Abstract
Running economy (RE) is highly reliable when measured in an unfatigued state; however, its reproducibility during prolonged exercise has not been investigated. RE is known to worsen during prolonged exercise (referred to as RE "durability"), and quantifying the reliability of RE durability measurements will determine our sensitivity to detect subtle differences between groups or conditions. This study aimed to assess the reliability of RE and other physiological parameters throughout a 90-min run in the heavy-intensity domain among well-trained runners. Fourteen male marathon runners (maximal oxygen uptake (V̇O(2max)): 63.1 ± 5.8 mL·kg(-1)·min(-1); marathon time: 02:47 ± 00:10 h:min) completed two 90 min treadmill runs at a speed of 10% Δ between lactate threshold one (LT) and two (LT2) (14.1 ± 0.9 km·h(-1)). Measurements were taken at 15 min intervals for RE (expressed as energy cost [EC] and oxygen cost [OC]), ventilation (V̇E), heart rate (HR), blood lactate (BLa), and rating of perceived exertion (RPE). Reliability was quantified using intraclass correlation coefficients (ICCs), coefficient of variation (CV), typical error (TE), and systematic differences. Both EC and OC demonstrated excellent reliability (ICCs 0.96-0.99; TEs 0.7%-1.4%; CVs 0.6%-1.2%) consistent across all time points (p > 0.15). V̇E and HR were highly reliable, whereas RPE and BLa exhibited greater variability, such that BLa became less reliable with run duration, especially after 60 min of exercise (time effect on BLa CV: p = 0.01). These results indicate that relatively modest changes in RE durability due to interventions or between conditions are likely to be detected, and provide critical information for future experiments aiming to measure and enhance the durability of endurance athletes.