Predicting occult paroxysmal atrial fibrillation (AF) in transient ischemic attack (TIA) or strokes

RelevanceAF accounts for 75,000 cases of strokes per year, yet 40% of strokes may have no apparent etiology. Etiology of these strokes may be occult PAF. The laboratory and echocardiography markers may help identify those at risk of PAF.PurposeParoxysmal AF (PAF) is a significant risk factor for TIA/stroke. However, routine Holters/ECG often fail to detect PAF. Our aim was to evaluate laboratory and echocardiographic parameters to predict PAF.Participants428 patients were enrolled, 220 males and 208 females, 51% and 49%, respectively, a mean age of 72.3 years. PAF present group 68 patients (16%) versus PAF absent group 360 patients (84%).MethodThe 24h-Holters recorded for evaluation of TIA/strokes were analyzed for PAF. Patients were divided into PAF present or absent groups.AnalysisMultivariate regression analysis was used to investigate BNP (Brain Natriuretic Peptide), D-dimer, mitral regurgitation (MR), left atrial size (LA), left ventricular hypertrophy (LVH) and diastolic dysfunction.ResultsBNP, MR, LA size, LVH, diastolic dysfunction were significantly higher in patients showing PAF on their Holters than those without PAF. Multivariate logistic regression analysis demonstrated BNP 400pg/ml (OR, 14.8; 95% CI 6.5–45, P<0.01), MR (OR, 8.1; 95% CI 3.12–26.2, P<0.001), LA size 4.0cm (OR, 5.2; 95% CI 2.01–14.6, P<0.002), LVH 1.2cm (OR, 4.9; 95% CI 1.2–6.88, P<0.001), diastolic dysfunction (OR, 6.7; 95% CI 2.3–27.2, P