Regional Differences in the Cerebral Blood Flow Velocity Response to Hypobaric Hypoxia at High Altitudes

Abstract

Symptoms of acute mountain sickness (AMS) may appear above 2,500 m altitude, if the time allowed for acclimatization is insufficient. As the mechanisms underlying brain adaptation to the hypobaric hypoxic environment are not fully understood, a prospective study was performed investigating neurophysiological changes by means of near infrared spectroscopy, electroencephalograpy (EEG), and transcranial doppler sonography at 100, 3,440 and 5,050 m above sea level in the Khumbu Himal, Nepal. Fourteen of the 26 mountaineers reaching 5,050 m altitude developed symptoms of AMS between 3,440 and 5,050 m altitude (Lake-Louise Score ≥ 3). Their EEG frontal beta activity and occipital alpha activity increased between 100 and 3,440 m altitude, i.e., before symptoms appeared. Cerebral blood flow velocity (CBFV) in the anterior and middle cerebral arteries (MCAs) increased in all mountaineers between 100 and 3,440 m altitude. During further ascent to 5,050 m altitude, mountaineers with AMS developed a further increase in CBFV in the MCA, whereas in all mountaineers CBFV decreased continuously with increasing altitude in the posterior cerebral arteries. These results indicate that hypobaric hypoxia causes different regional changes in CBFV despite similar electrophysiological changes.

INTRODUCTION

Nowadays an ever increasing number of people travel to high-altitude destinations for recreation or work. The incidence of acute mountain sickness (AMS) during an ascent on foot to altitudes between 2,800 and 5,336 m varies between 43% and 69% in different population studies. ¹ The incidence of possibly life-threatening forms of AMS-like high-altitude cerebral edema (HACE) and high-altitude pulmonary edema is much lower than for AMS: estimates range between 0.1% and 4.0%. ² Altitude-related illnesses have contributed to the death of 17% of British climbers attempting peaks > 7,000 m. ³

The main risk factors for developing high-altitude illness include the rate of ascent, individual susceptibility, and the altitude reached (sleeping altitude). ² Other risk factors are permanent residence < 900 m. ¹ Physical fitness is not a contributing factor, and people aged > 50 years may be even less susceptible. ² The reasons for the considerable variation in individual susceptibility are unclear, but most likely diverse interactions between genetic and environmental factors have a major role. ¹ The Lake-Louise Consensus Group defined AMS as the presence of headache in an unacclimatized person who has recently arrived at an altitude above 2,500 m in addition to one or more of the following: gastrointestinal symptoms (anorexia, nausea, or vomiting), insomnia, dizziness, and lassitude or fatigue. ¹ AMS appears 4 to 36 hours after ascent to high altitudes. ¹

It is not fully understood which factors are important for a good or poor adaptation to high altitude and how the brain reacts to these different states. Furthermore, it remains unclear which factors allow AMS to evolve directly into HACE. In this prospective study, we investigated early physiological and pathophysiological changes that occur at high altitudes. They include mechanisms of adaptation to high altitude such as brain activity measured by electroencephalography (EEG) and cerebral blood flow velocity (CBFV) measured by transcranial doppler sonography (TCD). Blood pressure, peripheral O₂ saturation (SpO₂), endtidal CO₂ (EtCO₂), and brain oxygen saturation, using near infrared regional saturation (NIRS), were measured as cofactors during a joint trekking trip to Everest Base Camp in the Khumbu Himal, Nepal. We hypothesized that the CBF increases because of the known decrease of peripheral and central (brain) oxygenation as compensatory responses to restore the blood and oxygen supply to the brain. Beginning brain hypoxia may lead to an increase of neuronal cell activity, resulting in an increased EEG background rhythm.

MATERIALS AND METHODS

Thirty-two healthy volunteers (12 females, 20 males; 43.5 ± 2.2 years; range 19 to 66 years; all hobby mountaineers) were examined during a joint, 22-day-long, high-altitude trekking tour in Sagarmatha National Park, Nepal. All participants were neurologically normal and had no history of neurological or psychiatric disorders, head trauma, or drug abuse. The most recent exposure to high altitude > 2,500 m had ended at least 6 months prior to the study.

The route began in Lukla (2,160 m) and ended at the so-called Silver Pyramid (Italian–Nepali Research Center Ev-K2-CNR, Lobuche, 5,050 m), which is located near Everest Base Camp. One group reached 5,050 m via Gokyo Ri (5,340 m), the other via Island Peak (6,200 m). The rates of ascent were as follows: from 2,860 to 3,440 m in 2 days and from 3,440 to 5,340 m in 4 days or 6,200 m in 6 days and back to 5,050 m in 2 to 3 days. EEG and TCD were performed at baseline level (100 m), at 3,440 m and at the Silver Pyramid 5,050 m. SpO₂, EtO₂, and Lake-Louise AMS Score were evaluated twice daily by an expedition doctor. Severe AMS was treated immediately according to the guidelines of the Wilderness Medical Society. A mountain guide accompanied and observed those mountaineers who descended to a lower altitude because of AMS.

The Lake-Louise Score was used to ascertain the presence of symptoms of AMS. It contains a self-rating and a foreign-rating scale (recorded by expedition doctor), which ranges from 0 to 29. ¹ Mountaineers with a score ≥ 3 were considered symptomatic if they had headache and one other AMS symptom.

Blood pressure was measured using the Riva–Rocci method, always by the same researcher and on the right arm, and after resting in a sitting position for at least 10 minutes. Systolic and diastolic blood pressure were recorded.

Oxygen saturation was measured using a finger clip on the right index finger, and EtCO₂ (Oridion Oxymetry and Capnography, Needham, MA, USA) by nasal probe. Measurements were made after resting in a sitting position for at least 10 minutes.

An INVOS 5,100 cerebral oximeter (Somanetics, Troy, MI, USA) was used to measure changes in regional cerebral oxygenation (rSO₂). Near infrared light (730 and 805 nm) is emitted through the forehead of the skull. Once the light passes through the cerebral cortex, the light that returns is detected at two distances from the light source (at 3 and 4 cm). The spectral absorption of blood in the brain can be determined on the basis of this principle and is defined as rSO₂. Before measuring, the skin was cleaned with a prewrapped skin-prep pad. Then two sensors were applied on the right and left sides of fronto-temporal region in accordance with the manufacturer's recommendations. To minimize the influence of extracerebral light, the forehead was covered with a black kerchief during the recording procedure. The rSO₂ data were recorded after a resting time of 3 minutes.

TCDs (DWL Elektronische Systeme, Sipplingen, Germany) were performed bilaterally of the middle cerebral artery (MCA) at 50-mm depth, bilaterally of the anterior cerebral artery (ACA) at 65-mm depth, and bilaterally of the posterior cerebral artery (PCA) at 65-mm depth using a 2-MHz sonde placed over the temporal scale. To avoid the well-known inter-rater variability in repeated TCD measurements, the same examiner performed all TCDs. To obtain the best possible results, the values were recorded after waiting for at least 1 minute for stable signals. All TCD measurements were made at 3,440 and 5,050 m the next day after arrival. As the study design focused on the assessment of relative changes in cerebral blood velocity, the baseline measurements at 100 m were set at 100%, and the relative changes were calculated.

EEGs (SIGMA Medizintechnik, Thum, Germany) were recorded for 10 minutes with 24 silver/silver chloride electrodes placed according to the 10 to 20 system using a sampling rate of 256 Hz; resistance was kept < 10 kOhm. Data analysis was performed using Fast-Fourier transformation of 15 × 4 seconds artefact-free EEGs. Constant vigilance during recording was ensured through irregular blinking. At 3,440 and 5,050 m, the EEG was performed the morning after arrival. FP2, FP1, F3, F4, and FZ were considered the frontal electrodes, and O1, O2, P3, P4, and PZ as the occipital electrodes. As the relative changes in the frontal background activity (beta band) and delta activity, as well as in occipital background activity (alpha band) and delta activity were of interest, the percentage changes of power activity in these bands were calculated for each altitude.

Although the investigators of NIRS, TCD, and EEG were not aware of the AMS Score, which was assessed by the expedition doctors independently, they were not blinded for clinical signs of AMS.

The data of left and right values were pooled if the differences of the latter did not reach the level of significance. Only data for those mountaineers who reached 5,050 m altitude were used in further statistical analyses. As there were no differences between the two mountaineering groups (based on the ascent route via Gokyo Ri (5,340 m) or Island Peak (6,200 m)), the data were assigned to the AMS or non-AMS group regardless of the ascent route taken.

Data were analyzed using the SPSS software (release 11.0, SPSS, Chicago, IL, USA). The Friedmann analysis was performed for several connected samples, and the analysis was continued only if the P value was < 0.05. This procedure can replace a Bonferroni correction to adjust the P value according to the number of analyses. ⁴ Statistical analysis of symptomatic vs. asymptomatic montaineers was performed using the exact non-parametric Mann–Whitney U-test. For assessment of symptomatic or asymptomatic mountaineers at different altitudes, the non-parametric Wilcoxon test was used. Correlation analyses were calculated with the Spearmann Correlation coefficient. A P value of < 0.05 was regarded as statistically significant. Values are given as mean (± s.e.m.).

The study protocol was approved in advance by the Ethics Committee of the University of Vienna (Austria) according to the ethical standards of the Helsinki Declaration of 1975 (and as revised in 1983). We confirm that written, informed consent was obtained from each participant.

RESULTS

All participants reached Namche Bazaar at 3,440 m on day 2, and 26 of the 32 participants, the Silver Pyramid (5,050 m) between 12 and 13 days later. Because of technical difficulties, EtCO₂, SpO₂, and TCD data of the PCA were not obtained in one mountaineer at 5,050 m.

AMS Lake-Louise Score

Fourteen of 26 mountaineers who reached 5,050 m altitude developed symptoms of AMS between 3,440 and 5,050 m altitude (Lake-Louise Score ≥ 3). The mean AMS Score in this group was 4 ± 0 (min = 3; max = 7) and the mean duration of illness, 1 ± 0.4 day (min = 1; max = 5).

Blood Pressure, Endtidal CO₂, and Peripheral and Regional Cerebral Oxygenation

All participants did not exhibit a change in blood pressure during the ascent from 100 to 3,440 m and 5,050 m (mean 120/80 mm Hg (±2.7/±1.8) to 130/80 mm Hg (±3.3/±1.9) to 127.5/80 mm Hg (±2.8/±1.4); the blood pressure also did not differ between AMS and non-AMS mountaineers.

Endtidal CO₂ levels decreased from 33 ± 0.5 to 31.5 ± 1.1 mm Hg at 3,440 m altitude (P < 0.0001) and to 25.0 ± 0.8 mm Hg at 5,050 m (P < 0.0001). SpO₂ decreased continuously from 100 to 3,440 m altitude from 99% ± 0.3 to 90% ± 0.6 (P < 0.05) and during further ascent to 5,050 m altitude to 84% ± 0.8 (P < 0.01). Regional cerebral oxygenation decreased continuously from 100 to 3,440 m by 69% ± 1.4 to 60% ± 1.3 (P < 0.0001) and further at 5,050 m by 57% ± 2.1 (P < 0.0001).

Cerebral Blood Flow Velocity

The CBFV in the ACA of all mountaineers showed an increase in blood flow velocity between 100 and 3,440 m altitude, 100% ± 0: 119% ± 9; P < 0.01. During further ascent to 5,050 m, this increase was followed by a decrease from 119% ± 9 to 112% ± 10 (see Figure 1A). Compared with the baseline of 12%, the increase still reached the level of significance (Figure 1A; P < 0.05).

Figure 1.

Changes in cerebral blood flow velocity (CBFV) for all mountaineers in the anterior cerebral artery (A), middle cerebral artery (B), and in the posterior cerebral artery (C). Differences in CBFV in the middle cerebral artery between mountaineers who developed acute mountain sickness (AMS) or did not are shown in panel (D). Different altitudes were compared using Wilcoxon test; differences between AMS and non-AMS mountaineers with the Mann–Whitney U-test.

CBFV in the MCA rose by 12% ± 10 from 100 to 3,440 m (P < 0.05) and was still 10% ± 10 higher at 5,050 m than at 100 m altitude (P < 0.05; see Figure 1B). During further ascent from 3,440 to 5,050 m altitude, a difference was observed between mountaineers with AMS and those without AMS (see Figure 1D): 14% ± 12 increase compared with 11% ± 17 decrease in CBFV (P < 0.05).

CBFV in the PCAs decreased continuously from 100 m over 3,440 m to 5,050 m; 100% ± 0: 80% ± 9: 64% ± 9, P < 0.05 in all mountaineers (see Figure 1C).

The CBFV in the ACA and the PCA correlated with EtO₂ levels, indicating that lower EtCO₂ levels (better hyperventilation) correlated with higher CBFV values (for ACA: r = −0.428, P < 0.05; and for PCA: r = −0.458, P < 0.05).

Electroencephalography Activity

The background activity of the frontal lobe is the beta rhythm and that of the occipital lobe, the alpha rhythm. Whereas the frontal EEG beta activity increased in AMS mountaineers between 100 and 3,440 m altitude by 24% ± 14 (P < 0.05; see Figure 2B), there was no such increase in non-AMS mountaineers. However, this group showed an increase by 27% ± 22 at 5,050 m altitude compared with baseline (P < 0.05; see Figure 2A).

Figure 2.

Changes in frontal beta activity in non-acute mountain sickness (non-AMS) (A) and AMS (B) mountaineers. Panel (C) shows changes in occipital EEG alpha activity in non-AMS and panel (D) in AMS mountaineers. Different altitudes were compared using Wilcoxon test; differences between AMS and non-AMS mountaineers with the Mann–Whitney U-test. EEG, electroencephalography.

An increase of 27% ± 33 in occipital alpha activity in AMS mountaineers also occurred between 100 and 3,440 m (P < 0.05; see Figure 2D) but not in those who did not develop AMS. The increase by 17% ± 36 in the non-AMS group between 100 and 5,050 m did not reach the level of significance (see Figure 2C).

Delta activity decreased in the frontal lobe by −34% ± 95 at 3,440 m and by −24% ± 188 at 5,050 m (not significant) as well as in the occipital lobe by −13% ± 15 at 3,440 m; a slight increase of +2% ± 13 was observed at 5,050 m (not significant).

DISCUSSION

Changes in Blood Pressure and Oxygenation

The blood pressure values did not change between the altitudes or between AMS and non-AMS mountaineers. These findings illustrate that the changes in CBFV did not contribute to a systemic increase in blood pressure. However, these results are limited, as we did not measure dynamic cerebral autoregulation using transfer-function gain, phase, and coherence between mean blood pressure and MCA velocity. ⁵ It is well known that air and O₂ pressure decrease with rising altitude, resulting in hypobaric hypoxia. NIRS revealed that these effects lead to a decrease of peripheral O₂ saturation as well as brain O₂ saturation. ⁶ We could not demonstrate a correlation between the decrease in rSO₂ and compensatory increase of frontal CBFV in our study. This may indicate that mechanisms other than those directly proceeding via hypoxia are involved in regulating the CBFV. ⁷ These results indicate that even an increase of CBF is not able to restore the oxygen level in the frontal cortex during an ascent to high altitude.

Changes of Cerebral Blood Flow Velocity

Regional differences. Low O₂ levels lead to an increase of CBFV in normoxic conditions, and low CO₂ levels, to a decrease of CBFV. This well-known mechanism is used in neurological intensive care units to reduce brain swelling and has been documented in all larger brain arteries. ⁸ However, in hypobaric hypoxia, CO₂ levels are low because of compensatory hyperventilation, resulting in competetive hypocapnia and hypoxia. It is well known that CBFV increases at high altitudes. ⁹ It is thought that the increase in the anterior parts of cerebral circulation is mediated by a decrease of oxygen, ¹⁰ which conflicts with the effects of hypocapnia because of hypoxic hyperventilation. ⁷ Previous reports on this effect were based on measurements made mainly in the MCA. ¹¹ Investigations comparing the anterior and posterior circulation are sparse. Willie et al. ⁸ reported no regional difference in the internal carotid artery or vertebral artery during a slow ascent to 5,050 m, whereas they measured a greater elevation in vertebral artery flow after a rapid ascent to 5,260 m than in the internal carotid artery. ¹² Thus, it seems upon sudden exposure to hypoxia that there is a preferential maintenance of brainstem blood flow. ¹¹ Here we report that a sojourn at high altitude led to a decrease of CBFV in the PCA. This finding contradicts results of the previously mentioned studies that showed an increase of CBFV in the vertebral and basilar arteries in the posterior part of the brain. A preliminary study in children has even shown that although elevations in velocity in the MCA and ACA occurred following ascent to 3,500 m from sea level, velocity in the PCA and basilar artery were otherwise unchanged. ¹³

Another limitation of our study was that we did not investigate the CBFV in the basilary or vertebral circuits. Skow et al. ¹⁴ showed a greater absolute cerebrovascular CO₂ reactivity in the MCA compared with the PCA in the human brain. Because of the diminished vasoreactivity of the PCA, the CBFV-increasing effect of hypoxemia might be less pronounced and result in an overbalance of the CO₂-induced decrease of CBFV. The increase in sympathetic activity at high altitudes ¹⁵ may be an additional factor relevant for the CBFV increase. The posterior parts of the blood vessels are less innervated by the sympathetic nervous system. ¹⁶ Whether this is another factor that can explain the subsequent decrease of CBFV in the PCA remains unclear. However, it is known that hypoperfusion in the PCA occurs in patients with hypoxic brain injuries. Cardiac events or asphyxation may cause brain lesions in very hypoxia-sensitive areas such as the basal ganglia or the posterior brain areas. ¹⁷

Relation to acute mountain sickness. CBFV in the MCA increased in the non-AMS mountaineers earlier (at 3,440 m altitude) than in the AMS group (at 5,050 m altitude). The literature has documented that CBFV increases in the MCA by 20% to 50% 12 to 24 hours after arrival at altitudes ranging between 3,475 and 4,559 m. ¹⁸ The subsequent rise in CBF, which acts as a compensating mechanism, may at least partially restore oxygen delivery to the brain, ¹⁰ but it does not cause AMS.^7,10 Such delayed oxygen delivery in the AMS mountaineers compared with non-AMS mountaineers may at least reflect their maladaptation to the hypoxic environment. Hypoperfusion of brain areas as occurs in stroke leads to rapid accumulation of diffusion-restricted water in the intracellular space (cytotoxic edema) because of the breakdown of the transmembrane pump and/or to microscopic brain pulsations. ¹⁹ Recent findings using diffusion-weighted magnetic resonance imaging (MRI) have shown that a decreased apparent diffusion coefficient directly predicted the appearance of AMS symptoms. ²⁰ If these effects are aggravated, AMS may evolve directly into HACE, further increasing vasogenic edema, which leads to dysfunction of the posterior parts of the brain. This results in ataxia, somnolence, vertigo, delirium, and hallucinations, which can be diminished by using dexamethasone to also reduce brain swelling.

Relation to high-altitude cerebral edema. A decrease of CBFV in the PCA seems inexplicable, as HACE is a vasogenic edema predominantly located in the posterior part of the brain. A similar syndrome occurring during normoxemia is the posterior reversible encephalopathy syndrome (PRES). It develops in patients with complex systemic conditions such as eclampsia, after transplantation, infection/sepsis/shock and autoimmune diseases, and after cancer chemotherapy. ²¹ Computed tomography or MRI studies show that the edema is often widspread but predominantly in the parietal and occipital regions. ²¹ Most studies have detected reduced brain perfusion in regions of PRES. ²¹ In agreement with our findings at high altitude, Brubaker et al ²² compared anterior with posterior hemispheric flow and demonstrated significant posterior brain hypoperfusion with increased mean transit time, reduced CBV, and reduced CBF in PRES. This could lead to further localized brain hypoxia, with upregulation of vascular endothelial growth factor because of tissue expression of the hypoxemia-inducible factor 1 alpha. ²³ These effects may stimulate angiogenesis and increase endothelial permeability. ²⁴ In a cross-sectional study comparing mountaineers who had experienced HACE, high-altitude pulmonary edema, severe AMS, or no altitude illness, susceptibility-weighted MRI detected microhemorrhages predominantly in the splenium of the corpus callosum. These microhemorages were a highly specific sign of HACE and correlated with the extent of the clinical presentation, indicating a leak in the blood–brain barrier. ²⁵ Interestingly, the splenium is at the watershed between branches of the ACA and PCA ²⁶ and may therefore be prone to regional hypoxic changes because of hypoperfusion of the PCA. This junctional/watershed nature has also been discussed in PRES. ²¹ These findings are compatible with the concept of increased capillary pressure because of a limited cerebral venous outflow at high altitude, which was suggested by Wilson et al. ²⁷ In a recent MRI study, it has been shown that susceptibility-weighted angiography visualizes hypoxia in cerebral veins. ²⁸ As changes in CBFV does not seem to be the cause of AMS, the absence of differences in posterior CBFV between AMS and non-AMS mountaineers in our study is not surprising.

Comparison with the native population living at high altitudes. We did not compare the mountaineers with natives of the Sagarmatha National Park or with other Himalayans. A review of the literature showed that Himalayans have slightly (6.2%) higher CBF values than subjects at sea level. ²⁹ In contrast, Andeans living in the same high-altitude range have CBF values that are substantially (18%) lower than those of the subjects at sea level, indicating a ~ 20% higher CBF, after correction for hematocrit and arterial oxygen saturation, in Himalayan compared with Andean high-altitude populations. ²⁹ Increased nitric oxide availability in Himalayans and/or hypo-metabolism of the brain of Andeans may have a role in this difference. ²⁹

Changes of Electroencephalography Activity

Neurovascular decoupling (increased neuronal activity despite reduced cerebral blood flow). Regional cerebral perfusion correlates with the activity and firing rate of the cortical neurons in the region. ³⁰ Thus reduced cerebral perfusion indicates reduced neuronal activity, i.e., downregulation. EEG activity of the cortical posterior areas showed, however, an increase of neuronal activity that resulted in an increase of alpha background rhythm. We propose that the increased neuronal activity at this stage despite reduced CBF is because of modulation of the ion channels by hypoxia that can result in depolarization and increased excitability of cells, including K, Ca ² , and Na channels. ³¹ In addition, positron emission tomography scan data of subjects with isocapnic hypoxia revealed that cortical blood flow is less responsive to hypoxia than blood flow in phylogenetically older areas of the brain. ³² This might cause the increase of frontal and occipital cortical EEG background activity and the absence of increased delta waves, representing a loss of cortical function because of hypoperfusion.

Relation to acute mountain sickness. It has been reported elsewhere that right regional temporal dysfunction signaled by an increase of delta activity in the EEG heralds the symptoms of AMS. ³³ AMS mountaineers showed an increase of the background activity (occipital alpha activity) between 100 and 3,440 m compared with non-AMS mountaineers, in whom such an increase was delayed until reaching higher altitudes between 3,440 and 5,050 m. Furthermore, low arterial PCO(2) is known to increase neural excitability. It was hypothesized that the low arterial PCO(2) resulting from ventilatory acclimatization causes an increase in cerebral oxygen metabolism. ³⁴ It has been shown that passive transport to 3,500 m altitude also leads to an increase of occipital alpha activity. ³⁵ Such an increase was observed in a small group of mountaineers with mild AMS symptoms, whereas mountaineers who developed severe AMS later showed a decrease of the background rhythm. ³⁶ This decrease may be because of a breakdown of compensatory mechanisms, indicating maladaptation.

A decrease of overall occipital alpha activity was reported in a pressure chamber study simulating a fast ascent (150 m/min and rest of 25 minutes every 1,000 m) to 6,000 m altitude. However, this study did not report any clinical signs of AMS, ³⁷ although they were probably present. A study investigating quantitative EEG changes revealed that hyperventilation caused an exponential increase in slow activity and a decrease in alpha power. ³⁸ The rate of ascent was moderate in our study, allowing the brain time to adapt to hypobaric hypoxia. Delta activity did not increase in the frontal or occipital lobe, perhaps because none of the mountaineers studie, developed HACE.

It has been suggested that the changes in increase of occipital alpha activity may be induced by sympathetic hyperactivity. Such hyperactivity might be mediated by the release of norepinephrine in the brainstem, thus influencing the reticular activation system, ³⁹ or induced via a central oxygen-chemosensitive network. ³¹

CONCLUSION

Two limitations of our study were the relatively small sample size and the sometimes difficult conditions in the field. The increase in CBFV in the anterior and MCA is a compensatory response to the hypobaric hypoxic environment, which restores blood and oxygen supply to the brain. The increase in CBFV in the MCA was delayed in mountaineers with AMS. Contrary to our hypothesis, CBFV decreased in the PCA with increasing altitude and hypoxia. This decrease was most likely because of differences in cerebrovascular reactivity to changes in CO₂, resulting in a different balance of hypoxic increase and hypocapnic decrease of CBFV.

In summary, the results of MRI studies of simulated ascents to very high altitudes in pressure chambers suggest that mild cytotoxic edema might be more prevalent in persons with symptoms of AMS. ²⁷ Such cytotoxic edema might occur if oxygen delivery is reduced because of the decreased perfusion in the posterior parts of the brain. This results in dysfunction of Na/K+ ATPase with consequently cytotoxic edema and AMS. If these effects are aggravated, AMS may evolve directly into HACE, which is characterized by dysfunction of the posterior parts of the brain. PRES may provide insights into interesting, analogous pathways in the pathophysiology of both conditions. It has been suggested that PRES develops because of the failure of autoregulation and hyperperfusion, but the theory of alternative endothelial dysfunction and hypoperfusion and vasoconstriction leading to altered integrity of the blood–brain barrier is favored. ²¹ Similar pathophysiological considerations might be valid for the development of HACE. Anatomical findings of the watershed area between the ACA and the PCA at the splenium might explain this predilection for microhemorrhages in HACE, in view of the continuous decrease of CBFV in the PCA.

These findings indicate that the physiological response of CBFV to hypobaric hypoxia differs in the anterior and posterior supratentorial parts of cerebral circulation despite their similar electrophysiological changes.