浦东新区社区居民血压测量值与身体测量指标关系的横断面研究

 目的 分析上海市浦东新区社区居民身体测量指标与血压测量值之间的关系，找到与高血压最相关的指标，为高血压高危人群的筛选提供最优预测指标、方法和依据。方法 2008年4~7月期间，采用三阶段抽样方法，从上海市浦东新区随机抽取15岁及以上社区居民5 927人进行问卷调查，同时进行身高、体重、腰围、臀围及血压的测量。使用SAS 9.1和Stata 10.0软件进行数据分析。受试者操作特征（receiver operating characteristics，ROC）分析用于评估每个身体测量指标作为高血压预测因子的准确性。结果 调查对象中高血压患病率达30.25%，男性略高于女性，高龄组显著高于低龄组（P=0.000 1）。随着体质指数（body mass index，BMI）、腰围（waist circumference，WC）、腰臀比（waist-to-hip ratio，WHR）或腰围身高比（waist-to-height ratio，WHtR）的增加，各年龄组高血压患病率均呈显著上升趋势（P=0.001）。除了20岁以下的男性，高血压患者的BMI、WC、WHR和WHtR显著高于血压正常者（P<0.05）。这些测量指标与血压测量值呈显著的正相关（P<0.000 1）。不同性别和年龄组人群中，高血压与这些指标的关联强度不同，每个十分位变化的OR值介于1.13~1.46，低龄组关联更强，存在显著的交互作用（P<0.05）。BMI、WC、WHR和WHtR预测高血压的效果总体上不太理想:WHtR较好，男性受试者操作特征曲线下面积（area under the ROC curve，AUC）达0.701 1，女性为0.723 9；WHR最差，男性AUC为0.652 0，女性AUC为0.670 1。各指标在不同性别和年龄组的预测效果有较大差异，但WHtR的预测效果优于WC和WHR。结论 年龄、性别、BMI、WC、WHR和WHtR都与高血压有显著关联，但使用身体测量指标预测高血压的价值有限。在社区进行高血压高危人群的筛选时可考虑根据性别和年龄选择相应的身体测量指标。

A cross-sectional study on the relationship between anthropometric indices and blood pressures among the residents of Pudong New Area of Shanghai

Objective To evaluate the relationship between anthropometric indices and measured blood pressures in the residents of Pudong New Area of Shanghai, and to compare the predictive values of the indices in the presence of hypertension in the population. Methods A total of 5 927 eligible residents above 15 years old were selected through a three-stage sampling and interviewed between April and July, 2008. Body height, weight, waist circumference (WC) hip circumference, systolic blood pressure (SBP) and diastolic blood pressures (DBP) were measured at the interview. Data analysis was conducted by SAS 9.1 and Stata 10.0 software. Receiver operating characteristics (ROC) analysis was applied to estimate the precision of body mass index (BMI), WC, waist to hip circumference ratio (WHR) and waist to height ratio (WHtR) in predicting the presence of hypertension. Results In this sample population, 30.25% subjects were diagnosed with hypertension. The prevalence was slightly higher in the males than in the females and significantly higher in the elder groups (P<0.000 1). Except for the males under 20 years old, the prevalence was observed to increase along with the increasement of BMI, WC, WHR and WHtR (P<0.05), and a significant positive correlation was observed for measured SBP and DBP with these anthropometric indices (P<0.000 1). The subjects with hypertension had higher BMI, WC, WHR and WHtR than normotensive pepole (P<0.000 1). The associations between the indices and hypertension varied across age and sex groups, with odds ratios (ORs) for each 10% increase of the indices ranging from 1.13 to 1.46. The associations were observed more pronounced in the younger groups, with P values for the interaction tests between age and the indices less than 0.05. None of the 4 anthropometric indices performed well in predicting the presence of hypertension. The highest area under the ROC curve (AUC) was observed for WHtR (0.701 1 in male and 0.723 9 in female), whereas the lowest (AUC) was obtained for WHR (0.652 0 in male and 0.670 1 in female). Although the best anthropometric index was different in different sex and age groups, WHtR performed better than WC and WHR consistently. Conclusions Age, sex, BMI, WC, WHR and WHtR are closely related with measured BP and the presence of hypertension. However, the anthropometric indices may not function as good predictors of hypertension. Selection of anthropometric indices in screening high-risk individuals for hypertension in the communities should take individual age and sex into consideration simultaneously.