Fat and fat-free mass index references in children and young adults: Assessments along racial and ethnic lines

Authors: SHYPAILO, ROMAN - Children'S Nutrition Research Center (CNRC); WONG, WILLIAM - Children'S Nutrition Research Center (CNRC)

Submitted to: The American Journal of Clinical Nutrition
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/7/2020
Publication Date: 5/29/2020
Citation: Shypailo, R.J., Wong, W.W. 2020. Fat and fat-free mass index references in children and young adults: Assessments along racial and ethnic lines. American Journal of Clinical Nutrition. https://doi.org/10.1093/ajcn/nqaa128.
DOI: https://doi.org/10.1093/ajcn/nqaa128

Interpretive Summary: Growth references are important tools in monitoring health and nutrition in children. BMI, a commonly used obesity reference, can often be misleading, since it is an index of body weight, not body fat. Fat-Free Mass Index (FFMI, fat-free mass divided by height squared) and Fat Mass Index (FMI, fat mass divided by height squared) are thought to be better health indicators. Race/ethnicity also influences body composition. Thus, our goal was to construct growth curves for FFMI and FMI for children based on race and ethnicity. We measured 1122 normal healthy children aged 2–21 y. Race/ethnicity for each child was recorded as black (B), white (W), or Hispanic (H). FFMI was similar for W and H children but was higher in B children. H boys had higher FMI. FMI in girls, though similar, was different among all three race/ethnicities. Race/ethnicities that did not differ in FFMI or FMI were combined to create growth models based on age. We then compared our growth curves with data published by other investigators. FFMI curves for W/H boys and girls agreed well with models based on children in the United Kingdom (UK), as did FMI curves for W girls and W/B boys. This showed the importance of race/ethnicity, since the UK children were predominantly W. Our curves did not agree well with curves based on NHANES data (U.S. National Health and Nutrition Examination Survey) that were generic in race/ethnicity. This could be explained by two factors: the NHANES curves did not include bone mineral in FFM, and the NHANES curves were adjusted to lower FFM. Once these factors were corrected, the NHANES curves seemed to agree with an average of our racial/ethnic groups. The use of FFMI and FMI overcomes the limitations of BMI by looking specifically at lean and fat mass. Racial/ethnic differences are evident in certain groups in both sexes. Models specific for race/ethnicity should be used when evaluating growth in children.

Technical Abstract: Fat-free mass index (FFMI) and fat mass index (FMI) are superior to BMI and fat percentage in evaluating nutritional status. However, existing references fail to account for racial/ethnic differences in body composition among children. Our goal was to produce age-based normative references for FFMI and FMI in children for specific racial/ethnic groups. Body composition, weight, and height were measured in 1122 normal healthy children aged 2-21 y. Bone mineral content measured by DXA, total body water by deuterium dilution, and total body potassium by whole-body ' counting were combined to calculate fat-free mass (FFM) and fat mass (FM) using equations based on the Reference Child and Adolescent models. FFMI and FMI were calculated by dividing FFM and FM by height squared, respectively. After outlier removal, the LMS (Lambda-Mu-Sigma) function within R's GAMLSS package was used to produce age-based FFMI and FMI growth curves for black (B), white (W), and Hispanic (H) children for each sex. Combined models were produced in cases where outcomes did not differ by race/ethnicity. Resulting models were compared with previously published FFMI and FMI models. FFMI and FMI models based on 1079 children, aged 2-21 y, were created for both sexes. FFMI models for B children showed higher values throughout. W and H children were combined to produce FFMI models for each sex. H boys were modeled individually for FMI, whereas W and B boys were combined. FMI models for girls were created for each race/ethnicity. Models agreed well with those based on children from the United Kingdom of comparable race/ethnicity. Race/ethnicity-specific references for FFMI and FMI will increase the accuracy of health and nutrition status assessment in children over race/ethnicity-generic references. The models allow the calculation of SD scores to assess health and nutrition status in children.