Running comfortably

From time to time I report on some interesting research paper I come along and which is (somehow) related to athletics. This time I would like to report on the findings of a Stanford team on the energetics of running. The title of their article is a catchy one: "Running in the wild, energetics explain ecological running speeds". (I tend to bristle whenever I see the word "ecological" but, since the article does not make a big issue out of this, one can surmise that it was put in the title in order to increase the probability of acceptance for publication).

For more than half a century (in fact since the seminal work of Margaria and Ceretelli) it was believed that since the energetic cost of running is proportional to the velocity, the amount of energy used to run a given distance is nearly the same whether one runs at a leisurely pace or at top speed. This would mean that the speed can be adapted to the task demands without any appreciable consequence as far as the total energy expenditure is concerned. 

The constant energetic cost theory is of course not valid when it comes to walking since we know that the energetic cost of walking is proportional to the square of the velocity. The consequence of this is that there is an optimal step frequency which allows to mimise the energetic cost of walking. At this point and before presenting the results for running I cannot refrain from presenting the results of an ingenuous experiment by a canadian team of physiologists. 

They used a robotic exoskeleton in order to shift the optimal step frequency of their subjects to higher or lower values of the preferred one. They found out that, after a few minutes adaptation, the subjects adapted the step frequency to the new energetic optimum.

Running, despite the high energetic cost it entails, has the advantage of allowing to adjust the speed to the task. It is believed that this is an adaptation of our hunting ancestors who would choose their speed so as to make it unsustainable to their pray. 

The Stanford group performed a very large scale experiment involving some 4600 runners. What they found is that each runner has a preferred speed, a comfortable pace, independent of the length of the run and which moreover corresponded to a minimal energy use per distance. And the results were independent of the participants sex, age or body mass index. (Although the mean velocity of women was roughly 85 % that of men, a tad on the low side, since women's competition speeds are around 90 % of that of men).

The distances of the experiment were in the 2-10 km bracket. It is expected that when the distance increases beyond 10 km and fatigue sets in the comfortable speed can no longer be maintained. Similarly when it comes to training athletes are not expected to stay within their comfort zone. 

The Stanford group investigated the optimal character of the preferred speed by performing laboratory measurements of the energetic cost (admittedly on a small cohort) of running on a treadmill. Their findings seem confort the conclusion that the preferred velocity is indeed the one that minimises the energetic cost.

At this point I must confess that I am a little perplexed. I am convinced that there is nothing wrong with the results of Margaria and his team: the energetic cost of running is proportional to the velocity. The problem is that the physiologists are using two different quantities: what they call "cost of locomotion" which is the energy expenditure in a unit of elapsed time and the "cost of transport" which is the energy expenditure in a unit of distance covered. Now, an elementary physics calculation tells us that if the former is linear in velocity the latter would be quadratic. 

This appears to be borne out by various measurements but I cannot fathom the origin of the minimum of the cost of transport at some finite velocity. The only explanation I can find is that the downgoing part of the curve of cost of transport in the two figures above corresponds to velocities which are at the transition region between walking and running and thus not really representative of running.

Be that as it may, humans, when running for recreational reasons, adopt speeds that minimise their energy expenditure. For people who do jogging in order to burn some calories this looks like somewhat counterproductive. So what can calorie-concerned runners do? The authors of the article advise to run with a running mate, one who has a higher preferred velocity. A sound advice indeed.

Another potentially great site, but ...

Some time ago I wrote about a potentially great Finnish site called Tilastopaja.  I pointed out that the site was user-unfriendly, since there was no way to know how much you had to pay in order to become a subscriber before registering.

The site I am going to write about today, Track and Field Statistics,

is not as complicated. Once you reach the home page you are asked to register. But if you happen to have the link to the page with the results you can continue witout registering. 

Clicking on an event gives a few salient results. One has free access to top-10 lists but the previous years' lists are reserved to (paying) members. The financial conditions are clearly stated. Personally I find the 100 $ per year subscription rather stiff. Moreover the results of major championships are freely available on the web, so, why pay for them?

The only thing I found interesting were the electronic books (mind you, they are not books but rather Excel files) which give statistical results from 1891 for men and 1921 for women. One book (well, part of a book), that of men's 800 m, is freely available and I must say that it is really interesting. So if one is looking for some athletics paleo-history, paying 10 $ for a book is not excessive (and you can have all 36 of them for 150 $). However, forking out 100 or 80 $ (there are two tiers) for an annual subscription, this is a no-no for me. 

So, another potentially great site, but...

The final word on testosterone by Ross Tucker

In a recent twitter thread Ross Tucker (a physiologist for whom I have a great respect) addressed the question of testosterone and the way it has been used in order to formulate policies for DSD and trans athletes. To my eyes his explanations settled once and for all the question (although the question of how one can deal with DSD and trans athletes is far from settled). Thus I decided to write a post presenting his ideas. But before doing this it is important to summarise the situation (although I have, on several occasions, written posts on this burning subject).

To my eyes women's sports, and in particular athletics, is something sacred that we should strive to preserve at all costs. I understand perfectly that men should be allowed to decide to pursue their lives as women, if this would bestow them a better equilibrium. I respect totally such decisions, just as I do for the opposite ones, women who decide to live as men. However I draw the line in the case of sports. A man should never be allowed to participate in women's competitions. Tucker explains perfectly the rationale behind this. While the case of transgender athletes is not negotiable, the case of DSD athletes is more complex.

In my post on sex verification I told the story of how, over the past century, there have been several occasions where misguided sport managers have tried (in some cases with success) to have men compete in women's events. Today, such an outright dishonest behaviour is practically impossible. But, unfortunately the peril for women's athletics is even bigger. It comes under the acronym DSD, that stand for Differences in Sexual Development. It is the politically correct version of what was called before "hyperandrogenism". 

Hyperandrogenism is a medical condition characterised by high levels of androgens. DSD is a more complicated situation and refers to congenital conditions affecting the reproductive system, in which development of chromosomal, gonadal, or anatomical sex is atypical. Persons with DSD are often referred to as "intersex". The case that introduced hyperandrogenism to the athletics vocabulary was that of C. Semenya.

I don't know how many times I have written on Semenya (too many) but, for the possible newcomers to this blog, I would like to summarise the situation.

Semenya came into the limelight when she won the 800 m in the 2009 World Championships with 1:55.45, a year-over-year improvement of her personal best by almost 9(!) seconds. The reactions didn't take long, the IAAF was constrained to launch an investigation and it resulted that Semenya was intersex. The counter-reactions did not take long either. The IAAF was accused of racism and imperialism. It goes without saying that I find these arguments scurrilous: the intersex condition of Semenya has nothing to do with the fact that she is black and african. 

Anyhow, Semenya was cleared to participate in women's competitions and she went on to win the 2011 and 2017 World championships as well as the 2012 and 2016 Olympics (to say nothing of African Championships or Commonwealth Games). In 2015 the IAAF introduced the testosterone rule aiming at blocking hyperandrogenic females from competing with non-hyperandrogenic women. Following an appeal of D. Chand, the CAS invalidated the rule allowing Semenya to participate in the 2016 and 2017 competitions. (Chand also profited from this in order to participate in major competitions, including the Olympics, her major success being a gold medal over 100 m in the 2019 Universiade).

The IAAF's petition to the CAS decision was re-examined in 2018 and a new testosterone rule was introduced forbidding participation to hyperandrogenic women in the events spanning the distances form 400 m to a mile. Semenya's appeal was rejected in 2019 limiting her choices to either undergo a pharmacological treatment in order to bring down her testosterone levels or move to longer distances. (She tried the latter without success). F. Niyonsaba, who won silver in Rio and London in the 800 m, being of a much lighter frame than Semenya opted for the longer distances where she has now personal bests of 14:25.34 in the 5000 m and 30:41.93 in the 10000 m. In Tokyo she was disqualified in the former distance during the semi-final, otherwise she would have certainly obtained a place on the podium. 

It was another DSD athlete who drew all the media attention in Tokyo. Early in the season, there were news of two 18 year old namibian athletes who registered fantastic times over 400 m, Christine Mboma and Beatrice Masilingi: 49.53 and 49.22 s. And then the verdict was handed down: both namibian teenagers were hyperandrogenic. Thus they were not allowed to race the 400 m any more. (Mboma's out-of-this-world time of 48.54 s was not homologated). Unfazed they converted to 200 m runners and went on to participate in Tokyo's olympic final where Masilingi was 5th (with 22.18 s) and Mboma second with an incredible 21.81 s personal best.

World Athletics is watching from the sidelines, Sir Sebastian having stated that Mboma's success suggests that the testosterone rules are vindicated. And when asked whether Mboma could break the world record he replied that this “probably” would then give the governing body more questions to contend with where the rules are concerned.

Before presenting Tucker's arguments let me remind you what does the "testosterone rule" stipulate (restricted events being the distances from 400 m to the mile including, hurdles relays and combined events):

If a female athlete wishing to participate in a Restricted Event at an international competition has a DSD that results in levels of circulating testosterone greater than 5 nmol/L, and her androgen receptors function properly, such that those elevated levels of circulating testosterone have a material androgenising effect, she must reduce those levels down below 5 nmol/L for six months (e.g., by use of hormonal contraceptives) before competing in such events, and must maintain them below that level until she no longer wishes to participate in Restricted Events at international competitions.

Without further ado that's what Ross Tucker has to say.

Testosterone and performance has become a kind of fixation, but  testosterone level per se is not the right place to look. Testosterone is the driver of the differences between men's and women's performances and so sports tried to use this cause-effect in order to exclude people from women's category.

But women's category in sport exists to exclude people who experience androgenisation during puberty and development.

People think that reversing T levels will remedy this. This is misguided. Once androgenisation has occurred it's too late. Some changes are irreversible, other only partly reversible and only a few can be undone. Lowering testosterone levels does not bring fairness into sport because testosterone has already done the work, primarily in puberty.

A "snapshot" of testosterone levels cannot be related to performance. And thus it is misleading. The important thing is whether androgenisation has occurred and testosterone levels are an excellent guide to whether it has or not. What matters is whether testosterone is high or not, the actual level is not important.

Here Tucker gives a great example. Measured VO2max does not correlate with marathon performance. All good marathoners have high VO2max. You need a high value to become a good marathoner but once you are there, there are other things that count. Of course there are people with not-high VO2max. They cannot become good marathoners.

So, the presence of high level of testosterone during life plus the ability to use it for androgenisation is the important thing. Androgenised people are in male category, unandrogenised are in female category. Lowering testosterone levels does not allow a male to become female.

Some male can be non-athletic and, were they to compete with women, they would be beaten. This is the same phenomenon as when an able-bodied athlete sneaks into a disabled category and still loses. But this failure does not invalidate the essential difference between categories.

And when people asked the question on how to implement segregation, Tucker replied that it should not be based on testosterone levels but rather on biological sex. 

That was the clearest discussion of the testosterone effect on performance that I have ever seen.

I am appalled by the fact that right now we are ready to sacrifice millions of women athletes in order not to look unfair towards the one transgender male who decided that, since he could not beat men, he would try his luck against women.

But then we are living in an era of woke-ism and these things should not astonish me any more. 

For the time being we are in the middle of an assault, in the name of fairness towards (erstwhile) oppressed categories, against what women have obtained through their hard-fought battles over more than a century, namely an equity in sports. Is there any hope for rule changes that would guarantee fairness in women's competitions?

When questioned, Lord Sebastian did not rule out the possibility of looking at the DSD rules governing other distances: “My responsibility, as uncomfortable as it is, is always to do whatever I can to maintain the integrity of competition and that level playing field. We’ve never said that there may not be advantages at lower levels elsewhere”. He also suggested that different classifications might yet come into sport in the future – including an “open” category for almost everyone and a restricted “biological female”. “I don’t think that’s where most of my council are at the moment,” he said. “But this is a debate at the moment. I’ve heard coaches and people who are interested in the sport, and more broadly from the sport, have discussed that”. 

Meanwhile, the WA site goes into ecstasies over C. Mboma's performances.

A brief history of the IAAF/WA scoring tables: 1912-1932

While combined events were introduced already in the antiquity, they were absent from modern athletics till roughly the middle of the 19th century. Then all-around competitions started being organised and a decathlon, not very different from the one we know today, made its appearance in Sweden with the first championship held in 1909. No serious olympic combined event existed before the 1912, Stockholm, Olympics. In my article "Olympic combined events at the turn of the century" I give a short summary of the situation concerning the 1904, St. Louis, Olympics, as well as an account of what happened during the 1906, Athens, Games and the pentathlon which was a copy of the ancient one. But while reading for the nth time Zarnowski's book I came across something that I totally ignored. A pentathlon was in the program of the 1900, Paris, Olympics. It was a farcical event, just as the rest of the Paris organisation. To begin with, despite being announced as pentathlon it was in fact a tetrathlon. The athletes had choices: they could run either 100 m or 400 m, 800 m or 1500 m, participate in any of the three jumps, long, high or pole and throw either the shot or the discus. And of course no details exist as to the scoring planned. But this is not a real problem as the event was never(!) contested. 

The, probably, first scoring table was one introduced in the US in 1893 (and not in 1884, as  World Athletics insists of giving as the date, despite the fact that Zarnowski has pointed out the mistake long ago). 

The Scandinavian countries, did also produce scoring tables, since they were already organising multi-event competitions. Contrary to the american tables, where 1000 points were attributed to the world record, these scandinavian tables were based on the national records of each country. Germany developed tables, also based on the world records, probably in preparation for the 1912 Olympics, (over 100 rather than 1000 points). 

All the aforementioned tables are plagued by two flaws. 

First, placing the 1000 points at the world (or national) record level, makes them non-adapted to the multi-event. Can one imagine modern decathlon tables where 1000 points for the 1500 m would correspond to 3:26.00? In case you wonder, in the scoring tables in effect one must run a 1500 m in 3:53.79 in order to obtain 1000 points. And, while it has never been done, it is not something that would appear out-of-this-world. Of course one can argue that the points attributed are calculated relatively to the 1000 performance and thus the fact that the latter corresponds to the world record is not crucial. However this statement makes the tacit assumption that all world records are equivalent, something we know not to be true. 

The second flaw is that all these tables are linear. One must of course be lenient. At the beginning of the 20th century the computing possibilities were quite restricted (but, of course, adding plus printing machines did exist) making the construction of the tables highly labour intensive. Being linear, i.e. neither progressive nor regressive, would not be a real flaw for scoring tables were it not for the fact that the linearity in the track events is as a function of time. As I have argued in this blog (and published papers), the proper quantity in which one should cast scoring is not time but, for running events, velocity. And when one examines the scoring of the linear tables as a function of the velocity one sees that the tables are, in fact, regressive.

With the 1912 Olympics approaching the organising committee set about preparing a new set of tables. They did not use the already existing swedish "Malmö" tables but attempted the construction of new ones. There was an initial attempt at producing progressive tables but it was abandoned due to the difficulties involved and the lack of time. Finally the tables adopted were linear with the 1000 points corresponding to the 1908 olympic record. In order to be able to attribute a scoring to every performance, down to cm precision, fractional numbers of points (with three decimals) were introduced, unnecessarily complicating the situation. 

Following the 1912 Olympics the tables were revised, with the 1000 points corresponding now to the 1912 olympic record. They were officially accepted after the war and were used in the Olympic Games from 1920 to 1932 (and, in fact, also during the first European Championships in 1934). 

When one reads the scoring tables history in the World Athletics document, one encounters a cryptic statement concerning the bottom of the tables i.e. the performance that obtains zero points. What is implied there is that the null-score for running would correspond to the time for walking the distance. Nothing is further from the truth. The null-score time for 100 m in the 1912 tables is 14.8 s. It corresponds to a velocity of 6.75 m/s. To make things clearer, were one to run a 10000 m race with such a mean velocity, the final time would be under 25 minutes! I think that this walking vs. running was added as an afterthought, once the people who were working on the tables in the IAAF/WA had heard this idea from my friend Gerry Purdy, who was the one to introduce a systematic approach to the null-score. But this is a story we shall tell another time.