Wilderness and Environmental Medicine, 17, 1 7 (2006)
Cerebrovascular Dynamics and Vascular Endothelial
Growth Factor in Acute Mountain Sickness
Joseph Palma, MD; Christian Macedonia, MD; Patricia Deuster, PhD; Cara Olsen, MPH;
B. Robert Mozayeni, MD; Kevin E. Crutchfield, MD
From the Departments of Military and Emergency Medicine (Drs Palma and Deuster), Biostatistics (Dr Olsen), Obstetrics and Gynecology
(Dr Macedonia), and Neurology and Military Emergency Medicine (Dr Crutchfield), and the Human Performance Laboratory
(Dr Deuster), Uniformed Services University of the Health Sciences, Bethesda, MD; and New Health Sciences Inc, Rockville, MD
(Drs Mozayeni and Crutchfield).
Objective.—To determine if serum vascular endothelial growth factor (VEGF) and ultrasonic mon-
itoring of vascular dynamics with dynamic vascular analysis at sea level and high altitude correlate
with acute mountain sickness symptoms.
Methods.—Nine volunteers participated in a staged ascent from sea level to 4300 m undergoing
complete transcranial Doppler studies with dynamic vascular analysis. Serum VEGF levels, Lake
Louise scores, Spielberger-1 scores, Subjective Exercise Experiences Scale positive scores, and Symp-
tom Checklist-90 surveys were collected after 24 hours at each altitude.
Results.—Symptom scores, index of pulsatility, and dynamic flow index differentiated the subjects
into 2 distinct groups. Symptomatic subjects had increased VEGF levels at sea level but decreased
levels at 4300 m. The dynamic flow index increased in symptomatic subjects at 4300 m compared
with the asymptomatic subjects. The mean flow velocity increased in both groups and could not be
used to differentiate the subjects.
Conclusions.—Altered vascular physiology is associated with acute mountain sickness. Increased
vascular permeability increases vascular capacitance, with an increase in dynamic flow index to meet
these demands. Altered vascular dynamics were associated with high-altitude cerebral edema in 1
subject. Dynamic vascular analysis demonstrated altered vascular pathophysiology associated with
acute mountain sickness. Changes in VEGF were meaningful when interpreted with the dynamic
vascular analysis findings. These physiological findings may help explain the vascular changes asso-
ciated with hypocarbic hypoxemia at altitude.
Key words: dynamic vascular analysis, transcranial Doppler ultrasound, acute mountain sickness, en-
dothelial dysfunction, blood brain barrier, vascular endothelial growth factor, altitude
Acute mountain sickness (AMS) represents an increas-
ing clinical problem at higher altitude1,2; if unrecognized
and untreated, it may lead to death in just a few hours.
Although well documented, no conclusive physiological
evidence exists to link increased cerebral blood flow
(CBF) secondary to hypobaric hypoxia to the develop-
ment of AMS or high-altitude cerebral edema. Serum
vascular endothelial growth factor (VEGF) levels have
been measured in subjects exposed to altitude; however,
Corresponding author: Kevin E. Crutchfield, MD, New Health Sci-
ences Inc, 6903 Rockledge Dr, Suite 230, Bethesda, MD 20817-1818
changes do not correlate with development of altitude
illness.3–5Hypoxemia constricts pulmonary vessels,6
whereas in the cerebrum it causes hyperemia.7However,
both vascular territories develop endothelial dysfunction
with edema at higher altitude. Human transcranial Dopp-
ler (TCD) studies at high altitude revealed increased
middle cerebral artery (MCA) mean flow velocities
(MFVs) secondary to hypoxemic drive,8yet changes in
MFV in the MCA do not predict subjects who develop
AMS.9Transcranial Doppler studies of the MCA at al-
titude have demonstrated diminished index of pulsatility
(PI) or index of resistance10–12but with unclear clinical
We investigated the impact of increasing altitude as a
Palma et al
function of graded hypoxemic stress by correlating CBF
dynamics informed by dynamic vascular analysis
(DVA)13of TCD-derived velocimetry data, serum
VEGF levels, and 4 sickness symptom scales. We uti-
lized DVA to assess cerebrovascular physiology, allow-
ing for sensitive monitoring of changes in cerebrovas-
cular physiology associated with changes in altitude.
The institutional review board at the Uniformed Services
University of the Health Sciences, Bethesda, MD, ap-
proved this prospective, longitudinal, single-arm, single-
blind pilot trial. Nine healthy, athletic male subjects who
were medical faculty and PhD candidates from the Uni-
formed Services University of the Health Sciences who
signed informed consent forms were taken from sea lev-
el to the summit of Pikes Peak, CO, at 4300 m (altitude).
The ascent up the mountain was staged for 5 days with
daily stops and testing at 1585 m, 2500 m, and 3415 m
and a final ascent to 4300 m (summit). All volunteers
were given the following questionnaires at each altitude
to measure severity of altitude illness: Lake Louise
score, Spielberger-1 score, Subjective Exercise Experi-
ences Scale positive score, and Symptom Checklist-90
surveys.14–17Transcranial Doppler data were acquired
by a single technologist (nonsubject) from recumbent
subjects at each altitude by well-documented handheld
techniques and depth of measurement18with a Pioneer
Companion system (Nicolet Vascular, Golden, CO) hav-
ing a 2-MHz probe. Velocimetry measurements were
taken from both MCAs M1 (50–60 mm), both anterior
cerebral arteries A1 (60–70mm), both posterior cerebral
arteries, both P1 (60–70 mm) and P2 (60–70 mm), both
vertebral arteries (60–65 mm), and the basilar artery
(80–100 mm). New Health Sciences Inc (Rockville,
MD) performed the DVA analysis according to previ-
ously reported methods.13This analysis now includes
the dynamic flow index (DFI ? MFV/PI), dynamic pres-
sure index (DPI ? natural log of systolic acceleration)/
PI), and dynamic compliance index (DCI ? natural log
of systolic acceleration/MFV). The DFI relates MFV to
resistance (PI) and reflects small vessel effects on flow
through conductance vessels. The DPI relates accelera-
tion or force of flow to the downstream resistance and
relates the effects of small vessel volume on the force
of flow or the pressure gradient on which the blood
flows. The DCI relates acceleration or force of flow to
MFV and reflects the elasticity and kinetic efficiency of
pulse wave propagation in the vessel segment insonated.
Blood samples were taken at each altitude and im-
mediately centrifuged at 3400g for 20 minutes to sepa-
rate the serum, which was subsequently flash frozen in
the field and sent for proteomic VEGF characterization
by the Advanced Concepts Laboratory of the National
Cancer Institute, National Institutes of Health, Bethesda,
MD (Dr Steven Libuttis).
All 9 subjects completed the study up to 3415 m. Only
8 subjects fully completed the study at summit altitude.
One subject experienced acute altitude illness 2 hours
after ascent, with ataxia, bilateral papilledema, confu-
sion, and disorientation to person, place, and time. He
recovered spontaneously without sequelae after emer-
gent descent to 3650 m. Transcranial Doppler veloci-
metry data were collected from this subject at the sum-
mit before descent, but no blood sample or symptom
scores were obtained. Symptom scores were all assumed
to be severe in this subject because he was incapable of
answering the questionnaires at altitude. To include him
in the analysis, he was assigned a total Lake Louise
score of 30, a Symptom Checklist-90 score of 90, and a
Spielberger-1 score of 60 at 4300 m. Nonparametric sta-
tistical methods based on ranks were used for all analysis
involving symptom scales, so the absolute numerical
score was not as important because the scores were high-
er than those recorded by the other subjects.
Mean flow velocity, systolic upstroke acceleration
(SA), PI, VEGF, and symptom scores were compared by
using the Wilcoxon signed rank test. A significant dif-
ference was considered at P ? .05. Symptom scores at
altitude were compared with MFV, natural log of systolic
acceleration, PI, VEGF, DFI, DPI, and DCI scores at all
altitudes by using Spearman rank correlation (r). Spear-
man rank correlations lie between ?1 and ?1, and a
correlation of 0 indicates no relationship. A positive cor-
relation indicates that an increase in that measure cor-
responds to an increase in symptoms, whereas a negative
correlation indicates that a decrease in that measurement
corresponds to a decrease in symptoms. The analysis
focused on the relationship to symptom scales at final
altitude because subjects experienced varied symptoms
at lower altitudes.
Symptom scores were compared with changes in the
other variables as follows:
● For each subject, the slope of the linear regression for
the PI at each altitude was calculated. A positive slope
indicated that PI increased with increasing altitude;
conversely, a negative slope indicated that PI de-
creased with increasing altitude.
● Spearman rank correlation was used to compare the
slope of the regression with a symptom score for ev-
ery subject. A negative correlation indicated that a
subject whose PI increased with increased altitude
(positive slope) had fewer symptoms.
● This analysis was repeated for every variable.
Dynamic Vascular Analysis of Change in Altitude
Both symptom groups showed increased MFV with altitude.
Effect of altitude on mean flow velocity (MFV).
symptom scores. The difference in symptom scales between
the symptomatic group (n ? 3) and the asymptomatic group
(n ? 6) became significant at peak altitude. Means and ranges
for Lake Louise scores are shown. All symptom scales showed
Effect of summit altitude (4300 m) on Lake Louise
Changes in PI significantly differed between the 2 symptom
groups with altitude (*P ? .05; paired-samples t test).
Effect of altitude on index of pulsatility (PI).
● For each TCD measurement, average change from sea
level to 4300 m was calculated for symptomatic (Sx)
and asymptomatic (Asx) subjects, and the 2 groups
were compared by using the paired-samples t test.
We observed changes in TCD values, confirming in-
creased MFV with increasing altitude in every subject
(Figure 1), with a holocephalic, median level increasing
from 46.64 cm·s?1at sea level to a median of 55.32
cm·s?1at altitude (P ? .008). There was no significant
cohort change in holocephalic PI with increasing alti-
tude: the PI increased in 5 subjects and decreased in 4
There was no cohort change in mean VEGF level with
altitude: 5 subjects experienced increased VEGF, 2 sub-
jects had no change, and 2 subjects had a peak value at
3415 m with a decrease in VEGF at summit altitude.
Symptoms scales increased for most subjects with in-
creasing altitude. The median number of reported symp-
toms increased from 1 at sea level to 15 at altitude (P
? .008) on the Symptom Checklist-90 scale, from 23 at
sea level to 25 at altitude (P ? .063) on the Spielberger-
1 scale, and from 1 at sea level to 9 at altitude (P ?
.008) on the Lake Louise questionnaire (Figure 2).
Subjects with a high PI at every altitude above sea
level reported fewer symptoms at final altitude (?0.04
? ? ? ?0.79). Subjects whose PI increased with in-
creasing altitude had far fewer symptoms than those sub-
jects whose PI decreased with increasing altitude (Figure
3). Because of this relationship, the group was subcat-
egorized into an Sx group with 3 subjects and an Asx
group with 6 subjects (Table 1).
Comparisons of TCD, DVA, and VEGF levels be-
tween the 2 groups at sea level and altitude, as well as
analysis comparing changes between sea level and alti-
tude within the Sx and the Asx groups, were performed
(Table 2). At altitude, we noted significant group differ-
ences (P ? .05) between the groups regarding PI (Figure
3) and DFI (Figure 4). Trending, but not a significant
difference, was noted in the DPI and the serum VEGF
level (Figure 5). We noted a dramatic drop in serum
VEGF level at high altitude in the Sx group. This drop
indicates complex kinetics of serum VEGF in hypoxic
Noninvasive monitoring of cerebrovascular dynamics
with changing altitude revealed that increased hypox-
Palma et al
Table 1. Ranges of symptom scales at sea level and altitude (4300 m) between the symptomatic (Sx) and asymptomatic (Asx)
Symptom Checklist-90 Sx
Symptom Checklist-90 Asx
Lake Louise total Sx
Lake Louise total Asx
Subjective Exercise Experiences Scale score Sx
Subjective Exercise Experiences Scale score Asx
Table 2. Transcranial Doppler, dynamic vascular analysis, and vascular endothelial growth factor (VEGF) changes noted from
sea level to altitude (4300 m) between the symptomatic (Sx) and asymptomatic (Asx) groups†
n Mean SD% change
nMean SD% change
MFV sea level
ln SA sea level
ln SA altitude
Pl sea level
VEGF sea level
DCI sea level
DFI sea level*
DPI sea level
*P ? .05.
†MFV indicates mean flow velocity; SA, systolic upstroke acceleration; PI, index of pulsatility; DCI, dynamic compliance index; DFI, dynamic
flow index; and DPI, dynamic pressure index.
emia in the Sx group was associated with increasing DFI
and diminished pulsatility, suggesting reduced vascular
resistance from expanding capacitance vessel volume.
The present study confirmed previous altitude studies
demonstrating that MFV increased in the MCA,7,11,19,20
increasing more in AMS.11,19,20Previous studies also re-
vealed that cerebral MFV increased with moderate ex-
ercise at altitude but decreased with intense exercise in
subjects developing AMS,20,21indicating diminished
metabolic reserve or hypoxemic tolerance. In animals,
hypoxic drive was demonstrated to have nonlinear re-
lationships to increased CBF22regardless of hypocarbia
from hyperventilation.23Previous altitude studies dem-
onstrated decreased carbon dioxide response or capaci-
tance vessel reserve20,24in all subjects, including native
Sherpas, regardless of AMS symptoms.24A clear rela-
tionship exists at altitude between MFV and capacitance
Capacitance vessel dilation from hypoxemic stress
will improve minute oxygen delivery24by increasing
overall CBF. Capacitance vessel reserve is mediated by
myogenic contraction not affected by hypoxia but cannot
override the potent effects of hypoxic vasodilation it-
self.25Vasodilation is affected by metabolic alterations
caused by hypoxemia, including increased cerebrospinal
fluid pH,26increased hematocrit,27increased hemoglo-
Dynamic Vascular Analysis of Change in Altitude
The DFI (mean flow velocity/index of pulsatility) significantly
increased in the symptomatic group compared with the asymp-
tomatic group (*P ? .05; paired-samples t test).
Effect of altitude on dynamic flow index (DFI).
growth factor (VEGF). Serum VEGF levels dropped initially
at 1585 m in the symptomatic (Sx) group, then the levels for
both the Sx group and the asymptomatic (Asx) group increased
until 3415 m, where the Sx group fell precipitously and the
Asx group had a slight decline. These findings were not sig-
Effect of altitude on serum vascular endothelial
bin, and decreased nitric oxide,28partially explaining ac-
climatization and stabilization of CBF at altitude. One
possible mechanism for the development of AMS is im-
pairment of endothelial gap junction function secondary
to hypoxemia3,29–32characterized by increased perme-
ability to macromolecules and increasing interstitial free
water, which may result in high-altitude cerebral ede-
ma33and high-altitude pulmonary edema as well as ex-
pansion of capacitance vessel volume.
High-altitude cerebral edema may be secondary to in-
creased endothelial permeability of cerebral capacitance
or small vessels.12,33Hypoxemic stress increases pro-
duction of VEGF3,34,35and VEGF receptors,5which in-
crease vascular permeability. Levels of VEGF correlate
with altitude symptoms compared with sea level con-
trols; however, at altitude, VEGF did not correlate with
these same symptoms.5This suggests variable responses
to increasing VEGF, as suggested by our cohort. Our
increase in VEGF did not reach significance but sug-
gested an oxidative stress. Vascular endothelial growth
factor increases vessel permeability.33Increased dys-
function of the endothelial tight junctions and adherens
junctions with increasing oxidative stress is known. A
recent study involving rats showed direct evidence of
increased permeability to larger molecular weight com-
pounds with hypoxia.36The increased fenestrations in
the endothelial barrier leak serum proteins into intersti-
tial spaces, pulling free water. This may explain the drop
in serum VEGF levels observed in our Sx subjects. If
these ‘‘vascular holes’’ open, then a drop in intravascular
pressure would occur in capacitance vessels, thereby di-
minishing capillary perfusion pressure and conductance
vessel resistance as observed in this study. As resistance
drops, flow increases to maintain adequate minute per-
fusion of the capillaries. Interestingly, if enough of these
gaps were open, a relative drop in intraluminal pressure
of a ‘‘closed’’ vascular system would occur because this
new state would represent an ‘‘open’’ system, causing
increased flow with less work provided by the conduc-
tance system to maintain the pressure gradient. This cor-
relates with our findings that the PI in conductance ves-
sels diminished, as did the vessel tone (decreased DCI)
secondary to increased vessel volume (increased DFI) in
the Sx group. The appropriate compensatory mechanism
would increase capacitance vessel myogenic tone to in-
crease intravascular pressure of this now-open system
and increase forward perfusion pressure, thereby in-
creasing the pressure gradient across the capillary beds.
This compensatory change was evident in the Asx group
but not in the Sx group.
Our study differs from other TCD-derived velocime-
try studies at altitude in that holocephalic (including the
posterior circulation) changes were monitored in contrast
to other studies of an isolated cerebral vessel segment.
This approach demonstrated that changes in CBF at al-
titude are holocephalic and that changes in the posterior
circulation are present. Cerebral blood flow increases as
a result of a medullary response to hypoxia37mediated
by a relay through the thalamus to dilate cortical vessels.
This underscores the importance of monitoring posterior
circulation flow when studying the pathophysiology of
In this cohort, changes in the conductance vessel SA
and DCI, as markers of perfusion pressure and vessel
tone, respectively, correlated with symptom scores and
serum VEGF levels. Less symptomatic subjects had
higher VEGF scores, with the ability to maintain cere-
Palma et al
brovascular compliance (ie, higher DCI and perfusion
pressure, SA). Symptomatic subjects had increased vas-
cular tone with lower VEGF levels, suggesting increased
vascular tone reflected as decreased DCI or PI. The most
striking finding in our study was the reverse relationship
of symptom scores to the PI. As PI diminished (lower
vascular resistance secondary to expansion of capaci-
tance vessel volume), symptoms increased.
This study was underpowered to show significant
group changes, a possible type 2 error. Although serum
VEGF levels were trending, this study suggests that
studies with more subjects would show significant dif-
ferences in VEGF levels between the symptom groups.
We run the risk of a type 1 error given the number of
subjects in this study. Pooling of all vessel data mini-
mized this risk.
In summary, this DVA-enabled study indicates that
acute mountain sickness is associated with altered ce-
rebrovascular kinetics and volume expansion as a func-
tion of dysfunctional capacitance vessels. At altitude, in-
creased permeability of the vascular endothelium from
opening of endothelial junctions in capacitance vessels
most likely precedes increasing symptoms and eventu-
ally high-altitude cerebral edema and high-altitude pul-
monary edema. Dynamic vascular analysis proved to be
a sensitive tool for measuring and elucidating the vas-
cular physiology behind changes in CBF dynamics as-
sociated with symptomatic altitude illness. Potential
therapies directed at maintaining the integrity of endo-
thelial junctions and capacitance vessel tone to prevent
acute altitude sickness or other hypoxemic states could
be noninvasively monitored with DVA.
The authors thank Alan Faden, MD, for his review and
comments regarding the manuscript and James Zimble,
MD, Barry Wolcott, MD, and Fred Cecere, MD, for their
support. Dynamic vascular analysis is a processing
method for Doppler data provided by New Health Sci-
ences Inc. Dynamic vascular analysis, dynamic flow in-
dex, dynamic pressure index, and dynamic compliance
index are trademarks of New Health Sciences Inc, Rock-
ville, MD 20850, (240)453-9191.
1. West JB. Oxygen enrichment of room air to improve well-
being and productivity at high altitude. Int J Occup En-
viron Health. 1999;5:187–193.
2. Burtscher M. High altitude headache: epidemiology, path-
ophysiology, therapy and prophylaxis. Wien Klin Woch-
3. Schoch HJ, Fischer S, Marti HH. Hypoxia-induced vas-
cular endothelial growth factor expression causes vascular
leakage in the brain. Brain. 2002;125(pt 11):2549–2557.
4. Appenzeller O, Minko T, Pozharov V, et al. Gene expres-
sion in the Andes; relevance to neurology at sea level. J
Neurol Sci. 2003;207:37–41.
5. Maloney J, Wang D, Duncan T, Voelkel N, Ruoss S. Plas-
ma vascular endothelial growth factor in acute mountain
sickness. Chest. 2000;118:47–52.
6. Mentzer RM, Rubio R, Berne RM. Release of adenosine
by hypoxic canine tissue and its possible role in pulmo-
nary circulation. Am J Physiol. 1975;229:1625–1631.
7. Jensen JB, Wright AD, Lassen NA, et al. Cerebral blood
flow in acute mountain sickness. J Appl Physiol. 1990;69:
8. Yang YB, Sun B, Yang Z, Wang J, Pong Y. Effects of
acute hypoxia on intracranial dynamics in unanesthetized
goats. J Appl Physiol. 1993;74:2067–2071.
9. Berre J, Vachiery JL, Moraine JJ, Naeije R. Cerebral blood
flow velocity responses to hypoxia in subjects who are
susceptible to high-altitude pulmonary oedema. Eur J Appl
10. Ter MA, Beydon L, Ursino M, Gardette B, Gortan C, Ri-
chalet JP. Doppler study of middle cerebral artery blood
flow velocity and cerebral autoregulation during a simu-
lated ascent of Mount Everest. Wilderness Environ Med.
11. Otis SM, Rossman ME, Schneider PA, Rush MP, Ringel-
stein EB. Relationship of cerebral blood flow regulation
to acute mountain sickness. J Ultrasound Med. 1989;8:
12. Hackett PH. The cerebral etiology of high-altitude cerebral
edema and acute mountain sickness. Wilderness Environ
13. Crutchfield KE, Razumovsky AY, Tegeler CH, Mozayeni
BR. Differentiating vascular pathophysiological states by
objective analysis of flow dynamics. J Neuroimaging.
14. Speilberger CD, Gorsuch RL. Manual for the State-Trait
Anxiety Inventory (STAI). Palo Alto, Calif: Consulting
Psychologists Press; 1970.
15. Hardt J, Gerbershagen HU, Franke P. The Symptom
Checklist SCL-90-R: its use and characteristics in chronic
pain patients. Eur J Pain. 2002;4:137–148.
16. Savourey G, Guinet A, Besnard Y, Garcia N, Hanniquet
A, Bittel J. Are the laboratory and field conditions obser-
vations of acute mountain sickness related? Aviat Space
Environ Med. 1997;68:895–899.
17. Kao WF, Kuo CC, Hsu TF, et al. Acute mountain sickness
in Jade Mountain climbers of Taiwan. Aviat Space Environ
18. Aaslid R, Markwalder TM, Nornes H. Noninvasive trans-
cranial Doppler ultrasound recording of flow velocity in
basal cerebral arteries. J Neurosurg. 1982;57:769–774.
19. Baumgartner RW, Bartsch P, Maggiorini M, Waber U,
Oelz O. Enhanced cerebral blood flow in acute mountain
sickness. Aviat Space Environ Med. 1994;65:726–729.
7 Download full-text
Dynamic Vascular Analysis of Change in Altitude
20. Jansen GF, Krins A, Basnyat B. Cerebral vasomotor re-
activity at high altitude in humans. J Appl Physiol. 2003;
21. Huang SY, Sun S, Droma T, et al. Internal carotid arterial
flow velocity during exercise in Tibetan and Han residents
of Lhasa. J Appl Physiol. 2003;73:2638–2642.
22. Jensen JB, Sperling B, Severinghaus JW, Lassen NA. Aug-
mented hypoxic cerebral vasodilation in men during 5
days at 3,810 m altitude. J Appl Physiol. 1996;80:1214–
23. Lassen NA. Increase of cerebral blood flow at high alti-
tude: its possible relation to AMS. Int J Sports Med. 1992;
24. Jansen GF, Krins A, Basnyat B, Bosch A, Odoom JA.
Cerebral autoregulation in subjects adapted and not adapt-
ed to high altitude. Stroke. 2000;31:2314–2318.
25. Lui Y, Harder DR, Lombard JH. Interaction of myogenic
mechanisms and hypoxic dilation in rat middle cerebral
arteries. Am J Physiol Heart Circ Physiol. 2002;283:
26. Weiskopf RB, Gabel RA, Fencl V. Alkaline shift in lumbar
and intracranial CSF in man after 5 days at high altitude.
J Appl Physiol. 1976;41:93–97.
27. Clench J, Ferrell RE, Schull WJ. Effect of chronic altitude
hypoxia on hematologic and glycolytic parameters. Am J
28. Schechter AN, Gladwin MT. Hemaglobin and the para-
crine and endocrine functions of nitric oxide. N Engl J
29. Abbott NJ. Astrocyte-endothelial interactions and blood-
brain barrier permeability. J Anat. 2002;200:629–638.
30. Brown RC, Davis TP. Calcium modulation of adherens
and tight junction function. Stroke. 2002;33:1706–1711.
31. Pachter JS, De Vries HE, Fabry Z. The blood-brain barrier
and its role in immune privilege in the central nervous
system. J Neuropathol Exp Neurol. 2003;62:593–604.
32. Van Hinsbergh VW, Van Nieuw Amerongen GP. Intracel-
lular signalling involved in modulating human endothelial
barrier function. J Anat. 2002;200:549–560.
33. Severinghaus JW. Hypothetical roles of angiogenesis, os-
motic swelling, and ischemia in high-altitude cerebral ede-
ma. J Appl Physiol. 1995;79:375–379.
34. Walter R, Maggiorini M, Scherrer U, Contesse J, Reinhart
WH. Effects of high-altitude exposure on vascular endo-
thelial growth factor levels in man. Eur J Appl Physiol.
35. Maloney J, Wang D, Duncan T, Voelkel N, Ruoss S. Plas-
ma vascular endothelial growth factor in acute mountain
sickness. Chest. 2000;118:47–52.
36. Witt K, Marks KS, Hom S, Davis TP. Effects of hypoxia-
reoxygenation on rat blood-barrier permeability and tight
junction protein expression. Am J Physical Heart Circ
Physiol. 2003 Dec;285(6):H2820–31.
37. Golanov EV, Christensen JR, Reis DJ. Neurons of a lim-
ited subthalamic area mediate elevations in cortical cere-
bral blood flow evoked by hypoxia and excitation of neu-
rons of the rostral ventrolateral medulla. J Neurosci. 2001;