NEW YORK (AP) – The test to detect synthetic testosterone – used on Tour de France winner Floyd Landis’ urine sample – is highly accurate, experts say.

Philippe Verbiest, the lawyer for cycling’s world governing body, confirmed Tuesday that a carbon isotope ratio test, which detects a subtle chemical difference between the natural and synthetic versions of testosterone, was done.

The New York Times has reported the test detected the man-made hormone, but Verbiest refused to give any details of the result.

The test focuses on carbon atoms in testosterone molecules. These atoms come in different types, called isotopes. The ratio of one particular isotope to another isotope is different in synthetic versus natural testosterone. The test detects that difference.

When the lab procedure is performed correctly it is very accurate, said Dr. Don Catlin, the director of the UCLA Olympic doping lab. He helped develop the test.

Dr. Gary Wadler, an internist and author with expertise in the area of drug use by athletes, said he’s not aware of anything other than synthetic testosterone that will produce a positive test result.

When athletes take synthetic testosterone to boost performance, it typically helps them get stronger, recover faster from workouts, prevent tissue breakdown and increase their assertiveness, said Wadler, a member of the World Anti-Doping Agency who stressed he was not speaking for that group.

Testosterone can be injected, he said, though athletes are moving toward taking it by skin patch or cream, which allows lower doses that might escape detection. Typically, athletes take the hormone daily for weeks or months during training, Wadler said, because it takes awhile for the benefits to appear.

Taking doses of synthetic testosterone would not raise the body’s production of the natural hormone, and if anything it might suppress that production, Wadler said. It also might suppress production of epitestosterone.

Landis tested positive for an unusually high ratio of testosterone to epitestosterone after a tough Tour Alpine stage on July 20 – when he staged a remarkable comeback just a day after a poor performance.

Michael Henson, a spokesman for Landis, confirmed Tuesday that the cyclist’s July 20 test turned up an 11:1 ratio – far above the 4:1 limit. The results of the backup sample will be announced Saturday.

People typically have a ratio of 1:1 or 2:1. Once an athlete’s level surpasses 4:1, the test to look for synthetic testosterone is administered.

Few people reach that level naturally, Catlin said.

“The human body gets (its) testosterone ratio at birth, and it stays with you for life. It could vary a little bit, but it will never waver more than 30 percent from your mark,” he explained.

“You can map out your levels through life, and if it spikes significantly, that’s a spike pattern and it’s not natural – something must have been taken to get into the testosterone metabolic chain.”

The lack of a positive test before July 20 might suggest Landis took a single dose and then immediately put in his stunning performance.

A person also can increase his body’s production of testosterone by taking another substance, called luteinizing hormone, Wadler said. But that substance also is banned.

There are legitimate reasons for an athlete to take synthetic testosterone. If an athlete’s body doesn’t make enough of the hormone naturally, such as after the testicles are removed surgically, he could seek an exemption from anti-doping rules, Wadler said. Landis does not fall into that category.

As for suggestions that Landis’ consumption of beer and whiskey could have raised his testosterone to epitestosterone ratio, that still would not explain that dramatic a spike, Catlin said.

“There are a few cases of people getting stoned drunk and they can get a bit of a spike in their ratio – up one or two – but that increase is rare,” Catlin said.