Sunday's Nautica New York City Triathlon resulted in two deaths, both from cardiac events that arose during the event's initial swimming leg. A 64-year-old man and a 40-year-old woman were pulled from the Hudson River before they could complete the 1.5-kilometer swim from a wharf near Manhattan's 96th Street down to the West 79th Street Boat Basin. Both athletes were taken to a nearby hospital in cardiac arrest. Police said 26 others were removed from the water needing assistance for minor injuries or pains that arose during the swim portion of the competition.

The man, identified as Michael Kudryk of Freehold, N.J., was competing as part of a three-person relay team but lost consciousness halfway through the swim. Rescuers managed to get Kudryk to one of the four New York City Fire Department boats stationed in the river, but he later died at Saint Luke's-Roosevelt Hospital Center. Kudryk was registered for another triathlon next month in Long Branch, N.J., the New York Daily News reported. Amy Martich,  of Elmhurst, Ill., died Monday morning at Saint Luke's.

Triathlons are categorized by distance. Sprint triathlons typically require a 750-meter swim, 20-kilometer bike ride and five-kilometer run. Olympic triathlons (including the NYC Triathlon) involve a 1.5-kilometer swim, 40-kilometer bike ride and 10-kilometer run. In "Iron Man" triathlons, the swimming portion can be as long as 3.9 kilometers, followed by up to a 180-kilometer bike ride and a 42.2-kilometer run (the same distance as a marathon).

Despite being the first leg and covering the shortest distance in any triathlon, swimming has proved to be the most deadly. Minneapolis Heart Institute cardiologist Kevin Harris last year published a study in JAMA: The Journal of American Medical Association analyzing the results of 2,971 USA Triathlon-sanctioned events held between January 2006 and September 2008, during which 14 participants died—13 of them while swimming and one while biking. Swimmers who died were between 28 and 65 years old; 11 were men (although it is worth noting that more men compete in triathlons than women).

Harris's research found that six of the swimming deaths occurred during sprint triathlons, four in Olympic triathlons and the final three in so-called "long course" triathlons, which include Ironman events. Eight swimmers were in distress and called for assistance while five were found motionless on the water. Six of the deaths occurred during swims in the open ocean, four in lakes, two in reservoirs and one in a river. Still, Harris's research indicates that deaths are rare during triathlons—nearly 960,000 athletes participated in the events studied.

Scientific American spoke with Harris about his research, the dangers of the swimming leg and what might be done to make triathlons less dangerous.

[An edited transcript of the interview follows.]

What prompted you to study the causes of sudden death in triathlons?
We have a large registry here at the Minneapolis Heart Institute Foundation that looks at sudden death in athletes. We were aware of the death rate in marathons. However, given that the triathlon has become so popular (and we had noted some triathlon deaths), we started to look at it formally. The USA Triathlon–sanctioned races allowed us to pull out the denominator of athletes competing and thus calculate the incidence of death [at about 1.5 per 100,000 participants].

Nearly all of the deaths you studied occurred during the swimming events. Did this surprise you, given that swimming is the first leg of the triathlon and, presumably, the athletes are not suffering from the heat or from exhaustion at this point in the competition?
Yes, exactly. We were first surprised by this but I note this trend continues beyond the end of our study (also in several non-sanctioned races we did not formally look at). While at first I was surprised, it does make sense for a number of reasons: first, the adrenaline surge and pure number of athletes entering the water at the same time; second the fact that I suspect many athletes come from a background in running or other sports and may be less adept at swimming; third, swimming in a triathlon is totally different sport than doing some laps in the pool due to variability of extremes of waves [as well as] people swimming around you and on top of you; fourth, the inability to rest properly if needed (or call for help) as you could do in the marathon and bike [segments]; and, fifth, the difficulties in being noticed if the swimmer is in trouble due to the number of athletes in a body of water, which is not transparent. I think these are some of the factors that are related.

It would appear that most of the deaths that you studied were the result of cardiovascular abnormalities. How is it possible to train for a triathlon without such problems arising well in advance of the competition?
We don't know the death rate while training for a triathlon. I know deaths have occurred but I agree it's probably less than the triathlon itself. I suspect the adrenaline rush with competition and the above factors contribute.

Why are the swimming events so dangerous from the perspective of cardiac arrest?
I think because of the delay in identification and rescue and the time to get to dry land to use resuscitation equipment. Time matters with resuscitation.

What can be done to improve the safety of triathlon competitors?
First, knowledge of risk of the race for the competitors will be helpful and hopefully allow people to take training seriously, especially for the swim. Second, if there are any concerns regarding suitability of the competitor, such as symptoms with training of chest pain, shortness of breath, syncope (passing out), family history of cardiac disease or—in athletes of older age—other cardiac risk factors. Each of the people in these groups should be fully evaluated before participation.

Whether races can mandate pre-op exams (or more testing) similar to what is done for high school and college sports and whether that would make a difference is unknown. Third, I know many race directors are interested in making races safer through the number of people on the water, staggered starts, et cetera. These efforts should help. Another question is whether some minimal proficiency for open-water swimming should in some way be evaluated prior to enrollment in the race.