Claire Chaisson

This study investigated within-session variability in treadmill running mechanics during repeated same-day runs across multiple speeds, with a secondary aim to assess how these mechanics changed with speed. Eleven team sport athletes (9 M, 2 F; 25.0 ± 3.2 years) completed three trials of 30-s treadmill runs at 10, 12, 14, 16, 18 and 20 km/h. The coefficient of variation (CV%) for step length (2.7 to 1.7%, β = -0.10, p = 0.004) and flight time (7.9 to 4.3%, β = -0.41, p < 0.001) decreased with speed, indicating greater consistency. Step length (β = 0.068), step rate (β = 0.054), flight time (β = 3.69) and horizontal work (β = 0.026) increased with speed, and contact time (β = –9.86), duty factor (β = –0.008) and vertical work (β = –0.041) decreased. These findings suggest movement patterns become more consistent at higher speeds.

INTRODUCTION: Duty factor (DF) is a simple marker of movement strategy, defined as the ratio of ground contact time to stride time during running (1). Athletes who adopt aerial-based running patterns (low DF) receive increased biomechanical loads on the foot, ankle, and calf muscles whereas terrestrial-based runners (high DF) place increased load on knees and hips. Given football players typically run 9-11 km in matches, including 111-255 m of sprinting (2), there is a need to better understand loading patterns regarding implications for optimal training regimens. The aim of this study was twofold: (1) to create a method that practitioners could implement easily into training routines and (2) to categorise running styles remotely during the maximum velocity phase of a controlled sprint. METHODS: 53 football players across the Women’s Pro Game Academy League performed controlled sprints as part of their normal training routine across the 2024-25 season (161 observations total; 3.0 ± 1.5 per player). As minimal disruption to training was prioritised, a method was needed to automatically detect the maximum velocity phase while considering different routines in place at each club (e.g., 30 m or 40 m sprints, repeated sprints). An algorithm (RStudio, Posit Software, Boston, MA) was created to automatically detect the start and end times of the maximum velocity phase of the controlled sprints from 10-Hz acceleration, velocity, and distance data from a foot-mounted inertial measurement unit (Playermaker, London, UK), already in use in training by clubs. These times were manually entered into the Playermaker dashboard, which allowed the mean contact time and stride time to be calculated for the entered bout. Contact time and stride time were adjusted using a prior validation study comparing the F-IMU system with an instrumented treadmill and used to calculate DF. DF values were used to identify two separate groups (high and low DF) and to compare between playing positions. RESULTS: Results showed low DF (0.290 ± 0.016) and high DF (0.330 ± 0.015) groups achieved similar maximal sprint speeds (6.59 ± 0.41 m·s-1 and 6.61 ± 0.48 m·s-1, respectively), indicating that DF reflects movement strategy and leg stiffness rather than sprinting ability. Playing position itself did not meaningfully influence DF, with substantial variation in sprinting styles existing within positions (e.g., central [0.258-0.351] vs wide [0.268-0.330] defenders). CONCLUSION: This study provided a novel proof of concept for assessing running styles across several Women’s Pro Game Academies during the maximum velocity phase of a controlled sprint. DF can help contextualise sprint exposure beyond velocity metrics when individualising training. In team sports where there is limited time, this offers one method of monitoring running styles with minimal burden to practitioners. 1. Hanley et al. (2022). Frontiers in Sports and Active Living, 4:939676 2. Savolainen et al. (2023). Biology of Sport, 40(4):1187–1195

The aim was to assess concurrent validity and test-retest reliability of spatiotemporal gait parameters from a thoracic placed inertial measurement unit (IMU) in lab- (Phase One) and field-based (Phase Two) conditions. Spatiotemporal gait parameters were compared (target speeds 3, 5 and 7.5 m·s-1) between a 100-Hz IMU and an optical measurement system (OptoJump Next, 1000 Hz) in 14 trained individuals (Phase One). Additionally, 29 English Premier League football players performed weekly 3x60-m runs (5 m·s-1; observations =1227; Phase Two). Mixed effects modelling assessed the effect of speed on agreement between systems (Phase One), and test-retest reliability (Phase Two). IMU step time showed strong agreement (<0.3%) regardless of individual or running speed. Direction of mean biases up to 40 ms for contact and flight time depended on the running speed and individual. Step time, length and frequency were most reliable (Coefficient of variation = 1.3-1.4%) but confounded by running speed. Step time, length and frequency derived from a thoracic-placed IMU can be used confidently. Contact time could be used if bias is corrected for each individual. To optimise test-retest reliability, a minimum running distance of 40 m is needed to ensure 10 constant-speed steps are gathered.