Comparison of methods to evaluate changes in relative body mass index in pediatric weight control. |
| |
Authors: | Rocco A Paluch Leonard H Epstein James N Roemmich |
| |
Affiliation: | Department of Pediatrics, School of Medicine and Biomedical Sciences, State University of New York at Buffalo (SUNY), Buffalo, New York 14214-3000, USA. |
| |
Abstract: | Our objective was to compare three BMI-based adiposity measures to assess change in pediatric weight control: LMS z-BMI, BMI sympercent, and percent overBMI. Comparison 1 presents changes of +1.0, -1.0, and -2.0 BMI units for 36 hypothetical children (7-, 11-, and 15-year-old children with BMI values from 19-39 kg/m(2)). Comparison 2 presents effect sizes over 12 months and the relationship between baseline and change for 140 8-12-year-old children with BMI values ranging from 21 to 37 kg/m(2). Comparison 1 showed smaller changes in z-BMI than BMI sympercent or percent overBMI for equal changes in BMI when initial BMI values are greater. Comparison 2 showed similar effect sizes for the three measures, since there is a reduction in both standard deviation and magnitude of LMS z-BMI values as the BMI values increase. The three measures perform differently when considering the relationship of initial value to change. Initial percent overBMI shows a negative relationship with change, as heavier children show larger changes, LMS z-BMI shows a positive relationship, as children with lower baseline values show larger changes, and BMI sympercent changes were inconsistently related to baseline BMI sympercent values. Although all three measures result in similar effect sizes when evaluating treatment over time, we conclude that LMS z-BMI is less appropriate for comparing individuals and percent overBMI is the only measure that shows heavier children have greater change. |
| |
Keywords: | |
|
|