A prediction model of high-risk population for cardiovascular diseases
10.19485/j.cnki.issn2096-5087.2024.03.007
- Author:
ZHOU Zimeng
;
HONG Xin
- Publication Type:Journal Article
- Keywords:
cardiovascular diseases high-risk population;
influencing factor;
nomogram
- From:
Journal of Preventive Medicine
2024;36(3):211-214
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To investigate the proportion of high-risk population for cardiovascular diseases (CVD) among residents at ages of 35 to 79 in Nanjing City, and establish a prediction model of high-risk population for CVD.
Methods:Residents at ages of 35 to 79 years were selected from Nanjing City using a multi-stage stratified cluster random sampling method from 2020 to 2021. Participants' demographic information, characteristics of lifestyle and blood biochemical index were collected using questionnaire surveys, physical examination and laboratory testing. The high-risk population for CVD were determined according to the Chinese Guidelines for Cardiovascular Disease Risk Assessment and Management, and the Chinese Guidelines for the Prevention and Treatment of Adult Dyslipidemia (2016 Revision). Predictive factors for high-risk population for CVD were screened using a multivariable logistic regression model. A nomogram was established and verified with receiver operation characteristic (ROC) curve. Hosmer-Lemeshow goodness of fit test was used to evaluate the fitting effect and Bootstrap method was used for internal verification.
Results:A total of 38 428 individuals were surveyed, including 17 970 males (46.76%) and 20 458 females (53.24%), and 25 714 individuals aged 35 to 59 years. There were 8 905 high-risk population for CVD, with a detection rate of 23.17%. Multivariable logistic regression analysis identified 9 factors affecting high-risk population for CVD. A prediction model was established for ln[P/(1-P)]=-7.305+2.107×age-0.366×gender+0.299×marital status-0.297×educational level+0.631×body mass index+0.013×sleep duration+0.096×edible salt intake+0.444×smoke-0.069×alcohol consumption. The area under ROC curve was 0.799 (95%CI: 0.794-0.805), the sensitivity and specificity were 0.731 and 0.753, indicating good differentiation. The nomogram based on the above factors indicated good calibration and stability.
Conclusion:The nomogram constructed by age, gender, marital status, educational level, body mass index, sleep duration, edible salt intake, smoking and alcohol consumption can be used to predict high-risk population for CVD.
- Full text:2024072311083077218心血管病高危人群预测模型研究.pdf
