Human Limits

Exploring performance and health with Michael J. Joyner, M.D.

Photo of Michael J. Joyner, M.D.

How Believable Are Running Records?

In my last post I discussed the recent analysis of how physiologically “believable” key mountain climbs by noted riders in the Tour de France have been over the last 30 or so years.  The idea is doping likely contributed to some of the most unbelievable performances and that by analyzing power outputs, suspicious performances can be identified.    That having been said, it is interesting to note that the one hour record for cycling has hardly budged since 1972 when Eddie Merckx went just over 49.4 km at Mexico City.   People have gone much farther with exotic aero bikes, but the current record set in 2005 by Ondrej Sosenka is 49.7 km with a standard bike.  Sosenka had a number of doping violations that ultimately ended his career, so this record is suspect.  Merckx had issues with stimulants, but he rode in the pre-epo era and there is no evidence that I know of that he blood doped with traditional auto-transfusions.   So perhaps his one hour record sets the gold standard for what is physiologically possible in cycling without manipulating oxygen transport using epo or blood doping.


All of this led to some interesting e-mail traffic and a discussion about what is physiologically believable in running.  This is harder to guesstimate because unlike cycling, measuring the actual mechanical work done and power generated during running is very challenging.  It is also harder to link running performances to estimates of oxygen consumption.   However, one can ask what the best performance ever is in the pre blood-doping/EPO era was, use it as a baseline and then go from there.   This is reasonable at some level because by the 1960s people were training as hard as they do now, and lab data from that time suggests the record holders then were comparable to today’s top performers.  Additionally, the first real suspicions of blood doping emerged with Lasse Viren in 1972 and 1976.  Here is an example using men’s marathon and 10,000m performances:


  • Derek Clayton ran 2:08:33 in 1969 and his best for 10k was relatively slow 28:45 (likely on dirt).   In that era the fast 5,000 and 10,000m runners did not move up to the marathon that often, and if they did it was at the end of their careers.  There was no real money to be made running and the races (except for Boston and the Olympics) were pretty low profile.  Clayton’s record lasted 12 years until shortly after the big city/big money marathons started and top track runners “moved up”.   So people with 10k PRs faster than Clayton started to run the marathon and they started to run more of them at the peak of their careers.


  • The fastest 10k time on dirt is 27:39 by Ron Clarke in 1965.  If you figure a synthetic track is worth 2-3% then Clarke might have run ~27:00 on a synthetic track.  This time is equal to about a 2:04 for the marathon if you plug it into one of the better race conversion calculators.  Clarke ran his 27:39 alone and like Jim Ryun’s 3:51:1 mile on dirt there was not a pacer.  Both would have run faster in a real race with a pacer.



  • So it looks like a human might be able to run 26:45 with limited suspicion of doping and that works out to about a 2:03:03 estimated marathon.   If 26:30 is possible without drugs you get 2:01:54.  26:20 gets you 2:01:08. The race conversion calculators are not perfect but they are pretty good and most of them (plus an old point table system from the early 70s that I have a hard copy of) give relatively convergent estimates.


  • You can do a similar analysis (with a dirt track conversion factor) for the mile, 3000m and 5000m and the records from the middle 60s look very similar to those up to about 1990.  Also the best performances since about 2008-10 (when drug testing got better) are slower than the current records and perhaps closer to the pre-doping values.


Ultimately, who knows which runner has been doing what over the last 20 or so years?  Who ran what race clean, who doped, and who was clean all of the time and who doped all of the time.   However, like the hour ride for Merckx, there is no evidence that the current elite runners are physiologically better than people from an earlier era when techniques to manipulate oxygen transport for a competitive edge were not available.





Leave a Reply