Influence factors and predictive ability of a risk prediction model for carotid atherosclerosis in a follow-up population
10.16462/j.cnki.zhjbkz.2019.04.003
- Author:
Qi WANG
1
;
Juan-sheng LI
;
Hong-quan PU
;
Ya-na BAI
;
Hai-yan LI
;
Ning CHENG
;
Zheng-fang WANG
;
Lei-jie ZHANG
;
Wan-qi ZHU
;
Yan. YUAN
Author Information
1. Department of Epidemiology and Biostatistics, School of Public Health, Lanzhou university, Lanzhou 730000, China
- Publication Type:Research Article
- Keywords:
Followed up population;
Carotid arteriosclerosis;
Influencing factors;
Model evaluation
- From:
Chinese Journal of Disease Control & Prevention
2019;23(4):382-386
- CountryChina
- Language:Chinese
-
Abstract:
Objective To explore factors influencing the incidence of carotid atherosclerosis in different genders so as to provide reference for the specific prevention of the disease. Methods A nested case-control study was conducted to analyze factors influencing the incidence of carotid atherosclerosis in Jinchang cohort population who were randomly selected through stratified sampling by age and followed up. A risk prediction model was established and the goodness of fit was evaluated by the area under the receiver operator characteristic curve (ROC). Results The standardized incidence of carotid atherosclerosis in this follow-up population was 12.32%, and the incidence rate of males (13.65%) was greater than that of females (11.29%). The difference was statistically significant ( 2=4.267, P<0.001). Age, education, elevated systolic blood pressure, and elevated low-density lipoprotein cholesterol were common risk factors for carotid atherosclerosis in both men and women. Elevated fasting plasma glucose (OR=2.556, 95% CI: 1.618-4.038) and elevated triglyceride (OR=1.535, 95% CI: 1.058-2.227) were only associated with men. Abdominal obesity (OR=1.414, 95% CI: 1.013-1.974) was only associated with women. The area under ROC of male and female prediction models was 0.835 (95% CI: 0.815-0.856) and 0.809 (95% CI: 0.788-0.831), respectively. The sensitivity was 78.0% and 78.9%, the specificity was 78.8% and 73.1%, and the diagnostic coincidence rate was 91.3% and 82.4%, respectively. Conclusions There are different risk factors for carotid atherosclerosis in males and females, and targeted prevention and control measures should be taken according to gender. The risk prediction model established by Logistic regression had certain guiding value.
