Causes and clinical conditions associated with hypomagnesemia

Hypomagnesemia, defined as a serum magnesium of <0.75 mmol/L (1.8 mg/dL) ( ) , with the common causes of hypomagnesemia shown in Table 6.2 .

Biochemical manifestations of hypomagnesemia

Hypokalemia . Hypokalemia is common in patients with Mg deficiency and vice versa. Indeed, patients with Mg depletion have a renal loss of potassium caused by an increased potassium secretion in the connecting tubule and the cortical collecting tubule. In the kidneys, K ⁺ is absorbed across the basolateral membrane via Na–K ATPase and secreted into the lumen of the connecting tubule and cortical collecting tubule. This process is mediated by renal outer-membrane potassium channels (ROMK), which require ATP. At normal intracellular Mg ion concentrations, these channels move more K ions into cells. With a lack of intracellular Mg ions, K ions move freely through the ROMK channels, increasing cellular potassium efflux that may be detected as increased potassium in the urine ( ) .

Hypocalcemia . Mg deficiency is often associated with hypocalcemia ( ) . Patients with combined hypocalcemia and hypomagnesemia often have low levels of PTH. This is caused by Mg deficiency inhibiting the release of PTH from the parathyroid gland. Because parenteral Mg stimulates PTH secretion, reduced PTH secretion contributes to hypocalcemia in Mg deficiency ( ) .

Evaluation of magnesium status in patients

Magnesium should be measured during the initial examination of all patients with poor food intake, malabsorption disorders, hypokalemia, or hypocalcemia, as well as those taking magnesium-depleting agents, such as diuretics, alcohol, or aminoglycosides. A patient with hypomagnesemia may show a variety of nonspecific symptoms, such as weakness, muscle cramping, and rapid heartbeat ( ) . While measurements of total magnesium concentrations in serum remain the usual diagnostic test for detection of magnesium abnormalities, it has limitations:

• •
  

  25%–30% of magnesium is protein-bound. Therefore, as with total and ionized calcium, total magnesium may not reflect the physiologically active free ionized magnesium. This is discussed later in the section “Ionized magnesium versus total magnesium.”

  
  
• •
  

  Because Mg is primarily an intracellular ion, serum concentrations will not necessarily reflect the intracellular status of magnesium ( ) . Even when tissue and cellular magnesium is depleted by as much as 20%, serum magnesium concentrations may remain normal.

The magnesium load test may be the definitive method for detecting body depletion of magnesium, as was shown in 13 patients with pancreatitis and a normal concentration of magnesium in serum ( ) . However, because a magnesium load test requires >48 h to complete, its clinical use is limited. After collection of a baseline 24-h urine, 30 mmol (729 mg) of magnesium in a 5% dextrose solution is administered intravenously (i.v.) as another 24-h urine is collected. Individuals with adequate body stores of magnesium will excrete 60%–80% of the magnesium load within 24 h, while magnesium-deficient patients will excrete <50% ( ) .

Urinary magnesium may help confirm the cause of a magnesium deficit. If it is >25–30 mg/d, renal loss is suspected; if <20 mg/d, nonrenal causes should be suspected, such as inadequate intake or GI loss.

In addition to renal loss, acute hypomagnesemia can result from intracellular shifts of magnesium after the administration of glucose or amino acids ( ) . This effect is pronounced after starvation or insulin treatment for hyperglycemia.

Treatment of hypomagnesemia

For patients at risk for magnesium depletion, magnesium supplements may be given to prevent hypomagnesemia. Magnesium replacement therapy may be warranted if serum Mg is <0.5 mmol/L (<1.2 mg/dL) ( ). However, because serum Mg does not necessarily reflects the total body Mg status, patients at risk for magnesium deficiency or with symptoms consistent with hypomagnesaemia should be considered for treatment even with normal serum Mg concentrations ( ) . In acute-care patients, hypomagnesemia is often associated with coronary artery disease and coronary bypass surgery, malignancy, chronic obstructive pulmonary disease, and alcoholism. Among chronic diseases, alcoholism, liver disease, and carcinoma were commonly associated with hypomagnesemia ( ) .

If magnesium infusion is necessary, as a guide, 25 mmol of MgSO ₄ solution given intravenously over 12 h increased serum magnesium concentration to about 1.5 mmol/L, while 50 mmol per 12 h increased serum magnesium to about 2 mmol/L ( ) . Renal function should be assessed to guide supplementation and prevent hypermagnesemia ( ) . In another study, administration of 1 g of MgSO ₄ in solution to critically ill patients increased total Mg by 0.11 mmol/L and ionized Mg by 0.05 mmol/L ( ) .

In general, mild hypomagnesemia with no or only mild symptoms can be treated with an oral supplement, whereas parenteral Mg supplementation is indicated if Mg concentration is <0.5 mmol/L or if the patient presents with significant symptoms ( ) . For critically ill patients with mild-to-moderate hypomagnesemia, empirically administering 1 g (4 mmol) of IV Mg will increase the serum Mg concentration by 0.08 mmol/L within 18–30 h. Trials of patients with acute myocardial infarction suggest an initial bolus (e.g., 2 g (8 mmol)) followed by continuous infusions of up to 16 g (65 mmol) over 24 h. Guidelines for treating severe hypomagnesemia are that lower doses (<6 g MgSO ₄ ) can be given over a period of 8–12 h, whereas higher doses may be administrated over a period of 25 h or longer ( ) . These timings are made especially important due to the combined slow distribution of Mg in tissues and the rapid renal excretion of magnesium ( ) .

When drugs that deplete magnesium, such as gentamicin, diuretics, cisplatin and cyclosporine are given, MgSO ₄ solution can be given intravenously once or more daily, as guided by serum magnesium measurements. Women with toxemia of pregnancy (eclampsia) are sometimes given very large doses of magnesium sulfate intravenously (175 mmol/24 h). This should be monitored with measurements of serum magnesium, because potentially dangerous hypermagnesemia to concentrations of 4.0 mmol/L and above can occur ( ) .

Causes of hypermagnesemia

Significant hypermagnesemia is almost always associated with impaired kidney function, because healthy kidneys can readily excrete excess magnesium loads. A well-functioning kidney handles a magnesium load by diminishing Mg absorption in the thick ascending limb of the loop, where over 50% of filtered magnesium is normally reabsorbed ( ) . While mild elevations in plasma magnesium may be seen in 10%–15% of hospitalized patients, such patients usually have some degree of impaired kidney function. More severe hypermagnesemia can occur when excessive doses of magnesium-containing antacids, enemas, or parenteral nutrition are administered to patients with renal insufficiency ( ) . Hypermagnesemia is sometimes seen in adrenal insufficiency, probably related to diminished aldosterone causing volume depletion and hemoconcentration. The most common cause of dramatically increased plasma magnesium without chronic renal failure is found in treatment for eclampsia of pregnancy, for which large amounts of magnesium salts are given.

Kidney impairment . As kidney function declines, plasma magnesium levels rise because the only regulatory system that is able to remove magnesium is urinary excretion ( , ) . In patients with end-stage kidney disease or on dialysis, magnesium intake must be carefully monitored to maintain control of the plasma magnesium concentration. For example, hypermagnesemia (defined as a serum magnesium greater than 1.5 mmol/L) can occur with magnesium intake as low as 280 mg/day, which is within the range of average intake by the general population of about 200–340 mg/day, depending on weight. Because symptomatic hypermagnesemia can be caused by magnesium given as antacids or laxatives in typical therapeutic doses ( ) , these agents should be avoided in patients with renal impairment.

Magnesium infusion . In pregnant women with severe preeclampsia or eclampsia, parenteral magnesium is commonly given intravenously to decrease neuromuscular excitability ( ) . The usual plasma concentration achieved is 6–8.4 mg/dL or 2.5–3.5 mmol/L, but much higher levels can occur. Frequent complications of hypermagnesemia are hypocalcemia and hyperkalemia, as hypermagnesemia can suppress the release of PTH ( ) . Neonates with hypermagnesemia may also have hyperkalemia and hypocalcemia, along with hypotonia, all of which increase the need for neonatal intensive care ( ) .

Oral ingestion . Ingestion of very large oral doses of magnesium may exceed renal excretory capacity, especially if either chronic kidney disease or acute kidney injury is present ( ) . Accidental overdose can also occur with over-the-counter products, such as Epsom salts (MgSO ₄ ) or laxatives ( ) .

Enemas . Enemas usually contain high concentrations of MgSO ₄ , so that substantial quantities of magnesium can be absorbed from the large bowel following an enema. Even in normal subjects, enemas containing 400–800 mmol of MgSO ₄ can raise the plasma magnesium concentration to 7.2–19.2 mg/dL (3–8 mmol/L). This amount of magnesium given as an enema to patients with kidney failure can be fatal ( ) .

Symptoms of hypermagnesemia

Most symptoms of hypermagnesemia are from neuromuscular effects, cardiovascular effects, and hypocalcemic effects. Symptomatic hypermagnesemia may occur with the following plasma concentrations of magnesium ( ) :

• •
  

  2–3 mmol/L (4.8–7.2 mg/dL): Nausea, flushing, headache, lethargy, drowsiness, and diminished deep tendon reflexes.

  
  
• •
  

  3–5 mmol/L (7.2–12 mg/dL): Somnolence, hypocalcemia, absent deep tendon reflexes, hypotension, bradycardia, and electrocardiogram (ECG) changes.

  
  
• •
  

  Above 5 mmol/L (12 mg/dL): Muscle paralysis leading to quadriplegia, apnea, and respiratory failure that usually precedes complete heart block and cardiac arrest.

Neuromuscular effects are the most common complication of hypermagnesemia, because increased magnesium decreases impulse transmission across the neuromuscular junction, and this can be detected by diminished deep tendon reflexes, which are usually noted when the plasma magnesium concentration reaches 2–3 mmol/L (4.8–7.2 mg/dL) ( ) .

Cardiovascular effects are related to excess intracellular Mg blocking cardiac K channels. Hypotension, conduction defects, and bradycardia begin to appear at a plasma magnesium concentration of 2–2.5 mmol/L (4.8–6 mg/dL). ECG changes may occur at concentrations of 2.5–5 mmol/L (6–12 mg/dL). Complete heart block and cardiac arrest may occur at a plasma magnesium concentration above 7.5 mmol/L (18 mg/dL).

Moderate hypermagnesemia can inhibit the secretion of PTH, leading to a decreased plasma calcium concentration. The fall in the plasma calcium concentration is usually transient and asymptomatic, although ECG abnormalities may occasionally occur.

Treatment of hypermagnesemia

The approach to therapy largely depends on kidney function. If kidney function is normal, stopping the source of magnesium will promptly restore normal plasma magnesium concentrations. If necessary, loop or thiazide diuretics can be used to increase renal excretion of magnesium. In patients with mild-to-moderate degrees of either chronic or acute kidney impairment (eGFR between 15 and 45 mL/min/1.73 m ² ), urinary elimination of magnesium may be limited. Again, the initial treatment begins with removing magnesium-containing medications followed by administering intravenous isotonic fluids such as saline plus a loop diuretic. In patients with end-stage kidney failure (eGFR less than 15 mL/min/1.73 m ² , dialysis may be the only option ( ) .

Diagnosis of magnesium abnormalities

While measurement of total magnesium concentrations in serum remains the usual diagnostic test for the detection of magnesium abnormalities, total magnesium concentrations have two limitations. First, about 30% of magnesium is protein bound. Therefore, as with total versus ionized calcium, total magnesium may not reflect the physiologically active ionized magnesium. Second, and probably of greater significance, because magnesium is primarily an intracellular ion, concentrations of either total or ionized magnesium in blood will not necessarily reflect the intracellular and overall body stores of magnesium. Symptomatic hypomagnesemia may not appear until serum total magnesium levels fall below 0.5 mmol/L (well below the typical reference limit), which make associations between serum magnesium concentrations and the benefit of treatment difficult ( ) .

The magnesium retention test assesses overall magnesium status, but its use may be restricted to special circumstances when there is a strong suspicion of magnesium deficiency despite a normal serum magnesium. A magnesium retention test requires over 48 h to complete, collection of two 24-h urines, and giving 30 mmol of MgSO ₄ intravenously over 12 h. If more than 60%–70% of Mg is excreted in the urine following an intravenous load, Mg deficiency is unlikely, while magnesium-deficient patients excrete less than 50% of the administered magnesium ( , ) . Normal renal handling of Mg is necessary for this test to be useful because increased magnesium losses due to diabetes, medications or alcohol ingestion may result in a false-negative test. If kidney function is compromised, one may see false-positive results, so the Mg retention test should not be used in patients with renal impairment or in transplant patients receiving cyclosporine or tacrolimus, both of which cause urinary Mg wasting ( ) .

Ionized Mg versus total Mg

The routine measurement of ionized magnesium has been available for many years since sensors for magnesium ion electrodes were developed from lengthy studies on over 200 ionophores produced by the late Dr. Wilhelm Simon and his colleagues ( ) . However, ionized magnesium measurements have more analytical variability than ionized calcium measurements because all Mg ion-sensitive electrodes are also sensitive to Ca ions. While this interference by Ca ion is corrected by simultaneous measurement of the Ca ion concentration, this additional test adds variability to the ionized magnesium result ( ) .

There are also a few studies relating intracellular levels of magnesium to clinical information. For example, one study concluded that red blood cell magnesium was the best parameter to measure because it gave a higher incidence of hypomagnesemia (37%) than did either ionized magnesium (22%) or total magnesium (16%) ( ) . However, regardless of its possible clinical value, at least for the foreseeable future, these measurements are not practical for clinical use.

While some studies conclude that measurements of ionized magnesium improve clinical accuracy, others find that either total or ionized magnesium provides about the same clinical value in detecting hypomagnesemia. Conclusions are often based on whether the total serum magnesium and ionized Mg do or do not correlate well in various disease states.

While it is logical to equate the physiologic activity of ionized Mg with better clinical reliability, there are other issues to consider. Total magnesium measurements have better analytical reliability and have far greater familiarity in clinical practice. However, because ionized magnesium measurements are on “blood gas” analyzers, they have the advantage of faster turnaround times and can be used in point-of-care settings. Hopefully, ionized magnesium will become more widely used and gain clinical familiarity.

Here are representative reports that conclude the total and ionized Mg results give equivalent clinical information.

Saha et al. reported strong correlations between total and ionized Mg concentrations in serum from hemodialysis patients, and in patients with intestinal disease, alcoholic liver disease, and chronic renal disease ( , ) .

Lanzinger et al. often found that, in patients during abdominal surgery who did not have massive transfusions, the changes in ionized and total serum Mg concentrations correlated closely ( ) . Even though total serum Mg slightly overestimated the prevalence of hypomagnesemia, it adequately screened for hypomagnesemia.

Koch et al. found a strong correlation between ionized and total Mg ( r = 0.90), with hypomagnesemia in 21% of ICU patients measured by ionized Mg and 18% of ICU patients measured by total Mg. They concluded that ionized Mg could be inferred from total Mg ( ) .

In a study on 286 patients of indigenous and nonindigenous Australian descent, with or without diabetes, Longstreet and Vick noted significant correlations between total and ionized magnesium measurements in all groups ( r = 0.75; P < .001). Among people with diabetes, the correlation coefficient ( r ) was 0.81 whereas in nondiabetics, the r was 0.66. Based on the strong correlation between total and ionized Mg concentrations in serum, they concluded that either gives an accurate assessment of magnesium status in health and chronic diabetes, irrespective of ethnicity ( ) .

Here are representative reports that conclude that ionized Mg measurements provide better clinical information.

A poor correlation between ionized and total Mg was noted in critically ill patients in studies by both Barrera et al. ( r = 0.57) ( ) , and by Johansson and Whiss (r = 0.59) ( ) . However, after magnesium supplementation, both ionized and total Mg increased: ionized by 0.05 mmol/L and total by 0.11 mmol/L ( ) .

Fiser et al. observed that ionized magnesium was decreased in many critically ill pediatric patients who had normal total magnesium concentrations ( ) . In this study, the use of albumin-corrected total magnesium was not reliable in estimating the ionized magnesium.

In patients undergoing cardiopulmonary bypass surgery, plasma ionized (but not total) magnesium was decreased significantly by 24 h after bypass ( ) .

In the report by Yeh et al. ( ) the ionized magnesium result was assumed to always be “correct.” For example, if any magnesium supplementation was given based on a low total magnesium (34 patients), it was called “inappropriate administration” in 28 of the patients who had normal ionized Mg results. The 28 with low total Mg but normal ionized Mg were considered erroneously low, and resulted in 27 additional and presumably unnecessary repeat Mg measurements and administration of 60 g of MgSO ₄ .

Variable correlations were reported by Hutton et al. ( ) . In the general ICU population, they reported a strong correlation (r = 0.95) between ionized and total Mg, and 144 of the 185 samples fell in the same category (hypo-, normo-, or hypermagnesemia) for both ionized and total Mg based on their corresponding reference intervals of 0.49–0.71 mmol/L and 0.70–1.0 mmol/L, respectively. Also, the regression slope of 0.71 corresponded with 70% of magnesium being in the ionized form. However, in citrate-anticoagulated patients treated with continuous venovenous hemofiltration (CVVH), half of these patients showed normal total Mg and decreased ionized Mg. Based on this, they concluded that ionized Mg was better than total Mg as a marker for hypomagnesaemia in these CVVH patients.

The results from the study by Wilkes et al. could be interpreted to favor either ionized or total Mg measurements. They emphasized that correcting the ionized Mg by magnesium supplementation during cardiopulmonary bypass (CPB) reduced postoperative risks. However, their figures show that both total and ionized Mg declined almost identically during CPB and increased almost identically when magnesium supplementation was given to correct the hypomagnesemia ( ) .

To conclude the discussion of whether total or ionized magnesium measurement is more clinically appropriate, Longstreet and Vink provided an interesting theory ( ) . The chronic or acute nature of the patients’ condition may be relevant. In most chronic conditions, a generalized magnesium deficiency would maintain a consistent decline in both the total and ionized magnesium pools. However, in acute conditions that often require critical care, the rapid onset could rapidly alter the binding of Mg ions in serum and possibly cause greater relative differences between total and ionized magnesium concentrations. Thus, the chronic or acute nature of the disease process may be most relevant in deciding to measure total or ionized Mg to assess magnesium status.