A longitudinal population-based study of prothrombotic factors in elderly subjects with atrial fibrillation: the Rotterdam Study 1990-1999

Background: A prothrombotic or hypercoagulable state in atrial fibrillation may contribute to stroke and thromboembolism. Results of longitudinal population-based studies in elderly people with atrial fibrillation are not yet available. Methods: In the Rotterdam Study, a population-based prospective cohort study, 162 participants with atrial fibrillation at baseline, aged 55 years and over, were matched for age and gender with 324 people in sinus rhythm. Associations were examined between three coagulation factors and the risk of total and cardiac mortality and stroke. Hazard rate ratios were calculated with 95% confidence intervals using Cox's proportional hazards model, adjusted for potential confounders. Results: Plasma von Willebrand factor was, age- and gender-adjusted, associated with cardiac mortality in the total population (relative risk 1.16; 1.06-1.27, per 10 IU dL(-1) increase), but statistical significance was lost after additional adjustments. A strong association (1.27; 1.08-1.50, per 5-unit increase) was found between soluble P-selectin (sP-sel) and cardiac mortality in atrial fibrillation patients but not in participants in sinus rhythm. Furthermore, the expected association between fibrinogen and cardiac mortality was observed only in those in sinus rhythm (2.60; 1.69-4.01, per unit increase), and not in atrial fibrillation. No associations were found between coagulation factors and stroke. Conclusions: In this population-based study, plasma levels of sP-sel predicted clinical adverse outcomes in atrial fibrillation, suggesting a role of platelets in the prothrombotic state associated with atrial fibrillation. Fibrinogen was a risk factor of cardiac and all-cause mortality in sinus rhythm, but not in atrial fibrillation.