A prospective randomised study comparing oral 13C-bicarbonate tracer technique versus indirect calorimetry for measurement of energy expenditure in adults

Abstract

Accurate assessment of energy expenditure (EE) is important in guiding nutritional therapy but current methods are unsatisfactory. This study compared the oral 13C-bicarbonate tracer (BT) technique using the IRIS® system (Wagner, Germany) against indirect calorimetry (IC, ventilated-hood) to measure CO2 output (VCO2) and thus estimate EE. Ten overnight-fasted healthy male volunteers were randomised to studies at rest or mild exercise in a crossover manner. During each study BT-IRIS® and IC were used simultaneously to measure VCO2 and thus EE. Participants ingested a drink labelled with 50mg 13C-bicarbonate and breath samples were collected every 5 min for 180 min and analysed using IRIS®. Bland-Altman plots were used to assess agreement between the two techniques in measurements of VCO2 (L/day) and estimates of EE (kJ/day). Mean ± SE age and BMI of participants were 21.1 ± 1.1 yrs and 23.6 ± 0.6 kg/m2. Both at rest and exercise, there was small bias but overall poor agreement between the two techniques as evident by the wide 95% limits of agreement in measurements of VCO2 and EE: rest VCO2 (bias 1.4, SD 93, 95% limits of agreement -180 to 183), rest EE (-8.3, 1830, -3595 to 3578), exercise VCO2 (49.3, 66.1, -80.4 to 178.9) and exercise EE (1083, 1944, -2727 to 4893). Furthermore, there was also evidence of systematic error in these measurements. Prior to clinical application, further optimisation of the BT-IRIS® system should be undertaken, given the poor agreement with IC in measuring VCO2 and estimating EE.