Comparison of Intravenous and Oral Magnesium Replacement in Hospitalized Patients With Cardiovascular Disease

Brent N. Reed; Su Zhang, J. S. Marron; Deborah Montague


Am J Health Syst Pharm. 2012;69(14):1212-1217. 

In This Article


Our study demonstrated that among patients with a baseline SMC of 1.4–1.8 mg/dL, the use of i.v. magnesium sulfate was associated with a greater increase in SMC than the use of oral magnesium oxide. However, because both magnesium formulations consistently increased SMC values from baseline, these results may be interpreted in different ways.

Clearly, i.v. magnesium was more effective in quickly (i.e., within 24 hours) increasing SMC values from baseline values of 1.4–1.8 mg/dL. However, raising the SMC so rapidly is rarely indicated outside the context of managing life-threatening ventricular arrhythmias or the prevention of such arrhythmias in patients at the highest risk (e.g., those with evidence of ventricular ectopy on telemetry).[9][11] Moreover, the long-term cardiovascular benefit of intermittently increasing the SMC during hospitalization is not well established. The doses of i.v. magnesium commonly used to correct low SMCs during an acute-care episode are unlikely to provide the amount of magnesium necessary to correct chronic hypomagnesemia. For example, in one analysis, magnesium sulfate dosages of >7 g daily for seven days were required for the adequate correction of total magnesium stores.[12]

While not as effective as i.v. magnesium in quickly increasing SMCs, oral magnesium oxide produced a consistent median SMC increase of 0.1 mg/dL from baseline SMCs of 1.4–1.8 mg/dL. Our findings suggested that when a rapid SMC change is not clinically indicated, oral magnesium oxide may be a suitable option for magnesium replacement. The initiation of oral magnesium replacement during an acute-care episode may also be a favorable strategy for managing and preventing hypomagnesemia pending a patient's transfer to the outpatient environment, where chronic hypomagnesemia has been associated with a variety of disease states and conditions, including ischemic heart disease and heart failure.[1]

The observed inverse relationship between the baseline SMC and the change in SMC after i.v. magnesium therapy indicated that the i.v. route may not provide a substantially greater benefit than the oral route for patients with baseline SMCs close to the normal range. Our data suggest that the mean changes in SMC produced by i.v. magnesium sulfate and oral magnesium oxide are likely to be approximately equal for patients with a baseline SMC of 2.0 mg/dL. However, because i.v. magnesium was rarely prescribed (relative to oral magnesium oxide) for patients with SMCs approaching 2.0 mg/dL in the evaluated cases, our analysis did not include treatment courses in the context of baseline SMCs of >1.8 mg/dL.

Additionally, we did not find a dose–response relationship between the total dose of oral magnesium and the change in SMC. We attribute this to the small numbers of evaluated courses involving the highest doses of oral magnesium (fewer than 20 courses each of 1200 and 1600 mg, compared with 127 courses of 800 mg) rather than the absence of a dose–response relationship.

In our linear regression analysis, the route of magnesium administration, the timing of SMC measurements, renal function, and the concomitant use of loop diuretics were significantly associated with the change in SMC. Surprisingly, drugs expected to affect magnesium absorption did not appear to affect the bioavailability of oral magnesium oxide (and thus the change in SMC) in our analysis. We had expected that renal function and the concomitant use of diuretics would affect SMC changes with both oral and i.v. forms of magnesium replacement, and those expectations were borne out by the study findings; our analysis indicated that neither route of magnesium delivery conferred a superior benefit in the presence of renal dysfunction or diuretic use.

The timing of SMC measurement was associated with the degree of change in SMC with both routes of magnesium administration. This relationship was more evident in patients who received i.v. magnesium sulfate and may reflect the pharmacokinetic processes involved in i.v. drug delivery. Intravenous magnesium sulfate is delivered immediately to the intravascular space and is subject to distribution between the intracellular and extracellular compartments, as well as renal elimination. With oral magnesium oxide, such processes also occur but are limited by the rate of absorption from the gastrointestinal tract. We attempted to control for these potential pharmacokinetic differences by analyzing only those SMC values obtained more than six hours after the completion of a magnesium course, as we expected that such values would better reflect equilibrium than SMC values obtained sooner after magnesium administration by either route. Because we did not evaluate serial measurements of SMC, our study could not determine the extent to which SMC changes from baseline might differ at specific time points.

In the cases evaluated in our study, a median of 8.25 hours elapsed between the completion of a course of oral magnesium and blood collection for SMC determination. At our institution, when magnesium oxide is scheduled for twice-daily administration, the second dose is usually scheduled for 10 p.m., approximately 6 hours before the following morning's blood samples are usually drawn. Complete oral absorption of a magnesium dose requires 6–8 hours,[13] with the full effect on SMC occurring even later, so our analysis likely did not capture the full SMC-modulating effect of magnesium courses in at least some cases.

As expected, the courses of oral magnesium oxide examined in this analysis did not appear to have a greater effect on the number of bowel movements than the evaluated courses of i.v. magnesium sulfate. Because relatively few courses involved total magnesium oxide doses of 1200 or 1600 mg, we cannot conclude that those higher-dose courses did not have a greater influence on the number of bowel movements relative to the 800-mg courses. Nonetheless, we hope that our findings might have some small effect in reducing the reluctance of some clinicians to prescribe oral magnesium for routine replacement. A more comprehensive evaluation would have considered both the number and the character of bowel movements. While the quality of bowel movements was occasionally noted in patient records, such documentation was inconsistent, and we were thus unable to analyze this aspect of magnesium therapy in our investigation. Other therapies may also affect bowel function (e.g., opiate analgesics, prokinetic agents, antibiotics), but given the nature and number of patients included in this analysis, we did not expect these agents to have a notable influence on our results. For future analyses, we suggest classifying both the quantity and quality of bowel movements with validated assessment scales.[14]

There were limitations to our analysis beyond those already mentioned. Most important, the SMC is a poor surrogate for total-body magnesium stores. However, because SMC values drive clinical decisions, we deemed it important to use this measure as the primary variable of analysis. The retrospective nature of our analysis did not allow for the inclusion of a placebo group, so we were unable to determine how SMC changes after i.v. or oral replacement compare with normal daily SMC fluctuations. Because decisions to administer i.v. versus oral magnesium in the evaluated cases were not made prospectively, we attempted to control for bias by conducting a preliminary analysis that identified the SMC range in which similar amounts of i.v. and oral magnesium were typically prescribed at our institution; that range (1.2–1.8 mg/dL) was consequently used for our primary analysis.

The primary data points in the analysis were SMC values. As a result, patients could be included in the analysis multiple times if their SMCs met the study inclusion criteria each day. Although this may have biased our results, we believe that is unlikely given the number of patients included in this analysis (n = 211).

Finally, the pharmacokinetic limitations of oral magnesium oxide have been reviewed previously.[13] The bioavailability of magnesium differs with the oral magnesium product used, and some formulations may produce a greater change in SMC than magnesium oxide; however, the only oral magnesium formulation available at our institution during the period of data collection for this investigation was magnesium oxide.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.