Study: Biometric sensors show promise in flagging injury risk

The American Association of Equine Practitioners on Tuesday released results from the first prospective research project to test whether wearable biometric sensors can detect early signs of musculoskeletal injury in racehorses.

Thoroughbred 2-year-olds wearing sensors that produced yellow- or red-flag readings after a breeze faced more than twice the odds of sustaining a musculoskeletal injury compared with horses flagged green, according to lead data analyst Dr. Peta Lee Hitchens of the University of Melbourne. Both red and yellow flags carried an odds ratio of 2.2.

The yearlong study followed 561 2-year-olds through the 2025 race season, recording 4,252 breezes. Six sensor companies – Alogo, Arioneo, Equimetrics, Garmin, Stable Analytics and StrideSafe – took part, each deploying proprietary sensors and algorithms. Four ultimately met the study's strict 48-hour reporting requirement and had data included in the final analysis: Alogo, Arioneo, Stable Analytics and StrideSafe.

Dr. Larry Bramlage of Rood & Riddle Equine Hospital said the central question was whether the sensors could overcome their design differences and identify an increased probability of musculoskeletal injury. "The answer is yes from the data," he said.

Hitchens's multivariable analysis identified prior injury during the study period as the largest risk factor, with an odds ratio of 27.9. Horses accumulating red flags over 90 days and those whose last work was an unofficial breeze also faced elevated risk. Injuries occurred at a median of the sixth breeze, roughly 10 weeks into training.

The study documented 342 musculoskeletal injury reports, tracing to 221 distinct injuries across 181 horses. Bone issues accounted for 142 cases and soft tissue for 79. Of the distinct injuries, 77% linked to the last unofficial breeze, 17% to the last official breeze and 7% to a race.

Dr. Sarah Reuss, equine technical manager at Boehringer Ingelheim, tied the project to the association's broader safety mission. Eleven Thoroughbred and veterinary industry organizations contributed more than $900,000 in funding: Breeders' Cup Limited, Fasig-Tipton, the Foundation for the Horse (the charitable arm of the AAEP), Keeneland Association, Kentucky Thoroughbred Association, the New York Racing Association, New York Thoroughbred Horsemen's Association, Oak Tree Racing Association, Ocala Breeders' Sales Company, Thoroughbred Owners and Breeders Association and the Grayson-Jockey Club Research Foundation.

A notable secondary finding was that 50.5% of study horses raced in 2025, compared with the Jockey Club Fact Book's baseline of 39.8% for all 2-year-olds. Hitchens cautioned that selection bias and other factors might have contributed to the gap.

Bramlage framed the sensors as diagnostic aids rather than scratching tools.

"This is not meant to be something that scratches horses," he said. "This is meant to be something that says, maybe you ought to have this horse looked at."

Hitchens said validation studies must come next before the sensors can deliver the sensitivity and specificity data needed for regulatory use. Sensor placement varied across companies – girth, saddlecloth and tail – and each firm set its own thresholds for flag categories.

Bramlage acknowledged the sensors cannot yet reliably predict musculoskeletal injury on their own.

"But at some point, we feel like they're likely to be, based upon what we saw," he said.