关键词: 脑血管障碍, 意外跌倒, 预测模型

Abstract: Objective To investigate the factors influencing the fall risk of independently ambulatory patients with cerebral small vessel disease(CSVD) based on neuroimaging characteristics and clinical factors, to develop a prediction model and to validate its efficacy. 
Methods In total, 315 independently ambulatory patients with CSVD were selected in the Third Hospital of Hebei Medical University from Sept. 2021 to Sept. 2024 and divided into modelling group (n=196) and validation group (n=119) according to the 6∶4 principle. The timed up and go test was used to assess the fall risk of CSVD patients who could walk independently. Univariate and multivariate analyses was used to analyze independent risk factors for falls in independently ambulatory CSVD patients, a fall risk prediction model was constructed and a nomogram was plotted. The area under the receiver operating characteristic (ROC) curve (AUC) and calibration curve were used to evaluate the differentiation and calibration degree of the model in the modeling population and the verification population respectively. 
Results Compared with the group without fall risk, the proportion of patients with advanced age, hypertension, history of fracture, cognitive impairment, moderate-to-severe white matter hyperintensity (WMH), moderate-to-severe perivascular spaces (EPVS), and lacunes was higher in the group with fall risk, showing significant differences (P＜0.05). Multivariate Logistic regression analysis showed that age (95%CI: 1.356-3.256), hypertension (95%CI: 1.119-6.682), cognitive impairment (95%CI: 1.146-7.423), moderate to severe WMH (95%CI: 1.487-8.363) and lacune (95%CI: 1.965-9.636) were independent risk factors for falls in independently ambulatory patients with CSVD (P＜0.05). Based on the above influencing factors, a nomogram model of the fall risk of independently ambulatory patients with CSVD was constructed. The AUC of the modeling population and the verification population was 0.855 and 0.921 respectively, indicating a high degree of model differentiation. The calibration curve showed that the prediction model was in good agreement with the actual observation results. 
Conclusion Advanced age, hypertension, moderate-to-severe WMH, lacune, and cognitive impairment are independent risk factors for falls in independently ambulatory CSVD patients, and the clinical prediction model developed based on this study can better predict fall risk in independently ambulatory CSVD patients. 

Key words: cerebrovascular disorders, accidental falls, prediction model