Using the Population Approach to Obtain Individual Estimates From Indirect Calorimetry

Abstract

Quantification of human energy encompasses several domains of research. Consumption of oxygen and release of carbon dioxide are used as an indirect measure of energy output. The goal of this research was to understand if mixed effects modeling could be used to identify structural parameter estimates and the covariates that help explain population variance from measures of indirect calorimetry. Indirect calorimetry was assessed in 160 individuals stratified by gender, aged 18‐30, and with BMI of 20‐25. Measurements were taken for 20 minutes at rest, during a stationary cycling to obtain maximum VO2, and at a load of 125 Watts on stationary cycle for 60 minutes. In the instances of rest and a 125 Watt load, O2, CO2, and the respiratory quotient were modeled as stationary with time. The covariates of gender and weight both explain relevant variance. During the maximum VO2 test, O2 and CO2 were described with a linear model with slope and intercept and included between subject variability, and the respiratory quotient was described as an EMAX model with baseline, EMAX asymptote, Hill Coefficient, and slope. This effort determined a mixed effects modeling approach could be used to assess indirect calorimetry measures. Future studies with additional covariates such as direct measure of body composition may be informative. A design focused on a broad general population to potentially identify sub‐groups would be ideal.