The prognostic impact of magnesium in acute heart failure is different according to the presence of diabetes mellitus

Cidade-Rodrigues, C; Cunha, FM; Elias, C; Carreira, M; Barroso, I; Bettencourt, P; Lourenço, P

http://hdl.handle.net/10400.26/44837

Use this identifier to reference this record.

Name:	Description:	Size:	Format:
pj9-7-e197.pdf		410.41 KB	Adobe PDF	Download

Send Feedback

Authors

Abstract(s)

Background: Hypermagnesemia predicts mortality in chronic heart failure (HF); however, in acute HF, magnesium does not seem to be outcome-associated. Diabetes mellitus (DM) frequently associates with altered magnesium status. We hypothesized that DM might influence the prognostic impact of magnesium in acute HF. Methods: This is a retrospective cohort study of hospitalized patients with acute HF. Patients without data on admission serum magnesium were excluded. Follow-up: 1 year from hospital admission. Primary end point: all-cause mortality. Patients were divided according to median serum magnesium (1.64 mEq/L). The Kaplan-Meier survival method was used to determine survival curves according to magnesium levels. The analysis was stratified according to the presence of DM. A multivariable Cox regression analysis was used to study the prognostic impact of magnesium. Results: We studied 606 patients. The mean age was 76 ± 12 years, 44.1% were male, 50.7% had DM, and 232 (38.3%) died during follow-up. Median magnesium was 1.64 (1.48-1.79) mEq/L. Patients with magnesium ≥1.64 mEq/L had higher 1-year mortality [141 (46.4%) vs 91 (30.1%), P < .001]. After adjustments for age, sex, history of atrial fibrillation, systolic blood pressure, heart rate, ischemic etiology, B-type natriuretic peptide, estimated glomerular filtration rate, alcohol consumption, antihyperglycaemic agents or glycated hemoglobin, admission glycemia, New York Heart Association class IV, and severe left ventricle systolic dysfunction, serum magnesium ≥1.64 mEq/L was associated with higher mortality only in patients with DM: HR 1.89 (95% confidence interval: 1.19-3.00), P = .007, and 1.27 (95% confidence interval: 0.83-1.94) and P = .26 for non-DM patients. The results were similar if magnesium was analyzed as a continuous variable. Per 0.1 mEq/L increase in magnesium levels, patients with DM had 13% increased risk of 1-year mortality. Conclusions: Higher magnesium levels were associated with worse prognosis only in HF patients with DM.