Analysis Of Risk Factors And Establishment Of Clinical Prediction Model For Cerebral Small Vessel Disease

Objective(s):The aim of this study was to explore and analyse the influential factors of small cerebral vascular disease,construct the nomogram prediction model for CSVD,and test the model’s predictive capability.Methods:A total of 191 patients admitted to the Department of Neurology of The Second Affiliated of Kunming Medical University from January 2021 to October2022 were collected as the research objects.According to whether they had CSVD,they were divided into CSVD group and non-CSVD group.The basic information,previous history,laboratory examination,head magnetic resonance examination,carotid duplex ultrasound and ambulatory blood pressure examination results of the two groups of patients were collected.Using the SPSS25.0 software to conduct data analysis.x~2test,t test,and Mami-Whitney U test were used for comparison between groups,univariate and multivariate Logistic regression analysis was used to screen out the risk factors and independent risk factors of CSVD.The nomogram function of R(R 4.2.0)software was used to draw the nomogram risk prediction model.Using the Bootstrap method for internal validation of the model,the Calibration plot was drawn,and drawing the ROC curve to calculate the area under the curve.Results:1.In this study there were a total of 191 hospitalized patients were enrolled.According to whether they had CSVD,they were divided into CSVD group and non-CSVD group.Fibrinogen(Fib)(P=0.034),alcohol consumption(P=0.03),mean arterial pressure(P=0.024),Hyperhomocysteinemia(HHcy)(P=0.024),serum phosphorus(P=0.021),Lipoprotein(a)(Lp-A)(P=0.016),Apolipoprotein A1(apo A-1)(P=0.008),alkaline phosphatase(ALP)(P=0.005),standard deviation systolic blood pressure(P=0.005),gender(P=0.003),neutrophil-to-lymphocyte ratio(NLR)(P=0.001),age(P=0.000),hypertension(P=0.000),proteinuria(P=0.000),carotid atherosclerosis(CAS)(P=0.000)and glycated hemoglobin A1c(Hb Alc)(P=0.000)were statistically significant differences between the two groups(P<0.05),which may be risk factors for CSVD.2.Univariate Logistic regression analysis of the above factors showed that NLR(P=0.032),standard deviation systolic blood pressure(P=0.031),alcohol consumption(P=0.031),HHcy(P=0.025),Apo A-1(P=0.016),Fib(P=0.011),ALP(P=0.007),gender(P=0.003),Hb Alc(P=0.003),proteinuria(P=0.000),hypertension(P=0.000),age(P=0.000)and CAS(P=0.000)were risk factors for CSVD.Logistic multiple factor regression showed that age(P=0.002),hypertension(P=0.001),proteinuria(P=0.000)and CAS(P=0.000)were independent risk factors for CSVD.3.According to the independent risk factors,the nomogram risk prediction model of CSVD was drawn.The calibration curve was drawn and it was found that the actual result curve was close to the predicted result curve,indicating that the calibration of the model was good.The area under curve(AUC)of ROC was 0.883(P=0.000,CI=0.836-0.931),illustrating that the nomogram prediction model occupied a strong predication performance.Conclusion(s):1.Age,hypertension,proteinuria and CAS are independent risk factors for CSVD.2.Gender,HHcy,drinking history,Hb Alc,alkaline phosphatase,apo A-1,Fib and NLR are risk factors for CSVD.3.The CSVD nomogram prediction model can well predict the risk and probability of CSVD,and provide a basis for clinical diagnosis and treatment of CSVD and decision-making.