Author + information
- Received June 3, 1996
- Revision received July 19, 1996
- Accepted December 19, 1996
- Published online April 1, 1997.
- Manfred Zehender, MDA,*,
- Thomas Meinertz, MDB,
- Thomas Faber, MDA,
- Anton Caspary, MDB,
- Andreas Jeron, MDA,
- Katja Bremm, MDA,
- Hanjörg Just, MDA,1,
- for the MAGICA Investigators
- ↵*Dr. Manfred Zehender, Innere Medizin III, Universitätsklinik Freiburg, Hugstetterstrasse 55, 79106 Freiburg, Germany.
Objectives. This study sought to assess potential antiarrhythmic effects of an increase in the daily oral intake of magnesium and potassium in patients with frequent ventricular arrhythmias.
Background. Magnesium and potassium contribute essentially to the electrical stability of the heart. Despite experimental and clinical evidence for the antiarrhythmic properties of the two minerals, controlled data in patients with stable ventricular arrhythmias are lacking.
Methods. In a randomized, double-blind study, 232 patients with frequent ventricular arrhythmias (>720 ventricular premature beats [VPBs]/24 h) confirmed at baseline and after 1 week of placebo therapy were subsequently treated over 3 weeks with either 6 mmol of magnesium/12 mmol of potassium-dl-hydrogenaspartate daily or placebo.
Results. Compared with placebo pretreatment, active therapy resulted in a median reduction of VPBs by −17.4% (p = 0.001); the suppression rate was 2.4 times greater than that in patients randomized to 3 weeks of placebo therapy (−7.4%, p = 0.038). The likelihood of a ≥60% (predefined criterion) or ≥70% suppression rate (calculated from the placebo-controlled run-in period) was 1.7 (25% vs. 15%, p = 0.044) and 1.5 times greater in the active than in the placebo group (20% vs. 13%, p = 0.085), respectively. No effect of magnesium and potassium administration was observed on the incidence of repetitive and supraventricular arrhythmias and clinical symptoms of the patients.
Conclusions. To our knowledge, this study is the first to provide controlled data on the antiarrhythmic effect of oral administration of magnesium and potassium salts when directed to patients with frequent and stable ventricular tachyarrhythmias. A 50% increase in the recommended minimum daily dietary intake of the two minerals for 3 weeks results in a moderate but significant antiarrhythmic effect. However, with the given therapeutic regimen, repetitive tachyarrhythmias and patient symptoms remain unchanged.
(J Am Coll Cardiol 1997;29:1028–34)
© 1997 by the American College of Cardiology
Electrical stability of the heart is maintained by electrolyte homeostasis. Potassium and magnesium are two of the most important minerals contributing to the generation and regulation of ion channel transport processes in myocytes and by this are essentially involved in creating normal cellular excitability, proper and uniform impulse propagation and regular ventricular recovery. Experimental and clinical studies ([1–3]) have provided increasing evidence that low plasma concentrations of the two minerals increase the risk of induction, facilitation or aggravation of ventricular tachyarrhythmias (VTs).
Therapeutic interventions based on this concept have primarily been directed at the prevention of electrolyte depletion. More recently, the administration of magnesium salts in particular has gathered increasing interest due to its safety profile, potential effects on preserving myocardial function and potent antiarrhythmic properties. In most small, but not all larger, studies (), the administration of magnesium salts in patients with cardiac arrhythmias resulting from chronic heart failure ([5–7]) or an acute ischemic syndrome ([8–10]) during perioperative coronary bypass graft surgery was found to be effective in reducing tachyarrhythmic complications.
The concept of preserving electrical stability by a positive magnesium and potassium balance is attractive because of its simplicity, cost-effectiveness and safety. However, only very few data and no controlled studies exist to prove any antiarrhythmic effect of the two minerals in patients with frequent VTs under stable conditions. Although this group of patients had usually not been directed to antiarrhythmic treatment in the mid-1990s, the proven antiarrhythmic effects of magnesium and potassium would provide a safe alternative to the present problems with class I antiarrhythmic agents. If proven efficacious, potassium and magnesium might also be used as adjunctive treatment in patients receiving antiarrhythmic drug medication.
The present double-blind, randomized, placebo-controlled study was aimed at a large series of consecutive patients presenting with frequent and stable VTs to assess the potential antiarrhythmic benefit of increasing the daily recommended minimal dietary intake of magnesium and potassium by 50% over a 3-week treatment period.
From October 1991 to September 1994, 307 consecutive patients who presented with frequent VPBs were enrolled in the study at 20 European cardiology centers. The study protocol was approved by the Institutional Review Board of the University of Freiburg. All participants gave written informed consent before enrollment in the study.
Inclusion criteria for the study were 1) presence of frequent VPB’s (>720 ventricular premature beats [VPBs]/24 h during baseline 24-h Holter recording); 2) confirmation of VPB frequency after 1 week of placebo therapy (analysis A). Patients without the second criterion were also randomized and treated but were only considered for a secondary analysis conducted in all patients without regard to arrhythmia variability (analysis B).
Patients had to be 18 to 75 years old and have a left ventricular ejection fraction >25% as well as potassium (range 3.5 to 5.5 mmol/liter) and magnesium plasma levels (0.75 to 1.25 mmol/liter) within the normal range. Exclusion criteria (alone or in combination) included the presence of malignant ventricular arrhythmias (>6 consecutive beats >120/min; history of sustained VT/ventricular fibrillation), atrial fibrillation, evidence or history of important renal (serum creatinine concentration >120 μmol/liter) or liver impairment, thyroid malfunction, New York Heart Association functional class IV, antiarrhythmic pretreatment, angiotensin-converting enzyme inhibitors or potassium-sparing or potassium-depleting diuretic drugs given <5 days before the study. Oral medication was optimized before entering the runin phase of the study and was kept constant throughout the 4-week study period. Of the 232 patients included in analysis A, the use of aspirin (44 patients), nitrates (39 patients), antihypertensive drugs (25 patients), digitalis (), calcium channel blocking agents (10 patients) and diuretic drugs (5 patients) was not different in the active and placebo groups; cardiovascular medication was changed during the study in only 4 patients.
1.2 Study design
Complete history, followed by physical examination and an extensive diagnostic workup, was taken in all patients qualifying for the study. Patients then entered the placebo-controlled 1-week run-in period. At day 7, all patients had repeat 24-h Holter monitoring and were reassessed by means of a standardized questionnaire for the presence of clinical symptoms possibly related to the clinical arrhythmia.
Patients were then randomized to 3 weeks of either active or placebo treatment. In the active treatment group, daily oral dosing consisted of 2.163 mg of magnesium-dl-hydrogenaspartate (6 mmol magnesium) and 2.162 mg of potassium-dl-hydrogenaspartate (12 mmol potassium) daily. The dose was chosen to increase the recommended minimal daily dietary intake of magnesium (12 to 15 mmol) and potassium (20 to 30 mmol) by ∼50% in addition to usual diet ().
After 21 days of active or placebo treatment, 24-h Holter registration, plasma levels of magnesium/potassium, laboratory chemistry and 12-lead surface electrocardiogram (ECG) were repeated. Clinical symptoms of the patients were assessed by repeating the baseline questionnaire (10 items to describe the clinical symptoms, from 0 = no complaints to 5 = very severe complaints). In addition, the Clinical Global Impression Scale was used by the physician to determine overall benefit of therapy.
All 24-h Holter recordings were assessed by amplitude-modulated systems and analyzed centrally within 48 h at the University Hospital Freiburg using a Marquette Laser SXP0 system. Supraventricular and ventricular arrhythmias were identified and interpreted by a physician. Representative tracings were reviewed and interpreted as 1) VPBs; 2) multiform VPBs; 3) ventricular pairs (2 repetitive VPBs); and 4) ventricular salvos (≥3 repetitive VPBs). Holter recordings with an analyzable registration time <22 h were repeated.
Comparison of drug response in the two study groups was based on the Holter recording obtained immediately before randomization and after 3 weeks of active or placebo treatment (considered for analysis A: patients with >720 VPBs/24 h during two 24-h Holter recordings separated by 1 week of placebo medication; analysis B: patients with >720 VPBs/24 h during baseline 24-h Holter recording). The frequency of VPBs was the only major end point; repetitive arrhythmia was not considered a relevant end point because of the low frequency and higher spontaneous variability of this type of arrhythmia. Two response criteria for suppression of VPBs were used. One criterion was empirically predefined by the investigator group (reduction rate of VPBs >60%, criterion 1). A second reduction criterion was calculated on the basis of the spontaneous arrhythmia variability in our particular patient cohort as assessed during the initial Holter recordings and those done during placebo treatment. Regression analysis was used to determine the one-tailed 95% confidence intervals of spontaneous variability in the examined patient cohort by comparing the frequency of VPBs at baseline and after 1 week of placebo therapy ([12, 13]). The calculated minimal suppression rate was 70.2% (5% significance level, criterion 2). Because of the low incidence and high spontaneous variability of repetitive arrhythmias, no reduction criterion was predefined or calculated.
To improve visualization of the antiarrhythmic effects of the two minerals, a previously described () graphic display was chosen that is more sensitive to smaller but, over the total patient cohort, constant changes in the arrhythmia frequency.
All blood samples for measuring magnesium and potassium plasma levels were obtained under similar, standardized conditions. Blood samples were allowed to clot and then centrifuged at 4°C. The serum was stored at −70°C until it could be centrally analyzed. Serum levels of magnesium and potassium were determined in duplicate by spectrophotometric measurements.
1.3 Statistical analysis
Calculations of sample size were obtained after 200 patients had completed the runin placebo period and indicated that a total of 300 patients were required to prove the hypothesis that active treatment would be effective in reducing VPBs, with a significance level of 0.05 and a power of 0.80.
Data analysis considered patients with >720 VPBs/24 h before and after 1-week of placebo treatment (analysis A), as well as all patients regardless of low or high spontaneous arrhythmia variability (analysis B). Statistical analysis was based on regression analysis according to previous recommendations by Andresen () and Sami () and showed a minimal suppression rate of 70.2% (one-tailed 95% confidence interval). The antiarrhythmic effect of active treatment was assessed by 1) comparing the effect of 1 week of placebo treatment versus 3 weeks of magnesium/potassium treatment and 1 versus 3 weeks of placebo therapy (two-tailed Wilcoxon test, two-tailed Mann-Whitney Utest); 2) the incidence of responders (VPB reduction >60% and >70%) after 3 weeks of placebo or active treatment (linear regression analysis); and 3) comparison of 3 weeks of randomized placebo or active treatment (two-tailed Mann-Whitney Utest). Results are given as median value (25, 75 quartiles), except for data showing a normal distribution (mean value ± SD).
Of 307 consecutive patients entering the placebo period, 232 patients (76%) presented with >720 VPBs/24-h registration period before and after 1 week of placebo therapy and were considered for inclusion in analysis A. Clinical and arrhythmia data were not different for patients randomized to active or placebo therapy (Table 1).
2.2 Ventricular and supraventricular tachyarrhythmias
In patients with a stable arrhythmia frequency, magnesium and potassium administration resulted in a median reduction of daily VPB rate from 7,651 to 4,421 (median reduction −17.4%) compared with 7,485 to 5,578 VPBs/24 h (−7.4%) in the placebo group (p = 0.038) (Table 2). The antiarrhythmic effect of magnesium/potassium administration was observed when 1 week of placebo and 3 weeks of active treatment (p < 0.001) (Fig. 1, top) was compared with 1 and 3 weeks of placebo treatment (p = 0.97) (Fig. 1, bottom), as well as when only the two randomized treatment periods (placebo vs. active treatment for 3 weeks) were compared (Fig. 2). The graphic display also allows easy recognition of the proportion of patients in the two groups with a >60% or a >70% reduction and patients with any increase or a >100% increase in arrhythmia frequency.
Using the predefined (>60%) and calculated (>70%) VPB suppression criteria in patients with stable arrhythmia frequency, active treatment resulted in a 1.7 (25% vs. 15%, p = 0.044) and 1.5 times (20% vs. 13%, p = 0.085) higher response rate than placebo treatment (Table 2). Antiarrhythmic effects of magnesium/potassium administration were not predictable by clinical variables, except for the presence or absence of a history of myocardial infarction (−51% vs. −8% in the placebo group, p < 0.05).
Repetitive arrhythmias present at randomization in 99 (active group) and 93 patients (placebo group) were similar in response to active or placebo therapy (51 vs. 26 episodes/24 h [−49%]; 37 vs. 20 episodes/24 h [−46%], respectively) (Table 2). Lown class decreased by ≥1 class in 22 (active group) and 18 (placebo group) patients, respectively. Similarly, supraventricular arrhythmias remained unchanged when magnesium/potassium (median VPBs/24 h [25, 75 quartiles]: 43 [10, 316] vs. 61 [20, 275], p = NS) and placebo therapy were compared (63 [12, 185] vs. 41 [8, 177], p = NS).
2.3 Plasma level of magnesium and potassium
Three weeks of treatment resulted in no significant change in serum levels of magnesium (active treatment: 0.85 ± 0.03 vs. 0.88 ± 0.04 mmol/liter; placebo treatment: 0.87 ± 0.02 vs. 0.89 ± 0.03 mmol/liter) or potassium (4.5 ± 0.2 vs. 4.6 ± 0.2 mmol/liter; 4.4 ± 0.2 vs. 4.6 ± 0.2 mmol/liter).
In the active treatment group, linear regression analysis failed to detect any positive correlation between the change in plasma concentration of magnesium (r = 0.32) or potassium (r = 0.36) and VPB frequency.
2.4 Side effects, clinical symptoms and Global Clinical Impression Score
Side effects were reported by 20 patients (20%) in the active treatment group (gastrointestinal symptoms in 18), and by 17 (15%, p = NS) in the placebo group (gastrointestinal symptoms in 15). Drug treatment was stopped because of side effects in four and two patients, respectively.
Clinical symptoms potentially related to ventricular arrhythmias were found to be not different for the two study groups (Fig. 3). The Global Clinical Impression Score as determined by the treating physician also demonstrated no difference between the two study groups (excellent response: 3.5 vs. 5.6% of patients; moderate response: 19.6 vs. 21.0%; no recognizable response: 76.8 vs. 73.1%). Systolic and diastolic blood pressure, heart rate and all ECG variables (PQ, QRS, QT intervals) were unaffected by active or placebo treatment.
2.5 Analysis of all patients independent of the presence of low or high arrhythmia variability (analysis B)
When the analysis was based on all 307 patients with >720 VPBs/24 h during baseline 24-h Holter recording, the arrhythmia suppression rate was −12.2% in the active group and +2.2% in the placebo group (p = 0.049). Similar to analysis A, the responder rate to both the predefined (>60% reduction of VPB, p = 0.008) and calculated VPB suppression criterion (>70%, p = 0.032) were found to be significant (Table 2).
The concept of preserving and improving the electrical stability of the heart by increasing the daily intake of magnesium and potassium salts has generated controversy in recent years. Experimental evidence and encouraging clinical data ([5–10]) stressing some antiarrhythmic effects of the two minerals are contrasted by a lack of controlled studies.
3.1 Antiarrhythmic effects of increased intake of magnesium and potassium in patients with stable, frequent ventricular arrhythmias
In a double-blind, randomized, placebo-controlled study, 307 patients with frequent ventricular arrhythmias were treated for 3 weeks by placebo or magnesium and potassium, to increase the daily recommended minimal dietary allowance of the two minerals by 50%.
Primary data analysis was based on those 232 patients who presented with low spontaneous arrhythmia variability (>720 VPBs/24 h before and after 1 week placebo treatment, mean frequency >300 VPBs/h). Secondary analysis included all 307 patients with >720 VPBs/24 h at baseline, independent from the observed spontaneous arrhythmia variability. In patients with stable ventricular arrhythmias, oral magnesium and potassium administration resulted in a significant reduction of arrhythmia frequency. This effect was manifest when 3-week magnesium and potassium therapy was compared with the 1-week placebo runin period, as well as when comparison was made only between the two randomized 3-week active or placebo treatment periods. The predefined ≥60% suppression criterion (selected to be proved an antiarrhythmic effect) and the calculated ≥70% suppression criteria (logistic regression analysis) were met by 25% and 20% of patients after oral magnesium and potassium administration compared with 15% and 13% of patients in the randomized placebo group. Comparison of the two 3-week treatment periods demonstrated a 2.4 times greater reduction rate after magnesium/potassium than after placebo administration. The antiarrhythmic response to magnesium and potassium was also significant when all patients were analyzed independent of their spontaneous arrhythmia variability. There are no other controlled data that can be compared with the present study; uncontrolled studies published in the late 1980s are well in line with our findings ().
Of all clinical variables analyzed, none were predictive of drug-induced arrhythmia suppression, except for the presence of a history of myocardial infarction. Plasma concentration of the two minerals was found to be unaffected after 3 weeks of active treatment, and there was no correlation between changes in measured plasma concentration and the response of ventricular arrhythmias during drug therapy. Neither finding was unexpected because only 1% to 2% of total body store of the two minerals is in the extracellular fluid ([3, 11]).
Clinical symptoms present at the time of randomization showed a significant improvement during drug treatment, but this effect was not different between active and placebo medication. In light of previous uncontrolled observations reporting an improvement in clinical symptoms after magnesium (), these data clearly emphasize the importance of a controlled study design.
3.2 Limitations of the study
All data analyses must be interpreted in the light of the dosage and duration of magnesium/potassium administration used in the present study. It is not expected that any antiarrhythmic effect of magnesium and potassium administration will be similar to that observed with potent class I antiarrhythmic agents, which are now known to cause significant side effects in some groups of patients ([15, 16]). With decreasing antiarrhythmic potency of the drug it becomes more difficult to prove any positive or negative drug effect. A controlled study design, placebo-controlled verification of arrhythmia frequency and the use of different statistical tests cannot completely exclude the influence of spontaneous arrhythmia variability; however, they are very helpful in minimizing this influence.
To our knowledge, the present study is the first to provide controlled data on the antiarrhythmic effects of oral administration of magnesium and potassium salts in patients with frequent ventricular tachyarrhythmias. An increase of 50% in the daily intake of magnesium and potassium salts for 3 weeks results in a moderate but significant antiarrhythmic effect. Although the likelihood of a 60% reduction of VPBs is increased from every seventh to every fourth patient, no significant effect of this therapeutic regimen on repetitive ventricular tachyarrhythmias or on clinical symptoms was observed.
The present study should not be considered an argument for treating patients with nonmalignant ventricular tachyarrhythmias. In addition, it is uncertain whether any suppression of Holter-documented ventricular arrhythmias will translate into any clinical benefit. However, in light of its simplicity, cost-effectiveness and safety, increasing the daily intake of magnesium and potassium salts may provide another first-line option for treating patients with frequent but not life-threatening ventricular tachyarrhythmias.
We thank B. Balzer, MS for help in the organization and performance of the MAGICA study.
A.1 Participating Investigators and Institutions for the MAGICA Study
Prof. Dr. G. Bachour, Cardiological Practice, Ahlen (Germany); Dr. C. P. Billing, Cardiological Practice, Essen (Germany); Prof. Dr. W. Enenkel, Krankenhaus der Stadt Wien, Vienna (Austria); Dr. J. Gadow/Dr. J. Wunderlich, Cardiological Practice, Berlin (Germany); Dr. F. Gulic, Allgemeines Krankenhaus, Maribor (Slovenia); Dr. F. K. Hagel, Cardiological Practice, Nürnberg (Germany); Dr. R. Häge, Cardiological Practice, Dillingen (Germany); Dr. D. Hüwel, Cardiological Practice, Dillingen (Germany); Prof. Dr. W. Klein, Ambulatory Practice for Long-Term ECG, Graz (Austria); Prof. Dr. E. Knapp, Universitätsklinikum, Innsbruck (Austria); Dr. W. Kotulla, Cardiological Practice, Waldkirch (Germany); Prim. Prof. Dr. G. Mähr, Landeskrankenhaus, Feldkirch, (Austria); Prof. Dr. T. Meinertz, Universitätsklinik Eppendorf, Hamburg (Germany); Dr. G. Prager, Cardiological Practice, Regensburg (Germany); Dr. H. Petri, Cardiological Practice, Hamburg (Germany); Prof. Dr. P. Rakovec, University Medical Centre, Ljubljana (Slovenia); Prim. Prof. Dr. S. Sailer, Landeskrankenanstalten, Salzburg (Austria); Prof. Dr. H. E. Schröder, Med. Akademie, Dresden (Germany); Dr. F. Skrabl-Mocnik, Bolnica Celje, Celje (Slovenia); Prim. Prof. Dr. K. Steinbach, Wilhelminenspital, Vienna (Austria); Dr. R. Wolf, Herzkreislaufklinik, Bad Bevensen (Germany); PD Dr. M. Zehender, Universitätsklinikum, Freiburg (Germany); Dr. A. Zemva, Klinicni Center, Ljubljana (Slovenia).
- electrocardiogram, electrocardiographic
- ventricular premature beat
- ventricular tachyarrhythmia
- Received June 3, 1996.
- Revision received July 19, 1996.
- Accepted December 19, 1996.
- The American College of Cardiology
- Gettes L
- Arsenian MA
- ISIS Collaborative Group
- Gottlieb SS,
- Fisher MI,
- Krichten C,
- Patten RD,
- Pressel M,
- Brackett J
- Yusuf S,
- Teo K,
- Woods K
- Muerarity SA,
- Stark JL
- Sami M
- Zehender M,
- Hohnloser S,
- Geibel A,
- et al.
- Ferguson JJ
- ↵Singh SN, Fletcher RD, Fisher SG, et al., for the Survival Trial of Antiarrhythmic Therapy in Congestive Heart Failure. N Engl J Med 1995;333:77–82.