研究目的:本研究探討常見的肥胖指標與新建立的肥胖指標對於第二型糖尿病與高血壓的預測能力。 研究方法:資料來自於國民營養健康狀況變遷調查(Nutrition and Health Survey in Taiwan, NASHIT)1993-1996。比較五種肥胖指標對於第二型糖尿病與高血壓的預測能力:一種與一般肥胖有關-身體質量指數(body mass index, BMI);四種與中央肥胖有關-腰圍(waist circumference, WC)、腰臀比(waist-hip ratio, WHR)、腰高比(waist-height ratio, WHtR)及新肥胖指標-體積重量比(volume weight ratio, VWtR)。本研究利用邏輯斯回歸分析與接受者操作型曲線下面積分析探討肥胖指標對於第二型糖尿病與高血壓的預測能力。 結果:利用邏輯斯回歸分析,男性罹患糖尿病的危險性會隨著腰臀比 (waist-hip ratio, WHR) 值的增加而上升,其危險對比值 (odds ratio, OR) 為4.20 (95 % CI: 1.53 ~ 11.49) 、5.23 (95 % CI: 1.59 ~ 17.23),女性罹患糖尿病的機率會隨著體積重量比 (volume-weight ratio, VWtR) 值的增加而上升,其OR為6.15 (95 % CI: 1.52 ~ 24.96) 。在高血壓方面,無論男性或女性身體質量指數 (body mass index, BMI) 值的增加會使罹患高血壓的機率上升,其OR分別為1.72 (95 % CI: 1.15 ~ 2.57) 、2.36 (95 % CI: 1.36 ~ 4.10) 及2.15 (95 % CI: 1.41 ~ 3.27) 、2.82 (95 % CI: 1.58 ~ 5.06) 。利用接受者操作型曲線分析,WHR在男性與女性糖尿病之預測效果較高,其AUC分別為0.74 (95 % CI: 0.69 ~ 0.78) 、0.84 (95 % CI: 0.81 ~ 0.87) ,體積重量比 (volume weight ratio, VWtR) 在女性糖尿病之預測效果亦高,其曲線下面積(area under the curve, AUC)為0.84 (95 % CI: 0.81 ~ 087) ;BMI在預測男性與女性高血壓危險性中,AUC分別為0.73 (95 % CI: 0.70 ~ 0.76) 、0.80 (95 % CI: 0.78 ~ 0.83) 。 結論:WHR對於男性糖尿病有較高的危險性,VWtR對於女性糖尿病有較高的危險性,BMI對於預測男女性高血壓有較強的危險性。
Purpose: The purpose of this study was to compare the performance of measures of central, general and the proposed new index to predict the risk of type 2 diabetes and/or hypertension for Taiwanese individuals. Method: The data were obtained from the Nutrition and Health Survey in Taiwan (NASHIT) 1993-1996. We compared the predictive performance of five anthropometric indexes: one related to general obesity ─ body mass index (BMI); and four related to central obesity ─ waist circumference (WC), waist-hip ratio (WHR), waist-height ratio (WHtR), and the new obesity index (volume weight ratio, VWtR). We used the unconditional logistic regression and the area under (AUC) receiver-operating characteristic (ROC) curves to compare the performance of the indexes. Multivariate adjusted odds ratios (ORs) were estimated from the unconditional logistic regression. Results: Subjects in male with WHR value ≥ 0.91 (OR = 5.23; 95% CI: 1.59 ~ 17.23) and 0.87 ~ 0.91 (OR = 4.20; 95% CI: 1.53 ~ 11.49) showed significantly increased risk of diabetes than subjects with WHR value < 0.87 (OR = 1.00) (test for trend: p = 0.01). Subjects in female with VWtR value ≥ 1.38 (OR = 6.15; 95% CI: 1.52 ~ 24.96) showed significantly increased risk of diabetes than subjects with VWtR value < 1.22 (OR = 1.00) (test for trend: p = 0.009). Subjects in male with BMI value ≥ 25.74 (OR = 2.36; 95% CI: 1.36 ~ 4.10) and 23.39 ~ 25.74 (OR = 1.72; 95% CI: 1.15 ~ 2.57) showed significantly increased risk of hypertension than subjects with BMI value < 23.39 (OR = 1.00) (test for trend: p = 0.003). Subjects in female with BMI value ≥ 26.63 (OR = 2.82; 95% CI: 1.58 ~ 5.06) and 23.73 ~ 26.63 (OR = 2.15; 95% CI: 1.41 ~ 3.27) showed significantly increased risk of hypertension than subjects with BMI value < 23.73 (OR = 1.00) (test for trend: p < 0.001). WHR ( AUC = 0.74 for male), VWtR ( AUC = 0.84 for female) and BMI ( AUC = 0.73 for male and AUC = 080 for female) had the highest area under the ROC curve to predict diabetes and hypertension, respectively. Conclusion: Our findings suggested that WHR and VWtR could be a better predictor of the risk of diabetes for male and female, respectively. BMI could be the better predictor of the risk of hypertension for both genders.