The Sports Gene: Inside the Science of Extraordinary Athletic Performance by David Epstein Page B

and bioethics in the Feinberg School of Medicine at Northwestern University and an authority on the history of sex testing in sports, told me: “The reason we have females separated in sports is because in many sports the best female athletes can’t compete with the best male athletes. And everybody knows that but nobody wants to say it. Females are structured like a disabled class for all sorts of, I think, good reasons.”
    The difficulty in determining who is granted access to that class was evident at the 2009 world track and field championships whenCaster Semenya, a young and unheralded South African 800-meter runner, looked over her muscled shoulder and tore away from the field en route to the world title. Semenya’s competitors derided her in the world media. “Just look at her,” sneered Russian Mariya Savinova, the fifth-place finisher, in reference to Semenya’s narrow hips and armored torso. Just looking at her, though, does not give an answer.
    After the world championships, it was reported that Semenya has internal testes and no ovaries or uterus, and high levels of testosterone. (Semenya never confirmed or addressed that report.) So where, if true, should that leave her? To start breaking down sport classifications by specific biological traits, “you’d have to run international competitions like the Westminster Dog Show, with competitions for every breed,” says Myron Genel, the Yale professor of pediatrics. María José Martínez-Patiño, the Spanish hurdler, had both a Y chromosome and an SRY gene, but because she was insensitive to testosterone she was ultimately allowed to compete against women.
    Before the 2012 London Olympics, faced with continuing controversy over the Semenya case, the IAAF and the International Olympic Committee announced that sex would be determined based on testosterone levels. Not just the amount that is produced, but the amount the body can use.
    Testosterone levels are not on a continuous spectrum. A typical woman will make less than 75 nanograms of testosterone per deciliter of blood. For men, the range is typically 240 to 1,200. So the low end of the male range is still more than 200 percent higher than the high end of the female range. In 2011, the NCAA—informed by a think tank held with the National Center for Lesbian Rights—determined that any man who undergoes sex reassignment surgery to become a woman must sit out a year while lowering her testosterone levels before she can compete on a women’s team. Thus, testosterone has been deemed the source of the male athletic advantage. Though it may not be the only one.
    When I spoke with endocrinologists who work with androgen-insensitive women, they all felt that XY women with androgeninsensitivity—that is, like Martínez-Patiño, they can use no testosterone at all—are
overrepresented
, not underrepresented, in sports.
    At the 1996 Atlanta Summer Olympics, the last that had cheek swabs, 7 women out of the 3,387 competitors—or about 1 in 480—were found to have the SRY gene and androgen insensitivity. The typical rate of androgen insensitivity is estimated to be between 1 in 20,000 and 1 in 64,000. Over five Olympic Games, an average of 1 in every 421 female competitors was determined to have a Y chromosome. So women with androgen insensitivity are vastly overrepresented on the world’s largest sporting stage. Perhaps, then, something about the Y chromosome other than testosterone may be conferring an advantage.
    Women with androgen insensitivity tend to have limb proportions more typical of men. Their arms and legs are longer relative to their bodies, and their average height is several inches taller than that of typical women. Like Erika Coimbra, a 5'11" Brazilian volleyball player and 2000 Olympic bronze medalist who is one of the few athletes with androgen insensitivity whose names have ever been made public. (Two of the endocrinologists I spoke with said that XY women are also overrepresented in