Mok Yejin, Dardari Zeina, Sang Yingying, Hu Xiao, Bancks Michael P, Mathews Lena, Hoogeveen Ron C, Koton Silvia, Blaha Michael J, Post Wendy S, Ballantyne Christie M, Coresh Josef, Rosamond Wayne, Matsushita Kunihiro
Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, USA.
Division of Cardiology, Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
J Am Coll Cardiol. 2024 Feb 6;83(5):562-573. doi: 10.1016/j.jacc.2023.11.028.
American College of Cardiology/American Heart Association guidelines recommend distinct risk classification systems for primary and secondary cardiovascular disease prevention. However, both systems rely on similar predictors (eg, age and diabetes), indicating the possibility of a universal risk prediction approach for major adverse cardiovascular events (MACEs).
The authors examined the performance of predictors in persons with and without atherosclerotic cardiovascular disease (ASCVD) and developed and validated a universal risk prediction model.
Among 9,138 ARIC (Atherosclerosis Risk In Communities) participants with (n = 609) and without (n = 8,529) ASCVD at baseline (1996-1998), we examined established predictors in the risk classification systems and other predictors, such as body mass index and cardiac biomarkers (troponin and natriuretic peptide), using Cox models with MACEs (myocardial infarction, stroke, and heart failure). We also evaluated model performance.
Over a follow-up of approximately 20 years, there were 3,209 MACEs (2,797 for no prior ASCVD). Most predictors showed similar associations with MACE regardless of baseline ASCVD status. A universal risk prediction model with the predictors (eg, established predictors, cardiac biomarkers) identified by least absolute shrinkage and selection operator regression and bootstrapping showed good discrimination for both groups (c-statistics of 0.747 and 0.691, respectively), and risk classification and showed excellent calibration, irrespective of ASCVD status. This universal prediction approach identified individuals without ASCVD who had a higher risk than some individuals with ASCVD and was validated externally in 5,322 participants in the MESA (Multi-Ethnic Study of Atherosclerosis).
A universal risk prediction approach performed well in persons with and without ASCVD. This approach could facilitate the transition from primary to secondary prevention by streamlining risk classification and discussion between clinicians and patients.
美国心脏病学会/美国心脏协会指南推荐了用于原发性和继发性心血管疾病预防的不同风险分类系统。然而，这两种系统都依赖于相似的预测因素（如年龄和糖尿病），这表明存在一种针对主要不良心血管事件（MACE）的通用风险预测方法的可能性。
作者研究了预测因素在有和没有动脉粥样硬化性心血管疾病（ASCVD）的人群中的表现，并开发和验证了一种通用风险预测模型。
在9138名社区动脉粥样硬化风险研究（ARIC）参与者中，基线时（1996 - 1998年）有ASCVD的参与者为609名，无ASCVD的参与者为8529名。我们使用Cox模型，以MACE（心肌梗死、中风和心力衰竭）为终点，研究了风险分类系统中的既定预测因素以及其他预测因素，如体重指数和心脏生物标志物（肌钙蛋白和利钠肽）。我们还评估了模型性能。
在大约20年的随访中，发生了3209例MACE（无既往ASCVD者为2797例）。无论基线ASCVD状态如何，大多数预测因素与MACE的关联相似。通过最小绝对收缩和选择算子回归及自助法确定的包含预测因素（如既定预测因素、心脏生物标志物）的通用风险预测模型，对两组都显示出良好的区分度（c统计量分别为0.747和0.691），并且风险分类显示出良好的校准，与ASCVD状态无关。这种通用预测方法识别出了一些无ASCVD但风险高于某些有ASCVD个体的人，并在动脉粥样硬化多民族研究（MESA）的5322名参与者中进行了外部验证。
通用风险预测方法在有和没有ASCVD的人群中表现良好。这种方法可以通过简化风险分类以及临床医生与患者之间的讨论，促进从一级预防向二级预防的转变。
