Abstract 14678: Lower Serum Eicosapentaenoic Acid Predicts the Incidence of Paroxysmal Atrial Fibrillation During Hospital Stay in Patients With Acute Coronary Syndrome
Background: Emerging evidences have suggested the relationship between the lower serum omega-3 fatty acid fractions and the incidence of atrial fibrillation (A-fib). However, it is still elusive whether the examination of serum fatty acid fractions would be a practical predictor of A-fib. We studied serum fatty acid fractions in patients with acute coronary syndrome (ACS) and investigated the impact of aberrant fatty acid fractions on the incidence of A-fib during hospital stay in ACS.
Methods: In consecutive 119 ACS patients, who admitted our hospital from December 2011 to January 2013, the serum lipid profiles which include eicosapentaenoic acid (EPA), docosahexanoic acid (DHA), arachidonic acid (AA), and dihomo-gamma-linolenic acid (DHLA), were measured within 7 days of admission. Patients who have already taken the supplement of ethyl icosapentate on admission were excluded from study entry. In patients with myocardial infarction (n=100) or unstable angina (n=19), the coronary angiography and/or coronary intervention were performed urgently. Incidence of A-fib during hospital stay was assessed with the 24 hour electrocardiogram monitoring. Three patients (2.5%) had previous history of A-fib.
Results: (1) A total of 24 (20.2%) patients were detected as A-fib during hospitalization. Although female (45.8% vs. 22.1% in non-A-fib patients, p=0.037) and older age (74.8 years old vs. 68.0 years old of non-A-fib patients, p=0.007) were significantly different, other patient characteristics including blood pressure, BMI, and previous history of ischemic heart disease or chronic kidney disease were not different between A-fib and non-A-fib patients. (2) A-fib patients exhibited lower serum EPA fraction (38.9 μg/ml vs. 55.1 μg/ml of non-A-fib patients, p<0.001) and lower EPA/AA ratio (0.29 vs. 0.37, p= 0.005), however DHA, AA, and DHLA were not different between A-fib and non-A-fib patients. (3) A logistic regression analysis also revealed that older age (p=0.013), lower EPA (p=0.005), and lower EPA/AA ratio (p=0.025) had significant correlation with A-fib.
Conclusion: Our results suggest that lower EPA and/or EPA/AA ratio may predict the incidence of A-fib during hospitalization in ACS patients.
- © 2013 by American Heart Association, Inc.