Ed Coyle publishes a study on Armstrong
AS: Actually, that's the perfect segue for us to start talking about Ed Coyle a little bit. Ed Coyle wrote a study in 2005 about the work he did with Armstrong from '92 to '99, where he claimed that he was able to improve his watts per kilo 18%. Armstrong himself holds this study up as proof that he didn't need dope to win the Tour. There were two ways Coyle went about it. One, Armstrong's efficiency improved, and two, his weight was reduced. How did you become aware of this paper?
MA: I guess I need to put it in a little bit of context. Muscle efficiency is sort of like a holy grail in physiology. To put it in context, a 1% improvement in efficiency has been calculated with various modeling techniques to give you about a one minute improvement in a 40k time trial. So an 8% improvement in efficiency is simply unheard of. It has never been measured before, and so naturally, when Ed Coyle published a paper reporting that, there was an enormous amount of interest. Not just in the lay media, but in scientific circles as well, because lots and lots of people have tried to see if they could change cycling efficiency with different training protocols, it has never been found to change.
And so when you report that not only has it changed, it changed by 8%, then obviously that seems a very unusual finding. As scientists, the first thing you want to do is say, "I want to read the paper, I want to satisfy myself with the methodology that he used, because, gee, this seems like a strange result."
AS: Let's just define the term real quickly, cycling efficiency. The idea is that if you're riding at 400 watts, your body is actually producing much more energy than that, and the efficiency is the percentage of cycling power vs. total power?
MA: Yeah, in very simple terms, it's how much of the energy production in your muscles actually go into the pedals and gives you propulsive force.
AS: And overall energy is measured by how much oxygen you're burning?
MA: Yeah, it's a laboratory test where the subject is put on an ergometer, you measure very carefully how much oxygen the body is using, and that will tell you how much energy is being burned, if you like, in very crude terms. And then you look at the ergometer, and you measure very carefully how much energy the athlete is producing, and the ratio of those two after some adjustments give you the index of efficiency.
So, how much oxygen is he using, versus how much energy is he putting into the bike gives you that index of efficiency.
AS: So Coyle was claiming that for a given oxygen consumption, Armstrong was producing more watts because he was making himself more efficient through training.
MA: That's right. The claim was that because Armstrong had been training for three to six hours on his bike over a period of years, that probably altered his muscle composition, which led him to having a greater efficiency. Now, the only, the glaring oversight there, is that Lance Armstrong is NOT the only cyclist that trains for three to six hours on his bike each day, it's pretty much routine for a professional cyclist.
So, many other professionals have been measured after they've done this same sort of training, but none of them have shown a change in efficiency, which immediately brings into question the basis of Coyle's claim.
And he speculates that this was due to his unique ability to convert his fast twitch muscles to slow twitch muscles.
Again, there was no data to substantiate it. It was his speculation, attempting to explain what he had measured. Again, it just comes down to a simple case of, well, ok, if that was the underlying biological reason, then you'd expect to see it happen over and over again. There's nothing, despite what other people want to believe, there's nothing unique about Lance Armstrong. He's a human being, and he responds as other human beings do to training. And no one else has ever measured those changes that Coyle speculated might've happened in Armstrong.
AS: As a matter of fact, you could take a biopsy of his muscles and analyze it, but he never did.
MA: To put it in context, this wasn't a carefully planned study. This was an opportunistic approach where, the students in the lab related to us, it was simply a matter of Lance Armstrong swinging by occasionally, and Ed Coyle would test him almost as a favor to him, to give him some data. If he came into the lab, great. If he didn't, then there was no data.
And I think that's borne out. If you look at the timing of when these test sessions took place, there really isn't any coherent pattern. For example, if you really wanted to show that your cycling efficiency had increased leading up to a Tour de France, then you want to measure him immediately before, or immediately after the Tour. You don't wait four or five months, like Coyle did in '99, when he's stopped training and was almost beginning the next season.
So, it wasn't carefully planned, the timing of the test sessions were opportunistic rather than carefully thought through.
AS: Let's go in chronological order. You became of aware of this study, and then did you and some of your colleagues lodge a complaint with the University of Texas?
MA: The way it happened is, obviously when the article came out, it spiked a lot of interest and discussion in the scientific community. And the way that scientists address those issues is to write letters to the journal that published the article. Essentially in the letter, you're raising some questions, and then the author is given an opportunity to respond. So both your concerns and his response are published side by side to inform of the rest of the scientific community so they can judge for themselves - well I agree with that, or I don't agree with that.
So that's the first thing that happened, there were two letters published in the journal that questioned very specific aspects of the study. One of those aspects was which ergometer did Coyle use to measure Armstrong's power during those seven years. Now, the reason that's so critical is you HAVE to use the same ergometer very carefully calibrated to make sure that when you measure, say, 300 watts in 1992, that seven years later in 1999, if the ergometer reads 300 watts, you want to be sure that that's correct, and not just an artifact of the ergometer that you're using is different, or uncalibrated, or whatever.
So, the first question related to that, because we had experience with these longitudinal studies, they're exceptionally difficult to carry out successfully. The first question that we had was, "Did you use the same ergometer to measure power output?"
And categorically Coyle responded, "Yes I did. The same ergometer was used for all tests." And, we had to take that at face value. When you question a scientist, they publish their response, and you are obliged to accept whatever they say. So we pretty much had to accept that. We still had reservations, but that was as much as we could know.
Then interestingly, the paper itself became involved in an arbitration hearing where I was asked to serve as an expert witness and interpret this paper for the hearing. In that process I did some background checking to verify to myself what was happening and could this data be relied on. And I was very surprised when we were given a photograph showing Lance Armstrong in the first test session on an ergometer that was definitely not the ergometer that Coyle claims he tested him on.
It was a very disturbing revelation and it was purely a fluke occurrence, where the journalist had been in the laboratory, was taking photos for this journal article, and happened to take a photo while Armstrong was being tested. So we had that reservation and several others that we still felt uneasy about, and we elected to take those directly to Ed Coyle confidentially. We spelled out our concerns, and we said, "Professor Coyle, with the greatest respect, we really don't believe this paper is worthy of publication, would you please retract it?"
And his response was characteristically vehement, and adamant, that he would not retract the paper. And so, from that point we had no option but to seek some other way of getting what we felt was a scientific error corrected. And the next step up was to go to the journal themselves, and say, "Look, these are our concerns, what do you propose we do about it?"
And it came after a lengthy round of discussions, that the journal weren't going to do anything until we made a formal application to the university of scientific misconduct, and that's a very serious step, when you actually go to the scientist's institution and formally claim that he has conducted himself in an inappropriate way. And you're formally asking the university to take action. So those were the steps that began with just an initial "Gee, this paper seems unusual" and gradually became more and more disturbing as more evidence came to light, and eventually resulted in lodging this allegation of misconduct.
.gif "IP Gear")
