Population-attributable risk assessment and risk prediction model of cardiovascular disease risk factors
10.3969/j.issn.1006-2483.2025.01.017
- VernacularTitle:心血管疾病危险因素的人群归因危险度评估及风险预测模型研究
- Author:
Yumei QIN
1
;
Guiqi CAO
1
;
Shiying JIANG
1
;
Yizhang XIAO
2
Author Information
1. School of Public Health and Wellness , Hubei University of Medicine , Shiyan , Hubei 442000 , China
2. Health Education Institute , Shiyan Center for Disease Control and Prevention , Shiyan , Hubei 442000 , China
- Publication Type:Journal Article
- Keywords:
Cardiovascular disease;
Population attributable risk percentage;
Risk prediction model;
Central and eastern regions
- From:
Journal of Public Health and Preventive Medicine
2025;36(1):74-78
- CountryChina
- Language:Chinese
-
Abstract:
Objective To explore the “contribution” of different exposures to cardiovascular diseases at the population level and to construct a risk prediction model for the effective allocation of prevention resources. Methods The CHNS (China Health and Nutrition Survey) database was used. In 2009, 2011 and 2015, 9 899 permanent residents aged 35 to 75 years in 10 provinces and cities in the central and eastern regions (Beijing, Liaoning, Heilongjiang, Shanghai, Shandong, Henan, Hubei, Hunan, Guangxi and Jiangsu) were selected as the research subjects. A single-factor analysis was conducted to examine the risk factors including sex, age, BMI, marital status, urban/rural area, sleep time, smoking, alcohol consumption, diabetes, education, and health insurance. The multifactor-adjusted population-attributable risk of certain risk factors was also estimated based on logistic regression analysis. The cardiovascular disease (CVD) risk prediction model was developed using a modeling group of 6 927 randomly selected individuals (70%) and a validation group of 2 974 individuals (30%). The model's differentiation and calibration were assessed using the receiver operating characteristic (ROC) curve and the Hosmer-Lemeshow goodness-of-fit test. Results The results showed that the adjusted population attributable risk and 95% confidence interval for BMI, sleep time, smoking, drinking and diabetes were 32.20% (27.67%-36.89%), 7.90% (1.68%-16.58%), 18.56% (11.35%-26.24%), 6.47% (0.11%-13.25%) and 5.73% (4.42%-7.03%). The results of multivariate adjusted population attributable risk percentage showed that BMI was the dominant cause of cardiovascular diseases, followed by smoking, sleep time, drinking and diabetes. The low-risk prevalence rate was 18.44%, the higher-risk prevalence rate was 14.19%, and the high-risk prevalence rate was 42.52%. The area under ROC curve AUC was 0.711, P<0.001, and Hosmer-Lemeshow goodness of fit test showed P=0.257. Conclusion In the future, it is important to focus on high-risk groups , control body mass index to the normal range, and reduce smoking , which is of great significance for the prevention of cardiovascular diseases. The risk prediction model has the value of good differentiation and practicability , and can provide certain prediction ability for the prevention of cardiovascular diseases.
