RESUMEN
BACKGROUND: Body mass index (BMI) for gender and age and percentage-weight-for-height (PWH) for gender and height have been frequently used to diagnose childhood obesity. However, few studies have examined the optimal cutoff of percent body fat (%BF) to predict metabolic abnormalities in obese children. METHODS: One thousand fifteen obese children aged 10 to 15 years were enrolled in this study. The %BF cutoff values were calculated by ROC analysis for metabolic abnormalities. We also calculated %BF cutoff values corresponding to a BMI of > 85th or 95th percentile for gender and age and a PWH of > 120% for gender and height based on the 2007 Korean children and adolescents growth standard chart. To define metabolic abnormalities in children, we used the criteria of metabolic syndrome recommended by International Diabetes Federation consensus for children and adolescents. RESULTS: The %BF cutoff corresponding to the group having more than one metabolic abnormality in this study population were 38.4% (sensitivity 40.1%, specificity 84.4%) in total, 38.4% (sensitivity 35.4%, specificity 84.8%) in boys and 39.5% (sensitivity 38.9%, specificity 90.9%) in girls. The %BF cutoff values corresponding to a BMI > or = 95 percentile were 38.1% (sensitivity 50.5%, specificity 72.7%), a BMI > or = 85 percentile were 34.5% (sensitivity 69.3%, specificity 74.2%), and a PWH > or = 120% were 36.4% (sensitivity 72.3%, specificity 64.4%) in total. CONCLUSION: The optimal cutoff value of percent body fat to predict metabolic abnormalities in obese children may be 38.4% (boys 38.4, girls 39.5%) and we suggest 34.5% as a cutoff value of %BF for screening childhood obesity.