Home | Back to List of Magnesium Articles

Critical Care Clinics Volume 17 � Number 1 � January 2001 Copyright � 2001 W. B. Saunders Company


Hypomagnesemic Disorders

Michael J. Dacey M.D.

Kent County Hospital, Warwick, Rhode Island

Address reprint requests to Michael J. Dacey, MD Department of Medicine Critical Care Medicine, 4th Floor Offices Kent County Hospital 455 Tollgate Road Warwick, RI 02886 e-mail:

Magnesium is a critical ion that is essential for life. It is intimately involved in over 300 enzymatic reactions. It also has important endocrine functions and is required for protein synthesis.[65] It has been estimated that 65% of critically ill patients develop hypomagnesemia during the course of their ICU stays.[54] Chernow and colleagues[11] found that hypomagnesemia was associated with higher mortality rates in critically ill patients. Many large university hospitals report doing over 80,000 magnesium assays each year.[65] Despite greater appreciation for the importance of magnesium in critical illness, there remains controversy as to the best way in which to deploy this new understanding to maximally benefit critically ill patients. This article discusses the physiology and metabolism of magnesium; the causes of magnesium deficiency in the critically ill patient; techniques for measuring magnesium; the clinical uses of supplemental magnesium in critical illness including asthma, preeclampsia, arrhythmias, acute myocardial infarction, and stroke; and magnesium replacement strategies in the critically ill patient.


The body contains about 0.33 mg/kg (1.32 mmol/kg) of magnesium or about 24 grams total for an average-size adult. A healthy person will ingest 200 to 350 mg daily, with about one half of this amount being absorbed in the jejunum and ileum by both active and passive mechanisms.[75] The kidney filters about 2.5 g of magnesium per day and excretes about 5% of this total. The remainder is reabsorbed by the renal tubules.[30] This resorption occurs along the renal tubules by way of sodium-dependent cotransport. Resorption occurs in the loop of Henle (50%) but also in the proximal convoluted tubule (25%). Only about 5% of magnesium is reabsorbed in the distal convoluted tubule. The level of circulating magnesium is the major factor that influences magnesium reabsorbtion, although a wide variety of drugs and substances, including catecholamines and diuretics, inhibit the reabsorption. Other factors that influence magnesium balance are outlined in Table 1 .

Factor Increasing Mg Reabsorption Factors Decreasing Mg Reabsorption
  Parathyroid hormone Intravascular volume expansion
Hypocalcemia Metabolic acidosis
Vitamin D Hypercalcemia
Metabolic alkalosis Phosphorus depletion
Intravascular volume depletion Diuretics
Magnesium depletion Hypermagnesemia

Magnesium is primarily an intracellular cation; over 99% of nonskeletal stores are found in the intracellular space. Approximately two thirds of magnesium stores are in bone, with the remaining one third residing in cardiac muscle, skeletal muscle, and liver. Interestingly, in contrast to calcium, the maintenance of magnesium homeostasis is highly dependent on dietary intake. There is no known regulatory system that functions to mobilize magnesium in bone or elsewhere to support circulating extracellular levels.[74] A number of factors cause shifts in the usual intracellular:extracellular ratio of magnesium. Both acidosis and ischemia promote release of magnesium from intracellular binding sites and lead to an efflux of magnesium from the cell. Stimulation of both alpha- and beta-receptors causes an efflux of magnesium from the cell; however, enhanced adrenergic tone promotes breakdown of lipids, many of which chelate magnesium, leading to a decrease in active levels. A number of commonly encountered situations in critical care lead to an acute shift of magnesium into cells. These include refeeding syndromes, insulin use, intravenous solutions containing glucose, and amino acid infusions.

Magnesium is intimately involved in maintaining the ionic cellular balance. An accumulating line of evidence strongly suggests that magnesium plays an essential role in the function of the cell membrane sodium-potassium ATPase pump.[22] Magnesium deficiency impairs the action of this pump and leads to reduced intracellular ATP and increased concentrations of sodium within the cell. There are also a number of ion channels that are dependent on magnesium for proper selectivity. These include channels that allow potassium to move into the cell in the presence of adequate intracellular levels of magnesium. If magnesium levels inside the cell are sufficiently low, these channels allow potassium ions to move out of the cell, which may effect conduction and other aspects of cellular metabolism.

Magnesium is a calcium channel blocker. As magnesium deficiency develops, a substantial rise in intracellular calcium occurs. Several investigators have shown that many calcium channels are magnesium dependent, with higher concentrations of magnesium inhibiting flux of calcium through both intracellular-extracellular channels and from the sarcoplasmic reticulum.[69] In smooth muscle, low magnesium concentration enhances the vasoconstrictive effects of catecholamines and angiotensin II.

Magnesium is an obligate ion that is essential for the activation of well over 300 enzymes. Included are those enzymes involved in glucose metabolism, fatty acid synthesis and breakdown, and DNA and protein metabolism.[74] In addition, magnesium is required for the activity of adenylate cyclase, which transmits extracellular hormonal signals to the intracellular apparatus. Adenylate cyclase is activated by G protein located in the cell membrane. The alpha subunit of G protein is itself activated by a magnesium-dependent guanine phosphorylase reaction.[55] Magnesium is essential for virtually all hormonal reactions that occur in the body.


The most common causes of hypomagnesemia are listed in Table 2 . Critically ill patients who develop hypomagnesemia fall into three broad categories: those with decreased intake, those with altered intracellular-extracellular distribution, and those with increased losses. Increased losses may occur from either the kidney or gastrointestinal tract.

Increased Loses
         Aminoglycosides [74]
         Albuterol and other beta agonists
         Loop diuretics
         Thiazide diuretics
         Osmotic agents
      Any acute renal tubular injury
      Nasogastric suction
      Short bowel syndromes
      Malabsorption syndromes
Acute Intracellular Shift
   Refeeding syndromes
   Glucose infusions
   Amino acid infusions
   Metabolic acidosis of any type
   Total parenteral nutrition
   Phosphorus depletion
   Alcohol abuse
   Hungry bone syndrome
   Citrated blood products
   Cardiopulmonary bypass[6]
   Intravascular volume expansion
   Hypoalbuminemia (although ionized levels may be normal)

Decreased Intake

Malnutrition is common in patients at initial presentation to the ICU and those who have been under the care of intensivists for some time. Several studies have demonstrated marked decreases in muscle stores of magnesium in these patients.[10] A number of factors influence the serum magnesium level. Dietary intake of magnesium is a critical determinant of magnesium levels. Because chlorophyll is the magnesium chelate of porphyrin, green leafy vegetables are excellent sources of magnesium. Unpolished grains and nuts also have high levels of magnesium, whereas meats, starches, and milks have somewhat less. Water can be a source of magnesium intake, but the amount varies depending on the hardness of the water in a given area. Overall intake of magnesium seems to be declining in the US population because of increased use of processed foods. The factors that regulate magnesium absorption in the gastrointestinal tract are poorly understood; to date there has not been a regulatory factor discovered that is akin to vitamin D for calcium absorption, although in fact, vitamin D and its metabolites enhance magnesium absorption by the distal small bowel to a small extent. Patients with alcoholism have extremely poor magnesium intake and are prone to total body magnesium depletion. The use of total parenteral nutrition (TPN) is often associated with hypomagnesemia. The traditional practice of adding 0.20 mmol/kg/day of magnesium to TPN solutions is simply not adequate for most critically ill patients.[18] The highly concentrated glucose and amino acid infusions drive magnesium into cells at the same time that concentrated intralipid solutions are chelating free magnesium in serum.

Increased Losses


Substantial amounts of magnesium can be lost from the gastrointestinal tract. Diarrhea, regardless of the cause, is one of the most common reasons for gastrointestinal magnesium losses in ICU patients. Nasogastric suctioning also removes significant amounts of magnesium from the body over several days' time. Various malabsorption syndromes and short bowel syndromes that occur after surgery can produce high losses of magnesium.[66] Pancreatitis can lead to hypomagnesemia because of sequestration of magnesium-rich fluid within the pancreas combined with losses through nasogastric suctioning and diarrhea.


Renal magnesium wasting occurs when urinary magnesium losses exceed 12 mg (0.5 mmol) per day in the presence of ionized hypomagnesemia.[74] Patients at risk for renal magnesium wasting include those with diabetes, alcoholism, hyperthyroidism, hypercalcemia, and hypophosphatemia.[55] In diabetics there is a strong relationship between insulin resistance and hypomagnesemia.[3] Glucosuria probably contributes significantly to renal magnesium wasting in these patients. Any acute renal injury, particularly renal tubular injury or disorders, may promote wasting of magnesium. Many medications promote the excretion of magnesium in the urine (see Table 2) . Most drugs induce magnesium wasting by inhibiting tubular reabsorption of magnesium. Chief among these are the diuretics, especially loop diuretics.[45] Although drugs such as amphotericin-B and platinum-based chemotherapy agents have well deserved reputations for inducing severe hypokalemia and hypomagnesemia, others like aminoglycosides are often underappreciated. One study found that almost 40% of patients receiving aminoglycosides developed hypomagnesemia.[73] Although alcoholics are at great risk for magnesium depletion primarily because of malnutrition, alcohol itself does promote renal magnesium wasting by an effect on renal tubular magnesium reabsorption.

Altered Intracellular-Extracellular Distribution

Acute intracellular shift of magnesium will occur in patients with metabolic acidosis; those with elevated levels of circulating catecholamines; those given exogenous glucose, insulin, or amino acid solutions; and those with refeeding syndromes. There also is evidence that the hypomagnesemia that occurs after cardiac bypass in a large majority of patients is caused by an acute intracellular shift of magnesium.[6] [27] Large-volume resuscitations with hypotonic fluids not containing electrolytes will promote hypomagnesemia as will citrate present in blood products (by chelation of magnesium).


Magnesium exists in four forms in the body.[21] Intracellular stores make up 99% of total body magnesium content and are found mostly in bone and muscle. Extracellular magnesium exists in three forms: protein-bound (30%), chelated to various anions (15%), and ionized or active (55%). A decrease in total magnesium levels is common in patients who are critically ill and have low albumin levels because of a decrease in the protein-bound fraction, much akin to the situation with calcium measurements. Both calcium and magnesium have significant protein-bound fractions, thus creating the potential for a large difference between total serum and ionized (active) levels. In contrast, the monovalent cations sodium and potassium exist in blood 99% in the ionized state. Unfortunately, there is no correction factor that can be used to correct for low albumin in estimating ionized magnesium levels. Furthermore, the measurement of nonionized magnesium is subject to interference from a number of sources. Both elevated bilirubin levels and red blood cell hemolysis will cause absorption of light similar to that used in spectrophotometric assays of magnesium, leading to an overestimation of magnesium concentrations. Additionally, acute acidemia will cause an acute release of magnesium from the intracellular compartment. Many critically ill patients present with acute acidosis, so total magnesium levels may be misleadingly high on admission. Once the acidosis is corrected, levels may drop precipitously.[20]

For many years there were no commercially available, ion selective electrodes for ionized magnesium determinations. This lack led some to the practice of measuring magnesium in serum ultrafiltrates;[6] [75] however, these methods required large blood samples, were fairly time consuming, and did not distinguish between ionized magnesium and magnesium bound to organic and inorganic ions and fatty acids.[21] The development of ion selective electrodes for magnesium was a major advance. An ion-selective electrode works on the principle that the electrical potential difference between two solutions is dependent on the different concentrations of a given ion in two solutions with the patients' sample being compared to a reference solution. Reference ranges for ionized magnesium are listed in Table 3 . The first-generation electrodes suffered from interference by calcium ions, which created a requirement for a calcium ion selective electrode to chemometrically correct for any calcium ion interference. Newer ionized magnesium analyzers are much more specific for magnesium.[65]

Total Ionized
Adults 1.7-2.40 mg/dL 0.90-1.30 mg/dL
1.4-2.00 mEq/L 0.80-1.10 mEq/L
0.7-1.00 mmol/L 0.44-0.59 mmol/L
*Data from Greenway D, Hindmarsh J, Wang J, et al: Reference interval for whole block ionized magnesium in a healthy population and the stability of ionized magnesium under varied laboratory conditions. Clin Biochem 29:515-520, 1996.

One study assessed the use of ionized magnesium concentrations in critically ill children.[21] As in the adult population, almost 60% of patients had ionized levels less than 0.4 mmol/L. Interestingly, of the children with ionized hypomagnesemia, 60% had normal total magnesium determinations. The authors found no correlation between ionized magnesium concentrations and ionized calcium, pH, albumin, potassium, or serum creatinine. Furthermore, the usual replacement doses of magnesium did correct total magnesium levels to the normal range but not ionized levels. Other studies have found no difference in ionized magnesium measured in whole blood, plasma, or serum.[2] [13]

The use of ionized magnesium levels is not universally accepted.[23] Despite the poor correlation between total and ionized magnesium, some have questioned the role of ionized levels.[23] They point out that other studies have shown that ionized hypomagnesemia may not correlate with total body magnesium deficiency based on results of magnesium loading tests.[32] In this test, a 24-hour urine magnesium collection is performed followed by a magnesium infusion, usually 30 mmol of MgCl2 over the next 24 hours. A second 24-hour urine collection is done concomitantly with the magnesium infusion. Excretion of less than 50% of the administered dose of magnesium suggests depletion from nonrenal causes.[56] If ionized hypomagnesemia does not correlate well with total body stores, then what is its clinical significance? Indeed, is ionized hypomagnesemia a normal response to illness or is it a predictor of adverse effects (e.g., cardiac arrhythmias, organ injury)?[23] These questions remain to be answered. Ionized magnesium testing is gaining wider acceptance in clinical practice. Measurement of intracellular magnesium has been done in research settings; however, there are no commercially available tests for clinical use.


Magnesium deficiency is associated with diverse effects as listed.

Clinical Effects of Magnesium Deficiency

Cardiovascular Dysrhythmias [4]

Ventricular tachycardia (especially torsades de pointes)
Atrial fibrillation
Supraventricular tachycardia
Electrocardiogram changes[23]
Prolonged QT interval
Prolonged PR interval
Wide QRS
Peaked T waves
ST depression
Muscle fasciculations
Mental status changes
Endocrine [12]
Stimulated parathyroid hormone (PTH) secretion (in mild cases)
Suppressed PTH secretion (in severe cases)
Insulin resistance[19]
Muscle weakness
Respiratory failure
Increased levels of proinflammatory cytokines[67]
Augmentation of cyclosporine-induced nephrotoxicity[40]
Augmentation of gentamycin-induced ototoxicity[28]
Serious neurologic effects are rare until total serum levels drop below 1.0 to 1.2 mg/dL. [74] Arrhythmias, both ventricular and supraventricular, occur more frequently in patients with hypomagnesemia.[4] Augmentation of both gentamycin- and cyclosporine-induced toxicities have been described.[28] [40] A host of metabolic changes are associated with magnesium deficiency and include insulin resistance.[48] Hypocalcemia, hypokalemia, and hypophosphatemia are commonly found in association with hypomagnesemia. Mild ionized hypomagnesemia stimulates PTH secretion. Severe ionized hypomagnesemia will suppress PTH secretion as will hypermagnesemia.[39] Furthermore, normal magnesium levels are required for the action of PTH.[38] Therefore, large swings in circulating magnesium levels are common causes of ionized hypocalcemia.[74] Muscle weakness is common and may lead to respiratory failure.



Magnesium as a treatment for asthma is not a new idea, but the more widespread use of intravenous magnesium during asthma exacerbations and the results of several clinical studies have engendered some debate regarding the merits of the therapy.

The relaxant effects of magnesium on smooth muscle were shown in a recent animal trial.[34] Hirota and colleagues[34] showed that in isolated animal preparations of trachea exposed to histamine-containing solutions, exposure to either magnesium sulfate or zinc sulfate caused a dose-dependent relaxation of smooth muscle, whereas exposure to sodium sulfate had no effect.

Clinical studies of magnesium in asthma have produced conflicting results.[31] [32] [33] [34] [35] Studies by Okayama and colleagues,[51] Noppen and colleagues,[50] and Skobeloff and colleagues [62] have all shown beneficial results. In the study by Okayama and colleagues,[51] giving 1.2 g of magnesium sulfate over 20 minutes in patients with mild asthma exacerbations produced a significant increase in forced expiratory volume-one second (FEV1 ) measurements before any other treatments were given. In contrast, Skobeloff and colleagues[62] gave 1.2 g of magnesium sulfate to patients with severe asthma attacks in whom previous traditional treatments had failed and found a significant decrease in the need for hospital admissions in those given magnesium compared with controls. These results were not duplicated in other studies in patients with severe refractory asthma.[25] [64] Likewise, Bloch and colleagues'[7] study of 135 patients did not show an improvement in flow rates or hospital admission rates. Subgroup analysis did show statistically significant increases in flow rates and decreases in hospital admission rates in those patients classified as having severe asthma (FEV1 < 25% predicted). Tiffany and coworkers[64] found similar results in their randomized controlled clinical trial of 48 patients who failed to improve after conventional therapy. There are case reports of rapid infusion of 2 g of magnesium sulfate over 2 minutes preventing the need for intubation with dramatic reversal of bronchospasm.[56] There is no study that defines a therapeutic magnesium level for the treatment of acute asthma exacerbations. There are studies suggesting that intracellular magnesium depletion may be associated with asthma and bronchial hyperactivity but not necessarily to acute flares of asthma.[16] [29]

Mangat and colleagues[44] investigated the role of inhaled magnesium sulfate in acute asthma attacks. They compared nebulized salbutamol (2.5 mg) with 3 cc of a 3.2% solution (95 mg) of magnesium sulfate in a randomized, double-blind trial involving 33 patients. They found that the increases in peak flow rates were comparable in both groups of patients with moderate to severe asthma exacerbations. Nannini and colleagues[49] studied the effects of 286 mOsm of inhaled magnesium sulfate on bronchoconstriction induced by metabisulfite in ten asthmatic subjects.[40] They found a significant attenuation of the bronchoconstriction effect of metabisulfite when magnesium was inhaled 5 minutes before the challenge. Sharma and colleagues[59] studied the airway resistance of 18 patients with stable asthma after the intravenous infusion of 10 mmol of elemental magnesium. They found a significant decrease in airway resistance and an improvement in specific conductance with only a small improvement in maximal expiratory flow rates. They concluded that magnesium had its major bronchodilatory effect on large airways in patients with asthma. Hill and colleagues[39] found that although magnesium was a weak bronchodilator when given in doses of 2 g over 20 minutes, it offered no protection against histamine-induced changes in airway resistance in subjects with stable asthma.

Skorodin and colleagues[63] studied the effect of giving 1.2 g of magnesium sulfate over 20 minutes by nebulizer to 72 patients with exacerbations of chronic obstructive pulmonary disease after they had already received albuterol. They found the difference in peak expiratory flow from initiation of the infusion to 30 and 45 minutes later was significantly different in the group given magnesium as opposed to controls (25 L/min versus 7.4 L/min; P=0.03). There was also a trend toward a reduced need for hospitalization in the magnesium group (28% versus 41%; P=0.25).

In summary, the medical literature presents moderately convincing evidence that magnesium is a bronchodilator; however, determining the clinical utility of this effect will require the results of larger, randomized, placebo-controlled clinical trials. For the present, there seems to be adequate data regarding the safety and efficacy of magnesium to recommend its use in those patients with asthma who have severe disease that is refractory to traditional therapy.


Preeclampsia is defined as a blood pressure greater than 140/90 accompanied by proteinuria or edema during the second half of pregnancy. If seizures then occur, the diagnosis of eclampsia is made. Preeclampsia complicates approximately 7% of pregnancies. The vast majority of both eclamptic and preeclamptic patients are treated with magnesium, and decades of experience support its use. The mechanism by which magnesium suppresses seizures is unknown but may be multifactorial, involving effects on neuronal membrane stabilization, improved cerebral blood flow, diminshed tendency toward cerebral artery vasospasm, and antiplatelet effects. Evidence-based medicine favors the prophylactic use of magnesium in preeclamptic patients. Lucas[43] studied more than 2100 patients with preeclampsia, comparing treatment with phenytoin versus magnesium. There were 10 eclamptic seizures in the phenytoin group versus none in the magnesium-treated patients. Eclampsia also is commonly treated with magnesium, although some authors recommend using phenytoin, especially if magnesium is already being infused. Prichard and colleagues'[52] series found magnesium therapy to be very efficacious in the reduction of recurrent seizures and mortality rates. The dose of magnesium most commonly recommended is 4 g given intravenously over the first 15 minutes then 2 to 3 g/h intravenously.[61] The target serum total magnesium level is 5 to 7 mg/dL. If seizures persist, some would push the magnesium dose and levels higher, although many would suggest adding another medication, usually phenytoin.[24] The managing physician must be vigilant for clinical signs of magnesium toxicity. In general, this therapy is well tolerated until total magnesium levels rise above 7 mg/dL. Patients with clinical magnesium toxicity develop a loss of deep tendon reflexes and progressive weakness, including the muscles of respiration leading to acute respiratory failure. Hypotension, complete heart block, and even cardiac arrest have all been described. Hypermagnesemia also increases sensitivity to muscle relaxant drugs. Opponents of magnesium therapy in eclampsia point to the lack of randomized, controlled clinical trials and the tocolytic effects of the ion. There are case series suggesting higher rates of cesarean section in those patients treated with magnesium for eclampsia. There is, however, a growing interest in the possible beneficial effects that magnesium may have in protecting the brains of premature infants from neuronal damage.[9] [35] It is certainly true that the ultimate treatment of eclampsia is delivery of the fetus. Decades of experience with magnesium have made it the drug of choice in the minds of most obstetricians.


The American Heart Association currently recommends magnesium sulfate as first-line treatment for torsade de pointes. Advanced cardiac life support (ACLS) guidelines suggest giving consideration to the use of magnesium for monomorphic ventricular tachycardia refractory to more traditional therapy. In both conditions, a 2 g bolus of magnesium sulfate is used and may be repeated in 5 minutes if needed.

There are several lines of evidence for a role for magnesium in the pathogenesis and treatment of supraventricular arrhythmias. Hypomagnesemia promotes digoxin-induced arrhythmias. Supplemental magnesium has been associated with a greater chance of conversion of atrial fibrillation to sinus rhythm.[8] DeCarli and colleagues[14] found that in subjects with hypomagnesemia, twice the dose of digoxin was required to control the ventricular response of patients in atrial fibrillation. Multifocal atrial tachycardia also has been successfully treated with supplemental magnesium therapy.[36]

Moran and coworkers[46] compared parenteral magnesium sulfate versus amiodarone in the therapy of atrial tachyarrhythmias in a prospective, randomized fashion. They found that in doses of 37-mg/kg bolus followed by 25 mg/kg/h, titrated to levels of 1.5 to 2.0 mmol/L, magnesium was superior to amiodarone in the conversion of acute atrial tachyarrhythmias. The study group went to great lengths to ensure normal potassium levels over the 24-hour study period in this group of mixed medical and surgical patients. Both therapies were very well tolerated. The probability of conversion to sinus rhythm was 78% for magnesium-treated patients versus 50% for those treated with amiodarone. Interestingly, these authors did not find a relationship between rhythm conversion and plasma magnesium concentration. Therapy with magnesium cost the hospital $31.00 for the first 24 hours versus $67.00 for amiodarone. The authors postulated that the antiarrhythmic action of magnesium may be mediated through blocking of slow calcium channels, activation of membrane sodium-potassium ATPase, or effects on sympathetic nervous system activity.

Studies of the prophylactic actions of magnesium in preventing supraventricular arrhythmias have been largely limited to the surgical literature. Fanning and colleagues[18] studied the effects of giving 48 mmol of magnesium sulfate by continuous infusion during the first postoperative day after cardiac surgery. They found fewer episodes of atrial fibrillation in the group given prophylactic magnesium compared with controls. England and coworkers[17] could not demonstrate a benefit to giving 2 g of magnesium chloride at termination of cardiac bypass. Overall, magnesium is a cheap and efficacious therapy for treating both supraventricular and ventricular arrhythmias. The authors recommend it as primary therapy for non-life-threatening cardiac arrhythmias. If patients fail to respond, consideration can be given to administration of an antiarrhythmic drug.


Magnesium has undergone extensive study with respect to its benefit in acute myocardial infarction (MI). It is postulated that supplemental magnesium therapy may benefit those with acute myocardial ischemia in several ways. First, magnesium may limit myocardial damage, perhaps by inhibiting calcium influx into ischemic myocardial cells.[19] Magnesium may reduce coronary artery tone, thus improving distal blood flow to ischemic myocardium.[1] Magnesium increases the threshold for depolarization of cardiac myocytes and may have antiarrhythmic effects beneficial to these patients who are at greatly increased risk for life-threatening arrhythmias.[31] Magnesium infusion reduces peripheral vascular resistance and subsequently may increase cardiac output without increasing cardiac work.[60] Finally, there is evidence that magnesium may inhibit the platelet aggregation so central in the pathogenesis of acute MI.[53] [60] Early studies of magnesium used in the setting of acute MI were very encouraging, suggesting mortality reductions as high as 50% in eight small randomized trials involving almost 1000 patients.

The first large-scale, randomized, controlled clinical trial that sought to assess the effect of magnesium administration in acute MI was the second Leicester intravenous magnesium intervention trial or LIMIT-2 involving over 2300 patients with suspected acute MI.[71] In this study, the treatment group received 8 mmol of magnesium sulfate over 5 minutes followed by 65 mmol over the next 24 hours. The average serum level achieved was 1.55 mmol/L. Importantly, this protocol called for the initiation of magnesium therapy at the start of reperfusion therapy. The authors found a 24% reduction in 28-day mortality (P=0.04; 95% confidence interval [CI] 1-43%). The rate of left ventricular failure was reduced by 25% (P=0.009; 95% CI=7-39%). They did not discover any significant difference in the incidence of heart block or serious arrhythmias. Interestingly, the authors could not identify a mechanism to explain the beneficial effects of magnesium in their study. They found no differences in the rates of arrhythmias in treated versus untreated patients. The use or omission of aspirin did not influence the effect that magnesium had on outcome, arguing against a role for platelet inhibition. Neither did previous diuretic use seem to affect outcome that indicated against a simple replacement of total body deficits. The hemodynamic effects of magnesium on afterload reduction may have been beneficial, however, beyond the immediate period of the bolus infusion, the authors did not find a sustained effect. Finally, it should be noted that the calculated 95% confidence intervals were quite broad.

The results of the ISIS-4 trial stand in stark contrast to those of LIMIT-2.[37] The ISIS-4 investigators studied over 58,000 patients with suspected acute MI, assessing the effects on 5-week mortality of angiotensin converting enzyme (ACE) inhibition, nitrate therapy, and a 24-hour magnesium protocol involving an initial 8-mmol bolus followed by 72 mmol over the next 24 hours. The major difference in the protocol compared to LIMIT-2 was that magnesium was started after reperfusion therapy, not concurrently as in LIMIT-2. Patients in cardiogenic shock were excluded.

The ISIS-4 investigators found no differences in rates of arrhythmias of any type including ventricular fibrillation. Likewise, they found no differences in hospital length of stay or mortality in patients given magnesium versus controls. They carried out very careful subgroup analysis and again found no differences in any parameter between treatment and control groups. The analysis included those 17,000 patients who did not receive reperfusion with thrombolytic therapy. The only marginally statistically significant effect was a slight excess of deaths in those patients presenting with bradycardia or low systolic blood pressure who received magnesium. Magnesium also was associated with small but significant increases in rates of heart failure (12 per 1000 treated; P < 0.001), cardiogenic shock (5 per 1000 treated; P < 0.01), and deaths attributed to cardiogenic shock (1.62% versus 1.26%; P < 0.001). The incidence of sinus bradycardia, but not heart block, was significantly increased in the magnesium-treated patients (P < 0.0001). The authors then carried out a pooled analysis of all patients studied in all trials of magnesium in acute myocardial ischemia. They found a mortality rate of 7.59% for those receiving magnesium versus 7.46% for controls.

The authors of ISIS-4[37] correctly point out that the low mortality of all patients in their study speaks to the excellent treatment that patients already have for acute MI. Thrombolysis carried out within 12 hours of symptom onset prevents 20 to 30 deaths per 1000 patients treated. It seems reasonable to assume that acute angioplasty carries at least this degree of benefit. One month of aspirin therapy prevents 25 early deaths per 1000 treated patients and 10 to 15 nonfatal reinfarctions or strokes per 1000 treated patients. Angiotensin converting enzyme inhibition started early after an acute MI and continued for 1 month will save 5 lives per 1000 treated.[37] [38] Despite the already excellent therapy available, the ISIS-4 trial was designed with sufficient power to detect a beneficial effect of magnesium and none was found. The accumulated data from evidence-based medicine does not support a role for magnesium in the treatment of acute MI. Despite this result, there are still those who raise questions regarding the timing of therapy in the ISIS-4 trial. Would the results of ISIS-4 have been more similar to that of LIMIT-2 had the ISIS investigators given the magnesium at the start of reperfusion therapy instead of waiting until thrombolytic therapy was concluded? This question currently remains unresolved, although at least one trial has suggested that in patients with acute MI who are not candidates for thrombolytic therapy, magnesium given early reduces mortality from 17% to 4%.[60] This result is interesting, especially given that almost twice as many patients in ISIS-4 received thrombolytic therapy (70%) compared with LIMIT-2 (36%). A cell protective effect of magnesium in acute MI has not been excluded by the previous studies and may yet provide a niche for magnesium in this setting.

Acute Cerebral Ischemia

The past several years have seen a substantial increase in research focusing on cell protection therapy for acute cerebral injury. Magnesium has been reported to increase regional cerebral blood flow by vasodilatation of cerebral arteries. In addition, reduction of extracellular magnesium is directly correlated with the intensity of cerebral vasospasm in experimental animals. Direct neuronal effects include blockade of the NMDA receptor ion channel, calcium antagonism at voltage gated channels, enhanced buffering of intracellular calcium ions, and enhanced regeneration of ATP.[47]

Initial interest in the use of magnesium was generated by several epidemiologic studies suggesting a decrease in stroke rates and death from stroke among those people with diets rich in magnesium.[5] [38] [72] A number of animal studies then followed showing that supplemental magnesium reduced the size of experimentally induced cerebral infarctions.[42] [57] [58]

Lee and coworkers[42] studied the effect of magnesium administration combined with mexiletine, a potent sodium channel blocker that decreases neuronal energy demands and prevents energy depletion during ischemia in rats following middle cerebral artery occlusion. Both treatments significantly reduced infarct size when given separately, but coadministration carried no additional benefit. Schmid-Elsaesser and colleagues[58] studied the effects of magnesium administration and tirilazad, an antioxidant, on experimentally induced cerebral ischemia in rats. They found that magnesium reduced infarct volume by 25% (P not significant). The combination of magnesium and tirilazad reduced infarct volume by 59% (P < 0.05). The same group combined these same agents with hypothermia and almost completely abolished (minus 99%) cortical infarction.[57]

Studies in humans are in the early stages. One group[41] demonstrated a significant correlation between cerebrospinal fluid magnesium concentrations and infarct size in humans (P < 0.0001). They also discovered a significant correlation between low cerebrospinal fluid magnesium and the intensity of the residual neurologic deficit.[41] A number of initial studies assessing safety and dosing have been done showing that magnesium is safe and well tolerated in acute stroke. Several small case series have shown promising trends favoring magnesium administration although definitive recommendations await the results of large multicenter trials already in progress and expected in about 2 years.


It is imperative to identify an underlying cause for the magnesium deficiency. If possible, all drugs that contribute to magnesium wasting should be discontinued. Severe hypomagnesemia with associated signs and symptoms is treated with aggressive intravenous replacement as outlined.[74]

Magnesium Replacement Recommendations

  1. Discontinue all magnesium wasting medications if possible
  2. 2 g (16 mEq) MgSO4 over 10 minutes intravenously
  3. Begin continuous infusion of MgSO4 at 0.5 g/h intravenously. As an alternative, give 1-2 g intravenously every 4 hours
  4. If renal insufficiency present, start continuous infusion at 0.25 g/hr
  5. Check ionized magnesium levels every 6 hours and adjust as needed for high normal circulating levels as a goal. If ionized magnesium measurement is not available, follow total serum magnesium levels and clinical examination. Keep total levels less than 3-4 mg/dL
  6. Monitor potassium and calcium levels frequently throughout replacement
  7. The continuous infusion should run for approximately 72 hours, after which the patient may be changed to maintenance doses as follows (for those with normal renal function):
    Magnesium oxide at 0.4 mEq/kg/day enterally
    MgSO4 at 0.1-0.2mEq/kg/day intravenously
  8. In those without adequate intravenous access, MgSO4 can be given intramuscularly in doses of 4 g every 4-6 hours


Note: MgSO4 : 1 gram = 98 mg elemental Mg = 4 mmol = 8 mEq

MgCl2 : 1 gram = 118 mg elemental Mg = 4.5 mmol = 9 mEq

Magnesium gluconate tablets = 500 mg = 32 mg elemental Mg = 1.2 mmol = 2.4 mEq

Magnesium oxide tablets = 111 mg elemental magnesium = 4.5 mmol = 9 mEq

Bolus doses of magnesium are rapidly excreted by the kidneys, making smaller dose continuous infusions a better choice in most situations. An initial dose of 2 g (16 mEq) is given over 10 minutes followed by a continuous infusion for the next 72 hours to allow for replenishment of intracellular stores. Once circulating levels have been maintained in the normal range, patients can be converted to maintenance dosing by the enteral route. Magnesium oxide is the preferred enteral agent because of its higher bioavailability.


1. Altura B, Altura B: Endothelium dependent relaxation in coronary arteries requires magnesium ions. Br J Pharmacol 91:449-451, 1987  

2. Altura B, Shirey T, Young C, et al: A new method for the rapid determination of ionized magnesium in whole blood, serum and plasma. Methods Find Exp Clin Pharmacol 14:297-304, 1992  

3. American Diabetes Association: Magnesium supplementation in the treatment of diabetes. Diabetes Care 15:1065-1067, 1992  

4. Arsenian MA: Magnesium and cardiovascular disease. Prog Cardiovasc Dis 35:271-310, 1993  

5. Ascherio A, Rimm E, Hernan M, et al: Intake of potassium, magnesium, calcium, and fiber and risk of stroke among US men. Circulation 98:1198-1204, 1998  

6. Aziz S, Haigh W, Van Norman G, et al: Blood ionized magnesium concentrations during cardiopulmonary bypass and their correlation with other circulating cations. J Card Surg 11:341-347, 1996  

7. Bloch H, Silverman R, Mancherje N, et al: Intravenous magnesium sulfate as an adjunct in the treatment of acute asthma. Chest 107:1576-1581, 1994  

8. Brodsky M, Orlcu M, Capparelli E, et al: Magnesium therapy in new onset atrial fibrillation. Am J Cardiol 73:1227-1229, 1994  

9. Caddell J, Graziani L, Wiswell T, et al: The possible role of magnesium in protection of premature infants from neurological syndromes and visual impairments and a review of survival of magnesium exposed premature infants. Magnes Res 12:201-216, 1999  

10. Caddell J, Suskind R, Sillup H, et al: Parenteral magnesium load evaluation of malnourished Thai children. J Pediatr 83:129-135, 1973  

11. Chernow B, Bamberger S, Stroiko M, et al: Hypomagnesemia in patients in the postoperative intensive care unit. Chest 95:391-397, 1989  

12. Cholst I, Steinberg S, Tropper P, et al: The influence of hypermagnesemia on serum calcium and parathyroid hormone levels in human subjects. N Engl J Med 310:1221-1225, 1984

13. Cook L, Mimouni F: Whole blood ionized magnesium in the healthy neonate. J Am Coll Nutr 16:181-183, 1997  

14. DeCarli C, Sprouse G, Larosa J: Serum magnesium levels in symptomatic atrial fibrillation and their relation to rhythm control by intravenous digoxin. Am J Cardiol 57:956-959, 1986  

15. Dickerson R, Brown R: Hypomagnesemia in hospitalized patients receiving nutritional support. Heart Lung 14:561-569, 1985  

16. Dominguez L, Barbagallo M, DiLorenzo G, et al: Bronchial reactivity and intracellular magnesium. Clin Sci (colch) 95:137-142, 1998  

17. England M, Gordon G, Salem M, et al: Magnesium administration and dysrhythmias after cardiac surgery. A placebo controlled double blind randomized trial. JAMA 268:2395-2402, 1992  

18. Fanning W, Thomas C, Roach A, et al: Prophylaxis of atrial fibrillation with magnesium sulfate after coronary artery bypass grafting. Ann Thorac Surg 52:529-533, 1991  

19. Ferrari R, Albertini A, Curello S, et al: Myocardial recovery during post-ischemic reperfusion: Effects of nifidepine, calcium and magnesium. J Mol Cell Cardiol 18:487-498, 1986  

20. Fiaccordori E, Del Canale S, Coffrini E, et al: Muscle and serum magnesium in pulmonary intensive care unit patients. Crit Care Med 16:751-760, 1988  

21. Fiser R, Torres A, Butch A, et al: Ionized magnesium concentrations in critically ill children. Crit Care Med 26:2048-2052, 1998  

22. Flatman P, Lew V: The magnesium dependence of sodium pump mediated sodium-potassium and sodium-sodium exchange in intact human red cells. J Physiol 315:421-446, 1981  

23. Foley C, Zaritsky A: Should we measure ionized magnesium? Crit Care Med 26:1949-1950, 1998  

24. Frakes M, Richardson L: Magnesium sulfate in certain emergency conditions. Am J Emerg Med 15:182-187, 1997  

25. Green S, Rothrock S: Intravenous magnesium for acute asthma: Failure to decrease emergency treatment duration or need for hospitalization. Ann Emerg Med 21:260-265, 1992  

26. Greenway D, Hindmarsh J, Wang J, et al: Reference interval for whole blood ionized magnesium in a healthy population and the stability of ionized magnesium under varied laboratory conditions. Clin Biochem 29:515-520, 1996  

27. Gries A, Bode C, Gross S: The effect of intravenously administrated magnesium on platelet function in patients after cardiac surgery. Anesth Analg 88:1213-1219, 1999

28. Gunther T, Rebentisch E, Ising H, et al: Enhanced ototoxicity of gentamycin in maternal and fetal rats due to magnesium deficiency. Magnes Bull 11:19-21, 1989  

29. Gustafson T, Boman K, Rosenhall L, et al: Skeletal muscle magnesium and potassium in asthmatics treated with oral beta 2 agonists. Eur Resp J 9:237-240, 1996

30. Hamil-Ruth R, McGory R: Magnesium repletion and its effect on potassium homeostasis in critically ill adults: Results of a double blind, randomized, controlled trial. Crit Care Med 24:38-45, 1996  

31. Haverkamp W, Hindricks G, Keteller T, et al: Prophylactic antiarrhythmic and antifibrillatory effects of intravenous magnesium sulphate during acute myocardial ischaemia. Eur Heart J 9:228, 1988  

32. Herbert P, Mehta N, Wang J, et al: Functional magnesium deficiency in critically ill patients identified using a magnesium loading test. Crit Care Med 25:749-755, 1997  

33. Hill J, Britton J: Effect of intravenous magnesium sulfate on airway calibre and airway reactivity to histamine in asthmatic subjects. Br J Clin Pharmacol 42:629-631, 1996  

34. Hirota K, Sata T, Hashimoto Y, et al: Relaxant effect of magnesium and zinc on histamine induced bronchoconstriction in dogs. Crit Care Med 27:1159-1163, 1999

  35. Hirtz D, Nelson K: Magnesium sulfate and cerebral palsy in premature infants. Curr Opin Pediatr 10:131-137, 1998  

36. Iseri L, Fairshter R, Hardemann J: Mg and potassium therapy in multifocal atrial tachycardia. Am Heart J 110:789-794, 1985  

37. Isis-4 Investigators: Isis-4: A randomised factorial trial assessing early oral captopril, oral mononitrate, and intravenous magnesium sulphate in 58,050 patients with suspected acute myocardial infarction. Lancet 345:669-685, 1995  

38. Iso H, Stampfer M, Manson J, et al: Prospective study of calcium, potassium, and magnesium intake and risk of stroke in women. Stroke 30:1772-1779, 1999  

39. Johannsson A, Raisz L: Effects of low medium magnesium concentration on bone resorption in response to parathyroid hormone and 1,25 dihydroxy vitamin D in organ culture. Endocrinology 113:2294-2298, 1983  

40. Kaiser W, Biesenhack G, Kramer E, et al: Hypomagnesemia: A feature of cyclosporin tubulotoxicity in renal transplant recipients. Magnes Bull 14:130-132, 1992

41. Lampl Y, Geva D, Gilad R, et al: Cerebrospinal fluid magnesium level as a prognostic factor in ischaemic stroke. J Neurol 245:584-588, 1998  

42. Lee E, Ayoub I, Harris F, et al: Mexiletine and magnesium independently, but not combined, protect against permanent focal cerebral ischemia in Wistar rats. J Neurosci Res 58:442-448, 1999  

43. Lucas M: A comparsion of magnesium sulfate with phenytoin for the prevention of eclampsia. N Engl J Med 333:210-215, 1995  

44. Mangat H, D'Souza G, Jacob M: Nebulized magnesium sulfate versus nebulized salbutamol in acute bronchial asthma. Eur Resp J 12:341-344, 1998  

45. Martin B, Milligan K: Diuretic associated hypomagnesemia in the elderly. Arch Intern Med 147:1768-1771, 1987  

46. Moran J, Gallagher J, Peake S, et al: Parenteral magnesium sulfate versus amiodarone in the therapy of atrial tachyarrhythmias: A prospective, randomized study. Crit Care Med 23:1816-1824, 1995  

47. Muir K: New experimental and clinical data on the efficacy of pharmacological magnesium infusions in cerebral infarcts. Magnes Res 11:43-56, 1998  

48. Nadler JL, Buchanan T, Natara R, et al: Magnesium deficiency produces insulin resistance and increased thromboxane synthesis. Hypertension 21:1024-1029, 1993

49. Nannini L, Hofer D: Effect of inhaled magnesium sulfate on sodium metabisulfite induced bronchoconstriction in asthma. Chest 111:858-861, 1997  

50. Noppen M, Vanmaele L, Impens N, et al: Bronchodilating effect of intravenous magnesium sulfate in acute severe bronchial asthma. Chest 97:373-376, 1990  

51. Okayama H, Aikawa T, Okayama M, et al: Broonchodilating effect of intravenous magnesium sulfate in bronchial asthma. JAMA 257:1076-1078, 1987  

52. Prichard J, Cunningham F, Prichard S: The Parkland Memorial Hospital protocol for treatment of eclampsia: Evaluation of 245 cases. Am J Obstet Gynecol 148:951-963, 1984  

53. Ravn H: Pharmacological effects of magnesium on arterial thrombosis: Mechanisms of action? Magnes Res 12:191-199, 1999  

54. Ryzen E, Wagers PW, Singer FR, et al: Magnesium deficiency in a medical ICU population. Crit Care Med 13:19-21, 1985  

55. Salem M, Munoz R, Chernow B: Hypomagnesemia in critical illness. Crit Care Clin 7:225-252, 1991  

56. Schiermeyer R, Finkelstein J: Rapid infusion of magnesium sulfate obviates the need for intubation in status asthmaticus. Am J Emerg Med 12:164-166, 1994  

57. Schmid-Elsaesser R, Hungerhuber E, Zausinger S, et al: Combination drug therapy and mild hypothermia: A promising treatment strategy for reversible, focal cerebral ischemia. Stroke 30:1891-1899, 1999  

58. Schmid-Elsaesser R, Hungerhuber E, Zausinger S, et al: Neuroprotective effects of combination therapy with tirilazad and magnesium in rats subjected to reversible focal cerebral ischemia. Neurosurgery 44:163-171, 1999  

59. Sharma S, Bhargava A, Pande J: Effect of parenteral magnesium sulfate on pulmonary functions in bronchial asthma. J Asthma 31:109-115, 1994

60. Shechter M, Hod H, Chouraqui P, et al: Magnesium therapy in acute myocardial infarction when patients are not candidates for thrombolytic therapy. Am J Cardiol 75:391-393, 1995  

61. Sibai B, Graham J, McCubbin J: A comparison of intravenous and intramuscular magnesium sulfate regimens in pre-eclampsia. Am J Obstet Gynecol 150:728-733, 1984  

62. Skobeloff E, Spivey W, McNamara R, et al: Intravenous magnesium sulfate for the treatment of acute asthma in the emergency department. JAMA 262:1201-1213, 1989  

63. Skorodin M, Tenholder M, Yetter B, et al: Magnesium sulfate in exacerbations of chronic obstructive pulmonary disease. Arch Intern Med 155:496-500, 1995  

64. Tiffany B, Berk W, Todd I, et al: Magnesium bolus or infusion fails to improve expiratory flow in acute asthma exacerbations. Chest 104:831-834, 1993  

65. Toffaletti J: Physiology and regulation: Ionized calcium, magnesium and lactate measurements in critical care settings. Am J Clin Pathol 104(suppl 1):588-594, 1995

66. Van Gaal L, Delvigne C, Vandewoude M, et al: Evaluation of magnesium before and after jujuno-ileal versus gastric bypass surgery for morbid obesity. J Am Coll Nutr 6:397-400, 1987  

67. Weglicki W, Phillips T, Freedman A, et al: Magnesium deficiency elevates circulating levels of inflammatory cytokines and endothelin. Mol Cell Biochem 110:169-173, 1992  

68. Whang R, Whang D, Ryan MP: Refractory potassium repletion: A consequence of magnesium deficiency. Arch Intern Med 152:40-44, 1992  

69. White R, Hartzell H: Effects of intracellular free magnesium on calcium current in isolated cardiac myocytes. Science 239:778-780, 1988  

70. Woods K: Possible pharmacological actions of magnesium in acute myocardial infarction. Br J Clin Pharmacol 32:3-10, 1991  

71. Woods K, Fletcher S, Roffe C, et al: Intravenous magnesium sulphate in suspected acute myocardial infarction: Results of the second Leicester intravenous magnesium intervention trial (LIMIT-2). Lancet 339:1553-1558, 1992  

72. Yang C: Calcium and magnesium in drinking water and risk of death from cerebrovascular disease. Stroke 29:411-414, 1998  

73. Zaloga G, Chernow B, Pock A, et al: Hypomagnesemia is a common complication of aminoglycoside therapy, Surg Gynecol Obstetr 158:561-565, 1984  

74. Zaloga GP, Roberts PR: Calcium, phosphorus and magnesium disorders. In Ayres SM, Grenvik NA, Holbrook PR, Shoemaker WC (eds): Textbook of Critical Care, 4th ed. Philadelphia, WB Saunders, 2000 pp 905-928  

75. Zaloga G, Wilkens R, Tourville J, et al: A simple method for determining physiologically active calcium and magnesium concentrations in critically ill patients. Crit Care Med 15:813-815, 1987