I'm going to have to leave shortly. The report actually goes on a little bit more:in case you haven't seen the full pdf posted above, here's the explanation given by the expert panel:
"After careful review and considering the most relevant scientific literature, the Expert Panel concluded univocally that Sample 2 HGB and OFFS values were still much too high to be explained by a stay and training at altitude in the two weeks preceding the sample collection, even considering the most favourable scenario to the Rider:
“According to the new information provided in Mr. Scott’s report, the athlete was at altitude until the 3rd of August 2018 e.g., the altitude exposure ended 14 days before Sample 2 was collected. To examine the potential impact of this altitude exposure on the blood values in Sample 2 we have first calculated the approximate altitude dosage according to Garvican-Lewis et al. 2016. Three pieces of altitude information each day from 07.07.2018 till 03.08.2.18 were presented in the report by Mr Scott; ‘accomodation altitude’, ‘average exercise altitude’ and ‘peak exercise altitude’. For the majority of days the altitude reported does not differ significantly between the three categories. Since we have no information of the average exercise time per day we have hypothesized that the athlete exercised for an average of 8 hours each day (which likely is overestimated, but will benefit the athlete in our calculation) except on the 16.07.2018 and 21.07.2018 where he rested (was anticipated to be at ‘accomodation altitude’) and that the remaining 16 hours were spent at the ‘accomodation altitude’. The total altitude dosage during the 28 days was calculated to 992 km.hr. There are several publications having examined the effect of altitude. Mr. Scott has mentioned some of these in his report, also some of which have reported results during the post-altitude period, which is the period of interest in this profile. Instead of picking only specific papers that potentially fits into the argumentation, a more scientifically correct way would be to use data from a meta-analysis, where all relevant data is examined together. Lobigs et al. 2018 provided such a paper in 2018. Here it is evident that the average increase in Hb 15 days after altitude is around 0.2 g/dL above the normal level, while the OFFscore on average is 3-4 points above baseline. Considering only values reflecting a similar altitude dosage (800-1000 km.hr) as the athlete, none of the subjects from the meta-analysis had elevations in Hbs above 1 g/dL or elevations in OFFscores of more than 20 points, 15 days post altitude. Considering all Hb values except Sample 2 and 6, the athlete’s average level is 15.0 g/dL and the average OFFscore is 99 (only considering samples analyzed with the XT1 excluding Sample 2 and 6). Hence the Hb in Sample 2 is 1.6 g/dL above average and the OFFscore is 24 points above average, which highly contrast the average increase of 0.2 g/dL and OFFscore 3-4 points and above the highest single values recorded in the Lobigs study."
Further to this, Mr. Scott argues that the elevated Hb could be due to a plasma volume reduction 5-7 days after a high exercise load (Miller et al. 2019). According to Mr. Scott, the Rider raced (and won) on 12 August 2018, i.e. 5 days prior to the collection of Sample 2. In the Miller paper which examined hematological changes after an Ironman competition (which might not be directly comparable to a cycling race) the Hb was actually not statistically different from the baseline 5 days after the Ironman race. In relative terms the Hb was approximately 0.5%, 4%, and 3% higher than baseline on days 5, 6, and 7 post-race, respectively. Hence even a 4% increase in Hb corresponds to 0.6 g/dL, which in combination with the altitude effect still falls short of the value observed in Sample 2. Furthermore, in the Miller paper an increase in %ret of 20-30% is reported in the 5-7 day period after the Ironman competition, which contrast the low %ret value observed in the Rider’s profile.
The Miller effect (0.6) and the altitude effect from Lobigs et al. (0.2) combine for the 0.8 value I referenced earlier. And again, remember that the acceptance limit was exceeded by 0.3.
I can't figure out how to copy and paste the relevant figure from Lobigs et al., but it is Figure 2, and you would be looking at the values marked in asterisks. Note that the Lobigs et al. changes are from immediately pre-altitude, and are not from the change in the mean as the expert report states. Sample 1 was taken more than a year before sample 1, so the expert panel has no idea what the rider's blood values were immediately before altitude exposure. To account for this it would be necessary to increase the variability to reflect the uncertainty in pre-altitude blood values, but of course the expert panel didn't do that.
The change in OFF score from Lobigs et al. referenced by the expert panel is the result of a linear fit, and is not the mean change in OFF-score at that time. The 2(!) studies that included a total of 13 (!) participants that were included in the meta-analysis actually had an increase of 11.2 in the OFF-score- and note the large variability.
The expert panel reported that Stannard's OFF-score was 24 points above average, but it was only 6.65 points above the acceptance limit. 8 of the 13 participants had a change in OFF-score greater than that.
It is only by ignoring the variability built into the biological passport that the expert panel found a breach of the limits. Note that variability is completely absent from the expert panel's explanation.
I have spent too much time on this, so this is probably the last post unless something obviously wrong is posted.
Last edited: