血清GLU和尿液mALB水平检测对白内障发生的预测价值评价  

Evaluation of Predictive Value of Serum GLU and Urine mALB Levels in Cataract Development

作　　者：吕亚玲 LÜYa-ling(Ningyi Ophthalmic Center of Nanjing,Nanjing 210000,China)

机构地区：[1]南京宁益眼科中心,南京210000

出　　处：《现代检验医学杂志》2020年第4期154-157,共4页Journal of Modern Laboratory Medicine

摘　　要：目的采用Logistic回归及受试者工作特征(ROC)曲线综合评价实验室常规检测项目对白内障发生的预测价值。方法收集2018年1~12月于南京宁益眼科中心就诊的50例白内障患者及同期50例非白内障患者,检测分析实验室常规检测项目尿素(BUN)、肌酐(CREA)、总胆固醇(total cholesterol,TC)、三酰甘油(triglyceride,TG)、血糖(GLU)、纤维蛋白原(fibrinogen,FIB)、活化部分凝血活酶时间(APTT)、超敏C反应蛋白(hs-CRP)和尿微量清蛋白(mALB)的水平,对有统计学意义的指标,先采用多因素Logistic回归模型分析,再进行ROC分析模型的预测价值。结果①白内障患者组GLU及mALB均高于非白内障组,差异具有统计学意义(t=3.68,P=0.000;t=2.06,P=0.042)。②Logistic回归分析,GLU(OR=2.532,95%CI:1.224~5.237)和mALB(OR=1.155,95%CI:1.070~1.247)与白内障发生发展存在相关性(P<0.05)。据此构建的白内障发生的风险预测模型为Y=1/[1+e-(0.929X1+0.144X2-7.852)]。③两者联合诊断白内障的ROC曲线下面积(AUC)大于GLU和mALB单项检测的AUC。结论GLU和mALB与白内障疾病发生发展显著相关,白内障风险预测模型有一定的预测价值。Objective To evaluate predictive value of routine laboratory testing items on cataract development using logistic regression and the receiver operating characteristic(ROC)curve.Methods 50 patients with cataract disease and 50 patients with non-cataract disease during the same period in the Ningyi Ophthalmic Center of Nanjing from January to December 2018 were collected.The values of routine laboratory test items including urea nitrogen(BUN),creatinine(CREA),total cholesterol(TC),triglyceride(TG),blood glucose(GLU),fibrinogen(FIB),activated partial thromboplastin time(APTT),high sensitivity C-reactive protein(hsCRP)and urine microalbuminuria(mALB)were analyzed.Multivariate Logistic regression analysis was established using the indicators with statistical significance,and then ROC analysis was also performed to evaluate the model value.Results The levels of GLU and mALB in cataract group were all significantly higher than those of non-cataract group(t=3.68,P=0.000;t=2.06,P=0.042).Logistic regression analysis showed that GLU(OR=2.532,95%CI:1.224~5.237)and mALB(OR=1.155,95%CI:1.070~1.247)might be enrolled as risk factors for cataract(P<0.05).The risk prediction model for cataract development was constructed Y=1/[1+e-(0.929X1+0.144X2-7.852)].The area under the ROC curve(AUC)of the two joint tests was significantly larger than that of the AUC of GLU and mALB separately in the diagnosis of cataract disease.Conclusion GLU and mALB were significantly correlated with the occurrence of cataract diseases,and the model of cataract risk has certain predictive value.

关 键 词：白内障 尿微量清蛋白 血葡萄糖 LOGISTIC回归 

分 类 号：R776.1[医药卫生—眼科] R446[医药卫生—临床医学]