Intense exercise is a hazard to racehorses in Ontario and has been linked to hundreds of deaths within the industry, according to a new study from the University of Guelph.

Peter Physick-Sheard, an emeritus professor of population medicine, examined 1,709 deaths in Ontario's horse racing industry between 2003 and 2015.

"Training and racing at top speed within a short amount of time and space is a health risk for horses," said Physick-Sheard of his study that was released this week in Equine Veterinary Journal.

Damage during exercise to the horses' musculoskeletal system, such as fractures, dislocations and tendon ruptures, is the most common underlying problem in the deaths, the study says. The study notes that the immediate cause of death for 97 per cent of those injured horses is euthanasia that occurs shortly after.

Dying suddenly ranks second and is closely affiliated with exercise, with those horses either dying on the track, or very shortly after a workout.

Since 2003, Ontario has maintained a registry of racehorse deaths that occur within 60 days of a race or trial entry, which provides insight into mortality rates, the study notes.

Physick-Sheard studied thoroughbreds, standardbreds and quarter-horses and looked at the differences in mortality both between the breeds and within them.

He and his team examined data from about three million race starts -- each time a horse starts a race -- and official recorded work events along with necropsy reports, the bulk of which were done at the university after a racehorse dies.

Thoroughbreds have the highest mortality, the study says, and notes that exercise represents "a far greater hazard for thoroughbreds than standardbreds" with a risk of dying up to eight times higher for thoroughbreds.

"The best example is the very high mortality for very young horses, especially stallions, on the thoroughbred side and is associated with exercise, which is an area we should focus on," he said.

On the standardbred side, Physick-Sheard said "it's the high mortality that tends to occur off the track. That's really interesting and suggests something procedural and cultural."

The study says the rate of deaths off the track for standardbreds is still about half that of thoroughbreds off track.

He said his study highlighted important structural factors within the industry that need to be examined more closely, especially those within the thoroughbred world.

An oft-stated euphemism within horse racing goes as: standardbreds race to win and thoroughbreds race to breed.

That has real world implications on the horses, Physick-Sheard said.

Standardbred horses, those used in harness racing, are focused on winning over a long period of time, he said, noting that most begin racing at three years old and can race until they are about 14 years old.

"Their type of work and training is much more progressive, much more gradual," he suggested as a possible reason for the lower death rates of standardbred horses.

Those horses will often run upwards of 1,000 kilometres before they enter their first race. Standardbred horses have a high mortality that tends to occur off the track, pointing to another area that should be examined, Physick-Sheard said.

"That's really interesting and suggests something procedural and cultural," he said.

Thoroughbreds are different.

"Thoroughbred racing is more intense and progressive, but shorter in duration," Physick-Sheard said. There is more money to be made in thoroughbreds through breeding, rather than from wining races, he said, and making money is more intense for thoroughbreds younger than five years of age.

Thoroughbreds generally train with relatively short distances to minimize damage that is punctuated by short, intense bursts on the racetrack.

The management of young thoroughbreds, especially those aged two years old, is another area the racing industry should examine further, he said.

Physick-Sheard said the study is the "first of its kind" to compare mortality in three racing breeds.

The death registry is held by the Alcohol and Gaming Commission of Ontario, a Crown agency that regulates the horse racing industry. The AGCO granted the professor access to the data.

The AGCO said it has just begun to examine the study.

"There is an initiative to review the rules of racing and we're looking at what can be improved and what can help," said Mike Wilson, the manager of racing operations with the AGCO.

"The findings (of the study) will be useful in the AGCO's current review."

The AGCO said an official investigates every time a horse dies or is injured or there is an incident.

"Any time a horse goes down, it's one horse too many," Wilson said.