不同时段尿白蛋白在诊断早期糖尿病肾脏损伤中的应用

Application of urinary albuminuria at different time in the monitoring of early impairment in kidney of diabetes mellitus

目的研究糖尿病患者不同时段尿白蛋白（urinary albumin）的排泌情况及尿白蛋白在诊断早期糖尿病肾脏损伤中的应用。方法收集中山医院门诊及住院糖尿病患者及健康对照组3d内不同时间段的尿液，分析尿白蛋白天内、天间的排泌变化情况；以24h尿白蛋白为标准判断肾脏早期损伤情况，比较不同时段尿及时间点尿与24h尿白蛋白的相关性、诊断特异度及敏感度；评估随机尿的诊断特异度及敏感度，推导随机尿最佳诊断水平。结果尿白蛋白天间变异较大，以尿Cr和尿量分别校正后可降低变异。糖尿病组中尿白蛋白使用尿Cr校正后变异系数（CV小于尿量校正（CV分别为49％±23％vs 64％±30％）。尿白蛋白天内排泌呈节律性变化。不同尿液留取方式中夜间尿尿白蛋白／尿Cr（ratio of urinary concentrations of albumin and creatinine，ACR）与24h尿白蛋白定量相关性最好（R2=0．976），优于晨尿ACR（R2=0．900）、午间餐后尿ACR（R2=0．584）和随机尿ACR（R2=0．791）。以24h尿白蛋白总量作为判断标准进行受试者操作特性曲线（ROC曲线）分析显示，随机尿ACR的判断值为27．7μg／mg尿Cr（存在男女性别差异：男性12．8μg／mg尿Cr傩女性27．0μg／mg尿cr）。最小阴性似然比0．011时推导随机尿ACR的排除判断值为13．0μg／mg尿Cr；最大阳性似然比481．000时推导随机尿ACR的确诊判断值为87．4μg／mg尿Cr。结论尿Cr较尿量能更好地降低尿自蛋白天内变异，但仍无法完全消除变异。夜尿ACR与24h尿白蛋白定量相关性最好，可替代24h尿白蛋白定量。随机尿ACR作为最方便留取的尿液标本亦可以较好地替代24h尿白蛋白定量，但应考虑引入尿Cr后带来的性别间差异。以13．0μg／mg及87．4μg／mg作为随机尿ACR的排除判断值及确诊判断值可以便于临床医师基本排除或确定白蛋白尿的出现。

Objective To investigate the urinary albumin excretion of the diabetes patients and application value in the monitoring of early impairment in kidney. Methods The random urine samples from diabetes patients and controls within three days were collected. The changes of urinary albumin excretion within day and between days were analyzed. 24-hour urine albumin was used as a standard to evaluate early kidney damage. The correlations between results of random urine albumin at the different time points and different periods were compared. The sensitivity and specificity of random urine albumin at the different time points and different periods was evaluated and compared to deduce the best diagnostic performance of the random urine albumin. Results There are greater variations of the levels of urinary albumin of patients with diabetes and control. After the correction with urine creatinine and urine volume the variations can be reduced （ CV：49% ± 23% and 6.4% ± 30% ）. Urinary albumin excretion rate change rhythmically within the 24 h in healthy and diabetes patients. We found the best correlation between overnight ratio of urinary concentrations of albumin and creatinine （ACR） and 24-hour urinary albumin （R2 = 0. 976 ）. It was superior to urina sanguinis （R2 = 0.900 ）, postprandial urine （R2 = 0.584 ） and random urine （R2=0. 791）. When 24 h urinary albumin was taken as the standard, receiver operating characteristic （ROC） curve analysis showed there was significant difference between male and female（ male 12. 8 μg/mg urine creatinine vs female 27.0μg/mg urine creatinine） ,and the the cut-off value of ACR was 27.7 μg/mg urine creatinine. When the smallest available negative likelihood ratio （0. 011 ） and the greatest positive likelihood ratio （481. 000） were obtained,the concentration of 13.0 μg/mg creatinine and 87.4μg/mg creatinine were set as the cut-off value of ACR. Concluslomg The correction with urinary creatinine can reduce the variation between-days compared with urine volume, but still can not completely eliminate the variability. The ACR of overnight urine has the best correlation with the 24 h urinary albumin and can replace 24 h urinary albumin. Random urine as the most convenient collecting urine samples can also replace 24-hour urinary albumin, but the gender discrepancy need to be considered. When the concentration of 13.0μg/mg and 87.4μg/mg was set as a random ACR exclusion value and the confirmative value, it can basically rule out and confirm the existence of micmalbuminuria.