Prognostic relevance of magnesium alterations in patients with a myocardial infarction and left ventricular dysfunction: insights from the EPHESUS trial

Abstract

Aims Magnesium changes are common in myocardial infarction (MI) complicated with left ventricular systolic dysfunction (LVSD) and/or heart failure (HF). The relation between serum magnesium and clinical outcomes is insufficiently elucidated in this population. Methods and results The EPHESUS trial randomized 6632 patients to either eplerenone or placebo. Hypomagnesemia and hypermagnesemia were defined as a serum magnesium 1.10 mmol/L, respectively. Linear mixed models and time-dependent Cox regression analysis were used to determine the effect of eplerenone on magnesium changes and the prognostic importance of magnesium. The co-primary outcomes were all-cause mortality and a composite of cardiovascular (CV) mortality and CV hospitalization. A total of 5371 patients had a post-baseline magnesium measurement. At baseline, 231 (4.3%) patients had hypomagnesemia and 271 (5.0%) patients had hypermagnesemia. During a median follow-up of 16 months, 682 (13%) developed hypomagnesemia and 512 (9.5%) hypermagnesemia. Eplerenone treatment did not result in a different magnesium level during follow-up (P = 0.14). After covariate adjustment hypo- and hypermagnesemia were not associated with a higher risk of CV events. Magnesium levels did not modulate the effect of a high potassium (>5 mmol/L) or low potassium (<4 mmol/L) on the clinical outcome. Baseline magnesium levels did not influence the treatment effect of eplerenone (P-interaction > 0.1 for all primary and secondary endpoints). Conclusion In patients with MI complicated by LVSD or HF, magnesium alterations were not associated with clinical outcomes nor did they influence the effect of eplerenone. Serum magnesium did not modulate the effect of potassium changes on clinical outcome or the treatment effect of eplerenone. ClinicalTrials.gov identifier NCT00232180.