Abstract
Abstract
The importance of arterial oxygen saturation for the prediction of acute mountain sickness (AMS) is still a matter of debate. Reasons for discrepancies may result from varying laboratory or field conditions and their interactions. Thus, we analyzed data from our prior high-altitude studies, including participants of a broad range of age of both sexes (20 males and 20 females, aged between 20 and 67 years) under strictly standardized conditions of pre-exposure and acute exposure to real high altitude (3480 m). A set of resting cardiovascular, respiratory, hematological, and metabolic variables were recorded at high altitude (Testa Grigia, Plateau Rosa, 3480 m; Swiss-Italian boarder) after performing pretests at low altitude (Innsbruck, 600 m, Austria). Our analyses indicate that (1) smaller changes in resting minute ventilation (VE) and a larger decrease of peripheral oxygen saturation (SpO2) during the first 3 hours of acute exposure to high altitude were independent predictors for subsequent development of AMS (90% correct prediction), (2) there are no differences of responses between sexes, and (3) there is no association of responses with age. Considering the independent effects of both responses (VE and SpO2) may be of clinical/practical relevance. Moreover, the presented data derived from a broad age range of both sexes might be of interest for comparative purposes.
Introduction
Alarge number of studies tried to establish differences in physiological responses between subjects developing acute mountain sickness (AMS) and subjects staying healthy when acutely exposed to high altitude. Various physiological parameters have been suggested as being potentially explanatory for the susceptibility to AMS, that is, ventilatory, cardiovascular, hematological, and metabolic parameters (Roach et al., 1998; Burtscher et al., 2004; Karinen et al., 2010; Richalet et al., 2012; Spliethoff et al., 2013; Ding et al., 2014; Faulhaber et al., 2014; Sutherland et al., 2017). Arterial or peripheral oxygen saturation (SaO2, SpO2) has been emphasized as being the most promising predictor variable by several (Roach et al., 1998; Burtscher et al., 2004; Karinen et al., 2010; Richalet et al., 2012; Faulhaber et al., 2014), but not all (Grant et al., 2002; O'Connor et al., 2004) studies. Reasons for the discrepancies may result from varying laboratory or field conditions and their interactions, timing of measurements, pre-exposures to high altitude, selected study populations, etc. (Burtscher et al., 2004). Among the points listed, the timing of the measurement due to the circadian rhythm of SpO2, and a different degree of acclimatization of the subjects are important causes of error (Tannheimer et al., 2017). We believe that the nonstandardized measurement of SpO2 and the inconsistent study situation are the main causes of error. Due to periodic breathing, SpO2 measurements at high altitude are considerably more complex than at sea level. The studies by Grant et al. (2002) and O'Conner et al. (2004) are examples of how limited and problematic SpO2 determination at high altitude can be. Therefore, we intended to arrange a unique sample from our prior high-altitude studies, including participants of a broad range of age of both sexes, applying strict inclusion and exclusion criteria, and homogenous conditions of pre-exposure and acute exposure to real high altitude (3480 m) (Burtscher et al., 2018). Based on the availability of a relatively large set of physiological data collected at low and high altitude, we expected to obtain a clearer picture on the predictive importance of SpO2 for AMS development and other potentially explanatory variables. In addition, such data derived from a broad age range of both sexes might be of interest for comparative purposes.
Methods
Participants
Data sets of 40 adult subjects (20 males and 20 females, aged between 20 and 67 years) who participated in former placebo-controlled [published, e.g., Burtscher et al. (1998) and unpublished] studies of our laboratory have been analyzed [see also Burtscher et al. (2018)]. For those studies, physically active and healthy participants with a history of headache at high altitude, who did not take any medication, and who were not exposed to altitudes higher than 2000 m at least 1 month before the experiments were recruited by word-of-mouth recommendation primarily around Innsbruck (600 m), Austria. Experiments have been performed at the same altitude and location (Testa Grigia, Plateau Rosa, 3480 m; Swiss-Italian boarder) after performing similar pretests at low altitude (Innsbruck). In this study, only subjects without pharmacological pretreatment (placebo groups) have been included. Data sets of 20 age-matched male participants (out of 27) were assigned to 20 available complete data sets of females (Table 1). Study participants were transported by bus and cable car to high altitude where they stayed for at least 24 hours. Submaximal exercise testing was performed between 2 and 5 hours after arrival at high altitude. AMS scores were assessed in the evening of the arrival day and the next morning. All subjects had a light meal before ascending by cable car, had the same dinner after exercise testing at altitude, and slept for about 8 hours. Nonalcoholic beverages (primarily tea and water) were freely available. Participants provided written informed consent, and all studies were approved by the Ethical Commission of the Medical University of Innsbruck.
Characteristics of Participants
Measurements
Beside medical routine examination, medical history, lifestyle characteristics, and anthropometric data, selected resting cardiovascular, respiratory, hematological, and metabolic variables were recorded. Measurements were performed after medical routine examination in the late afternoon at low altitude and 2–3 days later at high altitude (also in the late afternoon about 3 hours after arrival). Resting measurements were followed by submaximal exercise tests, and the results of the latter have been recently reported (Burtscher et al., 2018). Subjects were advised not to perform intense exercise during the 2 days before measurements and avoid heavy meals for at least 4 hours before assessment. Peripheral oxygen saturation (SpO2), heart rate (HR) (by pulsoximetry; Onyx, NONIN), and systolic and diastolic systemic blood pressure (BP; OMRON R3, Japan) were recorded after at least 10 minutes in a sitting position. SpO2 values were recorded for 2–3 minutes until the measurement value stabilized. Arterial partial pressure of oxygen and carbon dioxide (PaO2, PaCO2) and pH (AVL OPTI, AVL Scientific Corporation) were measured from arterialized capillary blood (earlobe), hemoglobin values (Hb), and blood glucose (BG) from venous blood samples. Resting minute ventilation (VE) and the respiratory exchange ratio (RER) (Oxycon Alpha, Jaeger, Germany) were measured during 10–15 minutes in a sitting position, and reported values represent 3-minute averages taken when VE had stabilized. Rate-pressure product was calculated as HR × systolic BP and arterial oxygen content (CaO2) as (Hb × 1.34 × SpO2) + (0.0031 × paO2). The alveolo-arterial oxygen difference (A-a O2 diff) was calculated as (FiO2 × (pB − pH2O)) − (pCO2/RER) + (pCO2 × FiO2 × (1 − RER)/RER) − paO2. (FiO2: inspiratory oxygen fraction; pB: barometric pressure).
The Lake-Louise-Score (LLS) is a self-assessment questionnaire considering five main symptoms (headache, nausea, dizziness, fatigue, and difficulty sleeping), each rated with a score from 0 to 3 (0 for no discomfort, 1 for mild symptoms, 2 for moderate, and 3 for severe symptoms) (Roach et al. 1993). The LLS was recorded in the evening and the next morning before breakfast. The highest LLS was used for AMS diagnosis. AMS was diagnosed when the symptom headache was present together with at least one other symptom and the total score was 3 or higher (Roach et al., 1993).
Statistics
Data are presented as means (±standard deviation, SD) or frequencies. Physiological data were normally distributed. Chi-squared tests were used to compare differences in frequencies, and paired t-test was applied to compare differences of characteristics and physiological responses to high altitude within sexes. A mixed-design ANOVA was performed to analyze differences of physiological responses to altitude exposure between sexes. Pearson correlation coefficients were calculated for the relationship between variables. Multivariate step-wise logistic regression analysis was applied to detect potential predictors for the development of AMS (binary independent variable). Anthropometric characteristics (age, body mass, height) and all physiological variables measured (Table 2) were considered potential predictors (independent variables). Statistical analyses were conducted by PASW Statistics 24 (IBM, Austria). A p-value of <0.05 indicates statistical significance.
Resting Physiological Data of Males and Females at Low Altitude (600 m) and After Acute Exposure to High Altitude (3480 m)
A-a O2 diff, alveolo-arterial oxygen difference; BG, resting blood glucose level; BP, systemic blood pressure; CaO2, arterial oxygen content; HA, high altitude; LA, low altitude; paO2, arterial partial pressure of oxygen; paCO2, arterial partial pressure of carbon dioxide; SD, standard deviation; SpO2, oxygen saturation (measured by finger pulse oximetry).
Results
These physiological data taken at low and acute high altitude are derived from a sex-matched sample of a broad age range (20–67 years). With regard to baseline characteristics, sex differences exist for height, body mass, and body mass index (Table 1).
Changes in the physiological responses from low to high altitude did not differ between sexes (Table 2). Information on the time of menstrual cycle was only available for about half of all females but there was no indication of a potential influence of the menstrual cycle on VE. This observation is in line with that reported from Muza et al. (2001). Significant changes from low to high altitude within both sexes occurred for HR, rate pressure product, VE, SpO2, paO2, paCO2, and BG. Systolic BP and pH changes were only different among males, and Hb and A-a O2 changes were only different among females. No association was found between age and any of the variables assessed.
Thirteen (7 males and 6 females) out of the 40 participants developed AMS (AMS+), and 27 people remained free of AMS (AMS−). At this specific elevation (3480 m), an SpO2 value of 87% was the best predictive value for identifying AMS susceptibility; 92% of AMS (+) subjects had SpO2 values below 87%, and 85% of AMS (−) subjects had SpO2 values above 86%. Physiological responses to high altitude in AMS (+) and AMS (−) subjects are shown in Table 3. Significantly different changes between these subgroups were found for HR, diastolic BP, VE, SpO2, CaO2, and paCO2. Compared with AMS (−), AMS (+) subjects showed a more pronounced increase of HRs, a decrease of diastolic BP, lower VE increase, more pronounced SpO2 and CaO2 decrease, and a diminished decrease of paCO2 values when acutely exposed to high altitude.
Resting Physiological Data of AMS (−) and AMS (+) Subjects at Low Altitude (600 m) and After Acute Exposure to High Altitude (3480 m)
AMS, acute mountain sickness; FBG, fasting blood glucose level.
Correlation analyses revealed significant relations between changes (low vs. high altitude) of Hb and CaO2 (r = 0.9), SpO2 and CaO2 (r = 0.6), VE and SpO2 (0.4), VE and paCO2 (−0.3), SpO2 and systolic BP (0.4) and diastolic BP (0.5), A-a O2 diff and paO2 (−0.7), and paO2 and paCO2 (−0.4).
Multivariate logistic regression analysis revealed that smaller VE changes (OR 1.9; 95% CI 1.1–3.7) and a larger decrease of SpO2 (OR 2.5; 95% CI 1.2–5.4) when acutely exposed to high altitude were highly predictive (90% correct prediction) for the development of AMS (Table 4).
Results of Multivariate Logistic Regression Analysis
Discussion
We present a set of resting cardiovascular, respiratory, hematological, and metabolic responses to acute high altitude derived from a sex-matched sample of a broad age range. The findings indicate that (1) smaller changes in resting VE and a larger decrease of SpO2 during the first 3 hours of acute exposure to high altitude were independent predictors for subsequent development of AMS, (2) there are no differences of responses between sexes, and (3) there is no association of responses with age. Observations (2) and (3) are in accordance with findings reported by Schneider et al. (2002).
Cardiorespiratory responses to acute high-altitude exposure are consequences of the reduced pO2 at altitude and countermeasures to maintain oxygen supply to tissues. In this regard, the present findings are well in accordance with the literature (Burtscher et al., 1998, 2004; Richalet et al., 2012; Richalet and Lhuissier 2015; Boos et al., 2016). Reduced BG values at high altitude are hard to interpret since they do not represent fasting values but may at least partly be explained by enhanced carbohydrate oxidation during the short-term high-altitude exposure (Goto et al., 2015). The lack of differences between sexes, at least regarding cardiovascular and ventilatory responses to acute exposure to high altitude or hypoxia, is consistent with earlier and recent studies as well (Muza et al., 2001; Boos et al., 2016, 2017). For instance, Muza et al. (2001) demonstrated a nearly identical time course of ventilatory acclimatization in female lowlanders who rapidly ascended to 4300 m compared with reports of male lowlanders. Boos et al. (2016) recently investigated various cardiopulmonary responses to acute normobaric hypoxia (simulated 4800 m) and did not detect any sex differences. In contrast, the same authors report differences between sexes with regard to HR variability indices when acutely exposed to various altitudes, which, however, were not predictive for AMS development (Boos et al., 2017). In line with our findings, similar resting (and exercising) cardiorespiratory responses to acute hypoxia have been reported in young and older subjects of a large cohort (Lhuissier et al., 2012; Richalet et al., 2012; Richalet and Lhuissier 2015). The only association between age and physiological responses found by these authors was a somewhat lesser deoxygenation in aging men during acute high-altitude exposure, probably due to their increased ventilatory response. Although this was not confirmed by our data, a slightly (not significant) lesser deoxygenation at high altitude was seen in males compared with females (Table 2).
The observed AMS incidence corresponds well to that previously reported by our research group for similar altitudes (Mairer et al., 2009). It is worth mentioning that in this and the present study “sleeping difficulties” was included for AMS diagnosis, which is no longer considered in the updated LLS (Roach et al., 2018). In addition, in the present study, neither age nor sex was associated with the AMS incidence, also being in accordance with other studies (Horiuchi et al., 2018; Wu et al., 2018). New and noteworthy is the finding that not only a pronounced desaturation but also a low ventilatory response during the first hours at high altitude represents an independent predictor of the subsequent development of AMS (Table 3).
The level of oxygen desaturation has been repeatedly reported as an important predictor of AMS, whereas the low ventilatory response has been considered a cause of desaturation rather than representing an independent AMS predictor (Roach et al., 1998; Burtscher et al., 2004; Richalet et al., 2012). However, correlation analysis revealed an only moderate relationship between changes (low vs. high altitude) of VE and SpO2 (r = 0.4). With regard to desaturation when acutely exposed to high altitude, a recent meta-analysis convincingly confirmed the significant association between SpO2 and the risk of developing AMS (Guo et al., 2014). It is well accepted that hypoxia represents the primary AMS trigger. The nonsignificant correlation between SpO2 and paO2 may be surprising but becomes understandable based on the sigmoid shape of the oxy–hemoglobin dissociation curve, which is additionally affected by temperature, pH, 2,3 diphosphoglycerate, and paCO2. Although paO2 represents a good indicator for pulmonary gas exchange, SpO2 is a more appropriate measure for systemic oxygen delivery to tissues and its association with AMS development. Unlike SpO2, the hypoxic ventilatory response (HVR), at least when measured at sea level before going to high altitude, may be a good determinant neither of VE at altitude nor of the susceptibility to AMS (Milledge et al., 1988; Bärtsch et al., 2002). It is rather the failure to increase HVR after arrival at altitude accompanied by impaired gas exchange accounting for the more severe hypoxemia in AMS (Bärtsch et al., 2002). The HVR during acute high-altitude exposure may be affected by several conditions such as exercise, diet, medication, etc. and represents the beginning of the acclimatization process. Although SpO2 values during early high-altitude exposure constitute the most important and easy-to-determine predictor, the independent contribution of a low ventilatory response at altitude could make sense when considering potential pathophysiological mechanisms for the development of AMS. Although we are far from fully understanding the mechanisms of individual AMS development, acute hypoxia and its effects on the brain are accepted triggers of AMS. Lower minute VE may be associated not only with more severe hypoxia but also with higher paCO2 levels, as indicated by the negative correlation shown between paO2 and paCO2. Changes of these blood gas values may affect cerebral hemodynamics and cerebral oxygen delivery with a potential impact on the association with AMS (Wolff, 2000; Bian et al., 2014; Liu et al., 2017).
Moreover, both the level of hypoxia and the hyper VE-related hypocapnia contribute to high-altitude diuresis (Hildebrandt et al., 2000), which has been suggested to prevent or improve AMS symptoms (Bärtsch et al., 1991).
The A-a O2 diff decreased or at least tended to decrease at high altitude, which was expected due to the lower inspired partial pressure of oxygen (Wagner et al., 1987). No decrease or even widening of the A-a O2 diff in hypoxia would indicate shunting, VE –perfusion mismatch, or diffusion limitation.
The practical implications of the present findings may be twofold: First, they help to explain why SpO2 values are not always excellent AMS predictors and second, they emphasize the preventive effects of hyperventilation other than only increasing oxygen saturation when acutely exposed to high altitude.
Study limitations to be mentioned involve the retrospective analysis of data, the relatively small sample size, and the selection of a limited number of potential predictors, which does not entirely exclude a residual confounding bias. Strengths of the study are the sex-matched sample of a broad age range and, in particular, the very strict control of pre-exposure and exposure conditions.
In conclusion, our findings suggest that smaller VE changes and larger decreases of SpO2 during acute high-altitude exposure are independent predictors for the subsequent development of AMS. Consideration of these independent effects of both responses are of clinical/practical relevance; they can be modulated, for example, by inducing ventilatory stimulation at acute high altitude by hypoxia pre-exposures, and/or the avoidance of ventilatory depression, for example, by alcohol. Moreover, the presented data derived from a broad age range of both sexes might be of interest for comparative purposes.
Footnotes
Author Disclosure Statement
No competing financial interests exist.