Objective: To screen the risk factors of white matter lesion (WML) and explore therelationship between WML with stroke and dementia.

Methods: In this study, a total of 260 patients with WML and 60 without WML underwentMRl were collected. The patients with WML were divided into 3 grades according toFazekas scale. Gender, age,education level, hypertension,diabetes, hyperlipemia,hyperhomocysteinemia, overweight, smoking and drinking history, and stroke history ofall enrolled patients were collected. Multiple ordered logistic regression analysis was usedto identify the risk factors of WML. Compare whether the degree of WMl in strokepatients is more severe than in non-stroke patients. MMSE scale was used to score theenrolled patients, analyzed the relationship between WML and cognition.

Results: Age, history of hypertension, hyperhomocysteinemia and smoking were correlatedwith the severity of WML, The Fazekas classification of stroke patients was statisticallydifferent from that of non-stroke patients(2=-10.305,P=0.000). With the increase ofFazekas grading, the MMSE score ofpatients decreased(r_s-0.511,P=0.000).

Conclusion: Age, hypertension, hyperhomocysteinemia and smoking are independent riskfactors for WML. WM may be a predictive signal of stroke. The more severe the degree ofWML, the higher the risk of dementia in patients.