Monday, July 31, 2006

Landis busted?

According to the New York Times (link) in a late-breaking story, Landis tested positive for exogenous testosterone via an IRMS test performed by the LNDD. The source of information comes from an unnamed individual with the UCI.

Hold on to your hat folks. Apparently Landis did not file for testing of sample B despite claims of doing so, but the UCI went ahead and requested that verification.

No official word from a single named source has yet clarified whether IRMS was used on sample A and/or whether it will be used on sample B. An earlier story from the New York Times (link) claimed that IRMS was used but the source of that information reported several inaccuracies on the subject of the test and, further, the NYT story itself was ambiguously worded.

Saturday, July 29, 2006

A Diversion: Why I Love Cycling

So when as why I love cycling, the only reason to come forward is that it’s like a best friend and like old friends I have had fights with my bikes, fights which have given me a concussion and several scars and numerous near death experiences. My bike rides have showed me more places and taught me more things than I can list; I have meet people though my bikes; I have also found jobs with them and vacationed with them; and just as friends do, it also has grown with me. The relationship now is far less innocent and is taken less for granted; nonetheless, what remains is the knowing that every time I cross a leg over the saddle and clip my shoes into the pedals, it’s going to be a fun time out, with a faithful friend. That is why I love cycling.
- Benjamin Lyon, "Why I Love Cycling" (

Friday, July 28, 2006

The WADA and IRMS: a positive is always positive, but a negative can be positive too

According to the WADA's guide on T/E testing, if you test positive using the IRMS test, you're definitely guilty. But if the IRMS test clears your sample, then you can still be held as guilty of cheating.

That's right folks. If the definitive test finds you guilty, you're definitely guilty, but if the definitive test clears you, the WADA can still say you're guilty anyway.

According to the WADA guidelines, the labs administering the T/E tests don't have to use the IRMS test.

The IRMS test, to refresh your memory, in effect distinguishes between naturally occurring testosterone levels and artifically-induced levels. They can count the differing types of two types of carbon isotopes. If the one's high and the other is low, you're clearly doping, but if the one is low and the other is high, your own body is producing those high levels appearing in the T/E test. The test is expensive, unlike the T/E test, but conclusive, unlike the T/E test.

Apparently, in the case of Floyd Landis, there are four decision points involved in choosing whether to use IRMS: (1) the WADA lab can choose to do the IRMS as part of the T/E test protocol, which they rarely do because it is relatively expensive; (2) the WADA lab, in its report to the UCI, can choose whether to recommend that an IRMS be performed on sample B; (3) regardless of the WADA recommendation coming from the LNDD, UCI can order the LNDD to perform the IRMS on sample B; and finally (4) I've heard that Landis can demand it. This latter point is not in the report, but apparently the UCI allows the riders the opportunity to be involved in the requirements setting.

If Landis doesn't request an IRMS, then he's either guilty or poorly informed, and if he does request an IRMS, he is at the very least maintaining an appearance of innocence.

Also indicated in the WADA guidelines is an estimation of the timeline and procedure should Landis' B sample be tested using the IRMS and come out negative. If he comes out positive using the IRMS, the UCI ruling should come swiftly. But if he comes out negative, there are a number of steps involved that can prolong this process as many as four to five months. The length of time will depend on the subjective evaluation of a previous testing history. If there's not much of a history, then they will surprise him with random tests over the course of 1 to 3 months to establish that history.

It is likely that the UCI already has a history of tests with Landis. What isn't clear from the WADA guidelines, however, is whether the history of tests must include a history of IRMS results, or, alternately, whether the follow-up tests if he has no history must also be IRMS tests.

The implication of this is that if Landis is truly suffering from a serious thyroid ailment needing immediate treatment and the UCI is deficient in testing historyon Landis, then Landis will either have to forego treatment of the thyroid problem and continue to suffer from it or get that treatment and doom himself to a de facto positive result.

Lab same as one in the Lance controversy

It was only a few months ago that the WADA, the organization running the lab testing Landis' sample A, came under attack for its handling of the Lance doping scandal. The UCI approved a scathing report (link) of the WADA lab, and the WADA threatened to sue the UCI (link) after releasing their own response to the UCI report (link).

The lab in question is known as the LNDD (Laboratoire National de Dépistage du Dopage), the French national anti-doping lab in the Paris suburb of Chatenay-Malabry.

And only five months later, the same lab and the two supposedly divided organizations are together to publically release the premature guilt of Floyd Landis.

The Vrijman report, in its critique of the WADA's handling of the Lance controversy, found that the WADA violated its own assumptions in order to attack Lance Armstrong. In the conclusion Vrijman writes,

WADA had also the intention that the research results, in combination with the additional information requested by WADA, be used for disciplinary purposes against individual athletes, directly contrary to its representation that the results would not be used “for any sanction purpose ”. In this sense one can speak of targeting by WADA of the participants of the 1998 and 1999 Tours de France (128).

Vrijman also found that "the LNDD violated applicable rules on athlete confidentiality by commenting publicly on the alleged positive findings, especially in relation with a particular rider, Lance Armstrong."

Nobody knows nuthin'

Apparently no one outside of the UCI and/or the WADA-based testing lab in France involved with the testing of sample A knows whether IRMS was used or not. This from the Providence Journal ("Tour winner fails drug test", link)

Landis is entitled to be present when his "B" sample is tested, but he said he wasn't sure if he would avail himself of that privilege. Testosterone is considered among the trickier drugs for testers to pinpoint because it occurs naturally in the body. WADA requires that samples that possess greater than a 4:1 ratio of testosterone to epitestosterone be subject to further analysis, which includes the gathering of at least three additional urine samples to show that the high ratio is indicative of drug use. This step can only be bypassed if the sample was analyzed with what is known as the carbon-isotope ratio (CIR) test, a more expensive and complex testing procedure that can differentiate between natural and synthetic testosterone.

Landis said he did not know which testing methods were used. The simple ratio test alone has been subject to extensive criticism.

If Landis does know then the only reasonable explanation I can think of is that he is lying to create as much doubt about his guilt as possible. If he is telling the truth, however, and he doesn't know whether IRMS was used, then his guilt remains an open question. Why hasn't the information been released publically? I suspect it is because, for starters, no one in the world of journalism has put enough energy into the story to realize that this is THE question of the moment, and so journalists seem not to be asking. Also, there may be a regulation restricting how much these organizations like UCI and WADA can say. Again, I must be clear with you that I'm speculating here. I don't know.

Something's amiss, WADA doc claims

Something may be amiss with the result of the test, both a former pro cyclist and a WADA physician speculate. The WADA is responsible for the lab in France that administers the tests on the riders of the Tour de France.

From "Mark Conley: We shouldn't be surprised by Landis' plight" (link)

Since the sample taken after that stage is the one that has implicated Landis, it's easy to connect the dots and speculate his super-hero performance was anything but au naturale.

At least one former pro rider, "Fast Freddy" Markum of Santa Cruz, believes that's a bit too easy of a connection.

"They had probably tested him a dozen times over the course of the race to that point, so the last one comes up positive?" Markum says. "I smell a rat."

Markum, a two-time Olympian who raced against Greg LeMond in his prime and Armstrong in his infancy, smells a French rat, to be specific.

"The French have been so anti-American since Lance's second or third Tour win, they've tried to tear him down for so long now," Markum says, referring largely to the newspaper L'Equipe. "It's just a witch hunt at this point. They're tearing their own sport apart."

A World Anti-Doping Agency doctor agreed Thursday something seems amiss, telling an Associated Press medical writer, "Something's missing here. It just doesn't add up." If Landis had been cheating all along, then why wasn't there a positive test earlier in the race? The idea that he could load up effectively between his 16th stage death ride and his 17th stage super-heroics is bunk, the doctor said.

Thursday, July 27, 2006

IRMS Apparently NOT Used on Sample A

For years it has been widely known that the measurement of T/E ratios in urine is subject to an extraordinarily high rate of false positives and false negatives. The IOC, WADA, and other organizations setting standards for drug testing of athletes have known this for years and have even tried to address it. Namely they've addressed it by setting standards for the use of IRMS: isotope ratio mass spectrometry. I discussed earlier a test to evaluate the 13C/12C ratios, and that test *is* the very test known as IRMS.

Now WADA and UCI know of and use IRMS. But, guess what? Despite the fact that unlike T/E testing it's relatively safe from massive rates of false positives and false negatives, they don't regularly use it.

And the word is that Landis' sample A was not tested using IRMS. In other words, this positive test result borders on slander and neglect. It is a disgrace that the UCI would not perform an IRMS test on sample A. Disgraceful, utterly. they have a definitive test at their disposal. Why not do it in the first place?

This spooks me, as I immediately begin to worry about either (a) overzealous bureaucratic crusaders hostile to athletes, just as law officers can become callous to all citizens, or (b) some external and possibly political motivation for not doing it right for the first time.

T/E ratios...IRMS...the public will only know "positive" or "negative." And regardless of any future IRMS Landis has already been disgraced. Let's hope that an IRMS is at least administered to sample B. Let's know the truth, or at least a pretty good estimator of the truth rather than this T/E junk.

Ethanol and T/E Tests

A comment referred to the amount of ethanol needed to mask. In looking this up I was astounded to find that ethanol actually elevates T/E ratios dramatically. Here's an article I found online and it supports what I've discovered today, that the T/E test is bad science:

The Effect of Alcohol Consumption on the Urinary Testosterone / Epitestosterone Ratio

By Dr Simon Davis B.Sc., Ph.D.


Intoxicating beverages contain a number of different forms of alcohol, the major constituent of which is ethanol. When a beverage is consumed the ethanol content passes through the stomach wall and digestive tract into the blood stream. Once the ethanol enters circulation it begins to alter the bodies’ biochemistry. One such reaction is to differentially increase the rates of testosterone (T) and epitestosterone (E) metabolism. The overall effect of this reaction is to increase the ratio of T to E excreted in the urine.

It has been reported that ethanol consumption can increase urinary T/E ratios by 30% - 277% in healthy individuals. Observed changes in plasma T/E ratios can occur with the consumption of less than 2 pints of lager. The ingestion of ethanol by an individual will increase the T/E ratio observed in a urine sample.

It follows that if the effect of ethanol on T/E ratios is calculated relative to urinary E concentrations, it can be seen that increases in the ratio are exponential as E concentrations decrease. Individuals with naturally low E concentrations could, therefore, experience increases in T/E ratios of ? 940% greater than increases experienced in an individual with normal E concentrations. Calculations estimate that in individuals with low urinary E concentration, ratios of 17 to 1 or higher could have resulted from ethanol consumption without any administration of exogenous T.

The current T/E ratio test as performed by Kings College Laboratory and approved by the UK Sports, the IWF and IOC cannot discriminate between a 13 to 1 T/E ratio resulting from ethanol ingestion or a 13 to 1 ratio resulting from endogenous T administration.

Is there a definitive test?

Cycling's ruling body, Union Cycliste Internationale or UCI, will administer a second test to a different sample from Floyd Landis. This is what is commonly known as a copunter-analysis. In other words, they will use a different method than the previous screen to determine whether Flandis' elevated ratio is either a naturally occurring phenomenon or one caused by illicit actions of Landis'.

So, OK, these people aren't completely crazy.

But what constitutes a definitive test? And does the UCI use a definitive test method in their counter-analysis?

Let's go back to the experts.

According to a study performed by the Swiss Laboratory for Doping Analyses in Lausanne [C Saudan, N Baume, N Robinson, L Avois, P Mangin and M Saugy, "Testosterone and doping control", British Journal of Sports Medicine 2006;40(Supplement 1):i21-i24)] the ratio test used initially is what's known as an "indirect method." An indirect method by definition isn't definitive; it merely detects things that are frequently but not necessarily associated with illicit activity.

The study also indicates that the 4:1 ratio test (known as the T/E ratio) can be masked if the steroid is administered with ethanol.

It can be safely said that someone with half a brain administering to himself anabolic steroids and undergoing testing would administer the drug with a bit of ethanol and pass the test.

Clearly, at least to me, the T/E ratio test for competitive athletes is a load of junk. Ir produces high false positive and false negative rates, and the false negative rate can be easily jacked through simple methods, such as administering doasges with ethanol. That false negative rate can also be jacked by taking epitestosterone. This test is, simply put, easy to beat--if you have a reason to worry about it, that is.

OK, this indirect test, and perhaps any indirect test for that matter, relies only on half-baked science. What about a *definitive* test?

According to the aforementioned Swiss study, "(d)irect evidence may be obtained with a method based on the determination of the carbon isotope ratio of the urinary steroids" (i21). In fact it is mandatory for the International Olympic Committee to determine definitively through a secondary test whether the steroid is a natural steroid or an artificial one. Even if the ratio test method is tuned to the specific athlete over a long period of time to detect anomalous fluctuations in ratios, "there is a lack of definitive proof for the exogenous application of natural steroids," they write. Further, the authors of the study add,
One possible way of solving this problem is the ratio of the two stable carbon isotopes 13C/12C, which can allow the differentiation of natural and synthetic steroids. As exogenous testosterone or precursors contain less 13C than their endogenous homologues, it is expected that urinary steroids with a low 13C/12C ratio originate from pharmaceutical sources.

Does the UCI actually use such a method? Do they test for carbon isotope ratios? Or do they simply re-test Floyd with the same or another indirect method?

Even if such a definitive test is used by the UCI it remains to be seen whether cortisone would alter it.

Questions Linger. Can the administration of cortisone elevate naturally occurring testosterone levels or lower naturally occurring epitestosterone? Can extreme pain do it? What about extreme athletic performance? How about looking at the frequency of statistical outliers?

We'll take a look at the World Anti-Doping Administration's guidelines on elevated T/E ratios.

From 6:1 to 4:1: The creation of an unfair testing standard?

Floyd Landis failed his drug test on the basis of levels of his urinary testosterone. Specifically the testers look for the ratio of testosterone to epitestosterone. Anything above 4:1, according to numerous sports organizations, from the Olympics to the NFL to the UCF, constitutes a positive indication of artificial testosterone use.

In the past the ratio of positive evidence was 6:1 but it has been lowered uniformly to 4:1.

Landis tested positive at a ratio of somewhere around 5:1 putting his test result in that range where he would have been negative before.

According to a highly-cited and definitive controlled study of the effects of testosterone administration (DEHENNIN L, MATSUMOTO AM
JOURNAL OF STEROID BIOCHEMISTRY AND MOLECULAR BIOLOGY 44 (2): 179-189 FEB 1993) the 4:1 ratio may be a bad measure leading to unfair false positive results. In fact the study shows that even a 6:1 ratio leads to false positives in a significant number of cases.

In a placebo administration to 10 study participants, the study showed that 20% of the participants started with levels above 4:1 even before the administration of a placebo. One of the participants started above a 6:1 ratio without the administration of anything whatsoever.

It is likewise important to note that the participants in the study were normal 28-30 year old men. They were not athletic outliers with bodies undergoing the dual stresses of a 2000 mile bike ride under extraordinary levels of pain. And they weren't taking cortisone, a steroid injection. Even 20% of these average Joes would have failed the test simply by sitting on a couch eating potato chips.

Questions linger:
Can either pain or the administration of cortisone induce elevated ratios?
Is there a definitive test, if the 4:1 ratio test neither confirms nor denies tha actual use of testosterone?
Would the administration of cortisone create a false positive even in a definitive test?

Free Floyd Landis

Apparently Floyd Landis has tested positive for use of steroids and he is at risk of losing his standing and credibility.

However Floyd may be a victim of a poor testing methodology rather than a rule-breaker.

On this blog I will investigate the validity of the test method as the Landis story develops.