Predicting silent atrial fibrillation in the elderly – NOMED-AF study

Background Asymptomatic (“silent”) atrial fibrillation is common and associated with poor outcomes. It is important to determine the risk factors that predispose elderly individuals from the general population to atrial fibrillation (AF). However, population-based data for silent AF (SAF) are limited. Design First, to study the risk factors for symptomatic AF and SAF in an elderly (≥65 years) general population. Second, to develop a risk stratification model for predicting SAF. Methods Continuous ECG monitoring was performed for up to 30 days using a vest-based system in a cohort from NOMED-AF, a cross-sectional study based on a nationwide population sample. The independent risk factors for AF and SAF were determined using multiple logistic regression. ROC analysis was applied to validate developed risk stratification score. Results From the total cohort of 3014 subjects, AF was diagnosed in 680 individuals (mean age, 77.5±7.9; 50.1% men) with AF, and of these, 279 (41%) had SAF. Independent associations with an increased risk of AF were age, male gender, coronary heart disease, thyroid diseases, prior ischemic stroke or transient ischemic attack (ICS/TIA), diabetes, heart failure, chronic kidney disease (CKD), obesity (BMI>30) and NT-proBNP >125 ng/ml. Prior revascularization was negatively associated with risk of AF. The main risk factors for SAF were age, male gender, prior ICS/TIA, diabetes, heart failure, CKD and NT-proBNP >125 ng/ml. We developed a simple clinical risk scale (MR-DASH score) which had good prediction in the derivation cohort (AUC 0.726) and the validation cohort (AUC 0.730). Conclusions SAF is associated with various clinical risk factors in a population sample of individuals ≥65 years. Stratifying individuals from the general population according to their risk for SAF may be possible using the MR-DASH score, facilitating targeted screening programs of individuals with high risk of SAF Funding Acknowledgement Type of funding sources: Public Institution(s). Main funding source(s): National Centre for Research and Development