不同肥胖判定指标及其组合对预测高血压患病风险的效率评价

Evaluating the efficiency of different obesity indicators and their combinations in predicting the risk of hy⁃pertension

目的探讨和评价单项肥胖判定指标和各项肥胖判定指标联合使用对高血压患病风险的预测价值。方法对调查地区3 216名成年居民的体质指数(BMI)、腰围(WC)、腰臀比(WHR)、腰高比(WHtR)等结果进行横断面分析;通过建立Logistic回归模型,确立预测高血压的最优肥胖指标;分析各肥胖判定指标和各肥胖判定指标联合使用对预测高血压的ROC曲线,比较各ROC曲线下面积(AUC),探索预测高血压患病风险能力的最佳组合。结果 BMI、腰围、腰臀比和腰高比均为检出高血压的危险因素,男性OR值分别为1.46(95%CI:0.98~1.89)、1.72(95%CI:1.10~2.52)、1.81(95%CI:1.18~2.62)、2.05(95%CI:1.40~3.05);女性OR值分别为1.84(95%CI:1.34~2.52)、1.57(95%CI:1.08~2.43)、2.02(95%CI:1.24~3.10)、1.86(95%CI:1.21~2.78)。单项指标不同性别间AUC最大的指标均为腰高比,分别达到0.685和0.718,其AUC与BMI、腰围和腰臀比差异均有统计学意义,除女性腰高比的AUC> 0.7,其余均在0.5~0.7之间,诊断准确度较低。对各指标组合后,AUC随指标组合增多而增大,准确度较高,BMI&腰围&腰高比与腰围&腰高比、腰围&腰臀比&腰高比差异无统计学意义,其余差异均有统计学意义,且男性和女性的AUC均在BMI&腰围&腰高比组合达到最高值,分别达到0.703和0.740。结论单项指标中,腰高比为预测高血压患病风险的最佳指标,但预测准确性不高;BMI&腰围&腰高比组合可能是预测高血压患病风险的最佳组合;以腰高比为必选项的组合指标有助于提高高血压预测能力,同时应在健康教育和健康促进工作中引起重视。

Objective To evaluate the potential of single obesity indicators and their combinations to pre⁃dict the risk of hypertension.Methods The body mass index(BMI),waist circumference(WC),waist⁃hip ratio(WHR)and waist⁃to⁃height ratio(WHtR)of 3,216 adult residents in the survey area were established a logistic regression model.Analyze the ROC curve and compare the area under each ROC curve(AUC).Results Abnormal BMI,WC,WHR and WHtR were all risk factors for hypertension.For male,the OR(95%CI)were 1.46(95%CI:0.98~1.89),1.72(95%CI:1.10~2.52),1.81(95%CI:1.18~2.62),2.05(95%CI:1.40~3.05),and for female,the OR(95%CI)were 1.84(95%CI:1.34~2.52),1.57(95%CI:1.08~2.43),2.02(95%CI:1.24~3.10),1.86(95%CI:1.21~2.78)respectively.WHtR presented the largest AUC in both males and females,reaching 0.685 and 0.718,respectively.AUC increased with the increase of the number of combined indicators,indicating a high accuracy.Among both males and females,the AUC of BMI&WC&WHtR was the largest,reaching 0.703 and 0.740,respectively.Conclusions Among the single indicators,WHtR was the optimal indicator for predicting the risk of hypertension incidence,however,its accuracy was low.The combination of BMI&WC&WHtR could be the optimal predictor for the risk of hypertension.WHtR being an indicator for combination could improve the ability to predict hypertension.