School of Public Health, Key Laboratory of Public Health Safety of Ministry of Education, Fudan University, Shanghai, 200032, China; School of Public Health, Department of Epidemiology, Shanxi Medical University, Shanxi, 030001, China.
School of Public Health, Key Laboratory of Public Health Safety of Ministry of Education, Fudan University, Shanghai, 200032, China.
Nutr Metab Cardiovasc Dis. 2021 Jun 30;31(7):2058-2067. doi: 10.1016/j.numecd.2021.04.009. Epub 2021 Apr 20.
The accuracy of various 10-year atherosclerotic cardiovascular disease (ASCVD) risk models has been debatable. We compared two risk algorithms and explored clustering patterns across different risk stratifications among community residents in Shanghai.
A total of 28,201 residents (aged 40-74 years old) who were free of ASCVD were selected from the Shanghai Survey in China. The 10-year ASCVD risk was estimated by applying the 2013 Pooled Cohort Equations (PCEs) and Prediction for ASCVD Risk in China (China-PAR). The agreement was assessed between PCEs and China-PAR using Cohen's kappa statistics. The mean absolute 10-year ASCVD risk calculated by PCEs and China-PAR was about 10.0% and 6.0%, respectively. PCEs estimated that 44.9% of participants [with a 95% confidence interval (CI):44.0%-45.8%] were at high risk, while China-PAR estimated only 16.7% (95%CI:15.8%-18.0%) were at high risk. In both models, the percentage of high ASCVD risk was higher for participants who were older, men, less educated, current smokers, drinkers and manual workers. Among high-risk individuals, almost all participants (PCEs:90.5%; China-PAR:98.6%) had at least one risk factor; hypertension being the most prevalent. The concordance between PCEs and China-PAR was moderate (kappa:0.428, 95%CI: 0.420-0.434) with a better agreement for women (kappa:0.503,95%CI: 0.493-0.513) than for men (kappa:0.211,95%CI: 0.201-0.221).
The proportion of participants with a 10-year ASCVD high risk predicted by China-PAR was lower than the results of the PCEs. The risk stratifications of the two algorithms were inconsistent in terms of demographic and life-behaviour characteristics.
各种 10 年动脉粥样硬化性心血管疾病（ASCVD）风险模型的准确性一直存在争议。我们比较了两种风险算法，并探讨了上海社区居民在不同风险分层中的聚类模式。
从中国上海调查中选择了 28201 名无 ASCVD 的居民（年龄 40-74 岁）。应用 2013 年合并队列方程（PCEs）和中国 ASCVD 风险预测（China-PAR）估算 10 年 ASCVD 风险。采用 Cohen's kappa 统计量评估 PCEs 与 China-PAR 之间的一致性。PCEs 和 China-PAR 计算的平均绝对 10 年 ASCVD 风险分别约为 10.0%和 6.0%。PCEs 估计 44.9%（95%CI：44.0%-45.8%）的参与者处于高风险，而 China-PAR 仅估计 16.7%（95%CI：15.8%-18.0%）的参与者处于高风险。在这两种模型中，年龄较大、男性、受教育程度较低、当前吸烟者、饮酒者和体力劳动者的 ASCVD 高风险比例较高。在高危人群中，几乎所有参与者（PCEs：90.5%；China-PAR：98.6%）都至少有一种风险因素；高血压最为普遍。PCEs 与 China-PAR 的一致性为中度（kappa：0.428，95%CI：0.420-0.434），女性的一致性较好（kappa：0.503，95%CI：0.493-0.513），男性的一致性较差（kappa：0.211，95%CI：0.201-0.221）。
China-PAR 预测的 10 年 ASCVD 高危患者比例低于 PCEs 的结果。两种算法的风险分层在人口统计学和生活行为特征方面存在差异。
