Abstract 18762: The Association Between Atrial Fibrillation and Sudden Cardiac Death is Confounded by Clinical History of Congestive Heart Failure
Introduction: Atrial fibrillation / flutter (AF) has been associated with overall cardiac mortality, and recent studies have reported an association with sudden cardiac death (SCD). Since AF and SCD share many of the same risk factors, we hypothesized that this association could potentially be explained by other clinical predictors, especially a clinical diagnosis of congestive heart failure (CHF).
Methods: In an ongoing prospective case-control study of SCD in a metro region of the Northwestern US (pop. approx. 1 million), cases of SCD were ascertained from the region’s EMS system, medical examiner, and hospitals and adjudicated by systematic review of arrest circumstances and medical records. SCD cases age ≥18 (2002-2012) with physician records and a cardiac evaluation by ECG or echocardiogram were compared to control subjects with coronary artery disease (CAD) from the same geographic region. AF was defined by physician-documented medical history or from a pre-arrest ECG or echocardiogram.
Results: Case subjects (n=935, age 71 ± 13 yrs, 64% male) were more likely than control subjects (n=679, age 66 ± 12 yrs, 66% male) to have a history of myocardial infarction (41% vs. 34%, p=0.004), CHF) (45% vs. 18%, p<0.0001), cerebrovascular accident (CVA) (22% vs. 10%, p<0.0001), and diabetes (41% vs. 32%, p=0.0004), but less likely to have documented hyperlipidemia (54% vs. 82%, p<0.0001). Among all subjects, cases were significantly more likely to have a history of AF than controls (29% vs. 18%, p<0.0001). However, AF was strongly associated with a history of CHF: AF was prevalent in 44% of subjects with CHF and 15% of subjects without CHF (p<0.0001). In multivariable models, AF was significantly associated with SCD when adjusting for age, gender, history of MI, hypertension, CVA, diabetes, and dyslipidemia (p=0.01), but when CHF was added to the model, the AF-SCD association was no longer significant (OR 1.05, 95% CI 0.8 - 1.4, p=0.74). In this model, CHF remained a significant predictor of SCD risk (OR 3.5, 95% CI 2.6 - 4.5).
Conclusions: In this population-based study, the AF-SCD association appears to be driven by a clinical history of CHF. These findings have potential implications for clinical and mechanistic approaches to AF and SCD.
- © 2013 by American Heart Association, Inc.