Cross-Validation of Anthropometric Measures against Body Mass Index for the Assessment of Obesity in Children

Abstract

Background: It goes without saying that central or upper body adiposity leads to increased risks of obesity-associated metabolic complications. Direct methods of measurement of central obesity are not practical for field studies due to technical difficulties and cost. Thus, the urge to depend on anthropometry is going larger. The purposes of the study were to evaluate the relations between different anthropometry and to predict and cross-validate these measures. Methods: Subjects participating in the study included only 71 boys from different grades (aged 6-12 years). The following anthropometric measurements were measured and calculated: body mass index(BMI), waist circumference(WC), waist to hip ratio(WHR), and waist to height ratio(WHtR). Results: WC was the best single predictor of obesity, explaining 67.4% (r=0.67)of its variance while WHR explained 0.6% (r=0.006)and WHtR explained 7.3%.(r=0.07) Following quadratic regression and cross-validation techniques, it was obvious that WC better explained 77.3%(r=0.77) with p<0.05 of the variance of BMI, while WHtR explained 11.5% (r=0.11) with p<0.961, and WHR explained 16.7% (r=0.16)with p<0.546. Conclusion: Collinearity between body mass index(BMI) and waist circumference (WC) does exist, but it is recommended to use Waist circumference associated with body mass index at a young age due to the period of intense growth.