Comparison of various anthropometric and body fat indices in identifying cardiometabolic disturbances in Chinese men and women

• Published in: PloS one 2013-08, Vol.8 (8), p.e70893-e70893
• Main Authors: Zhang, Zhe-qing, Deng, Juan, He, Li-ping, Ling, Wen-hua, Su, Yi-xiang, Chen, Yu-ming
• Format: Article
• Language: English
• Subjects: Abnormalities; Adipose tissue; Adiposity; Adult; Aged; Aged, 80 and over; Anthropometry; Bioelectricity; Biology; Blood pressure; Body composition; Body fat; Body mass; Body mass index; Body measurements; Body size; Body weight; Body Weights and Measures; Cardiovascular disease; Cardiovascular Diseases - epidemiology; Cardiovascular Diseases - metabolism; Cardiovascular Diseases - physiopathology; China - epidemiology; Cross-Sectional Studies; Dextrose; Diabetes; Diabetes mellitus; Dyslipidemia; Female; Food; Glucose; Health risks; Hip; Humans; Hypertension; Hyperuricemia; Identification methods; Laboratories; Male; Medicine; Men; Mens health; Metabolic disorders; Metabolic syndrome; Metabolism; Middle age; Middle Aged; NMR; Nuclear magnetic resonance; Nutrition; Obesity; Odds Ratio; Performance prediction; Population studies; Public health; Public Health Surveillance; Regression analysis; Studies; Systematic review; Urine; Womens health
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0070893

Although many adiposity indices may be used to predict obesity-related health risks, uncertainty remains over which of them performs best. This study compared the predictive capability of direct and indirect adiposity measures in identifying people at higher risk of metabolic abnormalities. This population-based cross-sectional study recruited 2780 women and 1160 men. Body weight and height, waist circumference (WC), and hip circumference (HC) were measured and body mass index (BMI), waist-to-hip ratio (WHR), and waist-to-height ratio (WHtR) were calculated. Body fat (and percentage of fat) over the whole body and the trunk were determined by bioelectrical impedance analysis (BIA). Blood pressure, fasting lipid profiles, and glucose and urine acid levels were assessed. In women, the ROC and the multivariate logistic regression analyses both showed that WHtR consistently had the best performance in identifying hypertension, dyslipidemia, hyperuricemia, diabetes/IFG, and metabolic syndrome (MetS). In men, the ROC analysis showed that WHtR was the best predictor of hypertension, WHtR and WC were equally good predictors of dyslipidemia and MetS, and WHtR was the second-best predictor of hyperuricemia and diabetes/IFG. The multivariate logistic regression also found WHtR to be superior in discriminating between MetS, diabetes/IFG, and dyslipidemia while BMI performed better in predicting hypertension and hyperuricemia in men. The BIA-derived indices were the second-worst predictors for all of the endpoints, and HC was the worst. WHtR was the best predictor of various metabolic abnormalities. BMI may be used as an alternative measure of obesity for identifying hypertension in both sexes.