A comparison of fat and lean body mass index to BMI for the identification of metabolic syndrome in children and adolescents

Abstract

Context: The use of body mass index (BMI) to assess risk for cardiometabolic disease in the pediatric population may be limited by a failure to differentiate between fat and lean body mass.

Objectives: The objectives of the study were to identify biologically based criteria for the definition of obesity using fat (FMI) and lean body mass index (LBMI) and to compare the ability of FMI and LBMI to BMI to identify the presence of metabolic syndrome (MetSyn).

Design: This was a cross-sectional study using National Health and Nutrition Examination Survey 1999-2006 data.

Participants: A total of 3004 participants aged 12-20 years with dual-energy X-ray absorptiometry body composition and fasting laboratory data participated in the study.

Main outcome measures: Adjusted odds ratios for MetSyn according to FMI and LBMI status and area under the curve for the identification of MetSyn were measured.

Results: Receiver-operating characteristic curve analyses identified the 80th percentile for FMI and the 74th percentile for LBMI as the optimal cut points for the identification of MetSyn. There was no difference in the area under the curve for FMI [0.867; 95% confidence interval (CI) 0.838-0.891] vs BMI (0.868; 95% CI 0.837-0.894) Z-scores for MetSyn discrimination. Separate multivariate regression models identified odds ratios for the identification of MetSyn of 6.2 (95% CI 3.3-11.5) for BMI-Z, 6.4 (95% CI 3.7-11.1) for FMI-Z, and 4.6 (95% CI 3.0-7.1) for LBMI-Z. Models containing both FMI-Z and LBMI-Z revealed that greater LBMI-Z was associated with greater odds of low high-density lipoprotein (1.5; 95% CI 1.2-1.9), high blood pressure (1.8; 95% CI 1.1-2.9), and insulin resistance (1.8; 95% CI 1.4-2.5), independent of FMI-Z.

Conclusions: The use of FMI and LBMI does not improve upon BMI for the identification of MetSyn in the pediatric population. Unexpectedly, higher LBMI was associated with greater odds of multiple cardiometabolic risk factors independent of FMI. The use of FMI and LBMI allow for the independent evaluation of relationships between body compartments and disease and warrants future research.

Figure 1.

Adjusted odds ratios of body composition measures for the identification of cardiometabolic risk factors in 3004 NHANES participants, with a mean age of 16.1 years. All models are adjusted for sex, age, height, and racial group, and models including one or more than one body composition measure are shown. Exact odds ratio and 95% CI as well as AUC values for the models depicted above are provided in Supplemental Table 2.