Ex vivo calibration and validation of in vivo equine bone strain measures

Summary Reasons for performing study : Data are required to confirm that strain gauges recording high bone strains in Thoroughbred racehorses provide an accurate record of bone strain. Objective : To test the accuracy and reliability of very high in vivo strain recordings made during fast exercise in Thoroughbred racehorses. Methods : Strains were recorded during exercise from rosette gauges implanted onto the mid‐shaft dorsal cortex of each third metacarpal bone (MC3) in 6 yearling and 6 mature Thoroughbreds in a previous experiment. Bulk elastic modulus (E US ) was calculated from ultrasound speed and single photon absorptiometry measures. Each cleaned MC3 with the original gauge in situ and new gauges placed in a region of strain similarity (as shown by photoelastic coating) was loaded in a materials testing system (MTS) and strains recorded during loading. Elastic moduli were calculated from strain measures from new rosette gauges on the medial, dorsal, lateral and palmar surfaces (E m ; E d ; E l ; E p ), and bulk moduli calculated from the displacement of the MTS machine heads during loading (E bt ). Peak loads were increased incrementally to failure. Results : Of 14 original gauges tested against new gauges, 11 recorded strains from 80–115% of the new gauges and 3 showed reduced function (31–40%). E bt were similar to E m , and E US were similar to E l and not significantly different from E d . Maximum strains at yield were recorded by the medial gauges and ranged from −7500 to −16,000 μɛ. Conclusions and clinical relevance : Similarities between recordings from gauges used in vivo and new gauges confirmed the reliability and likely accuracy (or possible underestimate) of very high strains (exceeding −6000 μɛ) recorded in exercising Thoroughbred racehorses. The similarity between E bt and E m confirms that the gauges measured the true distortion of the bone in the MTS. These results confirm that mammalian bone may withstand much greater compressive loads than −4000 μɛ under some conditions at least.