Journal Logo

SPECIAL COMMUNICATIONS: Contrasting Perspectives

Hypoxic Training Is Not Beneficial in Elite Athletes


Author Information
Medicine & Science in Sports & Exercise: February 2020 - Volume 52 - Issue 2 - p 519-522
doi: 10.1249/MSS.0000000000002141
  • Free

Hypoxic training can follow the live high–train high (LHTH), live high–train low (LHTL), or live low–train high (LLTH) strategy. Below, we explain why we believe that the available evidence does not justify recommending any of these strategies to athletes.


Although LHTH has been used for decades, there is to our knowledge only one controlled study supporting a positive effect on sport specific performance of endurance athletes: In elite swimmers, 400-m performance was more improved 4 wk after LHTH than after a training camp near sea level (SL), although this effect was not observed in shorter swimming distances or with training effort as a covariate in the analysis (1). In the absence of compelling scientific evidence, one can argue that the popularity of LHTH supports its benefits. Nevertheless, as athletes enrolling in LHTH must have positive expectations, they are certainly susceptible to a placebo effect. Furthermore, observational data suggests that athletes considerably increase training load during LHTH (2). It is hence unclear whether performance improvements following “real-world” LHTH are indeed a consequence of altitude exposure.

Theoretically, LHTH could improve performance through the effects of passive altitude acclimatization and/or by enhancing the effects of training. However, hypoxia does not augment the effects of endurance training (3,4), leaving passive acclimatization as the mechanism. Altitude acclimatization triggers numerous physiological responses, but the one with the largest ergogenic potential is expansion of total red cell volume (RCV). Nevertheless, the 3- to 4-wk exposure to 2000 to 2500 m that characterize a typical LHTH intervention appear insufficient to consistently increase RCV, particularly in elite athletes: A meta-analysis of 66 altitude studies found no significant effect of up to 4 wk at altitudes <3000 m on RCV, and that RCV expansion is even slower if the initial RCV is high, as is the case in endurance athletes (5). However, we acknowledge that the RCV response to both passive altitude exposure (6) and LHTH (1) is subject to great interindividual variability, which might explain why despite seemingly insufficient altitude exposure RCV expansion can occur during LHTH (1,7). The problem is that an individual athlete’s RCV response to altitude is not reproducible and hence impossible to predict (8,9). This unpredictable effect on RCV should not only be weighed against the time and financial investments associated with LHTH but also against the adverse impacts that altitude exposure can exert on performance and/or health: Altitude exposure induces tonic chemoreflex activation that increases muscle sympathetic nerve activity and pulmonary ventilation. These responses promote vasoconstriction and may thereby limit perfusion of the locomotor musculature. Chemoreflex activation transiently persists after return to SL and can hence not only impair training during LHTH but also performance in ensuing competitions if insufficient washout time is provided. Reduced sleep quality and nocturnal periodic breathing at altitude further enhance chemoreflex activation and could impair recovery. Finally, hypoxic pulmonary vasoconstriction increases the already high right ventricular afterload that athletes face during intense training, which may promote adverse remodeling of the pulmonary vasculature and right ventricle (10).


This strategy aims to increase RCV while preventing the hypoxia-induced decrease in training speed/workload that occurs at altitude. However, the frequent descents do not only impose logistic constraints but also reduce the hypoxic dose and hence presumably the chance for RCV expansion. As most of the day, and particularly the night, are still spent at altitude, other adverse effects of altitude exposure likely persist, potentially explaining why performance can remain unchanged or decrease following LHTL even if RCV increases (9). Nevertheless, support comes from the pioneering study by Levine and Stray-Gundersen (7), where 4 wk of LHTL enhanced RCV and improved time trial (TT) performance of runners more than a training camp near SL. What complicates the interpretation is that throughout the intervention training effort was approximately 50% larger in the LHTL group, which may, at least in part, explain the larger increases in RCV and TT performance. Three further controlled LHTL studies using natural altitude are available. In one by our opponents, changes in 3K running performance of triathletes following LHTL did not significantly exceed those in the control group despite approximately 30% higher training effort (11). In the absence of a statistically significant difference between the groups, the authors used magnitude-based inference to support a positive effect of LHTL, but this approach was heavily criticized (12) and hence recently banned from this journal. In the second, a positive effect of LHTL (specifically: live high and train high and low) on TT performance of elite swimmers was observed (1). Also here, training effort was higher in the LHTL than in the control group, although the effect of LHTL persisted with training effort as covariate. The mechanism underlying the performance gains remained, however, unclear, as LHTL did not increase RCV. In the third study, training effort was similar between the LHTL and control group, leading to almost identical increases in RCV and TT performance (13).

To avoid shuttling between high and low altitudes, athletes can use normobaric hypoxia and there are more studies that investigated this form of LHTL. In some, LHTL improved V˙O2max or exercise economy, but such adaptations are only relevant if they translate into better performance in a sport specific task, that is, a TT. Although a positive effect of normobaric LHTL on TT performance was observed in two very recent studies (14,15), this is an uncommon finding: Last year, a review reported that out of nine controlled studies none could demonstrate that normobaric LHTL improves TT performance more than SL training, at least with conventional statistics (16). Even if the numerical outcomes rather than the P values of the reviewed studies are considered the picture does not improve. The TT improvements after normobaric LHTL numerically exceeded those in the control group in only four out of nine studies and in two of those the training effort was approximately 40% higher in the LHTL group (9,11). Interestingly, when one of the latter studies conducted the analysis with training effort as covariate they found a negative effect of LHTL on TT performance (9), again, potentially reflecting the adverse effects of altitude exposure. Although these results are discouraging enough, it must be emphasized that neither the studies reporting a significant effect of LHTL on TT performance (14,15), nor those reporting a numerically larger improvement than in the control group, were placebo-controlled. Although normobaric LHTL allows double-blinded group allocation, only two studies have taken advantage of this (6,17); neither of them observed a positive effect of LHTL on TT performance.

A final but important concern regarding both LHTL and LHTH is that, as the effect on RCV, the effect on performance is highly variable. A retrospective analysis of the studies of the Levine group illustrates that out of 39 athletes exposed to these forms of altitude training, 17 experienced a clear improvement in TT performance, whereas 7 experienced only minimal improvements and 15, a deterioration (18). Even for individual athletes, the effect can qualitatively differ from one intervention to the next: When eight highly trained runners completed two identical LHTL interventions, only two experienced a TT improvement after both interventions, whereas four experienced an improvement after one, and a deterioration after the other intervention (9). Such unpredictable outcomes make the use of altitude training before competitions a risky gamble!


Originally, LLTH meant that athletes perform endurance training in hypoxia while spending the rest of the day in normoxia, but this produces no additional benefits (3,4). However, our opponents have popularized a novel approach, where athletes perform repeated sprints in hypoxia (RSH), interspersed with incomplete recovery. They postulate that RSH enhances repeated sprint ability (RSA) more than corresponding training in normoxia (RSN). Although some studies support this, others (19–23), including two by our opponents themselves (22,23), as well as the only one with a crossover-design (21), do not. Sometimes, RSH had other positive effects, for example, on primarily aerobic tasks (19,22) or agility (23), but also these occurred inconsistently. Recently, our opponents published a meta-analysis concluding on a barely significant P value (P = 0.05) that RSH increases RSA more than RSN (24). However, the largest standardized mean difference between RSH and RSN illustrated in figure 2 of that article (extracted from the study by Kasai et al.) seems to be calculated based on standard errors, rather than standard deviations (as stated in the methods). If this mistake is corrected, the meta-analysis finds no significant difference between RSH and RSN.

Also here, the inconsistent effects of RSH should not only be weighed against the costs of a hypoxic training facility but also against potential health risks: A sprint markedly increases systolic pressure (25), and this effect is augmented during repeated sprints presumably as metabolite accumulation increases metaboreflex activation. In normoxia, the pressure increase is met by cerebral vasoconstriction that protects the brain against overperfusion, but hypoxia overrules this effect and causes vasodilation to preserve cerebral O2 delivery (25). High pressures combined with increased vessel diameters enhance the mechanic stress against the vessel walls, and it is unclear whether frequent exposure to this stress is tolerated without structural damage (25).


In the following, we point out what we perceive as flaws in Millet and Brocherie’s (26) argumentation:

  1. Maladaptive effects of hypoxia: Our opponents see no evidence that the maladaptive effects of hypoxia outweigh the benefits of hypoxic training (HT). They do not acknowledge that on an individual athlete’s level reductions in performance following LHTH/LHTL are common (18). They ignore a study reporting that LHTL had a negative effect on performance if differences in training effort were considered (9) or studies, in which LHTH (7) or LHTL (9,11,27) facilitated expansion of total hemoglobin mass (Hbmass) but not the increase in performance that results if Hbmass expansion is induced by, for example, blood doping. Finally, how do they explain the unblinded LHTH/LHTL studies that found no positive effect on performance despite the placebo effect that must have arisen from the athletes’ positive expectations?
    • Our opponents also show little concern regarding adverse health effects, which is surprising given that the moderate hypoxia used for HT is clearly sufficient to induce, for example, nocturnal periodic breathing with intermittent hypoxemia and sleep state disruption (28), raised heart rate, and arterial pressure indicative of sympathoexcitation (29) and pulmonary vasoconstriction (30). With regard to RSH our concern regarding cerebral overperfusion is amplified if, as promoted by our opponents (31), hypoxia is created by hypoventilation, where the resulting hypercapnia represents an additional mechanism for cerebral vasodilation.
    • Clearly, the onus is squarely on the promoters of HT to objectively investigate the significance of these concerns to the performance and health of athletes in their care. Unfortunately, they have so far either ignored them or only studied them with superficial methodologies as illustrated by, for example, a study of sleep disordered breathing during LHTL that merely assessed breathing frequency and lacked the EEG required for sleep monitoring (32), or the evaluation of LHTL-induced sympathoexcitation based on heart rate variability (15).
  2. Variability: Our opponents ignore that LHTL and LHTH produce highly variable and unpredictable performance outcomes. As with the maladaptive effects, proponents of these interventions have failed to even attempt to phenotype athletes to reduce this unacceptable level of inefficiency.
  3. Athletes perception versus evidence: Our opponents’ question whether we would trust a physician not asking how we felt after a treatment is justified. However, would they trust a caregiver prescribing an expensive, time-consuming treatment that has never been proven superior to a placebo, and the safety of which remains uninvestigated despite indications for adverse effects?
  4. Studies supporting HT: This paragraph in our opponents’ statement refers to evidence that LHTH/LHTL increase Hbmass. However, the ultimate goal of these interventions is not Hbmass expansion, but to improve performance. The fact that they do not mention a single study demonstrating that LHTH/LHTL improve endurance performance more than SL training speaks for itself and leaves nothing for us to debate.
    • Regarding RSH, we acknowledge the placebo-controlled studies supporting a positive effect on RSA. It should, however, be considered that improvements in RSA are unlikely to translate as directly into the performance of team sports as, for example, improvements in endurance capacity translate into race performance. It hence remains to be demonstrated that the small improvements in RSA that can result from RSH are worth the logistic effort.
  5. Evidence that HT is not beneficial: In this paragraph, our opponents ignore that out of two controlled LHTH studies one found no effect on performance (7), whereas the small effect observed in the other vanished when corrected for training effort (1). They also ignore that the few studies reporting a positive effect of LHTL on performance are contrasted by a larger number of studies that have failed to confirm this effect (16). It is unclear why our opponents do not consider any of this as evidence that LHTH/LHTL are not beneficial.
    • Our opponents further declare that there is little doubt that RSH improves RSA. This is a bold statement considering that even some of their own studies (22,23) and their meta-analysis (24) (if data is inserted correctly) have not detected this effect. Their claim that out of 25 RSH studies only two have found no positive effects must be interpreted with caution, since they consider each study showing any beneficial effect as positive: As a result, studies with completely contradictory outcomes (e.g., reporting that RSH improves RSA but not aerobic performance [33], or the exact opposite [19]) are claimed to support each other.
  6. Unjustified conclusions: Our opponents’ conclusions exclusively contain arguments that we find invalid. They state that most maladaptive effects of hypoxia are irrelevant for the altitude and duration used in HT, although nobody has seriously investigated this. They argue that altitude-induced improvements in oxygen transport capacity occur in most athletes, ignoring that this is irrelevant if it does not improve performance. Finally, they declare without specification that the robustness of most studies reporting no effect of particularly LHTL is questionable, whereas the few studies supporting their own view all lack placebo control, mostly rely on a heavily criticized statistical approach and are often limited by a larger training effort in the LHTL than in the control group.


We propose that LHTH and LHTL have been vigorously promoted for many years without solid evidence for their justification. What has been revealed is the extreme interindividual and intraindividual variabilities in performance outcomes suggesting that these practices promise little better than a coin flip—even worse, they steal valuable time away from legitimate training practices in normoxia. Supporters of LHTH/LHTL have often been negligent in their experimental designs, have not conducted legitimate investigations into the potentially adverse effects and have perpetuated a highly inefficient practice by ignoring athlete phenotypes as a basis for individualizing treatments. With regard to RSH, there is evidence for a beneficial effect on RSA, which is, however, far less consistent than propagated by our opponents, and the relevance of which for competitive performance is currently unclear.


1. Rodriguez FA, Iglesias X, Feriche B, et al. Altitude training in elite swimmers for sea level performance (altitude project). Med Sci Sports Exerc. 2015;47(9):1965–78.
2. Sharma AP, Saunders PU, Garvican-Lewis LA, et al. Training Quantification and Periodization during Live High Train High at 2100 M in Elite Runners: An Observational Cohort Case Study. J Sports Sci Med. 2018;17(4):607–16.
3. Robach P, Bonne T, Flück D, et al. Hypoxic training: effect on mitochondrial function and aerobic performance in hypoxia. Med Sci Sports Exerc. 2014;46(10):1936–45.
4. Hoppeler H, Klossner S, Vogt M. Training in hypoxia and its effects on skeletal muscle tissue. Scand J Med Sci Sports. 2008;18(1 Suppl):38–49.
5. Siebenmann C, Hug M, Keiser S, et al. Hypovolemia explains the reduced stroke volume at altitude. Physiol Rep. 2013;1(5):e00094.
6. Siebenmann C, Robach P, Jacobs RA, et al. “Live high-train low” using normobaric hypoxia: a double-blinded, placebo-controlled study. J Appl Physiol (1985). 2012;112(1):106–17.
7. Levine BD, Stray-Gundersen J. “Living high-training low”: effect of moderate-altitude acclimatization with low-altitude training on performance. J Appl Physiol (1985). 1997;83(1):102–12.
8. McLean BD, Buttifant D, Gore CJ, White K, Kemp J. Year-to-year variability in haemoglobin mass response to two altitude training camps. Br J Sports Med. 2013;47(1 Suppl):i51–8.
9. Robertson EY, Saunders PU, Pyne DB, Aughey RJ, Anson JM, Gore CJ. Reproducibility of performance changes to simulated live high/train low altitude. Med Sci Sports Exerc. 2010;42(2):394–401.
10. La Gerche A, Rakhit DJ, Claessen G. Exercise and the right ventricle: a potential Achilles’ heel. Cardiovasc Res. 2017;113(12):1499–508.
11. Hauser A, Schmitt L, Troesch S, et al. Similar Hemoglobin Mass Response in Hypobaric and Normobaric Hypoxia in Athletes. Med Sci Sports Exerc. 2016;48(4):734–41.
12. Sainani KL. The problem with “magnitude-based inference.” Med Sci Sports Exerc. 2018;50(10):2166–76.
13. Robach P, Hansen J, Pichon A, et al. Hypobaric live high-train low does not improve aerobic performance more than live low-train low in cross-country skiers. Scand J Med Sci Sports. 2018;28(6):1636–52.
14. Czuba M, Fidos-Czuba O, Płoszczyca K, Zając A, Langfort J. Comparison of the effect of intermittent hypoxic training vs. the live high, train low strategy on aerobic capacity and sports performance in cyclists in normoxia. Biol Sport. 2018;35(1):39–48.
15. Schmitt L, Willis SJ, Fardel A, Coulmy N, Millet GP. Live high-train low guided by daily heart rate variability in elite Nordic-skiers. Eur J Appl Physiol. 2018;118(2):419–28.
16. Bejder J, Nordsborg NB. Specificity of “live high-train low” altitude training on exercise performance. Exerc Sport Sci Rev. 2018;46(2):129–36.
17. Bejder J, Andersen AB, Buchardt R, Larsson TH, Olsen NV, Nordsborg NB. Endurance, aerobic high-intensity, and repeated sprint cycling performance is unaffected by normobaric “live high-train low”: a double-blind placebo-controlled cross-over study. Eur J Appl Physiol. 2017;117(5):979–88.
18. Chapman RF, Stray-Gundersen J, Levine BD. Individual variation in response to altitude training. J Appl Physiol (1985). 1998;85(4):1448–56.
19. Galvin HM, Cooke K, Sumners DP, Mileva KN, Bowtell JL. Repeated sprint training in normobaric hypoxia. Br J Sports Med. 2013;47(1 Suppl):i74–9.
20. Goods PS, Dawson B, Landers GJ, Gore CJ, Peeling P. No additional benefit of repeat-sprint training in hypoxia than in normoxia on sea-level repeat-sprint ability. J Sports Sci Med. 2015;14(3):681–8.
21. Montero D, Lundby C. No improved performance with repeated-sprint training in hypoxia versus normoxia: a double-blind and crossover study. Int J Sports Physiol Perform. 2017;12(2):161–7.
22. Brechbuhl C, Brocherie F, Millet GP, Schmitt L. Effects of repeated-sprint training in hypoxia on tennis-specific performance in well-trained players. Sports Med Int Open. 2018;2(5):E123–32.
23. Brocherie F, Girard O, Faiss R, Millet GP. High-intensity intermittent training in hypoxia: a double-blinded, placebo-controlled field study in youth football players. J Strength Cond Res. 2015;29(1):226–37.
24. Brocherie F, Girard O, Faiss R, Millet GP. Effects of repeated-sprint training in hypoxia on sea-level performance: a meta-analysis. Sports Med. 2017;47(8):1651–60.
25. Curtelin D, Morales-Alamo D, Torres-Peralta R, et al. Cerebral blood flow, frontal lobe oxygenation and intra-arterial blood pressure during sprint exercise in normoxia and severe acute hypoxia in humans. J Cereb Blood Flow Metab. 2018;38(1):136–50.
26. Millet GP, Brocherie F. Hypoxic training is beneficial in elite athletes. Med Sci Sport Exerc. 2020;52(2):515–8.
27. Robach P, Schmitt L, Brugniaux JV, et al. Living high-training low: effect on erythropoiesis and aerobic performance in highly-trained swimmers. Eur J Appl Physiol. 2006;96(4):423–33.
28. Latshang TD, Lo Cascio CM, Stowhas AC, et al. Are nocturnal breathing, sleep, and cognitive performance impaired at moderate altitude (1,630–2,590 m)? Sleep. 2013;36(12):1969–76.
29. Stowhas AC, Latshang TD, Lo Cascio CM, et al. Effects of acute exposure to moderate altitude on vascular function, metabolism and systemic inflammation. PLoS One. 2013;8(8):e70081.
30. Gabry AL, Ledoux X, Mozziconacci M, Martin C. High-altitude pulmonary edema at moderate altitude (< 2,400 m; 7,870 feet): a series of 52 patients. Chest. 2003;123(1):49–53.
31. Trincat L, Woorons X, Millet GP. Repeated-sprint training in hypoxia induced by voluntary hypoventilation in swimming. Int J Sports Physiol Perform. 2017;12(3):329–35.
32. Saugy JJ, Schmitt L, Fallet S, et al. Sleep disordered breathing during live high-train low in normobaric versus hypobaric hypoxia. High Alt Med Biol. 2016;17(3):233–8.
33. Gatterer H, Philippe M, Menz V, Mosbach F, Faulhaber M, Burtscher M. Shuttle-run sprint training in hypoxia for youth elite soccer players: a pilot study. J Sports Sci Med. 2014;13(4):731–5.
Copyright © 2019 by the American College of Sports Medicine