Power-Based Running Durability in Trained Runners: Preliminary Evidence of Speed-Adjusted Mechanical Drift as a Marker of Fatigue.

This study examined wearable-derived running power and biomechanical variables as indicators of durability during prolonged dual-intensity exercise in trained trail runners. Twelve trained trail runners completed a 216-min treadmill protocol of three repeated 72-min blocks (60 min steady-state running at 85% of lactate threshold speed plus a 12-min uphill time trial [+12% incline]), plus an unfatigued control trial. A foot-mounted running power metre (Stryd Next Gen, Stryd Inc., CO, USA; firmware version 2.1.32.1.1) continuously recorded power output, duty factor, ground contact time, form power, cadence, vertical oscillation and leg stiffness. Repeated-measures ANCOVA with running speed as a covariate compared biomechanical across time trials at a common, model-adjusted running speed. During the uphill time trials, duty factor, ground contact time, leg stiffness, vertical oscillation and form power changed significantly after controlling for speed (all p < 0.05; η2p = 0.23-0.41), reflecting neuromechanical fatigue and a shift towards a lower-stiffness, higher-contact-duration gait. In contrast, absolute power, normalised power and %CP declined in proportion to running speed and did not reach significance after covariate adjustment. During steady-state running, biomechanical and physiological variables remained stable except duty factor, which increased modestly but significantly from 60 to 180 min (71.3 ± 1.6% to 72.7 ± 1.7%; p = 0.014, η2p = 0.32), the sole wearable-detected indicator of mechanical drift at submaximal intensity. Running cost remained preserved throughout both exercise conditions. These findings suggest that the duty factor is the most sensitive wearable-derived marker of fatigue-related mechanical reorganisation during prolonged trail running, and that a wearable device can capture both the neuromechanical and performance dimensions of durability.