Magnesium in headache (original) (raw)

Abstract

Magnesium's role in migraine pathogenesis is well-described, with deficiencies known to promote cortical spreading depression, alter nociceptive processing and neurotransmitter release, and encourage the hyperaggregation of platelets, all major elements of migraine development. Research on magnesium has found it to be a potentially well-tolerated, safe and inexpensive option for migraine prevention, while it may also be effective as an acute treatment option for headaches including migraines, tension- type headaches and cluster headaches, particularly in certain patient subsets. This chapter will review the various aspects of migraine in which magnesium plays a part, as well as numerous studies on the use of magnesium in both headache prophylaxis and in the acute treatment of headaches, offering recommendations in its use in clinical practice.

Magnesium in the Body

Magnesium (Mg), the second most abundant intracellular divalent cation, is a cofactor of many enzymes and is involved in a plethora of cellular functions. It plays a central role in both glucose metabolism and in ATP function. Over 300 enzymes require the presence of magnesium ions for their catalytic action, including all enzymes utilizing or synthesizing ATP, or those that use other nucleotides to synthesize DNA and RNA. ATP exists in cells as a chelate of ATP and a magnesium ion. Because of the important interaction between phosphate and magnesium ions, magnesium ions are essential to the basic nucleic acid chemistry of life, and thus are essential to all cells of all known living organisms.

Magnesium is involved in the formation of phospholipids and the insertion of proteins into the phospholipid membrane, and is therefore critical to membrane stabilization (Durlach et al., 1987). It also contributes to contraction of the cytoskeleton at the myoneural junction, playing a vital role in the function of both skeletal, cardiac and other smooth muscles.

Magnesium is absorbed in the gastrointestinal tract, via intestinal epithelial channels in the ileum as well as by the renal system’s thick ascending limb, distal tubule, and loop of Henle of the nephron (Wagner, 2007). It facilitates calcium absorption via the thick ascending limb, and the absorption of both ions is regulated by the parathyroid hormone (PTH) secreting cells of the parathyroid gland (Bapty et al., 1998). The calcium/magnesium sensing receptor within the parathyroid gland regulates absorption of both ions by detecting their levels in ionized form, and then controlling PTH secretion, thereby maintain- ing calcium homeostasis (Brown et al., 1993). Dietarily, absorption is affected by protein intake as well as phosphate, phytate and fat. Absorbed dietary magnesium is largely excreted through the urine, although most iatrogenically admin- istered oral magnesium is excreted through the faeces.

Adult human bodies contain approximately 24 grams of magnesium, with 67% located in the skeleton, 31% intracellularly (20% in skeletal muscle), and only 1-2% extracellularly. Of this amount, one half is ionized, and 25-30% is protein bound. As a result, levels found on routine serum testing, which only reflects that magnesium found in the extracellular space, is not represent- ative of true total body magnesium stores (Moe, 2008).

Serum levels are typically 0.7–1.0 mmol/L or 1.8– 2.4 mEq/L. Serum magnesium levels may appear normal even in cases of underlying intracellular deficiency, and true hypomagnesemia is common, possibly due to decreased intake or absorption, increased loss via the urine or diarrhea, or genetic factors (Henrotte, 1982).

Primary and Secondary Hypomagnesemia

Familial hypomagnesemia with secondary hypo- calcemia has been studied in various kindreds, and heredity has been found to be X linked in some families, and autosomal recessive in others (Walder et al., 1997). There are currently more than 30 known mutations in the TRPM6 gene that are associated with familial hypomagnesemia and hypocalcemia. Another hereditary form of hypo- magnesemia, tubular hypomagnesemia/hypo- kalemia with hypocalciuria (Gitelman's synd- rome), is hypothesized to be due to two different types of genetic transmission, one autosomal recessive and one autosomal dominant with high phenotype variability (Bettinelli et al., 1995).

A population study in Germany found the prevalence of serum hypomagnesemia to be 14.5%, with even higher frequencies in females (Schimatschek and Rempis, 2001). Additionally, chronic disease is associated with hypo- magnesemia, including diabetes, asthma, cardio- vascular disease, sickle cell anaemia, pre- eclampsia and eclampsia (Laires et al., 2004). 10- 20% of hospitalized patients are deficient in magnesium, and up to 65% of patients in intensive care units are hypomagnesemic. Alcoholism is also associated with inadequate magnesium levels, in part due to poor nutrition (Bohmer and Mathiesen, 1982). It has also been implicated in patients with end-stage renal disease suffering from hemodialysis headache (Goksel et al., 2006), and is often seen in conjunction with electrolyte abnormalities including hypokalemia, hyponatremia, hypo- calcemia and hypophosphatemia (Whang et al., 1985). A number of medications such as diuretics, aminoglycosides and digoxin are associated with hypomagnesemia, and patients with refractory hypocalcemia and hypokalemia should be evaluated for hypomagnesemia (Innerarity, 2000).

Magnesium Imbalances

Clinical symptoms of hypomagnesemia include hallucinations, depression, delirium, lethargy, weakness, paresthesias, tremors, premenstrual syndrome, cold extremities, leg or foot cramps, seizures, ventricular arrhythmias and congestive heart failure (Douban et al., 1996). However, since total body stores are not accurately represented by serum levels, routine blood testing and even erythrocyte Mg concentrations may reveal normal levels, particularly in patients with low free (ionized) magnesium levels. Urinary fractional excretion or the oral magnesium load test can estimate the total body magnesium status. Intravenous magnesium loading tests are likely the most accurate and practical assessment, whereby total excretion of urinary magnesium is calculated over a 24 hour period, following administration of a loading dose; a retention of 20% or more indicates deficiency (Arnaud, 2008).

Hypermagnesemia, on the other hand, is a rare condition due to the nephron’s rapid response to increased levels. It usually develops only in people with kidney failure who are given magnesium salts or who take drugs that contain magnesium such as laxatives or antacids. Clinical symptoms include nausea, muscle weakness, lethargy, confusion, hypotension and arrhythmias. In mild cases, withdrawing magnesium supple- mentation is often sufficient. In more severe cases, intravenous calcium gluconate and diuretics or dialysis may be required.

Magnesium Levels in Migraineurs

For many decades, it was postulated that magnesium deficiency played a role in migraine pathogenesis. However, the lack of simple and reliable measures of magnesium levels prevented further research to prove this theory. While low serum, cerebrospinal fluid and cerebral tissue levels of magnesium have been found to be low in patients with migraine (Jain et al., 1985; Ramadan et al., 1989; Schoenen et al., 1991), these results have been inconsistent, with both normal and low levels detected in the same tissues of some patients. The variability of results may be due to the need to measure ionized magnesium as a true reflection of magnesium metabolism, and the development of an ion selective electrode for ionized magnesium in whole blood, serum, plasma and aqueous samples has made an accurate and rapid measurement of ionized magnesium levels possible (Altura et al., 1992).

A study measuring ionized magnesium levels in 40 patients during an acute migraine attack found that 50% had levels below 0.54 mmol/l (normal adult range 0.54-0.65 mmol/l) (Mauskop et al., 1995), with all subjects having total serum magnesium levels within normal limits. Intravenous administration of 1g of magnesium sulfate was most effective in those with low ionized magnesium, with 86% of patients reporting sustained pain relief over 24 hours in those found to have low serum ionized magnesium, while this was the case in only 16% of patients with normal levels. This finding was extended to patients with various headache types, including migraine without aura, cluster head-ache, chronic migraine and chronic tension type headache (Mauskop et al., 1996), with most patients demonstrating low ionized magnesium levels. In addition, high serum ionized calcium to magnesium ratios were found in all headache types except for in those patients with chronic tension-type headaches. Based on these findings, it has been suggested that tension type headache may possibly be discriminated from chronic migraine based on serum ionized magnesium levels (Mauskop et al., 1994).

Migraine Pathogenesis

Migraine is the most common form of disabling primary headache that afflicts patients, affecting approximately 12% of Western populations (Lipton et al., 2007). It is clear that there is genetic transmission of the disorder, although specific genes for most forms of migraine have not yet been identified. Additionally, although the exact etiology remains to be defined, current theories centre on hyperexcitability of the cortex and trigeminovascular complex. In migraineurs, headache triggers stimulate the release of neuropeptides from the trigeminovascular neurons, including calcitonin gene-related peptide and substance P. Additionally, stimulation of the trigeminal ganglion increases cerebral blood flow due to the release of vasoactive intestinal peptide by the facial nerve (Goadsby and Macdonald, 1985). The vasodilation is accompanied by mast cell degranulation, blood vessel edema and increased vascular permeability, resulting in meningeal neurogenic inflammation.

The trigeminal nerve transmits this information to the brainstem trigeminal nucleus caudalis, and then on to the thalamic nuclei and the cortex, where the pain is ultimately perceived (Moskowitz, 1984). Other structures, including the periaqueductal gray matter and the locus coeruleus and the dorsal raphe nuclei, modulate pain transmission and therefore its perception (Martin and Behbehani, 2001).

Recent research has elucidated the aura phase of migraine, which affects up to 5% of the adult population (Agostini and Aliprandi, 2006). Migraine aura is the presentation of charac- teristic neurological symptoms usually developing prior to the onset of the painful phase of a migraine headache, believed to be due to a phenomenon known as cortical spreading depression (CSD). CSD was originally described by Leao (1944) and is an intense depolarization of neuronal and glial membranes, with alterations in membrane resistance and ion flow. There is subsequent massive release of glutamate and potassium as well as an increase in intracellular sodium and calcium. This results in a strong wave of depolarization that spreads across contiguous neuronal tissue. It can be triggered by depolarization of a small region of brain tissue or by direct application of excitatory amino acids, and activation of the N-methyl-D aspartate (NMDA) receptor can evoke CSD (Gorji et al., 2001). There are characteristic alterations in cerebral blood flow, with an initial brief oligemia followed by a profound hyperemia, and a mild, long-lasting oligemia (Otori et al., 2003). Precisely how the aura phase of a migraine evolves into the painful phase remains unknown, and it has been theorized that it is due to the action of a number of inflammatory proteins including calcitonin gene-related peptide (CGRP), nitric oxide and vasoactive peptide (Goadsby et al., 1990), which feed into the trigeminal nerve and generate pial artery dilation and CSD, and, ultimately, headache.

Using phosphorus nuclear magnetic resonance spectroscopy (MRS), low levels of magnesium have been found in the cerebral tissue of some migraineurs both during attacks and interictally (Ramadan et al., 1989). Another study utilized the same technology to assess the brain cytosolic free magnesium concentration and free energy released by the hydrolysis of adenosine triphosphate, an index of cellular bioenergetics in both migraineurs and patients with cluster headaches (Lodi et al., 2001). Cytosolic free magnesium and the free energy released by the reaction of ATP hydrolysis were significantly reduced in the occipital lobes of patients with all types of migraine as well as in cluster headache patients. The authors of this study took these results to lend support to their hypothesis that the reduction of free magnesium in tissue with mitochondrial dysfunction is due to a bioenergetics deficit, as magnesium is essential for mitochondrial membrane stability and the coupling of oxidative phosphorylation. Phosphorus nuclear MRS demonstrated reduced magnesium in the occipital cortices of patients with hemiplegic migraine, with decreases correlating with the severity of neurological complaints (Boska et al., 2002). Interestingly, this study also found increased magnesium in the brains of patients with migraine without aura, which the authors attributed to a possible decrease in intracellular potassium, which may occur in neuronal tissue prone to hyper-excitability.

Various hypotheses abound as to the reason behind the hypomagnesemia observed in migraineurs. Some suggest that during a migraine headache, excessive amounts of magnesium are excreted due to stress. Others propose that stress triggers excretion of magnesium, with secondary hypomagnesemia causing a migraine (Durlach, 1976). During attacks as well as interictally, magnesium levels in both serum and saliva are decreased (Gallai et al., 1992), perhaps an indicator of low cerebral magnesium levels and therefore a decreased threshold for migraine development (Ramadan et al., 1989). Interictal studies on intracellular and serum levels in patients with migraines and tension-type headaches have shown inconsistent results. However a study of migraineurs utilizing the magnesium load test found that after loading with magnesium lactate, there was retention of the magnesium, suggesting a systemic deficiency. Further, interictal levels of erythrocyte magnesium are lower in both adult (Schoenen et al., 1991) and juvenile migraineurs with and without aura (Soriani et al., 1995). In support of these studies, one in which migraineurs with low erythrocyte magnesium levels as well as decreased ionized lymphoycyte magnesium levels were given mineral water containing magnesium over a two-week period. Both erythrocyte and lymphocyte magnesium levels rose (Thomas et al., 2000). These suggest that assays of erythrocyte magnesium may be a useful and easily available tool to assess magnesium deficiency in migraineurs.

Magnesium in Migraine Pathogenesis

Magnesium is believed to be involved in a number of the aspects of migraine patho- physiology, and deficiency has been linked to cortical spreading depression (Mody et al., 1987), platelet aggregation (Baudouin-Legros et al., 1986), release of substance P (Weglicki and Phillips, 1992), neurotransmitter release (Coan and Collingridge, 1985) and vasoconstriction (Altura and Altura, 1982).

NMDA receptors are associated with nociception and the resulting neuroplastic changes in the trigeminal nociceptive neurons, as well as with regulation of cerebral blood flow (Foster and Fagg, 1987). Magnesium ions may block the NMDA receptor, thereby preventing calcium ions from moving intracellularly, and stopping the calcium's effects on neurons and cerebral vasculature (Coan and Collingridge, 1985). Decreased magnesium levels therefore facilitate the NMDA receptor, increasing its effects on CSD, as well as the effect of glutamate on the NMDA receptor. The NMDA receptor has been shown to have a part in both the initiation and spread of cortical depression (Ferrari, 1992). Magnesium has been shown to block CSD induced by glutamate, and CSD is more easily initiated with decreased magnesium levels (Mody et al., 1987).

Nitric oxide (NO) plays a role as a synaptic modulator, affecting nociceptive processing (Meller and Gebhart, 1993) in addition to its involvement in the regulation of blood flow both intracranially and extracranially. It augments the NMDA receptor-evoked currents, thereby facilit- ating glutaminergic transmission (Choi and Lipton, 2000), which, as previously discussed, can be inhibited by magnesium. Production of NO can be inhibited by decreased magnesium levels.

CGRP, a neuropeptide, is released from activated trigeminal sensory nerves, is involved in the dilation of intracranial blood vessels and may also increase nociceptive transmission in the brain- stem and spinal cord. It is therefore believed to play a central part in the development of migraines. A positive correlation has been demonstrated between migraines and serum CGRP levels, and after the pain of a migraine subsides, levels are observed to return to normal (Coderre et al., 1993). It has hence been hypothesized that inhibition of the release of CGRP either centrally or from the trigeminal nerve, may inhibit intracranial vasodilation thereby aborting migraine attacks. CGRP antagonists lack vasoconstrictive properties and have an advantage over current acute migraine treatment, triptans, which are contraindicated in patients with cardiovascular risk factors due to their vasoconstrictive effects.

Circulating CGRP levels have been shown to be decreased by the administration of magnesium in patients with primary Raynaud’s phenomenon (PRP). A study followed CGRP levels in 12 women with PRP and 12 controls before and after the administration of intravenous magnesium sulfate. While there was no significant difference between baseline circulating CGRP in the two groups, following the infusion of magnesium sulfate there was a significant reduction in CGRP levels in the women with PRP only as well as a significant increase in RBC magnesium levels in the women with PRP but not in the control subjects (Myrdal et al., 1994).

Serotonin, released from platelets during migraines, promotes cerebral vasoconstriction as well as triggering nausea and vomiting. Cerebral vascular muscle serotonin receptors may develop increased affinity if serum ionized magnesium falls and the ratio of serum ionized calcium to magnesium increases. This may lead to further cerebral vasoconstriction and facilitates the release of serotonin from neuronal storage sites (Altura and Turlapaty, 1982). Pretreatment with magnesium has been shown to reduce serotonin- induced vasoconstriction (Goldstein and Zsoter, 1978).

Oral Magnesium Supplementation

A double-blind placebo-controlled study was performed in which 24 women with menstrually- related migraine were given supplementation with 360 mg of magnesium pyrrolidone carbox- ylic acid, divided into 3 divided daily doses taken from ovulation to the first day of menstrual flow (Facchinetti et al., 1991). The treatment was well tolerated, with only one patient dropping out of the study due to side effects (magnesium- induced diarrhoea), and significant reductions were observed in number of days of headache, total pain index and in Menstrual Distress Questionnaire score.

A larger double-blind, placebo-controlled, rand- omized study on 81 migraineurs receiving 600 mg of trimagnesium dicitrate taken once daily, showed attack frequency reduction of 41.6% in the magnesium group and 15.8% in controls. Three patients dropped out of the study due to side effects, 18.6% of patients complained of diarrhoea and 4.7% of gastric irritation (Peikert et al., 1996).

A third placebo-controlled double-blind trial showed no effect of magnesium on migraine (Pfaffenrath et al., 1996), possibly attributable to the use of poorly-absorbed magnesium salt, as almost half of the treatment group complained of diarrhoea.

A recent double-blind placebo-controlled, randomized study investigated the effects of 600 mg of magnesium citrate supplementation per day in patients with migraine without aura. A combination of clinical assessments, visual- evoked potentials (VEPs) and brain single-photon emission computerized tomography (SPECT) were performed to assess neurogenic and vascular mechanisms of action (Koseoglu et al., 2008). Supplementation was associated with significant decrease in migraine attack frequency and severity, as well as decreased P1 amplitude on VEPs, and increased cortical blood flow to the inferolateral temporal, inferofrontal and insular regions seen on SPECT. The authors suggested that magnesium may interfere with both neurogenic and vascular mechanisms of migraine and may hence be an effective prophylactic treatment. Other studies using SPECT have shown conflicting results (Ramadan et al., 1991; Ferrari et al., 1995; Olesen et al., 1982).

Women with menstrually-related migraine (MRM) may be particularly prone to developing magnesium deficiency. A prospective study involving 270 women, 61 of whom had MRM, showed that 45% had ionized magnesium deficiency during MRM attacks, 15% during non- menstrual attacks, 14% during menstruation without migraine and 15% between menstruation without migraine (Mauskop et al., 2002). Serum ionized calcium/magnesium levels were elevated in MRM, with normal ionized calcium levels.

Magnesium deficiency has also been found in paediatric migraine, with decreased levels of serum, RBC and mononuclear blood cell magnesium found in paediatric migraineurs with or without aura compared with patients with tension type headache and controls (Soriani et al., 1995; Mazzotta et al., 1999). Magnesium supplementation may be a safe and well- tolerated option for migraine prophylaxis in the prevention of paediatric migraine. However, while a double-blind placebo-controlled, randomized trial of patients aged 3 to 17 years found a statistically significant downward trend in headache frequency in those treated with magnesium oxide, the difference in the slopes of the two lines was not statistically significant (Wang et al., 2003). It could therefore not be determined if oral magnesium was superior to placebo in preventing frequent migraines in children and adolescents.

While oral magnesium is generally well tolerated, the most prominent side-effect is diarrhoea. Magnesium toxicity leads to loss of deep tendon reflexes as well as generalized muscle weakness. Severe toxicity can manifest as cardiac muscle weakness, respiratory paralysis and death, and patients with renal disease are at greater risk for developing toxicity. Although the diarrhoea itself may prevent the development of toxicity, patients should be cautioned regarding excessive intake.

Intravenous Magnesium

Studies examining the use of intravenous magnesium in the treatment of acute migraine have been conflicting. A study on 40 patients with acute migraine attacks showed an 85% correlation between levels of serum ionized magnesium (measured during an attack) and clinical response to 1g of intravenous magnesium sulphate (Mauskop et al., 1995). Although the study was neither double-blinded nor placebo- controlled, both researchers and subjects were blinded to ionized magnesium levels. A further study on various headache types found 1g of magnesium sulphate to provide rapid relief in patients with low serum ionized magnesium levels (Mauskop et al., 1996).

A single-blind placebo-controlled, randomized trial involved 30 patients with migraines who were randomized to receive either magnesium sulphate 1g or placebo (Demirkaya et al., 2001). After 30 minutes, patients in the placebo group who had ongoing pain, nausea or vomiting were given magnesium sulphate 1g. Treatment was superior to placebo in terms of both response rate (100% for magnesium sulphate vs 7% for placebo) and pain-free rate (87% for magnesium sulphate vs 0% for placebo) and those treated did not experience headache recurrence within 24 hours. 87% complained of flushing or a burning sensation in the face and neck.

A double-blind placebo-controlled, randomized study evaluated the efficacy of magnesium sulphate 1g on the pain and associated symptoms of migraine with and without aura (Bigal et al., 2002). In subjects with migraine without aura, although there was a significant decrease in the intensity of photophobia and phonophobia, no significant differences were observed in pain relief or nausea. However subjects with migraine with aura had significant improvement in pain and all associated symptoms.

An emergency room-based double-blind placebo- controlled, randomized study of 44 subjects with acute migraines tested a combination of magnesium sulphate 2g and metoclopramide 20mg versus metoclopramide 20mg alone at 15 minute intervals for up to 3 doses, or until pain relief occurred (Corbo et al., 2001). Pain intensity was recorded using a standard visual analogue scale (VAS). Although both groups experienced more than 50mm improvement in the VAS score, improvement was smaller in the magnesium group, both comparing VAS score improvements and evaluating normal functional status. The authors suggested that magnesium may diminish the efficacy of metoclopramide in decreasing migraine pain.

A study involving 22 patients with cluster headaches who were treated with magnesium sulphate 1g found that 41% of patients reported 'meaningful relief' (defined as complete cessation of attacks or relief for more than 3 days) after treatment (Mauskop et al., 1995).

Conclusion

Magnesium is central to numerous physiological functions, and the role it plays in the various aspects of migraine pathogenesis is well described. Although some studies have shown an association between migraines and magnesium deficiency, it is difficult to assess this with routine blood testing and serum magnesium levels are a poor reflection of body stores of the cation. Therefore, treatment should be based on clinical suspicion, with both oral and intravenous magnesium available as simple, safe, inexpensive and well-tolerated options for the management of migraines. In patients with symptoms suggestive of hypomagnesemia such as pre- menstrual syndrome, cold extremities and foot or leg cramps, we suggest daily magnesium supplementation with 400mg of chelated magnesium, magnesium oxide or slow-release magnesium. While some patients may require doses of up to 1000mg, diarrhoea and abdominal pain may be limiting factors. Intravenous magnesium may be used in patients who are unable to tolerate or absorb oral magnesium or who are non-compliant with daily dosing. It may also be used for the treatment of acute migraines, or as a monthly prophylactic infusion, often administered premenstrually.

References