Abstract 47: Atrial Fibrillation is Associated With Increased Risk of Incident Venous Thromboembolism: The Atherosclerosis Risk in Communities Study

Abstract

Background: It is well-established that atrial fibrillation (AF) is associated with thrombus formation in the left atrium, which can lead to ischemic stroke. Case reports, autopsies, and transesophageal echo data have indicated that clot formation also occurs in the right atrium (i.e. right-side intracardiac thrombosis) of AF patients, which could lead to pulmonary embolism (PE). However, it is unclear whether this occurrence is common. Objective: Test the hypotheses that individuals with incident AF are at elevated risk of developing venous thromboembolism (VTE), and that the association will be stronger for those presenting with PE alone versus PE and deep vein thrombosis (DVT) or DVT alone. Methods: A total of 15,205 Atherosclerosis Risk in Communities (ARIC) study participants, aged 45-64 years, were followed from baseline (1987-1989) to 2011 for incidence of AF and VTE (median follow-up 19.8 years). Incident AF and VTE events were identified via active surveillance and defined by relevant hospital discharge ICD codes. VTE events were validated by medical record review. Multivariable-adjusted Cox proportional hazards regression models were used, with AF modeled as a time-dependent covariate. We also evaluated separately risk of PE without evidence of DVT, DVT without PE, and events presenting with both PE and DVT. Results: At baseline participants were on average 54 years old, 55% female and 26% black. In the absence of AF there were 678 VTE events, for an incidence rate of 2.6 per 1000 person-years. After an AF diagnosis there were 77 events, with an incidence rate of 7.1 per 1000 person-years. In multivariable-adjusted models, having AF (versus no AF) was associated with a greater risk of incident VTE; the HR (95% CI) was 2.10 (1.65-2.68) after adjustment for demographics, 1.82 (1.42-2.32) additionally accounting for numerous AF and VTE risk factors, and 1.97 (1.53-2.53) after further adjusting for time-dependent anticoagulant use. When we restricted to PE events without evidence of DVT there were 188 events in total, of which 19 occurred following a diagnosis of AF. The HR for AF (versus no AF) was 1.53 (0.92-2.56) in fully adjusted models. For DVT alone there were 384 events in total, of which 48 occurred after AF diagnosis; the HR for AF was 2.43 (1.77-3.33). Among the 116 events presenting with both DVT and PE, 10 occurred after AF diagnosis, and the HR for AF was 1.36 (0.67-2.75). Conclusions: Diagnosis with AF was associated with a nearly 2-fold increased risk of incident VTE. The association was not stronger when isolated to those with PE without DVT, suggesting that higher risk of VTE among AF patients may be due to either the coagulation abnormalities that accompany AF, or shared risk factors that were not fully accounted for in this analysis.