Lance's 2009 Blood Values - Much Ado about Nothing

Recently Lance Armstrong published his doping test results in an effort to demonstrate that he's clean. I've said in the past, that doing this is probably a mistake, especially for someone like Lance, because someone is bound to come along and find this or that detail suspicious.

At any rate, we have this data from the Tour de France (extracted from a table available here).

Date Hemoglobin Hematacrit Reticulocyte Comments
7/2/09 14.3 42.8 0.5 pre-tour, in France
7/10/09 14 41.3 0.5 stage 7, 1st mountain day, Contador attacks
7/11/09 13.7 40.7 0.5 stage 8, 2nd day in Pyrenees
7/13/09 no tests rest day
7/14/09 14.4 43.1 0.7 stage 10, sprinters holiday
7/20/09 14 41.7 0.5 rest day
7/25/09 14.5 43 0.7 stage 20, Mount Ventoux

In this case the first one of note to rise to the occasion was Jakob Mørkebjerg, a Norwegian blood researcher. I first heard of the story at nyvelocity.com which has good coverage of the story. The one link they seem to leave out is a more recent article at cyclingnews.com. Dr. Mørkebjerg has made three major points about Lance's blood values. First, that his hematacrit and hemoglobin remain very constant throughout the tour, rather than dropping as expected. Second, that his highest hematacrits both occur after rest days, with the implication being that he could hvae added a bit of blood on the rest days, and third that his reticulocyte counts are low, especially in light of his steady hematacrit and hemoglobin.

Everybody knows your crit drops during a Grand Tour

Lots of things that everybody knows turn out to be wrong. In this case, I make no claims. But that cyclingnews article mentions a study where this was demonstrated. The study involved seven athletes. Maybe it was a tour that was hot at the beginning and cooler later, in which case hydration issues would cause an apparent drop in crit during the tour. I'm not specifically making this claim, I'm just saying that I want to see a lot more data than one study of seven riders before making any conclusion at all.

And before saying that a lack of a drop is a sign of doping, I'd want to know how big of a drop is typical, and how much variation there is. For this kind of statistical data, which is absolutely required to draw this kind of conclusion, you'd need multiple studies each involving scores of athletes. Otherwise there's no statistical basis to decide what is normal and what is not. I admit at this point that I haven't done any research here. I have no idea of the number or quality of studies out there on this subject. But presumably Mørkebjerg is familiar with the studies, and yet he only mentioned a single study with seven riders. It may just be what fit in the article, but it doesn't inspire confidence.

Increases after rest days

Recently I had the misfortune of having some medical issues, and had my blood drawn twice in one week. A complete blood count was performed in both cases. On Monday, my crit was 47, and on Thursday it was 43. The doctor was neither surprised nor concerned. There was no reason to suspect any actually loss of blood or red blood cells. My assumption is that this difference is attributed to hydration, because on Sunday I hadn't been drinking enough, maybe only 20 ounces or so. The rest of the week I spent in my maximum hydration mode that I use for illness, probably 80-100 ounces of fluid every day. There were no other issues like diarrhea that would have had a larger impact on my fluid levels.

Whether you like my hydration assumption or not, the fact that crit can change that much in four days without my doctor thinking it's significant says volumes about what these numbers mean.

In the entire month of the tour, the difference between Lances highest and lowest values was 2.4. Again, if mild dehydration accounted for my crit changes, just simple randomeness of hydration - when did Lance last pee and drink, how much is he sweating - could account for the variations we see. So on the issue of increases in crit after rest days, this data lies completely within the measurement noise. It would certainly be hard to argue for any kind of useful or measurable performance gain from such a small increase in blood values.

But even so one has to wonder what the natural effect of crit is from rest days. Certainly the obvious result during a rest day would be a drop in crit: your body no longer has all that waste to flush, and one presumes the athlete hydrates well, so the body will retain more fluid. This retained fluid translates into a larger blood volume without any increase in red blood cells. In other words, a lower crit. But then after another hard day of riding, does the crit jump back up to pre-rest-day values? Does it take a couple of days to return to those values? Or does it have an extra-quick rebound as the body is fully hydrated and willing to release fluids more quickly? I don't know the answer, but the point is, I'd hardly be surprised to see an influence on normal crit values from a rest day.

Reticulocytes

Reticulocytes are new blood cells, and the measurement is the percentage of his blood that is new. It's no surprise to see a low value going into the tour - he was coming off of high-altitude training, and presumably his body would not be producing many new red blood cells at this point, as it would have already built up a sufficient volume from the altitude. Really though his crit isn't very hight for someone who had been training at altitude.

For the lack of attention this one has gotten, for my money it's still the most interesting of the three arguments. This probably says more about how uninteresting the other arguments are. At any rate, normal reticulocycte counts are around 1%, but with a lot of variation.

So the most interesting bit is that his reticulocytes are consistently low and yet his crit doesn't drop. Still, this isn't much of a sign of doping with crits this low. And, we have five days of tour blood values, meaning we don't know what his reticulocyte count looked like the other 17 days.

Red blood cells are produced in the marrow, and the body can request an increase in production by producing EPO. This tells the body to produce more red blood cells. But it also tells the body to release the cells that are already in production, even though they may not be "ripe". So it's entirely possible, even normal, to see a higher reticulocyte count for a few days, followed by a lower-than-normal count for a few days. It's a good argument for testing some of these blood parameters every single day during a tour, so you could actually make some meaningful statements about how the blood is trending.

Conclusions

If someone wants to pursue this to the last detail, and ignore the signal- to-noise issues, there's still some interesting questions to ask here. Which samples were taken before the stage, and which after? What kind of stage was ridden that day and the day before? What kind of weather conditions? How did Lance perform? In my opinion, even that is a waste of time. There's no way to account for a rider doing a better job of hydrating one day versus the next, simply because of perception of thirst, or mood. There's no way to account for random luck of the body in how the electrolytes are balanced one day versus the next.

There's no way to account for the fact that there's always some complex natural variation that for any practical purposes might as well be considered random.

The more interesting open questions would relate to statistical data on how much blood values are expected to drop during a grand tour, and how much that normally varies. In the absence of that data, the most interesting thing to say about Lance's published blood values is that they look very very normal.

Reader Comments (Experimental. Moderated, expect delays. Posts may be edited or ignored. I reserve the right to remove any or all comments, at any time.)

1 comments:

At 2009/09/22 14:40
wrote:

End Comments

Add a comment

Tom Fine's Home Send Me Email