Atrial fibrillation after stroke in the elderly.

To examine the relationship between atrial fibrillation and mortality after stroke, we studied 186 men and 167 women from the Waikato Stroke Registry whose mean age was 75.2 +/- 7.5 years. Twenty-three percent (82 of 353) had atrial fibrillation or flutter on their admission electrocardiogram. This group differed significantly from that with sinus rhythm in three respects: 1) They were older (p less than 0.01); 2) they had more severe current stroke deficit as evidenced by lower limb power (p less than 0.05) and Mini-Mental State Score (p less than 0.001), higher incidence of homonomous hemianopia (p less than 0.05), and lower incidence of lacunar syndrome stroke (p less than 0.001); and 3) they had a significantly higher incidence of cardiomegaly and congestive heart failure (p less than 0.01). Functional outcome was insignificantly better in the group with sinus rhythm. During a mean follow-up period of 18 months, mortality was significantly higher in the group with atrial fibrillation (p = 0.001). Proportional hazards modeling, however, showed that the apparently poorer survival in those patients with atrial fibrillation could be explained by factors other than cardiac rhythm, such as age, Mini-Mental State Score, level of consciousness, and interstitial edema on admission chest radiograph. Thus, atrial fibrillation was not an independent predictor of survival after stroke.