In this article, by using a process of logic, it is shown why altitude training is counter-productive.

Tuesday, April 6, 2010


"It is the customary fate of new truths to begin as heresies and end as superstitions." T X Huxley

The history of orthodox medicine reveals a number of procedures that were once believed to be beneficial but were later found to be useless - or even harmful - and consequently abandoned. Examples include: - blood-letting and leeching to lower blood pressure; the administration of mercury to cure syphilis; trepanation, lobotomy and electric shock therapy to cure mental problems; etc. It's quite probable that some current medical procedures/beliefs will at some future time undergo the same fate. Anyone for homoeopathy, chiropractic, acupuncture, or magnet therapy?

The history of athletics coaching also contains a number of fads and beliefs that were later discredited and abandoned. However there is one in particular that remains, is alive and well, and is rarely seriously questioned or challenged. It is the claim that there is a competitive advantage for distance runners who engage in altitude training. It is a belief held - widely and almost universally - among coaches, runners, and exercise physiologists for many decades now. However in all that time no one has ever supplied irrefutable proof of its validity. Rather than stating definitively that altitude training confers an advantage, studies tend to qualify their findings with phrases such as “it seems” or “it appears” (to confer an advantage).
When it first became fashionable, it was believed that for runners to benefit from altitude training, they needed to both live at altitude and train at altitude. In recent years, alternative permutations have been postulated, e.g. live at sea-level and train at altitude. Most recently the belief is that for the best outcome runners should live at altitude/train at sea-level. It seems to be just a matter of time before the advice will be to live at sea-level/train at sea-level!
However for that to happen an awful lot of coaches and exercise physiologists will have to accept that they were wrong for a very long time. But as that would result in a diminution in their professional reputation, it appears unlikely to happen any time soon. Instead, it seems, they will continue to hold on tightly to some component of altitude training, however illogical. What else can explain the current thinking that it is best to live at altitude/train at sea-level?

What follows is a definition of altitude training; then a brief history of how the fad originated, and a rebuttal of the thinking that led to it. Then a section outlining why altitude training is counter productive. After that, a section questioning why then, if it is counter productive, that altitude trained runners are the more successful. Finally, a consideration of an instance when it is necessary to engage in altitude training. Yes! believe it or not, there is an occasion when it is necessary!

Anyone who wishes to comment, to correct, to agree or to disagree on anything, please use the comment email address -
Write the word “altitude” in the subject box. All relevant comments will be published.

Because of the physiological changes that may occur, altitude training is believed to confer an advantage. It usually involves the runner living and training, during the out-of-competition periods, at an altitude which is generally in the range of between 5,000 and 8,000 feet. The length of the period spent at altitude will vary depending on the circumstances of each individual athlete but is usually of the order of 2 to 3 months. (Some athletes, aside from competitions, try to spend as much time as possible at altitude.) Following this, the runner will compete in a major competition, or series of competitions. (With only a few exceptions, practically all major competitions are held at low altitude, i.e. close to sea-level.) Following this, the runner returns for another period of altitude training and the cycle repeats.

As stated earlier, the most recent thinking advises the athlete to train at sea-level and live at altitude. As this is practically impossible to carry out under natural conditions, a number of contraptions have been invented which reproduce the rarefied air which exists at altitude. The runner uses one of these devices when not training and/or when sleeping.

Prior to the mid-1960s it is doubtful if any distance runners deliberately sought to train at altitude. The top African runners of that time, principally Kenyans and Ethiopians, were training simply where they lived, which happened to be at a high altitude. It was the startling success of one of these runners, the Ethiopian Abebe Bikila, in the Olympic marathons of 1960 and '64, along with events occurring at the distance races of the 1968 Games - held at the 7,300 feet altitude of Mexico City - that started the trend for, and the belief in, altitude training. For example: -
● It was postulated that runners who are born and live mainly at high altitude have an inherent advantage that is denied to those fated to be born and raised at a low altitude. (With the gap in the level of performance between the present crop of top East Africans and the best of the rest of the world now at an all time high, that belief is perhaps as strong and prevalent as ever.)
● Adding to this view were the failures (one spectacularly so) of two of the favourites for gold medals at Mexico City: - the 1,500 meters and the mile world-record holder, the American Jim Ryun; and also the 5K and 10K world-record holder, the Australian Ron Clarke. In the 1,500 meters, Ryun was beaten into second place by the Kenyan, Kipchoge (Kip) Keino, while Clarke finished 5th in the 5K. In the 10K Clarke managed only 6th and required immediate medical attention and the administration of oxygen.
● And while at those particular Games Bekila failed to finish in the marathon, his Ethiopian compatriot, Mamo Wolde, took the gold in his stead.

● What many failed to appreciate regarding those failures of Jim Ryun and Ron Clarke are the special circumstances which caused them. Ryun had in fact undergone a period of altitude training prior to the Games. However he was convinced that the 1,500 meters would be won in a time of no faster then 3:39 and so planned to run the race accordingly. When the pack, led by Kip Keino, passed the bell in 2:54, Ryun was 6 seconds adrift. Realizing his miscalculation and quickening on the final lap he managed to come second in 3:37.89 to Keino's 3:34.91. Had Ryun not underestimated Keino, a former 3K and 5K world record holder who would go on to win the steeplechase at the Munich Olympics and take silver there at the 1,500 meters, he might have taken the gold. In Ryun's case it is probably fair to say that rather than being beaten by altitude or a failure to altitude-train, he was beaten by a combination of his own miscalculation, and by a superb athlete.
● Significantly, Ron Clarke did not undertake any altitude training prior to the Games. It also later emerged that he had an underlying heart condition which was possibly exacerbated by racing at altitude. So in Clarke's case, it is probably fair to say that he was not so much defeated by altitude trained runners, but rather defeated by the altitude.
● As regards the successes of Bekila and Wolde in those Olympic marathons, Ethiopia had earlier sent a team of athletes to the 1956 Games in Melbourne. Those altitude-trained runners failed to win any medals there. In Bikila's case he wasn't even scheduled to go to Rome but was picked after the first choice runner got injured. Had that not happened, it is probable that the gold medal would have been won by a non-altitude-trained runner and the myth of altitude training might never have developed. As regards Wolde, he had in fact taken part in three previous Games without medal success before getting gold at Mexico where the high altitude was to his advantage.

Why altitude training is counter-productive
Because it is practically impossible to conduct a scientific study that concludes irrefutably whether altitude training or sea-level training is more beneficial - there are just too many variables that can't be controlled - it is necessary to approach the subject from a theoretical perspective. With this approach there will be only one variable: - altitude.

Now it should not be necessary to state this, but the best training for someone seeking to be a successful distance runner, is neither cycling, or rowing, or swimming, but running; lots and lots of it, (but without over-doing it). And that can be narrowed down to two basic types: - Long Steady Distance (LSD), to develop stamina/endurance; and Speed-work, to perfect racing speed. (Sure, a serious runner will have a programme that involves other exercise components e.g. flexibility, weights etc. but no one is suggesting that there is a benefit to be gained by doing them at altitude.)

So now let's assume we have two runners who are equal in every way: - size, weight, lung capacity, motivation etc. etc. etc. Most importantly, they have an equal capacity to undergo, to benefit, and to recover from a particular training load. Let's call them Runner A who will be the altitude trained guinea pig, and Runner S who will train at sea-level. Now we give them an identical training programme, sufficiently taxing - without being excessive - to expect them to improve. Runner S remains and trains at sea-level while Runner A is sent to an altitude of 25,000 feet. Now, obviously an altitude of 25,000 feet on this planet will have an extremely cold climate, so for the purposes of this theoretical perspective we give it the same climate as that at sea-level; the only difference therefore between the two locations is altitude (and, of course, their resultant different atmospheres).

Effect of altitude on LSD running
Let's now send the two runners on a LSD run that has an identical training load/effort. For a 2 hour run at a particular training load (pace), Runner S will cover a certain distance. At 25,000 feet Runner A will experience the same training load while just walking, or perhaps doing the lightest of light jogs. He most certainly will not be able to sustain any sort of run for the 2 hours, and will cover very little distance. If then after a period (say 3 months) of LSD training only, we match the two runners in a distance race at sea-level, there is no doubt that the winner will be Runner S. In contrast Runner A will actually be less fit (from the perspective of running at sea-level) than before he went to altitude. While he has undergone a similar training load during the 3 month period, effectively he has been de-training (as a competitive runner at sea-level). [In essence, the result is somewhat similar to matching a top class runner against a top class swimmer (or cyclist, or rower, etc.) in a foot race. The runner will win; but reverse the sport, and the swimmer will win in a swimming race, as indeed will the rower in a rowing race.]

So now instead of having had Runner A go to 25,000 feet, let's assume we had sent him instead to 15,000 feet. At that altitude the air is still pretty thin but even so he may be able to sustain a steady run for 2 hours. However the pace he can maintain and consequently the distance he can cover will be substantially less than Runner S can, for the same training effort. Less mileage at a slower pace is not going to produce a faster runner. When after a period of altitude training he races against Runner S (at sea-level) he will still come out second-best, although by a smaller margin than had he been training at 25,000 feet.

Next let's look at a scenario where instead of having Runner A go to either of the two altitudes discussed above, instead he goes for a training period to 6,000 feet - a popular height for altitude training. Here the effect of altitude is much less pronounced and many runners can maintain a pace close to their sea-level potential. Nevertheless for the same training effort, Runner A will still cover a shorter distance than Runner S in a set time. (Anyone who does not accept this need only examine the times set by those distance-runners who compete regularly at both altitude and sea-level.) When the two runners are matched in a distance race at sea-level, following the pattern set in the other examples above, Runner S will still be the winner although by a fairly small margin. (Needless to say if both runners had undergone an identical training programme and training load at sea-level, then when they are matched in a race it will result in a dead heat.)

Effect of altitude on Speed-work
Next let's consider the effect of altitude on the speed-work component of training. There are many different patterns of speed-work e.g. fartleck, tempo work, etc. - but basically it can be simplified down to 3 types of workout: - sprint repetitions; middle-distance repetitions; and long-distance time-trials. For the purposes of this examination we will presume 'altitude' to be in the range of 5,000 to 8,000 feet; that 'sprints' are in the range of 100 to 200 meters; that 'middle-distance' is from 400 to 2,000 meters; and that long-distance time-trials are in the range of 3K to 10K, (and even longer for marathon competitors).
Let's look at how each type of workout effects a runner at altitude compared to an identical runner performing the same workout at sea-level. Let's consider long-distance time-trials first. It's assumed that these will be carried out at maximum effort, or close to, the intention being to reproduce race pace. As shown above for LSD, Runner A will not cover a set distance at altitude as fast as Runner S at sea-level, for the same training effort. Over a period of time it will accumulate to result in a law which states that: - all other things being equal, the same amount of speed-work at a slower pace will not produce a faster runner. While Runner A's heart/lungs will undergo the same training load, his joints, and perhaps other parts of his musculoskeletal system, will experience a slightly lesser stress than that of Runner S. On the face of it this might appear to be a recommendation for altitude-training, but it is hardly sufficient to outweigh the disadvantages.
As regards the middle-distance repetitions, it is assumed that they will be carried out at maximum or near-maximum effort. Because of the thinner air at altitude, Runner A - at maximum effort - will take longer to complete each of the reps compared to Runner S. (At near-maximum effort he may be able to match Runner S for the time taken for each rep, but only by exerting a greater effort resulting in a greater oxygen debt that will take longer to recover from in order to be ready for the next rep. Of course this breeches the terms of this theoretical analysis; his training load/effort must match that of Runner S, not exceed it.)
As regards sprint repetitions, it is generally accepted that a top class sprinter can cover a distance of somewhere between 250 and 300 meters at maximum speed before oxygen debt forces him to slow down. That is why while the world record for the 200 meters is practically double that for the 100 meters, the 400 meters has to be run at a somewhat slower average speed. (Actually the 200 meters is usually completed at a slightly faster average pace than the 100 meters for the reason that for the second half the runners enjoy a flying-start, unlike the 100 meters where they have to start and accelerate from stationary.) While distance runners may not be able to travel as far at their top speed as sprinters before suffering oxygen debt, especially at altitude, it's assumed that they can manage reps of up to 200 metres. Inevitably Runner A will take longer to recover than Runner S from each individual rep (because of the greater oxygen debt incurred at altitude) and so will not complete as many in a fixed time period - the only way to determine that they are exposed to the same training load/effort. Less speed-work (for the same training load/effort) will not produce a faster runner; just the opposite. It can however be argued that Runner A's sprints are actually accomplished at a faster pace than those of Runner S because the thinner atmosphere gives less air resistance allowing for faster movement. It is doubtful if this is of enough practical significance to warrant training at altitude. Even if it is, it is not necessary to go to altitude to reproduce the conditions for a faster pace. It can be done at sea-level by running with a following wind, or by simply doing the reps on a slightly downward slope.

So considering the scenarios outlined above, it is clear - although the difference is small - that the runner who trains at sea-level will benefit more from the same training load/effort than the runner who trains at altitude. The inevitable conclusion therefore is that altitude training is counter-productive.

Why altitude trained runners are the more successful
Successful altitude trained runners can be categorized under one of two groups: - those born and usually living near sea-level but who go to altitude for periods of training; and those born and living - except for competitions - at altitude. Of the two groups the latter one, consisting mainly of the Ethiopian and Kenyan distance-runners, is by far the more successful. Let's look at the reasons. [N.B. It can be argued that the East Africans do not actually partake in altitude training; they are simply training where they live, which happens to be at altitude. However a trend seems to be developing there in that many of them go to spend periods training at an even higher altitude - within their own countries - to where they usually live.]

● The first point to appreciate is that it is not due to the fact that they are born and usually live at altitude. If it were, then we would expect to see equally good runners emerging from the many other countries around the world where large numbers of their populations live at altitude. However few of those other high-altitude locations are at or near the equator. Kenya and Ethiopia being on or near the equator, results in their high-altitude regions enjoying a year round climate of warm days, that are rarely excessively hot; and cool, but not cold, nights (a surprisingly benign climate when discovered by those when they first go there). This allows for training and competition with the minimum of kit at anytime of the year. In most of those other countries with populations living at altitude, the climate tends to make it unattractive for year-round running, because winters produce snow storms and freezing temperatures.
● For children in countries like Ethiopia and Kenya, running is practically the only game in town, the only sport they can afford because it costs practically nothing. In the absence of running kit and shoes, kids there will train and race in their bare feet and their everyday clothes. An attraction to take up the sport of running is that those countries have a running heritage, producing many champions and record-breakers with production line regularity for over half a century now. It offers a route for people to possibly get out of poverty. Add to this that children there have a hardiness that makes them better equipped to handle the hardships and training loads that are necessary to make it as a champion; and then some. Where kids in the developed world will just give up because of the hardship and dedication required, and/or maybe pursue another sport if they don't get fairly early success, the East Africans will keep on trying. Allied to this is that there are so many sports that kids can select from here in the developed world - e.g. tennis, football, rugby, rowing, cycling, sailing, motor-sport, squash, cricket, basketball etc. etc. - that the pool of potential running champions here is consequently greatly reduced.
● Another factor is that the East African runners appear to have a lower proportion of body-fat to those from developed countries. In general, and with the caveat that too low a level of body-fat may in certain circumstances be a potential health risk, less body-fat makes for a more efficient/faster running machine. Allied to this is that the diet of East Africans is perfect for distance runners: - high in carbohydrates, sufficient proteins, and low in saturated fats. In contrast the diet of many in the developed world is somewhat the opposite. Could it perhaps be that the arteries of many runners from developed countries are, by their early twenties, suffering the early stages of narrowing, with the consequent lessening of an ability to pump blood to vital muscles compared to their East African rivals?
● And - though there is no evidence either way, and while personally remaining to be convinced - the East Africans may have a genetic component that has nothing to do with altitude but nevertheless gives them an advantage as distance runners. (A factor which has been noted of East African runners, and may contribute to their success, is that they generally have a thinner lower leg - from the knee to the ankle - than those from other ethnic groups. The fundamentals of the mechanics of motion shows that, all other things being equal, having a lighter lower leg (lever) will confer an advantage in distance running. If this turns out to be the factor that gives East African runners the edge, then no amount of altitude training by non-Africans will ever bridge the performance gap!)

As regards the relative success of those born and usually living at/near sea-level who spend periods undergoing altitude training when compared with those who don't, well the reason is simple. They happen to be the best runners among their sea-level fellows. They are therefore more likely to be selected/choose to go for altitude training. A period spent training at altitude is likely to have little practical effect on their eventual performances; and they may actually show an improvement, as a result of having trained harder than usual during their periods spent at altitude. And of course there is the extra motivation that arises from living and training among other top runners and top coaches.
{And what can't be ruled out is the Placebo Effect - if an athlete strongly believes that altitude training will result in an improvement in performance, then it probably will.}

When it is necessary to altitude train
If the over-riding intention is to win a competition that involves a handicap, then it is prudent to train as much as possible under the conditions of that handicap. For example, a runner who plans to run the steeplechase is extremely unlikely to do well if he/she does not train at jumping the hurdles. Similarly, a runner aiming to win at cross-country is not doing him/herself any favours by just training on the roads. Put simply, training should be adjusted to match the conditions expected to be encountered in the race.
Let's say a race is to be held with the stipulation that the runners must compete with a 5lb weight attached to each leg, then all other things being equal, the runner who has spent a sufficient period training under that condition will beat the runner who didn't. (Of course, if the race is held without weights then the runner who has been training with them, will come second-best to the runner who didn't.)
The same logic applies for a race that is to be held at altitude. For a runner who doesn't live and train at altitude, the conditions at altitude can present a handicap. So therefore, it is necessary to train for that handicap by undergoing a sufficient period of altitude training. However as practically all major competitions are held at/near sea-level, then - with the exception of events such as the African Championships when they are held at venues such as Addis Ababa and Nairobi - altitude training is imprudent and will be counter productive. (In theory there should be no necessity to altitude train for an event to be held at altitude. However in the real world most non altitude-trained runners will misjudge the correct pace to adopt for optimum performance there, usually by adopting and trying to maintain the pace they would normally use at sea-level – a pace that is faster than they can maintain at altitude.)

So, anyone for running-training while wearing 5 lb weights? It makes just about as much sense!

Send comments etc. to: -
N.B. Write the word "altitude" in the subject box.
Comment received on 29th October 2010
Peter in California, U.S.A. writes: - I enjoyed your article on altitude training. I totally agree with your comment “The history of athletics coaching also contains a number of fads and beliefs that were later discredited and abandoned.” The problem is that good scientific methodology isn’t used as a yardstick to evaluate any innovation. I predict the use of compression socks during competition will be added to this list: no-one evens knows at this point if they actually improve performance.
Reply: - Peter you may very well be correct. It would not surprise me.

Comment received on 2nd November 2011
Sam Callan writes: - Michael, you offer no scientific studies to dispute the claim. You might make a stronger case if you would spell Jim Ryun's name correctly and also if you would not give an example of someone training at 25000 feet (an altitude that is pretty close to the "death zone"). How about publishing some peer reviewed papers.

Your blog also ignores the live high, train low studies. perhaps Dr. Chapman will come to rip you a new one, from a purely scientific standpoint.

Altitude training alone will not be successful. Just going to live/train at altitude is only part of the equation. A solid training program still needs to be implemented. No one will dispute that a proper training plan is critical and that poor training/living at altitude will be damaging.

I would also point out that in the Levine, Stray-Gundersen, "live high, train low" paper 3 groups of runners participated. One group lived at trained at about 9000 feet (Deer Valley, UT). A second group lived and did most of their training in Deer Valley, but traveled to Salt Lake City for the higher intensity training. The third group lived and trained in Chula Vista, CA (maybe 100 feet). When they all returned to Dallas, both the UT groups outperformed the Chula Vista group. According to you that should not have happened.
Reply: - Sam, good to hear from you even if you disagree. You are obviously a true believer.
Thanks for alerting me to spelling Jim Ryun's name incorrectly; here in low altitude Ireland we spell that surname a little differently. I will correct it presently.
As regards picking an altitude of 25,000 feet. I was simply giving an example of a height at which - I think you would agree - it can be shown that altitude training would be completely counter-productive. (Of course this raises the question for those who believe in it, at what precise height does altitude training start to become counter-productive? For me it is basically at any altitude above sea-level.)
About peer reviews; well you are one of the first to peer review my piece and so I will include your comments on the website. Just bear in mind that the peer review process is far from infallible.
Finally, as regards the example you supply of an experiment which matched selected groups of altitude trained versus non-altitude trained runners. Well this hardly clinches it. For example how can we be sure that the runners in each group equaled in every way the runners in each of the other groups? And that they all experienced an identical training load? It's just impossible to ensure that requirement.
But, taking the example you give, let's see if I can shake your faith. For that experiment, and others like it, to be considered in any way conclusive then the experimenters would need to have reverse-engineered the process after the runners were tested in Dallas. In other words they should now have sent the CA group to UT, and sent the UT runners to CA for a similar period of training, and then tested them all a second time.
So Sam, which group do you think would now perform better?
A final word to all those who accept the altitude training hypotheses; when you are looking to prove what you believe, you will always find some evidence.

Hit Counter

Blog Archive