# Endurance eliminations can be predicted, say researchers

Researchers who analysed the performance of more than 4000 endurance horses have shown it is possible to predict likely eliminations as races progress.

The French research team, writing in the open-access journal, *PLOS ONE*, noted that nearly half of the horses participating in endurance events were eliminated at a vet gate.

“Detecting unfit horses before a health problem occurs and treatment is required is a challenge for veterinarians but is essential for improving equine welfare,” Eric Barrey and his colleagues said.

Barrey, joined in the study by Mohamed Younes, Céline Robert and François Cottin, proposed it would be possible to detect unfit horses earlier in a race by measuring key variables, including heart-rate recovery. Their hypothesis proved correct.

Using their model, it would be possible for riders or trainers to be told at a vet check that there was a 70 percent chance of their horse being eliminated at the next vet gate.

The researchers crunched the data on 4070 different endurance horses that undertook a total of 7032 starts in high-level national and international endurance events between 2007 and 2011.

The events were raced over distances of 80km to 160km. They were rated one-star to three-star events, and all were run under FEI rules in either France, the United Arab Emirates, Spain or Portugal.

The races in question were split into four, five or six successive phases of 30–40km each, depending upon the total distance of each race. Each horse underwent a veterinary inspection after each leg, which it must pass before being allowed to continue, as required under the sport’s rules.

The research team analysed key data related to heart rate, cardiac recovery time and average speed recorded at the previous vet gate, as well as factoring in the age of the horse and the race distance. They found that it was possible to predict the probability of elimination of a horse as it continued with the race.

Overall, 39 percent of the horses that started an event were eliminated, mostly due to lameness (64%) or metabolic disorders (15%).

The percentage of correctly predicted eliminations at each vet gate, from 2 to 5, ranged between 62% and 86%, depending upon the gate.

For a horse to be considered fit enough to continue an event, its heart must be below 65 beats per minute within 20 minutes of arrival, although the exact thresholds vary according to the level of the event.

Overall, they found that horses have a 70% chance of being eliminated at the next gate if their cardiac recovery time is longer than 11 minutes at vet gate 1 or 2, or longer than 13 minutes at vet gates 3 or 4.

“Heart-rate recovery and average-speed variables measured at the previous vet gate(s) enabled us to predict elimination at the following vet gate.”

The researchers suggested that these variables should be checked at each veterinary examination, in order to detect unfit horses as early as possible.

“Our predictive method may help to improve equine welfare and ethical considerations in endurance events,” the researchers said, noting that elimination rates appeared to have increased in recent years.

This increase was a source of concern for the sport’s ethics and image, they suggested.

Lameness, dehydration and metabolic disorders were the main causes of elimination in 160km events, they noted. Elimination rates in endurance events varied from one geographical area to another.

For example, eliminations for metabolic disorders occurred more frequently in hot and humid countries. However, lameness remained the most common cause of elimination in all countries.

“Given that a significant proportion (12%) of eliminated horses require emergency medical treatment, the rapid detection of tired horses is a challenge for veterinarians and a key issue in the improvement of equine welfare,” they said.

Heart rate, they said, was a key parameter for success in endurance events because the cardiac recovery time was taken into account when calculating the average speed of each phase of an event.

In order to check the validity of their predictive calculation, they computed the probability of elimination for an independent dataset of 80 horses having competed in 1, 2 and 3-star endurance events organized in 2014.

The 80 horses were randomly selected so that the dataset contained 10 qualified and 10 eliminated horses for each phase. In all, 75% of the 80 horses were classified correctly with the system.

Lastly, they determined the threshold values of average speed and cardiac recovery time that corresponded to a 70% probability of elimination at a vet gate due to a heart rate of 65 or more.

They presented the results using a colour scale in order to show the probability of elimination when the cardiac recovery rate and average speed were above the computed thresholds at the previous vet gate.

“Briefly, a horse that needs more than 11 minutes to recover – that is, to achieve a heart rate below 65 beats per minute – at vet gates 1 or 2 has a 70% probability of being eliminated at vet gates 2 or 3, respectively.

“At vet gates 3 or 4, a horse that needs more than 13 minute to recover has a 70% probability of being eliminated at vet gates 4 or 5, respectively.

“Although the probability of elimination rises with increasing average speed for all phases, the phenomenon is more pronounced in phases 2 and 3.”

The researchers, from a range of tertiary institutions, said their results confirmed the relationship between cardiac recovery rate, heart rate and average speed in predicting elimination.

Changes in these paramaters during the event are clearly associated with an increased probability of elimination at the next vet gate. “The higher the speed, the higher the probability of elimination. The average in phases 2 and 3 appears to be strongly associated with the probability of elimination at vet gates 3 and 4.

“Most metabolic changes occur during the first half of an event; when combined with high speed, the effect of these changes increases the probability of elimination later in the event.

“Elevated values of cardiac recovery time and heart rate were also associated with a greater probability of elimination.”

The data processed in the study represented a quarter of the FEI-registered endurance events organized between 2007 and 2011.

“Although several studies have highlighted the risks associated with excessively high speed in endurance events, the present study may be the first to have demonstrated the relationship between average speed, cardiac recovery time and the probability of elimination.”

Discussing their findings, the researchers said that, given the statistical nature of the prediction and the presence of false positive predictions, their model should not be used directly as an FEI rule for eliminating a horse. It could, however, serve as a tool for improving the efficiency of the vet gates.

“For example, the probability of elimination could be given to the rider and the trainer as a warning. They could therefore decide whether to retire, slow down (in order to decrease the risk of elimination at the next vet gate) or maintain the same racing strategy (despite the high risk of elimination at the next vet gate).”

They concluded: “We suggest that the cardiac recovery time should be checked at each vet gate, in order to identify unfit horses earlier and thus improve their welfare.”

In practice, cardiac recovery time threshold values of 11 minutes (at vet gates 1 and 2) and 13 minutes (at vet gates 3 and 4) are associated with a 70% probability of elimination at the next vet gate.

“We encourage the veterinarians involved in endurance events and the FEI to apply this new knowledge, in order to improve the welfare of endurance horses and prevent health problems during or after endurance events.”

Younes M, Robert C, Cottin F, Barrey E (2015) Speed and Cardiac Recovery Variables Predict the Probability of Elimination in Equine Endurance Events. PLoS ONE 10(8): e0137013. doi:10.1371/journal.pone.0137013

The full study can be read here.

• Receive a notification when a new article is posted: