Prediction Factors of Recurrent Stroke among Chinese Adults Using Retinal Vasculature Characteristics

Abstract

Objective: To identify the characteristics of retinal vessels that are associated with recurrent cerebral infarction and to establish predictive model for recurrent cerebral infarction in further study.

Methods: Up to 196 patients with cerebral infarction were included in our study, including 137 patients with new onset cerebral infarction and 59 patients with recurrent cerebral infarction. Retinal vessel characteristics of all patients were analyzed and reported by an automated retinal image analysis system. Retinal and clinical characteristics were compared between recurrent and new onset subjects. Multivariate logistic regression was used to identify risk factors associated with recurrent ischemic stroke within the year after an initial stroke onset.

Results: Every unit increase in the vessel caliber measures of central retinal artery and vein equivalent was found to give coronary heart diseases an additional 4.28 times (95% confidence interval 1.204-15.215) the risk of stroke recurrent whereas every increased unit of arterial and venous angles would be associated with 3.9 and 13.7 times increase in such risk among atrial fibrillation subjects. Measures such as bifurcation coefficient of arterioles showed opposite effects when interacted with different clinical factors. Negative integrations were found between venous asymmetry index and various factors such as atrial fibrillation (odds ratio .73, 95% confidence interval .61-.88), coronary heart diseases (odds ratio .79, 95% confidence interval .70-.89), and bifurcation coefficient of arterioles (odds ratio .38, 95% confidence interval .21-.72), indicating a decreased recurrent risk among these subjects when venous asymmetry level increases.

Conclusions: Retinal vessel characteristics interacted with each other as well as traditional clinical risk factors in affecting the risk of stroke recurrence.

Keywords: Cerebral infarction; recurrent infarction; retinal vessels; risk prediction.