Dodgy Science, Sport and Sex

Self-described Defenders of Women's Sport Misrepresent Research on Testosterone to Advance their Political Agenda

The challenge of how to determine who is eligible to compete in women’s competitive or elite athletic competition is an issue that ignites passions. It is also an issue on which reasonable people can disagree. However, one thing that everyone should agree on is that in making arguments about this issue, scientific research should be accurately reported, and certainly not misrepresented.

Today, I was surprised to see on Twitter that Nancy Hogshead-Makar, both an Olympic champion and also a champion in the fight against the sexual abuse of athletes, was badly misrepresenting science and policy associated with testosterone and female athletes. When I dug deeper, things only got worse.

Here is one of her Tweets. In it Hogshead-Makar asserts, falsely, that “An XX person w testosterone in the male range is disqualified from the women’s category. She is doping.”

Nancy Hogshead-Makar, JD, Oly @Hogshead3Au

@leakylike @TheDrakeGroup @WrstleLikeAGirl @iChampionWomen @FemMajority @MsMagazine @Martina @WeCOACH @WomenLeadersCS @edwinmoses @diananyad It is testosterone that creates the sex-linked advantages that are important for sport. 👉🏽An XX person w testosterone in the male range is disqualified from the women’s category. She is doping. See science on testosterone advantage: web.law.duke.edu/sites/default/…

This claim is simply false. No anti-doping agency evaluates any woman (“an XX person”) for testosterone levels “in the male range.” Anti-doping agencies certainly do check to see if both men or women show evidence of using testosterone not produced by their bodies (called “exogenous”). If such a substance is detected, then it is an anti-doping violation, and potentially sanctionable. The so-called “male range” of testosterone is just not a thing in anti-doping.

It is bad enough that a leading advocate for women’s sport is promoting false information, but when I visited the website of her advocacy organization related to trans athletes, I found much more of concern.

The “Women’s Sports Policy Working Group,” of which Hogshead-Makar is a member, includes some notable figures such as Martina Navratilova, Greg Louganis and Diana Nyad. The group seeks to motivate Congressional legislation that would effectively establish a separate but equal category of competition for trans athletes, apparently based on testosterone thresholds.

Issues related to sex and gender are complicated, and won’t be addressed here. What I will document unequivocally here is that a core claim advanced by the Working Group related to biological sex is based on misrepresenting scientific research. And that is never a good way to engage a policy debate. In fact, it should be disqualifying in this case.

The core claim has to do with whether an individual’s testosterone levels can be used to determine if he/she is a male or a female. According to the Working Group, the answer is yes: in all cases testosterone is a perfect proxy for biological sex. This is simply incorrect. The reality is that scientific research — the very research that the Working Group cites — says otherwise.

In a FAQ provided by the Working Group, they ask and answer: “Don't some healthy females produce testosterone in the "male" range? No.” They cite this definitive answer to a paper published in 2020 in the Journal of the American Medical Association.

But in fact, that paper’s finding are completely to the contrary (emphasis added):

After cessation of the age-related increase in testosterone for female youths (at 14 years), there was an intersection of testosterone concentration distributions between the lowest (first) percentile of male youths and the uppermost (99th) percentile of female youths (≥100 ng/dL), which includes 8 of 949 samples (<1%) for female youths.

Less than 1% may be small, but it is not zero. This is not horseshoes and hand grenades. Even statistical outliers are real people, deserving of real respect. The Working Group should have known this because one of its members, Doriane Coleman of Duke University, happens to be a co-author of the JAMA paper being mis represented.

But it get worse.

The Working Group also cites a 2018 paper in the journal Clinical Endocrinology to assert that there is no overlap between female and male testosterone ranges. Readers who click on the link to that paper will notice that there is a prominently-featured link at the top of the article to “corrections to this article.”

I know this because I’m the person who pointed out to the authors and to the journal fundamental errors in the 2018 paper. Identification of errors resulted in the subsequent publication of a long Erratum, characterized as a “massive correction.” In its FAQ, the Working Group cited the Clinical Endocrinology paper but failed to acknowledge the “massive correction.” Not good.

The correction acknowledged that there was in fact an overlap between male and female testosterone ranges. More generally, as Olympian and science studies expert Madeline Pape and I explained, all of the work being done in defining so-called male and female testosterone ranges actually occurs in how people are classified before measurements are actually taken. It turns out to be a perfectly circular argument.

I am sure that there are many areas of policy related to the delineation of competition categories that the Women’s Sports Policy Working Group and I would agree on, and perhaps we would find more areas of agreement than disagreement.

But one area where we will never agree is that it is acceptable to misrepresent evidence in pursuit of political aims.