diet on cortical spreading depression in young rats. Neurosci. Lett. 434, 66–70 (2008).

    PubMed  Google Scholar 

  222. 222.

    Gerich, F. J., Hepp, S., Probst, I. & Müller, M. Mitochondrial inhibition prior to oxygen-withdrawal facilitates the occurrence of hypoxia-induced spreading depression in rat hippocampal slices. J. Neurophysiol. 96, 492–504 (2006).

    PubMed  Google Scholar 

  223. 223.

    Takano, T. et al. Cortical spreading depression causes and coincides with tissue hypoxia. Nat. Neurosci. 10, 754–762 (2007).

    CAS  PubMed  Google Scholar 

  224. 224.

    Angelova, P. R. et al. Functional oxygen sensitivity of astrocytes. J. Neurosci. 35, 10460–10473 (2015).

    CAS  PubMed  PubMed Central  Google Scholar 

  225. 225.

    Chen, S.-P. et al. Inhibition of the P2X7-PANX1 complex suppresses spreading depolarization and neuroinflammation. Brain 140, 1643–1656 (2017).

    PubMed  Google Scholar 

  226. 226.

    Bolay, H. et al. Intrinsic brain activity triggers trigeminal meningeal afferents in a migraine model. Nat. Med. 8, 136–142 (2002). The first study to show that CSD in rats can activate the trigeminovascular system.

    CAS  PubMed  Google Scholar 

  227. 227.

    Karatas, H. et al. Spreading depression triggers headache by activating neuronal Panx1 channels. Science 339, 1092–1095 (2013). This study showed that CSD-induced trigeminovascular activation is mediated by the opening of neuronal pannexin-1 channels.

    CAS  PubMed  Google Scholar 

  228. 228.

    Feuerstein, D. et al. Regulation of cerebral metabolism during cortical spreading depression. J. Cereb. Blood Flow Metab. 36, 1965–1977 (2016).

    CAS  PubMed  Google Scholar 

  229. 229.

    Yuzawa, I. et al. Cortical spreading depression impairs oxygen delivery and metabolism in mice. J. Cereb. Blood Flow Metab. 32, 376–386 (2012).

    CAS  PubMed  Google Scholar 

  230. 230.

    Viggiano, E. et al. Cortical spreading depression produces a neuroprotective effect activating mitochondrial uncoupling protein-5. Neuropsychiatr. Dis. Treat. 12, 1705–1710 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  231. 231.

    Shatillo, A. et al. Cortical spreading depression induces oxidative stress in the trigeminal nociceptive system. Neuroscience 253, 341–349 (2013).

    CAS  PubMed  Google Scholar 

  232. 232.

    Diener, H.-C. et al. CGRP as a new target in prevention and treatment of migraine. Lancet Neurol. 13, 1065–1067 (2014).

    CAS  PubMed  Google Scholar 

  233. 233.

    Durham, P. L. Calcitonin gene-related peptide (CGRP) and migraine. Headache 46 (Suppl. 1), S3–S8 (2006).

    Google Scholar 

  234. 234.

    Holland, P. R. et al. Acid-sensing ion channel 1: a novel therapeutic target for migraine with aura. Ann. Neurol. 72, 559–563 (2012).

    CAS  PubMed  Google Scholar 

  235. 235.

    Benemei, S., Fusi, C., Trevisan, G. & Geppetti, P. The TRPA1 channel in migraine mechanism and treatment. Br. J. Pharmacol. 171, 2552–2567 (2014).

    CAS  PubMed  PubMed Central  Google Scholar 

  236. 236.

    Kozai, D., Ogawa, N. & Mori, Y. Redox regulation of transient receptor potential channels. Antioxid. Redox Signal. 21, 971–986 (2014).

    CAS  PubMed  Google Scholar 

  237. 237.

    Martins-Oliveira, M. et al. Neuroendocrine signaling modulates specific neural networks relevant to migraine. Neurobiol. Dis. 101, 16–26 (2017). An experimental study that showed that systemic metabolic changes can modulate firing of nociceptors in the spinal trigeminal nucleus.

    CAS  PubMed  PubMed Central  Google Scholar 

  238. 238.

    Ma, W., Berg, J. & Yellen, G. Ketogenic diet metabolites reduce firing in central neurons by opening K(ATP) channels. J. Neurosci. 27, 3618–3625 (2007).

    CAS  PubMed  PubMed Central  Google Scholar 

  239. 239.

    Kawamura, M., Ruskin, D. N. & Masino, S. A. Metabolic autocrine regulation of neurons involves cooperation among pannexin hemichannels, adenosine receptors, and KATP channels. J. Neurosci. 30, 3886–3895 (2010).

    CAS  PubMed  PubMed Central  Google Scholar 

  240. 240.

    Ghasemi, M., Mayasi, Y., Hannoun, A., Eslami, S. M. & Carandang, R. Nitric oxide and mitochondrial function in neurological diseases. Neuroscience 376, 48–71 (2018).

    CAS  PubMed  Google Scholar 

  241. 241.

    Kruuse, C., Thomsen, L. L., Birk, S. & Olesen, J. Migraine can be induced by sildenafil without changes in middle cerebral artery diameter. Brain J. Neurol. 126, 241–247 (2003).

    Google Scholar 

  242. 242.

    Bolaños, J. P., Peuchen, S., Heales, S. J., Land, J. M. & Clark, J. B. Nitric oxide-mediated inhibition of the mitochondrial respiratory chain in cultured astrocytes. J. Neurochem. 63, 910–916 (1994).

    PubMed  Google Scholar 

  243. 243.

    Abad, N. et al. Metabolic assessment of a migraine model using relaxation-enhanced 1H spectroscopy at ultrahigh field. Magn. Reson. Med. 79, 1266–1275 (2018).

    CAS  PubMed  Google Scholar 

  244. 244.

    Dong, X. et al. Abnormal mitochondrial dynamics and impaired mitochondrial biogenesis in trigeminal ganglion neurons in a rat model of migraine. Neurosci. Lett. 636, 127–133 (2017).

    CAS  PubMed  Google Scholar 

  245. 245.

    Fried, N. T., Moffat, C., Seifert, E. L. & Oshinsky, M. L. Functional mitochondrial analysis in acute brain sections from adult rats reveals mitochondrial dysfunction in a rat model of migraine. Am. J. Physiol. Cell Physiol. 307, C1017–C1030 (2014).

    CAS  PubMed  PubMed Central  Google Scholar 

  246. 246.

    Goadsby, P. J., Edvinsson, L. & Ekman, R. Vasoactive peptide release in the extracerebral circulation of humans during migraine headache. Ann. Neurol. 28, 183–187 (1990). By showing that CGRP levels in the blood in the jugular vein are increased during migraine attacks, this study paved the way for the novel migraine therapies that block CGRP neurotransmission.

    CAS  PubMed  Google Scholar 

  247. 247.

    Lassen, L. H. et al. CGRP may play a causative role in migraine. Cephalalgia 22, 54–61 (2002).

    CAS  PubMed  Google Scholar 

  248. 248.

    Khan, S., Olesen, A. & Ashina, M. CGRP, a target for preventive therapy in migraine and cluster headache: systematic review of clinical data. Cephalalgia 39, 374–389 (2019).

    PubMed  Google Scholar 

  249. 249.

    Hansen, J. M., Hauge, A. W., Olesen, J. & Ashina, M. Calcitonin gene-related peptide triggers migraine-like attacks in patients with migraine with aura. Cephalalgia 30, 1179–1186 (2010).

    PubMed  Google Scholar 

  250. 250.

    Bai, Y.-X., Fang, F., Jiang, J.-L. & Xu, F. Extrinsic calcitonin gene-related peptide inhibits hyperoxia-induced alveolar epithelial type II cells apoptosis, oxidative stress, and reactive oxygen species (ROS) production by enhancing notch 1 and homocysteine-induced endoplasmic reticulum protein (HERP) expression. Med. Sci. Monit. 23, 5774–5782 (2017).

    PubMed  PubMed Central  Google Scholar 

  251. 251.

    Dang, H.-X. et al. CGRP attenuates hyperoxia-induced oxidative stress-related injury to alveolar epithelial type II cells via the activation of the Sonic hedgehog pathway. Int. J. Mol. Med. 40, 209–216 (2017).

    CAS  PubMed  Google Scholar 

  252. 252.

    Holzmann, B. Antiinflammatory activities of CGRP modulating innate immune responses in health and disease. Curr. Protein Pept. Sci. 14, 268–274 (2013).

    CAS  PubMed  Google Scholar 

  253. 253.

    Smillie, S.-J. et al. An ongoing role of α-calcitonin gene-related peptide as part of a protective network against hypertension, vascular hypertrophy, and oxidative stress. Hypertension 63, 1056–1062 (2014).

    CAS  PubMed  Google Scholar 

  254. 254.

    Hothersall, J. S., Muirhead, R. P. & Wimalawansa, S. The effect of amylin and calcitonin gene-related peptide on insulin-stimulated glucose transport in the diaphragm. Biochem. Biophys. Res. Commun. 169, 451–454 (1990).

    CAS  PubMed  Google Scholar 

  255. 255.

    Morishita, T. et al. Effects of islet amyloid polypeptide (amylin) and calcitonin gene-related peptide (CGRP) on glucose metabolism in the rat. Diabetes Res. Clin. Pract. 15, 63–69 (1992).

    CAS  PubMed  Google Scholar 

  256. 256.

    Leighton, B. & Foot, E. A. The role of the sensory peptide calcitonin-gene-related peptide(s) in skeletal muscle carbohydrate metabolism: effects of capsaicin and resiniferatoxin. Biochem. J. 307, 707–712 (1995).

    CAS  PubMed  PubMed Central  Google Scholar 

  257. 257.

    Holland, P. R., Saengjaroentham, C. & Vila-Pueyo, M. The role of the brainstem in migraine: potential brainstem effects of CGRP and CGRP receptor activation in animal models. Cephalalgia 39, 390–402 (2019).

    PubMed  Google Scholar 

  258. 258.

    Yi, C.-X. et al. Pituitary adenylate cyclase-activating polypeptide stimulates glucose production via the hepatic sympathetic innervation in rats. Diabetes 59, 1591–1600 (2010).

    CAS  PubMed  PubMed Central  Google Scholar 

  259. 259.

    Lisicki, M. et al. Age related metabolic modifications in the migraine brain. Cephalalgia 39, 978–987 (2019).

    PubMed  Google Scholar 

  260. 260.

    Liveing, E. On Megrim, Sick-headache, and Some Allied Disorders: A Contribution to the Pathology of Nerve-storms. (Churchill, 1873). The formulation of the concept that the migraine attack results from a disruption of brain homeostasis that the attack itself resolves.

  261. 261.

    Loder, E. What is the evolutionary advantage of migraine? Cephalalgia 22, 624–632 (2002).

    CAS  PubMed  Google Scholar 

  262. 262.

    de Tommaso, M. et al. Altered processing of sensory stimuli in patients with migraine. Nat. Rev. Neurol. 10, 144–155 (2014).

    PubMed  Google Scholar 

  263. 263.

    Noseda, R. et al. Migraine photophobia originating in cone-driven retinal pathways. Brain J. Neurol. 139, 1971–1986 (2016).

    Google Scholar 

  264. 264.

    Evers, S. et al. EFNS guideline on the drug treatment of migraine – revised report of an EFNS task force. Eur. J. Neurol. 16, 968–981 (2009).

    CAS  PubMed  Google Scholar 

  265. 265.

    Holland, S. et al. Evidence-based guideline update: NSAIDS and other complementary treatments for episodic migraine prevention in adults: report of the Quality Standards Subcommittee of the American Academy of Neurology and the American Headache Society. Neurology 78, 1346–1353 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  266. 266.

    Marmura, M. J., Silberstein, S. D. & Schwedt, T. J. The acute treatment of migraine in adults: the American Headache Society evidence assessment of migraine pharmacotherapies. Headache 55, 3–20 (2015).

    Google Scholar 

  267. 267.

    Silberstein, S. D. et al. Evidence-based guideline update: pharmacologic treatment for episodic migraine prevention in adults: report of the Quality Standards Subcommittee of the American Academy of Neurology and the American Headache Society. Neurology 78, 1337–1345 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  268. 268.

    Tfelt-Hansen, P. C. Evidence-based guideline update: pharmacologic treatment for episodic migraine prevention in adults: report of the Quality Standards Subcommittee of the American Academy of Neurology and the American Headache Society. Neurology 80, 869–870; correction 871 (2013).

    PubMed  Google Scholar 

  269. 269.

    Pringsheim, T. et al. Canadian Headache Society guideline for migraine prophylaxis. Can. J. Neurol. Sci. 39, S1–S59 (2012).

    PubMed  Google Scholar 

Download references

Author information




E.C.G. was responsible for the literature search and the main composition of the manuscript, including the majority of display items. M.L. edited the manuscript and provided additional text and display items. D.F. and P.S.S. edited the manuscript. J.S. was responsible for the design of the manuscript, edited in depth the manuscript and display items and provided additional text and citations. All authors proofread the final manuscript prior to submission.

Corresponding author

Correspondence to Jean Schoenen.

Ethics declarations

Competing interests

E.C.G. is the founder of KetoSwiss. E.C.G. and D.F. are the inventors of patent WO/2018/115158 held by the University Children’s Hospital Basel (UKBB) and the University of Basel for the use of β-hydroxybutyrate in migraine prevention. M.L., P.S.S. and J.S. declare no competing interests.

Additional information

Peer review information

Nature Reviews Neurology thanks A. Carolei and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Gross, E.C., Lisicki, M., Fischer, D. et al. The metabolic face of migraine — from pathophysiology to treatment. Nat Rev Neurol 15, 627–643 (2019).

Download citation

Further reading


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing