不同体格测量指标对血糖异常预测的分析

Analysis on the Predictive Value of Different Physical Measurement Indexes for Blood Glucose Abnormality

目的比较不同体格测量指标对于血糖异常的预测价值,探索适宜预测血糖异常的体格测量指标。方法通过绘制受试者工作特征(Receiver Operating Characteristic,ROC)曲线,比较某社区1087名35岁及以上人群的体质指数(BMI)、腰臀比(WHR)、腰高比(WHtR)、腰围(WC)等体格测量指标对血糖异常的预测价值,并计算各指标预测血糖异常的最佳截断值及其相应灵敏度和特异度。结果四个体格测量指标对于血糖异常均有一定预测价值(AUC为0.639~0.653),约登指数依次为WHtR(26.03%)、WHR(22.52%)、WC(22.35%)、BMI(21.59%),敏感性依次为BMI(84.85%)、WHtR(76.86%)、WC(59.23%)、WHR(48.76%)。结论体格测量指标的异常可以早期预测血糖异常;BMI、WHR及WC预测血糖异常有性别差异;WHtR无性别差异,约登指数及敏感性均较高,对于血糖异常的预测价值优于BMI、WHR及WC。

Objective To compare the predictive value of different physical measurement indexes for abnormal blood glucose,and explore the physical measurement indexes suitable for predicting abnormal blood glucose.Methods By drawing Receiver Operating Characteristic(ROC)curves,the predictive value of abnormal blood glucose between Body Mass Index(BMI),Waist To Hip Ratio(WHR),Waist To Height Ratio(WHtR),Waist Circumference(WC)among 1087 people aged 35 and above in a community were compared,and the optimal cut-off values and their Sensitivity and Specificity of each index for predicting blood glucose abnormality were calculated.Results The four indexes had certain predictive value of abnormal blood glucose(AUC was 0.639~0.653).The Jordon index from the largest to the smallest was WHtR(26.03%),WHR(22.52%),WC(22.35%),BMI(21.59%),and sensitivity in proper order were BMI(84.85%),WHtR(76.86%),WC(59.23%),WHR(48.76%).Conclusion The abnormal physical measurement indexes can predict abnormal blood glucose early.The predictive value of abnormal blood glucose abnormality has significant gender difference in BMI,WHR,WC,but there was no gender difference in WHtR,and the Jordan index and sensitivity were higher than the others’.The WHtR’s predictive value for abnormal blood glucose was better than BMI,WHR and WC.