Altitude training is known for more than 50 years. The international scientific discussion started with the Olympic Games in Mexico City. Athletes from all over the world prepared for competition under hypoxic conditions at 2300m. In these years altitude training was primarily used for competition at natural altitude. Afterwards it became also relevant for endurance performance capacity at sea level and thus an important research area in sport science. In addition to natural hypoxia artificial hypoxia plays more and more an important role. There are various methods of artificial hypoxia like hypoxic chambers, rooms, tents or houses under normobaric conditions, portable hypoxic breathing masks or hypobaric chambers.

During the last 40 years, several methodical approaches have been developed to foster altitude training. Among the classical form of altitude training „Sleep high – Train high“ the methods of „Sleep low – Train high“ and „Sleep high – Train low“ have been established too (Friedmann & Bärtsch, 1997; Friedmann, 2000; Navot-Mintzer, Epstein & Constantini, 2003). These procedures apply under natural and under artificial hypoxic conditions.

In recent years, artificial hypoxia, differentiating between hypobaric and normobaric hypoxia, was subject to investigations regarding the methods of intermittent hypoxia. Here, the athletes are not permanently exposed to the stimulus of oxygen reduced air. In some of the approaches the subjects stay in rest a part of the day in oxygen reduced environment (Katayama, Shima, Sato, Qui, Ishida, Mori & Miyamura, 2001; Katayama, Sato, Matsuo, Ishida, Iwasaki & Miyamura, 2004), in other studies training takes place in hypoxia (Hendriksen & Meeuwsen, 2003; Levine 2002; Meeuwsen, Hendriksen & Holewijn, 2001).

Results of several investigations demonstrate the difficulty of choosing the correct individual exercise stimulus under hypoxic conditions. The results indicate significant improvements and stagnation of performance as well as substantial decreases of performance capacity of the measured parameters.

An early group controlled crossover study published by Adams et al. (1975), for example, did not show any effects on oxygen uptake after training under hypoxic conditions. In contrast, performance enhancing effects were revealed by a study conducted by Stray-Gundersen, Chapman & Levine (2001) who studied 22 long distance runners of national and international level. Fourteen men and eight women had to stay at an altitude of 2500m for 27 days and practiced a high intensive training at 1250 m. The results show a significant increase in running-time regarding the 3000 m competition of 1.1 %, a significant improvement of maximal oxygen uptake (VO2max) of 3 %, a significant enhancement of hematocrit (HCT) and erythropoietin (EPO).

In literature in most of the studied group based designs have been used. As there are significant individual differences in measured physiological reactions (Friedmann & Bärtsch, 1997; Osterburg, 2004) it makes sense to take not only group-statistical procedures into account but also specific single-cases. A study published by Friedmann, Frese, Menold, Kauper, Jost & Bärtsch (2005), tested 16 junior swimmers of international level who stayed in a training camp for 3 weeks at an altitude of 2100-2300 m (SH –TH). The results demonstrate the individuality of effects of hypoxic training. The results of single athletes of the incremental swimming test range from a decrease of maximal performance of 1.5 % to an increase of 5 % as well as a decrease of performance at anaerobic threshold of 1 % up to an improvement of 10 % between pre and post tests. Truijens, Toussaint, Dow & Levine (2003) found differences between individual athletes in VO2max in a post test after altitude training of more than 20 % (-6 % bis +16 %) in comparison to the pre test. The 18 swimmers of the controlled double-blind placebo controlled study trained for four weeks under artificial hypoxia. An older study of Mader et al. (1980) analysed the performance of ten rowers on a rowing-ergometer before and after an altitude training camp and found individual changes between -12.1 % and +25.5 %.

The parameters that are used for characterizing training and for testing the biological response as well as the performance can be assigned to different categories at a “four-level-model of diagnosis” (Fig. 1).

The first level covers daily input data and basic markers like training protocol, resting heart rate, body weight and psychological variables. At the second level parameters of physiological markers (blood count, reticulocytes, EPO etc.) are used. Directly or indirectly performance related markers measured in lab are represented at level three. The fourth level describes the competition performance measured objectively (e.g. 5000 m running time).

The significance of the parameters at the four levels regarding the competition performance increases from the bottom to the top without explaining the variance of performance completely. The most significant criterion that defines the potential of aerobic performance and associated motivational parameters is competition.

The aim of the following study was to investigate possible changes of endurance performance as a result of an intermittent training in a normobaric hypoxic chamber (2500 m) at different levels of biological response.

The group of subjects comprised seven athletes (age: 24 ± 1.8 years, height: 176.6 ± 3.4 cm, body weight: 69 ± 6.3 kg) of different sports (Tab. 1).

Table 1: Characteristics of the subjects

The study with a total duration of 23 weeks started with a two-week specification of the baseline, followed by the first test week (Fig. 2). After another two weeks the first training block with a three-week training phase in each case was set up at sea level and under hypoxic condition. Each sea level and altitude phase was followed by a test week. After this week the second training block which was build according to the first block was carried out. Finally, a post test phase with a duration of 3 weeks and the last test week took place.

The training was practiced on a treadmill (PPS 55-med, Woodway, Weil/Rhein) in a normobaric hypoxic chamber (Fa. Hypoxic) at an artificial altitude of 2500 m (PO2 15 %). The intensity was set to 70 % of the anaerobic threshold which was determined in each case by an incremental test under sea level or altitude conditions. At the second block, the Training volume was increased; whereas the volume in the block remained unaffected.

Physiological basic parameters (Level 1)
Every day after, the subjects carried out measurements of resting heart rate (Polar, Finnland), body temperature and body weight and documented sleeping hours.

Two to three times a week venous blood samples were drawn in the morning to determine the concentration of haemoglobin (HGB), erythrocytes (RBC), leukocytes (WBC) (Sysmex KX-21N, Sysmex Deutschland GmbH, Düsseldorf), reticulocytes (Advia 120, Bayer AG, Leverkusen) and ferritin (Elecsys 1010, Roche Diagnostics GmbH, Mannheim).

During the test week and at the beginning of each training phase incremental tests on a treadmill (PPS 55-med, Woodway, Weil/Rhein) were carried out. The velocity of the treadmill was increased by 0.5 m/s every five minutes until exhaustion. Before, after every stage and three minutes after the last stage, 20 µl blood samples were taken from the earlobe to determine the lactate concentration (Ebio Plus, Eppendorf AG, Hamburg). In addition, heart rate was registered (Polar, Finland) and respiratory parameters were analyzed by breath using the ZAN 600 (ZAN Messgeräte GmbH, Oberthulba).

During the testing week a competition of 5000 m running on a 400 m-track took place as well. During the test the running heart rate (Polar, Finnland) was recorded and lactate concentration was determined from earlobe before, after and three minutes post to the end of the test.

Statistical analysis was done by using a statistic software package (Statistica for windows, 6.0, Statsoft, Tulsa, USA). As normal distribution could not be assumed, the data of the group were compared by the non-parametric Wilcoxon test for dependent samples. The level of significance was set at p=.05.

In order to analyze the biological response of individuals single-case-studies were applied. The fitting of time series was carried out by a 4253H-filter.

There were no significant differences of resting heart rate during the different phases of the study. At the first altitude phase the group median of resting body temperature decreased significantly in comparison to the first sea level phase. In the second altitude phase the median of the body weight of the group is lower than in the baseline. During the same phase sleeping hours increased compared to post test phase (Tab. 2).

At the group level the blood parameters did not show any significant differences during the phases of the study.

In the single-case of P 1 erythrocytes increased at the first altitude phase, levelled off at sea level, but decreased after the second altitude phase (Fig. 3). Ferritin values increased at the end of the first altitude phase up to the end of the second altitude phase and remained at this level.

Both, the group results and the results of the single-cases show clear variations of the running velocity at anaerobic threshold between the incremental tests before and after the first training block. Figure 4 indicates the significant increases of the group median of test 3 SL (pre altitude phase) to test 6 SL (post altitude phase).

Test 4 H and 5 H (acute and chronic hypoxia tests) demonstrate significant differences of the performance at 4 mmol/l lactate concentration compared to test 3 SL and 6 SL (Tab. 3).

In six single-cases the anaerobic threshold curves were shifted to the right at sea level after the altitude training (6 SL). The curves under hypoxic conditions (4 H & 5 H) were left-shifted. Moreover, the chronic hypoxia test (5 H) showed a right-shift after three weeks of altitude training which is exemplarily presented by P 1 (Fig. 5).

At the second altitude phase only sea level tests 10, 11 indicated a significant difference to the altitude tests 8 and 9, while test 7 SL only diverged from 8 H. The differences occurred at a decrease of velocity at anaerobic threshold under hypoxic conditions (Tab 1 & 2).

Also for the single-cases a left-shift of performance in hypoxic tests in comparison to sea level tests at second altitude phase could be seen. The changes of performance after altitude training were different according to individual conditions.

At test 3 SL in comparison with 6 SL no significant differences in VO2max could be detected. The VO2max (Test 4 H & 5 H) that was determined under hypoxic conditions was significant lower compared to VO2max which was measured at sea level (Tab. 3).

The single-cases showed contradictory results. For example, VO2max of P 4 increased at the first training block at sea level tests as well as at hypoxic tests.

Also, the second training block indicated no improvements of VO2max in the group median after the training at hypoxia. A significant decrease was seen in test 8 in comparison to 10 SL and to 11 NT as well as in comparison to 9 H and to 10 FL (Tab. 3).

The results of group-statistic analyses showed no significant changes during the whole study period. According to the results of the single-cases the male subjects achieved clear improvements of running time, whereas the performance of the female athletes remained unchanged or even decreased.

P 1 increased his competition performance after the first training block by 26 seconds. The competition performance of P 4 did not improve in the course of the study.

The results of the physiological basic markers could be explained by cumulative effects of training but this cannot be related to hypoxic effects. The significant decrease of body weight during the second altitude phase compared to baseline could be explained by the high physical workload over a long period of time combined with relatively lower energy consumption.

The significant lower body temperature group during the first altitude phase is hard to explain. It is a fact that that the body temperature increases by intensive physical workload because the sympathetic nervous system tone increases (Hollmann & Hettinger, 2000). The literature, however, does not describe a contrary effect.

The higher number of sleeping hours during the second altitude phase compared to the post test phase could be traced back to the high training load and the necessary regeneration. The subjects reported on disturbed sleep during the altitude phases. Shephard and Shek (1997) described the phenomenon of disturbed sleep in the context of high physical exercises.
The stimulus of oxygen reduced inspired air during the training phases under hypoxic conditions is associated with a higher relative physical workload for the athletes. During the altitude training, e.g., the VO2max and heart rate are reduced, whereas lactate concentration increases more rapidly than at sea level.

Different studies have shown that erythropoiesis is stimulated by hypoxic training (Friedmann & Bärtsch, 1997; Levine & Stray-Gundersen, 1997) that may lead to an improvement of aerobic performance capacity. However, there were no changes in blood parameters during or after altitude training. Furthermore, Rodríguez et al. (1999) and Venutra et al. (2003) came to the same conclusion. These findings could lead to the explanation that the stimulus of the intermittent altitude training was too small to produce changes of these blood parameters (Ventura et al., 2003). This assumption is supported by the results of the blood parameters of P1. He trained with a higher training volume compared to the rest of the subjects and his erythrocytes increased during the first altitude phase. On the other hand, the results reveal that the reactions of erythropoiesis to the stimulus of hypoxia are individual (Friedmann et al., 2005). In this context, Chapman, Stray-Gundersen and Levine (2002) use the term of responder and non-responder.

The increase of ferritin could be seen as an answer to the increasing exercise volume in the second altitude phase. On the other hand, on examination of the enhanced ferritin values the increased loss of erythrocytes caused by the increased running performance leading to an improvement of iron supply, has to be taken into account (Hollmann & Hettinger, 2000).

After the first training block, a performance improvement at the anaerobic threshold occurred in the group and in all single-cases. Thus, the observed right-shift of anaerobic threshold curves indicates an increased performance of the subjects.

The left-shifted lactate curves of tests under hypoxic condition in comparison to the sea level test confirm the so called lactate paradox (Friedmann & Bärtsch, 1997; Lundby, Saltin & van Hall, 2000). This proves that the lactate concentrations in acute hypoxia (Test 4 H) rises faster than at sea level, while maximal lactate concentrations remain unchanged. After acclimatisation (Test 5 H) a right-shift of the lactate curves occurred, that was located between the results of the acute altitude and the sea level tests. The maximal lactate concentration is reduced as well. The physiological aspects of the lactate paradox are not resolved entirely (Lundby et al., 2000).

The determination of VO2max does not show any improvements at the group examination after altitude training. Different authors (Julian et al., 2004; Ventura et al., 2003) confirm these results. In contrast, Meeuwsen et al. (2001) demonstrated significant enhancements that could be found in six single-cases of this study too.

The lower VO2max which was measured under hypoxic conditions is documented by several studies (Friedmann & Bärtsch, 1997; Koistinen, Takala, Martikkala & Lappäluoto, 1995).

For the male subjects the competition performance improved during the study. Due to the different training status of the subjects the results are very heterogeneous. P 1 ran a minimum time of 17:05 minutes in the 5000 m run on an evidently higher level than P 3. He performed the distance in 21:30 minutes. These variations could be associated with the sports of the two subjects (P 1: long distance runner, P 3: soccer player) and, in addition, to the different training volume. The considerable improvements of both athletes in spite of the variations could be explained as follows: P 1 spend a long time in oxygen reduced air whereas P 3 performed at a low trainings level that approved faster adaptations, associated with improvements of performance (de Marées, 2002).

The stagnation or the decrease of performance capacity of female subjects in competition does not comply with the results of the incremental tests. Therefore, other variables - such as motivation - have to be taken in to account.

Within the study two altitude phases were completed in order to find out whether the altitude stimulus is reproducible. Table 4 gives an overview of the effects of the altitude training of the group and the single-cases on anaerobic threshold, VO2max and competition performance after the first and second training block in each case compared to the values before the altitude phases.

These findings show greater improvements after the first altitude phase than after the second hypoxic intervention. This could lead to the conclusion that the training stimulus induced by the second altitude phase was not individually effective. As a result, the described parameters did not increase continuously or even decreased. The exceptions (Tab. 4) of some single-cases regarding the VO2max, P 1 and P 2, trained with a greater training volume than most of the other athletes (cp. Tab.1). As they had a bigger performance capacity, a higher training age and more training experiences they probably could better deal with the increased hypoxic stimulus.

The few statistically significant results of group-statistical analyses could be associated with the heterogeneous and small sized group. Many studies and own experiences indicate that it is difficult to find a homogeneous group of competitive athletes. Moreover the athletes must be willing to take part in a controlled training according to a required design of study. Only Levine und Stray-Gundersen (1997) succeeded in examining a homogeneous group of 39 subjects.

Therefore, single-case analyses could be helpful methods. In addition, the individuality of adaptation revealed through this study may be considered. The study has shown that intermittent hypoxia training may lead to individual increase of performance. However, it could not be shown whether the improvements where caused by training under hypoxic conditions or by the training only. On the one hand, this has to be seen in the context that generally an intensive training stimulus is difficult to prove in its effects and altitude training qualifies as such. One reason may, inter alia, be that training leads to individual reactions to special workloads which are often delayed. On the other hand, there is no reason to doubt the physiological effects of hypoxic stimulus. There are no adequate protocols with regard to the question which hypoxic stimulus (altitude) in combination with which training load can lead to the best individual results. If it is taken into consideration that the training age of the athletes differs, in some cases a biological effect for the purpose of an improvement of performance caused by altitude training can be assumed. Further studies of this topic are in process.

Meta-analyses of more than 100 international studies in the last 50 years and own results show a controversial picture. Some practical experiences as well as controlled studies indicate performance enhancement effects, whereas others do not. Acute and chronic hypoxia induce well-known physiological effects in gas exchange, hematology etc.

Performance enhancement, however, may occur but it is in onset, magnitude and duration very individual. The re-adaptation to sea level is quite fast; the duration of positive effects is scientifically unclear. The effect of all options live high/low – train high/low are not proven sufficiently. The criteria for individual input, like training load at altitude, are often insufficient. Results show that there are high-low responders, early-late responders and non-responders.