Multiple regression analysis of the risk factors to predict different recurrent stroke types after initial ischemic stroke

Objective To respectively analyze the patterns and possible predictors of recurrent strokes among patients with initial ischemic stroke. Methods Three hundred and sixty-one patients with recurrent strokes (acute ischemic stroke or intracerebral hemorrhage) after initial ischemic strokes were collected from Jan 2004 to Dec 2009. The data about conventional risk factors such as smoking, heavy alcohol drinking, hypertension, diabetes, hyperlipidemia, heart diseases, head trauma, migraine, family history of cardiovascular disease, and the use of preventive medications were collected and analyzed among patients with different types of recurrent strokes. Results Patients (n=361) were divided into ischemic stroke group (n=321) and hemorrhagic stroke group (n=40) according to the recurrent stroke type. The ischemic stroke group was further divided into the anterior circulation stroke subgroup (n=234) , the posterior circulation stroke subgroup (n=75) and watershed cerebral infarction or multiple infarction subgroup (n=12) . Multivariate logistic regression analysis revealed that older age at initial stroke onset (OR=1.036,95%CI 1.006—1.067，P=0.02) and hyperlipidemia (OR=2.253,95%CI 1.092—4.647，P=0.028) were both the independent risk factors for the recurrent ischemic stroke. Comparing the subgroups, multivariate logistic regression analysis showed that atrial fibrillation (OR=4.217,95%CI 1.489—11.942，P=0.007) was the independent risk factor for the recurrent ischemic stroke in the posterior circulation territory. Conclusion Aging and hyperlipidemia are possible predictors of recurrent ischemic stroke after the initial ischemic stroke which would be useful for individualized secondary prevention of stroke. Key words: Brain ischemia; Stroke; Recurrence; Risk factors