腰围和腰围身高比预测中心性肥胖的效果与差异

Contrast and efficacy of waist circumference and waist-to-height ratio in predicting central obesity

目的评价用腰围（WC）和腰围身高比（wHtR）预测中国中年人群中心性肥胖的效果和差异。方法选用国家“八五”和“九五”攻关课题的两次横断面调查共30630名35～59岁人群资料，按不同身高分组，根据诊断中心性肥胖的WC切点（男性≥85cm，女性≥80cm）和WHtR的切点（≥0．50），比较预测中心性肥胖的一致性，并计算心血管病危险因素（包括高血压、空腹血糖异常、血清总胆固醇升高及血清高密度脂蛋白胆固醇降低）聚集（危险因素个数≥2）的灵敏度和特异度等诊断学指标。结果WC与WHtR在男女性人群中预测中心性肥胖的一致性较高，Kappa值分别为0．805和0．816。但在身高较低（男性〈160cm，女性〈150cm）或身高较高（男性≥180cm，女性≥170cm）人群中，两指标预测的一致性较差（Kappa值均≤O．6）。以WC诊断切点预测危险因素聚集，在男女性身高较低人群中，其灵敏度均较低，而在身高较高的人群中，特异度均较低；而用WHtR诊断切点预测危险因素聚集，在不同身高组中预测的灵敏度（男性为56．1％-64．1％，女性为64．7％～73．2％）和特异度（男性为70．0％～74．5％，女性为59．2％～75．9％）均较好，且波动范围较小。结论在全人群中，男女性WC和WHtR对预测心血管病危险因素聚集，评价中心性肥胖预测效果相当；但在身高较低或较高的人群中，WHtR的预测效果优于WC。

Objective To study the efficacy of waist circumference （WC） and waist-to- height ratio （WHtR） in predicting central obesity among the Chinese adult population. Methods A total of 30 630 participants aged 35-59 from different areas in China's Mainland were surveyed for the risk factors of cardiovascular diseases （CVD） in two independent cross-sectional studies that were carried out in 1992-1994 and 1998, respectively. In subgroups with different heights, consistency analysis for central obesity diagnosed by WHtR （≥0.50） and WC （ ≥85 cm for men, ≥80 cm for women） were conducted. Sensitivity and specificity for predicting the clustering of risk factors （number ≥2） would include hypertension, abnormal glucose, high serum total cholesterol and low serum high density lipoprotein cholesterol and they were also calculated to evaluate the efficacy of prediction, with the two indices in the different height subgroups as well. Results The consistency of diagnosis on central obesity by WC and WHtR was good in the whole population （the Kappa value was 0.805 in men and 0.816 in women） ,but poor （all Kappa values≤0.6） for those with tall （men＇ s height ≥ 180 cm and women＇ s height≥170 cm） or with short statures （men＇ s height 〈 160 cm, and women＇s height 〈 150 cm）. Sensitivity in the shorty subgroups and specificity in the tall subgroups appeared poor in both genders, by using WC criteria to predict the clustering of risk factors. However, the sensitivity （ranged from 56.1% to 64.1% for men and 64.7% to 73.2% for women） and specificity （from 70.0% to 74.5% for men, 59.2% to 75.9% for women） seemed good and stable in all the subgroups as well as in both genders by using the WHtR criteria. Conclusion WC and WHtR could both be applied in predicting the clustering of risk factors of CVD and in evaluating the central obesity in the whole population. With satisfactory efficacy, WHtR seemed to be better than WC in the prediction of central obesity ,both in men or women with tall or short statures.