Salgado, R. M., Parker, D. L., & Quintana, R. (20009). The effects of hypoxic manipulation on VO2max and sea-level performance: A meta-analysis. ACSM 56th Annual Meeting, Seattle, Washington. Presentation number 2789.

The purpose of this meta-analysis was to identify the effects of hypoxic manipulation versus normoxic training on VO2max and sea-level performance. An online database and the reference lists of peer-reviewed journal articles were used to find pertinent journal articles. A total of 74 articles were found, of which 32 met the inclusion criteria and 16 of the 32 were reviewed for this preliminary study. The analyzed journal articles had to meet the following criteria: 1) use a control group in the design, and 2) report means and standard deviations with the results. The dependent variables were VO2max and exercise performance. Exercise performance was defined as time trial performance, peak power during a GXT, or total work capacity. The independent variable was hypoxic manipulations and included traditional altitude training, live-high and train-low, live-low and train-high, and intermittent hypoxic exposure. There were 43 extracted data points: VO2max (N = 19 control and N = 24 hypoxic manipulation) and performance (N = 20 control and N = 23 hypoxic manipulation). Effect size was calculated using Cohen’s d: ES= (posttest mean-pretest mean)/pretest SD.

The mean effect sizes for control and hypoxic manipulation were .28 and .33 for VO2max and .29 and .50 for performance, respectively. There was no difference in effect size for VO2max but a statistically higher effect size was observed for performance in hypoxic manipulation group.

Implication. Hypoxic manipulation was found to be related to performance improvements but not VO2max when a meta-analysis was performed on a number of studies. [The criteria for evaluating original studies might not have been sufficiently stringent to disqualify confounded studies. The combining of very different and often contradictory experiences into the "hypoxic manipulation" category is dubious, because varying effects and response characteristics for different manipulations have been noted in the research literature. As well, within a subset of hypoxic manipulations, the control of influential variables has often been absent. For example, control groups should experience living in hypoxic tents the same as experimental groups with the only difference being the level of hypoxia in the tent atmosphere. That is very different to the common in-tent and not-in-tent studies. As concluded in another meta-analysis, there is a difference in response to hypoxic manipulation between trained elite athletes and lesser individuals, with the latter groups showing a response when the former does not. The meta-analysis here could have combined the two groups biasing the differences to yield an overall effect (which might have only been exhibited by a subset of Ss). The massing of studies with different experimental variables and labeling them into a similar category is a very questionable practice.]

Return to Table of Contents for this issue.