In the Journals - Sports doping

Not, I think, peer-reviewed science this time, but two articles from Nature and Science respectively, both on the subject of sports doping, a subjetc which is of course very topical given the Beijing Olympics, due to start tomorrow, 8th August.  The first is a Commentary article from Donald Barry on the statistical significance of doping tests, focussing on the case of Floy Landis, who was stripped of his 2006 Tour de France victory following a positive test for testosterone.  The second article is one of a series of News articles in Science related to the upcoming Olympics, and questions whether performance enhancing drugs really do benefit performance.

The figure on the left is included in the article - it shows delta notation of isotope ratios of 167 samples tested at the LNDD (who tested the 2006 TdF samples, including Landis'), those considered positive are in red, those negative are in green.   I don't think these samples are necessarily derived from the Tour, if any, since a good number are considered positive.

------------------------------------------------------------------------------------------------ 

The Science of Doping

Donald A. Berr

Nature 454, 692 - 693 (2008) | doi:10.1038/454692a.

Full Text (requires subscription)

Berry presents a Commentary piece which sets out the stall for his opinion that sports dope-testing lacks rigour with repect to the statistical treatment of test result data.  In a sense the essential problem is also identified in the Science article discussed below: how do we know definitively which test results in a sample set correspond to natural and which to doped sampls?  Obviously doped athletes aren't going to confess!  Add to this the possibility that the characteristics that  make a sportsperson champion material are in some way due to abnormal hormomonal state (in comparison to "Joe Public"), and you can potentially find a confusing situation. I'm not really competent to comment of the statistics - but will observe that a lot of the discussion seems to me to hinge upon specificites which are just pulled out of the hat to make an argument.

The figure on the left is included in the article - it shows delta notation of isotope ratios of 167 samples tested at the LNDD (who tested the 2006 TdF samples, including Landis'), those considered positive are in red, those negative are in green.   I don't think these samples are necessarily derived from the Tour, if any, since a good number are considered positive.

What I find a little confusing is that what's being plottedare values that reflect isotopic ratios - I don't think it's stated, but presume it is the ratio of 12C to 13, and which is indicative of exogenous steroid, generally because it's derived from plant sources.

Looking at these scatter plots, on has to say they look like a mess.  But given that these data samples will come from different athletes, competing in different sports and disciplines, and taking (or not) a variety of supplements or doping products, it's not surprising that the data look a little ragged.  My understanding from the article is that these data are of the type that are the second round of analysis, once to testosterone:epitestosterone ration has been examined.

Berry is of course right to say that organisers and testers should be more open about what the cut-off criteria are.  He's also right that rigourous double-blind testing would help here.  The difficulty is of course that finding real "negatives" and real "positives" would be well-nigh impossible. 

------------------------------------------------------------------------------------------------

Science at the Olympics: Does doping work?

Martin Enserink

Science 1 August 2008:
Vol. 321. no. 5889, p. 627
DOI: 10.1126/science.321.5889.627a

This news item makes the claim that the bulk of proscribed drugs make little difference to athletic performance, other than a select few:

anabolic steroids - increase muscle mass, benefiting sports requiring muscular strength.

EPO - enhances haematocrit (the proportion of red blood cells in the blood), thereby increasing aerobic endurance.

amphetamines - useful for enhancing short term explosive efforts such as sprinting.

That there have been effects of doping practises can be inferred (if not proven) by looking at historical records - such as the graph accompanying thisarticle, showing annual best hammer throw records 1970-2007:

On the other hand, one might have expected a greater decline through the 80s had the post-1970 rise been solely attibutable to steroid use.  Anyway, the figure makes a general point.

The article makes the observation that double-blind trials are rare in sport - for pretty obvious reasons.  It also quotes Harm Kuipers (Maastricht University) as saying that many of the substances listed are there only because it was known that athletes were using (or reputedly using) them.  "Un-banning" them might make sense to some, but presumably would encourage their use.  One has to bear in mind that many of these pharmaceuticals can have unexpected or dangerous effects, particularly those that are hormones.  Finally some substances are banned because they are thought to be used as masking agents or to help remove substances from the system.

Of course we now have the spectre of genetic manipulation ahead of us...