Diet Can Elevate Test Results, But How Much Is Unknown
Does that mean that if Floyd Landis ate a heaping helping of sugarcane after stage 16 he could have triggered a false positive for the IRMS?
Saudan, et al (2006) speculate that such a diet couldn't do such a thing. However they admit that there's much to be learned, and they also admit in their study that diet in experimental subjects was not monitored, logged, etc. So, in other words, the study assumes that there is some increase in dietary 13C but makes no strides to monitor the correlation between intake of certain foods and the degree to which 13C is elevated in tests. As the test was conducted, people who lived in Africa elevated their 13C by as much as 3%, certainly not enough to trigger a false positive.
But it is interesting to note that the % increase in 13C increased linearly the entirety of the time the test subjects were on this Kenyan diet. There's no decay in the increase, even over a 3 month period. Ostensibly, then, if one were on this diet for a year, then one might see a 12% increase in 13C, two years, 24% increase, and so on. The study lends itself to this interpretation.
If Floyd Landis ate somewhat like a Kenyan for three months, consuming some amount of sugarcane and millet, he could have elevated his 13C levels by 3%. That would have put his test results at 30% above test error, still leaving him 5% away from dubious test results. But if he's been eating loads of millet sweetened with sugarcane for a few years, it very well may have put him in false positive territory.
Several questions remain: how much does the intake of *any* amount of exogenous steroids elevate one's testable 13C levels? In other words, how uncommon is it for someone taking any amount of steroids to test as low as Landis did? Another question: how much sugarcane or millet or pineapple would it take to trigger a false positive? Some cyclists are fanatical about their diets, and many aspects of their diets are aspects that have witnesses. If the guy ate millet breakfast lunch and dinner there would be many individuals who could confirm it. Further, what effect does alcohol, a product of fermented sugar, have on the IRMS?
Whil;e the IRMS is a very good test, there are ways a false positive may be arrived at. The ways in which a false positive result is possible seem apparent yet they are poorly understood.
One thing is clear: if I ever get my knee healthy and try to compete at cycling, I'm going to eat lots and lots of millet and pineapple and give myself a natural means of supplementing with androsterone.
Seriously, though, the moral should be that if the UCI wants to test cyclists they should bear the burden of keeping record of each cyclist's diet.
(Short communication: Christophe Saudana, Matthias Kamberb, Giulia Barbatic, Neil Robinsona, Aurélien Desmarcheliera, Patrice Mangina and Martial Saugya. Longitudinal profiling of urinary steroids by gas chromatography/combustion/isotope ratio mass spectrometry: Diet change may result in carbon isotopic variations. Journal of Chromatography B, Volume 831, Issues 1-2 , 2 February 2006, Pages 324-327.)