Computer modeling of obesity links theoretical energetic measures with experimental measures of fuel use for lean and obese men.

The goal of this research was to use a computational model of human metabolism to predict energy metabolism for lean and obese men. The model is composed of 6 state variables representing amino acids, muscle protein, visceral protein, glucose, triglycerides, and fatty acids (FAs). Differential equations represent carbohydrate, amino acid, and FA uptake and output by tissues based on ATP creation and use for both lean and obese men. Model parameterization is based on data from previous studies. Results from sensitivity analyses indicate that model predictions of resting energy expenditure (REE) and respiratory quotient (RQ) are dependent on FA and glucose oxidation rates with the highest sensitivity coefficients (0.6, 0.8 and 0.43, 0.15, respectively, for lean and obese models). Metabolizable energy (ME) is influenced by ingested energy intake with a sensitivity coefficient of 0.98, and a phosphate-to-oxygen ratio by FA oxidation rate and amino acid oxidation rate (0.32, 0.24 and 0.55, 0.65 for lean and obese models, respectively). Simulations of previously published studies showed that the model is able to predict ME ranging from 6.6 to 9.3 with 0% differences between published and model values, and RQ ranging from 0.79 to 0.86 with 1% differences between published and model values. REEs >7 MJ/d are predicted with 6% differences between published and model values. Glucose oxidation increases by ∼0.59 mol/d, RQ increases by 0.03, REE increases by 2 MJ/d, and heat production increases by 1.8 MJ/d in the obese model compared with lean model simulations. Increased FA oxidation results in higher changes in RQ and lower relative changes in REE. These results suggest that because fat mass is directly related to REE and rate of FA oxidation, body fat content could be used as a predictor of RQ.