论文部分内容阅读
背景用肾小球滤过率(eGFR)和蛋白尿评估法来定义慢性肾脏病及其分期,还存在着实质性的争议。我们采用荟萃分析法评估eGFR和蛋白尿与死亡率的独立及合并相关性。方法在一般人群的联合荟萃分析中,我们纳入的研究至少含1 000个受试者,且有eGFR和尿蛋白水平基线信息,集中了全因和心血管死亡率标准化数据。校正潜在影响因素,采用Cox比例风险模型评估全因和心血管死亡率与eGFR和尿蛋白相关危害比(HRs)的关联性。发现有14项研究,105 872名受试者(730 577人年数)做了尿蛋白肌酐比值(ACR)检测;有7项研究,1 128 310名受试者(4 732 110人年数)做了尿蛋白检测(试纸浸渍法)。在做ACR检测的研究中,eGFR水平在75～105 ml.min-1.1.73 m-2时与死亡风险无相关性;若eGFRs低于此范围死亡风险增加。与eGFRs为95 ml.min-1.1.73 m-2相比,全因死亡率校正危害比:eGFR60 ml.min-1.1.73 m-2为1.18〔95%CI(1.05,1.32)〕;45 ml.min-1.1.73 m-2为1.57〔95%CI(1.39,1.78)〕;15 ml.min-1.1.73 m-2为3.14〔95%CI(2.39,4.13)〕。在双对数坐标上ACR与死亡风险呈线性相关,无阈值效应。与ACR 0.6 mg/mmol相比,全因死亡校正危害比:ACR 1.1 mg/mmol为1.20〔95%CI(1.15,1.26)〕;3.4 mg/mmol为1.63〔95%CI(1.50,1.77)〕;33.9 mg/mmol为2.22〔95%CI(1.97,2.51)〕。eGFR和ACR与死亡风险增加相关,但未获得交互作用证据。心血管死亡率和试纸浸渍检验研究也记录了相似的发现。解释eGFR<60 ml.min-1.1.73 m-2和ACR≥1.1 mg/mmol(10 mg/g)为一般人群死亡风险的独立预测因子。本研究为风险评估及慢性肾脏病的定义和分期提供了定量数据。
Background There is substantial controversy over the definition of CKD and its stage by using glomerular filtration rate (eGFR) and proteinuria assessment. We used a meta-analysis to assess the independent and coherent associations of eGFR and proteinuria with mortality. Methods In a co-meta-analysis of the general population, we included at least 1,000 subjects with baseline information on eGFR and urinary protein concentrations and standardized standardized data on all-cause and cardiovascular mortality. Potential influencing factors were corrected and Cox proportional hazards model was used to assess the association between all-cause and cardiovascular mortality and eGFR and urinary protein-associated hazard ratio (HRs). A total of 14 studies and 105,872 subjects (730,577 person-years) were tested for urinary protein creatinine ratio (ACR); 7 studies, 1,128,310 subjects (4,732,110 person-years) were performed Urine protein test (dipstick dip). In the ACR test, there was no correlation between the eGFR level and death risk at 75-105 ml.min-1.1.73 m-2. If the eGFRs were below this range, the risk of death increased. The adjusted hazard ratio for all-cause mortality compared with eGFRs of 95 ml.min-1.1.73 m-2 was 1.18 [95% CI (1.05, 1.32)] for eGFR 60 ml.min-1.1.73 m-2; 1.57 [95% CI (1.39, 1.78)] for ml.min-1.1.73 m-3; 3.14 [95% CI (2.39, 4.13)] for 15 ml.min-1.1.73 m-2. There is a linear correlation between ACR and death risk in double logarithmic coordinates, with no threshold effect. The hazard ratio of all-cause death correction was ACR 1.1 mg / mmol 1.20 [95% CI (1.15,1.26)] compared to ACR 0.6 mg / mmol; 1.63 [95% CI 1.50,1.77] at 3.4 mg / mmol ; 33.9 mg / mmol 2.22 [95% CI (1.97, 2.51)]. eGFR and ACR were associated with an increased risk of death but no evidence of interaction. Similar findings were also recorded in studies of cardiovascular mortality and dipstick test strips. Interpretation eGFR <60 ml.min-1.1.73 m-2 and ACR ≥ 1.1 mg / mmol (10 mg / g) are independent predictors of mortality in the general population. This study provides quantitative data for the definition and staging of risk assessment and chronic kidney disease.