腰围与腰围身高比值筛查糖尿病前期和未诊断糖尿病患者的效能比较

Optimal screening tool for prediabetes and undiagnosed diabetes using waist circumstance and waist-to-height ratio

目的：探讨和比较上海地区人群腰围、腰围身高比值（WHtR）筛查空腹血糖受损（IFG）、糖耐量异常（IGT）、未诊断糖尿病的效能。方法采用2013年上海市慢性病及其危险因素监测研究的资料，共入选35岁及以上人群18062人，用Pearson相关分析方法分析腰围、WHtR与空腹血糖（FPG）、服糖后2 h血糖（2 hPG）的相关关系；受试者工作特征（ROC）曲线分析腰围、WHtR筛查IFG、IGT、未诊断糖尿病的曲线下面积（AUC）、最佳切点下的灵敏度和特异度；按目前推荐的切点，分析腰围、WHtR及与BMI串联、并联试验筛查三种血糖异常的灵敏度和特异度。结果腰围、WHtR与FPG存在相关关系（r＝0.161，0.153，均P〈0.05），两者与2 hPG也存在相关关系（r＝0.182，0.225，均P〈0.05）。腰围、WHtR筛查总的三种血糖异常的AUC分别是0.620（95%CI：0.611～0.629）、0.636（95%CI：0.628～0.645），差异有统计学意义（Z=2.549，P〈0.05）。腰围、WHtR在筛查IGT的AUC分别是0.617（95%CI：0.605～0.629）、0.646（95%CI：0.635～0.658），差异有统计学意义（Z=-3.418，P〈0.01），但在筛查IFG和未诊断糖尿病上未见统计学差异（均P〉0.05）。性别分层显示，对女性IGT筛查时， WHtR的AUC〉腰围（0.663比0.636，Z=-2.386，P〈0.05）。腰围最佳切点男女相差3？5 cm，WHtR男女相差0.01。根据推荐切点的四种组合筛查方案中，腰围联合BMI串联筛查IFG、IGT、未诊断糖尿病及总的血糖异常的特异度均最高，分别为67.61%、69.17%、68.94%、72.46%，WHtR联合BMI并联筛查时灵敏度最高，分别为76.10%、82.79%、83.89%、81.63%。结论 WHtR对女性IGT的筛查效能优于腰围，且WHtR在筛查IFG、IGT、未诊断糖尿病的最佳切点男女差异小。

Objective To investigate and compare the optimal screening tool for impaired fasting glucose（IFG） subjects, impaired glucose tolerance（IGT） subjects and undiagnosed diabetes, using waist circumstance（WC） and waist-to-height ratio（WHtR） in Shanghai. Methods A total of 18 062 participants （aged ≥35 years） were enrolled from the Shanghai Chronic Disease and Risk Factor Surveillance in 2013. Correlational analyses of WC and WHtR with IFG and IGT were conducted using Pearson correlation method. The receiver operating characteristic （ROC） analysis was used to compare their predictive ability for IFG, IGT and undiagnosed diabetes. According to the recommended cut-off value, screening of the three abnormal glucose status was performed using WC, WHtR and/or body mass index （BMI）. Results Both WC and WHtR were positively correlated with FPG （r=0.161, 0.153, P〈0.05 all above） and 2 hPG （r=0.182, 0.225, P〈0.05 all above）. The AUCs of WC and WHtR in screening IGT, IFG and undiagnosed diabetes were 0.620（95%CI：0.611-0.629）,0.636 （95%CI：0.628-0.645, Z=2.549,P〈0.05）, respectively. The same result was obtained among female patients in screening IGT （0.663 vs 0.636, Z=-2.386, P〈0.05）. The AUC of WC in screening IGT （0.617, 95%CI：0.605-0.629） was significantly lower than that of WHtR（0.646, 95%CI：0.635-0.658, Z=-3.418, P〈0.01）. However, no difference between WC and WHtR in screening IFG and undiagnosed diabetes was shown （P〉0.05）. The difference of the best cut-off value of WC and WHtR between male and female subjects was 3-5 cm and 0.01, respectively. Among 4 screening algorithms by combining BMI in parallel or in serial, the specificity of WC combing BMI in serial was the highest for the screening of IFG, IGT, undiagnosed diabetes, and abnormal glucose subject （67.61%, 69.17%, 68.94%, and 72.46%, respectively）. The sensitivity of WHtR combing BMI in parallel was the highest （76.10%, 82.79%, 83.89%, and 81.63%）. Conclusion WHtR is a better option than WC in screening IGT for females. There is almost no gender difference in the optimal cut off values of WHtR for screening of IGT, IFG and undiagnosed diabetes.