109例2型糖尿病患者子女患病风险预测分析

目的利用自身容易获得的患者临床资料,建立一种简单易行的评价方法,对2型糖尿病患者子女进行糖尿病或糖尿病前期风险预测。方法选择2型糖尿病患者子女共109名,统计其年龄,测定其空腹血糖及口服葡萄糖耐量试验(oral glucosetolerance test,OGTT)2 h血糖、身高、体质量、颈围(neck circumference,NC)、腰围(waist circumference,WC)并计算腰臀比(waist-hip ratio,WHR)及体质量指数(body mass index,BMI)。分别以年龄、颈围、腰围、腰臀比、体质量指数为筛查指标,以空腹血糖及OGTT 2 h血糖作为诊断糖尿病或糖尿病前期的金标准,绘制受试者工作特征曲线(receiver operating curve,ROC),计算ROC曲线下面积,并根据分析结果建立风险预测决策树。结果①以上筛查指标中,以腰围筛查糖尿病时,ROC曲线下面积(area underreceiver operating curve,AUC)最大(AUC=0.770,P=0.000)。②以腰围筛查糖尿病或糖尿病前期时,ROC曲线下面积亦最大(AUC=0.704,P=0.000)。③糖尿病患者子女患糖尿病风险预测决策树的ROC曲线下面积(AUC=0.784,P=0.000),灵敏度为61.1%,特异度为95.6%,阳性预测值为68.8%,阴性预测值为93.5%。④糖尿病子女患糖尿病或糖尿病前期风险预测决策树的ROC曲线下面积(AUC=0.727,P=0.000),灵敏度为56.5%,特异度为88.9%,阳性预测值为78.8%,阴性预测值为73.7%。结论①在年龄、颈围、腰围、腰臀比、体质量指数等5项临床资料中,反映中心性肥胖的腰围是患糖尿病或糖尿病前期的最主要危险因素。②用本决策树对糖尿病患者子女患糖尿病或糖尿病前期进行风险预测简便易行,具有一定的灵敏度和特异度。

Analysis of 109 filial generation of T2DM in predicting diabetes and prediabetes

Objective Using some data that were easy to access at home, to establish and evaluate a simple tool to identify the type 2 diabetes mellitus(T2DM) and prediabetes(PDM) in the filial generation of T2DM. Methods A total of 109 filial generation of T2DM patients participated in the test. Measuring their blood sugar, age, neck circumference(NC), waist circumference(WC), hip circumference(HC), height, weight. Their waist-hip ratio(WHR) and body mass index(BMI) were calculated. Receiver operating curve(ROC) was used in the analysis and according to the analysis results to establish classification tree. Results For detecting T2DM, the area under receiver operating curve(AUC)of WC was the highest among all of the five risk factors, the AUC=0.770, P=0.000. For detecting T2DM and PDM, AUC of WC was the highest among all of the five risk factors, the AUC=0.770, P=0.000. AUC, sensitivity, specificity, positive predictive value, negative predictive value were 78.4%, 61.1%, 95.6%, 68.8% and 93.5%, respectively for detecting T2DM. AUC, sensitivity, specificity, positive predictive value, negative predictive value were 72.7%, 56.5%, 88.9%, 78.8% and 73.7% respectively for detecting T2DM and PDM. Conclusion WC, which reflects central obesity, has the biggest AUC, can best predict the possibility of suffering T2DM and PDM and is used as the first bifurcation in the classification tree. The classification tree is an easy and feasible tool to detect T2DM and PDM.