Author(s): HORNER, K., O'NEILL, J.E.R., Institution: UNIVERSITY COLLEGE DUBLIN, Country: IRELAND, Abstract-ID: 2559

INTRODUCTION:
Resting metabolic rate (RMR) measurement is crucial for athletes to assess energy requirements and energy availability [1]. The ratio of measured to predicted RMR (RMR ratio), is being increasingly used as a proxy indicator of energy availability, with a ratio <0.9 suggesting low energy availability [2, 3]. A recent meta-analysis shows variability in RMR prediction equation accuracy, suggesting potential issues with using standard equations for RMR suppression detection [4]. A more suitable use may be in longitudinal monitoring, and interpretation of directly measured RMR and body composition [4]. For accurate longitudinal monitoring, it is crucial that the indirect calorimeter exhibits exceptional reliability. This study evaluates the reliability of the Q-NRG (Cosmed, Italy) indirect calorimeter (IC) against the established Quark CPET (Cosmed, Italy) for RMR measurement.
METHODS:
Thirty athletes from multiple sports (19M, 11F) underwent two RMR assessments a week apart using Q-NRG and CPET ICs, with the order randomised initially and then reversed. Outcomes were compared between test visits using paired t-tests and inter-day CVs were calculated. Additionally, RMR measurements were evaluated against recommended prediction equations for use in athletes [4] (Cunningham (1980) [5], Harris-Benedict (1918) [6], Ten-Haaf FFM (2014), Ten-Haaf BW (2014) [7].
RESULTS:
Between testing sessions 1 and 2, there were no significant differences in Q-NRG RMR (P=0.80), VO2 (P=0.79), VCO2 (P=0.86) and RQ (P=0.47), or CPET RMR (P=0.96), VO2 (P=0.82), VCO2 (P=0.58) and RQ (P=0.45). Both ICs were highly reproducible between test days with a slightly lower CV of 2.4 ± 1.6% (range: 0.2-5.9%) for the Q-NRG, and 4.0 ± 3.6% (range: 0.1-13.7%) for the CPET. A significant mean bias of -148 ± 124 kcal/24h was observed for RMR with the Q-NRG compared to CPET (1684±239 kcal/24h, range: 1169-2385 vs. 1831±294, range: 1271-2511 kcal/24h, P<0.0001). Significant mean biases were also observed between Q-NRG and CPET VO2 (P<0.0001), VCO2 (P<0.0001), and RQ (P=0.03). When comparing measured RMR to predicted values, the CPET measured a greater percentage of participants to be within10% of predicted values versus the Q-NRG (CPET: 50% (30/60); M, 39.5% (15/38); F, 68% (15/22) vs. Q-NRG: 33% (20/60); M, 16% (6/38); F, 64% (14/22)).
CONCLUSION:
The within-individual variability of both ICs was excellent, with the Q-NRG producing superior reproducibility, indicating that both are suitable for use in longitudinal athlete monitoring. The Q-NRG had a bias when compared to the CPET, highlighting that it is essential to use the same indirect calorimeter for repeated monitoring. Ongoing work is establishing the accuracy of the devices independently via ethanol burn testing and metabolic simulator.

[1] D. T. Thomas et al. 2016
[2] S. Staal et al. 2018
[3] T. Sterringer et al. 2022
[4] J. E. R. O’Neill et al. 2023
[5] J. J. Cunningham et al. 1980
[6] J. A. Harris et al. 1918
[7] T. ten Haaf et al. 2014