Cardiovascular Risk Based on ASCVD and KDIGO Categories in Chinese Adults: A Nationwide, Population-Based, Prospective Cohort Study
Yu Xu,Mian Li,Guijun Qin,Jieli Lu,Li Yan,Min Xu,Tiange Wang,Zhiyun Zhao,Meng Dai,Di Zhang,Qin Wan,Yanan Huo,Lulu Chen,Lixin Shi,Ruying Hu,Xulei Tang,Qing Su,Xuefeng Yu,Yingfen Qin,Gang Chen,Zhengnan Gao,Guixia Wang,Feixia Shen,Zuojie Luo,Li Chen,Yuhong Chen,Yinfei Zhang,Chao Liu,Youmin Wang,Shengli Wu,Tao Yang,Qiang Li,Yufang Bi,Jiajun Zhao,Yiming Mu,Weiqing Wang,Guang Ning,for the China Cardiometabolic Disease and Cancer Cohort Study Group,
DOI: https://doi.org/10.1681/asn.2020060856
IF: 14.978
2021-03-04
Journal of the American Society of Nephrology
Abstract:Background The Kidney Disease Improving Global Outcomes (KDIGO) clinical practice guideline used eGFR and urinary albumin-creatinine ratio (ACR) to categorize risks for CKD prognosis. The utility of KDIGO’s stratification of major CVD risks and predictive ability beyond traditional CVD risk prediction scores are unknown. Methods To evaluate CVD risks on the basis of ACR and eGFR (individually, together, and in combination using the KDIGO risk categories) and with the atherosclerotic cardiovascular disease (ASCVD) score, we studied 115,366 participants in the China Cardiometabolic Disease and Cancer Cohort study. Participants (aged ≥40 years and without a history of cardiovascular disease) were examined prospectively for major CVD events, including nonfatal myocardial infarction, nonfatal stroke, and cardiovascular death. Results During 415,111 person-years of follow-up, 2866 major CVD events occurred. Incidence rates and multivariable-adjusted hazard ratios of CVD events increased significantly across the KDIGO risk categories in ASCVD risk strata (all P values for log-rank test and most P values for trend in Cox regression analysis <0.01). Increases in c statistic for CVD risk prediction were 0.01 (0.01 to 0.02) in the overall study population and 0.03 (0.01 to 0.04) in participants with diabetes, after adding eGFR and log(ACR) to a model including the ASCVD risk score. In addition, adding eGFR and log(ACR) to a model with the ASCVD score resulted in significantly improved reclassification of CVD risks (net reclassification improvements, 4.78%; 95% confidence interval, 3.03% to 6.41%). Conclusions Urinary ACR and eGFR (individually, together, and in combination using KDIGO risk categories) may be important nontraditional risk factors in stratifying and predicting major CVD events in the Chinese population.
urology & nephrology