BACKGROUND: Multiple screening tools, such as BMI and skinfold-derived percent body fat (%BF), are available to identify youth at risk of excess adiposity. However, poor classification agreement among these tools can be problematic for those interpreting test results. PURPOSE: The purpose of this study was to investigate the validity of using BMI as an estimate of %BF in youth and to identify optimal BMI thresholds for identifying at-risk children and adolescents based on %BF. METHODS: Percent body fat was derived from the skinfold thicknesses of children aged 5-18 years from three cross-sectional waves of the National Health and Nutrition Examination Survey (NHANES [1999-2004]; N=8269). Stature and body mass from the same data set were used to calculate BMI. Receiver operating characteristic (ROC) analysis was employed to determine the optimal BMI thresholds for detecting previously created %BF standards. RESULTS: The optimal BMI percentile associated with the low risk %BF threshold was the 83 rd and 80th in boys and girls, respectively. The selected BMI percentiles associated with the higher risk threshold were the 92 nd and 90th in boys and girls, respectively. Overall, classification accuracy when using BMI percentiles to identify the two %BF risk groups ranged from 86.9% to 89.1%. CONCLUSIONS: BMI and skinfold-derived %BF demonstrate reasonable agreement when used to classify adiposity status in children and adolescents.
BACKGROUND: Multiple screening tools, such as BMI and skinfold-derived percent body fat (%BF), are available to identify youth at risk of excess adiposity. However, poor classification agreement among these tools can be problematic for those interpreting test results. PURPOSE: The purpose of this study was to investigate the validity of using BMI as an estimate of %BF in youth and to identify optimal BMI thresholds for identifying at-risk children and adolescents based on %BF. METHODS: Percent body fat was derived from the skinfold thicknesses of children aged 5-18 years from three cross-sectional waves of the National Health and Nutrition Examination Survey (NHANES [1999-2004]; N=8269). Stature and body mass from the same data set were used to calculate BMI. Receiver operating characteristic (ROC) analysis was employed to determine the optimal BMI thresholds for detecting previously created %BF standards. RESULTS: The optimal BMI percentile associated with the low risk %BF threshold was the 83 rd and 80th in boys and girls, respectively. The selected BMI percentiles associated with the higher risk threshold were the 92 nd and 90th in boys and girls, respectively. Overall, classification accuracy when using BMI percentiles to identify the two %BF risk groups ranged from 86.9% to 89.1%. CONCLUSIONS: BMI and skinfold-derived %BF demonstrate reasonable agreement when used to classify adiposity status in children and adolescents.
Authors: Andreia Pelegrini; Diego Augusto Santos Silva; João Marcos Ferreira de Lima Silva; Leoberto Grigollo; Edio Luiz Petroski Journal: Rev Paul Pediatr Date: 2015-01-31