Due to different time and frequency of blood sampling by different researchers during and after altitude training, complete unification of results from studies on EPO kinetics is difficult, but the data presented in the reviewed literature (see Table 1) allowed us to detect certain trends. During altitude training, the EPO level increases significantly after the first to third days/nights at altitude (Asano et al., 1998; Friedmann et al., 1999; Ashenden et al., 2000; Stray-Gundersen et al., 2001; Dehnert et al., 2002; Heinicke et al., 2005; Wehrlin et al., 2006; Schuler et al., 2007; Clark et al., 2009; Garvican et al., 2012a; Neya et al., 2013; Chapman et al., 2014; Czuba et al., 2014). After the peak, the EPO level start to fall gradually, but stays above initial values for a few days to weeks. Changes of EPO level become non-significant compared to baseline values after the first (Schuler et al., 2007; Garvican et al., 2012a; Chapman et al., 2014), second (Stray-Gundersen et al., 2001; Dehnert et al., 2002), or third week (Wehrlin et al., 2006; Czuba et al., 2014) at altitude.

Following a few weeks of altitude training, after returning to sea level, EPO concentration returns to baseline immediately or after a few days (Asano et al., 1998; Friedmann et al., 1999; Ashenden et al., 2000; Wehrlin et al., 2006; Czuba et al., 2014) or even declines below the initial value (Stray-Gundersen et al., 2001; Mounier et al., 2006; Clark et al., 2009; Garvican et al., 2012a; Chapman et al., 2014).

Some of the studies show that LH-TH, LH-TL, and HiHiLo protocols all have beneficial effects on the hematological variables (Table 1). Following altitude training, RCV (Mounier et al., 2006; Wehrlin et al., 2006; Chapman et al., 2014) and tHb_mass increase relative to the pre-altitude values (Wehrlin et al., 2006; Clark et al., 2009; Garvican et al., 2012a; Pottgiesser et al., 2012; Neya et al., 2013). Also hemoglobin concentration ([Hb]), hematocrit value (Hct), red blood cell count (RBC), and percent of reticulocytes (%Ret) are elevated after altitude training (Stray-Gundersen et al., 2001; Wehrlin et al., 2006; Schuler et al., 2007; Nadarajan et al., 2010; Chen et al., 2014; Czuba et al., 2014).

However, there are studies in which hematological variables do not improve after altitude training. The results obtained by Friedmann et al. (1999); Dehnert et al. (2002); Heinicke et al. (2005), and Robach et al. (2006) indicated that tHb_mass and RCV did not increase after the LH-TH or LH-TL protocol. In some studies also other hematological variables did not change significantly (Asano et al., 1998; Ashenden et al., 2000; Dehnert et al., 2002; Heinicke et al., 2005; Robach et al., 2006).

It also happened that even if a significant increase in EPO level during altitude training was observed, after return to sea level, improvement of hematological variables was not observed (Asano et al., 1998; Ashenden et al., 2000; Dehnert et al., 2002; Heinicke et al., 2005; Robach et al., 2006).

Studies have shown that EPO level increases following several dozen minutes or several hours of acute hypoxic exposure (Eckardt et al., 1989; Knaupp et al., 1992; Rodríguez et al., 2000; Mackenzie et al., 2008). The increase in the EPO level is higher when there is a greater decline in partial pressure of oxygen in arterial blood (PaO₂) (Eckardt et al., 1989). Acute changes in EPO level following exposure to moderate altitude are presented in Figure 2. During continuous exposure to altitude, EPO reached a peak after 1–3 days (Berglund et al., 1992; Jelkmann, 2011). This observation is confirmed by the results of the studies presented in Table 1.

In subsequent days of staying at altitude, after reaching the peak, the EPO level starts to fall gradually. The rate of decline may vary, but it seems that it does not depend on the altitude training protocol which was used. A factor that may affect the rate of decline of EPO is the hypoxic level. With an altitude below the threshold of ~2,000 m, even if it is sufficient to stimulate EPO production, the EPO level may start to drop rapidly. Li et al. (2002) observed a significant increase in EPO levels following 6 h of hypoxia at different altitudes (1,780–2,800 m). After 24 h, the EPO level continued to increase, but only for 2,454 and 2,800 m, whereas at altitudes 1,780 and 2,085 m, EPO started to decline. A similar comparison was made by Chapman et al. (2014), who documented a rise in EPO levels after 24 h at altitude, but after 72 h in the 1,780 m group, EPO returned to baseline levels, while it remained elevated in those individuals staying at a higher altitude.

The mechanisms of the reduction in EPO levels during altitude training are still unclear. On consecutive days staying at altitude, a gradual fall of EPO level with an opposite tendency of peripheral oxygen saturation (SpO₂) changes is observed (Ashenden et al., 2000; Czuba et al., 2018). Furthermore, the kinetics of EPO level in response to hypoxia are similar to hypoxia-inducible factor 1 (HIF-1) protein level kinetics. Lundby et al. (2009) pointed out that in response to hypoxic exposure, the HIF-1 protein level peaks within the first hours and gradually falls toward the initial values. Therefore, a progressive decline of EPO level and the degree of this decline are probably associated with the phenomenon of stabilization of HIF-1α. These data show that progressive acclimatization may be one of the causes of reduced EPO production in successive days and weeks during altitude training.

On the other hand, it is possible that production of EPO is not inhibited. Rusko et al. (2004) observed that serum EPO levels reflected a balance between renal EPO production and its consumption in the bone marrow. In the first hours at altitude, EPO production is higher than its consumption. The decline in EPO concentration results from elevated erythropoiesis that leads to greater EPO consumption during chronic altitude exposure. The increase of reticulocyte production accompanying the EPO decrease at altitude (Mairbäurl, 1994; Stray-Gundersen et al., 2001; González et al., 2006) supports this argument.

One of the most expected benefits of chronic hypoxia is to improve hematological variables. Exposure to a terrestrial or simulated hypoxic environment leads to a reduction in the PaO₂ as well as in arterial oxygen saturation (SaO₂) (Rusko et al., 2004). Data obtained by Eckardt et al. (1989) indicate that a higher hypoxic level causes a greater decline of PaO₂. The decrease in PaO₂ stimulates EPO production (Jelkmann, 2011). Consequently, the exposure to hypoxia induces an increase in EPO production by the kidneys, which in turn accelerates erythropoiesis. This mechanism has not been completely explored, but it is claimed that HIF-1α is the major and pivotal mediator of this erythropoietic cascade (Caro, 2001; Prabhakar, 2001; Wilber et al., 2007). The physiological threshold for acceleration of erythropoiesis has been described to be at PaO₂ levels of ~70 mmHg, which corresponds to exposure to an altitude of circa 1,600 m (Weil et al., 1968). However, Chapman et al. (1998) reported individual variability in the altitude-induced erythropoietic response ranging from 8.0 to 37.5 mU/ml at 2,500 m, with further dependency upon exposure time.

A rise in erythropoiesis is possible if the erythropoietic cascade is started. Previous studies have shown that the increase of the EPO level and acceleration of erythropoiesis depend on the “hypoxic dose.” Both level and duration of hypoxic exposure can influence the hematological response to altitude training. The results of the meta-analysis carried out by Gore et al. (2013) indicated that tHb_mass increases at ~1.1%/100 h of altitude for LH-TL. The anticipated rise of tHb_mass was ~3.4% after 2 weeks of LH-TH training (>2100 m). The LH-TL protocol is just as effective as LH-TH when the total hours of hypoxia are matched, and altitude increases to ~3,000 m. Likewise, Rusko et al. (2004) suggested that the minimum time to reach an acclimatization effect and improve blood oxygen-carrying capacity appears to be >12 h per day for at least 3 weeks at an altitude of 2,100 to 2,500 m. Similarly, data obtained by Li et al. (2002) showed that altitudes of >2,100 to 2,500 m are needed to stimulate EPO production. Wilber et al. (2007) argued that the altitude of 2,000 to 2,500 m is sufficient if the daily dose of hypoxic exposure is higher than 22 h for a minimum of 4 weeks. Fewer hours (12–16 h) may be also used, but exposure to a higher altitude (2,500 to 3,000 m) is necessary to reproduce the erythropoietic effects.

The above data show that the hematological response may not be triggered when the altitude is too low or when the total time of hypoxic exposure is too short. Studies presented in Table 1 indicate that when the total duration of hypoxic exposure was <200 h (Ashenden et al., 2000; Dehnert et al., 2002; Robach et al., 2006), a hematological response was not elicited. In turn, other research indicate that total time in hypoxia of about 210 h was sufficient to increase RCV (Mounier et al., 2006) or tHb_mass (Neya et al., 2013). However, in these studies the altitude to which the athletes were exposed was higher, reaching ~3,000 m. The results of some studies (Asano et al., 1998; Dehnert et al., 2002) also show that an altitude <2,000 m may be ineffective for improving hematological variables. However, the data presented by Chapman et al. (2014) and Nadarajan et al. (2010) indicate that it is possible to improve RCV or %Ret after training performed at an altitude below 2,000 m. Other studies support these observations. Garvican-Lewis et al. (2015) indicated that after 3-week training at 1,800 m, tHb_mass increased by 3% and it was higher than in the control group, which trained at 600 m. Furthermore, Frese and Friedmann-Bette (2010) carried out an experiment in which two training camps were held at altitudes of 1,300 and 1,650 m (LH-TH), separated by training at sea level. They reported a non-significant increase in tHb_mass after the first 3 weeks of altitude training, but after the second training camp tHb_mass increased by 5.1% compared to the pre-test values. In this experiment, the cumulative time at altitude was ~1,000 h and perhaps compensated for the lower altitude to elicit a hematological response.

These observations indicate that the total dose of hypoxia resulting from the combination of altitude and time of exposure seems to determine the hematological response to altitude training. As Garvican-Lewis et al. (2015) point out, further research is needed to better understand the relationship between hypoxic dose and improvement of hematological variables, and to determine the minimally effective altitude.

Attempts to optimize different doses of altitude exposure are still underway. This optimization would make it possible to establish a minimum threshold efficacy of hypoxic exposure. Garvican-Lewis et al. (2016) proposed a new metric, with hypoxic dose termed “kilometer hours” and defined as km·h (“km” denotes elevation of the exposure and “h” is the total duration of the exposure). This model integrates the role of exposure time and altitude. However, Millet et al. (2016) stressed the large interindividual variability in the physiological responses to a specific hypoxic dose; therefore, they proposed a metric based on the magnitude of the stimulus rather than on the altitude. This metric has been termed saturation hours and defined as %·h = (98/s - 1) x h x 100, where “s” is the saturation value (in %) and “h” is the time (in hours) sustained at any second level.

As Millet et al. (2016) rightly pointed, the metric proposed by Garvican-Lewis et al. (2016) concerns the hypoxic stimulus, while ”saturation hours“ refers to the hypoxic response. Such an approach could help to individualize training. For coaching practice, the use of the “saturation hours” metric means that to obtain the same “hypoxic dose,” those athletes with a higher saturation value or a faster rise in SpO₂ during altitude training should extend their training time or increase exposure to a higher altitude to reduce saturation. This approach would increase the efficacy of altitude/hypoxic training by considering the individual response to applied stimuli.

Undoubtedly, standardization of hypoxic dose metrics would help establish clear guidelines for altitude training prescription, which includes the potential for a targeted approach based on individual response patterns, whilst providing a stronger framework for interpreting and comparing different training protocols employed in research and practice.

The positive effects of altitude training in athletes are temporary (Wehrlin et al., 2016). Even if immediately after or a few days after a return to sea level, hematological variables are elevated, some data indicate that nearly all of the hypoxia-induced changes may be lost within 1–2 weeks (Pottgiesser et al., 2012). Chapman et al. (2014) noted a decrease in RCV, while Brugniaux et al. (2006) observed a drop in tHb_mass to the baseline level after 14 to 15 days following the completion of altitude training. This loss of effect may be why some studies did not detect improvements in hematological variables despite applying the recommended hypoxic dose. When the only post-altitude measurement is made after 14 days, the effects may not be observed (Heinicke et al., 2005; Robach et al., 2006). On the other hand, in a study conducted by Brocherie et al. (2015), the tHb_mass was significantly higher 3 weeks after altitude training compared to pre-altitude levels and this result remains consistent with the model proposed by Gore et al. (2013), where gains in tHb_mass values are estimated for ~3 weeks.

Interestingly, even if hypoxia promotes erythropoiesis, it remains uncertain whether this will lead to an increase in RBC, as in certain conditions the reticulocytes may be targeted for early destruction (Hahn and Gore, 2001). This is likely linked to the phenomenon of neocytolysis. Neocytolysis is a physiological process that controls red cell mass by down-regulation if it is excessive. Because the absence of hypoxic stress after return to sea level causes a decline of the EPO level, neocytolysis is accelerated, which consequently leads to selective hemolysis of the youngest circulating red blood cells (Alfrey et al., 1997; Rice et al., 2001; Rice and Alfrey, 2005). A significant decrease in tHb_mass, fall of erythrocyte survival time, lower iron turnover and reduction of bone marrow production of erythroid cell lines are observed after the return to sea level of both native highlanders and those individuals who stayed at altitude temporarily (Levine, 2002; Prommer et al., 2010). These results suggest that neocytolysis may be one of the causes of the lack of improvement in hematological variables after return to sea level even if altitude training favored erythropoiesis.

For athletes, the duration of maintaining the post-altitude training effects is a very important issue, because it determines the moment of the return to sea level before the competition. Schmidt and Prommer (2008) noted that tHb_mass is reduced again by 50% after 3 weeks following the return to sea level. Therefore, the time between altitude training and the competition should be <20 days. Dick (1992) claimed that the maximum performance occurs between 15 and 24–28 days, whereas the first week following altitude training is characterized by poor performance. Conversely, Heinicke et al. (2005) found that the re-acclimatization period prior to competition should be very short (1–3 days). Chapman et al. (2014) concluded that interaction between hematological and other physiological parameters influences the optimal time to compete after altitude training. Furthermore, Turner et al. (2017), suggested that additional normobaric hypoxic exposure (4,200 m and above, 2 h per day) after an altitude training camp can prevent a sudden decline in EPO levels and help maintain an improved tHb_mass level. This problem is yet to be explained, and practical application should be individually adapted to athletes.

Based on the majority of the reviewed literature, it would seem logical to suppose that the conflicting and inconsistent reports concerning tHb_mass after altitude training might be, at least partially, attributed to different methodological approaches used by researchers. It was indicated that not all techniques had the same precision and were suitable for the research concerning training in hypoxia (Wehrlin et al., 2016). Furthermore, Garvican et al. (2012b) and Millet et al. (2017) provided evidence that the use of some tHb_mass measurements may lead to either overestimated or underestimated tHb_mass values, which makes it difficult to achieve unequivocal interpretation of the results and evaluation of the actual effects of altitude training. Based on the comments by the above mentioned authors, one may suppose that among currently available techniques the most precise method for tHb_mass estimation is the CO-rebreathing method with refinements that allow for minimization of measurement errors and determination of individual responses to altitude/normobaric hypoxia exposure (Garvican et al., 2012b).

The rate of EPO production is related to the level of hypoxic stress, but the increase in EPO induced by altitude is characterized by inter-individual variability. The difference of EPO level after exposure to the same altitude can be up to several hundred percent between subjects (Table 2). The study conducted by Li et al. (2002) showed that some subjects achieved about a 100% increase in EPO levels after exposure to 1,780 m, whereas others did not respond to the altitude of 2,805 m. The individual EPO response may affect the level of hematological adaptation and performance in athletes after altitude training. Chapman et al. (1998) collected data from 39 athletes living at an altitude of 2,500 m. The athletes performed a 4-week training block at altitudes between 1,200 and 3,000 m. Based on the improvements in sea-level performance during a 5,000 m run, the study group was retrospectively divided into responders and non-responders to altitude training. It was found that the responders had significantly larger increases in mean EPO concentration after 30 h of hypoxia compared to the non-responders. After 14 days of exposure to altitude, EPO remained elevated only in responders. It is likely that for non-responders, greater hypoxic stimuli may be needed to induce a sufficiently large release in EPO and augment red cell production (Chapman et al., 1998). These observations showed that individual aptitudes of athletes play an important role in the efficiency of altitude/hypoxic training.

Individual variability in response to altitude training is also important in the context of the efficiency of anti-doping tools such as the Athlete Biological Passport (ABP). The results of a meta-analysis conducted by Lobigs et al. (2018) indicated that due to the complex nature of the response to altitude training, altitude still cannot be directly included in the ABP algorithm. However, better knowledge of the effects of altitude training on blood markers can make it easier for experts to make a judgment on a dubious ABP profile. We believe that the information presented in our review may also be a guide for experts concerning the impact of altitude training on erythropoietic response and hematological variables.

Various mechanisms have so far been suggested as responsible for individual variability, but many factors associated with the altitude response seem to be genetically determined (Chapman et al., 1998; Li et al., 2002). Attempts to find genetic determinants of the hypoxic response are in progress.

For instance, Witkowski et al. (2002) suggested that a specific haplotype of the EPO gene may exist, which can be useful in predicting the erythropoietic response to altitude. Further analysis (Jedlickova et al., 2003) provided more information on this issue. In this study, based on the level of increase in the EPO level after hypoxic exposure, athletes were divided into low responders, intermediate responders, and high responders. The obtained data demonstrated that the 185-bp allele of the EPO gene was most prevalent in the group of high responders. The 185-bp allele showed a significant correlation with the EPO hypoxic response. When all study participants were considered, the increase in EPO in the group of people with the 185-bp allele was 135%, whereas in the group without this allele, the EPO level rose by 78%.

In other studies, the association between polymorphism of the HIF-1α gene and hypoxic response was analyzed. Tanimoto et al. (2003) demonstrated that the T allele (Ser582) compared to the C allele (Pro582) is related to higher HIF-1α activity. In turn, Liu and Hu (2006) demonstrated a higher SaO₂ during exercise performed in hypoxia in subjects with the CT genotype. Furthermore, following 4 weeks of hypoxic exposure and hypoxic training, changes in maximal oxygen uptake (VO₂max) were higher for the CT genotype compared to the CC genotype.

However, there are data that do not confirm the above observations. Hennis et al. (2010) examined whether the presence of the T allele impacts the EPO level, hypoxic ventilatory response (HVR) and SaO₂. However, the results demonstrated that the CT/TT genotype was not associated with an improved response for 8 h hypoxic exposure. No differences were observed in the increase in EPO levels following 4 and 8 h of exposure or in the levels of SaO₂ and HVR between the CC genotype and CT/TT genotypes. Also, Richalet et al. (2009) did not detect a correlation between HIF-1α genotypes and HVR during exercise in normobaric hypoxia.

The potential relationship between the ACE gene polymorphism (I/D) and hypoxic response was also analyzed. Angiotensin II is involved in the regulation of EPO production and it can increase EPO level (Freudenthaler et al., 2000; Gossmann et al., 2001). Presence of the I allele of the ACE gene reduces enzymatic activity and, consequently, decreases the level of angiotensin II. This suggests that the DD genotype may be associated with a greater increase in EPO level after hypoxic exposure. However, González et al. (2006) reported no significant differences in EPO levels between the DD and ID/II genotypes. The responders and non-responders had similar genotypic profiles. These results led to the conclusion that polymorphism of the ACE gene does not influence the EPO response in athletes during exposure to hypoxia.

It turns out that the I allele of the ACE gene is identified in mountaineers more often than the D allele (Martin et al., 2010). Woods et al. (2002) found that presence of the I allele can produce benefits at high altitude due to prolonged maintenance of SaO₂. Furthermore, Bigham et al. (2008) observed higher values of SaO₂ both at rest and during exercise at high altitudes (~4,000 m) in subjects with the I/I genotype compared to other genotypes. However, other research, conducted by Patel et al. (2003) and Richalet et al. (2009), revealed no differences between ACE genotypes in the values of HVR and SaO₂ during exercise in hypoxia.

Current reports on genetic determinants of the hypoxic response are few and ambiguous. Therefore, further research is needed, because such data may be useful in predicting the response to altitude training.

Due to the individual variability of EPO production and many factors affecting the hypoxia response, identification of responders and non-responders among athletes prior to altitude training seems to be useful in optimizing the training process.

As suggested by Chapman et al. (1998), the number of non-responders can be minimized by screening of the erythropoietic and training velocity response to acute hypoxic exposure. In studies conducted by Friedmann et al. (2005), the EPO increase following 4 h of exposure to normobaric hypoxia was significantly correlated with the acute EPO rise during altitude training (after the first and second day). These findings indicate that short hypoxic exposure can be used to identify athletes who respond to altitude training. Based on the results of previous research (Figure 2), it can be assumed that the EPO level after 24 h of exposure to moderate altitude should be at least doubled in relation to baseline. The smaller rise may suggest that the response to altitude/hypoxic training will also be weak.

However, as we observed on the basis of the reviewed studies, even a significant increase in EPO level during altitude training does not guarantee improvement of hematological variables after the return to sea level (see Table 1). This observation is in line with previous reports in which correlations were not found between the EPO levels following exposure to altitude and tHb_mass (Rusko et al., 1999; Friedmann et al., 2005). Therefore, it is doubtful whether the improvements of hematological variables after altitude training can be predicted using the EPO response to acute hypoxic exposure. Determining efficacious methods to identify better and worse responders prior to altitude training remains a topic for further research.

Regardless of attempts to find an effective tool for identification of responders and non-responders prior to altitude training, daily measurement of SpO₂ at rest during altitude training may by a useful and inexpensive method to analyze the hematological response to chronic hypoxia. Millet et al. (2016) noted that the SpO₂ could be easily measured without limitations or inconvenience to athletes through development of sportswear incorporating an oximeter inside the textile. It is known that the increase in EPO levels is directly proportional to the level of hypoxia and decline in SpO₂ (Eckardt et al., 1989). In our recent study Czuba et al. (2018) conducted during successive weeks of altitude training, SpO₂ in the altitude training group showed an upward tendency that was opposite to changes in EPO levels. Consequently, the effect of hypoxia on blood EPO levels became weaker with increasing acclimation, which was caused by the improvement in blood oxygen-carrying capacity. Since the desaturation level depends not only on altitude but is highly individualized and also dependent on the training background of athletes or previous altitude experiences (Chapman et al., 1998, 2014; Millet et al., 2016), the measurement of SpO₂ during altitude training will allow for the measurement of the time the athlete has spent at a given saturation level and determine the actual magnitude of the stimulus individually for each athlete (Millet et al., 2016).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.