Construction Of Cardiovascular Disease Risk Prediction Model Based On Physical Examination Population

Objective:To construct a cardiovascular disease risk prediction model for physical examination population,stratified accurate management of low-risk,medium-risk and high-risk groups of cardiovascular disease in physical examination population,and early intervention of risk factors,so as to provide a basis for effective prevention and treatment of cardiovascular diseases.Methods:In this study,physical examination population aged 20-74 years old in the physical examination center of a Class III hospital in Changzhi City was selected as the research objects.Convenient sampling method was adopted to conduct a questionnaire survey after informed consent of the physical examination subjects in accordance with the principle of voluntary participation from March to October 2022,and general demographic data,behavior and lifestyle,and mental health data were collected.Obtain physical measurement and physiological and biochemical test data from the health information platform of healthy people.Logistic regression was used to analyze the independent influencing factors of cardiovascular disease in the physical examination subjects,and then four cardiovascular risk assessment models were selected: Risk prediction was conducted by Framingham Cardiovascular Composite Risk Score(2008revision),10-year Ischemic Cardiovascular Disease Risk Score(ICVD),10-year ASCVD Overall Risk Score of Chinese Lipid Guidelines(2016)and 10-year ASCVD Risk Prediction Tool of Chinese population(2016 China-PAR).Kappa consistency evaluation was used to compare the evaluation results,behavioral factors and psychological factors were added into the two models with the highest consistency,and the area under the subjects’ working characteristic curve was used to evaluate the prediction effect of the models.Results:1.Univariate analysis showed that there were statistically significant differences among the general demographic characteristics: residence region,sex,age,marital status,education level,occupation category,family history of cardiovascular disease,history of diabetes,whether to take antihypertensive drugs,whether to receive antihypertensive treatment and whether to suffer from cardiovascular disease(P<0.05).Physical measurement and biochemical indicators: body weight,waist circumference,SBP,DBP,TC,HDL-C,LDL-C and whether there was cardiovascular disease were statistically significant differences(P<0.05);Lifestyle: There were statistically significant differences between smoking,drinking,frequency of consumption of pickled products,edible oil intake,actual sleep duration at night and whether there was cardiovascular disease(P<0.05);Dietary status: There were statistically significant differences in fruit intake,milk or dairy intake,fish intake,egg,tea intake,food types,frequency of drinking sugar-sweetened beverages and whether there was cardiovascular disease(P<0.05);In terms of exercise,there was a statistically significant difference between the duration of severe activity and whether there was cardiovascular disease(P<0.05).Psychological factors: The total score of DASS-21,social support and whether suffering from cardiovascular disease were significantly different(P<0.05).2.The results of multivariate analysis showed that among the behavioral factors,alcohol consumption,actual sleep duration at night,intake of milk or dairy products,intake of fish food,frequency of drinking sugar-sweetened beverages,amount of tea and duration of heavy activity were independent influencing factors of cardiovascular disease in the physical examination group(P<0.05).Among psychological factors,social support was an independent factor affecting cardiovascular disease in the physical examination group(P<0.05).3.The consistency between the lipid guideline risk assessment tool and the China-PAR model was the highest,and the Kappa value was 0.580.The consistency between the two models was moderate,and the difference was statistically significant(P<0.001).Among them,the proportion of low-risk and high-risk groups showed by the China-PAR model was lower than that of the lipid guideline risk assessment tool.4.Based on the China-PAR model,behavioral factors(alcohol consumption,actual sleep duration at night,intake of milk or dairy products,intake of fish foods,frequency of consumption of sugar-sweetened beverages,amount of tea,duration of heavy activity)and psychological factors(social support)were added to the cardiovascular disease risk prediction model of physical examination subjects.The area under ROC curve was 0.923(95%CI: 0.882 ～ 0.965),the specificity was 92.9%,the sensitivity was 88.2%.Conclusion:The cardiovascular disease risk prediction model established in this study has good prediction ability and can accurately predict the physical examination population with different risks,which is of great value for the development of personalized intervention measures,reducing the risk of cardiovascular disease and improving the quality of life,and can provide a basis for the precise prevention and treatment of cardiovascular diseases in the future.