Daniels, J., & Oldridge, N. (1970). The effects of alternate exposure to altitude and sea level on world-class middle distance runners. Medicine and Science in Sports, 2(3), 107-112.

This is the paper that is often used to support the argument that training at altitude enhances sea-level performance, particularly in distance athletes. The attractiveness of the study is that one of its subjects was Jim Ryan, a world-record holder in the late 1960s.

The intent of the study was to evaluate whether or not brief returns to sea-level during altitude adaptation were detrimental to acclimatization. Six subjects, including Jim Ryan, alternated training at altitude and training/competing at sea-level. The design entailed the following:

  1. Considerable preparatory sea-level training (test Sea-level 1a/b).
  2. Two weeks at altitude preparing for the National AAU Championships to be held at Bakersfield (test Altitude 1).
  3. Five days at sea-level and competing in National AAU Championships (test Sea-level 2).
  4. Two weeks at altitude (test Altitude 3).
  5. Five days at sea-level to compete in the USA versus the Commonwealth Track and Field meet held early in July, 1967 in Los Angeles (test Sea-level 3).
  6. One week at altitude (test Altitude 5).
  7. Eleven days at sea-level (test Sea-level Post).
  8. Some subjects returned to altitude for a further week and then to sea-level.

This is a "pre-control" design study, and is therefore unacceptable for inferring causes to observed phenomena. The authors claim that training at altitude remained "normal" and was equivalent to sea-level training.

"Normal sea-level training was continued at altitude with the addition of various exposures of several hours each to altitudes up to 3300 meters. Hard training sessions were included from the outset . . ." (p. 108).

"In the present study rigorous training, equal in intensity to normal sea-level training in preparation for important international competitions, was continued throughout the study" (p. 111).

These descriptions lead one to assume that training did not change from the sea-level to altitude situation. Such an assumption would be incorrect. The first altitude exposure centered on "sharpening-up" for the ensuing National AAU Championships, and can be presumed to constitute a taper. The second altitude session included recovery from the AAU Championships plus preparations for the Los Angeles dual meet between the USA and a Commonwealth team. Added to those quality/quantity reductions, travel to and from Alamosa (2,300 m) also forced days of inactivity (unavoidable recovery). It is generally recognized that sea-level equivalent quality/quantity work cannot be performed on initial exposure, if ever, to altitude. This latter feature is supported by the reported data where performances and physiological capacities were markedly reduced at altitude (e.g., VO2max decreased by 14%). Therein lies a major weakness with the study, the altitude training was not sea-level equivalent. It constituted a disguised or unintended taper. Consequently, any performance changes at sea-level could be attributed to the improved rested state of the athletes just as logically as could be attributed to altitude. "Unintended" tapers (recovery periods) are one of the commonest contaminants of altitude/sea-level performance research. The failure to control for consistent quality and quantity of training at both sea-level and altitude means that any performance changes could be attributed to the change in training program.

A second weakness of the study is the reliability of testing procedures. The protocols for repeating physiological assessments were altered. Specifically, those alterations were as follows:

"In the pre-altitude tests they ran at speeds equal to the mean velocity of several "control" performances; the post-altitude pace was closer to their best effort" (p. 111).

The change of test speed could decrease the reliability of testing since athletes could alter their approach to the task in terms of effort application and efficiency. If procedures are unreliable, then there should be some notable variance in repeated assessments. As an example, in altitude tests (1, 3, and 5) of maximum aerobic power, subject TR changed by approximately 5 ml/kg/min between each test over a period of six weeks (see Table 3), changes that would be unreasonable to expect in an already physically trained athlete. Not only did VO2max values change in the subjects, but the trends of the changes also varied. A reanalysis of the data in Table 3 in the article is included below. It shows anything but a consistent effect across subjects of the experience in the study.

Test Comparison                            Nature of Change
                               No Change         Worse           Improved

Altitude 1 - Altitude 3           0                1                5
Altitude 1 - Altitude 5           2                1                3
Altitude 3 - Altitude 5           2                3                1
Sea-level 1 - Sea-level 2         4                0                2
Sea-level 1 - Sea-level Post      4                0                2

With regard to VO2max adaptation at altitude, some athletes improved from the first to the second exposure but then regressed from the second to the third exposure. There is no justification for concluding that there was a general trend of adaptation to either sea-level or altitude conditions in terms of maximum aerobic power over the term of the study.

One of the attractive features of the study was that it contained Jim Ryan and included his world record runs after coming down from altitude. At Bakersfield he broke the WR for the mile and then at the next sea-level exposure he broke the WR for 1500 m. As well as his WRs, the other athletes also performed a series of personal best performances. One would expect good performances to occur at national championships, for those usually are the focus of a year's training. It is often asserted, and this is concluded in this study, that altitude training contributed to the performance enhancement. However, this reanalysis concludes that the good results were more likely due to the nature of training at altitude rather than any altitude adaptation "benefit." The data do show that performances at altitude did improve over initial time-trial results. That is something to be expected from adaptation.

Something of note that was not discussed fully, was that on return to sea-level at the end of the investigation, the economy of running had worsened. Post-altitude submaximal work required more oxygen than pre-altitude work. This suggests that the specificity of altitude training eventually interfered with the specificity of sea-level training. That interference gradually developed and was only evident after sufficient exposure.

Another weakness in the study is the comparison of performances and their designation as the "control" value. Performance comparisons were made to the average of three race times performed in the preparatory training phase. During the study, one national championship and one international meet were attended. It would be expected that elite athletes would perform well in the major meets, certainly to a better level than in the pre-study competitions. The use of the word "control" for the pre-study performances is misleading. That level of performance can be used for comparative purposes, but is not a control in the classical experimental sense.

The absence of adequate intraindividual controls or an intergroup control group does not allow inferences to be made about causal effects. It is wrong to infer causality to any factor if it cannot be shown that it was the only factor that varied over the course of the study.

VEmax was elevated out of proportion to VO2max, a common observation in altitude adaptation. This may be explained as a consequence of chronic hyperventilation at altitude which leads to a lowering of plasma bicarbonate levels for the restoration of blood acid-base balance disturbed by developing hypocapnia. A lowered buffering capacity of the blood may explain partially why the metabolic carbon dioxide production furnishes greater stimulus to the respiratory center at altitude. This "hypersensitivity" of the respiratory center recedes slowly after return to normal sea-level conditions. During this transient period, athletes breathe more air for any given work intensity than at sea-level prior to altitude adaptation. The additional work involved in moving this greater volume requires more oxygen which is provided at the expense of oxygen demands of the muscles used in the sporting exercise (p. 110). A factor such as that most probably would produce an effect upon high-level performance. Thus, ventilation altered by altitude adaptation may reduce performance capacity at sea-level in elite athletes.

Intermittent exposures to sea-level while undergoing altitude adaptation could promote the maintenance of normal competitive technique and muscle power the loss of which has been reported by Balke, Daniels, and Faulkner (1967) as a result of prolonged altitude training. [Balke, B., Daniels, J., & Faulkner, J. A. (1967). Training for maximum performance at altitude. In R. Margaria (Ed.), Exercise at altitude. Excerpta Medica Foundation.]

It could be hypothesized that altitude training would benefit subsequent anaerobic sea-level performance as athletes withstand more discomfort than usual. This could reflect a greater utilization of the anaerobic capacity which may be beneficial to sea-level anaerobic performances. This is only a possibility for it was not investigated in this study.

The individuality of responses reported is particularly noteworthy. Some adapted well to altitude, others showed little reactivity, and one showed a negative reaction. Before any assumptions are made about altitude adaptation, the responsiveness to altitude stress for each individual should be known.

Conclusion. This study is flawed and does not substantiate its own conclusions. It is not possible to assert that altitude training shows considerable promise for benefiting sea-level running performance. However, altitude training does promote improvements in altitude performance and brief visits to sea-level do not appear to disrupt that adaptation process. The fact that other researchers have not been able to develop conclusions of altitude training benefits supports the results of this reassessment.

Return to Table of Contents for this issue.