Altitude training: no impact on running economy in 13 pooled trials
Endurance Track, AI-assisted writing ·
A 2026 systematic review with meta-analysis asked whether training in low-oxygen (hypoxic) conditions changes running economy and performance in competitive middle- and long-distance runners, compared with training at normal oxygen levels. It reviewed 16 randomized trials and pooled 13 of them, with 182 runners for running economy, 95 for VO2max, 134 for time trial and 81 for time to exhaustion[1]. The abstract reports no impact on running economy, VO2max or time trial, and altitude training better than normal-oxygen training for time to exhaustion[1].
Key findings
| Outcome | Standardized Mean Difference | 95% Confidence Interval | P-value | I-squared |
|---|---|---|---|---|
| Running economy | -0.2 | -0.58 to 0.19 | 0.31 | 33% |
| VO2max | 0.33 | -0.72 to 1.37 | 0.54 | 82% |
| Time trial | -0.34 | -0.80 to 0.13 | 0.16 | 42% |
| Time to exhaustion | 0.89 | 0.12 to 1.66 | 0.02 | 56% |
Running economy: the standardized mean difference was -0.2 (95% confidence interval -0.58 to 0.19; p = 0.31), so no impact was shown. Because the interval includes zero, a small benefit or a small harm is not ruled out[1].
VO2max: 0.33 (95% confidence interval -0.72 to 1.37; p = 0.54), no impact, with results that varied a lot between trials (I-squared 82%) [1]. Time trial: -0.34 (95% confidence interval -0.80 to 0.13; p = 0.16), no impact [1].
Time to exhaustion: altitude training was better than normal-oxygen training, 0.89 (95% confidence interval 0.12 to 1.66; p = 0.02) [1]. The runners were tested under normal-oxygen conditions[1].
Conclusion
The authors conclude that the available randomized trials do not support an improvement in running economy, VO2max or time trial from altitude training compared with normal-oxygen training, and that future trials should specifically study its effect on running economy in elite athletes[1].
What this does not show
The abstract does not say which altitudes, oxygen levels, doses or training programmes the trials used, so it cannot guide any particular camp or device[1]. The groups were small (81 to 182 runners per outcome) and mostly men (4% to 19% women)[1]. The abstract does not report race results, injuries or health[1].