Author(s): WALKER, A., MCKAY, A.K.A., BURKE, L.M., Institution: AUSTRALIAN CATHOLIC UNIVERSITY, Country: AUSTRALIA, Abstract-ID: 2098

INTRODUCTION:
Many athletes can be at risk of low energy availability (LEA) caused by high exercise energy expenditure (EEE) and/or inadequate energy intake (EI). Accurate calculation of EEE is important for determining energy availability (EA), however, various methods can be used. Currently, it is unclear how different calculation methods affect both EEE and EA, particularly in real world, free-living settings. This study compared the values of EEE (kcal) derived from different calculation and prediction methods and the effects on estimates of EA (kcal/kg/FFM).
METHODS:
Elite rowers (n=46; Tiers 4-5) attended a 3-week altitude training camp in which day-to-day energy availability was measured. This study focused on six different methods for EEE calculation across all rowing and bike ergometer sessions: (1) measured gross efficiency (GE) for each individual and modality, (2) estimated GE for each modality, (3) heart rate derived calculation, (4) heart rate derived calculation that accounted for VO2max, (5) metabolic equivalent score (METs) estimation and (6) linear regression prediction from GE. Predicted RMR was estimated using three different equations, with basal energy expenditure during exercise either included or excluded to determine its effect. Finally, EA was calculated as EI minus EEE, relative to fat free mass (FFM). Linear mixed models were used to determine the effects of EEE and RMR calculation methods on both EEE and EA. Method 1 was used as a benchmark to which the other methods were compared to.
RESULTS:
There were significant differences between the values derived from each of the six methods of EEE calculation (p<0.001). Compared to individualised gross efficiency, METs calculations overestimated EEE by 3.2% whereas the heart rate (HR) derived calculations showed 33.5% and 26.8% underestimations of EEE respectively. The removal of RMR significantly affects EEE values (p<0.001). The transfer of EEE values into EA calculations also produced significantly different results (p<0.001). The lowest EA values were derived from METs values (7.4 kcal/kg/FFM lower than benchmark) while the highest EA values came from HR derived EEE calculations (7 kcal/kg/FFM higher than benchmark).
CONCLUSION:
These results have practical implications for EA calculations in both research and applied settings. The use of METs may overestimate EEE while HR derived calculations (at least in athletic populations) may underestimate EEE compared to power output derived estimates using individualised GE outcomes. Differences in the method to estimate EEE from the same exercise sessions translated into significant differences in calculations of EA, as did the removal of RMR from EEE. Although the results are statistically significant, experts need to decide whether such changes may have a clinical or physiological effect on athletes in a real-world environment.