bs_bs_banner
Journal of Diabetes 7 (2015) 386–392
O R I G I N A L A RT I C L E
Evaluation of different obesity indices as predictors of type 2 diabetes mellitus in a Chinese population Xianchao XIAO,1 Yujia LIU,1 Chenglin SUN,1 Xiaokun GANG,1 Jie CHENG,1 Suyan TIAN,2 Ying GAO,1 You LV,1 Zhonghua SUN,1 Yazhen LI,1 Ping HE,3 Yang LIU,1 Gang WANG,1 Yuan GAO,1 Liwen ZHU,1 Yan LIU1 and Guixia WANG1 Departments of 1Endocrinology and Metabolism, 2Epidemiology and 3Gastroenterology, the First Hospital of Jilin University, Changchun, Jilin, China
Correspondence Guixia Wang, Department of Endocrinology and Metabolism, the First Hospital of Jilin University, 71 Xinmin St. Changchun, Jilin, China. Tel: +86 431 8878 2866 Fax: +86 431 8878 6066 Email: [email protected] Received 17 June 2014; revised 1 July 2014; accepted 21 July 2014. doi: 10.1111/1753-0407.12201
Abstract Background: The aim of the present study was to compare correlation coefficients between anthropometric indices and blood glucose level, and to determine optimal cutoff points of obesity indices in a Chinese population. Methods: 2419 males and 5140 females participated in this study in Changchun city. Weight, height, WC (waist circumference) and hip circumference were measured; and BMI (body mass index), WHtR (waist height ratio), WHR (waist hip ratio) and BAI (body adiposity index) were calculated by formulas. FPG (fasting plasma glucose), PPG (post-load plasma glucose), HbA1c (glycosylated hemoglobin) and FSI (fasting serum insulin) were measured. Results: WHtR had the highest partial correlation coefficients with blood glucose and HbA1c in both males and females; however, BAI had the lowest coefficients. AUC (area under receiver operating characteristic curves) for WHtR in screening type 2 diabetes mellitus were higher than other indices. We confirmed that the optimal cutoff points of obesity for WHtR, WHR, WC, BMI and BAI were 0.55, 0.92, 88.1, 25.9 and 28.1 in males, and 0.52, 0.85, 81.6, 24.6 and 31.0 in females, respectively. Conclusions: Central obesity correlates more closely to glucose metabolism in Chinese population than general obesity. WHtR is the best index to predict type 2 diabetes mellitus in Chinese adults. Keywords: diabetes mellitus, obesity, receiver operating characteristic curve, type 2.
Significant findings of the study: Waist to height ratio is a better obesity index predicting type 2 diabetes mellitus in a Chinese population. What this study adds: Optimal cutoff points for waist to height ratio are 0.55 and 0.52 in men and women respectively in Chinese adults.
Introduction The prevalence of overweight and obesity in adults reached 34.4% and 12.0%, respectively worldwide in 2008.1 In China, the China National Diabetes and Metabolic Disorders Study Group reported that this prevalence reached 31.4% and 12.2% according to diagnostic 386
criteria of the Working Group on Obesity.2 Anthropometric indicators measurements, such as body mass index (BMI), waist circumference (WC), waist hip ratio (WHR) and waist to height ratio (WHtR) are costeffective methods to screen obesity. Usually, BMI represents general obesity, and others are used to analyze central obesity. Lately, Richard defined a new parameter
© 2014 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd
X. XIAO et al.
named body adiposity index to reflect percentage of body fat more accurately for Mexican American and African American adults.3 But Pavel et al. demonstrated that BAI was not valid enough to take the place of BMI to assess body fat mass in a Caucasian population.4 It reflected that ethnic differences exist in the relationship of anthropometric measurements and obesity. Obesity is associated with cardiovascular diseases, musculoskeletal disorders, cancers, and significantly with diabetes. The risk of type 2 diabetes in adults increases correspondingly with increasing obesity, and decreases with weight loss.5 A meta-analysis performed with 32 studies showed that BMI, WC, and WHR had similar correlations with diabetes incidences.6 However, in recent years, more and more studies have indicated that central adiposity might be a better predictor for diabetes mellitus. In the British Women’s Heart and Health study, measurements of central adiposity were more strongly associated with diabetes than BMI.7 In Asian studies, WHtR was more effective to screen diabetes than BMI.8,9 Another meta analysis supported the superiority of indices of central obesity over BMI for detecting cardiovascular risk factors, like diabetes.10 Until now, it is still controversial which anthropometric measurement of obesity is the most important predictor of risk of type 2 diabetes mellitus. Because of different ethnicities, the body composition of the Chinese population may be not the same as other countries. With the recent high prevalence of type 2 diabetes in China,11 specific studies should be focused on diabetic risk of obesity in the Chinese population. Between June 2011 and January 2012, a cross sectional epidemiological survey was completed in a Chinese elderly population in Changchun, China. In this study, we aimed to analyze the correlations of different body fat indices with type 2 diabetes mellitus. Further, we determined the optimal cutoff points of the better predicting index according to their risk for diabetes. Methods The present work was one part of the baseline survey from the REACTION study investigating the association of diabetes and cancer, which was conducted among 259 657 adults, aged 40 years and older in 25 communities across mainland China, from 2011 to 2012.12–14 A total of 8062 permanent residents, ≥40 years of age and without diagnosed diabetes mellitus, were randomly recruited to participate in this investigation in six communities in Changchun, China. Exclusion criteria were acute cardiovascular accidents, serious trauma, infection, and history of blood disorders. Complete data including anthropometric and plasma glucose data were
Comparing obesity indices
available for 7599 adults, including 2419 males and 5140 females. The Institutional Ethics Committee at the First Hospital of Jilin University approved the study protocol. Written informed consent was obtained from each participant before data collection. Processes, including anthropometry measurements, questionnaire, and blood drawing, of this survey were all conducted in community health service centers. Participants came to the health service centers after an overnight fasting of 8–14 h. Participants wore light clothes, without shoes when anthropometric indices were obtained. Weight, height, WC and hip circumference were measured twice with the average taken, and these measurements were finished by the same examiner. Body weight was measured on an electronic body scale to the nearest 0.1 kg. Height, WC, and hip circumference were measured to the nearest 0.5 cm in standing position.15 Waist measurements were made at the level of the umbilicus at the end of a normal expiration and hip circumference measurements were taken around the widest portion of the buttocks. We calculated BMI as body weight (kg) divided by squared height (m), WHR as WC (cm) divided by hip circumference (cm), WHtR as WC (cm) divided by height (cm), and BAI as (hip circumference (cm)/height1.5 (m) −18).3 Blood pressure was measured by electronic sphygmomanometer (OMRON, HEM-7220, Liaoning, China), which would not be recorded until both systolic pressure and diastolic pressure were close to 5 mmHg between two measurements. Investigators completed questionnaires by asking participants questions about demographic characteristics, disease history, and family history. All surveys were performed by trained staff. Each participant accepted an oral glucose tolerance test (OGTT). After fasting, venous blood samples were drawn, and they drank 300 mL water containing 75 g glucose. A second venous blood sample was drawn 2 h after glucose load. Fasting plasma glucose (FPG) and 2 h post-load plasma glucose (2 h PPG) concentrations were measured by the glucose oxidase method (Bayer ADVIA2400, Germany) in the Center laboratory of the First Hospital of Jilin University on the same day. HbA1c was measured by using the high performance liquid chromatography method (Bio-Rad Laboratories, Hercules, CA, USA), and fasting serum insulin (FSI) concentration was measured by electrochemiluminescence assay (Roche Diagnostics, Basel, Switzerland), which were both completed in Shanghai Institute of Endocrinology and Metabolism.16 Homeostasis model assessment of insulin resistance (HOMA-IR) was calculated by formula: FSI (μU/mL)/ (22.5e-ln FPG (mmol/L)).17 Diabetes mellitus was diagnosed as FPG ≥7.0 mmol/L, and/or 2 h PPG ≥11.1 mmol/L, according to a 1999
© 2014 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd
387
Comparing obesity indices Table 1
X. XIAO et al.
Baseline characteristics for males and females
Age (years) Systolic pressure (mmHg) Diastolic pressure (mmHg) Weight (kg) Height (cm) Waist circumference (cm) Hip circumference (cm) BMI (kg/m2) WHR (cm/cm) WHtR (cm/cm) BAI (cm/m1.5) FPG (mmol/L) 2h PPG (mmol/L) HbA1c (%) FSI (μU/mL) HOMA-IR
Males (n = 2419)
Females (n = 5140)
P-value
57.9 ± 10.6 143.1 ± 21.0 83.5 ± 11.7 72.3 ± 11.3 168.4 ± 6.1 87.6 ± 9.3 98.9 ± 7.2 25.5 ± 3.4 0.886 ± 0.056 0.521 ± 0.054 27.3 ± 3.5 5.90 ± 1.34 7.64 ± 3.36 5.84 ± 0.79 8.63 ± 6.16 2.32 ± 1.89
56.0 ± 9.8 138.2 ± 22.0 78.8 ± 11.8 61.9 ± 9.1 158.0 ± 5.8 82.1 ± 8.9 96.7 ± 6.8 24.8 ± 3.3 0.849 ± 0.060 0.521 ± 0.060 30.8 ± 4.1 5.70 ± 1.10 7.66 ± 3.02 5.87 ± 0.73 9.62 ± 5.98 2.51 ± 2.14
Journal of Diabetes 7 (2015) 386–392
O R I G I N A L A RT I C L E
Evaluation of different obesity indices as predictors of type 2 diabetes mellitus in a Chinese population Xianchao XIAO,1 Yujia LIU,1 Chenglin SUN,1 Xiaokun GANG,1 Jie CHENG,1 Suyan TIAN,2 Ying GAO,1 You LV,1 Zhonghua SUN,1 Yazhen LI,1 Ping HE,3 Yang LIU,1 Gang WANG,1 Yuan GAO,1 Liwen ZHU,1 Yan LIU1 and Guixia WANG1 Departments of 1Endocrinology and Metabolism, 2Epidemiology and 3Gastroenterology, the First Hospital of Jilin University, Changchun, Jilin, China
Correspondence Guixia Wang, Department of Endocrinology and Metabolism, the First Hospital of Jilin University, 71 Xinmin St. Changchun, Jilin, China. Tel: +86 431 8878 2866 Fax: +86 431 8878 6066 Email: [email protected] Received 17 June 2014; revised 1 July 2014; accepted 21 July 2014. doi: 10.1111/1753-0407.12201
Abstract Background: The aim of the present study was to compare correlation coefficients between anthropometric indices and blood glucose level, and to determine optimal cutoff points of obesity indices in a Chinese population. Methods: 2419 males and 5140 females participated in this study in Changchun city. Weight, height, WC (waist circumference) and hip circumference were measured; and BMI (body mass index), WHtR (waist height ratio), WHR (waist hip ratio) and BAI (body adiposity index) were calculated by formulas. FPG (fasting plasma glucose), PPG (post-load plasma glucose), HbA1c (glycosylated hemoglobin) and FSI (fasting serum insulin) were measured. Results: WHtR had the highest partial correlation coefficients with blood glucose and HbA1c in both males and females; however, BAI had the lowest coefficients. AUC (area under receiver operating characteristic curves) for WHtR in screening type 2 diabetes mellitus were higher than other indices. We confirmed that the optimal cutoff points of obesity for WHtR, WHR, WC, BMI and BAI were 0.55, 0.92, 88.1, 25.9 and 28.1 in males, and 0.52, 0.85, 81.6, 24.6 and 31.0 in females, respectively. Conclusions: Central obesity correlates more closely to glucose metabolism in Chinese population than general obesity. WHtR is the best index to predict type 2 diabetes mellitus in Chinese adults. Keywords: diabetes mellitus, obesity, receiver operating characteristic curve, type 2.
Significant findings of the study: Waist to height ratio is a better obesity index predicting type 2 diabetes mellitus in a Chinese population. What this study adds: Optimal cutoff points for waist to height ratio are 0.55 and 0.52 in men and women respectively in Chinese adults.
Introduction The prevalence of overweight and obesity in adults reached 34.4% and 12.0%, respectively worldwide in 2008.1 In China, the China National Diabetes and Metabolic Disorders Study Group reported that this prevalence reached 31.4% and 12.2% according to diagnostic 386
criteria of the Working Group on Obesity.2 Anthropometric indicators measurements, such as body mass index (BMI), waist circumference (WC), waist hip ratio (WHR) and waist to height ratio (WHtR) are costeffective methods to screen obesity. Usually, BMI represents general obesity, and others are used to analyze central obesity. Lately, Richard defined a new parameter
© 2014 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd
X. XIAO et al.
named body adiposity index to reflect percentage of body fat more accurately for Mexican American and African American adults.3 But Pavel et al. demonstrated that BAI was not valid enough to take the place of BMI to assess body fat mass in a Caucasian population.4 It reflected that ethnic differences exist in the relationship of anthropometric measurements and obesity. Obesity is associated with cardiovascular diseases, musculoskeletal disorders, cancers, and significantly with diabetes. The risk of type 2 diabetes in adults increases correspondingly with increasing obesity, and decreases with weight loss.5 A meta-analysis performed with 32 studies showed that BMI, WC, and WHR had similar correlations with diabetes incidences.6 However, in recent years, more and more studies have indicated that central adiposity might be a better predictor for diabetes mellitus. In the British Women’s Heart and Health study, measurements of central adiposity were more strongly associated with diabetes than BMI.7 In Asian studies, WHtR was more effective to screen diabetes than BMI.8,9 Another meta analysis supported the superiority of indices of central obesity over BMI for detecting cardiovascular risk factors, like diabetes.10 Until now, it is still controversial which anthropometric measurement of obesity is the most important predictor of risk of type 2 diabetes mellitus. Because of different ethnicities, the body composition of the Chinese population may be not the same as other countries. With the recent high prevalence of type 2 diabetes in China,11 specific studies should be focused on diabetic risk of obesity in the Chinese population. Between June 2011 and January 2012, a cross sectional epidemiological survey was completed in a Chinese elderly population in Changchun, China. In this study, we aimed to analyze the correlations of different body fat indices with type 2 diabetes mellitus. Further, we determined the optimal cutoff points of the better predicting index according to their risk for diabetes. Methods The present work was one part of the baseline survey from the REACTION study investigating the association of diabetes and cancer, which was conducted among 259 657 adults, aged 40 years and older in 25 communities across mainland China, from 2011 to 2012.12–14 A total of 8062 permanent residents, ≥40 years of age and without diagnosed diabetes mellitus, were randomly recruited to participate in this investigation in six communities in Changchun, China. Exclusion criteria were acute cardiovascular accidents, serious trauma, infection, and history of blood disorders. Complete data including anthropometric and plasma glucose data were
Comparing obesity indices
available for 7599 adults, including 2419 males and 5140 females. The Institutional Ethics Committee at the First Hospital of Jilin University approved the study protocol. Written informed consent was obtained from each participant before data collection. Processes, including anthropometry measurements, questionnaire, and blood drawing, of this survey were all conducted in community health service centers. Participants came to the health service centers after an overnight fasting of 8–14 h. Participants wore light clothes, without shoes when anthropometric indices were obtained. Weight, height, WC and hip circumference were measured twice with the average taken, and these measurements were finished by the same examiner. Body weight was measured on an electronic body scale to the nearest 0.1 kg. Height, WC, and hip circumference were measured to the nearest 0.5 cm in standing position.15 Waist measurements were made at the level of the umbilicus at the end of a normal expiration and hip circumference measurements were taken around the widest portion of the buttocks. We calculated BMI as body weight (kg) divided by squared height (m), WHR as WC (cm) divided by hip circumference (cm), WHtR as WC (cm) divided by height (cm), and BAI as (hip circumference (cm)/height1.5 (m) −18).3 Blood pressure was measured by electronic sphygmomanometer (OMRON, HEM-7220, Liaoning, China), which would not be recorded until both systolic pressure and diastolic pressure were close to 5 mmHg between two measurements. Investigators completed questionnaires by asking participants questions about demographic characteristics, disease history, and family history. All surveys were performed by trained staff. Each participant accepted an oral glucose tolerance test (OGTT). After fasting, venous blood samples were drawn, and they drank 300 mL water containing 75 g glucose. A second venous blood sample was drawn 2 h after glucose load. Fasting plasma glucose (FPG) and 2 h post-load plasma glucose (2 h PPG) concentrations were measured by the glucose oxidase method (Bayer ADVIA2400, Germany) in the Center laboratory of the First Hospital of Jilin University on the same day. HbA1c was measured by using the high performance liquid chromatography method (Bio-Rad Laboratories, Hercules, CA, USA), and fasting serum insulin (FSI) concentration was measured by electrochemiluminescence assay (Roche Diagnostics, Basel, Switzerland), which were both completed in Shanghai Institute of Endocrinology and Metabolism.16 Homeostasis model assessment of insulin resistance (HOMA-IR) was calculated by formula: FSI (μU/mL)/ (22.5e-ln FPG (mmol/L)).17 Diabetes mellitus was diagnosed as FPG ≥7.0 mmol/L, and/or 2 h PPG ≥11.1 mmol/L, according to a 1999
© 2014 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd
387
Comparing obesity indices Table 1
X. XIAO et al.
Baseline characteristics for males and females
Age (years) Systolic pressure (mmHg) Diastolic pressure (mmHg) Weight (kg) Height (cm) Waist circumference (cm) Hip circumference (cm) BMI (kg/m2) WHR (cm/cm) WHtR (cm/cm) BAI (cm/m1.5) FPG (mmol/L) 2h PPG (mmol/L) HbA1c (%) FSI (μU/mL) HOMA-IR
Males (n = 2419)
Females (n = 5140)
P-value
57.9 ± 10.6 143.1 ± 21.0 83.5 ± 11.7 72.3 ± 11.3 168.4 ± 6.1 87.6 ± 9.3 98.9 ± 7.2 25.5 ± 3.4 0.886 ± 0.056 0.521 ± 0.054 27.3 ± 3.5 5.90 ± 1.34 7.64 ± 3.36 5.84 ± 0.79 8.63 ± 6.16 2.32 ± 1.89
56.0 ± 9.8 138.2 ± 22.0 78.8 ± 11.8 61.9 ± 9.1 158.0 ± 5.8 82.1 ± 8.9 96.7 ± 6.8 24.8 ± 3.3 0.849 ± 0.060 0.521 ± 0.060 30.8 ± 4.1 5.70 ± 1.10 7.66 ± 3.02 5.87 ± 0.73 9.62 ± 5.98 2.51 ± 2.14
