Wednesday 18 October 2017

“Gender doping”: should women with high testosterone be allowed to compete in athletics?

At the University of Essex we have just started a new MSc. in Sports and Exercise Science [1]. As part of my drugs in sport module, I am getting my students to examine a “hot” research paper each week. They do the background reading and then we discuss the paper together asking the questions: Why was the research done? What did the research find? What were the problems and limitations of the study? What are the consequences for sport?

I am really enjoying this type of small group teaching. The nine students are a mix of UK and overseas, some straight from undergraduate degrees and others with real world experience. The discussions are lively and they all bring different ideas to the table.

The first paper we discussed [2] was one about a condition called hyperandrogenism  - where someone has excess levels of the natural anabolic steroid testosterone. The paper is about to play a key role in the upcoming (any day now!) verdict from the Court of Arbitration of Sport with regards to the IAAF’s hyperandrogenism regulation. Approved in 2011 [3], the “Regulations Governing Eligibility of Females With Hyperandrogenism to Compete in Women's Competition” state that any female athlete with a level of testosterone above a stated value (i.e. in the “male range”) will not be allowed to compete until they had taken measures to reduce this value (usually this would require hormone treatment). The IAAF argument was that the difference in sporting performance between elite men and women is predominantly due to higher levels of testosterone in men. This is why women tend to get excess benefits by doping with anabolic steroids (such as testosterone) when compared to men. Therefore it is “unfair” for someone with a natural high level of testosterone to compete as it gives them benefits akin to doping. Note that at no point did the IAAF blame a “hyperandrogeic” athlete nor argue that they were not a woman - just that they are not allowed to compete in elite women’s sport (there is a distinction here if you look hard enough I guess).

An Indian sprinter – Dutee Chand – refused to take treatment to lower her testosterone and challenged the legality of the regulations at the Court of Arbitration for Sport (CAS). On 24 July 2015 CAS issued an Interim Award in the arbitration procedure [4]. They suspended the regulations for two years to give the IAAF the opportunity to provide scientific evidence about the quantitative relationship between enhanced testosterone levels and improved athletic performance in hyperandrogenic athletes. I can recommend reading the full award; it is a fascinating discussion of this difficult area [5]. However, to summarise a long case, essentially the IAAF had argued that natural (endogenous) testosterone gave a female athlete the equivalent benefit of unnatural added (exogenous) testosterone. The IAAF did not dispute this idea in principle, but wanted to see proof. Hence the two-year study that culminated in the research paper I discussed with my students [1, 2]. In the meantime with the regulations suspended Dutee Chand and other hyperandrogenic athletes were allowed to compete with their natural (high) levels of testosterone [6].

The IAAF and WADA funded study was published in the British Journal of Sports Medicine [2]. The study looked at over two thousand observations from elite athletes competing at the 2011 and 2013 IAAF World Championships. Measurements were made of blood levels of testosterone and comparisons made to performance. I have some potential issues with the statistics used in that they seemed a bit too generous towards finding a performance association, but perhaps that is a story for another day. The key finding, that no doubt will be forwarded to CAS, is that after dividing the athletes into thirds – high, medium and low testosterone – women in the highest testosterone category performed significantly better in 400m, 400m hurdles, 800m, hammer throw, and pole vault with margins of 2.73%, 2.78%, 1.78%, 4.53%, and 2.94%, respectively. This pattern was not seen in any male events.  

So what did my students think of it? Well they made some great observations. Perhaps the most interesting was that the effects were restricted to so few events. Results in many power-based events showed no association with increased testosterone. Indeed in the 100m, 100m hurdles and 200m those with the highest testosterone performed worse than those with the lowest testosterone. The difference was not scientifically significant, but clearly is heading in the wrong direction for the IAAFs case. However, it could be legally significant given Dutee Chand  - who bought the case after all – is a 100m sprinter! 

Actually what was not observed is almost as important as what was. Across female track and field events there was no association of testosterone levels with discipline i.e. elite female athletes in strength events such as shot putters and sprinters do not have higher testosterone than those competing in distance events like the 10,000m and marathon. Even more surprising (at least to me as an oxygen “expert”), the levels of the oxygen carrying protein hemoglobin were not greater in the “aerobic” events. This result is hidden away in Supplementary Table 8 if you want to find it. There is a trend for sure, but I would have expected the difference to be more evident, given that we know that long distance runners are doing all they can within the rules (altitude training, sleeping in low oxygen tents) and without the rules (EPO, blood doping) to increase their hemoglobin; something sprinters don’t bother with at all.

The fact that a biochemical marker like hemoglobin that athletes can manipulate and that is known to affect performance shows such a poor association with performance, does question the whole idea of these kind of “association studies” even when they include so many athletes.

Of course the real problem with this study is that it is – quite rightly - limited by ethics. None of the female testosterone values in this study were in the “male” range. Otherwise the athletes would have been banned at that time of course. The best way to prove that hyperandrogenism is equivalent to testosterone doping is to dope elite athletes with testosterone for a short time. Then look at their performance when they are on and off the steroids. Does the increase and decrease mirror those seen when a hyperandrogenic athlete comes on and off hormone therapy to vary their natural levels of testosterone? Good luck getting that study through a university ethics committee!

One final point is that the paper also looked at male athletes. Unlike with women, the study did find that male sprinters had increased testosterone levels compared to other events. However, in no event was there any association of an individual’s testosterone level with performance in that event. In the discussion the authors’ suggest there is “sigmoidal” curve associating testosterone levels and performance. For women this means that increasing testosterone can have a large effect on performance (the association curve is steep), but at the higher values seen in men the relationship smooths out and is undetectable. Of course this begs the question. Why is taking testosterone and other anabolic steroids banned for male athletes, if the performance benefit is so marginal as to be unmeasurable? I am not sure the IAAF and WADA can have things both ways. If associations with performance can be used to ban female hyperandrogenic athletes from performing, then surely the lack of such an association can be used to unban male anabolic steroid dopers? I think Shakespeare called this being hoist on your own petard [see Hamlet Act 3, scene 4).

What ever CAS decides, at least my students enjoyed discussing the paper. “Much better than my undergraduate teaching said one”. Given I taught him as an undergraduate, this was a somewhat back handed compliment, but I’ll take it anyway!  

Onwards and upwards ….next week we are discussing a controversial paper that suggests  that  blood doping does not increase performance in long distance aerobic sports events such as cycling [7]. Can’t wait……


[1] MSc. in Sports and Exercise Science, University of Essex, UK


[2] Bermon S, Garnier PY (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. 51:1309-1314.


The paper can be downloaded from the IAAF web page: https://www.iaaf.org/news/press-release/hyperandrogenism-research


[3] IAAF to introduce eligibility rules for females with hyperandrogenism ; IAAF news April 2011


[4]  CAS media release: CAS suspends the IAAF hyperandrogenism regulations Lausanne, July 27, 2015


[5]  Dutee Chand v IAF and IAAF, Interim Arbitral Award delivered by the Court of Arbitration of Sport,  July 27, 2015


[6] The most famous current elite female athlete is the South African World and Olympic 800 m champion, Caster Semenya. Since the lifting of the IAAF regulations her success has come under some criticism from fellow athletes.


However, it was notable that this is not necessarily shared by the athletics watching public. I was at the London Olympic Stadium two months ago when the crowd were right behind her during her race, in her post racer interview and during the medal ceremony.



[7]  [Heuberger J, Rotmans JI, Gal P et al (2017) Effects of erythropoietin on cycling performance of well trained cyclists: a double-blind, randomised, placebo-controlled trial. Lancet Haematol 4:e374-e386.