Biometric Sensors Present Promise in Detecting Early Injury Risk in Racehorses

The American Association of Equine Practitioners (AAEP) has released findings from a landmark prospective study evaluating wearable biometric sensors for early detection of musculoskeletal injuries in Thoroughbred racehorses. The research, which followed 561 two-year-old horses over the 2025 race season, demonstrates that sensor-generated alerts can identify horses at significantly higher risk of injury, offering trainers and veterinarians a valuable tool for proactive health management.

Study Design and Methodology

The AAEP Wearable Biometric Sensor Research Project is the first prospective study of its kind to assess sensor effectiveness in real-world training conditions. Six technology companies—Alogo, Arioneo, Equimetrics, Garmin, Stable Analytics, and StrideSafe—participated by deploying proprietary sensors and algorithms on participating horses. Sensors were worn during high-speed exercise sessions, known as breezes, and provided real-time feedback categorized into three status levels: Green (low risk), Yellow (moderate risk), or Red (high risk).

To ensure data reliability, the study required participating companies to submit reports within 48 hours of data collection. Only four companies—Alogo, Arioneo, Stable Analytics, and StrideSafe—met this threshold and contributed data to the final analysis. Over the course of the year, researchers recorded 4,252 breezes and documented 342 musculoskeletal injury reports across 221 distinct injuries in 181 horses.

Key Findings on Injury Risk Detection

Horses that received a Yellow or Red sensor reading after a breeze faced more than twice the odds of sustaining a musculoskeletal injury compared to those flagged Green. Specifically, both Yellow and Red alerts were associated with an odds ratio of 2.2 for injury occurrence. This finding held true across sensor platforms, indicating that despite differences in proprietary algorithms, the core biometric signals were consistently linked to elevated risk.

Further analysis revealed that risk increased with cumulative alerts. Horses receiving multiple Yellow or Red designations over time showed a progressively higher likelihood of injury. Horses whose most recent workout was an unofficial breeze, or those that accumulated Red flags over a 90-day period, were identified as having elevated risk profiles.

The study similarly identified prior injury during the observation period as the strongest predictor of future injury, with an odds ratio of 27.9. Injuries typically occurred at a median of the sixth breeze, approximately 10 weeks into the training cycle.

Implications for Equine Health Management

According to Dr. Peta Lee Hitchens of the University of Melbourne, the study’s lead data analyst, the sensors demonstrated clear potential as a complementary tool in injury prevention. “The answer is yes from the data,” said Dr. Larry Bramlage of Rood & Riddle Equine Hospital, affirming that the technology could overcome design variations among devices to detect meaningful physiological signals.

By providing objective, real-time data on locomotor patterns and physiological responses, biometric sensors enable trainers to make informed decisions about adjusting training loads, scheduling veterinary evaluations, or implementing rest periods. This approach supports a shift from reactive treatment to proactive prevention, potentially reducing welfare risks and economic losses associated with training downtime.

Limitations and Future Research

While the results are encouraging, the study acknowledges limitations. Sensor data alone cannot diagnose specific injuries, and false positives may lead to unnecessary interventions. The researchers emphasize that sensor alerts should be used alongside veterinary examinations and trainer expertise, not as replacements.

Future work will focus on refining algorithms to improve specificity, expanding monitoring to older horses and other disciplines, and integrating sensor data with other health metrics such as heart rate variability and gait symmetry. Long-term studies are needed to assess whether sensor-guided interventions reduce actual injury incidence over multiple seasons.

Conclusion

The AAEP’s research confirms that wearable biometric sensors hold meaningful promise for early detection of injury risk in racehorses. By translating complex physiological data into actionable alerts—Green, Yellow, or Red—the technology offers a practical framework for enhancing equine welfare and performance optimization. As sensor technology advances and validation efforts continue, these tools may become a standard component of responsible racehorse training and care.

Frequently Asked Questions

What types of injuries were monitored in the study?

The study tracked musculoskeletal injuries, which include damage to bones, joints, tendons, ligaments, and muscles commonly associated with high-intensity training and racing.

How do the sensors work?

Wearable biometric sensors collect data on movement patterns, stride characteristics, and physiological responses during exercise. Proprietary algorithms analyze this data to assess injury risk and assign a color-coded status (Green, Yellow, Red).

Can sensors prevent injuries?

Sensors do not prevent injuries directly but can flag increased risk, allowing trainers to modify training regimens or seek veterinary evaluation before a minor issue becomes serious.

Are all sensor brands equally effective?

The study found that despite differences in technology, four of the six tested systems produced consistent risk signals when they met data quality thresholds, suggesting core biometric indicators are robust across platforms.

Where can I read the full study?

Details of the AAEP Wearable Biometric Sensor Research Project are available through the American Association of Equine Practitioners’ official publications and news releases.