The Interactive Association of General Obesity and Central Obesity with Prevalent Hypertension in Rural Lanzhou, China

OBJECTIVE: To evaluate the interactive association between obesity with different anthropometry indices and prevalence of hypertension in rural Lanzhou. METHODS: A cross-sectional survey was conducted in rural Lanzhou from April to July in 2013. The available information of 1275 rural residents aged more than 35 years was collected with a unified questionnaire and their blood pressure and anthropometry indices were measured in the field. The male-to-female ratio was 1:1.1. A generalized estimate equation (GEE) linear model was used to determine the association between obesity with different indexes and hypertension. RESULTS: There was a moderate prevalence of general obesity (~11%) and very high prevalence of central obesity (53.2~67%) among the adults of rural Lanzhou. The prevalence of hypertension approximated 28%. GEE linear models showed that obesity with any one of anthropometry indices was associated significantly with the increased prevalence of hypertension among both males and females. In females, general obesity increased the prevalence of hypertension by 37% (0.37, 95%CI: 0.27,0.47) but in males by 23% (0.23, 95%CI: 0.12,0.35). The hypertensive effect of all central obesity was much lower than that of general obesity but approximately comparable to that of overweight. In addition, the interactions of the classified body mass index (BMI) and central obesity showed that when general obesity or overweight coexisted with any one of central obesity, the prevalence of hypertension was increased significantly, and this effect was a little higher than the corresponding main effect of general obesity or overweight in females but was much higher in males. In addition, general obesity or overweight which did not coexist with central obesity was not significantly associated with the increased prevalence of hypertension, nor were the other situations of central obesity in the normal weight or underweight except for the situation of central obesity with waist-to-hip ratio in the males of normal weight or underweight. CONCLUSION: In rural Lanzhou, higher BMI had larger associations with the increased prevalence of hypertension than central obesity indices. Only when general obesity or overweight coexisted with central obesity, the prevalence of hypertension was significantly increased. So, central obesity indices should be used jointly with BMI in evaluating the risk of hypertension.