| Background:It is generally recognized that overweight and obesity are major independent risk factors for hypertension. It remains controversial that whether waist circumference (WC), waist:hip ratio (WHR) and waist:stature ratio (WSR) predict incidence of hypertension better than body mass index (BMI) or add to the prediction of incidence of hypertension. Most studies on optimal obesity index cut-off values were based on cross-sectional samples. The association between change in BMI and incidence of hypertension remains unclear in Chinese population, and the impact of change in overweight status defined by the potential cut-off value on the incidence of hypertension has not been well understood.Objectives:To examine the relative contribution for the predictive capacity of hypertension by WC, WHR or WSR with that by BMI; To ascertain whether WC, WSR or WHR, combined with BMI, improves the prediction of incident hypertension; To determine the cut-off values of anthropometric markers for detecting incident hypertension in Chinese adults; To assess change in obesity indices values and its impact on the incidence of hypertension,and to assess the impact of change in overweight status based on this cut-off value on the incidence of hypertension in Chinese adults aged 18 to 65.Methods:Two waves of data were collected in 2006 and 2011 by the China Health and Nutrition Survey (CHNS) with samples drawn from nine provinces in China. Individual BMI, WC, WHR and WSR were assessed. To identify the obesity index that best predict incidence of hypertension, the receiver operating characteristic (ROC) curves and the area under the ROC curve (AUC) of the obesity indices in relation to incidence of hypertension were plotted and calculated and the corresponding values for the maximum of the Youden indices were considered the optimal cut-off values. The logistic regression model was used to examine the association between change in obesity indices and the incidence of hypertension, and odds ratio (OR) and 95% confident interval (95% CI) were calculated. Additionally, overweight-specific incidences of hypertension in 2006 and 2011 were further investigated using logistic regression analysis.Results:1. Five-year cumulative incidences of hypertension were 21.5%(95% CI:19.4-23.6) in men and 16.5%(95% CI:14.7-18.2) in women and there was a significant increased trend on incidence of hypertension with increasing BMI, WC, WHR or WSR (P for trend< 0.0001, respectively) in both men and women.2. Models with WHR had lower AUC compared with models with BMI for men (P < 0.001). Models with WC and WSR showed similar AUC compared with models with BMI for men (P> 0.05). Models with WC, WHR or WSR showed similar AUC compared with models with BMI for women (difference in AUC≤3%; P> 0.05). When adding WC, WHR or WSR to a model with BMI, respectively, no significant improvemet in model fit was observed (P> 0.05 in all of the models). Among men, the AUC for BMI was 0.65 and the AUCs for WC, WHR, and WSR were 0.63,0.57, and 0.63, respectively. Among women, the AUC for BMI was 0.63 and the AUCs for WC, WHR, and WSR were 0.62,0.60, and 0.63, respectively.3. For men, the Youden index indicated that the optimal BMI, WC, WHR and WSR cut-off values were 23.53 kg/m2 (sensitivity,0.555; specificity,0.676),83.7 cm (sensitivity,0.552; specificity,0.648),0.90 (sensitivity,0.442; specificity,0.699), and 0.51 (sensitivity,0.497; specificity,0.693), respectively. For women the optimal BMI, WC, WHR and WSR cut-off values were 24.25 kg/m2 (sensitivity,0.503; specificity, 0.715),79.9 cm (sensitivity,0.601; specificity,0.595),0.85 (sensitivity,0.591; specificity,0.585), and 0.52 (sensitivity,0.534; specificity,0.676), respectively. The Youden index for BMI was highest in men and women among the measured obesity indices.4. There was a significant increased trend on incidence of hypertension with increasing quartiles of the difference values in all obesity indices between 2006 and 2011 in men (P for trend< 0.05) and of the difference values in BMI, WC and WSR between 2006 and 2011 in women (P for trend< 0.05). The participants were further stratified by overweight or central obesity status in 2006 and in 2011. The ORs of incident hypertension were all significantly higher for overweight or central obesity subjects versus non-overweight or non-central obesity ones in 2011, regardless of their overweight or central obesity status at baseline (P< 0.05).Conclusions:The waist-related indices do not perform better than BMI in predicting hypertension risk, whether used alone or combined with BMI in Chinese adults. Our study suggested that the cut-off values for BMI (24 kg/m2), developed by the Working Group on Obesity in China (WGOC) and issued in the standard of WS/T 428-2013 (China), may be the appropriate upper limit for predicting incidence of hypertension in Chinese adults. The cut-off values for WC (80 cm) in women, recommended by the WGOC and the standard of WS/T 428-2013 (China), may be the appropriate upper limit for predicting incidence of hypertension, while the optimal WC cut-off value in men was 84 cm, which is lower than the recommended cut-off value for WC (85 cm). The WHR cut-off values recommended by the World Health Organization (WHO) may be the appropriate upper limits for predicting incidence of hypertension in Chinese adults. Our study also supported the hypothesis that a global WSR cut-off value of 0.50 may be the appropriate upper limit for predicting incidence of hypertension in Chinese adults. Five-year change in BMI, WC, WHR and WSR were associated with the incidence of hypertension in men, and change in BMI, WC and WSR were associated with the incidence of hypertension in women. Findings from this study provide insight into recommendation of the optimal cut-off values of multiple obesity indices for predicting incidence of hypertension in Chinese adults. It is of public health implication that effective intervention of overweight, obesity or central obesity contributes to curb hypertension epidemic among adults in China. |
PDF Full Text Request