目的：探讨高龄老年人群中几种常用估算肾小球滤过率 (estimated glomerular filtrations rate, eGFR)公式和尿白蛋白肌酐比(urinary albumin to creatinine, UACR)与心脑血管事件、肾脏终点事件和死亡的关系。方法：收集2015年1月至2018年12月上海交通大学医学院附属瑞金医院老年病科诊治的年龄≥65岁且有1年以上随访资料的患者。使用Logistic回归分析评价由无种族差异的慢性肾脏疾病流行病学协作组（Chronic Kidney Disease Epidemiology Collaboration，CKD-EPI）肌酐（creatinine，Cr）公式(CKD-EPI_Cr公式)、CKD-EPI胱抑素C（cystatin C，CysC）公式 (CKD-EPI_Cys公式)、无种族差异的CKD-EPI_Cr-Cys公式和柏林倡议研究（Berlin Initiative Study，BIS）公式2（基于Cr和CysC）计算的eGFR与预后关联。结果：本研究共纳入475例老年患者，中位年龄及随访时间分别为83（76~87）岁及76（68~91）个月。校正前，使用CKD-EPI_Cys、CKD-EPI_Cr-Cys及BIS2公式计算eGFR时，eGFR下降[即eGFR<60 mL/（min∙1.73 m²）]者发生死亡风险较高，但经多个协变量校正后，3个公式与死亡的相关性均消失。校正前，4种eGFR估算公式均提示eGFR下降者发生心脑血管事件的风险升高，其中BIS2公式比值比（odds ratio，OR）最高，仅BIS2公式经性别、年龄及既往病史等因素校正后OR仍有统计学意义（P=0.038）。CKD-EPI_Cr-Cys公式计算eGFR时，经性别、年龄及既往病史等因素校正后，eGFR下降者发生肾脏终点事件风险较高（P=0.023），其他公式均未显示两者相关性。校正前，4种公式均提示eGFR下降者发生复合事件的风险升高，其中使用BIS2公式的OR最高，但经性别、年龄校正后，CKD-EPI_Cr（P=0.030）、CKD-EPI_Cys（P=0.044）和BIS2（P=0.034）OR具有统计学意义。多因素校正后，与UACR正常者（即UACR<30 mg/g）相比，UACR升高者发生死亡、心脑血管事件、肾脏终点事件及复合终点的风险均显著提高，且除肾脏终点事件外，UACR升高与其他终点事件相关性均高于eGFR下降。结论：UACR升高与死亡、心脑血管事件及复合终点相关性均较强。不同eGFR公式中，使用BIS2公式估算的eGFR下降与心脑血管事件相关性较强，而CKD-EPI_Cr-Cys与肾脏终点事件相关性较强。

Objective To investigate the association of several commonly used estimated glomerular filtration rate (eGFR) formula and urinary albumin to creatinine ratio (UACR) with cardiovascular events, renal endpoint events, and mortality in an elderly population.Methods The patients aged ≥ 65 who were treated and had follow-up data more than 1 year in the Department of Geriatrics of Ruijin Hospital, Shanghai Jiao Tong University School of Medicine from January 2015 to December 2018 were selected. Logistic regression analyses were used to evaluate the association of eGFR derived by chronic kidney disease epidemiology collaboration(CKD-EPI) creatinine（Cr） formula (CKD-EPI_Cr), CKD-EPI cystatin C formula( CKD-EPI_Cys), the racially neutral CKD-EPI_Cr-Cys formula and Berlin Initiative Study (BIS) formula 2 (based on Cr and cystatin C) with prognosis.Results Totally 475 elderly patients were recruited with a median age of 83(76-87) years and a follow-up time of 76(68-91) months. Before adjustment, those with a decreased eGFR[<60 mL/(min·1.73 m²)] had a higher risk of death when eGFR estimated by CKD-EPI_Cys, CKD-EPI_Cr-Cys, and BIS2, but the association was negative after adjusting by multiple covariates. Before adjustment, those with a decreased eGFR by all equations had a significantly higher risk of cardiovascular events. The odds ratio was highest for eGFR estimated by BIS2. After adjusting for gender, age, and past medical history, the association remains statistically significant between eGFR estimated by BIS2 and cardiovascular events(P=0.038). After adjusting for gender, age, and past medical history, the risk of renal endpoint events was higher in those with decreased eGFR estimated by CKD-EPI_Cr-Cys (P=0.023). None of the other equations showed a significant correlation with renal endpoint events. Before adjustment, those with a decreased eGFR by all equations had a significantly higher risk of composite events. The odds ratio was highest for eGFR estimated by BIS2. After adjusting for gender and age, the association remains statistically significant between composite events and eGFR estimated by CKD-EPI_Cr(P=0.030) fit without race, CKD-EPI_Cys (P=0.044) and BIS2(P=0.034). After multivariate adjustment, those with UACR ≥ 30 mg/g had a significantly higher risk of death, cardiovascular events, renal endpoint events, and composite endpoints compared with those with UACR < 30 mg/g. Elevated UACR was more strongly associated with outcomes than decreased eGFR but not with renal endpoint events.Conclusions Elevated UACR was strongly associated with death, cardiovascular events, and composite endpoints. Among different eGFR formations, decreased eGFR using BIS2 was more strongly associated with cardiovascular events, whereas CKD-EPI_Cr-Cys was more strongly associated with renal endpoint events.