Eur J Appl Physiol (2007) 100:193205 DOI 10.

1007/s00421-007-0417-8

O R I G I N A L A RT I C L E

Normo or hypobaric hypoxic tests: propositions for the determination of the individual susceptibility to altitude illnesses
Gustave Savourey Jean-Claude Launay Yves Besnard Angelique Guinet-Lebreton Antonia Alonso Fabien Sauvet Cyprien Bourrilhon

Accepted: 29 January 2007 / Published online: 24 February 2007 Springer-Verlag 2007

Abstract Assessment of individual susceptibility to altitude illnesses and more particularly to acute mountain sickness (AMS) by means of tests performed in normobaric hypoxia (NH) or in hypobaric hypoxia (HH) is still debated. Eighteen subjects were submitted to HH and NH tests (PIO2=120 hPa, 30 min) before an expedition. Maximal and mean acute mountain sickness scores (AMSmax and mean) were determined using the self-report Lake Louise questionnaire scored daily. Cardio-ventilatory (f, VT, PetO2 and PetCO2, HR and nger pulse oxymetry SpO2) were measured at times 5 and 30 min of the tests. Arterial (PaO2, PaCO2, pH, SaO2) and capillary haemoglobin (Hb) measurements were performed at times 30 min. Hypoxic ventilatory (HVR) and cardiac (HCR) responses, peripheral O2 blood content (CpO2) were calculated. A signicant time effect is found for DSpO2 (P = 0.04). Lower PaCO2 (P = 0.005), SaO2 (P = 0.07) and higher pH (P = 0.02) are observed in HH compared to NH. AMSmax varied from 3 to12 and AMSmean between 0.6 and 3.5. In NH at 30 min, AMSmax is related to PetO2 (R = 0.61, P = 0.03), CpO2 (R = 0.53, P = 0.02) and in HH to CpO2 (R = 0.57, P = 0.01). In NH, AMSmean is related to Df (R = 0.46, P = 0.05),

HCR (R = 0.49, P = 0.04), CpO2 (R = 0.51, P = 0.03) and, in HH at 30 min, to VT (R = 0.69, P = 0.01) and a tendency for CpO2 (R = 0.43, P = 0.07). We conclude that HH and NH tests are physiologically different and they must last 30 min. CpO2 is an important variable to predict AMS. For practical considerations, NH test is proposed to quantify AMS individual susceptibility using the formulas: AMSmax = 9.47 + 0.104PetO2 (hPa)0.68CpO2 (%), (R = 0.77, P = 0.001); and AMSmean = 3.91 + 0.059Df + 0.438HCR0.135CpO2 (R = 0.71, P = 0.017). Keywords Hypoxic tests Hypoxia Acute mountain sickness Altitude illnesses

Introduction Unacclimatized people going to high altitudes may develop altitude illnesses such as acute mountain sickness (AMS) but also pulmonary, cerebral edemas or subacute mountain sickness. AMS is characterized by headaches, gastro-intestinal symptoms, ataxia and insomnia and may lead to more severe pathologies (subacute mountain sickness, pulmonary or cerebral edema). It is well known that a large variability is observed among individual for the development of high altitude rtsch et al. 2004). This illnesses (Robinson et al. 1971; Ba variability is consecutive to an individual susceptibility to hypoxia leading to a mal-adaptation to high altitude. Consequently, it appears that the determination of this individual susceptibility, especially to AMS, is of a great interest. Thus studies have been conducted in the past to determine the relationships between AMS and physiological/biological characteristics observed or measured

G. Savourey (&) J.-C. Launay Y. Besnard A. Guinet-Lebreton A. Alonso F. Sauvet Departement des Facteurs humains, Pole tolerance climatique et vetement, Centre de recherches du service de sante des armees, BP 87 38702 La Tronche cedex, France e-mail: gsavourey@crssa.net C. Bourrilhon Institut de medecine aerospatiale du service de sante des armees, BP 73 91223 Bretigny sur Orge cedex, France

123

194

Eur J Appl Physiol (2007) 100:193205

before high altitude exposures. These measurements were performed either in hypoxia or in normoxia. The main results reported from normoxic tests show that maximal oxygen uptake (Richalet et al. 1988; Milledge et al. 1991; Savourey et al. 1995a), pulmonary functional measurements such as forced vital capacity and forced expiratory volume by second (Savourey et al. 1995a), percentage of body fat content (Savourey et al. 1995a) were not related to AMS. Body mass index (BMI) has been inversely related to AMS susceptibility by Hirata et al. (1989) and Kayser (1991) but not by Savourey et al. (1995a). Cold pressor test has also been used by Viswanathan et al. (1978) who showed that Indian soldiers having developed high altitude pulmonary edema (HAPE) presented a lower response to cold pressor test. However, Richalet et al. (1988) and Savourey et al. (1995a) have not observed such a relationship in AMSsusceptible subjects. Normoxic tests give consequently poor results and hypoxic tests are thus more often used to study the physiological responses to acute hypoxia before an ascent at high altitude for determining the AMS susceptibility. Lower hypoxic ventilatory responses (HVR) has been described in AMS subjects compared to controls by King and Robinson (1972), Hyers et al. (1979), Hackett et al. (1982), Hu et al. (1982), Mathew et al. (1983) and by Moore et al. (1986). Richalet et al. (1988) and Rathat et al. (1992) showed that, during a retrospective study, the most AMS-clinically subjects presented at least one abnormal response (ventilatory or cardiac response to hypoxia at FIO2 = 0.115), especially during exercise. Recently, Lanfranchi et al. (2005) reported that an abnormally high low-frequency component of systolic blood pressure response during 10 min of hypoxia, indicating an exaggerated chemoreex vasoconstrictive response to hypoxia, may predict AMS. On contrast, Sutton et al. (1976), Milledge et al. (1988, 1991) and Savourey et al. (1995a) did not nd any correlation between AMS and HVR but these last authors showed that both sea level and altirtsch tude PetO2 (4,500 m) were predictive for AMS. Ba et al. (2002) reported that the susceptibility to AMS cannot be predicted by low altitude HVR. Additionnaly, Hohenhaus et al. (1995) found that HVR was lower in HAPE but not in AMS subjects. The discrepancies observed in the literature concerning the relationships between HVR and AMS could be explained (Milledge et al. 1988; Savourey et al. 1995a) by results that are not statistically different as in King and Robinson (1972) or Moore et al. (1986), by a small number of subjects as in the study of Hu et al. (1982) but also by the different techniques used to measured HVR (hy poxia at rest or exercise, pokilocapnic or isocapnic technique, duration of the test). Moreover, pokilocap-

nic hypoxic tests are performed either in normobaria (Richalet et al. 1988, Rathat et al. 1992) or in hypobaria in a hypobaric chamber (Savourey et al. 1995, 2003). Recently, Savourey et al. (2003) have showed that physiological differences were observed between hypobaric and normobaric hypoxia at a same PIO2 of 120 hPa. These differences were characterized by a greater hypoxemia, hypocapnia and a lower O2 arterial saturation in hypobaria. Consequently, the type of the hypoxic test could be very important to consider for the determination of the individual susceptibility to altitude illnesses. Today, this point is unexplored. Nevertheless, it appears that AMS-susceptible subjects could present ventilatory disturbances caused by multifactors more or less intricate such as the chemosensibility to hypoxia evaluated by HVR, the breathing pattern in hypoxia or unknown factors due to the reduction in the barometric pressure. Ward et al. (2000) wrote it would seem that although susceptibility to HAPE is associated with a low HVR, susceptibility to AMS is not. These authors (Ward et al. 2000) also reported possible gas exchange abnormalities in AMS susceptible subjects as studied by Ge et al. (1997) and by Roach et al. (1996) who showed that a low SaO2 on arrival at altitude is an indicator for the later development of AMS. Other factors may occur in the development of AMS such as the cerebral blood ow and intra cranial pressure (Singh et al. 1969; rtsch et al. 2004) or Hackett 1999; Ward et al. 2000; Ba impairments in uid balance (Ward et al. 2000). Indeed, a mild diuresis is normally observed in altitude, whereas AMS susceptible subjects present an antidiuresis leading to body weight increase (Singh et al. 1969; Hackett et al. 1982; Loeppky et al. 2005) attesting probable disturbances of extracellular volume, especially plasma volume (Ward et al. 2000). In summary, development of high altitude illnesses, especially AMS, appears multifactorial including at least ventilatory disturbances, impairment of gas exchanges and disturbances of uid balance as main factors. In these conditions, hypoxic tests performed before an ascent seem today the better choice to determine the individual AMS susceptibility, waiting further ndings (genetic data for example). However, from the review of the literature reported above, it appears that these hypoxic tests do not permit to determine the individual susceptibility to high altitude illnesses with accuracy since their use and the results obtained are still debated. This could be due to the fact that all the current tests do not take into account the main factors implied in the development of high altitude illnesses but only few of them (HVR, HCR or BMI as in Richalets tests). Moreover, whether or not the best hypoxic test must be normo or hypobaric

123

Eur J Appl Physiol (2007) 100:193205

195

remains to be dened because physiological differences have been found between normo and hypobaric hypoxia (Savourey et al. 2003). The duration of the test must also be standardized because this one varies among the different studies (5 min for four stages of the Richalets test, for example). Another criticism of the actual hypoxic tests is represented by the fact that these tests determine only the high altitude individual susceptibility without information concerning the severity of AMS or/and the degree of mal-adaptation to altitude. For example, an individual going to high altitude may present a severe AMS (high maximal AMS score) during the rst days of the stay but no more after, indicating a good adaptation as shown by a decrease in mean AMS score calculated throughout the sojourn. Inversely, an individual presenting a moderate AMS will have a moderate maximal AMS score but a possible high mean AMS score in the case of a lesser adaptation. Thus, it seems to us that the severity and the degree of mal-adaptation must be studied together as they are more indicative for practical medical consideration than the determination of the individual susceptibility alone (AMS+, AMS). Consequently, we hypothesized that a combination of the main physiological variables implied in the development of altitude illnesses such as cardio-ventilatory variables (ventilatory volume, HVR, breathing pattern, heart rate, HCR), blood variables resulting from gas exchanges (arterial blood gases PaO2, PaCO2, CaO2, SaO2, SpO2) or plasma volume changes (indirectly represented by changes in haematocrit or hemoglobin concentration) measured during a hypoxic test could be related both to severity (maximal AMS score) and/or to the degree of the mal-adaptation to altitude (mean AMS score) observed during various high altitude expeditions. In order to differentiate not only the effects of the type of the hypoxic test (hypo and normobaric hypoxic test) but also its duration, we have realized 40 min normo and hypobaric tests (PIO2 = 120 hPa reached in 10 min), and we have measured and compared the physiological variables described above at 5 min and at 30 min of the PIO2 = 120 hPa steady state. These two tests were performed in 18 subjects before two expeditions (Mera Peak for 7 subjects and Alpamayo for 11 subjects). During the expedition, AMS was scored daily. Thereafter, we have studied the relationships between the physiological variables measured at 5 and 30 min of the hypo/normobaric hypoxic tests and, rst, the maximal AMS scores and, second, the mean AMS scores in order to determine algorithms describing the individual susceptibility to AMS in terms of severity (maximal AMS scores) and level of mal-adaptation (mean AMS scores).

Methods Subjects Eighteen healthy subjects (1 female, 17 males) not acclimated or recently exposed to altitude participated in the study after the protocol was approved by the Grenoble University Ethics Committee. They signed an informed consent after medical examination. Their biometrical characteristics were given in Table 1. Body mass (P) was measured with an electronic balance (EC240 with E/03-E3300, Sauter, Ebingen, Germany) with an accuracy of 10 g. Body surface area (AD) was calculated using the equation of Du Bois and Du Bois (1916). Lean body mass and percentage of body fat content were calculated from skinfold thickness measured in two sites (scapula and triceps brachialis) with a Holtain calliper (Crymich, UK) following the equation of Lohman et al. (1975). Body mass index (BMI) was calculated as the ratio between body mass in kg and height in m2. General protocol Each subject was submitted in a randomized order to a hypobaric hypoxic test (HH) and to a normobaric hypoxic test (NH) at an ambient O2 partial pressure (PO2) equal to 120 hPa reached in 10 min (4,500 m) in a hypobaric chamber located in La Tronche (220 m, PB = 992 hPa). Thereafter, subjects took part in a high altitude expedition in the Himalayas for a rst group (n = 7) or in the Andes for the second one (n = 11). During the expeditions, AMS was scored daily using the Lake Louise self-report questionnaire. Thereafter, AMS maximal score attesting the severity of AMS and AMS mean score attesting the degree of mal-adaptation were calculated for each subject. These scores were then correlated to biometrical characteristics and to physiological data observed at times 5 and 30 min of the PIO2 = 120 hPa steady state during HH and NH tests, in order to determine the relationships between these variables and AMS scores. Hypoxic tests Each subject was submitted to an HH and to an NH test before the expeditions. These NH and HH tests have been previously described by Savourey et al. (2003). Briey, each experiment lasted 85 min, but the hypoxic exposure lasted 40 min. Experiments were realized in a random order 2-week apart in the morning after a standardized breakfast in a comfortable and controlled ambient climatic environment (Tdb = 22C,

123

196 Table 1 Biometrical characteristics of the subjects Subjects Age Body mass Height Arm skinfold Scapula skinfold (years) (kg) (m) thickness (mm) thickness (mm) S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 Mean SD 24 24 21 25 24 23 26 24 23 24 22 32 42 42 32 29 32 43 28.44 7.24 84.3 71.0 74.4 70.4 73.5 86.8 58.6 79.5 61.6 67.2 66.0 69.2 64.2 75.2 71.7 75.0 72.0 89.2 72.77 8.25 1.83 1.75 1.82 1.79 1.80 1.85 1.60 1.73 1.73 1.80 1.74 1.77 1.71 1.73 1.81 1.83 1.75 1.88 1.77 0.06 8.10 7.00 5.90 3.90 6.40 11.90 13.60 4.20 6.00 4.70 6.90 5.35 3.60 6.38 5.45 5.45 5.63 6.08 6.47 2.56 10.95 8.00 8.00 9.00 10.10 14.70 10.20 9.75 8.70 7.20 8.70 9.55 5.95 8.13 8.53 11.00 8.20 8.80 9.19 1.87 Lean body mass (kg) 69.99 59.94 63.43 60.16 61.60 69.28 45.90 67.78 51.86 57.90 55.34 58.39 56.06 63.84 60.98 62.86 61.25 75.88 61.25 6.94

Eur J Appl Physiol (2007) 100:193205

Body fat Surface content (%) area (m2) 17.0 15.5 14.8 14.5 16.2 20.2 21.7 14.8 15.8 13.9 16.2 15.6 12.6 15.1 14.9 16.2 14.9 14.9 15.8 2.1 2.05 1.84 1.96 1.87 1.92 2.10 1.60 1.93 1.72 1.87 1.76 1.85 1.75 1.89 1.91 1.96 1.87 2.15 1.89 0.13

Body mass index (kg m2) 25.17 23.18 22.46 21.97 22.68 25.36 22.89 26.56 20.58 20.74 21.80 22.08 21.96 25.12 21.89 22.40 23.51 25.24 23.09 1.72

RH = 50%), the subject being lightly clothed. Prior to the experiments, subjects were familiarized with the staff and the equipment of the laboratory and biometrical characteristics were measured during the medical examination. Each experiment was divided into three phases. During the rst phase, the subject was instrumented with the physiological sensors in the hypobaric chamber: he remained seated for 30 min in a comfortable chair breathing ambient air at ambient barometric pressure and listening to music through headphones to avoid disturbances from the environment. Then, during the second phase, the subject was connected to a system described below delivering dry air (O2 = 21%, N2 = 79%, CO2 = 0%) for 15 min. The mean of the data recorded during the last 5 min of this phase was the time 0 min of the HH and NH tests. Immediately after, hypoxia began (third phase of the experiment) by lowering PB (HH test) without change in the inspired dry air, or by lowering the O2 fraction at the level of the gas mixture system without change in PB (NH test) in order to reach in 10 min PO2 equal to 120 hPa. This level of PO2 was monitored by a mass spectrometer and maintained for 30 min by manipulating the vacuum (HH test) in the chamber or the O2 fraction (NH test). At the 40th min of hypoxia, the physiological measurements were stopped but hypoxia was maintained for a few minutes until the physician accompanying the subject sampled arterial blood from a radial artery and capillary blood from a puncture of a nger. Thereafter, the subject was de-equipped. As far as possible the subject was not aware of the type of the

test (HH or NH) because in each experiment he breathed a dry gas mixture, and the chamber was ushed with medical air gas both to maintain O2 = 21%, CO2 = 0% and to create the same noisy environment. The gas mixture system This system was located near the hypobaric chamber and was designed in our laboratory (Witt, Valence, France). It can create dry gas mixtures determined both in composition (N2, O2, CO2, or rare gas) in fraction and in speed of mixing by a computer. In our experiment, O2 and N2 were only used. The subject breathed the dry gas mixture injected into a 2 l balloon located inside the hypobaric chamber through a tube connected to a three-way Hans Rudolf valve and a low dead space face mask (Hans Rudolph, Kansas City, MO, USA). The gas mixture is continuously analyzed in the balloon using CO2 and O2 Servomex 1400 analyzers (Servomex, Crowborough, UK) and at the subjects mouth level by mass spectrometry (Mediex SX 200, VG Instruments, Cheshire, UK) previously calibrated with accurate gases (Air Liquide, Bonneuil, France). Measurements _ Tidal volume (VT), minute volume VE and respiratory frequency (f) were measured with an ultrasonic owmeter (fr 41, BDRL Flowmetrics, Birmimgham, UK) inserted between the face mask and the three-way

123

Eur J Appl Physiol (2007) 100:193205

197

valve, the dead space on the whole being 150 ml. The owmeter was calibrated before and after each test. These volumes were corrected to body and pressure saturated (BTPS). Duration of inspiration (TI), expiration (TE) were thereby calculated by the computer from owmeter data. End tidal fractions of O2 and CO2 FetO2 ; FetCO2 were measured at the mouths subject with the mass spectrometer in order to calculate the end tidal partial pressures of O2 and CO2 PetO2 and PetCO2 : The delay in response of the mass spectrometer was maintained constant during ascent by an additional differential pump. Heart rate (HR) was monitored by three thoracic electrodes and O2 blood saturation (SpO2) was measured by nger pulse oxymetry (Supermon 7210, Kontron, Watford, UK). O2 and CO2 arterial blood pressures (PaO2 and PaCO2), blood pH, O2 arterial saturation (SaO2) and content (CaO2) were measured from the arterial blood sampled with an arterial blood syringe placed in ice and immediately analyzed with a blood gas analyzer previously calibrated (Radiometer ABL 500, Copenhagen, Denmark). Moreover, nger capillary blood was also sampled (Autolet system and microcapillary tube) after the arterial blood sample to analyze haemoglobin (Hb) concentration in peripheral capillary blood. The expeditions at high altitude and AMS scoring A rst group of seven subjects spent 20 days in the Himalayas in order to climb Mera Peak (6,476 m) after trekking for 8 days from 2,800 m to the base camp (5,300 m). Mera Peak was reached by ve subjects. The second group of 11 subjects (10 males and 1 female) spent 20 days in the Andes (Peru) in order to climb Alpamayo (5,997 m) and Huascaran (6,768 m) after trekking for 5 days from 3,000 m to the base camp (4,800 m). The summits were reached by 9 subjects. No serious medical problem was encountered during these expeditions except AMS of various severity as described in the Results section. All the subjects were refrained to take medications for AMS or other. Acute mountain sickness was scored daily at 6 p.m. using the Lake Louise self-report questionnaire (Roach et al. 1993). The questionnaire was answered by the subjects themselves who were previously familiarized with this one. AMS scoring was realized during 20 consecutive days for the two groups between 2,800 and 5800 m for the rst group (Mera Peak) and between 3,000 and 5,800 m for the second one (Peru). After return, maximal AMS score (AMS max) dened as the highest daily AMS score and mean AMS score (AMS mean) dened as the sum of the daily AMS

scores divided by the number of days of the expeditions were determined for each subject. Calculations The hypoxic ventilatory response (HVR) was calculated at each min during the NH and HH test as follows : HVR _ DVE DSpO2

_ _ where DVE and DSpO2 are changes in VE (l min1) and in SpO2 (%) between values observed at each min of the tests and values observed at time 0. The hypoxic cardiac response (HCR) was also calculated at each min during the NH and HH test as follows : HCR DHR DSpO2

where DHR and DSpO2 are changes in heart rate HR (bpm) and in SpO2 (%) between values observed at each min of the tests and values observed at time 0. Peripheral blood O2 content (CpO2) was calculated as follows: CpO2 HbSpO2 1:34 where [Hb] (g/dl) is the capillary haemoglobin concentration, SpO2 (%) the peripheral O2 saturation measured by pulse nger oxymetry and 1.34 the oxyphoric power of 1 g of Hb. This calculation was justied by the fact that the measurement of CaO2 needs an arterial sample and because we hypothesized that this indirect measurement, easy to perform in eld conditions, could reect CaO2 and could be related to AMS (see Discussion section). Difference in alveolo-arterial partial pressure of O2 DPA - aO2 and CO2 DPA - aCO2 were calculated at the end of the tests as follows : DPA aO2 PetO2 PaO2 DPA aCO2 PetCO2 PaCO2 Statistical analysis Results are presented as means standard deviations. Physiological data were measured each min during the HH and NH test. However, data observed only at 15 and 40 min of the hypoxic tests (corresponding, respectively, to 5 and 30 min of the steady state of

123

198

Eur J Appl Physiol (2007) 100:193205

PIO2 = 120 hPa) were used to study the effect of time and to compare HH and NH tests at the same time. A two-way ANOVA with repeated measures (time and NH/HH tests) was used to perform this analysis. A Tukey post hoc test was used to locate the statistical signicance between means when a global statistical signicance was found. Thereafter, linear relationships between the variables observed at these two times and AMS scores (AMS max and AMS mean) were studied using, in a rst phase, single linear correlation analysis in order to locate variables related to AMS scores and then, in a second phase, multiple linear regressions were performed to establish the relationships between these selected variables and AMS scores. Statistica for Windows v6.0, (Statsoft France, Maisons-Alfort, France) was used to realize these statistical analyses. Null hypothesis was rejected at P 0.05.

Biometrical characteristics of the subjects and AMS scores Biometrical characteristics of the subjects are given in Table 1. No signicant linear relationship was found between the biometrical characteristics studied and either AMS max or AMS mean, especially for BMI (P = 0.50 and 0.52 for AMS max and AMS mean, respectively). However, AMS max and AMS mean tend to be related to the percentage of body fat content (R = 0.43, P = 0.07 and R = 0.43, P = 0.07, respectively). Physiological and biological variables measured or calculated during the hypoxic tests and AMS scores Means and standard deviations of the physiological variables observed at time 5 and 30 min of the steady state of the hypoxic tests (NH and HH tests) are described in Table 3. Biological variables measured or calculated at the end of the NH and HH tests are presented in Table 4. Physiological variables

Results AMS scores during the expeditions Maximal AMS scores (AMS max) and mean AMS scores (AMS mean) determined from the daily Lake Louise self-report questionnaires for each subject during the expeditions are given in Table 2. AMS max varied between 3 and 12, and AMS mean varied between 0.6 and 3.5 attesting a large AMS score variability among individuals. A linear relationship was found between AMS max and AMS mean scores (R = 0.78, P 0.0001).
Table 2 Maximal AMS scores (AMS max) and mean AMS scores (AMS mean) observed during the expeditions

Subjects AMS max S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 Mean SD 6 8 6 8 3 6 12 6 6 9 4 5 5 4 4 6 6 5 6.06 2.12

AMS mean 1.80 0.75 1.00 2.35 0.75 1.35 3.50 1.95 1.25 2.10 0.60 0.94 0.95 1.14 0.95 1.40 1.21 2.00 1.44 0.73

A large inter-individual variability is observed for each variable both at time 5 and 30 min during the NH or the HH tests (Table 3). No signicant time effect is observed, except for DSpO2 where a global signicant time effect is found (P < 0.05). PetCO2 tends to be lower in NH at 30 min compared to 5 min (P = 0.08). The comparisons between HH and NH at the same time (Table 3) show that, at time 5 min, f is higher in HH (P = 0.03), whereas VT ; TI ; PetCO2 ; SpO2 are lower in HH compared to NH (P < 0.05). At time 30 min, no statistical signicant difference is observed between HH and NH excepted for DSpO2 which is higher in HH (16.15% (4.59) vs. 13.33% (5.76), P < 0.05). Single linear correlations observed between the physiological variables at 5 and 30 min of the HH and NH tests and AMS max and AMS mean are presented in Table 5. This table reports only the variables for which a signicant (or very near of the statistical signicance) linear relationship is found. For all the other variables studied and not presented in this table, no signicant correlation was found. AMS max scores observed during the two expeditions are signicantly related in NH to PetO2 both at 5 and 30 min (R = 0.48, P = 0.04 and R = 0.61, P = 0.03, respectively), to DPA - aO2 and to CpO2 at 30 min (R = + 0.61, P = 0.03 and R = 0.53, P = 0.02, respectively). In HH, AMS max is only signicantly related to CpO2 at 30 min (R = 0.57, P = 0.01).

123

Eur J Appl Physiol (2007) 100:193205 Table 3 Cardio-ventilatory variables observed at times 5 and 30 min of the hypobaric (HH) and normobaric (NH) hypoxic test at PIO2 = 120 hPa

199

5 min f(min1) Df (min1) NH HH P NH HH P NH HH P NH HH P NH HH P NH HH P NH HH P NH HH P NH HH P NH HH P NH HH P NH HH P NH HH P NH HH P NH HH P 12.24 (3.80) 15.73 (4.64) 0.03 0.60 (2.75) 1.88 (6.11) NS 0.88 (0.22) 0.72 (0.25) 0.03 10.14 (1.51) 10.49 (2.59) NS 2.99 (0.98) 1.94 (0.65) 0.01 2.39 (0.89) 2.09 (0.87) NS 0.31 (0.07) 0.41 (0.20) 0.05 79.56 (11.94) 72.50 (6.58) 0.08 48.87 (5.53) 44.09 (6.38) 0.05 0.03 (0.15) 0.05 (0.21) NS 69.62 (9.95) 70.32 (9.91) NS 8.97 (7.93) 9.11 (8.10) NS 87.11 (4.81) 83.03 (4.49) <0.05 10.32 (7.35) 15.59 (3.99) <0.01 0.63 (0.79) 0.61 (0.55) NS

30 min 13.76 (4.47) 14.77 (4.17) NS 2.12 (3.25) 0.91 (2.82) NS 0.86 (0.34) 0.83 (0.37) NS 10.78 (1.67) 10.70 (1.93) NS 3.00 (1.16) 2.40 (1.25) NS 2.31 (1.17) 1.98 (0.84) NS 0.30 (0.08) 0.39 (0.17) 0.06 76.09 (11.61) 73.15 (7.16) 0.08 46.13 (6.61) 43.43 (6.02) NS 0.07 (0.22) 0.09 (0.17) NS 70.67 (12.07) 69.50 (12.07) NS 8.18 (10.06) 10.59 (12.15) NS 85.50 (4.84) 82.49 (4.39) 0.06 13.33 (5.76) 16.15 (4.59) <0.05 0.79 (0.78) 0.52 (0.74) NS

P for time effect NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS 0.08 NS NS NS NS NS NS NS NS F1 = 4.72, P = 0.04 17 (global effect for time)

VT (l) _ VE l

TI (s) TE (s) VT/TI (l s1) PetO2 (hPa) PetCO2 (hPa) HVR (l %1)

HR (bpm)

D HR (bpm)

SpO2 (%) DSpO2 (%)

HVR and HCR are, respectively, hypoxic ventilatory and cardiac responses to hypoxia

HCR (bpm %1)

NS NS

For AMS mean scores in NH, no signicant relationship is found with the physiological variables studied at 5 min whereas at 30 min, AMS mean scores are related to Df (R = 0.46, P = 0.05), HCR (R = 0.49, P = 0.04), CpO2 (R = 0.51, P = 0.03) and pH (R = 0.61, P = 0.04). In HH at 5 min, AMS mean scores are signicantly related to VE (R = 0.53, P = 0.02), TI (R = 0.50, P = 0.03), HR (R = 0.48, P = 0.04). In HH at 30 min, AMS mean scores are signicantly related to f (R = 0.58, P = 0.05) and to VT (R = 0.69, P = 0.01).

Biological blood variables Compared to NH, PaCO2 is lower in HH (46.3 hPa (6.5) vs. 52.2 hPa (4.2), P = 0.005) and SaO2 tends to be also lower (81.09% (7.76) vs. 85.48% (5.63), P = 0.07), whereas pH is higher in HH (7.45 (0.04) vs. 7.44 (0.04), P = 0.02). A linear relationship is found between CaO2 measured at the end of both HH and NH tests and CpO2 calculated at the end of the two tests as shown in Fig. 1 (R = 0.69, P 0.001).

123

200

Eur J Appl Physiol (2007) 100:193205

Table 4 Biological blood data observed or calculated at the end of the normobaric (NH) and hypobaric (HH) hypoxic tests Subject S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 Mean SD Mean SD P Test NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH PaO2 (hPa) 61.5 62.1 75.4 65.5 86.0 79.9 64.7 84.1 63.2 60.9 60.0 NV NV 60.1 72.3 56.0 63.8 62.4 58.9 46.4 82.4 70.1 53.2 51.0 61.2 62.2 NV 63.9 NV 54.3 53.1 NV 54.5 42.4 NV NV 65.0 10.3 61.4 11.1 NS PaCO2 (hPa) 55.4 51.5 55.1 52.2 47.5 32.5 44.4 32.2 55.4 47.2 54.5 NV NV 50.0 47.5 49.3 53.7 50.0 49.4 52.0 47.2 44.6 53.8 48.0 53.4 43.3 NV 41.2 NV 49.4 58.5 NV 55.1 50.8 NV NV 52.2 4.2 46.3 6.5 0.005 pH 7.42 7.44 7.41 7.41 7.44 7.51 7.49 7.58 7.42 7.45 7.43 NV NV 7.39 7.42 7.43 7.43 7.44 7.48 7.47 7.44 7.44 7.41 7.44 7.43 7.48 NV 7.48 NV 7.43 7.56 NV 7.41 7.41 NV NV 7.44 0.04 7.45 0.04 0.02 SaO2 (%) 84.50 85.10 90.80 86.30 94.50 93.30 87.30 96.20 87.40 NV 82.00 NV NV 80.10 89.40 81.80 87.30 83.30 84.80 72.00 92.60 89.30 77.75 71.80 84.40 77.20 NV 84.10 NV 70.90 75.10 77.10 78.90 76.50 NV 72.40 85.48 5.63 81.09 7.76 0.07 CaO2 (ml/dl) 14.18 15.62 19.33 17.39 18.05 19.77 17.70 17.84 16.14 NV 16.61 NV NV 11.97 16.36 14.85 16.69 15.47 15.81 13.02 18.12 17.55 15.55 14.70 17.70 16.40 NV 18.70 NV 16.30 16.20 18.50 14.90 15.80 NV 17.30 16.67 1.39 16.32 2.07 NS Hbc (g/dl) 12.20 13.30 15.40 14.70 13.80 15.40 14.70 13.50 14.10 14.10 14.70 14.10 11.05 11.00 13.30 13.20 13.90 13.50 13.60 13.20 14.20 14.30 14.85 15.40 14.65 15.62 16.40 18.82 16.39 16.25 15.50 18.40 13.80 15.68 16.70 15.81 14.51 1.51 14.81 1.90 NS CpO2 (ml/dl) 13.20 14.86 18.23 16.36 16.58 15.96 17.67 16.23 16.92 14.76 17.76 15.30 12.06 11.58 15.73 15.04 16.10 14.66 16.15 13.60 16.98 16.12 16.74 17.01 15.93 17.96 20.40 22.81 17.99 17.79 16.34 20.18 15.75 15.98 18.91 19.10 16.63 1.93 16.40 2.54 NS DPA aO2 (hPa) 3.50 11.68 13.40 6.53 3.92 3.10 15.27 5.34 5.79 2.39 31.61 NV NV 15.75 1.83 17.36 2.74 7.60 40.01 13.03 6.40 7.03 9.80 12.00 15.80 13.44 NV 22.32 NV 12.00 8.30 NV 24.00 31.25 NV NV 11.61 13.09 11.34 8.71 NS DPA aCO2 (hPa) 13.96 8.25 14.63 10.63 7.69 0.46 5.62 7.90 8.49 2.60 10.03 NV NV 6.10 0.73 6.48 6.25 4.94 4.49 2.09 7.40 3.74 7.20 3.60 3.70 12.81 NV 1.45 NV 6.26 8.30 NV 8.70 3.94 NV NV 6.63 5.42 3.12 5.73 NS

NV No value

No signicant linear relationship between AMS max scores and arterial blood analyses was found at the end of both NH and HH tests. AMS mean scores are signicantly related to pH (R = 0.61, P = 0.04) in NH and to capillary Hb in HH (R = 0.72, P = 0.01). Results of the multiple linear correlations Multiple linear relationships between physiological or biological variables and AMS scores (AMS max and AMS mean) are presented in Table 6.

For AMS max scores, a signicant multiple linear relationship is only found in NH at time 30 min with PetO2 and CpO2 (R = 0.77, P = 0.001). For AMS mean scores, no signicant multiple linear relationship is found in NH at time 5 min. Also in HH at time 5 min, no signicant multiple linear relationship is found despite single relationships were observed. However, AMS mean score in NH at time 30 min appears related to Df and HCR (R = 0.63, P = 0.02), to HCR and CpO2 (R = 0.67, P = 0.012) and, nally, to Df and HCR and CpO2 (R = 0.71, P = 0.017). In HH at

123

Eur J Appl Physiol (2007) 100:193205

201

Table 5 Single linear signicant correlations (or near the statistical signicance level) between AMS max or AMS mean scores and the physiological/biological variables AMS max 5 min
f(min ) Df (min ) VT (l) _ VE (l) TI (s) VT/TI (l s ) PetO2 (hPa) HVR (l %1) HR (bpm) HCR (bpm % ) DPA aO2 hPa CpO2 (hPa)
1 1 1 1

AMS mean 30 min 5 min

R = 0.45, P = 0.06

30 min
R Y R Y = = = = 0.58, P = 0.05 1.61 0.01f 0.46, P = 0.05 1.28 + 0.08 Df

NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH NH HH

R = 0.44, P = 0.07

R = 0.42, P = 0.09 R = 0.69, P = 0.01 Y = 1.25 + 0.23 VT R = 0.42, P = 0.09 R = 0.53, P = 0.02 Y = 2.99 0.15 VE R = 0.50, P = 0.03 Y = 0.36 + 0.56 TI R = 0.48, P = 0.04 Y = 0.79 + 0.086 PetO2 R = 0.55, P = 0.06 R = 0.61, P = 0.03 Y = 0.60 + 0.087 PetO2 R = 0.45, P = 0.06 R = 0.43, P = 0.08 R = 0.45, P = 0.06 R = 0.48, P = 0.04 Y = 1.05 + 0.036 HR R = 0.49, P = 0.04 Y = 1.81 + 0.46 HCR R = 0.61, P = 0.03 Y = 5.24 + 0.07 DPA aO2 NV NV NV NV R Y R Y = = = = 0.53, P = 0.02 15.85 0.59 CpO2 0.57, P = 0.01 13.94 0.48 CpO2 NV NV NV NV R Y R Y R = = = = = 0.51, P = 0.03 4.65 0.19 CpO2 0.43, P = 0.07 3.46 0.12 CpO2 0.61, P = 0.04

pH

NH HH

For the other physiological/biological variables not presented in this table, no signicant linear correlation was found

30 min, AMS mean score is only signicantly related to VT and to CpO2 (R = 0.60, P = 0.037).

Discussion This study shows that the duration of the hypoxic test (HH or NH) is important, as indicated by the different values observed at time 5 and 30 min, especially for DSpO2 for which a global statistical time effect is found (P = 0.04). In a same way, PetCO2 tends to be lower in NH at time 30 min compared to 5 min (P = 0.08). More variables were expected to be different according to the times of the tests. This is not surprising because, in our study, times 5 and 30 min correspond to the time of the hypoxic steady state (PIO2 = 120 hpa, 4,500 m) whereas these times correspond to times 15 and 45 min from the beginning of the tests. This procedure is necessary to control exactly the installation of the hypoxia in order to compare the NH and HH tests but, in contrast, could explain why DSpO2 only differed according to the duration of the tests since the steady

state is already obtained. Moreover, the large individual variability observed in the measurements (Table 3) could also be an explanation. This seems to suggest that a duration of 30 min for steady state hypoxic test is necessary to study the individual susceptibility to altitude illnesses. This point is also conrmed by the fact that single linear correlations between AMS max or mean scores and measurements at 5 and 30 min are somewhat different for the same test (HH or NH), as shown in Table 5. This study also shows that physiological and biological measurements are different following the hypoxic test used (NH or HH) for a same PIO2 = 120 hpa. Those differences are particularly clear at time 5 min and are characterized in HH compared to NH by a greater breathing frequency (P = 0.03), a lower tidal volume (P = 0.03) without signicant change in minute volume leading to a decrease in PetCO2 (P = 0.05), PetO2 (P = 0.08) and consequently in SpO2 (P < 0.05) without signicant change in HVR, HR and in HCR. These results attest that HH induces a greater dead space ventilation associated with a ventilatory

123

202
CpO 2
24.00 22.00 20.00 18.00 16.00 14.00 12.00 10.00 10.00 y = 0,8115x + 3,1359 R = 0,69 n = 30 P < 0.001

Eur J Appl Physiol (2007) 100:193205

12.00

14.00

16.00

18.00

20.00

22.00

CaO 2

Fig. 1 Linear correlation between arterial blood O2 content (CaO2, ml/dl) measured at the end of the normobaric (NH) and hypobaric (HH) hypoxic tests and peripheral O2 content (CpO2, ml/dl) calculated at the end of HH and NH tests

alkalosis. The arterial blood measurements performed at time 30 min reveal that these disturbances are still observed at the end of the HH tests. Indeed, a greater decrease in PaCO2 (P = 0.005), in SaO2 (P = 0.07) and a signicant increase in pH (P = 0.02) attest a greater hypoxemia, hypocapnia and arterial blood alkalosis in HH compared to NH at the end of the hypoxic tests. However, the physiological differences performed at
Table 6 Synthesis of the linear (single and multiple) relationships between AMS max, AMS mean scores and the physiological/biological variables of the NH and HH tests

time 30 min are less conclusive from a ventilatory point of view because breathing frequency, minute volume and PetCO2 are not signicantly different between HH and NH despite a signicant greater decrease in DSpO2 (P < 0.05) in HH. These results conrm the results described by Savourey et al. (2003) who reported that during similar hypoxic tests, HH induced a greater hypoxemia, hypocapnia, blood alkalosis and a lower arterial saturation, compared to NH. For these authors, the observed physiological differences could be related to the barometric pressure reduction including the expansion of the pulmonary water vapor and it was proposed to group these particular responses to acute hypobaric hypoxia under the term the specic response to hypobaric hypoxia. From this study which conrms this specic response, it appears that the physiological differences are mainly observed at the beginning of the test during the reduction of the barometric pressure and at the beginning of the steady state of hypoxia at PIO2 = 120 hPa. Indeed, physiological measurements differed essentially at time 5 min whereas at time 30 min, DSpO2 only differs signicantly. However, at this time, arterial blood measurements are more altered in HH attesting severe and delayed biological effects of the reduction of the barometric pressure despite a pseudo normalization of the ventilatory responses. These delayed effects could

AMS max NH 5 min R= Y= 30 min R = Y= R= Y= R= Y= R= Y= 0.48; P = 0.04 0.79 + 0.086 PetO2 0.61; P = 0.03 0.60 + 0.087 PetO2 0.61; P = 0.03 5.24 + 0.07 DPA aO2 0.53; P = 0.02 15.85 0.59 CpO2 0.77; P = 0.001 9.47 + 0.104 PetO2 0.68CpO2

AMS mean NS R = 0.46; P = 0.05 Y = 1.28 + 0.08 Df R = 0.49; P = 0.04 Y = 1.81 + 0.46 HCR R = 0.51; P = 0.03 Y = 4.65 0.19 CpO2 R = 0.63; P = 0.02 Y = 1.65 + 0.88 Df + 0.49 HCR R = 0.67; P = 0.012 Y = 4.61 + 0.40 HCR 0.17 CpO2 R = 0.71; P = 0.017 Y = 3.91 + 0.059 Df + 0.438 HCR 0.135 CpO2 R = 0.53; P = 0.02 _ Y = 2.99 0.15 VE R = 0.50; P = 0.03 Y = 0.36 + 0.56 TI R = 0.48; P = 0.04 Y = 1.05 + 0.036 HR Multiple correlations = NS R = 0.58; P = 0.05 Y = 1.61 0.01 f R = 0.69; P = 0.01 Y = 1.25 + 0.23 VT R = 0.60; P = 0.037 Y = 3.91 + 0.96 VT 0.199 CpO2

HH 5 min

NS

30 min R = 0.57; P = 0.01 Y = 13.94 0.48 CpO2

123

Eur J Appl Physiol (2007) 100:193205

203

have detrimental medical incidences and could explain the earlier and greater development of acute mountain sickness in hypobaric compared to normobaric hypoxia as reported by Roach et al. (1996). All the above observations also suggest indirectly that a hypoxic test needs at least a duration of 30 min of hypoxic steady state to take into account these biological changes. A longer duration could be interesting to know the time at which the physiological and/or biological variables reach or do not reach the same values in HH and NH. However, 30 min of steady state of hypoxia at PIO2 = 120 hPa seems, from our study, sufcient for practical consideration. As physiological/biological differences between NH and HH at a PIO2 = 120 hPa are conrmed by this study, it seems important to discuss the more appropriate hypoxic test that one must use for the determination of the individual susceptibility to acute mountain sickness (AMS). We have tried to answer this question by studying the linear relationships between AMS max or AMS mean scores determined during two different high altitude expeditions, and the physiological/biological variables observed at times 5 and 30 min of the HH and NH tests. Lake Louise selfreport questionnaires have been used to quantify AMS. This procedure is today well established and usually used. Moreover, Savourey et al. (1995b, 1997) have shown that this scoring system was related to other AMS questionnaires both in laboratory or in eld conditions. More original is the characterization of AMS from a quantitative and not from a qualitative point of view. Indeed, today subjects are divided in AMS+ or AMS subjects (Richalet et al. 1988; Rathat et al. 1992; Lanfranchi et al., 2005). This procedure is very interesting but, for medical considerations, it is also interesting for a subject found susceptible to AMS to know the potential AMS severity and the risk of mal-adaptation that he could have during an ascent. We have therefore dened the AMS max and the AMS mean score as representative of the AMS severity and the level of mal-adaptation, respectively. However, the signicant linear relationship found between AMS max and AMS mean scores (R = 0.78, P = 0.0001) shows that AMS mean is affected by the score of AMS max. This is not surprising because AMS mean score is dened as the mean of the daily AMS scores observed during the expedition. Today there is no consensus concerning physiological or biological markers of the mal-adaptation to altitude. Nevertheless we may suggest from a medical point of view that the analysis of those two scores taken together could be useful to elaborate pertinent medical advices. For example, an individual going to high altitude may have a severe

AMS (high AMSmax score) during the rst days of the stay but no more after, indicating a good adaptation attested by a low AMSmean score. Inversely, an individual presenting a moderate AMS (low AMSmax score) may present a worse adaptation (high AMSmean score). Thus, the medical analysis of the various combination of those two scores taken together, waiting other physiological/biological markers of the maladaptation, could improve the pertinence of the medical advices. As shown in Table 2, AMS max scores varied from 3 to 12 and AMS mean scores varied from 0.6 to 3.50 indicating a large variability among individual for the severity of AMS and for the degree of mal-adaptation, respectively. No signicant linear relationship was found between AMS max or AMS mean scores and the biometrical data of the subjects, especially for body mass index. This conrms the studies of Savourey et al. (1995) but not those of Hirata et al. (1989) and Kayser (1991). On the other hand, AMS max scores in NH at 30 min of PIO2 = 120 hPa appear signicantly and positively correlated to PetO2, DPA - aO2 (R = 0.61, P = 0.03 and R = 0.61, P = 0.03, respectively) and inversely correlated to CpO2 (R = 0.53, P = 0.02). The correlation between AMS max and DPA - aO2 could attest an impairment of the gas exchanges as reported by Ward et al. (2000) but this is unlikely because no relationship is found between PaO2, SaO2 and AMS max scores. Relationship found between CpO2 and AMS max scores both in NH and in HH tests at 30 min (P = 0.02 and P = 0.01, respectively) attests the importance of this quantitative variable to predict AMS max scores. Moreover, CpO2 appears as a good estimation of CaO2 because a linear relationship is found between these two variables as shown in Fig. 1 (R = 0.69, P 0.001). CpO2 depends rst on qualitative measurements represented by SpO2 and, second, on hemoglobin concentration, this last variable being indirectly related to extracellular volume changes because the duration of the hypoxic test is not sufcient to induce erythropoiesis. As reported by Ward et al. (2000), plasma volume reduction inducing an increased Hb may be considered as a factor of acclimatization to hypoxia before that erythropoietic response occurs. Finally, a multiple linear regression is found between AMS max scores in NH at 30 min and PetO2 and CpO2 (R = 0.77, P = 0.001), whereas AMS max scores are only related to CpO2 in HH test at 30 min. Differences are also observed between NH and HH tests at time 30 min to predict AMS mean scores. In NH and in HH, AMS mean scores are related to ventilatory variables (Df in NH, VT in HH) and to CpO2 whereas in NH, AMS mean scores are also

123

204

Eur J Appl Physiol (2007) 100:193205

related to HCR (P = 0.04). An exaggerated chemoreex activity could explain the relationships between these cardioventilatory variables and AMS means scores, as suggested by Lanfranchi et al. (2005). Consequently, CpO2 appears still to be an important variable to predict AMS mean scores in our experimental conditions. The multiple linear regression described to predict the level of the mal-adaptation to altitude with the intervention of cardio-ventilatory and CpO2 variables suggest that the development of this mal-adaptation is multifactorial as described in the Introduction section. As this study shows that not only the type of the hypoxic test (NH or HH) but also its duration must be taken into account to determine the individual susceptibility to AMS (severity and state of mal-adaptation), it is important from medical and practical points of view to propose a guidance for a future hypoxic test. First, it appears as an evidence that an NH test must be used for practical reasons. Second, the duration of this NH test must be at least of 30 min of steady state at a PIO2 = 120 hPa corresponding to a FIO2 = 12% with the measurements of PetO2, Df, VT, HCR and CpO2. Determination of the potential gravity of AMS (AMS max) could be done with the following algorithm : AMSmax 9:47 0:104 PetO2 0:68 CpO2 with PetO2 in hPa and CpO2 in ml/dl. And the potential level of the mal-adaptation with the following algorithm : AMSmean 3:91 0:059 Df 0:438 HCR 0:135 CpO2 with Df in min1, HCR in bpm %1 and with PetO2 in hPa and CpO2 in ml/dl. In a next step, it will be necessary to study the sensibility and the specicity of this test from a larger number of subjects. In conclusion, this study conrms and claries the physiological differences observed between HH and NH at a same PIO2. A new approach (quantitative) for the determination of the individual susceptibility to AMS is proposed in terms of potential gravity (AMS max) or level of mal-adaptation to high altitude (AMS mean). This procedure could optimize the medical advices given before a high altitude sojourn. Moreover, new hypoxic tests (HH or NH) are proposed with the physiological measurements needed (especially CpO2) and the algorithms for the determination of AMS max and mean scores are given. A 30 min NH test at a PIO2 = 120 hPa seems the best proposition because of practical considerations. However, the specicity and

the sensibility of this new test must be studied in the future for a denitive validation.
Acknowledgments The contributions of subjects, especially the subjects of the Mountain Club of the ESSA LYON-BRON (Peru expedition) and the subjects participating in the expedition to Mera Peak, are acknowledged as well as the technical assistance of L. Vachez-Collomb, N. Piccarreta, J. Denis, N. Clerc and V. Leroux. We particularly thank Dr. P. Arvers for his important contribution for the revision of the statistical analysis.

References
Bartsch P, Swenson ER, Paul A, Julg B, Hohenhaus E (2002) Hypoxic ventilatory response, ventilation, gas exchange, and uid balance in acute mountain sickness. High Alt Med Biol 3(4):361376 Bartsch P, Bailey DM, Berger MM, Knauth M, Baumgartner RW (2004) Acute mountain sickness: controversies and advances. High Alt Med Biol 5(2):110124 Du Bois D, Du Bois EF (1916) Clinical calorimetry. Tenth paper: a formula to estimate the approximate area if height and weight be known. Arch Intern Med 17:863871 Ge RL, Matsuzawa Y, Takeoka M, Kubo K, Sekiguchi M, Kobayashi T (1997) Low pulmonary diffusing capacity in subjects with acute mountain sickness. Chest 111(1):5864 Hackett PH (1999) The cerebral etiology of high-altitude cerebral edema and acute mountain sickness. Wilderness Environ Med 10(2):97109 Hackett PH, Rennie D, Hofmeister SE, Grover RF, Grover EB, Reeves JT (1982) Fluid retention and relative hypoventilation in acute mountain sickness. Respiration 43(5):321329 Hirata K, Matsuyama S, Saito A (1989) Obesity as a risk factor for acute mountain sickness. Lancet 2(8670):10401041 Hohenhaus E, Paul A, McCullough RE, Kucherer H, and Bartsch P (1995) Ventilatory and pulmonary vascular response to hypoxia and susceptibility to high altitude pulmonary oedema. Eur Respir J 8(11):18251833 Hu ST, Huang WY, Chu SC, Pa CF (1982) Chemoreexive ventilatory response at sea level in subjects with past history of good acclimatization and severe acute mountain sickness. In: Brendel W, Zink RA (eds) High altitude physiology and medecine. Springer, New York, pp 2832 Hyers TM, Scoggin CH, Will DH, Grover RF, Reeves JT (1979) Accentuated hypoxemia at high altitude in subjects susceptible to high-altitude pulmonary edema. J Appl Physiol 46(1):4146 Kayser B (1991) Acute mountain sickness in western tourists around the Thorong pass (5400 m) in Nepal. J Wilderness Med 2:110117 King AB, Robinson SM (1972) Ventilation response to hypoxia and acute mountain sickness. Aerosp Med 43(4):419421 Lanfranchi PA, Colombo R, Cremona G, Baderna P, Spagnolatti L, Mazzuero G, Wagner P, Perini L, Wagner P, Perini L, Wagner H, Cavallaro C, Giannuzzi P (2005) Autonomic cardiovascular regulation in subjects with acute mountain sickness. Am J Physiol Heart Circ Physiol 289(6):H2364H2372 Loeppky JA, Icenogle MV, Maes D, Riboni K, Hinghofer-Szalkay H, Roach RC (2005) Early uid retention and severe acute mountain sickness. J Appl Physiol 98(2):591597 Lohman TG, Boileau RA, Massey BH (1975) Prediction of lean body mass in young boys from skinfold thickness and body weight. Hum Biol 45:245262

123

Eur J Appl Physiol (2007) 100:193205 Mathew L, Gopinathan PM, Purkayastha SS, Sen Gupta JS, Nayar HS (1983) Chemoreceptor sensitivity and mal adaptation to high altitude in man. Eur J Physiol 51(1):137144 Milledge JS, Thomas PS, Beeley JM, English JSC (1988) Hypoxic ventilatory response and acute mountain sickness. Eur Respir J 1(10):948951 Milledge JS, Beeley JM, Broome J, Luff N, Pelling M, Smith D (1991) Acute mountain sickness susceptibility, tness, and hypoxic ventilatory response. Eur Respir J 4(8):10001003 Moore LG, Harisson GL, McCullough RE, McCullough RG, Micco AJ, Tucker A, Weil JV, Reeves JT (1986) Low acute hypoxic ventilatory response and hypoxic depression in acute altitude sickness. J Appl Physiol 60(4):14071412 Rathat C, Richalet JP, Herry JP, Larmignat P (1992) Detection of high-risk subjects for high altitude diseases. Int J Sports Med 13(Suppl 1):S76S78 Richalet JP, Keromes A, Dersch B, Corrizi F, Medhioui H, Pophillat B, Chardonnet H, Tassery F, Herry JP, Rathat C, Chaduteau C, Darnaud B (1988) Caracteristiques physiologiques des alpinistes de haute altitude. Sci Sport 3:89108 Roach RC, Bartsch P, Hackett PH, Oelz O (1993) The Lake Louise acute mountain sickness scoring system. In: Sutton JR, Coates G, Houston C (eds) Hypoxia and molecular medicine, Lake Louise, Alberta, Canada. Queen City Printers, Burlington, pp 272274 Roach RC, Loeppky JA, Icenogle MV (1996) Acute mountain sickness: increased severity during simulated altitude compared with normobaric hypoxia. J Appl Physiol 81(5):1908 1910 Robinson SM, King AB, Aoki V (1971) Acute mountain sickness: reproducibility of its severity and duration in an individual. Aerosp Med 42:706708

205 Savourey G, Moirant C, Eterradossi J, Bittel J (1995a) Acute mountain sickness relates to sea-level partial pressure of oxygen. Eur J Appl Physiol 70(6):469476 Savourey G, Guinet A, Besnard Y, Garcia N, Hanniquet AM, Bittel J (1995b) Evaluation of the Lake Louise acute mountain sickness scoring system in a hypobaric chamber. Aviat Space Environ Med 66(10):963967 Savourey G, Guinet A, Besnard Y, Garcia N, Hanniquet AM, Bittel J (1997) Are the laboratory and eld conditions observations of acute mountain sickness related? Evaluation of the Lake Louise acute mountain sickness scoring system in a hypobaric chamber. Aviat Space Environ Med 68(10):895899 Savourey G, Launay JC, Besnard Y, Guinet A, Travers S (2003) Normo- and hypobaric hypoxia : are there any physiological differences? Eur J Appl Physiol 89(2):122126 Singh I, Khanna PK, Srivastava MC, Lal M, Roy SB, Subramanyam CS (1969) Acute mountain sickness. N Engl J Med 280(4):175184 Sutton JR, Bryan AC, Gray GW, Horton ES, Rebuck AS, Woodley W, Rennie ID, Houston CS (1976) Pulmonary gas exchange in acute mountain sickness. Aviat Space Environ Med 47(10):10321037 Viswanathan R, Subramanian S, Lodi ST, Radha TG (1978) Further studies of pulmonary oedema of high altitude. Abnormal responses to hypoxia of men who had developed pulmonary oedema at high altitude. Respiration 36(4):216 222 Ward P, Milledge JS, West JB (2000) Acute mountain sickness. In: High altitude medicine and physiology. Arnold Edts, London, pp 215231

123