BMI Does Not Track Body Composition Changes in Overweight Children

Abstract

Childhood obesity is increasing due to a pervasive toxic environment that facilitates increased caloric intake and reduced physical activity. Childhood obesity is predictive of type 2 diabetes mellitus (T2DM). Little is known about the determinant of T2DM in children, but interventions that promote physical activity and loss of adipose improve health outcomes. Historically body mass index (BMI) has been accepted as the standard clinical screening tool for children to determine their risk status. However, the relationships between BMI and laboratory measurement of body fat and lean tissue mass are not clear for overweight children. It is also unclear whether BMI can track changes in body composition for this population. PURPOSE: 1) To determine the relationships between BMI and body composition in overweight children and 2) determine if BMI can track changes in %FAT, fat mass and non-bone lean mass. METHODS: 1231 children (mean±SD) age = 11.7±2.9 yrs, BMI = 30.1±8.0, %FAT = 37.5±6.8%, lean body mass = 36.3±14.0kg, fat mass = 33.9 ± 15.1 kg were evaluated. Criterion body fat was measured by DXA. Whole body scans were performed with a Norland XR-36 bone densitometer before and after a 1 year intervention designed to improve body composition. RESULTS: For female subjects at baseline (n=642, 52%) the correlation between BMI and %FAT and non-bone lean mass were r = 0.61 and r = 0.65, respectively. For male subjects at baseline (n=588, 48%) the correlation between BMI and %FAT and non-bone lean mass were r = 0.50 and r = 0.69, respectively. For the second objective, 375 subjects that were re-scanned following a 1year intervention promoting increased activity and healthy eating, there was no difference in BMI Z score (p=0.81) and BMI percentile for age (p=0.59). However, there was a significant decrease in %FAT (p< 0.001), fat mass (p< 0.001) and increase in non-bone lean mass (p< 0.001) for the whole body and the abdominal region (p=0.006). CONCLUSIONS: The correlations between BMI and %FAT and non-bone lean mass support the relationships at baseline. BMI actually had a stronger correlation with lean mass than with fat mass. However, the results indicate that actual decreases in %FAT and fat mass, accompanied by increases in non-bone lean mass resulting from a 1 year intervention were not reflected or supported by changes in BMI or BMI Z-score.