PURPOSE: The present study investigated clinical characteristics including plasma brain natriuretic peptide (BNP) among transient ischemic attack (TIA) and stroke patients with atrial fibrillation (AF). METHODS: We prospectively enrolled 227 consecutive patients with AF within 24 h of onset of TIA or stroke, and plasma BNP was measured on admission. Patients were divided into 2 groups: TIA and stroke groups. The factors associated with TIA were investigated by multivariate logistic regression analysis. RESULTS: 21 patients (9.3%) were diagnosed with TIA, and 206 patients (90.7%) with stroke. The plasma BNP level of the TIA group was significantly lower than that of the stroke group [median (interquartile range) 86.5 (72.7-189.0) pg/ml vs. 269.0 (146.0-432.0) pg/ml, p = 0.0002]. The optimal cutoff level, sensitivity, and specificity of BNP levels to distinguish the TIA group from the stroke group were 120 pg/ml, 79.1 and 66.7%, respectively. Multivariate logistic regression analysis demonstrated that pre-admission warfarin use (OR 3.7; 95% CI 1.178-11.570, p = 0.0250), glucose of ≤120 mg/dl (OR 5.1; 95% CI 1.629-16.238, p = 0.0052) and a plasma BNP of ≤120 pg/ml (OR 6.1; 95% CI 1.967-18.931, p = 0.0017) were independently associated with TIA. CONCLUSIONS: In AF patients, the BNP value on admission is lower in those with TIA than in those with stroke. Thus, cardiac function may be associated with neurological severity at the onset of TIA and stroke.
PURPOSE: The present study investigated clinical characteristics including plasma brain natriuretic peptide (BNP) among transient ischemic attack (TIA) and strokepatients with atrial fibrillation (AF). METHODS: We prospectively enrolled 227 consecutive patients with AF within 24 h of onset of TIA or stroke, and plasma BNP was measured on admission. Patients were divided into 2 groups: TIA and stroke groups. The factors associated with TIA were investigated by multivariate logistic regression analysis. RESULTS: 21 patients (9.3%) were diagnosed with TIA, and 206 patients (90.7%) with stroke. The plasma BNP level of the TIA group was significantly lower than that of the stroke group [median (interquartile range) 86.5 (72.7-189.0) pg/ml vs. 269.0 (146.0-432.0) pg/ml, p = 0.0002]. The optimal cutoff level, sensitivity, and specificity of BNP levels to distinguish the TIA group from the stroke group were 120 pg/ml, 79.1 and 66.7%, respectively. Multivariate logistic regression analysis demonstrated that pre-admission warfarin use (OR 3.7; 95% CI 1.178-11.570, p = 0.0250), glucose of ≤120 mg/dl (OR 5.1; 95% CI 1.629-16.238, p = 0.0052) and a plasma BNP of ≤120 pg/ml (OR 6.1; 95% CI 1.967-18.931, p = 0.0017) were independently associated with TIA. CONCLUSIONS: In AFpatients, the BNP value on admission is lower in those with TIA than in those with stroke. Thus, cardiac function may be associated with neurological severity at the onset of TIA and stroke.