Tour, Giro, Vuelta:Rapid Progress in Cycling Performance Starts in the 1980s

After analyzing historic records (1892-2008), El Helou et al. [1] reported a distinctive 6.38% improvement in speed in European professional road racing from 1993 onwards, a period which coincides with the years of the ‘epo epidemic’ in professional cycling. We aim to show that this improvement might be spurious, since El Helou et al. did not account for the influence of confounding variables on riders’ speed progression over time. We scrutinized archival data provided by the French Association Mémoire du Cyclisme [12] and assessed winning riders’ kilometers per hour (kph) and time performances, demonstrated in the Tour, Giro and Vuelta from 1903 to 2011(N = 256). We next classified these measures in ten time periods, accounting for El Helou et al.’s ‘critical’ year 1993. We further assessed the distances and brutality rates of the races, as well as the number of stages in the races and included these variables as covariates in the study, because we expected them to influence riders’ achievements. Analyses of covariance (ANCOVA) showed that the variables included in the model explained R2adj. = .89–.98 of the variation in riders’ performances. The three covariates indeed influenced riders’ performances to a greater or a lesser extent over the years. Time performances appeared to be more valid to appraise riders’ speed progression than kph performances, because the former variable is not biased by the distances of the races. After adjusting for the influence of the covariates, multiple comparisons between time periods indicated that time performances in El Helou et al.’s critical years did not significantly differ from performances displayed by riders in immediate foregoing or succeeding years. Furthermore, the 1970s appeared to be key in riders’ evolution in performance over time. We next calculated the proportional progress (%) in time performances per period as ANCOVA follow–up. Across races, we obtained an improvement of 3.18% in time performance beyond the 1990s that does not deviate from the range of expected variability in performance progress over time. Using the 1970s as a baseline, findings further showed a significant linear and curvilinear progress in time performance within and across the three Grand Tours. Inconsistent with El Helou et al.’s conclusion, however, the rapid linear progress originates in the 1980s, not in the 1990s, and gradually levels off from the 1990s onwards in all three multi–stage races. Findings strongly question opinions about the effects of the ‘epo epidemic’ on cyclist