A Nomogram Model For The Risk Of Post-stroke Depression Based On Clinical Characteristics And DNA Methylation

Objective:The purpose of our study is to use the clinical data that are easy to be collected in the early stage of patients and their DNA methylation levels to build a comprehensive line chart model to predict the risk of PSD,to help doctors and patients prevent the occurrence of PSD as soon as possible.Methods: According to the diagnostic criteria of stroke in the fourth academic Conference on Cerebrovascular Diseases in China,226 patients with acute stroke(onset time within 14 days)were collected from the Department of Neurology of Xiangya Hospital of Central South University from June 2019 to May 2021.DSM-5 was used to diagnose depression one month after onset.There were 112 cases in depression group and 114 cases in non-depression group.After obtaining informed consent,the basic clinical data of the patients were carefully questioned and collected.Including age,sex,education level,occupation,work status,nature of work,marital status,number of children,lifestyle,smoking history,drinking history,interest love history,time from onset to hospitalization,hypertension,diabetes history,hyperlipidemia history,stroke area,NIHSS score,BI score,Rankin score a total of 20 clinical indicators,and the next morning fasting peripheral venous blood of patients to extract DNA preservation.10 PSD and 10 non-PSD samples were selected for 850 K methylation sequencing,a PSD methylation spectrum was established,and different methylation sites,fragments and genes were found.On this basis,the sample size was expanded on the basis of Methyl Target secondgeneration sequencing.Finally,a nomogram model for predicting the risk of PSD and a nomogram model for predicting the risk of post-stroke major depression were established by using the 4 differentiated methylation sites and 24 predictive factors of the previous 20 clinical indicators.Result:1.Gender,education level and smoking history are the risk factors of PSD.Among them,women,lower education level and non-smoking patients have a higher risk of PSD;2.The methylation level of cg02550950 site on HECW2 gene may be related to the occurrence of PSD,and it is an independent risk factor for PSD,and the higher the methylation level,the greater the risk of PSD;3.NIHSS score,stroke location and time from onset to hospitalization were independent risk factors for PSD.The higher NIHSS score,the anterior circulation of stroke site,and the longer the time from onset to hospitalization,the higher the risk of PSD;4.24 predictive factors were screened by LASSO regression analysis,and it was found that seven predictive factors had predictive ability.The PSD line chart model was established by R software(version 3.6.1).The C index was as high as 0.898(95%CI0.857-0.939)and the area under the ROC curve was 0.902 by II.It is proved that the model has excellent calibration and differentiation;Conclusion: The nomogram model established to predict the risk of PSD based on seven predictive factors: gender,smoking history,education level,time from onset to hospitalization,location of stroke,NIHSS score and cg02550950 site methylation level has excellent calibration,differentiation and high clinical practical value.