Ryan Shay: Performance analysis suggests that he did not use doping products

Even as we continue to mourn the tragic and sudden death of Ryan Shay during the USA Men's Olympic Marathon Trials, there have been occasional murmurs that Shay's death was somehow related to the use of performance-enhancing drugs. Shay never had a positive test for any banned substance, although these days that means little as more and more doped athletes admit to past wrongs without ever having been caught. A toxicology report would do much to silence any critics, but all of us still await any news from the NYC medical examiner's office. Their report should be conclusive, but even in its absence, we're confident that based on an analysis of his performances, Ryan Shay was unlikely to be using performance-enhancing drugs.

No prior history

With each new Marion Jones, David Millar, and Erik Zabel that comes forward and admits to past doping, the fact that an athlete is the "most tested athlete in the history of sport” means less and less. Still, this small point falls in Ryan Shay's favor.

The absence of a "breakthrough" performance or season

One aspect of the model we use to understand doping in sport is that those athletes who start doping only later in their careers will produce "breakthrough" performances or seasons that indicate the introduction of drugs to their program. This phenomena has been seen before most notably with Barry Bonds, who at the age of 36 became the most productive offensive player in history. In track and field, Franke and Berendonk published the data of an East German female shot putter. This athlete trained drug-free for 14 years before using steroids, and over the next ten weeks her performance improved by 9%. (We reviewed this study here).

So it is an athlete's "breakthrough" that first should raise the alarm that something might be different. Typically, those athletes who are the best in their sport tend to be near the top at each successive level as they move through the junior and amateur and professional ranks. Adopting new (and legitimate) training techniques and indeed make a difference, but these translate generally into small improvements and not career building breakthroughs.

If we examine Ryan Shay's performances over his running career, we see that his 10,000 m time between the years of 2000-2002 (while he was running at Notre Dame) did not change by more than 1-2%, especially after the 2000 season. His "peak" performances for the Big East and NCAA Championships were always with 20-30 s of each other year in and year out, as show in the graph below:


Even after he graduated from Notre Dame in 2002, his 10,000 m performances did not change by very much in 2004, again shown on the graph. Had he used performance-enhancing drugs after leaving the NCAA competitive arena, we would have expected to see a significant improvement in his 10,000 m times during 2003-04, yet instead we see he ran similar times as all his prior performances, and in 2003 he finished 3rd in 14:06 at the USA 5km champs.

Later in 2002 Shay made his marathon debut at the La Salle Bank Chicago Marathon and posted a respectable 2:14:30. In a big city marathon like Chicago that put him way off the pace (he finished 15th), but showed he was on par with his American colleagues and perhaps with more experience he might improve that time by a few minutes. In fact Shat won the USA Men's Marathon champs in 2003, yet his record in the marathon is remarkably similar to his record in the 10,000 m in that he never really made a breakthrough performance:


Although it might appear that Shay's marathon times were coming down over his last three performances, in reality his slower times must be a function something else, for he ran his best time of 2:14:08 at the 2004 NYC Marathon, which was only 22 s faster than his debut at Chicago in 2002. So again his marathon performances do not exhibit any kind of "spike" or breakthrough that would suggest the use of performance-enhancing drugs. Had he suddenly starting using something, we could expect his marathon time to drop by 2-5%, which equates to a 2.5-6 min improvement. Instead, over the six marathons he ran Shay produced remarkably similar times.

What the performance data cannot tell us

These data unfortunately represent circumstantial evidence and cannot with 100% accuracy determine if Shay (or any athlete) did or did not use performance-enhancing drugs. However, what we can say here is that the data do not suggest he was taking anything, or perhaps even if he were then it certainly was not improving his performance and does not follow the other available evidence from Barry Bonds and our East German shot putter.

So for now we must continue to await the toxicology report from the NYC medical examiner's office. It will say what it will, although at this point the evidence from his performances suggests that he was not taking anything.