Triathlons could be risky for athletes with heart disease, potentially leading to cardiac arrest or sudden death, according to a new Minneapolis Heart Institute Foundation study.

The competitive, three-segment sport has grown increasingly popular in the U.S. since its inception in 1970, with nearly half a million participants taking part in 3,200 USA Triathlon-sanctioned events in 2015, the study, published this week in the Annals of Internal Medicine, stated. Since past preliminary research has questioned the safety of such events, lead investigator Kevin Harris, MD, and a team analyzed data of 135 athletes who suffered sudden death during competition in an effort to dissect the risks of the sport.

In their study, Harris—a two-time triathlete himself—and colleagues assembled a study group from USA Triathlon (USAT) records, the U.S. National Registry of Sudden Death in Athletes and personal reports from between 1985 and 2016. The team identified 107 cases of race-related sudden death, 13 instances of resuscitated cardiac arrests and 15 trauma deaths among the study pool. A large majority—85 percent—of subjects were men, and the average age of the athletes was 47.

“I have long been interested in sports and, in particular, endurance sports,” Harris told Cardiovascular Business. “I did a couple triathlons over the years, and one was in my early career as a cardiologist in Minneapolis. Although I’m a strong swimmer, I really found the swim portion to be challenging with all the chaos of swimming fast surrounded by others in very close proximity. It was difficult to get adequate breaths. Later when attending a sports medicine conference, I became interested in the fact that people were dying during the marathon.”

The researchers found the bulk of mortalities took place during the swimming segment of the race, making up 67 percent of all unexpected cardiac events. Whereas 90 victims suffered in the water, just 22 had difficulties during the bike portion and 15 had difficulties during the running portion of the triathlon. Eight victims experienced trauma post-race.

Nearly 40 percent of these participants were participating in their first triathlon, Harris and co-authors wrote, and of those new-timers, most were running a sprint race, rather than an intermediate- or long-distance course. Course distance didn’t prove to be a make-or-break variable, though; death risk appeared to be similar regardless of race length. Younger athletes also proved to be a vulnerable group—they were much more likely than seasoned participants to die in trauma-related incidents, which made up 15 of the 22 deaths during the bike segment.

While the land portions of a triathlon give racegoers ample space to watch competitors, trouble was harder to detect when athletes were active in oceans, lakes, rivers and other natural bodies of water, which made up 87 percent of all swim venues. Just 3 percent of those venues were pools. A third of all victims who experienced trouble during the swimming segment were discovered unconscious in the water; nine signaled for help, nine were found by other competitors or rescue personnel, and four collapsed immediately after exiting the water.

“As I personally experienced, the swim portion is really unique and difficult to prepare in advance for,” Harris said. “The athlete enters water with many other swimmers. There is potential for body collision, inability to get in breathing rhythm and inherent difficulty for the swimmer to rest or signal for help. Yes, the environmental conditions play a role. Although there are certain guidelines, there is a lot of unpredictability with nature. I have a strong appreciation about how the ocean can treat a swimmer. Swimming laps in the pool is likely not enough to prepare for the stress of the triathlon.”

Presence of established cardiovascular disease or abnormalities could be a factor in predicting the health and abilities of a triathlon participant, the authors wrote. Of 61 autopsy reports they were able to obtain, 44 percent of subjects had cardiovascular abnormalities; 18 individuals had significant atherosclerotic coronary artery disease. Other patients showed evidence of hypertrophic cardiomyopathy, mitral valve prolapse, ascending aortic dissection, arrhythmic right ventricular cardiomyopathy and congenital coronary anomaly. Participants also died from heat stroke, rhabdomyolysis and trauma injuries. Harris and colleagues attributed the 13 recoveries from cardiac arrest to “timely” cardiopulmonary resuscitation and defibrillation.

The researchers wrote the sex differences in their data could be attributed simply to the fact that existing coronary disease was more prominent in male triathletes than in female ones.

“In most studies of death in athletes, male deaths predominate,” Harris said. “This may relate to earlier development of coronary disease in men. It could also relate to other protective factors or inherent differences in the way men and women perceive symptoms in the throes of competition.” He said he was also surprised by the marked increase in mortality in males that came with each decade after they turned 40.

Despite gender and age, Harris doesn’t recommend anyone with a background of heart disease compete in a triathlon—especially males.

“Anyone with risk factors should consult their provider before competing,” he said. “Certainly anyone with cardiovascular symptoms needs to discuss those with their provider prior to competition.”