This, from the NYTimes:
Landis said last week that he was expecting the worst because backup samples, or B samples, almost always confirm the initial result. But [Landis's doctor, Brent] Kay said the B sample could come back negative.
“The carbon isotope was only mildly elevated,” he said. “We know, from a statistical standpoint, that the first result could have been a false positive.”
posted by Patrick Herron at 10:47 PM
I'm a former (very) amateur cyclist with debilitating arthritis in my left knee. I live my dreams of cycling glory vicariously through people like Lance and Floyd.
My bias in Floyd's favor is offset by the familiarity I developed with performance-enhancing drugs while in high school; I *know* the abuse of performance-enhancing drugs is far more prevalent than is being reported. Later in life I had a good friend who was advised by his college head coach that if he wanted to make the NFL he should take steroids. He did take steroids, and he did make the NFL. That coach is now a division I coacn at a big sports program. If Floyd used, it wouldn't be shocking. Cycling has been dirty for over two decades.
Having said that, I have serious and well-founded doubts that organization ssuch as the WADA or UCI can be effective at making determinations about drug use, at least not without checks and balances and good independent oversight.
My understanding of the underlying issues goes beyond the mere anecdotal. I've worked professionally as a researcher in gene toxicology at the NIEHS and later helped start two organizations in the US federal government that evaluate governmental test method standards both in the US and internationally. Let's just say I'm accustomed the the bureaucratic influences involved in setting standards like the ones used for detecting drug use in pro athletes. I have been first-hand witness to a phenomenon worse than groupthink: political machinations, dictated by greed, deliberately undermining the quality of well-researched test standards.
In December 2006 I completed graduate school. I studied the adoption of innovations by pharmaceutical companies for my master's thesis. My academic study focused on biomedical informatics: consumer health, evidence-based medicine, clinical and biotech data mining, and biomedical text mining. I took the Summer of 2006 off from school, working as an enterprise architect responsible for the reorganization of the SAS corporate support site, optimizing findability for the 4 million or so SAS users. I performed a great deal of language use analysis geared towards making sense out of large collections of expert domain documents, the domain of expertise in this case being statistical analysis.
I am currently seeking employment and have recently enjoyed some good interviews. I haven't found the right match yet. I want to get it right.
In my spare time I write poetry. I've written five books to date but have stopped writing for the moment. The second book of mine to be published will be released this fall by Effing Press. The forthcoming book, written by my puppet Lester, has enjoyed a cult status among poets for years.
12 Comments:
The key paragraph for me was,
The test determines whether the testosterone in the athlete’s urine has less carbon-13 than another naturally occurring hormone in the urine, like cholesterol. The test is considered positive when the carbon isotope ratio — the amount of carbon-13 compared to carbon-12 — is three or more units higher in the athlete’s testosterone than it is in the comparison hormone. It is evidence that the testosterone in the urine was not made by the athlete’s body. Landis’s difference was 3.99, according to his own doctor.
Apparently the IRMS is tricky enough to run that a 25% error might be possible and this could be refuted with the B sample test. If so, then that would lead right back to the T/E ratio, about which there could be a lot of jockeying and argument.
So, without more solid IRMS findings on the B sample, this could be a lengthy appeal process.
So 25% of 3 is 75 right? So a reading of 3.75 or lower is within the margin of error. 3.99 is not. Looks like he's toast.
Liberty Seguros have not been cleared, Communidad Valencia have, and since they rose from the ashes of proven dopers Kelme, they cant complain too much.
The guilty first innocent later issue has arisen because of the scandalous record of cycling regard to doping. Maybe its not justice as we are used to it, but the sport only has its self to blame.
I'm an editor with WSJ.com and wanted to point out Carl Bialik's (The Number's Guy) new piece on alcohol and testosterone.
http://online.wsj.com/public/article/SB115444904804023475.html?mod=blogs
I want to believe, I still believe he is not guilty........
From VeloNews:
Manzano: ‘Testosterone effects almost immediate'
Jesús Manzano, the ex-pro who revealed insider details of doping in a series of paid interviews two years ago, said that testosterone can be taken during competition and its effects can be felt "almost immediately."
In an article with his byline in the Spanish daily AS, Manzano outlined the way riders can take testosterone during a competition, undercutting the argument that testosterone is a substance taken over weeks and months to slowly build strength and resistance.
Manzano's interviews helped draw attention to controversial Spanish doctor Eufemiano Fuentes, who is now at the center of an ongoing doping investigation in Spain.
"Its effects are felt almost immediately," Manzano wrote. "It gives you a lot of force and produces a sort of euphoria."
Manzano outlined three ways to take testosterone during a competition.
First are patches, called AndroGel, which are applied during a light massage usually during the evening mealtime. Manzano said the patches had to be used less than two hours to not risk going above the T/E ratio threshold and risking a positive doping test. Manzano also said clandestine suppositories are used as well as intravenous injections, called Rastandol, taking about 20 minutes to 1 hour before competition.
Another option are pills of Andriol, called "beans" in the cycling underworld - or "Ferrari's jellybeans" by some - but these usually take several days to kick in and he said it's not likely riders would use them during competition.
So, the NYT article today says "three or more units" higher for C-13 concentration, for the meat-based testosterone, than for the soy-based testosterone used in synthesis.
What are these units? Parts per million? Parts per billion? Percentage points? It makes a difference -- the tinier the concentrations, the more likely the test is to be wrong.
Apparently the differences are based on the fact that plants like soybeans absorb atmospheric C-13 less readily than animals (incl. humans) do. About 1% of terrestrial carbon is C-13, so there's plenty of it around.
The basis for the IRMS test is the ratio of 13C to 12C in the urinary testosterone metabolites. This ratio is often expressed as the d13C-value, which is defined as parts per thousand deviation of isotope ratio of the sample from the standard:
d13C-value = 13C/12Csample - 13C/12Cstandard x 1000
13C/12Cstandard
When the difference between the sample and standard is 3.0 parts per thousand, the result is determined to be “consistent with exogenous steroid administration.”
The precursor of all steroid hormones, including testosterone, is cholesterol. So, for most humans the primary urinary steroids will have originated from cholesterol consumed in animal foods, i.e., meat, dairy products, and eggs. In contrast, synthetic steroid hormones are made from sterols isolated from plants, usually phytosterols from soybeans. Plants vary in their 13C/12C ratio depending on which photosynthetic pathway is used to take CO2 out of the atmosphere and use it to make glucose and, ultimately, sterols. Plants that use the Calvin Cycle (C3) pathway incorporate relatively more 12C than 13C compared to plants that use the Hatch-Slack (C4) pathway. As a result of the different photosynthetic pathways, the d13C-value of C3 plants ranges from -35 to -22 compared to -20 to -8 parts per thousand for C4 plants. Soybeans, the predominant source of sterols for synthetic steroids, are C3 plants; hence, the lower d13C-value in the urinary steroids of exogenous testosterone users.
There is significant ethnic variation in the d13C-values for urinary steroids. For example, Caucasian athletes (non-steroid users) were reported to have values that were 1-2 parts per thousand lower than those of Japanese subjects (Ueki, 1999). These differences may be due to differences in diet. Urinary d13C-values are higher in individuals residing in Africa compared to those living in western countries. This difference can be attributed to the relatively greater consumption of C4 plants, such as maize and sugar cane, in Africa compared to western nations. A recent longitudinal study of 3 elite male runners reported that the d13C-value of urinary steroids is altered by diet. In this case, the d13C-values increased by up to 2.5 parts per thousand after 5 weeks living in Kenya or South Africa compared to usual values determined while the subjects were living in Switzerland the remainder of the year (Saudan, 2006). In this study, the changes in d13C-values resulting from diet produced values that differed from the standard by 0.7-0.9 parts per thousand. In other words, these results would not have produced a test consistent with exogenous steroid administration. However, the upper limit of the diet effect on d13C-values in urinary steroids is not known.
Since the comparison is between the c13/12 ratio of testosterone and another naturally occurring, but non-performance enhancing substance IN THE SAME SUBJECCT, ie Floyd, the variation based on ethnicity is immaterial.
BTW, 3.99 is 33% greater than 3. He's over the limit by a significant amount.
Here's an interesting Usenet post:
A clue of the Landis affair, might be the case Theunisse in 1990. Theunisse tested positive after la FlPche Wallone: his T/E ratio was far above the 6:1 allowed in those days. Of course, Theunisse protested his innocence and claimed to have a unusually high T/E ratio, which was fluctuating in a exceptionally way. Almost nobody believed him, but he was probably right - which by the way didn't mean he was innocent. About a year later a Belgian journalist let me listen to a tape recording of a interview with one of the soigneurs of PDM (the team of Theunisse). It was a very interesting story. The politics of PDM in these days was: we're allowed to give our riders as much testosterone as long as their T/E ratio doesn't exceed the 6:1 ratio. Of course, they tested the team regularly to keep from unexpected surprises. Consequently, it was a great surprise that Theunisse had already once tested positive on testosterone. They assumed they must have made a mistake, but they didn't want to take any risk. So when Theunisse in 1990 believed to be capable to win the FlPche Wallone, three days before the race the whole team was tested by the team doctor after a long training ride in the Ardennes. Theunisse's T/E ratio was below the 6:1, so they had nothing to fear, especially because of course they were not so stupid to take a new dose testosterone before the race. And yet, after the FlPche Wallone (I think he finished third) Theunisse tested positive. No wonder he felt cheated, proclaimed his innocence and suggested he must be suffering from a congenital physical abnormality. The explanation (which was corfirmed by a specialist I interviewed later) is rather simple. Under normal conditions it certainly would have been a freak result, but pro's are riding in exceptional conditions in which the body is sometimes reacting exceptionally, especially because it is already functioning in a unusual way, because riders are usually taking several different products which might have strong effects (that's the intention anyhow).
Of course it's pure speculation, but what could have happened is that Landis had taken the standard set of products: steroids, hgh, epo, testosterone, etc., and in such quantities that he seemed to have nothing to fear. However, under the exceptional conditions - exceptionally hard mountain stages, the heat, dehydration, etc.- his body might have reacted in the same unexpected way as Theunisse's 16 years before.
To Roadchick,
Yes, the ratio of 13C to 12C can be changed by your diet. In fact, it has been used to differentiate between vegetarians and omnivores. Vegetarians eat more C3 plants which have less 13C in them. As a result, their ratio of 13C/C12 is lower than that of omnivores. It is not, however, appear to be low enough to cause a positive test for exogenous testosterone.
Floyd posted a message on his blog this afternoon reiterating his denial of doping. He also states that in his t:e ratio, his t level was normal and his e level was unusually low. This is the first actual confirmation of this fact. Before this it was only speculation or rumor, from what I'd been able to find.
His statement seems to be influenced by his lawyers, but he still maintains his innocence in a Floyd-like way. Perhaps the miniscule chances of error, sabotage, freak of nature, etc. have aligned against him, but it's getting harder to believe. I still hope he can prevail.
roadchick, the answer is a definite yes: diet can lead to a false positive. i found an article two nights ago on the subject, and propam in her comments here cited the very same article.
Post a Comment
<< Home