Atrial fibrillation associated with higher rates of sudden cardiac death

Researchers analyzed two large cohorts of the general population and found a strong association between incident atrial fibrillation (AF) and sudden cardiac death (SCD) and between incident AF and non-sudden cardiac death (NSCD). But editorial writers offered several caveats about the analysis.

The results were published online Nov. 26 in Archives of Internal Medicine.

AF is one of the most common cardiac complaints, and sudden cardiac death accounts for about half of the mortality from cardiovascular disease, with more than 300,000 deaths annually in the U.S. (Circulation 2010; 122[22]:2335-2348).

To investigate the hypothesis that incident AF is associated with increased rates of SCD in the general population, Lin Y. Chen, MD, of the University of Minnesota Medical School in Minneapolis, and colleagues analyzed two study populations: the Artherosclerosis Risk in the Community (ARIC) study, a cohort of 15,792 people aged 45-64 years at time of enrollment from communities in Maryland, Minnesota, Mississippi and North Carolina; and the Cardiovascular Health Study (CHS), a cohort of 5,201 participants aged 65 and older at recruitment from four centers in California, Maryland, North Carolina and Pennsylvania.

Baseline in the ARIC study was 1987-1989, and participants were interviewed every year thereafter. There were two baselines in the CHS study. Most recruitment occurred between 1989-1990, but between 1990 and 1994 black participants were added in an attempt to increase minority representation in the study cohort.

Lin et al analyzed data from the baseline of ARIC through Dec. 31, 2001, and both baselines of CHS through Dec. 31, 2006. They identified incidents of AF through electrocardiograms and hospital discharge records. Both studies defined SCD as “a sudden pulseless condition presumed to be due to a ventricular tachyarrhythmia in a previously stable individual without evidence of a non-cardiac cause of cardiac arrest.” The researchers counted SCD only if the participant did not have a life-threatening non-cardiac comorbidity and was not under hospice care.

The authors ran two models, one of which adjusted for age, race, gender and geographic location. The second model adjusted for 11 baseline covariates, among them smoking status, diabetes mellitus, body mass index and certain prescription drugs. 

Among the ARIC study group, participants who had experienced an incident AF were at significantly higher risk of SCD than those who had not, with a hazard ratio (HR) of 5.4. When adjusted for cardiac risk factors, the increased risk was still apparent (HR 3.26).  The risks were similar for women and men, (HR 4.12 vs. HR 3.12), but higher for blacks than for non-blacks (HR 5.77 vs. HR 2.49). A sensitivity analysis restricting the population to definite cases of SCD showed that an incident AF doubled the risk of SCD. A sensitivity analysis to control for confounding by left-ventricular systolic dysfunction by adjusting for left ventricular fractional shortening on two dimensional echocardiogram found that the HR for SCD was 13.59 and for NSCD it was 10.74.

Among the CHS study participants, incident of AF carried an HR 2.14 for SCD and 3.10 for NSCD. The risk was similar for women and men (HR 2.49 vs. HR 1.99), and researchers found no significant race-based differences. Sensitivity analyses found the risk doubled for participants who experienced definite SCD (HR 2.25) and adjusting by left-ventricular ejection fraction on two dimensional echocardiogram showed an HR of 2.07 for SCD and 2.92 for NSCD.

Overall, the researchers showed the hazard risk for participants who experienced incident AF was 2.47 for SCD and 2.98 for NSCD. “To the best of our knowledge, the present study is the first to demonstrate that incident AF is associated with an increased risk of SCD in two independent population-based cohorts. This association was observed in men and women, as well as in blacks and non-blacks,” the authors wrote.

As strengths of their study, the authors pointed to the large population-based cohorts and the long-term follow-up, and the fact that their analysis produced similar results in both cohorts. As limitations the authors noted that they could not include asymptomatic AF or AF managed in an outpatient setting, could not adjust for left-ventricular systolic function in the entire ARIC cohort, and the potential for residual confounders.

In an accompanying editorial, Kyndaron Reinier, PhD, and Sumeet Chugh, MD, of Cedars Sinai Medical Center in Los Angeles, questioned whether the authors demonstrated an association between AF and SCD that will have relevance to clinical practice. They noted that the study demonstrated an association between AF and both SCD and NSCD, which may be due to shared risk factors.

The editorialists acknowledged the study authors attempted to tease out distinctions by their sensitivity analyses, but, “It is difficult to account for diastolic dysfunction or heart failure with preserved ejection fraction, conditions that are as likely (or more likely) to be associated with AF than shortening or decreases in shortening fraction or left ventricular ejection fraction. Therefore, a causal association is difficult to determine.”

The editorial also noted the many cases of undetected, asymptomatic AF, and the fact that cerebrovascular accidents can present as sudden, unexplained death. Therefore, it is possible that the researchers did not capture all of the AF, and may have overstated the SCD.

However, Reinier and Chugh agreed that the results are worthy of further study, and pointed out the need to identify the mechanism of SCD and stratification of risk among vulnerable populations.