Brain natriuretic peptide as a predictor of delayed atrial fibrillation after ischaemic stroke and transient ischaemic attack


Dr. Y Okada, Department of Neurology, Matsuyama red-cross hospital, 1 Bunkyo-cho, Matsuyama City, Ehime, 790-8524, Japan (tel.: +81 89 924 1111; fax: +81 89 958 7081; e-mail:


Background and purpose:  We investigated whether the brain natriuretic peptide (BNP) level can serve as a predictive biological marker of delayed atrial fibrillation (AF).

Methods:  Two hundred and thirty seven consecutive patients admitted to our institution with acute ischaemic stroke or transient ischaemic attack (TIA) within 24 h of onset were enrolled. The patients were classified according to the presence or absence of AF upon admission [AF and sinus rhythm (SR) groups]. The SR group was subdivided based on the development of AF after admission (new- and non-AF groups). We compared the characteristics between the AF and SR groups, and between the new- and non-AF groups. The factors associated with new-AF were investigated by multivariate logistic regression analysis.

Results:  Amongst the enrolled patients, 72 (30.4%) had AF upon admission (AF group), and 13 (5.5%) developed AF thereafter (new-AF group). The plasma BNP level was significantly higher in the AF, than in the SR group (401.7 vs. 92.1 pg/ml, P < 0.001). Moreover, the plasma BNP level was significantly higher in the new-, than in the non-AF group (184.7 vs. 84.1 pg/ml, P < 0.001). The optimal cutoff BNP level required to distinguish new-, from non-AF groups was 85.0 pg/ml, and the sensitivity and specificity was 83.3% and 76.2%, respectively. On multivariate logistic regression analysis, plasma BNP level >85.0 pg/ml (odds ratio, 7.20; 95% confidence interval, 1.71 to 30.43, = 0.007) was an independent factor associated with new-AF.

Conclusion:  High plasma BNP level should be a strong predictor of delayed AF after ischaemic stroke or TIA.