한빛사논문
Yejin Mok PhD, MPH a, Zeina Dardari MS b, Yingying Sang MS a, Xiao Hu MHS a, Michael P. Bancks PhD, MPH c, Lena Mathews MD, MHS b, Ron C. Hoogeveen PhD d, Silvia Koton PhD a,e, Michael J. Blaha MD b, Wendy S. Post MD a,b, Christie M. Ballantyne MD, PhD d, Josef Coresh MD, PhD a, Wayne Rosamond PhD f, Kunihiro Matsushita MD, PhD a,b
aDepartment of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, USA
bDivision of Cardiology, Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA
cDepartment of Epidemiology and Prevention, Wake Forest University School of Medicine, Winston-Salem, North Carolina, USA
dDepartment of Medicine, Section of Cardiovascular Research, Baylor College of Medicine and Houston Methodist DeBakey Heart and Vascular Center, Houtson, Texas, USA
eStanley Steyer School of Health Professions, Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
fDepartment of Epidemiology, Gillings School of Global Public Health, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA
Address for correspondence: Dr Kunihiro Matsushita
Abstract
Background: 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).
Objectives: 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.
Methods: 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.
Results: 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).
Conclusions: 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.
논문정보
관련 링크