Abstract 17403: Serum Magnesium Levels and In-Hospital Mortality in Patients With Acute Myocardial Infarction
Introduction: Despite lack of data on the benefit of routine administration of magnesium (Mg) in patients with acute myocardial infarction (AMI), guidelines recommend keeping Mg levels > 2.0 mg/dL to prevent malignant arrhythmias. Moreover, optimal serum Mg levels during AMI and the association between Mg levels and mortality following AMI have not been adequately studied.
Hypothesis: Both low and high serum Mg levels are predictive of in-hospital mortality in patients with AMI.
Methods: We examined the association of Mg levels with in-hospital mortality in AMI patients at 77 US hospitals (2005-15) in the Cerner Health Facts database. Serum Mg levels were measured at the discretion of local clinicians. We used multivariable Cox regression analysis, stratified by hospital to evaluate the association of serum Mg (analyzed as a time-varying covariate) with in-hospital mortality, adjusting for covariates.
Results: Among 10,806 patients with AMI, 63% had serum Mg checked (hospital range 10% to 98%) during a median hospital stay of 3.1 days, with a median of 2 Mg checks per patient. When admission Mg levels were divided into quartiles, patients with admission Mg levels <1.8 and >2 had higher mortality than those in the middle quartiles (Fig A; p<0.001). After adjusting for covariates and allowing Mg levels to vary with time, the most recent Mg level was still associated with in-hospital mortality in a U-shaped manner (Fig B; p <0.001). Lowest hazard of death appeared to be with Mg level ~1.8 with increased hazard <1.7 and >1.9.
Conclusions: Serum Mg levels were not evaluated in over a third of AMI patients, with high variability across hospitals. Among Mg levels checked, the lowest mortality was observed in patients with Mg levels between 1.7 and 1.9 mg/dL. Our results support routine monitoring of serum Mg levels, suggest that the optimal range of Mg may be lower than previously defined in AMI patients (>2mg/dL), and highlight potential risks associated with both low and high Mg levels.
Author Disclosures: A. Shafiq: Research Grant; Modest; T-32 grant (Award Number T32HL110837). A. Goyal: None. P.G. Jones: None. S. Sahil: None. M. Hoffman: None. M. Qintar: Research Grant; Modest; T-32 grant (Award Number T32HL110837). D. Buchanan: None. M. Kosiborod: Research Grant; Modest; AstraZeneca, Gilead Sciences, Genentech, Sanofi. Consultant/Advisory Board; Modest; AstraZeneca, Sanofi, GSK, Ely Lilly, Takeda, Amgen, Boehringer-Ingelheim, ZS Pharma. S.V. Arnold: None.
- © 2016 by American Heart Association, Inc.