Stephen D. Nugent, Gilbert R. Kaats &Harry G. Preuss
Journal of the American College of Nutrition Published online
February 9, 2018
Objective: A general assumption is that the body mass index (BMI) reflects changes in fat mass (FM). However, it fails to distinguish the type of weight that is lost or gained—fat mass (FM) or fat-free mass (FFM). The BMI treats both changes the same although they have opposite health consequences. The objective of this study was to propose a more precise measure, a body composition change index (BCCI), which distinguishes between changes in FM and FFM, and this study compares it with using the BMI as an outcome measure.
Methods: Data were obtained from 3,870 subjects who had completed dual-energy x-ray absorptiometry (DEXA) total body scans at baseline and end-of-study when participating in a variety of weight-loss interventions. Since height remained constant in this adult cohort, changes in the BMI corresponded with scale weight changes (r = 0.994), allowing BMI changes to be converted to “lbs.” to match the statistic used for calculation of the BCCI. The BCCI is calculated by scoring increases in FFM (lbs.) and decreases in FM (lbs.) as positive outcomes and scoring decreases in FFM and increases in FM as negative outcomes. The BCCI is the net sum of these calculations. Differences between scale weight changes and BCCI values were subsequently compared to obtain “discordance scores.”
Results: Discordance scores ranged from 0.0 lbs. to >30.0 lbs. with a mean absolute value of between the two measures of 7.79 lbs. (99% confidence interval: 7.49-8.10, p <0.00001), SD = 7.4 lbs. Similar discordance scores were also found in subgroups of self-reported gender, ethnicity, and age.
Conclusions: A significant difference of 7.79 lbs. was found between the BCCI and the BMI to evaluate the efficacy of weight loss interventions. If assessing changes in body composition is a treatment goal, use of the BMI could result in significantly erroneous conclusions.