Not by gender, not by sex, but by testosterone saith the IAAF: International athletics and the new Female Eligibility Regulations

Published 22 May 2018 By: Dr Ben Koh, Daryl Adair, Liam Elphick

Athletics runner on track

On 23 April 2018, the International Association of Athletics Federations (IAAF) released its “Eligibility Regulations for the female classification: Athletes with differences of sex development“ (DSD Rule) which comes into effect on 1 November 2018.1 The stated purpose of the DSD Rule is for athletics to be inclusive and to respect the dignity of individuals with differences of sex development (DSD),2 through, paradoxically, the creation of discriminating criteria that affects only female athletes.

Although the IAAF claims that the DSD Rule is in no way intended to be a judgement on the sex or gender identity of any athlete, but is instead intended to preserve the dignity of athletes with DSDs, the new IAAF Rule nevertheless requires that otherwise healthy female athletes with DSD must reduce their naturally elevated testosterone levels by medical means. This intervention specifically excludes surgical anatomical changes, but chemotherapy, hormonal castration, and iatrogenic irradiation are all presumably acceptable means for reducing natural testosterone production in women subject to the DSD Rule.

This article explores the history leading to the DSD Rule, the apparent scientific basis for the criteria, as well as the justification and proportionality of the DSD Rule. Based on this review, the authors question what they view as the continued discrimination towards females in its many guises, and indeed specifically of intersex athletes, in an age when sex is understood to exist along a spectrum, and when gender identity is fluid.


The IAAF has long been engaged with processes to establish the "authenticity" of female competitors. As early as 1946, the IAAF required medical certificates from women in order to validate their sex status. The International Olympic Committee (IOC) followed the IAAF’s lead two years later. By 1966, international athletics featured a combination of "nude parades" or even gynecological examinations to establish, to the satisfaction of sport authorities, that the competitors were indeed women, rather than men masquerading as women.3

By 1968, sex testing via chromatin analysis was introduced at the Olympics with the support of the IAAF, along with drug testing of both male and female athletes. At this point the primary interest was with the impact of exogenous androgens (i.e. doping) on women: there was anecdotal evidence that Eastern Bloc female athletes were being virilised by the imposition of Anabolic Androgenic Steroids (AAS). Sport authorities were, by this time, more anxious than ever that the women’s category of competition needed protection, whether from faux females, people who were not "quite" women, or women who, in an effort to enhance athletic performance, were prepared to virilise using AAS.4

The chromatin test was abandoned in 1988 after it dawned on the IAAF that there are variations beyond the typical chromosomes of XX female and XY male. A Spanish female athlete, Maria Jose Martinez-Patino, who had been provided with the mandated "Certificate of Femininity" was later found, on the basis of a subsequent chromatin test, to have XY chromosomes. Ms Patino was androgen insensitive such that her endogenous testosterone levels have no impact on her body and everything about her life, gender and sex organs was typical of a woman. Despite this, her chromosomes did not conform to the XX female mould.

Ms Patino’s case led to a growing realisation that biological sex is on a spectrum, with a very small proportion of people (in this case women) who do not fit conventional expectations of neat binary categories of female and male. As such, the IOC and the IAAF abandoned sex testing just before the 2000 Olympics. The surveillance and intrusion on women seemed disproportionate to concerns about the "authenticity" of female athletes.5

By 2006, the IAAF – while assuring women that there would be no compulsory or regular sex testing during track and field events – announced that if a female athlete had "suspicion" about the "authenticity" of a fellow competitor they could institute a "challenge", this allowing the IAAF’s medical panel to evaluate the individual. This process could involve gynecologists, endocrinologists, internal medicine specialists, psychologists and others.6 That policy was applied three years later in what was arguably the most controversial moment in the history of women’s sport. Caster Semenya, a teenage athlete from South Africa, had performed beyond expectations at the IAAF World Championships, with her time prompting concerns of doping. After repeat excellent performances, it became obvious that Semenya was an authentic champion – there was no hint of doping. In the meantime, though, commentary by fellow competitors, combined with salacious media reporting, produced a discourse that Semenya was not an "authentic" woman. Put crudely, the charge was that she was virilised compared to a typical female athlete, this giving her an advantage over track competitors. It may be a natural part of who Semenya is, but critics insisted that this was unfair on other women athletes.7

During this cacophony of speculation and accusations, Semenya remained outwardly calm, even if inwardly distressed. The IAAF, meanwhile, was obliged to investigate the "suspicion" based on a "challenge" by one or more athletes. By mid 2010, the IAAF’s "panel of experts" had concluded that Semenya was free to compete in athletics with immediate effect.8 The case seemed closed. However, ten months later it became apparent that the IAAF was re-evaluating its position. A new "Hyperandrogenism" (H) Policy was introduced in 2011 with the focus of "authenticity" for the female category based on the levels of endogenous (naturally occurring) Testosterone (T).9

In the general population, women of reproductive age typically have T levels within a range that is less than 1.0 nmol/L,10 whilst healthy men have a higher level of around 10-30 times that of women.11 The inference appeared simple enough: a fundamental physiological difference between women and men, as well as a prime driver of athletic performance, is endogenous T. Women who participate in IAAF competition should, as a consequence, have a level of natural T that is commensurate with the broader female population. The assumption is that high levels of endogenous T provide female athletes with an "unfair" (albeit natural) advantage over competitors in what the IAAF deems the "normal" range.12

The 2011 H Policy, to the delight of some female athletes, was set to impact Caster Semenya who, it was widely understood, had high endogenous T (reportedly as a consequence of a congenital condition). Like other athletes deemed hyperandrogenic, Semenya had two choices: medical (including surgery) intervention or drop out of IAAF events. Semenya chose to continue, though it is unknown what intervention she adopted. Competing at the London 2012 Olympics she won silver, though later this was upgraded to gold after the Russian winner was found - ironically enough - to have doped. Semenya was still racing well, though observers noted that her times were a little slower than her best.13

Perhaps surprisingly, the IAAF’s H Policy was not officially challenged by Semenya. Instead, little known Indian sprinter, Dutee Chand, was the catalyst for an appeal to the Court of Arbitration for Sports (CAS).14 Chand had been disqualified from competing by the Athletics Federation of India, which ruled that she contravened the H Policy. A demure and non-striking physical figure, Chand was contesting a ruling that – as far as most commentators were concerned – was intended to "reign in" female athletes who gave the "appearance" of virilisation and an unfair (albeit natural) advantage.15

In 2015 CAS determined that the H-Policy was discriminatory and that the IAAF had not provided adequate evidence to support or justify such discrimination. In particular, the CAS held that the IAAF had not established that the the “degree of advantage that androgen-sensitive hyperandrogenic females enjoy over non-hyperandrogenic females16 was sufficient to justify the H-Policy. In other words , the IAAF had not established that the H-Policy was appropriate or necessary to ensure fair competition within the female category.17 CAS also noted that the difference in competition timing between the best male and female athletes in common athletic events was around 10-12%, and that the timings achieved by H women were nowhere near that of men. CAS therefore ruled the IAAF H Policy was invalid for a period of two years.18 The IAAF could present further evidence to justify the H-Policy within that two-year period, but if it failed to do so the H-Policy would be declared permanently invalid.19 Dutee Chand was therefore able to compete at the Rio Olympics, where she enjoyed the thrill of participation even though she did not get past the early heats. In the meantime, Caster Semenya enjoyed immense success, winning the 800m comfortably with the 5th fastest time among women in Olympic history.20

Sport policy and body politics

The IAAF did not come back to CAS within the two-year time frame, but the tribunal generously offered a six months extension. Curiously, though, the IAAF chose not to challenge Dutee Chand’s position: she and other high T female sprinters thus remain free to compete in the 100m and 200m without restriction. Instead, the IAAF identified a limited number of events for which it claims there is demonstrable proof of "unfair" performance advantage derived by women with naturally high testosterone levels. According to the IAAF, its scientific researchers were able to establish substantive differences in key events: the 400m (including hurdles), the 800m, 1500m, one mile, pole vault and hammer throw.21

Surprisingly, though, the IAAF deemed only track events to be of interest: it now mandated, through the April 2018 DSD Rule, that the 400m, (including hurdles), the 800m, 1500m and one mile are "restricted" events. Whether by coincidence or design, these are distances in Caster Semenya’s performance range. In these "restricted" events, female competitors with DSD will under the new rule need to demonstrate that their endogenous T is 5.0 nmol/L or lower22 at all times.23 The other requirement is that a competitor in the female category must be deemed, under law, to be either female or intersex.24 In essence, therefore, the IAAF has reconfigured the categories of male and female in athletics to be about levels of endogenous T, but only in the female category.25

The litmus test to be accepted as a female participant in 2018 is a maximum level of endogenous T.26 Athletes who are identified or present with high T will in real terms have three choices: (1) undergo treatments to lower their endogenous T and thus be eligible to compete at international levels of athletics in the female category; (2) refuse to alter their endogenous T levels and be eligible to compete only at national (or lower) levels of athletics; or (3) request participation in men’s athletic events.27

Science-based policy and "fair play" in scholarship

The 2018 DSD Rule claims to be based on broad medical and scientific consensus that high levels of endogenous testosterone circulating in athletes can significantly enhance their sport performance.28 However, the citations provided in support of this purported scientific consensus excluded independent empirical research suggesting the contrary,29 used unpublished research where its methodology and validity cannot be determined,30 and included a review paper where the authors specifically stated that the effects of elevated endogenous androgen concentrations on athletic performance among female athletes remains controversial and unresolved.31

The DSD Rule relies ultimately on a single published study (2017 Study)32 authored by Dr. Pierre-Yves Garnier and Dr. Stéphane Bermon, both affiliated with the IAAF. It was funded by the IAAF and was part of the evidence the IAAF prepared for its follow-up defence33 to its initial defeat of the 2015 Dutee Chand case at CAS.34 The stated purpose of the 2017 Study was to explore the possible relationship between serum androgen concentration, particularly serum testosterone concentration, and athletic performance. To do this, the authors performed a cross-sectional (i.e. a snapshot in time) retrospective statistical analysis of blood samples taken from female athletes at the 2011 and 2013 IAAF World Championships and compared for each athlete the best performance (heats or finals) achieved during the corresponding World Championships.35

In total, there were 1332 samples from female athletes and of these 17.3% of female athletes were counted twice - in 2011 and 2013. The 2017 Study found that twenty-four female athletes had a high testosterone level (>3.08 nmol/L) of which nine had DSD and thus presumably naturally occurring endogenous testosterone levels; nine were found to have been doping and presumably used exogenous androgens; and six were found to be impossible to classify.36 The failure to differentiate exogenous and endogenous androgens makes any conclusion problematic.37 No attempt was made by the authors to compare the endogenous DSD group with the exogenous doping group.

To test for the association between the serum androgen concentration and the type of athletic event, all athletes were firstly clustered into tertile38 groups based on their free testosterone (fT) level. No explanation was given as to why the groups were divided into three instead of two groups (a low-fT and high-fT), or otherwise simply leaving them as a continuous variable.39 This lack of explanation in its methodology may easily lead some to question the validity and practical value of any statistical effect recorded.40

Next, the samples were then assigned to one of the seven following groups:

It is unclear how an athlete (if any) who has participated in more than one of these events was classified, and whether blood samples were taken from only one, from some, or from all the events that athlete participated in. For example, if a female athlete who only participated in the 2011 championship and competed in the 400m and 800m events had only one blood sample taken, would that data be classified into the sprinting group or middle distance group or both (double counting of the data)?

Finally, the serum androgen concentration tertiles were statistically compared with the type of athletic event groups. From this analysis the authors found that when the athletes in the highest-fT and lowest-fT tertiles were compared, a statistically significant but small improvement was seen only in the following:41

The authors did not report if the highest-fT tertile or lowest-fT tertile were any different from the middle-fT tertile (i.e. the middle group that was neither high nor low in fT).

Given that the study was cross-sectional in design, it cannot attribute causation, nor correlation. Additionally, since there was at least a double-counting in 17.3% of cases, this risks overstating any evidence,42 despite the authors’ assertion that it is a non-issue. It is also important to note that because the granular data were not reported for individual athletes who had both 2011 and 2013 samples, it is unclear if the levels in those athletes had changed in that time, and whether those changes (if they occurred) correlated with a change in performance of that athlete.

Even if we put aside the numerous methodological inadequacies43 and assume a true effect due to the naturally elevated testosterone that exists in DSD females, there was only a statistically modest 1.78%-4.53% increase in performance, and in just a handful of events. Certainly, the improvement was not in the order of 10-12% between males and females - a level CAS has determined to be the degree required for an unfair advantage and a legitimate separation for competition between males and females.44

The level of methodological flaws evident in the 2017 study and the DSD Rule’s reliance on it makes one question if confirmatory bias was at play and whether the conclusions drawn were potentially a strained attempt to force a square data peg into a round policy hole. The assertion of subjective scientific conclusions unrelated to objective empirical data provides a challengeable basis, either scientifically or legally, on which the DSD Rule now sits. Furthermore, robust scientific evidence of causation such as that required to satisfy the CAS Panel in the 2015 decision usually involves multiple and independent studies conducted over many years. Only three years ago, CAS formed the view that there was insufficient scientific evidence to show that female athletes with elevated T levels benefited from such a performance advantage to require that they be excluded from female competition.

According to skeptics, the onus is now on the IAAF to prove that any DSD Rule is a robust and proportionate justification for requiring high T women to medicate in order to compete in female events. No new emerging scientific evidence over the last three years, nor in the 2017 study, appear to have added anything to alter the 2015 CAS decision.

Many social science critics and human rights activists assert that the IAAF scientific research is, rather, a smokescreen for a prime objective: to curtail the athletic performances of Caster Semenya.45,46 Indeed, there has been vocal discontent among (mainly non-African) athletes who have lost against Semenya in competition, some decrying that it is "unfair" they be forced to compete against someone they believe has an unreasonable, even if natural, advantage by way of physical performance.47 If the science underpinning the DSD Rule is unconvincing, the IAAF risks being perceived as unfairly discriminatory, yet again.48

Legal, social and health-related concerns

Considering the significant concerns surrounding the research underpinning the IAAF’s DSD Rule, it is likely that a legal challenge will come. There are various legal arenas in which the DSD Rule could be challenged: namely the IAAF’s own tribunal, a national court or, most likely, CAS.49 While this breadth of potential jurisdictions makes it difficult to pinpoint the arguments that might be raised in any legal challenge, several key points can be made at this early stage by placing potential legal arguments in their social and health-related contexts.

First, the right to freedom of non-discrimination is a well-recognised and universal human right. Article 14 of the European Convention on Human Rights and Article 26 of the International Covenant on Civil and Political Rights both provide that all persons should be able to exercise their rights free of any discrimination on a number of grounds, including sex, religion, race, and "other status". The European Court of Human Rights has held that ‘other status’ protection extends to non-discrimination on the basis of gender identity,50 but no European case has yet considered whether this protection extends to intersex status.51 As the European Court of Human Rights has found that "other statuses" protected include any "differences based on an identifiable, objective, or personal characteristic, or “status”, by which individuals or groups are distinguishable from each other",52 it is likely that intersex status is protected.

The IAAF Constitution and the Olympic Charter also prohibit discrimination. The IAAF Constitution reads at article 4.4: “The Objects of the IAAF are […] To strive to ensure that no gender, race, religious, political or other kind of unfair discrimination exists, continues to exist, or is allowed to develop in Athletics in any form, and that all may participate in Athletics regardless of their gender, race, religious or political views or any other irrelevant factor.53 Similarly, the Olympic Charter, to which the IAAF is also bound,54 prohibits discrimination. Principle 6 of the Olympic Charter provides that "The enjoyment of the rights and freedoms set forth in this Olympic Charter shall be secured without discrimination of any kind, such as race, colour, sex, sexual orientation, language, religion, political or other opinion, national or social origin, property, birth or other status."55

The right to non-discrimination has also been enshrined in domestic law in numerous countries through anti-discrimination legislation. In Australia, for example, intersex status is a protected ground under federal anti-discrimination legislation.56 This means that treating an intersex athlete less favourably by reason of their intersex status or associated characteristics,57 for example by excluding them from female competition unless they alter their natural physiology, would likely be unlawful discrimination.58

There is also a human rights issue in regards to sex discrimination. The DSD Rule sets an upper testosterone limit for female international events, but no limits or bounds are set for male international events. This may raise the question of whether the IAAF is also discriminating against women on the ground of sex, considering their limitations are stricter than for men. If testosterone is to be the decisive factor for determining sex in international athletics events, it may even be discriminatory against men to not allow those men with low testosterone to compete in women’s events.59

Secondly, if the IAAF is indeed discriminating against intersex and/or women athletes, any court or tribunal determining the substantive legality of the DSD Rule would likely need to consider whether this policy is necessary and proportionate in pursuing legitimate aims - as is the standard legal approach when human rights are undermined or clash.60 As already noted, this type of proportionality approach underpinned the 2015 Dutee Chand decision by CAS. The IAAF would likely argue that the aim of the DSD Rule is to ensure fairness for competing female athletes at international athletics events, such that any "unfair advantage" is removed. While this is a legitimate aim, the means used to achieve this end may not be convincing nor proportionate. The IAAF would be questioned as to why 5.0 nmol/L is the specific benchmark set, why the DSD Rule applies only to track events from 400m to one mile and not in hammer throw or pole vault where testosterone has an even greater performance-related advantage,61 and indeed as to whether high endogenous testosterone is scientifically proven to create a substantial athletic advantage in female events. The significant individual harm62 the DSD Rule causes to athletes such as Caster Semenya renders the proportionality requirement even harder to satisfy.

If the DSD Rule is, though, proven proportionate, the broader question of whether the rule is necessary in pursuing a legitimate aim will be the next difficult hurdle to jump over. In considering the legitimate aim of removing "unfair advantage" from competitive athletics, the concept of "advantage" would be distinguished from "unfair advantage". Taller athletes tend to have an advantage over shorter athletes, yet few people would argue this is "unfair" - perhaps because height differences have been an ever-present sporting fixture, while intersex and transgender athletes competing openly has been a relatively recent phenomenon. In contrast, weight categories are common in sports such as boxing in order to ensure a "fair" fight and the safety of participants. To prove the DSD Rule is necessary to achieve the aim of "fairness", therefore, the IAAF would need to establish why higher naturally-occurring testosterone levels, which are as physiologically intrinsic to a person as most other athletic features, are "unfair". As one of the key features of a successful sporting career is using physiological attributes, whether acquired genetically or through rigorous training, to one’s advantage, such an argument would be nuanced. As Buzuvis notes, "[s]aying that no one can use natural advantage is antithetical to sport."63 The IAAF’s argument may well be, then, that natural advantages apply only within each sex category, and that high testosterone levels unfairly cross this boundary for the female sex category. However, drawing again on norms of non-discrimination on the ground of sex, if naturally-occuring physiology is used as a distinguishing feature for women such that they become "unfairly" advantaged, why would naturally-occurring physiology64 not be used as a distinguishing feature for men? Such questions may well turn on scientific data that, as noted above, simply are not yet available to us.

Finally, while international athletics remains segregated into only two categories characterised by sex for most events,65 excluding female-identifying athletes from women events is unjust and could lead to further liability under torts law for health-related concerns. While the DSD Rule does not explicitly exclude intersex athletes from being able to compete in "restricted events", health risks associated with testosterone reduction may well have the practical effect of excluding many intersex athletes from competing at international women events. This is particularly so in the absence of a separate category under which they could compete, with the IAAF flagging a potential new "intersex" category in the DSD Rule.66 While athletes can undergo treatment to lower their testosterone levels to below 5.0 nmol/L, this crude and dogmatic approach to a complex medical issue can cause significant issues to individual athletes and their health.67 Indeed, many researchers have discovered, and expressed significant concern at, adverse health effects caused by treatment where testosterone levels are reduced below those naturally produced and required by an individual’s body.68 Forcing athletes to do so in order to compete in female international events could therefore open up the IAAF to negligence suits by such affected athletes, were their health to deteriorate or suffer as a result of testosterone-reducing treatment.69

Concluding remarks

While any female athletes with testosterone over 5.0 nmol/L could seek treatment to reduce this level in order to compete under the IAAF’s DSD Rule, the health consequences of doing so may exclude many from being able to undertake this course of action. Exclusion from competition is arguably the most severe punishment that any athlete could face in a sporting context; such an intervention must only occur for the strongest of justifications and in the most extreme of cases. Moroever, in an era where we repeatedly combat doping in sport and seek to exclude those who enhance their performance by pharmacological means it is highly ironic that some athletes could be required to pharmacologically alter their body in order to be permitted to compete.

Far stronger reasoning than that presented by the IAAF and the research it has relied upon is required before we allow the IAAF to bar a group of athletes from being able to compete in their chosen events. In the absence of further justification by the IAAF as to why the DSD Rule applies in such a targeted way, and in a developing world where social norms now reflect a more fluid definition of sex and gender, the fall-back position of international athletics must be non-discrimination. While the Olympic motto reads "citius altius fortius", meaning "faster, higher, stronger", it is another Latin maxim that must prevail when assessing intersex eligibility in international athletics: primum non nocere, "first, do no harm".


1† IAAF, Eligibility Regulations for the female classification: Athletes with differences of sex development (23 April 2018), available to download here: (last accessed 22 May 2018)

2† Ibid [1.1(a)], [1.1(c)].

3† Wackwitz, Laura A. 1996. “Sex Testing in International Women’s Athletics a History of Silence.” Women in Sport and Physical Activity Journal 5 (1). Human Kinetics: 51–68. doi:10.1123/wspaj.5.1.51; Wiederkehr, S. (2009). ‘We shall never know the exact number of men who have competed in the Olympics posing as women’: Sport, gender verification and the Cold War. The International Journal of the History of Sport, 26(4), 556-572.

4† This altering the traditional expectation of what a woman ‘is’, either in sport or society: Olsen-Acre, H. K. (2006). The use of drug testing to police sex and gender in the Olympic Games. Mich. J. Gender & L., 13, 207-236.

5† Ritchie, I. (2003). Sex tested, gender verified: Controlling female sexuality in the age of containment. Sport History Review, 34(1), 80-98; Ljungqvist, A., Martínez-Patiño, M. J., Martínez-Vidal, A., Zagalaz, L., Díaz, P., & Mateos, C. (2006). The history and current policies on gender testing in elite athletes. International SportMed Journal, 7(3), 225-230.

6† Berry, E. E. (2012). Respect for the Fundamental Notion of Fairness of Competition: The IAAF, Hyperandrogenism, and Women Athletes. Wis. JL Gender, & Soc'y, 27, 207; Davis, P., & Edwards, L. (2014). The new IOC and IAAF policies on female eligibility: old Emperor, new clothes?. Sport, ethics and philosophy, 8(1), 44-56.

7† Nyong'o, T. (2010). The unforgivable transgression of being Caster Semenya. Women & Performance: a journal of feminist theory, 20(1), 95-100; Camporesi, S., & Maugeri, P. (2010). Caster Semenya: sport, categories and the creative role of ethics. Journal of Medical Ethics, 36(6), 378-379.

8† Schultz, J. (2011). Caster Semenya and the “question of too”: Sex testing in elite women's sport and the issue of advantage. Quest, 63(2), 228-243.

9† Crincoli, Shawn. (2011). The IAAF Hyperandrogenism Regulations and Discrimination, World Sports Law Report, June, 9(6), 3-5; Schultz, J. (2012). New standards, same refrain: the IAAF's regulations on hyperandrogenism. The American journal of bioethics, 12(7), 32-33.

10† Although an upper reference level of 5.0 nmol/l is sometimes accepted in females to not indicate any pathological conditions: Jane, K., et al. (2007) Measurement of serum testosterone in women; what should we do? Annals of Clinical Biochemistry, 44(1), 5-15. doi:10.1258/000456307779595896

11† Travison, T. G., et al. (2017). Harmonized Reference Ranges for Circulating Testosterone Levels in Men of Four Cohort Studies in the United States and Europe. The Journal of Clinical Endocrinology & Metabolism, 102(4), 1161-1173. doi:10.1210/jc.2016-2935. Note that although the cohorts included in these analyses were diverse in morbidity, age, and geographic location, they were largely comprised of men who identified as ‘white’ within a US or European social context. Elsewhere, significant geographic and racial differences in sex-steroid levels have been reported by researchers, which might have important implications for clinical evaluation and, by extension, sports policy. There is a further caveat: the general population values may not be applicable in the athlete cohort, where substantial variations within sex categories and occasional overlap of the testosterone range was reported among male and female participants: Healy, M. L., Gibney, J., Pentecost, C., Wheeler, M. J., & Sonksen, P. H. (2014). Endocrine profiles in 693 elite athletes in the postcompetition setting. Clinical Endocrinology, 81(2), 294-305. doi:10.1111/cen.12445. For further discussion see: ‘Testosterone, sex and gender differentiation in sport – where science and sports law meet’,, 14 Oct 2014, last accessed 22 May 2018,

12† Bermon, S., & Garnier, P. Y. (2017). Serum androgen levels and their relation to performance in track and field: mass spectrometry results from 2127 observations in male and female elite athletes. Br J Sports Med, bjsports-2017.

13† Cooky, C., & Dworkin, S. L. (2013). Policing the boundaries of sex: A critical examination of gender verification and the Caster Semenya controversy. Journal of sex research, 50(2), 103-111; Davis, P., & Edwards, L. (2014). The new IOC and IAAF policies on female eligibility: old Emperor, new clothes?. Sport, ethics and philosophy, 8(1), 44-56.

14† Dutee Chand v Athletics Federation of India (AFI) & The International Association of Athletics Federations (IAAF) CAS 2014/A/3759

15† Franklin, Simon, Jonathan Ospina Betancurt, and Silvia Camporesi. 2018. “What Statistical Data of Observational Performance Can Tell Us and What They Cannot: the Case of Dutee Chand v. AFI & IAAF.” British Journal of Sports Medicine 52 (7): 420–21. doi:10.1136/bjsports-2017-098513.

16† Camporesi, Silvia. 2016. “Ethics of Regulating Competition for Women with Hyperandrogenism.” Clinics in Sports Medicine 35 (2). Elsevier: 293–301. doi:10.1016/j.csm.2015.10.003.

17† Patel, S. (2015). The IAAF’s hyperandrogenism regulations suspended. World Sports Law Report, 13(8), 8-11.

18† Dutee Chand v Athletics Federation of India (AFI) & The International Association of Athletics Federations (IAAF) CAS 2014/A/3759 at [548].

19† Ibid.

20† A recent study raises further doubt on the matter by showing that the percentage difference in performance between women with and women without hyperandrogenism did not even reach a 3% difference, meaning it also does not reach the widely accepted 10% range of difference in performance between men and women: Ospina Betancurt, Jonathan, Zakynthinaki, Maria S., Martinez-Patiño, Maria Jose, & Cordente Martinez, Carlos. (2018). Hyperandrogenic athletes: performance differences in elite-standard 200m and 800m finals. J Sports Sci, 1-8. doi: 10.1080/02640414.2018.1464620

21† 2018. “IAAF Introduces New Eligibility Regulations for Female Classification| News |” Accessed May 14.

22† This representing a 50% reduction on the previous 2011 High T policy.

23† Thus, the rule applies both in and out of competition. It is unclear if the choice of 5 nmol/l as a cut-off level was based on an absolute upper level found in the reported studies or, for example, three or more statistical standard deviations from the mean levels in reported studies (to theoretically capture more than 99.99% of the population). In terms of the DSD Rule, the IAAF noted that, from a paper yet to be published, women (including female athletes) with DSDs can have serum levels of testosterone above 5 nmol/L and well into (or even above) the normal male range: Handelsman, Hirschberg and Bermon (2018), Circulating Testosterone as the Hormonal Basis of Sex Differences in Athletic Performance, Endocrine Reviews (publication pending). For completeness, it should be noted that in the general population, women with hyperandrogenic symptoms presenting for medical review may have testosterone levels up to a maximum of 4.62 nmol/l: Escobar-Morreale, H. c. F., Sanchón, R., & San Millán, J. L. (2008). A Prospective Study of the Prevalence of Nonclassical Congenital Adrenal Hyperplasia among Women Presenting with Hyperandrogenic Symptoms and Signs. The Journal of Clinical Endocrinology & Metabolism, 93(2), 527-533. doi:10.1210/jc.2007-2053.

24† From an IAAF perspective, the significance of an intersex category is twofold: (1) some people prefer that status; (2) it is common for hyperandrogenic women to have an intersex status (even though it may be a private matter).

25† Stéphane Bermon, speaking recently on radio, confirmed that binary sex categories still exist in IAAF events, but that conditions of participation for athletes in the women’s category are now determined by level of endogenous T. He insisted that the IAAF had “no right” to make determinations about the sex or gender of athletes, but if they chose to participate in the female category of athletics then the defining eligibility factor is endogenous T. Bermon argued that this allowed intersex people who may have XX or XY chromosomes, with some presenting traits normally associated with men, such as testes, to take part in the female category of athletic competition. According to Bermon, the IAAF has been tracking the performances of intersex athletes for many years: ABC Radio, The Ticket, 29 April 2018 (Podcast available via iTunes). It is unclear how the IAAF would know whether an athlete presented with an intersex status, unless their participation fell under the previous H policy, which was in operation from 2011-2016. It is also not known publicly how the IAAF has ‘identified’ athletes with high T: the answer may lie in anti-doping. All competitors are subject to the World Anti-Doping Agency (WADA) testing for doping and this included the Athlete Biological Passport (ABP) process of serial blood testing. Is there sharing of endogenous hormone profiles between WADA (or its agencies) and the IAAF? Have chromatin tests been covertly conducted on the ABP and made known to IAAF? Have confidential medical records of athletes been cross-referenced with the ABP by WADA, the various National Anti-doping Agencies (NADAs) or the IAAF? [see Koh, B. (2013). ‘Anti-doping and medical privacy’,,]. Whatever the case might be, the process for ‘identifying DSD’ athletes under the IAAF’s policy ought to be spelt out in the interests of transparency and fairness. Whilst the identification process should be transparent, the biological information should remain confidential to protect an athlete’s identity and status.

26† This 2018 DSD Rule policy is similar to the previous 2011 H policy (that was disapproved by CAS) but at a maximum level of endogenous T that is half of what the IAAF insisted upon previously.

27† There is a fourth possibility in future: athletes deemed to be subject to the DSD Rule would be free to partake in intersex athletic competition, should such a category one day emerge.

28† IAAF, Eligibility Regulations for the female classification: Athletes with differences of sex development (23 April 2018) [1.1(d)].

29† Ibid A-9

30† Handelsman, Hirschberg and Bermon, 'Circulating Testosterone as the Hormonal Basis of Sex Differences in Athletic Performance' (2018).

31† Grace Huang and Shehzad Basaria, 'Do anabolic-androgenic steroids have performance-enhancing effects in female athletes?' (2018) 464 Molecular and Cellular Endocrinology 56.

32† Stéphane Bermon and Pierre-Yves Garnier, 'Serum androgen levels and their relation to performance in track and field: mass spectrometry results from 2127 observations in male and female elite athletes' (2017) 51(17) British Journal of Sports Medicine 1309.

33† IAAF, Press Release: Levelling the playing field in female sport - new research published in the British Journal of Sports Medicine (3 July 2017). <>.

34† IAAF, Press Release: IAAF comments on interim award issued by the CAS regarding the IAAF Hyperandrogenic Rules (27 JUL 2015) <>; Dutee Chand v Athletics Federation of India (AFI) & The International Association of Athletics Federations (IAAF) CAS 2014/A/3759.

35† The study also included male athletes participating in the 2013 championship.

36† This may suggest that some athletes who were females and did not have DSD and were not doping had naturally high T.

37† It is beyond the scope of this paper to discuss the scientific evidence of endogenous versus exogenous testosterone levels on the physical body of an individual, and the further problems of hypothesising any resulted physical changes translate into actual sporting performance. It should be noted, however, that in Dutee Chand the CAS held that the relationship between testosterone (exogenous and endogenous) and female sporting performance is unclear (Dutee Chand v Athletics Federation of India (AFI) & The International Association of Athletics Federations (IAAF) CAS 2014/A/3759 at [155], [158], [162], [171], [187], [199-200]) especially for endogenous testosterone above 10.0 nmol/L (at [521-522]), but nevertheless that the athlete had failed to establish on the balance of probabilities that exogenous and endogenous testosterone have different effects on the body (at [528]).

38† A statistical method whereby an ordered distribution from lowest to highest fT levels are divided into three equal parts to give a low, middle and high-fT group.

39† Which makes a statistical analysis more robust.

40† A statistical “significance” does not necessarily equate to any real world “true” effect especially when a continuous variable is artificially converted into an ordinal variable without proper scientific justification.

41† Incidentally, the study reported that in male elite athletes, no significant difference in performance was noted when comparing the lowest and the highest fT tertiles. If true, then the corollary question is whether medically- recommended T treatments should be reconsidered for male athletes with low levels?

42† Stephen J. Senn, 'Overstating the evidence – double counting in meta-analysis and related problems' (Pt BioMed Central) (2009) 9 BMC Medical Research Methodology 10.

43† Many of which were not explored in this discussion.

44† Dutee Chand v Athletics Federation of India (AFI) & The International Association of Athletics Federations (IAAF) CAS 2014/A/3759, at [522].

45† Sheridan, D. (2018, 11 April). Critics of Caster Semenya forget what she has achieved, from so little. SMH.

46† There is also a risk of tautological reasoning in the 2017 Study: Semenya has reported to have high-T and is likely to be included in the study (2011 Sample). Because Semenya was performing well in competitions (winning silver in the 2011 championship), the results of the study will show individuals with high-T performed well.

47† Kessel, A. (2018, 18 February). The unequal battle: privilege, genes, gender and power. The tense debate around Caster Semenya and Dutee Chand demonstrate the intersection between race, gender and medical imperialism. The Guardian. Retrieved from

48† Sheridan, D. (2018, 11 April). Critics of Caster Semenya forget what she has achieved, from so little. SMH.; Koh, B. (2013). Unheard voices of racism and sexism in sport: Is anti-doping but another “ism” in fairer clothes?,

49† It should be noted that Clause 5 of the DSD Rule limits any appeal against the regulations to CAS, and limits the appeal to largely procedural grounds. Whilst the validity of this clause may itself be challenged, CAS would still appear the most likely forum for any appeal - with substantive grounds, such as human rights, likely to be raised regardless of Clause 5.

50† Identoba & Others v. Georgia, no. 73235/12, [96], ECHR 2015.

51† Christa Tobler, ‘Equality and Non-Discrimination Under the ECHR and EU Law’ (2014) 74 ZaöRV 521, 538.

52† Novruk & Others v. Russia, no. 31039/11, [90], ECHR 2016.

54† IAAF Constitution, Article 4.11, provides that it is an object of the IAAF to "affiliate to the IOC and play a leading role in the achievement of the aims of the Olympic Movement."

55† International Olympic Committee, 'Olympic Charter' Fundamental Principles of Olympism (2 August 2015) at 14 <>

56† Sex Discrimination Act 1984 (Cth) s 5C.

57† Which would include high T levels for female-identifying intersex athletes. For more detailed discussions of what is meant by ‘discrimination’ beyond the simplistic ‘less favourable treatment’ definition, see Neil Rees, Simon Rice and Dominique Allen, Australian Anti-Discrimination & Equal Opportunity Law (The Federation Press, 2018, 3rd ed) 86-92.

58† That such intersex athletes could compete in male competitions would likely not change the outcome of any discrimination claim. If an intersex athlete defines their gender identity as female, and/or their sex characteristics align more with female sex than male sex, then their exclusion from female competition would likely be defined as discrimination.

59† Koh, B., Sonksen, P., & Adair, D. (2014). Testosterone, sex and gender differentiation in sport – where science and sports law meet. Law in Sports.

60† See, eg, Vicki C. Jackson and Mark Tushnet (eds), Proportionality: New Frontiers, New Challenges (Cambridge University Press, 2017).

61† One possibility is that the absolute values of T in the hammer throw and pole vault events in the highest fT tertile was not as markedly elevated as compared to the running events: the mean (SD) results in nmol/L for 400m, 400m hurdle, 800m, hammer throw and pole vault were 7.08 (11.64), 1.68 (1.63), 3.26 (6.60), 1.13 (0.36), 0.98 (0.31) respectively. However, this would not explain for the decision to include the 1500m event where the absolute values of T in the highest fT tertile was also not markedly elevated: the mean (SD) results in nmol/L was 1.12 (0.62). These values are comparable to the hammer throw and pole vault. Additionally, there was no statistically significant performance advantage for 1500m between this highest fT tertile compared to the lowest fT tertile.

62† Including but not limited to physical, mental, emotional, social and economic harm.

63† Buzuvis, E. E. (2010). Caster Semenya and the Myth of a Level Playing Field. The Modern American, 6(2), 36.

64† For example, hypogonadal males with low T.

65† Which is a discussion for another time: see, eg, Nancy Leong, ‘Against Women’s Sports’ (2018) 95 Washington University Law Review 1; cf Kate Lambeski, ‘Sorry senator, women don’t want to play with men’, The Sydney Morning Herald (online), 23 February 2018 <>.

66† Although on this issue it must be noted that a so-called “intersex” athlete may not identify as intersex. She may identify as female and forcing such an athlete to publicly identify and compete as intersex, or not compete at all, in circumstances where she does not identify as intersex or chose not to publicly identify as intersex is a cruel choice to impose on a human being.

67† Indeed, Dutee Chand expressed significant concern at potential adverse health effects of reducing naturally-occurring testosterone levels in the 2015 CAS case: Dutee Chand v Athletics Federation of India (AFI) & The International Association of Athletics Federations (IAAF) CAS 2014/A/3759, at 8-9.

68† See, eg, Peter Sonksen et al, ‘Medical and Ethical Concerns Regarding Women With Hyperandrogenism and Elite Sport’ (2015) 100(3) The Journal of Clinical Endocrinology & Metabolism 825; Patrick Fenichel, ‘Molecular Diagnosis of 5a-Reductase Deficiency in 4 Elite Young Female Athletes Through Hormonal Screening for Hyperandrogenism’ (2013) 98(6) The Journal of Clinical Endocrinology & Metabolism 1055.

69† There is little doubt that the IAAF would owe a duty of care to its athletes when implementing new policies and regulations. A negligence suit could turn on whether this duty of care has been breached by the IAAF, such that it causes harm to particular athletes.

Related Articles


Dr Ben Koh

Dr Ben Koh is a medical doctor with a Masters in Sports Medicine and a Masters in Psychology and has clinical and educational training in surgery, sports medicine, emergency medicine and critical care.

Daryl Adair

Daryl Adair

Daryl Adair is a historian and Associate Professor of Sport Management at UTS Business School, University of Technology Sydney, Australia. His recent research focuses on policy challenges in sport, such as integrity processes and the politics of elite participation.

  • This email address is being protected from spambots. You need JavaScript enabled to view it.

Liam Elphick

Liam Elphick is a Lecturer at the University of Western Australia Law School, and has worked in Australian Football League (AFL) scouting, recruitment and list management for the past decade. His research focuses on anti-discrimination law, particularly in the sporting context, and he is a member of the Discrimination Law Experts Group.