Simulating changes in regional body composition with weight change using STELLA

Abstract

With the increase in human obesity, it would be beneficial to investigate contributive effects of diet and physical activity to changes in regional body composition. We used STELLA to develop and test mathematical models of energy balance (energy intake vs energy expenditure), body weight, and the fat and lean composition of 8 regions of the body (head, upper arms, lower arms, upper trunk-chest, mid trunk-waist, lower trunk-hips/buttocks, upper legs, and lower legs). The effect of weight change on regional body composition was predicted and assessed using the models. The efficacy of using subject-reported energy intake and expenditure values to predict weight change was investigated as well as methods that might be used to correct for errors in self-reported data. Changes in body composition were best simulated when holding the percentage of body fat in a particular region constant. Correction for self-reported energy values was best accomplished by calculating a correction factor that matched reported body weight changes during a portion of the study period and using that correction factor in simulations for the remaining study period. Supported by a Brigham Young University Mentored Environment Grant.