I posted
this in the Jan Ullrich thread, but it's of course more appropriate here. Let the historical revisionist project begin!
Lance Armstrong’s Era of Performance – Part I: Are his time trial performances much different from other winners?
Abstract
In the aftermath of USADA’s doping charges, Lance Armstrong eventually acknowledged the use of banned substances during his professional cycling career. Reckoning his confessions, we decided to evaluate Armstrong’s sportive accomplishments by comparing his winning time trial achievements with achievements demonstrated by other riders in similar races over the years. In time trial racing, there are no collaborating riders on the course, making opportunities to profit from other riders’ efforts through drafting impossible. Time trial performances thus solely depend on the strength and endurance of the individual rider. Accordingly, we argue that an examination of the ‘historic’ variation in these individual performances will increase chances to detect the influence of illicit doping aids on Armstrong’s performances. In view of his doping use, we expected that his performances would be faster compared to performances of his counterparts in foregoing and succeeding years. We scrutinized archival records of the cycling sport and retrieved information concerning Armstrong’s winning time trial performances (N = 7), realized in the Tour de France (1999–2005), as well as performances of other riders (N = 55) who, from 1934 to 2010, won races in the three European Grand Tours (Tour de France, Giro d’Italia, and Vuelta a España) and all faced time trial distances comparable to Armstrong’s (50–61 km). We examined our research question by analysis of covariance (ANCOVA) with riders as the independent variable (Armstrong vs. the other riders) and mean km/h performances as the dependent variable in which we controlled for the influence of year of competition (i.e., the year in which riders won their time trial) and distances of the trials on riders’ speed. ANCOVA initially revealed that Armstrong (Mkm/h = 49.37) indeed raced faster relative to the other riders (Mkm/h = 44.67, p ≤ 0.05). However, this main effect disappeared (p = 0.80) after controlling for the influence of competition year on riders’ performances, b = 0.20 km/h, p ≤ 0.001. Distance did not have a significant influence, b = -0.03 km/h, p = 0.84. ANCOVA further indicated that all but one of Armstrong’s performances fell within the bandwidth of the 68% confidence interval. Reckoning the historic variation in riders’ performances, Armstrong’s achievements do not appear to be outstanding or atypical, implying that effects of the performance–enhancing doping aids he resorted to are limited. Alternatively, his performances can also be plausibly explained by a gradual progress in speed over time, which is characteristic for professional cycling races such as the Tour de France.
Incredibly, the authors cite a recent paper that concluded there is no evidence that EPO has a performance-enhancing effect. I believe that paper was discussed on this forum a while back, and thoroughly rebutted, though I can't find the thread. They also cite a paper by that giant of anti-doping research Hein Verbruggen, who says that the effects of other PEDs are over-rated. I guess this is Heinie's fallback position from "never, never, never".
A few comments on the paper itself:
1) Nothing about how LA’s winning TT times compared to what he accomplished in his earlier years. You would think if one were going to use data like these to argue that there isn’t much effect of doping—which is exactly what these authors are doing—you would want to make a comparison like this.
2) The riders used for comparison are not described very well. In the Abstract, it says they “won races in the three European Grand Tours”. Does this mean they won the GT itself, or that they simply won an ITT? In the article itself, the authors say they won ITTs, but do not say whether they won the GT. Based on the number (55), and that all GTs were analyzed for a period of about 70 years, one might assume they were all GT winners. There would have been a total of several hundred ITTs in the three GTs during this period. But apparently only long ITTs, 50-60 km, were analyzed. So I'm guessing the 55 winners represented all winners of ITTs of this length, whether or not they won the GT. This is a critical point that should have been clarified, as ITT winners who did not win the GT might have been specialists, and not appropriate for comparison with someone who was winning ITTs as well as climbing stages. LA's performance as continuing a historical trend of ITT improvement becomes more remarkable in that he was also the best climber in his Tours.
3) All three GTs were analyzed. Since the level of competition is generally lower in the Giro and Vuelta than in the TDF, it could be argued that this historical database of ITT times is somewhat biassed towards lesser riders.
4) A central claim of the paper is that LA’s ITT times are not unusually fast when compared with his contemporaries. They are better than riders of earlier eras, but so were ITT times in general of his era. But of course if doping enhanced LA’s performance, it would enhance that of his contemporaries, too.
However, as far as I can tell from the data (Fig. 1), the recent times do not indicate a big increase in speed that could be attributed to PEDs. This is in contrast to analyses of climbing times or power measurements that I have seen.
5) Given the effects of weather, technical aspects of the course, presence of climbing, etc., interpretation is always going to be difficult. If, e.g., the ITTs that LA won were hillier than those in the past, his times would be slower. Even if they weren’t, these factors put a sizeable error margin into times.
Facebook
Twitter
Instagram