海南省心血管疾病高危人群预测模型的构建OA

Objective:Conducting high-risk screenings for cardiovascular disease(CVD)among the permanent residents of Hainan Province and developing a predictive model for identifying individuals at high risk for CVD will provide a foundational refer-ence for the prevention and management of this condition.Methods:A total of 71 819 permanent residents of Hainan Province were sampled in a questionnaire survey,physical examinations,and laboratory tests from 2016 to 2023,.The data were collected using a multi-stage stratified sampling method,and subsequently divided into a training set of 57 455 cases and a validation set of 14 364 cases,following an 8∶2 ratio.LASSO regression was employed for variable selection,while multi-factorial logistic regres-sion models were utilized to examine the determinants influencing CVD risk groups within the training dataset.Plotting line graphs for a risk prediction model aimed at individuals at risk of CVD,based on statistically significant influencing factors.Incorporate the training and validation datasets into the model,and evaluate its performance and utility using ROC curves,calibration curves,and decision curves.Results:12 937 high-risk group for CVD were screened,resulting in a detection rate of 18.01%(19.00%for males and 17.38%for females).The age group of 60 to 70 years had the highest proportion of high-risk individuals.The influenc-ing factors for the CVD high-risk groups were analyzed using LASSO regression and logistic regression.The following meaningful risk factors were identified for inclusion in the model:ethnicity,smoking status,dyslipidemia,systolic blood pressure,diastolic blood pressure,triglyceride-glucose index(TyG),and total cholesterol to high-density lipoprotein ratio(TC/HDL).Column-line plots were generated to illustrate these findings.The area under the curve(AUC)and 95%confidence intervals(CI)for the training and validation sets were 0.797(0.792~0.802)and 0.788(0.778~0.799),respectively.Conclusion:The prevalence of CVD detection among high-risk populations residing permanently in Hainan Province is notably low.The predictive model devel-oped in this research demonstrates commendable accuracy and stability,thereby facilitating the implementation of personalized CVD risk assessments.