Repeatability of gait analysis measurements in Thoroughbreds in training.
Authors: Sepulveda Caviedes M F, Forbes B S, Pfau T
Journal: Equine veterinary journal
Summary
# Editorial Summary: Repeatability of gait analysis measurements in Thoroughbreds in training Inertial measurement units (IMUs) offer promise for identifying emerging injuries in racing Thoroughbreds through longitudinal gait monitoring, yet the day-to-day and week-to-week variability of these measurements in training horses had not been thoroughly characterised. Sepulveda Caviedes and colleagues fitted 14 racing Thoroughbreds with IMUs at the poll, sacrum and both tuber coxae, then measured head and pelvic symmetry during in-hand trot on concrete over daily and weekly intervals, calculating intraclass correlation coefficients (ICC) and the distribution of absolute differences in vertical displacement parameters. Whilst median daily differences were modest (4–7 mm), the 90th percentile of daily variation reached 9–16 mm and weekly variation extended to 11–19 mm; ICC values averaged 0.73 for daily repeats and 0.65 for weekly repeats, substantially lower than previously reported within-day repeatability. For practitioners planning injury surveillance programmes using gait analysis, these findings indicate that meaningful change detection requires clear baseline establishment and awareness that measurement variation can obscure subtle asymmetries—particularly important when horses vary in training intensity and existing movement quality. The authors appropriately caution that population-specific repeatability studies should precede any long-term monitoring protocol, as ICC values this moderate demand careful interpretation of apparent gait shifts in individual horses.
Read the full abstract on PubMed
Practical Takeaways
- •If using IMU gait analysis to monitor Thoroughbreds in training, expect natural variation of 9-16 mm daily and 11-19 mm weekly in movement parameters before assuming injury-related changes
- •Establish baseline measurements and repeatability thresholds specific to your training group, as ICC values suggest moderate to substantial but not high reliability for detecting subtle asymmetries
- •Consider that gait assessment reliability differs significantly between clinical lameness examinations and routine training monitoring, so don't apply the same interpretation thresholds to both contexts
Key Findings
- •Median absolute daily differences in IMU gait parameters ranged from 4-7 mm, with weekly differences of 4-8 mm in Thoroughbreds in training
- •ICC values averaged 0.73 for daily repeats (range 0.40-0.92) and 0.65 for weekly repeats (range 0.27-0.91), lower than previously reported within-day values
- •90% of daily differences fell between 9-16 mm and 90% of weekly differences between 11-19 mm across head and pelvis movement measurements
- •Repeatability of gait symmetry measurements in training horses is lower than in lameness examination settings, requiring discipline-specific validation studies