Tomorrow, the eyes of the horse racing world will turn to the 133rd annual Preakness Stakes, the middle jewel of U.S. horse racing's Triple Crown.

But the dust has barely settled from the tragedy at the Kentucky Derby two weeks ago, where vets were forced to euthanize the promising gray thoroughbred filly, Eight Belles, when she collapsed on the track after completing the race at Churchill Downs, suffering from two shattered ankles in her front legs.

The death sent a chill through the world of thoroughbred racing, which just over a year ago lost beloved Barbaro from complications of surgery to repair a massively fractured leg suffered at the Preakness eight months before. Critics immediately condemned the racehorse industry's emphasis on breeding for speed and quick maturation, alleging that it compromised the overall health of the horses.

Thoroughbred racing "is in a moral crisis, and everyone now knows it," sportswriter Sally Jenkins wrote in The Washington Post.

Veterinary experts are not dismissing the notion that modern horses might pay a genetic toll for their precocity. "Those of us who have seen how the breed has changed do believe that's possible," says Rick Arthur, equine medical director for the California Horse Racing Board. But hard evidence remains scarce.

In Eight Belles's case, an autopsy released yesterday by the Kentucky Horse Racing Authority found no preexisting bone disease or injuries that might have contributed to the injury, making it hard to pin the accident on any one cause. The horse's owner has requested additional tests to prove that she had not been given steroids.

Even as they look for hints of declining vigor among racehorses, vets and racing officials have taken steps they hope will reduce the chances of fatal accidents. Preliminary evidence from 34 tracks, collected since last June as part of the on-track injury reporting system, suggests that switching from dirt tracks to a synthetic blend of wax-coated silica sand, fibers, and rubber particles may improve horse safety.

Mary Scollay, an associate veterinarian at Gulfstream Park Racing & Casino and at Calder Race Course, both in south Florida, who coordinates the on-track project, reported at the March 17 Welfare and Safety of the Racehorse Summit that dirt tracks such as Churchill Downs had seen 2.03 fatal injuries per 1,000 starts compared with 1.47 per 1,000 on synthetic tracks. Scollay cautions that the number reflects less than a year's worth of data so it should not be taken as definitive.

The belief that horses face hidden health risks comes from changes in breeding incentives since the 1960s and 1970s, Arthur says, when rich owners bred and raced their own contenders. Horse breeding is now big business, with top thoroughbreds easily commanding six-figures per mating. To recoup their investments, buyers want horses that will mature rapidly and win often.

Anecdotally, trainers believe that selection for speed and youthful vigor has made horses unable to race as frequently as in the past, says Ed Bowen, president of the Grayson-Jockey Club Research Foundation in Lexington, Ky., the cradle of U.S. horse breeding.

In October 2006 the Jockey Club convened the first Welfare and Safety of the Racehorse Summit. The group set up its Durability Index, which shows that the number of starts made by the average U.S. racehorse declined from 11.3 per year in the 1960s to 6.4 in 2006. The goal, Bowen says, is to allow breeders to identify thoroughbreds possessing above-average durability, or ability to race frequently over multiple years.

There is strong evidence that horses inherit racing prowess, if imperfectly, says Ernest Bailey, professor of genetics at the Gluck Equine Research Center at the University of Kentucky in Lexington and coordinator of the Horse Genome Project. But he says a decline in durability would likely require changes in multiple genes.

"I don't understand how such a change could have come about within 50 years on a genetic basis for a population that may be one million worldwide," he says. More plausible, he adds, is that horses might inherit particular types of fragility. Researchers are mining the horse genome, assembled last year, looking for genes that might contribute to disorders such as osteoarthritis or respiratory ailments.

"There are horses that are notorious for having certain types of problems," Arthur says. He cites a "prominent stallion" that had congenitally constricted airways but who "could have had three ears and they would have bred him … his horses were so great."

Bailey, however, notes that four generations of selection for a gene for HYPP (hyperkalemic periodic paralysis), which bestowed quarter horses with large, well-defined buttocks but led to injuries, spread the gene to only two percent of horses—descendents of Impressive (1968 to 1995), the stallion that carried the gene.

Up in the air, as well, is whether the Durability Index would reflect limits imposed by genetic defects or by changes in the way animals are regarded and cared for. "In the old days it was not uncommon to run-run-run-run the horses," Arthur says. "Maybe they were having the same problems but they ran them anyway."

Modern horses receive exceptional care and treatment, Scollay says, but by the same token, their owners demand wins. "One needs to be very careful when you increase the physical demands on a person or an animal … to keep them performing at that level," she says. Horses in top condition, such as Eight Belles, are at particular risk, she says, for microscopic tears in their bone and cartilage that may build up over months of hard racing without causing the horse any discomfort.

"We've got very good evidence that bone disease and microdamage cause all fractures," says Wayne McIlwraith, an equine orthopedic surgeon and director of the Equine Orthopaedic Research Center at Colorado State University in Fort Collins. Eight Belles suffered fractures in the sesamoid and metacarpal condyle bones of the fetlock, or ankle joint.

"If you look at a horse finishing a race, they're usually very hyperextended in that joint," which is the site of almost all fatal fractures, McIlwraith says. Although the autopsy uncovered no preexisting problems, Scollay says that a microscopic analysis of the bone tissue was apparently not done.

McIlwraith says he and his colleagues are currently working on a blood test to screen for microdamage. They enrolled 200 thoroughbred racehorses in training and sampled their blood every month to check for seven biological markers—portions of collagen and sugar-flecked protein molecules—exposed when bone and collagen start to degrade.

When a horse suffered a fracture, they compared its blood markers with those of two foals (horses around one-year old) matched in age and sex. McIlwraith says they found they could have predicted 75 percent of the fractures up to two months in advance based on the pattern of markers.

He says his group is working with a veterinary company in hopes of commercializing a screening test within two years. "If we could diagnose that microdamage, we could potentially prevent catastrophic injuries," he says. "Wayne's test would be a tremendous asset," Scollay agrees.

Bailey advocates a step-by-step gathering of the evidence to address possible risks to racehorses. "Frankly, the most productive approach for this problem will be to identify all the discrete questions" and "eliminate the causes of fragility that have a hereditary basis … one by one. It's the only approach that will work, anyway."