Factors associated with ischemic stroke during aspirin therapy in atrial fibrillation: Analysis of 2012 participants in the SPAF I-III clinical trials

Background and Purpose - Nonvalvular atrial fibrillation (AF) is a strong, independent risk factor for stroke, but the absolute rate of stroke varies widely among AF patients, importantly influencing the potential benefit of antithrombotic prophylaxis. We explore factors associated with ischemic stroke in AF patients taking aspirin. Methods - We performed multivariate logistic regression analysis of 2012 participants given aspirin alone or in combination with low, inefficacious doses of warfarin in the Stroke Prevention in Atrial Fibrillation I-III trials followed for a mean of 2.0 years, during which 130 ischemic strokes were observed. Results - Age (relative risk [RR] = 1.8 per decade, P<0.001), female sex (RR = 1.6, P=0.01), history of hypertension (RR=2.0, P<0.001), systolic blood pressure > 160 mm Hg (RR=2.3, P<0.001), and prior stroke or transient ischemic attack (RR=2.9, P<0.001) were independently associated with increased stroke risk. Regular consumption of ≥14 alcohol-containing drinks per week was associated with reduced stroke risk (adjusted RR=0.4, P=0.04). Among SPAF III participants, estrogen hormone replacement therapy was associated with a higher risk of ischemic stroke (adjusted RR=3.2, P=0.007). With the use of these variables, a risk stratification scheme for primary prevention separated participants into those with high (7.1%/y, 22% of the cohort), moderate (2.6%/y, 37% of the cohort), and low (0.9%/y, 41% of the cohort) rates of stroke. Ischemic strokes in low-risk participants were less often disabling (P<0.001). Conclusions - Patients with AF who have high and low rates of stroke during treatment with aspirin can be identified. However, validation of our risk stratification scheme is necessary before it can be applied with confidence to clinical management. Postmenopausal estrogen replacement therapy and moderate alcohol consumption may additionally modify the risk of stroke in AF, but these findings require confirmation.