Poor Association between Facial Expression and Mild Lameness in Thoroughbred Trot-Up Examinations.
Authors: Anderson Katrina A, Morrice-West Ashleigh V, Wong Adelene S M, Walmsley Elizabeth A, Fisher Andrew D, Whitton R Chris, Hitchens Peta L
Journal: Animals : an open access journal from MDPI
Summary
# Editorial Summary: Facial Expression as an Indicator of Mild Lameness in Thoroughbreds Detecting pre-clinical lameness in racehorses remains challenging despite its potential to prevent catastrophic injuries, prompting researchers to investigate whether facial expression scales—specifically the Horse Grimace Scale (HGS) and Facial Expressions in Ridden Horses (FEReq)—might serve as reliable indicators of subtle orthopaedic pain during pre-race trot-ups. Anderson and colleagues analysed 380 images from 21 mildly lame and 17 non-lame Thoroughbreds using objective lameness measurement via Equinosis Lameness Locator®, with two independent observers scoring facial features and interobserver reliability assessed using statistical models. Whilst some associations emerged—notably that exposed sclera correlated with lameness (p = 0.045) and orbital tightening showed links to head amplitude measures—the findings were inconsistent and sometimes counterintuitive, with moderate mouth strain and upper lip tension paradoxically associated with *lower* lameness likelihood. The moderate interobserver reliability (κ 0.45) and weak predictive power of facial expression parameters suggest these scales cannot reliably identify mild lameness in fit-to-race horses, indicating that practitioners should continue relying on established kinetic and kinematic assessments rather than facial expression analysis for pre-race lameness screening.
Read the full abstract on PubMed
Practical Takeaways
- •Facial expression assessment alone is insufficient for detecting mild lameness in pre-race trot-ups and should not replace objective lameness detection methods like the Equinosis Lameness Locator
- •If using facial expressions clinically, exposed sclera may be the most reliable indicator of orthopaedic pain, but this finding alone is too weak to guide decision-making
- •Continue relying on objective gait analysis and physical examination parameters rather than subjective facial expression scoring for identifying pre-clinical musculoskeletal injury in racehorses
Key Findings
- •Interobserver reliability for facial expression scoring was only moderate (κ 0.45; 95% CI 0.36, 0.55)
- •Exposed sclera was the only facial expression parameter consistently associated with lameness (p = 0.045)
- •Moderate mouth strain and tense upper lip were paradoxically associated with lower likelihood of lameness (p = 0.042 and p = 0.027)
- •Inconsistent associations between facial expression scales (HGS and FEReq) and mild lameness limit their predictive utility for pre-race examinations