The effect of interruptions during training on the time to the first trial and race start in Thoroughbred racehorses.

Few studies have investigated the effect of having interruptions during training on future training and racing performance in Thoroughbred racehorses. The aim of this paper was to investigate the effect of having an interruption before the first trial on starting in a trial or a race. A prospective cohort study was used to record the training activity of a cohort of Thoroughbred racehorses, over two racing seasons. Fourteen racehorse trainers recorded information on the distances worked at canter and at fast speeds (<15s/200 m) and provided reasons for horses not training, or for having interruptions (break from training). Trial and racing results were obtained from the New Zealand Thoroughbred Racing online database. A Cox proportional hazards regression model was used to investigate two outcome measures of performance: (1) time to the first trial and (2) time to the first race. The type of interruption that had occurred before the first trial was the main exposure of interest, and was grouped into: no interruption, voluntary (no known condition or disease present) and involuntary interruptions (due to the presence of a condition or disease). A total of 160/200 (80%) horses started in at least one trial and 100/205 (48%) horses started in at least one race during the study period. The median time to starting in a trial or a race differed significantly (p<0.001) with the type of interruption. The hazard of starting in a trial was lower for horses experiencing voluntary and involuntary interruptions (p<0.001) but there was no association with starting in a race, after adjusting for confounding variables. As age at the start of training increased the hazard of starting in a trial decreased. Horses accumulating longer distances at 15s/200 m had a higher hazard of starting in a trial, whilst horses accumulating fewer events at high speed and fewer trials had a reduced hazard of starting in a race. There was significant clustering at the trainer level for both the outcomes investigated. Interruptions to training had an effect on the time to, and hazard of, a trial but not a race start. The timing of these interruptions may have implications for future racing success and career longevity.