How the practice of sex-testing targets female Olympic athletes

Doctors inspected the women’s genitalia and found some atypical features. But genetic analysis and physical exams revealed something more unusual: all four had XY chromosomes rather than the usual female XX, and their internal reproductive glands were testes, not ovaries.

The athletes were diagnosed with 5-alpha reductase deficiency. Because of a genetic quirk, their bodies do not produce an enzyme that converts testosterone into a secondary hormone involved in the development of male characteristics. These women were chromosomally male, but had lived their lives, and competed athletically, as women.

The women, the doctors reported in a 2013 journal article, had “many questions” about the biological landmarks of female life: child-bearing, menstruation, sex. But from a competitive perspective, there was a different concern. Their testosterone levels were considered by the International Olympic Committee and International Association of Athletics Federations to be in the male range, and therefore too high to compete fairly.

The doctors proposed a solution: gonadectomies, or surgical removal of their reproductive organs; clitoridectomies, or partial surgical removal of their clitorises; and, later, surgeries to feminize their vaginas and estrogen replacement therapy. These measures were not medically necessary, but after a year, the women would be cleared to compete. All four agreed.

Criticism from bioethicists, sports activists and intersex scholars was loud and furious.

Had these perfectly healthy women really just undergone clitoral mutilation for the purposes of sport? The athletes were competing with the hormones their bodies produced naturally, not doping — why is natural genetic variation policed in female athletes, but celebrated in men, like freakishly tall basketball players? What evidence is there, anyway, that testosterone is the sole predictor of superior athletic performance?

If the particulars were new, the fight itself was not. Ever since women began competing at elite levels, sports governing bodies like the Olympics have been searching for a way to cleanly divide men from women. Supporters say these boundaries are necessary to create a level playing field for female athletes.

Critics charge that these divisions are just metastasized social anxieties about gender. It is certainly true that crucial leaps in our understanding of sex differentiation were the result of studying animals and humans that warp our binary definitions. Even hard-nosed scientists admit that the field floats atop a fathomless public fascination: what makes a man, and what makes a woman?

But the more research we do, the less easy it is to answer that with a clear biological boundary. The public is by now used to the idea that sex is biological while gender is a social construct. But where in our biology does sex reside: in our genes, in our genitals, in our hormones? And is it even possible to separate biological sex from the environmental influence of gender?

“It is very difficult to come up with an absolute line,” says Arthur Arnold, distinguished professor of integrative biology and physiology at the University of California Los Angeles. “The Olympic committees have struggled with this and have had different lines, all of which have broken down in various ways.”

Last year, the IAAF was forced to lift its rules on female athletes and testosterone, and the IOC has begrudgingly followed suit. When the Games begin in Rio on Aug. 5, there will be no policy in place at all. It’s not clear what tests or policies will eventually replace the revoked rules — or even if there should be a policy.

“It’s kind of a mess,” says Arnold.

The hormone red herring

Frank Lillie was already a highly regarded embryologist — his 1908 book on chick development was a classic of the field — when the manager of his hobby farm north of Chicago, where Lillie kept a herd of purebred cattle, alerted the professor to the birth of a freemartin.

Freemartins are the barren female twins of normal male calves. While their bull brothers develop typically, freemartins often bear odd-looking genitals, and their sterility had stood as a mystery for centuries. Lillie, a Toronto-born scientist who rose to prominence at the University of Chicago, was intrigued. He sent word to a local abattoir, and by 1916 had collected and examined 41 cow uteruses bearing bovine twins.

His “Theory of the Freemartin” jolted the study of sex differentiation, which examines how male and female traits develop. Because the twinned calves shared a blood source, Lillie proposed, the male calf’s sex hormones coursed through the female embryo, masculinizing her internal reproductive organs into testes-like structures, regardless of her own genetic sex.

A decade prior, scientists had discovered the sex chromosomes, the XX or XY that “determine” whether an embryo is male or female.

But Lillie’s work lifted the locks on a barely understood factor in human development — sex hormones — and turned a trickle of interest into a powerful current that would bend the fates of those four female athletes.

By the 1940s, after the French researcher Albert Jost castrated fetal rabbits and found that without testosterone they all developed as females, the paradigm of sex differentiation was set.

A fertilized egg receives an X chromosome from its mother and either an X or a Y from its father. But for the first several weeks, a human embryo is a sex-neutral lump: the ridge that will develop into gonads is “bipotential,” and the precursors of both female and male sex ducts exist.

In males, at about eight weeks’ gestation, a gene on the Y chromosome causes the testes to emerge and secrete a cascade of hormones that push the embryo down a masculinizing pathway. One suppresses the female sex ducts. Others, androgens, prompt the male ducts to differentiate into internal sex organs, like the seminal vesicle and the prostate. Later on, an enzyme converts testosterone to dihydrotestosterone (DHT), which masculinizes the external genitalia.

Testes and the hormones they produce, Jost and others thought, are the “active sex differentiator”: they “take over the control” of sex characteristics, as Jost wrote, and “impose masculinity on the whole genital sphere.”

Female development is a passive process, by contrast. The lack of androgens and other male sex hormones allows the female ducts to develop into Fallopian tubes, the uterus and other sex organs.

The exceptions seemed to prove the rule. For 30 years, researchers hunted for the on switch that triggered the production of testes. They finally found it by examining the genes of individuals with XY (male) chromosomes who developed as females. They were missing a gene called SRY. Researchers who spliced this gene into XX mice and observed that the animals developed as male confirmed it: SRY was the command to start the masculine hormonal cavalcade.

But soon researchers discovered that SRY mutations only explain a fraction of “intersex” traits — what researchers used to call hermaphroditism, and now called Disorders of Sex Development.

They found that other genetically male (XY) women had a double dose of a gene on the X chromosome, DAX1, that antagonizes male development, even if SRY is working. Then they found WNT4, another “anti-male” gene on the X chromosome that, if over-expressed, overwhelms the “pro-male” SRY. Rather than a singular bully that “imposed” masculinity, sex differentiation began to look more like a negotiation.

Other tent poles of the paradigm began to list.

In zebra finches, males sing courtship songs that females do not, and the brain regions controlling song are five to six times larger. But when UCLA’s Arthur Arnold began experimenting on the finches by manipulating their sex hormones, their brain structure didn’t respond in expected ways.

By studying animals, including intersex “gynandromorph” birds whose plumage is male on one side and female on the other, Arnold and his colleagues believe that genes on the sex chromosomes don’t just permanently fix an embryo’s tissues via hormones and then fall silent. They are expressed in other cells of the adult body — the brain, the heart, the liver — and act on those tissues directly.

From an on-off switch, biological sex has become a complicated series of parallel pathways. “It used to be that biological sex was pretty simple,” Arnold says. “Now there are more things to think about.”

Meanwhile, social scientists want to know how we ended up with ideas about a dominant male essence and a submissive female essence in the first place. Rebecca Jordan-Young, a women’s, gender and sexuality studies professor at Barnard College, specializes in science and gender. When she examined the literature on hormones and sex differences in the brain, she found a “really interesting pattern.”

“When claims coincide with cultural ideas about male and female nature, and an absolute binary … scientists as well as lay people are much more ready to accept that,” Jordan-Young says. Research that contradicts our gender norms has a tendency to drop out of view.

The sex verification test

At the 1936 Olympics, officials said they had conducted a “sex check” on an American gold-medal sprinter, Helen Stephens. Stephens had narrowly beaten a Polish runner, Stella Walsh, and Polish newspapers questioned the American’s true gender.

Stephens “passed” the first known sex verification test. Officials, apparently, did not feel the need to examine Walsh.

In 1980, Walsh was murdered in a robbery in Cleveland. An autopsy revealed she had ambiguous genitalia, including a tiny, nonfunctioning penis. The revelation caused an uproar. In the Star, a headline declared “Olympic star Stella was a man.” Along with media worldwide, the paper repeated that as fact over the decades.

But the coroner actually found that Walsh’s cells were mismatched: some carried XY chromosomes, and some carried chromosomes with one X, not two. This type of “mosaicism” is caused by embryonic cell division errors, and is also called “mixed gonadal dysgenesis.”

Walsh had mixed chromosomes, mixed internal sex organs and mixed external genitalia. As for her gender identity, the coroner, Samuel Gerber, offered his own assessment. “Socially, culturally and legally, Stella Walsh was accepted as a female for 69 years. She lived and died a female.”

Western anthropologists have long fixated on “third” genders in other cultures: the Samoan Fa’afafine, the South Asian Hijra. But in international sports, authorities found increasingly laboured ways to split the two sexes, eventually mandating sex testing in the 1960s.

“Every generation of sex testing has been accompanied by a generation of activists and critics, and they’ve all pretty much had the same criticisms — which is that sex cannot be differentiated so easily,” says Sandy Wells, a doctoral candidate at the University of British Columbia who studies the history of these policies.

At first, athletes were asked to appear naked before a panel of doctors. The “nude parades” were criticized as humiliating and abandoned.

Critics today would add they are diagnostically useless. Features like clitoromegaly — an atypically enlarged clitoris — are associated not only with intersex conditions but with premature birth and nerve tumours. “Guess what — everyone’s body is different,” says Wells. “More often than you’d think, the opinion of the panel would be contradictory.”

Barr Body testing, more dignified and supposedly more scientific, used cells from a cheek swab, stained and examined under a microscope, to detect a bundled, silenced X chromosome, found in most cells of XX females and visible as a dark blot.

Maria Jose Martinez-Patino, a Spanish hurdler, failed a Barr Body chromosome test in 1985. She was stripped of her national records and had her sports scholarship revoked. “I felt ashamed and embarrassed,” Martinez-Patino wrote decades later. “I lost friends, my fiancé, hope and energy. But I knew that I was a woman.”

Martinez-Patino has androgen insensitivity syndrome. Her chromosomal sex is XY and she has male internal sex organs, but a genetic mutation means her androgen receptors don’t work. “When I was conceived, my tissues never heard the hormonal messages to become male,” she wrote.

She developed as a typical female, and she and her family never doubted her gender. As an athlete, she could not make use of her body’s testosterone. Armed with medical evidence, the hurdler fought the IAAF and won permission to race again. Absent from competition for three years, however, she failed to qualify for the next Olympics.

Through the decades, scientists joined in criticizing the tests as inaccurate and unethical. Nevertheless, athletes who had always known themselves to be women continued to be traumatically “outed.” Between three and 15 athletes at every Games failed a sex test, a figure that does not count those who slunk away from sport to avoid humiliation.

The IAAF finally sponsored two medical-scientific symposia, which concluded what critics said all along: female athletes with sex-related genetic abnormalities, including mosaicism, androgen insensitivity syndrome and 5-alpha reductase deficiency do not enjoy any unfair physical advantage beyond the range of normal human variability. Blanket sex-testing is arbitrary and discriminatory.

The IAAF dropped the practice 1991.

But the IOC lumbered on. It replaced Barr Body testing with DNA tests to locate the newly discovered SRY gene on the Y chromosome. At the 1996 Games, eight female athletes were found positive for SRY. All had androgen insensitivity syndrome or 5-alpha reductase deficiency, and they were eventually cleared.

Dozens of scientific and medical societies had by this point called for the elimination of sex testing, including the American Medical Association and the Canadian and Australian genetics societies.

Whether it was the censure of scientists, disapproval from many athletes, or the cost of a program that had never caught a man masquerading as a woman — its stated purpose — the IOC finally dropped blanket sex testing in 1999.

It was a short-lived victory. Both the IAAF and IOC had retained an important condition: they could medically evaluate an athlete if necessary. A decade later, the case of Caster Semenya would convulse the sports world all over again.

‘Just look at her’

When Caster Semenya, the South African 800-metre runner, posted dominant wins in two championships in 2009, word leaked that officials had singled her out for a sex test. As her white competitors publicly questioned the muscular black woman’s femininity — “just look at her,” one said — Semenya underwent a battery of invasive physical and psychological exams.

Again, an athlete who had lived her life as a woman was barred from competition as her gender was publicly, humiliatingly scrutinized. Semenya was cleared to run again in 2010, and in Rio she will be looking to add to her Olympic silver.

In the wake of that furor, the IAAF adopted new regulations on “female hyperandrogenism.” The IOC followed suit. Female athletes with testosterone levels “within the male range” would be unable to compete, unless they lowered their levels with drugs or surgery. (Those with androgen insensitivity syndrome were excused.)

The sports bodies said the new policies were not sex tests — they were tests for unfair advantage. Observers were dismayed, if not incredulous.

“I’m very proud of my association with the Olympic movement,” says Bruce Kidd, who participated in numerous Games as an athlete, journalist and social scientist, and is now the principal of University of Toronto Scarborough. But, Kidd says, “for a movement that has long tried to link science and scholarship to its mission, it’s an embarrassment.”

Testosterone is a natural place to park suspicions about the source of athletic advantage. Men’s testosterone levels are about 10 times higher than women’s, and elite male athletes are demonstrably faster and stronger than elite female athletes.

IAAF researchers found that women with XY-chromosome disorders of sex development, who have significantly higher testosterone than is typical, are “dramatically” overrepresented among elite athletes compared to the general population.

What’s more, anabolic steroids — the drug of choice for East Germany’s state-sponsored doping program, Canadian sprinter Ben Johnson, and baseball’s Barry Bonds — usually consist of synthetic testosterone.

But no study has ever shown that performance — say, finishing times in a race — correlates to athletes’ levels of testosterone. Testosterone is a complicated, dynamic substance: finishing a marathon causes it to plummet in men, while positive feedback from a coach can cause it to shoot up.

Like every other biological marker, testosterone’s ability to cleanly divide two binary sexes is muddled, especially in athletes.

In one study of nearly 700 Olympians participating in 15 sports, 13.7 per cent of women had natural testosterone levels above the typical female range, 4.7 per cent were within the male range, and 16.5 per cent of men had levels below the typical male range, with some falling into the female range. An IAAF study on 849 female athletes, excluding competitors with intersex traits and other genetic sex diversity, still found outliers.

If hyperandrogenic women are overrepresented in elite sports, they themselves are still vastly outranked by female athletes who are winning with typical hormones.

As for the cheaters, there is little evidence that synthetic or “exogenous” testosterone behaves the same way as natural or “endogenous” testosterone, which is produced by and responds to a dynamic system of hormones.

“There is every reason to be really cautious about how we fill in those gaps, because we tend to fill them in a way that fits really pre-scientific ideas about masculinity and femininity,” says Barnard College’s Rebecca Jordan-Young.

Kidd was part of a team that took the hyperandrogenism rules to the Court of Arbitration for Sport after a young Indian sprinter with hyperandrogenism, Dutee Chand, was barred from competition.

“We have said all along that we support the anti-doping prohibitions against exogenous . . . pharmaceutically created testosterone. But we are bitterly opposed to the idea that a woman can’t run on her own chemicals,” Kidd said.

In July 2015, the court found that indirect evidence indicates testosterone probably confers an advantage. But it was not convinced that testosterone explained the entire performance gap between men and women, a difference in the range of 10 to 12 per cent. If testosterone confers an advantage in the realm of 1 to 3 per cent, it is no different from other variables that conspire to produce a superior athlete: height, muscle mass, oxygen uptake.

The ruling suspended the IAAF’s regulations, citing lack of scientific evidence. But it raised questions that venture well beyond science: what is fairness, really, in the world of elite sport?

Those who decried the court’s ruling cited a desire to maintain a level playing field for women. Female athletes themselves want that: the British world-record-holding marathoner, Paula Radcliffe, testified in support of the IAAF, and Maria Jose Martinez-Patino, the banned hurdler, co-authored an article that concluded some rule must exist.

But why are women the only ones who need the playing field levelled? Is it fair to compete against LeBron James, Michael Phelps or Usain Bolt? Sports celebrates outliers on the spectrum of genetic diversity. No one is hunting for male athletes who produce vastly more testosterone than their peers, or enjoy any other freakish advantage.

We don’t know the names of the four young female athletes who travelled to France, or anything about their athletic or personal lives, only that they were from developing countries. But because of the court’s decision — affirmed by the IOC — their irreversible gonadectomies would no longer be necessary to compete. And of course their partial clitoridectomies, a cosmetic procedure, were never necessary.

Nobody knows what will happen next. The IAAF has two years from last July’s ruling to produce evidence supporting the hyperandrogenism regulations. Even some supporters don’t think they will: the science of sex and athletic advantage is just too elusive. Should declaring a gender identity be enough? Should we do away with divisions entirely?

Kidd points out we don’t actually know how much of the 10- to 12-per-cent male advantage is the result of social context rather than biology.

“Will women run as fast as men in some events if genuine equality is achieved — if the genetic pool out of which female Olympians are drawn is the same, and there is the same motivational encouragement, the same financial encouragement?” he asks.

“We’re living in an experiment.”