Evaluation of Horse Fitness for Exercise: The Use of a Logit-Log Function to Model Horse Postexercise Heart Rate Recovery

The aim of this study was to test the adjustment of the logistic-log model with postexercise heart rate (HR) recovery data in horses and introduce the logit-log model as a parsimonious model, with the half recovery time as a measure of horse fitness. Postexercise HR (bpm) measurements at different timings were obtained from 32 polo ponies in two different exercise regimes: n = 16 “full chukker” and n = 16 in “half chukker.” Measurements were taken at rest, immediately after exercise and then 2, 4, 6, 10, and 20 minutes after exercise. The HR variable was transformed into a logit (HR) variable, and time was transformed into log (time). Means of logit (HR) at the different log (times) were obtained, creating two time series of transformed variables that were then adjusted by simple linear regression. The degree of adjustment of the model is high with values of r² = 0.989 for the full chukker and r² = 0.998 for the half chukker. Full chukker ponies have a half recovery time of 5.7 minutes and half chukker ponies of 11.2 minutes, with the former showing to be fitter than the latter, which is why they are chosen to play for longer periods. The logit-log model is parsimonious, and the half recovery time can be easily determined. The half recovery time has potential to be used as a measure of fitness degree, allowing comparison between horses.