Altitude Training for Elite Endurance Performance

A 2012 Update

Abstract

Altitude training is commonly used by endurance athletes and coaches in pursuit of enhancement of performance on return to sea level. The purpose of the current review article was to update and evaluate recent literature relevant to the practical application of altitude training for endurance athletes. Consequently, the literature can be considered in either of two categories: performance-led investigations or mechanistic advancements/insights. Each section discusses the relevant literature and proposes future directions where appropriate.

Introduction

Altitude training is a common practice undertaken by endurance athletes in pursuit of an enhancement of subsequent sea level (or altitude) performance. There are several forms of altitude training that have been previously described (^3,26,36,43): live high-train high (LHTH), live high-train low (LHTL), live moderate-train moderate (LMTM), and live low-train high (LLTH). In a 2009 meta-analysis of sea level performance after altitude exposure (³), it was found that in elite athletes, enhancement of maximal endurance power output was only possible (i.e., ≥50% chance of enhancement) with natural LHTL (4.0%; 90% confidence limits ±3.7%) but unclear (i.e., >5% chance of increase and >5% chance of decrease) with LHTH (1.6%; 90% confidence limits ±2.7%) and LLTH (0.6%; 90% confidence limits ±2.0%). Interestingly, those authors (³) reported that, by enhancing protocols with appropriate manipulation of study characteristics, such as the degree of altitude, number of days of exposure, and the test day (after altitude exposure), there may be clear positive effects with LHTH (5.2%; 90% confidence limits ±4.1%) and LHTL (4.3%; 90% confidence limits ±4.1%) in elite athletes. Although merely statistical analysis, such manipulations do extend the implication that refinement of protocol of altitude exposure can enhance the performance gain experienced by athletes. It also serves to highlight the need to pay close attention to such factors when planning altitude exposure within an athlete’s training program, particularly when combining altitude training methods, such as time, altitude, and training program (for a review, see Millet et al. (²⁶)). Nevertheless, such practices/manipulations do require further scientific examination. The scope of the current review, therefore, was to examine new findings that are relevant to altitude training, in particular from an applied perspective. Where appropriate, future directions are discussed.

Literature Review

The following sections address performance-led investigations and mechanistic advancements in sequence and discuss individual study findings, relevance, and possible implications for athletes undertaking altitude training.

Performance-Led Investigations

Ultimately, any hypoxic training intervention undertaken by athletes must provide performance enhancements, either in the short or long term, for the investment to be worthwhile. Table 1A illustrates relevant performance outcome driven studies. A common protocol used by coaches and athletes is to use a combination of altitude exposures to gain a competitive edge (²⁶). Robertson et al. (³⁵) examined an LHTL-and-train-high approach in well-trained endurance runners (Table 1A). They found that LHTL and train high significantly increased maximal oxygen uptake (V˙O_2max) (4.8% ± 2.8%, mean ± SD), total hemoglobin mass (Hb_mass) (3.6% ± 2.4%), and 3-km time trial performance time immediately after altitude (−1.1% ± 1.0%) but not 2 wk after (−0.4% ± 1.3%). The train high group had small changes in performance immediately after altitude (−0.1% ± 1.0%) and 2 wk after (−0.6% ± 1.1%) but did display significant improvements in V˙O_2max (2.2% ± 1.8%). The authors compared these results (³⁵) to a further study carried out by the same research group (³⁴) that examined a LHTL protocol as the subjects were of a similar caliber. This essentially allowed comparison of LHTL, LHTL with train high, train high alone, and a control protocol. Robertson et al. (³⁴) found that well-trained endurance runners (Table 1A) who completed two repeat 3-wk LHTL blocks had reproducible improvements in V˙O_2max (2.1% ± 2.1% and 2.1% ± 3.9%) and Hb_mass (2.8% ± 2.1% and 2.7% ± 1.8%) after each block compared to trivial, but consistent changes observed in the control group for V˙O_2max (0.9% ± 2.8% and 0.7% ± 3.1%) and Hb_mass (1.4% ± 2.7% and −1.5% ± 1.5%). Frese and Friedmann-Bette (14) also found that repeat exposures of moderate altitude (Table 1A) augmented Hb_mass during the course of two altitude camps, although no performance measures were identified. Robertson et al. (34) noted that 4.5-km time trial performance changes were less consistent with the LHTL group, substantially faster after the first training block, but marginally slower after the second block (−1.4% ± 1.1% and 0.7% ± 1.3%). In contrast, the control group demonstrated only trivial changes in performance after both training blocks (0.5% ± 1.5% and -−0.7% ± 0.8%). Examined together, these investigations (^34,35) suggest that LHTL with train high elicits a substantially greater increase in V˙O_2max and Hb_mass than 3 wk of LHTL or TH alone, yet this is not necessarily reflected in proportionally greater improvements in time trial performance. Furthermore, repeat exposure seems to consolidate observed physiological improvements in V˙O_2max and Hb_mass, but this too does not transfer into proportional improvements in time trial performances in running. In swimming, comparable results have been reported with Gough et al. (²⁰), observing a similar increase in Hb_mass during LHTH and LHTL altitude camps (Table 1A) relative to a control group (∼4% increase) alongside a reduction in competitive swim performance measured on days 1 and 7 (LHTH = 1.4% ± 1.3% and LHTL = 1.6% ± 1.6%) after hypoxic exposure with a similar level of performance measured on days 14 and 28 after altitude exposure compared to a control group.

Table 1: Characteristics of study groups sorted by study area.

These findings (^20,34,35) are unexpected given that it has been shown repeatedly that erythrocythemia (induced either by red cell infusion or erythropoietin (EPO) administration) increases both V˙O_2max and endurance performance in direct proportion to increases in arterial oxygen content (^2,4,12,45). The corollary also has been demonstrated via blood loss or partial blocking of hemoglobin (^5,13,23). Garvican et al. (¹⁸) conducted an elegant study to determine the contribution of Hb_mass and nonhematological effects to cycling performance after LHTL by preventing Hb_mass increases in an experimental group (Clamp) via periodic phlebotomy (Table 1A) compared to a control group (Response). After 26 nights of LHTL, Hb_mass increased significantly in the Response group (5.5% ± 2.9%), but no increase was observed in the Clamp group (−0.4% ± 0.6%) as result of repeated phlebotomy. As expected, V˙O_2peak increased in the Response group (3.5% ± 2.3%) but not in the Clamp group (0.3% ± 2.6%). What is most interesting is that, despite no increase in Hb_mass and V˙O_2peak in the Clamp group compared to the Response group, there was a similar increase in cycling maximal 4 min effort (MMP_4min) and no difference detectable between groups (P = 0.58). There was, however, a substantial difference between groups in a time-to-exhaustion test at peak power output (T_lim) with the Clamp group significantly worse than the Response group (−37.6%; 90% CL = −58.9% to −5.0%, P = 0.07). These data are noteworthy because they support the contention that alternative, nonhematological adaptive mechanisms exist as a result of altitude exposure, which contribute to enhanced performance (19). More specifically, Garvican et al. (18) postulate that anaerobic metabolism is enhanced as a result of hypoxic training, but the exact mechanisms remain to be elucidated. Given the study design (i.e., T_lim was conducted 10 min after MMP_4min), the findings also support a role for an increase in Hb_mass in improving aerobic metabolism, as shown by recovery processes in repeat maximal cycling as T_lim was substantially reduced in the Clamp group compared to the Response group. Aside from the acute mechanistic implications of these results (i.e., that an increase in Hb_mass may have a significant role in recovery processes), these results may suggest that an increase in Hb_mass could have a bearing on long-term athlete development because athletes would presumably be able to complete more training and/or a greater quality of training during a training program, although this is speculative.

Robertson et al. (³³) evaluated a long-term coach-prescribed altitude training program for the preparation of 18 elite swimmers for National Championships (NC) and Commonwealth Games (CG). Swimmers were monitored during a 21-wk period and were assigned to an altitude training group or control group by the coaches (Table 1A). The altitude group participated in three to four 2-wk altitude blocks, based on selection for the relative competitions. The LHTL altitude training exposures lasted 9 to 10 h per night at a simulated height of 2,600 m and the LMTM consisted of living and training at a natural height of 1,350 m. Hb_mass was measured throughout, and performance was assessed using the official race times from the competitions. After three to four altitude training blocks, there was a small increase in Hb_mass of ∼1%; however, as with previous studies, there was a large individual variation indicating that some swimmers may benefit from altitude training while some may not. Overall, the altitude group did not swim substantially faster in the NC compared with the previous year and swam slower at the CG 6 wk later. The control group did swim substantially faster at the NC but also were slower at the CG. There were no substantial differences between the groups at either the NC or CG. For this particular group of athletes, it could be proposed that further refinement of altitude training methodology is required owing to the absence of performance improvement. This may be due to the period of altitude exposure being too short, the LMTM altitude not being high enough, and/or the LHTL exposure of 9 to 10 h per night for 10 nights being insufficient to elicit significant physiological changes (³⁶).

In contrast, Siebenmann et al. (⁴¹) investigated the effect of LHTL in a double-blinded, placebo-controlled design, in 16 national-level endurance athletes with a hypothetically (³⁶) sufficient hypoxic dose (Table 1A). The main finding was that LHTL in highly trained endurance cyclists did not affect Hb_mass, 26-km time trial performance, V˙O_2max, or exercise economy in normoxia. This may be explained by less-than-expected hypoxemia. Morning arterial oxygen saturation (SaO₂) estimated by pulse oximetry (92% ± 2%), which may indicate the extent of desaturation, may be inadequate under the restricted and sometimes sedentary living arrangements in the hypoxic rooms to induce an adequate erythropoietic response. Under terrestrial conditions, the subjects are likely to complete everyday tasks in a hypoxic state, so a greater hypoxemia may be induced. The study was successful in blinding the subjects to the condition they were in, and because there were no performance gains or physiological enhancements, it may be suggested that any previously reported benefits of LHTL could have been, in part, a result of a placebo effect. However, given the novelty of the findings, further investigations are required using a blinding protocol but with a greater degree of hypoxia.

Another factor of importance to the outcome of an altitude training strategy on performance is the training undertaken during the intervention period. Burtscher et al. (⁶) sought to investigate the effect of LLTH with consideration of training per se on economy, expressed as the relative oxygen consumption (mL·min⁻¹·kg⁻¹) taken from three submaximal running speeds, in well-trained endurance runners (Table 1A). The novelty of this particular investigation is that it examined repeat exposure to altitude in a manner that reflects standard protocols used by elite coaches. The main finding was that running economy improved during the 13-wk study period in the LLTH group and the control group, but only during the first 5-wk training phase of LLTH when compared to training alone. The authors (6) suggest that LLTH is more likely to affect running economy positively in “preseason” training compared to a precompetitive phase, and as such, if manipulation of economy is the goal of the training program or hypoxic exposure then this needs to be planned appropriately in an athlete’s training program.

The preceding discussion may be summarized by the following main points:

• Mixed altitude strategies (such as LHTL plus train high) appear to increase Hb_mass, V˙O_2max, and endurance performance greater than typical LHTL (^34,35).
• Absolute changes in Hb_mass and V˙O_2max do not necessarily translate into proportional changes in endurance performance (^20,34,35).
• Absolute changes in Hb_mass and V˙O_2max appear to be repeatable as a result of LHTL, but performance improvements are far more variable (³⁴).
• LHTL may induce additional nonhematological improvements that transfer into cycling performance improvement (¹⁸).
• Enhancement of Hb_mass as a result of LHTL may improve repeat maximal cycling performance (¹⁸).
• In elite athletes, sedentary and confined living conditions as a result of living in an altitude house may not provide an adequate stimulus for enhanced erythropoietic activity, leading to a substantial increase in Hb_mass (⁴¹).
• Consideration of current training phase is required when implementing an altitude training program designed to enhance specific endurance performance determinants, in particular, running economy (⁶).

Mechanistic Advancements

There is widespread acceptance among coaches and athletes that altitude training can enhance endurance performance with an almost Darwinian-like progression in protocols used. From a scientific perspective, the concept of altitude training is generally supported, but the induced mechanisms remain inconclusive and the subject of a polemic in relation to the ethics of use (^{3,26,36,43,44}). This section, therefore, focuses on advancements in our understanding of mechanisms involved in altitude adaptation (Table 1B).

Although unlikely to be the sole factor that determines improved performance at sea level after altitude training (^18,19), changes in Hb_mass reflect the major adaptation (³⁸). The optimized carbon monoxide rebreathing method for measurement of Hb_mass (³⁷) is, therefore, a particularly important tool for assessing the effectiveness of altitude exposure. Given the novelty of the technique, there remain a number of unanswered questions (³⁸). Garvican et al. (¹⁷) studied the seasonal variation of Hb_mass in internationally competitive female endurance cyclists (Table 1B) as the extent of variation is unknown, in particular, because of manipulations in training load. Their study revealed that Hb_mass varied by 3.3% during a competitive season, an effect that could partly be related to changes in training load because it was found that a 10% change in training load during a 6-wk period was associated with a 1% change in Hb_mass. A further investigation (¹¹) sought to determine within-subject variation in Hb_mass in elite athletes during the period of 1 year with the additional effect of altitude training on Hb_mass. It was found that the variation in males and females was 4% and that altitude training can potentially increase or decrease Hb_mass by a further 3%, which is in line with previous investigations (Table 1B). Two investigations (^15,42) examined to what extent Hb_mass increases with endurance training from adolescence to adulthood. It was found that endurance training from the ages of 15 to 17 years (¹⁵) and from 21 to 28 years (⁴²) has no effect on Hb_mass, whereas between the ages of 16 and 21 years (⁴²), it appears to have a significant effect. These results do appear to be in contrast to the aforementioned investigations (^10,17). However, training load was not measured/estimated in either study (^15,42), and the investigation by Steiner and Wehrlin (⁴²) was not a longitudinal analysis, which may explain the discrepancy in results to a certain extent, particularly if the training loads in those studies (^10,17) were significantly higher. Thus, in elite athletes, there does appear to be variation of approximately 3% to 4% in Hb_mass throughout a training year (^10,17) that exceeds the typical methodological noise (approximately 2%) reported for the carbon monoxide rebreathing method (¹⁹). The exact mechanisms for an increase in Hb_mass relative to increases in training load remain inconclusive; however, it may be the case that a certain “threshold” of training load has to be reached before any increases are realized (e.g., Lawler et al. (²⁴) and Maassen and Pries (²⁵)). The exact “threshold” is undetermined but is likely to have a significant individual variation. Indeed, there is considerable individual variability in the EPO response to altitude (⁸).

The exact mechanisms for individual variability in EPO response to altitude are undetermined. Chapman et al. (⁸) sought to determine whether the hypoxic ventilatory response to altitude may be related to the magnitude of EPO released from the kidney because ventilatory differences may alter the degree of arterial hypoxemia sensed by the kidney (Table 1B). It was found that there was no correlation between changes in EPO at altitude and hypoxic ventilatory response measured at sea level. This suggests that peripheral chemoresponsiveness may not be responsible for the variability in EPO response in elite distance runners, and the likely mechanisms are likely to be downstream from the lung (⁸). The same group (⁷) also sought to investigate the relationship between SaO₂ and impairment in 3,000-m run time at altitude in elite runners (Table 1B). It is commonly observed that athletes with the largest sea level V˙O_2max appear to have the largest decline of V˙O_2max when exercising at altitude (²⁴). This is likely due to limitations in pulmonary gas exchange that may be the result of a number of factors that collectively act to reduce SaO₂ (30). Although a clear mechanistic link appears to exist between reduction in SaO₂ and decline in V˙O_2max at altitude, this has not been linked to actual race performance at altitude. As expected, Chapman et al. found (⁷) a significant relationship between the degree of impairment in 3,000-m race time at altitude compared to normoxia and reduction in SaO₂, which also manifested itself as an inability to maintain oxygen uptake during the run. Although Chapman et al. (⁷) have demonstrated a link between the ability to maintain SaO₂ and endurance performance at altitude, further work is required to distinguish the exact mechanisms contributing to a reduced SaO₂.

As mentioned previously, there are likely to be alternative, nonhematological, mechanisms for enhanced endurance performance at sea level after altitude exposure (^18,19). Although EPO is well known for its effect on red cell mass, it has only recently been studied extensively owing to the potential effects it may have on nonerythroid tissues, in particular the brain, heart, smooth muscle, and vascular endothelium (¹), mediated by specific cell surface receptors (EpoR). It has emerged as a major tissue-protective factor from hypoxia-induced cell damage in various nonhematopoietic organs and, as a result, has been investigated mainly in view of clinical applications (¹). Nevertheless, the role of EPO in exercise performance has generated some attention (²⁵). For example, in healthy subjects, EPO treatment appears to increase mood (²⁷) and perception of physical conditioning (²⁸), suggesting that EPO may have the capacity to modulate central fatigue or competition strategy (²⁹). As a result, Rasmussen et al. (³¹) investigated whether EPO administration, with sufficient dose to pass the blood-brain barrier, would improve exercise capacity (Table 1B). It was concluded that high-dose EPO administration had no effect on time to exhaustion in hypoxia, cognition, and voluntary activation but actually increased RPE. However, it must be noted that using a time-to-exhaustion test as a means of examining performance without completing log-log modeling may result in less reliability of the test than a time trial performance test (²¹). In the investigation by Rasmussen et al. (³¹), although exercise performance was statistically unchanged (423 ± 90 vs 436 ± 55 kJ), this does represent an approximately 3% improvement in work capacity. It is unclear whether the small change in endurance capacity observed after EPO administration (³¹) will transfer to a real-life race situation in elite athletes (²²). A further question may be what happens when EPO is reduced significantly, as is the case when athletes immediately return from altitude training to sea level, a process that may take up to 10 to 12 d to return to normal baseline levels (^9,16). Aside from the potential loss of red blood cells (i.e., neocytolysis (³²)) a fall in circulating EPO may potentially have a negative effect on mood during that period and as a result exercise performance (²⁹). This may have a contributing role to play in the reduction of performance often observed in elite athletes returning from altitude training, usually seen on days 3 to 10 (^3,26,36,43). EPO appears to take up to 3 d to alter mood and cognition when administered exogenously (27); that it may take a similar period to negatively affect mood and cognition and may partly explain the often unexplained underperformance of athletes during that period. This is, however, speculative and requires further investigation.

EPO-EpoR signaling is just one mechanism that enables an organism to adapt to hypoxia. When faced with low arterial oxygen content a wide range of (short- and long-term) adaptive responses take place at the systemic level, at the tissue level, and at the cellular level to regulate oxygen homeostasis. At the center of these adaptations is hypoxia-inducible factor 1 (HIF-1). It is beyond the scope of this review to fully explore HIF-1 (for a comprehensive review, see Semenza et al. (⁴⁰)), but once tissue hypoxia induces HIF-1 activity, a cascade of cellular events that serve to upregulate a number of key adaptive processes, such as erythropoiesis, angiogenesis, and glucose and energy metabolism and their associated gene transcripts is initiated. Schmutz et al. (³⁹) sought to examine the specific alterations of muscle transcriptomes during repeated LLTH during a 6-wk training block (Table 1B). Given the study design (Table 1B), this investigation allows the examination of adaptations at the cellular level, which may enhance endurance performance in addition to hematological adaptations as a result of prolonged altitude exposure. The authors found that the gene expression response of myocellular energy pathways to endurance work is graded with regard to metabolic stress and training state. In particular, they conclude that improvements in oxidative energy metabolism and maximal aerobic power output with endurance training are specifically enhanced by the incorporation of a hypoxic stimulus. However, given the effect of training state and associated metabolic perturbations on gene expression (as a result of training load), it is unclear if there would be a similar response in elite athletes, as presumably, the activation threshold for these gene oncologies would be different. Nevertheless, this study (³⁹) significantly advances our understanding while, at the same time, also raises our awareness of the complexities of the cellular events activated by training and hypoxia (in isolation or in combination).

The preceding discussion may be summarized by the following main points:

• In elite athletes, there appears to be variation of approximately 3% to 4% in Hb_mass throughout a training year (^10,17).
• Variability in EPO response at altitude is not associated with hypoxic ventilatory response measured at sea level (⁸).
• The degree of impairment in 3,000-m race time at altitude compared to normoxia is associated with SaO₂ (⁷).
• EPO administration had no effect on time to exhaustion in hypoxia, cognition, and voluntary activation but actually increased RPE (³¹). However, although exercise performance was statistically unchanged (423 ± 90 vs 436 ± 55 kJ), this does represent an approximately 3% improvement in work capacity.
• LLTH training over a 6 wk period alters myocellular gene expression (³⁹).

Final Remarks

This article sought to highlight recent advances in the scientific application of altitude training for endurance performance. Major advances have been made from a performance and mechanistic perspective. Future understanding of the mechanistic basis of altitude training and adaptation, in particular, myocellular events and nonhematological adaptive mechanisms, will most likely provide a major stimulus for further performance-based experimentation. In the meantime, elite athletes and coaches will continue to strive for enhancement of sea level endurance performance using a combination of current scientific literature and applied experience.

The authors declare no conflicts of interest and do not have any financial disclosure.