As the incidence of atrial fibrillation (AF) increases and the US population diversifies, the need for a clinically available and applicable prediction tool is indeed warranted. To that end, Schnabel et al1 add an important piece to this literature in validating the Framingham-derived AF prediction algorithm across 2 racially diverse populations. As described by Marcus2 in his insightful commentary on this study, several clinical characteristics including family history of AF and B-type natriuretic peptide levels (which have been shown to improve risk stratification for the prediction of AF development), would improve the discriminatory power of this prediction algorithm. Another important variable that should also be considered is a patient history of obstructive sleep apnea or a validated clinical or diagnostic correlate (ie, the Epworth sleepiness scale3 or polysomnography). With levels of obesity and AF sharply rising, obstructive sleep apnea is associated with poorer secondary prevention of AF with less freedom from AF after AF catheter ablation4 and is thus an important factor to consider in the development and prediction of incident AF.
Mazurek JA. Improving the Prediction of Incident Atrial Fibrillation. Arch Intern Med. 2011;171(12):1125. doi:10.1001/archinternmed.2011.266