Methods for the non-invasive assessment of bone quality were used to monitor the osteopenia induced by immobilising a forelimb in a cast. These techniques included the measurement of ultrasound velocity, single photon absorptiometry and radiographic photodensitometry. Serial measurements were made before, during and after an eight week period of immobilisation of the left forelimb of four adult standardbreds in a light fibreglass cast which included the foot and extended above the carpus. The measurements continued for 12 weeks after the removal of the cast. The results indicated a trend to decreasing bone strength and mineral content in both forelimbs, with the decreases being greater in the limb with the cast. The values tended to recover slowly after the removal of the cast, then more quickly after the horses returned to the paddock four weeks later. The differences between the limb with the cast and the opposite limb were significant at only a few points of comparison.

Summary The current methods for non‐invasive measurement of bone quality are reviewed. In the horse this has traditionally involved the use of radiography, but there are now two other modalities available for the critical evaluation of cortical bone quality and strength. These utilise single photon absorptiometry and ultrasound velocity. Photon absorptiometry gives a direct measurement of bone mineral content, by using a monoenergetic radionuclide source, and transverse ultrasound velocity in bone gives a measure of bone stiffness or elasticity. They can both be used conveniently on the metacarpus of the conscious horse. Both ultrasound velocity and bone mineral content can be used as accurate indicators of skeletal maturity. In addition, the effects of disuse on bone and certain types of lameness can be monitored accurately. Preliminary data show an association with exercise in young and mature horses. There also appears to be considerable scope for in vivo research of bone changes in horses produced by immobilisation, weightlessness, exercise and nutrition.

The in vivo precision of ultrasound velocity measurement and single photon absorptiometry for the assessment of equine bone quality is discussed. In vivo precisions for ultrasound velocity measurements were less than 0.5 per cent, whereas cortical cross-sectional area, compact bone density and modulus of elasticity were around 1 per cent, and bone mineral content and density were just over 2 per cent. Except for ultrasound velocity, substantial improvements could be achieved by taking the mean of five readings for each measurement. The long-term precision of the techniques was also high, with all precision values being less than 3 per cent. The possible sources of variation in ultrasound velocity measurements were also investigated. The method for combining ultrasonic and photon absorptiometric data for equine bone quality assessment has been proven to be highly accurate and precise. There appears to be no reason why the same principles cannot be applied with equal success to the noninvasive assessment of bone quality in humans.

Summary Transmission ultrasound speed was determined in the third metacarpal bone of 347 Thoroughbred racehorses in training. In addition to direct measurement of apparent ultrasound speed and lateral bone diameter, several derived parameters were estimated (cortical ultrasound speed, corticomedullary ratio and cortical cross‐sectional area). Multiple regression equations were developed for each ultrasound measurement using the explanatory variables: sex, age, duration and intensity of training, preparation number and previous shin soreness. The horses (114 females, 82 entire males and 151 geldings ranging in age from 20 to 116 months) had been in continuous training for an average of 10.0 weeks when measurements were made. All measurements except corticomedullary ratio were influenced by the sex of the horse. Geldings had significantly lower ultrasound transmission speed than entire males and females, reflecting a lower bone density. However, the geldings had larger cross‐sectional areas (ie bone size). We suggest that the relationship between low ultrasound speed (ie low bone density) and large cross‐sectional area maintains the overall mechanical integrity of the metacarpus. This relationship (high ultrasound speed and low cross‐sectional area) was also found in each sex group. Ultrasound speed and corticomedullary ratio, but not cross‐sectional area, were affected by age. Both duration and intensity of training influenced the ultrasound speed measurements. This study indicated that sex, age and state of training affect ultrasound results, but this explained only about 30% of the variance in the data. Other factors, ie genetic variation in bone density and size, differences in soft tissue thickness and distal limb temperature, presumably account for most of the variance. Acknowledgements. This project was sponsored by the Australian Equine Research Foundation and partly through a special grant from the Department of Sport and Recreation of Victoria. We thank Dr C. Savage and Dr H. Davies for assistance and advice; Mrs M. Jones and Mr A. Alder for technical help with the ultrasound measurements; and the many trainers who allowed access to the horses in their care.