Validity and reliability of a new whole room indirect calorimeter to assess metabolic response to small calorie loads.

We overview of our whole room indirect calorimeter (WRIC), demonstrate validity and reliability of our WRIC, and explore a novel application of Bayesian hierarchical modeling to assess responses to small carbohydrate loads. To assess WRIC validity seven gas infusion studies were performed using a gas blender and profiles designed to mimic resting and postprandial metabolic events. Sixteen participants underwent fasting and postprandial measurements, during which they consumed a 75-kcal drink containing sucrose, dextrose, or fructose in a crossover design. Linear mixed effects models were used to compare resting and postprandial metabolic rate (MR) and carbohydrate oxidation. Postprandial carbohydrate oxidation trajectories for each participant and condition were modeled using Bayesian Hierarchical Modeling. Mean total error in infusions were 1.27 ± 0.67% and 0.42 ± 0.70% for VO2 and VCO2 respectively, indicating a high level of validity. Mean resting MR was similar across conditions ([Formula: see text] = 1.05 ± 0.03 kcal/min, p = 0.82, ICC: 0.91). While MR increased similarly among all conditions (~13%, p = 0.29), postprandial carbohydrate oxidation parameters were significantly lower for dextrose compared with sucrose or fructose. We provide evidence validating our WRIC and a novel application of statistical methods useful for research using WRIC.

Validity and reliability of a new whole room indirect calorimeter to assess metabolic response to small-calorie loads.

Objective: To provide an overview of our whole room indirect calorimeter (WRIC), demonstrate validity and reliability of our WRIC, and explore a novel application of Bayesian hierarchical modeling to assess responses to small carbohydrate loads. Methods: Seven gas infusion studies were performed using a gas blender and profiles designed to mimic resting and postprandial metabolic events to assess WRIC validity. In a crossover design, 16 participants underwent fasting and postprandial measurements, during which they consumed a 75-kcal drink containing sucrose, dextrose, or fructose. Linear mixed effects models were used to compare resting and postprandial metabolic rate (MR) and CO (CO). Bayesian Hierarchical Modeling was also used to model postprandial CO trajectories for each participant and condition. Results: Mean total error in infusions were 1.27 ± 1.16% and 0.42 ± 1.21% for VO2 and VCO2 respectively, indicating a high level of validity. Mean resting MR was similar across conditions (x¯=1.05±0.03 kcal/min, p=0.82, ICC: 0.91). While MR increased similarly among all conditions (~13%, p=0.29), postprandial CO parameters were significantly lower for dextrose compared with sucrose or fructose. Conclusions: Our WRIC validation and novel application of statistical methods presented here provide important foundations for new research directions using WRIC.