Body mass index percentiles versus body composition assessments: Challenges for disease risk classifications in children.

Background: Identifying at-risk children with optimal specificity and sensitivity to allow for the appropriate intervention strategies to be implemented is crucial to improving the health and well-being of children. We determined relationships of body mass indexes for age and sex percentile (BMI%) classifications to actual body composition using validated and convenient methodologies and compared fat and non-fat mass estimates to normative cut-off reference values to determine guideline reliability. We hypothesized that we would achieve an improved ability to identify at-risk children using simple, non-invasive body composition and index measures. Methods: Cross-sectional study of a volunteer convenience sample of 1,064 (537 boys) young children comparing Body Fat Percentage (BF%), Fat Mass Index (FMI), Fat-Free Mass Index (FFMI), determined via rapid bioimpedance methods vs. BMI% in children. Comparisons determined among weight classifications and boys vs. girls. Results: Amongst all subjects BMI% was generally correlated to body composition measures and indexes but nearly one quarter of children in the low-risk classifications (healthy weight or overweight BMI%) had higher BF% and/or lower FFMI than recommended standards. Substantial evidence of higher than expected fatness and or sarcopenia was found relative to risk status. Inaccuracies were more common in girls than boys and girls were found to have consistently higher BF% at any BMI%. Conclusions: The population studied raises concerns regarding actual risks for children of healthy or overweight categorized BMI% since many had higher than expected BF% and potential sarcopenia. When body composition and FMI and FFMI are used in conjunction with BMI% improved sensitivity, and accuracy of identifying children who may benefit from appropriate interventions results. These additional measures could help guide clinical decision making in settings of disease-risks stratifications and interventions.