The limit of human performance: How much faster?

Swifter, higher, stronger...up to a point?  Or beyond?  The limit of human performance

Today we revisit a topic that seems to run like carousel, popping up once every few months on the site - how much faster can human beings run?  How close to a "ceiling" in performance are we?

The latest discussion is inspired by a few articles, and a recent round-table discussion between some of the all-time greats of 800m running, and the new world record holder, David Rudisha.  For those who haven't heard, Rudisha was appropriately named the IAAF Male Athlete of the Year for his two world records in 2010.

But preceding that announcement was a panel discussion between Alberto Juantorena, Sebastian Coe, Wilson Kipketer And Rudisha in Monaco.  To watch the discussion and read some commentary on it, check out Letsrun.com here.

One of the topics that came up was how much faster the record would still go.  Rudisha came agonizingly close (1/100th of a second) to dipping under 1:41.00, so it's safe to assume that barrier is in reach.  But how much lower?  Here are three quotes attributed to those legends:

"This record belongs to the future - 1:41.01." and "You never know the limit of a human being." - Alberto Juantorena

"I don't think we're anywhere near them (the limits of the event) ..." - Seb Coe

"In 800 it is possible to run under 1:40. It is still coming." - Wilson Kipketer

Then, an interesting piece by the Guardian discussed the same issues around whether we are near the limits, with the conclusion that it will take many years before we know the answer to this question.  I will point out that saying this is not quite the same thing as saying that we are not near the limit, but rather that we just do not know where it is.  There is a ceiling somewhere above us, but we don't know when we will hit our head!  You can read that piece here.

The physiological basis for "limits"

In trying to evaluate these arguments, I think it's important to understand a little bit about the physiological basis for why there may be a limit to performance.  And here, the most important thing to recognize is that we don't fully know what that basis might be.  It certainly varies by event, and is more complex than I want to go into now, but those keen for more might consider reading our series on Fatigue and Performance

For example, in 100m sprinting, some limits are the metabolic changes in the muscle, which affect its force-producing capacity, combined with mechanical factors such as muscle-tendon elasticity, ability to apply force to the ground, the force and torque on joints, and limits to how quickly neural signals can reach the muscle from the brain.

Some really interesting work by a colleague at my university has confirmed that the rate of force production and muscle relaxation drops over time during maximal exercise, even when the electrical signal to activate the muscle doesn't change.  In other words, there is a drop in the force that the muscle can produce, and that's why even events as short as 10 seconds show signs of pacing - if you go too hard early, you fall away at the end.  But, the peripheral factors, like acidosis in the muscle, depletion of ATP, accumulation of calcium and phosphate are only part of the problem, and the Guardian article talks more of the mechanics of sprinting, which affect the force production on the ground.

As you move up in distance, other factors play more of a role.  The rate of energy supply becomes a factor in middle-distance events, oxygen availability is a potential limiter (though whether it is to skeletel muscle or to the brain (more likely) is a debate).  So too, chemical changes such as a drop in pH may be regulated or responsible for a decline in performance.  Then, as you reach marathon distances, fuel availability becomes important (hitting the wall being the obvious sign of not getting this right), and the ability to burn fat to preserve glycogen is part of the elite athlete make-up.  Heat storage is another limiting factor, because fatigue occurs when the body temperature reaches what has been called a critical level of hyperthermia.

The point is that performances are not limited by one thing only, but rather a complex interaction between all the physiological systems, whose weighting depends on the type of event, and the external conditions for the event on the day.

Those interested in a more academic discussion of the topic might consider the following review articles:

Distribution of cycling power output: Impact and mechanisms

Physiological regulation of pacing strategies (and hence exercise performance)

Model for performance regulation by the brain

Applying this to performance limits

In any event, you may be wondering what this has to with a debate over whether an athlete can run under 1:40 for 800m, sub 9.40s for 100m, or break the 2-hour barrier in the marathon?

Well, in my view, knowing that performance is limited by physiological changes in the muscle, lungs, heart, brain, body temperature, there is an obvious "barrier" that cannot be broken without causing harm to the athlete.  We cannot simply head out and run or cycle ourselves to the limit - our brain controls the degree of muscle activation so that we are protected against, quite literally, exercising ourselves to death.  So the brain will, for example, detect the rate of heat storage early on and then reduce the exercise workrate through adjustments in muscle activation, the result of which is a slowing in pace, but also a drop in heat storage and avoidance of that limiting body temperature.  The same presumably happens with oxygen availability, glucose supply (probably both to the brain), blood flow, cardiac output, osmolality - any number of "homeostats" that have to be defended in order for us to survive

This is the reason, incidentally, that it is possible to predict the maximal sustainable power output by a cyclist during a mountain climb in the Tour de France.  We did this in July during the Tour, to some criticism, but I'm confident in saying that the physiological basis is sound, and so too is the prediction that power outputs of 6.1 W/kg to 6.3 W/kg represent a maximal power output that is possible given human physiology.

In other words, exercise performance is limited by a capacity in oxygen delivery, a capacity in heat storage and body temperature, a capacity in the rate of ATP supply, a capacity in the total energy availability.  Short of finding a human being who exceeds everything we know of physiology, or finding that individual who possesses the maximal combination of every single physiological attribute (this individual doesn't exist except in theory), the records will not "leap" forward, they will inch forward incrementally, and I do believe that we are quite close to the limit, when world records will become more and more infrequent, and eventually no longer be broken, unless we start measuring down to the nearest thousandth of a second.

The sub-2 hour marathon as an example

Let's look at the sub-2 hour marathon as an example.  This came up last week again, when Ed Coyle suggested that he was "confident" it could happen and predicted a 1:58 as the limit.  It won't be in my lifetime, that's for sure, but I'll get onto that shortly.

The 1:58 prediction, incidentally, is based around this paper, which applied much the same process as we did to cycling in the Tour to suggest the 6.2W/kg limit.  It works on the premise that performance in the marathon is limited (I would rather say regulated, but that's debated in our fatigue series) by oxygen delivery, lactate threshold and running economy.  It concludes that 1:57:58 is possible for "a hypothetical subject with a VO2max of 84 ml/kg/min, a lactate threshold of 85% of VO2max, and exceptional running economy".

The word "hypothetical" is important, because performance is not hypothetical.  That study was done in 1991, and knowledge of the limiting factors has evolved a little.  The role of the brain has become recognized, as have the mechanical factors such as energy storage and return in the tendons.  The key for me is that the athlete with a VO2max that high never has an exceptional running economy, so it is much like trying to find a motor vehicle with a 6 liter engine that also gets you 80 miles to the gallon (and those are the kind of unrealistic figures we're talking in combination) 

Another key point is that performance is determined by more than just the VO2 max and lactate threshold.  Athletes regularly "under-perform" given their physiological stats - perhaps it is because they lack racing nous or desire, they don't have the discipline to train, they are too big and thus heat storage becomes a factor for them.  Perhaps they are injury prone and so despite having the highest VO2max and economy in the history of sports science, they can't run 40km a week before breaking down.  The point is that for this "hypothetical athlete" to have a shot at a sub-2 hour marathon, there have to be hundreds of them, because the set of characteristics needed is not limited to three, and is extremely rare.

However, I would be surprised if there was even one such athlete,  let alone hundreds, because he would have been found by now.  An arguable point, certainly, but I believe that kind of physiology to be so rare that this person would stand out instantly and therefore, given the "free-market" that is sport, they would have been seen already. 

The performance spectrum - why one performance doesn't exist in isolation

The other reason the sub-2 hour marathon is, in my opinion, unlikely, is because it has implications for what happens at shorter distances.  I've described this before and would encourage you to read this post if you're interested, but the summary of it is that in order for a marathon to be run in under 2 hours, that athlete must possess a half marathon that lies closer to 57 minutes, and more tellingly, a 10,000m in closer to 25 minutes.  This once again comes back to the issue that all performances in these complementary events are limited by similar physiological "regulators", so that the physiology of a great 10,000m runners is often easily transferred up to the marathon (Haile Gebrselassie, Paul Tergat and Sammy Wanjiru are exhibits A, B and C).

So don't ask whether a sub 2 hour marathon is possible, rather ask whether a 25-xx minute 10km is possible.  Improvements in the 10,000m world record have declined in recent years, to the point that even a sub-26 minute time seems unlikely for a long, long time.  Therefore, even though mathematical predictions for the marathon have suggested that if performance continues to improve at the same rate, we'll see the 2 hour marathon in 2021, it seems unlikely given how improvements in the 10,000m event have dropped off recently.

Also, there are less "scientific" reasons why it will take much longer than predicted, and one is the lack of courses where a 2:03 time is possible, combined with commercial interests.  Right now, only Dubai, Berlin and possibly Chicago have the kind of course and money to drive a world record attempt.  Then it requires perfect conditions - 1 or 2 degrees too hot, a slight headwind, too cold, wet, and the record possibility disappears.  So unlike Ed Coyle, I'm not at all "confident" we'll see a sub-2 hour marathon.  Certainly not in our life-times, if at all.

The sub-4 minute mile response

So the question that comes up many times in response to this kind of opinion is that back in 1953, people suggested that the mile world record was at its limit and that the 4-minute mile would not be broken.  History clearly proves that to have been foolish, as the record is now 17 seconds faster, and many high-school athletes are breaking the barrier.  So therefore, is there not a chance that the same applies to the 2-hour marathon?

Of course there is.  But, there are some fundamental differences between that situation and the current one.  What we know of physiology now says that the 4- minute mile was always going to happen.  So if in 50 years, athletes are running under 2-hours in the marathon, then it will be because we have missed something in our understanding of the physiology today.  What are the chance of that?  Pretty high, of course, it would be arrogant to say that we know everything, we simply cannot.

However, I don't believe there are fundamental physiological principles that have not yet been discovered.  Performance is limited by the physiological regulators, and things like VO2max, running economy, threshhold running pace and thermoregulation are known to be regulators.  So we're either wrong, or we're still waiting for that one-of-a-kind human being who possesses physiological stats never seen before.  That wasn't the case in the 1950s - they were good athletes with exceptional but expected physiology, and it was lack of professionalism and training/diet, along with "vision" of what might be possible that limited them

Today, with money to be made, advancements in training, globalization of the sport (back then, Kenyans may have been running 3:50 for the mile, who knows?  They weren't competing enough), and a shifting of the horizon in terms of limits, we know much more what is possible.  We know what kind of physiological specimens exist, and I believe, we know what doesn't exist.  Genetic engineering may change that, but I really do believe we're approaching those limits.

In the 100m, Bolt came along and blew away the record books, but he hasn't done anything that mathematical models suggested would be impossible - they have the record limit at 9.48s, based on hundreds of years of data.  He just took us closer to it long before anyone thought it might happen.

Similarly, in the 800m, Rudisha has edged us towards 1:40 (only by 1/10th of a second, compared to Kipketer), and it does seem possible that this record will be improved again.  In the marathon, we have to find four minutes, from the same populations we're working with now, with limited opportunities for the record to be broken.  Physiologically, hypothetically, 1:58 is possible, but I don't share the confidence, and I don't believe that the hypothetical athlete exists, and I'd be very surprised if the record dips below 2:03 in the next fifteen years, and perhaps then we'll have a better idea of where the ceiling is, if we haven't hit our heads against it by then!

Ross

More detailed reading for those interested:


The Sub-2 hour marathon:  The link between 10,000m times and marathon times, and why it'll take longer than expected to crack 2 hours


The physiology of a sub-2 hour marathon runner