J Appl Physiol 121: 11511159, 2016.

First published September 22, 2016; doi:10.1152/japplphysiol.00394.2016.

Pulmonary artery pressure and arterial oxygen saturation in people living at

high or low altitude: systematic review and meta-analysis
Rodrigo Soria,1 Matthias Egger,2,3 Urs Scherrer,1,4 Nicole Bender,2,5* and Stefano F. Rimoldi1*
1
Department of Cardiology and Clinical Research, Inselspital, University of Bern, Switzerland; 2Institute of Social and
Preventive Medicine (ISPM), University of Bern, Switzerland; 3Division of Epidemiology and Biostatistics, School of Public
Health and Family Medicine, University of Cape Town, Cape Town, South Africa; 4Facultad de Ciencias, Departamento de
Biologa, Universidad de Tarapac, Arica, Chile; and 5Institute of Evolutionary Medicine, University of Zurich, Switzerland
Submitted 26 April 2016; accepted in final form 21 September 2016

Soria R, Egger M, Scherrer U, Bender N, Rimoldi SF. Pulmo- MORE THAN 140 MILLION people worldwide live at high altitude,
nary artery pressure and arterial oxygen saturation in people living at defined as 2,500 m above sea level or higher (32). At high
high or low altitude: systematic review and meta-analysis. J Appl altitude, arterial hypoxemia related to decreased barometric
Physiol 121: 11511159, 2016. First published September 22, 2016; pressure leads to a series of physiological responses (24),
doi:10.1152/japplphysiol.00394.2016.More than 140 million peo-
including increased pulmonary artery pressure (PAP) and pul-
ple are living at high altitude worldwide. An increase of pulmonary

Downloaded from http://jap.physiology.org/ by 10.220.32.247 on November 25, 2017

artery pressure (PAP) is a hallmark of high-altitude exposure and, if monary vascular resistance (45). If pronounced, these pulmo-
pronounced, may be associated with important morbidity and mortal- nary vascular changes may contribute to the increased morbid-
ity. Surprisingly, there is little information on the usual PAP in ity and mortality associated with living at high altitude (4, 23).
high-altitude populations. We, therefore, conducted a systematic re- Although pulmonary vascular maladaptation to high altitude
view (MEDLINE and EMBASE) and meta-analysis of studies pub- may have important consequences on long-term health and
lished (in English or Spanish) between 2000 and 2015 on echocar- disease, there is little information on the usual PAP and arterial
diographic estimations of PAP and measurements of arterial oxygen oxygen saturation in populations living at high altitude. This is
saturation in apparently healthy participants from general populations related, at least in part, to the fact that measurement of PAP
of high-altitude dwellers (2,500 m). For comparison, we similarly necessitated right heart catheterization that due to its invasive
analyzed data published on these variables during the same period for nature could not be used to study larger groups of apparently
populations living at low altitude. Twelve high-altitude studies com-
healthy high-altitude dwellers under field conditions.
prising 834 participants and 18 low-altitude studies (710 participants)
fulfilled the inclusion criteria. All but one high-altitude studies were The advent of echocardiography as a reliable noninvasive
performed between 3,600 and 4,350 m. The combined mean systolic tool to estimate PAP in humans has changed this situation and
PAP (right ventricular-to-right atrial pressure gradient) at high altitude allowed investigators to examine normal values of PAP at high
[25.3 mmHg, 95% confidence interval (CI) 24.0, 26.7], as expected altitude (28). When comparing groups of subjects, echocardio-
was significantly (P 0.001) higher than at low altitude (18.4 mmHg, graphic estimation of PAP has convincingly been shown to be
95% CI 17.1,19.7), and arterial oxygen saturation was significantly accurate vs. right heart catheterization (2, 8, 9), thus applicable
lower (90.4%, 95% CI 89.3, 91.5) than at low altitude (98.1%; 95% CI to population studies. Since the beginning of this century, an
97.7, 98.4). These findings indicate that at an altitude where the very increasing number of studies have reported PAP data on
large majority of high-altitude populations are living, pulmonary echocardiographic estimations of PAP in healthy high-altitude
hypertension appears to be rare. The reference values and distributions dwellers (17, 21, 35, 39). However, since these studies gener-
for PAP and arterial oxygen saturation in apparently healthy high-
altitude dwellers provided by this meta-analysis will be useful to
ally include a relatively small numbers of participants, PAP in
future studies on the adjustments to high altitude in humans. this population remains ill defined at present.
To fill this gap, we systematically reviewed and meta-
high altitude; systolic pulmonary artery pressure; meta-analysis; echo- analyzed data published during the past 15 yr on echocardio-
cardiography graphic estimations of PAP and measurements of arterial
oxygen saturation in apparently healthy participants from gen-
eral populations of high-altitude dwellers. For comparison, we
similarly analyzed data published on these variables during the
NEW & NOTEWORTHY
same period for populations living at low altitude.
Many millions of people are living at high altitude. An increase
MATERIALS AND METHODS
of pulmonary artery pressure (PAP) is a hallmark of high-
altitude exposure and may be associated with important mor- We conducted a systematic review and meta-analysis of studies
bidity, however there is little information on it. The findings of reporting echocardiographic estimates of PAP in general populations
this systematic review and meta-analysis show that pulmonary living at high or low altitude. We limited our search to the past 15 yr,
hypertension in healthy high-altitude dwellers is rare. The i.e., the period echocardiography became the gold standard to estimate
provided reference values and distributions for PAP and SaO2 PAP in healthy people. Reviewing and reporting were performed
will be useful to future studies on high-altitude (mal)adaptation according to the recommendations of the Cochrane Collaboration (15)
in humans. and PRISMA (25), respectively.
Data Sources and Searches
* N. Bender and S. F. Rimoldi contributed equally to this work.
Address for reprint requests and other correspondence: S. F. Rimoldi, Dept. We searched the MEDLINE and EMBASE databases from January
of Cardiology and Clinical Research, Inselspital, Univ. of Bern, Freiburgstr., 2000 to November 2015 on the PubMed and Ovid platforms. We used
BHH, A143, CH-3010 Bern, Switzerland (e-mail: [email protected]). free text words and specific thesaurus terms (MeSH in MEDLINE and

http://www.jappl.org 8750-7587/16 Copyright 2016 the American Physiological Society 1151

1152 Pulmonary Artery Pressure at High Altitude Soria et al.

EMTREE in EMBASE), including pulmonary artery, pressure, diography in general populations living at high (2,500 m above sea
and altitude and terms denoting study designs and publication types. level) or low altitude (1,500 m) were included. Participants had to
Separate searches to identify eligible studies from low- and high- have lived at the study altitude for at least 1 yr. We included
altitude regions were developed. We also examined the bibliographies randomized controlled trials, cohort studies, and cross-sectional stud-
of relevant articles to identify eligible studies that may have been ies published in English or Spanish. Studies reporting data in 10
missed by the database searches. participants and case reports and studies reporting duplicate data or
Study Selection data in participants with morbidities (e.g., systemic hypertension,
diabetes mellitus, etc.) were excluded. The abstracts of retrieved
Articles that reported the mean value and standard deviation (SD) articles were first checked for eligibility, and ineligible studies were
[or standard error (SE)] of estimations of PAP by Doppler echocar- excluded at this stage. The full text of potentially eligible articles was

High Altitude Low Altitude

Identification

Records Records
identified identified

Downloaded from http://jap.physiology.org/ by 10.220.32.247 on November 25, 2017

through through
database database
searching searching
(n=388) (n=2117)
Duplicates
Duplicates removed (n=368)
removed (n=52)
Screening

Excluded based on
titles and abstracts
Titles and abstracts Titles and abstracts (n=1610)
Excluded based on screened (n=336) screened (n=1749) Reasons:
titles and abstracts Studies in animals
(n=313) (n=310)
Reasons: Other languages
Studies with animals (n= 62)
(n=63) Catheterization (n=5)
Other languages Different pathologies
(n= 21) (n=885)
Studies at low Case report (n=12)
altitude (n=55) Without variables of
Different pathologies interest (n= 90)
(n=48) Meta-analysis (n=8)
Case report (n=6) Information not
Without variables of available (n=23)
interest (n= 68)
Eligibility

Other reasons (n=52)

Full text articles Full text articles Other reasons
(n=215)
assessed assessed
(n=23) (n=139)
Full text articles
excluded
Fulltext articles
(n=121)
excluded
Reasons:
(n=11)
Control group was
Reasons:
mixed (n=51)
Groups with <10
Withoutall variables
participants
(n=52)
(n=3)
Groups with <10
Groups considered as
participants
healthy had
(n=7)
Included

comorbidities (n=6)
Measurement of
Intermittent
pulmonary artery
exposure to high
pressure with
altitude (n=2) Studies included in Studies included in
different method
the the (n=6)
analysis analysis Measurement were
performed at high
(n=12) (n=18) altitude (n=5)

Fig. 1. Flow of information through the different phases of systematic review. n, Number of studies.

J Appl Physiol doi:10.1152/japplphysiol.00394.2016 www.jappl.org

 Pulmonary Artery Pressure at High Altitude Soria et al. 1153
then examined, and depending on whether eligibility criteria were met the analysis. Discrepancies between the two reviewers were resolved
or not, articles were included or excluded. by discussion and consultation with the senior authors (ME, US, or
SFR).
Data Extraction
Two reviewers (RS, NB) extracted data, using a data extraction Methodological and Reporting Quality
sheet developed and piloted for this review. The two reviewers Methodological and reporting quality of studies was assessed by
extracted the mean, SD, or SE of systolic PAP and, if reported, the the following criteria: 1) Were there clearly defined inclusion and
mean, SD, or SE of arterial oxygen saturation. In this study, we used exclusion criteria? 2) Was the process of recruiting study participants
the right ventricular-to-atrial pressure gradient as an estimate of clearly described? 3) How many participants were excluded from the
systolic PAP. In the majority of the included studies, PAP was derived analysis, for example, due to missing data? 4) Based on the answers
from continuous wave Doppler sampling of the tricuspid regurgitation to the first three questions, was the study population representative of
jet, using the simplified Bernoulli equation. In a few studies where the general population?; 5) Was the echocardiography performed
instead of systolic, mean PAP was reported, we used the following according to established standards (12)?; 6) Were potential sources of
formula to calculate systolic PAP (11): Systolic PAP (mean PAP bias addressed in the Discussion?
2)/0.61.
If atrial pressure had been added to the pressure gradient, we Statistical Analysis
subtracted these estimates to obtain a comparable set of data for
meta-analysis. We recorded bibliographic details, number, sex, age, We combined the data on systolic PAP and oxygen saturation using
and ethnicity of the study participants and whether they had lived for random-effects meta-analysis, stratified by altitude (high or low).

Downloaded from http://jap.physiology.org/ by 10.220.32.247 on November 25, 2017

at least 1 yr at high, respectively, low altitude before study, study Statistical evidence for heterogeneity between studies was assessed by
location and its altitude, and the number of participants excluded from the I2 statistic (16). We converted SD to SE by dividing the SD by the

Table 1. Characteristics of study participants and settings

First Author (year) Subjects Man Woman Mean (SD) age, yr* Study Location (Country) Altitude, m Ethnicity

High-altitude studies
Hoit (2005) (17) 57 20 0 33 (9) Panam Xiang (China) 4,200 Tibetan
0 37 29 (12) 4,200
Schwab (2008) (41) 34 24 10 36 (9) La Paz (Bolivia) 3,600 Aymara
34 21 13 38 (7) La Paz (Bolivia) 3,600 Caucasian
Stuber (2008) (43) 200 105 95 9.5 (3.6) La Paz (Bolivia) 4,000 Aymara
3,600
77 42 35 9.7 (3.7) La Paz (Bolivia) 3,600 Caucasian
Richalet (2008) (39) 15 15 0 44 (9) Cerro de Pasco (Peru) 4,300 Quechua
Stuber (2010) (44) 32 32 0 46 (11) La Paz (Bolivia) 3,600 Aymara
Jayet (2010) (21) 90 55 35 14 (7) La Paz (Bolivia) 3,600 Aymara
Hoit (2011) (18) 75 53 0 32 (9) Amhara (Ethiopia) 3,700 Amhara
0 22 30 (8) 3,700
Pratali (2012) (35) 26 26 0 48 (7) La Paz (Bolivia) 3,600 Aymara
Groepenhoff (2012) (13) 15 9 6 41 (7.7) Cerro de Pasco (Peru) 4,350 Quechua
Pratali (2013) (34) 40 40 0 48 (8) La Paz (Bolivia) 3,600 Aymara
Bruno (2014) (7) 95 30 65 33.7 (13.8) Khumbu Valley (Nepal) 3,800 Tibetan
Faoro (2014) (11) 13 10 3 25 (3.6) Khumbu-Sagarmatha National Park 5,050 Tibetan
Low-altitude studies
Aessopos (2000) (1) 53 19 34 38.9 (12.9) Athens (Greece) 338 Caucasian
Grnig (2000) (14) 11 11 0 37 (11) Heidelberg (Germany) 100 Caucasian
Lindqvist (2003) (26) 20 10 10 46 (12) Sweden Caucasian
Ricart (2005) (38) 17 17 0 33.1 (7.2) Barcelona (Spain) 12 Caucasian
Modesti (2006) (29) 20 17 3 3253 (range) Florence (Italy) 50 Caucasian
Senn (2006) (42) 26 13 13 28 (11) Zurich (Switzerland) 490 Caucasian
Reichenberger (2007) (36) 14 12 2 37 (mean) Giessen (Germany) 171 Caucasian
Huez (2007) (20) 13 2 11 54 (12) Brussels (Belgium) 13 Caucasian
Stuber (2008) (43) 29 16 13 8.8 (2.6) Bern (Switzerland) 450 Caucasian
Kriemler (2008) (22) 20 10.7 (1.1) Bern (Switzerland) 450 Caucasian
Huez (2009) (19) 15 7 8 36 (12) Brussels (Belgium) 13 Caucasian
Baillie (2009) (5) 42 26 16 22.4 (6.3) Edinburgh (Scotland) 47 Caucasian
41 18 23 21.2 (2.3)
Naeije (2010) (31) 30 15 15 35 (mean) Brussels (Belgium) 13 Caucasian
Bates (2011) (6) 42 26 16 21.5 (2.7) Edinburgh (Scotland) 47 Caucasian
20 11 9 21.2 (3)
Hoit (2011) (18) 46 21 0 32 (9) Cleveland (USA) 282 Caucasian
0 25 38 (12)
54 45 0 34 (9) Zarima (Ethiopia) 1,200 Amhara
0 9 28 (9)
Allemann (2012) (3) 118 63 55 11 (2) Bern (Switzerland) 568 Caucasian
Groepenhoff (2012) (13) 15 9 6 35 (11.6) Brussels (Belgium) 13 Caucasian
Bruno (2014) (7) 64 24 40 36.2 (12.4) Pisa (Italy) 4 Caucasian
*Mean (SD) is shown unless otherwise indicated.

J Appl Physiol doi:10.1152/japplphysiol.00394.2016 www.jappl.org

1154 Pulmonary Artery Pressure at High Altitude Soria et al.

square root of the number of participants included in the analysis. We of 139 full-text low-altitude articles were examined, and 18
used random-effects meta-regression to assess the effects of age, articles included. We excluded 121 studies, mostly because
ethnicity, and altitude of measurement on the results. Results from participants with comorbidities, (e.g., diabetes mellitus, arterial
meta-analyses are expressed as mean values and 95% confidence hypertension) were studied.
intervals. Based on the results of the meta-analysis (using the point
estimates and the mean of the standard deviations), we plotted normal
distributions of systolic PAP and oxygen saturation for hypothetical Characteristics of Included Studies
populations at low and high altitude. The statistical package Stata
High-altitude studies. Twelve studies with a total of 834
(version 11.2; Stat, College Station, TX) was used for all analyses.
participants fulfilled the inclusion criteria. The number of
RESULTS
participants included ranged from 200 in the largest (39) to 13
in the smallest study (35) with mean ages of the participants
Identification of Eligible Studies ranging from 9.5 to 48 yr (Table 1). With the exception of one
study that was performed at 5,050 m, all others were performed
We identified a total of 2,505 articles, 388 studies from between 3,600 and 4,350 m. Slightly more than half of the
high-altitude settings and 2,117 from low altitude. The flow of participants studied (54%) had an Andean indigenous back-
the selection of studies and the reasons for exclusion are shown ground; the remaining participants were made up by Asians
in Fig. 1. (24%), Caucasians (13%), and Africans (9%).
High-altitude studies. Database searching revealed 388 arti- Low-altitude studies. Eighteen studies including a total of

Downloaded from http://jap.physiology.org/ by 10.220.32.247 on November 25, 2017

cles. Based on titles and abstracts we excluded 52 duplicates 710 participants with a mean age ranging from 8.8 to 54 yr
and 313 articles that did not meet inclusion criteria. We fulfilled the inclusion criteria (Table 1). The vast majority of
examined 23 full-text articles and included 12 articles. Eleven the participants were Europeans (86%), the remainder Africans
articles were excluded because they did not meet the inclusion (8%) and North Americans (6%).
criteria (10 participants, comorbidities, high-altitude expo-
sure 1 yr) (Fig. 1). Methodological and Reporting Quality
Low-altitude studies. Database searching revealed 2,117
articles. Based on titles and abstracts, we excluded 368 dupli- Due to incomplete reporting of the recruitment or eligibility
cates and 1,610 articles that did meet inclusion criteria. A total criteria of participants, it was not always possible to determine

Table 2. Assessment of methodological quality of included studies

Was the process of According to the previous
Were there clearly recruiting study How many participants questions, was the study Was echocardiography Were potential sources
defined inclusion/ participants clearly were excluded from population representative done according to of bias addressed in
exclusion criteria? described? the analysis (%)? for the general population? established standards? the Discussion?

High-altitude Studies
Hoit 2005 (17) yes unclear 8 (9.1) unclear yes yes
Schwab 2008 (41) yes yes 0 (0) yes yes no
Stuber 2008 (43) yes no 14 (5) unclear yes no
Richalet 2008 (39) yes no 0 (0) unclear yes yes
Stuber 2010 (44) yes no 1 (3.1) unclear yes no
Jayet 2010 (21) yes unclear 6 (6.7) unclear yes yes
Hoit 2011 (18) yes unclear unclear unclear yes no
Pratali 2012 (35) yes no unclear unclear yes yes
Groepenhoff 2012 (13) yes no unclear unclear yes no
Pratali 2013 (34) yes no 0 (0) unclear yes yes
Bruno 2014 (7) yes yes unclear unclear yes yes
Faoro 2014 (11) yes no 0 (0) unclear yes yes
Low-altitude Studies
Aessopos (2000) (1) yes yes no unclear yes yes
Grnig (2000) (14) yes yes unclear unclear yes yes
Lindqvist (2003) (26) unclear no 0 (0) unclear yes yes
Ricart 2005 (38) no no 3 (17.6) unclear yes no
Modesti 2006 (29) unclear yes 0 (0) unclear yes no
Senn 2006 (42) unclear no 10 (38.5) unclear yes no
Reichenberger 2007 (36) yes yes 0 (0) yes unclear no
Huez 2007 (20) no no 0 (0) unclear yes unclear
Stuber 2008 (43) unclear no 1 (3.4) unclear yes no
Kriemler 2008 (22) yes yes unclear unclear yes yes
Huez 2009 (19 unclear no 9 (60) unclear yes no
Baillie 2009 (5) no no 16 (19.3) unclear unclear yes
Naeije 2010 (31) yes no 8 (26.7) unclear yes yes
Bates 2011 (6) unclear no 7 (11.3) unclear yes yes
Hoit 2011 (18) yes unclear unclear unclear yes no
Allemann 2012 (3) yes no 5 (4.2) unclear yes yes
Groepenhoff 2012 (13) yes no unclear unclear yes no
Bruno 2014 (7) yes yes unclear unclear yes yes

J Appl Physiol doi:10.1152/japplphysiol.00394.2016 www.jappl.org

 Pulmonary Artery Pressure at High Altitude Soria et al. 1155
whether or not participants were truly representative of the pected, arterial oxygen saturation also differed significantly
general population. In two studies, one done at high (41) and (P 0.001) between high and low altitude. The combined
the other at low altitude (36) the study population likely were mean arterial oxygen saturation at rest was 90.5% (95% CI
representative of the general population. In almost all studies, 89.3, 91.6) at high and 98.1% at low altitude (95% CI 97.8,
the echocardiography was performed according to the quality 98.4). At high altitude, arterial oxygen saturation varied be-
criteria of the European Association of Echocardiography (12). tween 83.3% among Tibetans living at 4,200 m and 93.6%
Roughly half of the studies discussed potential sources of bias among Caucasians living at 3,600 m (Fig. 3), whereas at low
(Table 2). altitude it varied between 96.0 and 100%. In meta-regression
analyses of high- and low-altitude studies no statistically sig-
Meta-analysis of PAP and Oxygen Saturation at High and nificant associations (P 0.26) were evident between PAP or
Low Altitude oxygen saturation and ethnicity or age. Oxygen saturation
showed a significant association with altitude at low altitude
At high altitude, the point estimate from meta-analysis of the
(P 0.018) and at high altitude (P 0.001).
mean systolic PAP at rest was 25.3 mmHg [95% confidence
interval (CI) 24.0, 26.7], whereas at low altitude the point Simulated Distributions of PAP and Oxygen Saturation at
estimate, as expected, was significantly (P 0.001) lower High and Low Altitude
(18.4 mmHg, 95% CI 17.1, 19.7). At high altitude, mean
systolic PAP ranged from 22.0 to 30.0 mmHg and at low The simulated distribution of systolic PAP at high altitude

Downloaded from http://jap.physiology.org/ by 10.220.32.247 on November 25, 2017

altitude from 12.1 to 26.0 mmHg (Fig. 2). All high-altitude was shifted toward higher values compared with low altitude
studies were performed by experienced echocardiographers (Fig. 4A). Furthermore, high-altitude values showed a wider
having published extensively in the field. Eight of the 18 distribution (SD 5.3 mmHg) than low-altitude values (SD 4.2
low-altitude studies included in this meta-analysis were per- mmHg). The distribution of oxygen saturation was shifted to
formed by these same experienced investigators. To test for the left at high altitude compared with low altitude, with the
potential bias related to the multiexaminer approach in low- high-altitude values showing a flatter distribution (SD 2.9%)
altitude studies (and possible inclusion of studies performed by than low-altitude values (SD 1.8%) (Fig. 4B).
less experienced ultrasonographers), we separately analyzed DISCUSSION
low-altitude PAP data reported in these 11 studies; mean
systolic PAP was comparable (18.6 mmHg, 95% CI 17.1, 20.1) Increased PAP is a hallmark of the cardiovascular response
with mean systolic PAP of all low-altitude studies. As ex- to high-altitude exposure, which, if pronounced, may lead to

Fig. 2. Meta-analysis of mean systolic pul-

monary artery pressure at rest and its 95%
confidence interval (CI) in healthy high- and
low-altitude dwellers. I-squared shows the
heterogeneity.

J Appl Physiol doi:10.1152/japplphysiol.00394.2016 www.jappl.org

1156 Pulmonary Artery Pressure at High Altitude Soria et al.

Fig. 3. Meta-analysis of mean arterial oxy-

gen saturation at rest and its 95% CI in
healthy high- and low-altitude dwellers.
I-squared shows the heterogeneity.

Downloaded from http://jap.physiology.org/ by 10.220.32.247 on November 25, 2017

cardiovascular morbidity and mortality. Surprisingly, there is agreement between invasive and noninvasive measurements
little information on the usual PAP in the general population when comparing groups of subjects (2, 8, 28). All high-altitude
living at high altitude. Here, we meta-analyzed for the first time studies and 8 of the 18 low-altitude studies were performed by
all published data between 2000 and 2015 on echocardio- experienced investigators who have published extensively in
graphic estimates of pulmonary artery pressure and measure- the field of echocardiographic estimation of PAP. The multi-
ments of arterial oxygen saturation in 834 people living at high examiner approach at low altitude does not appear to have
altitude. For comparison, we also meta-analyzed data pub- resulted in bias, since the mean values for systolic PAP of the
lished on these variables during the same period in 710 people studies performed by experienced investigators and of all
living at low altitude. Meta-analysis revealed a mean resting studies meta-analyzed were almost identical. In line with these
systolic PAP of 25 mmHg and a mean arterial oxygen satura- observations, invasive measurements of systolic PAP, dating
tion of 90.5% in apparently healthy individuals from the from the pre-echocardiography area, between 33 and 38 mmHg
general population living between 3,600 and 5,050 m, com- in small numbers of presumably healthy subjects living be-
pared with 18 mmHg and 98.1%, respectively, in individuals tween 3,400 and 4,300 m (17), after subtracting 5 to 10 mmHg
living below 1,200 m. Using the current definition of pulmo- for atrial pressure, are in agreement with the noninvasive
nary hypertension as a mean PAP 25 mmHg (ERS) or 30 echocardiographic estimations of PAP meta-analyzed in the
(high-altitude consensus) the findings of our meta-analysis present study.
(systolic PAP of 25 mean PAP of 19.5 mmHg) indicate that High altitude is defined as an altitude 2,500 m. Surpris-
in apparently healthy individuals living at high altitude (3,600- ingly, our search did not reveal any study reporting PAP
4,200 m, i.e., an altitude where the major high-altitude popu- measurements in high-altitude dwellers living between 2,500
lations live worldwide) pulmonary hypertension appears to be and 3,600 m that fulfilled the eligibility criteria. Thus, there is
a very rare event (40). a need for future studies to fill this important gap. It is likely
The present data represent by far the largest dataset on PAP that this lack of information on PAP between 2,500 and 3,600
in high-altitude dwellers reported so far. Echocardiography is m explains the absence of a significant relationship between
the standard technique to assess PAP in high-altitude popula- PAP and altitude among the high-altitude studies. More inter-
tions, since for ethical and logistic reasons, invasive measure- estingly, the absence of such a relationship suggests that for
ments are not feasible in population-based studies. Echocar- altitudes between 3,600 and 4,200 m (the study altitudes of all
diographic estimations have been validated against invasive but one of the articles meta-analyzed), there is little variation of
measurements of PAP at high and low altitude, with good PAP in healthy high-altitude dwellers. This could imply that

J Appl Physiol doi:10.1152/japplphysiol.00394.2016 www.jappl.org

 Pulmonary Artery Pressure at High Altitude Soria et al. 1157
would be important to test for this association in larger studies
and other high-altitude populations in future studies.
Genetic analyses provide evidence for differences in the
evolutionary adaptation to high altitude between Andean, Ti-
betan and Ethiopian high-altitude populations that may have
resulted in different human high-altitude phenotypes (10, 27).
For example, it has been suggested that Tibetans may have
lower PAP than Andean high-altitude populations. The present
meta-analysis does not allow us to confirm this suggestion. If
one uses a correction for hematocrit (11), as recommended in
the consensus statement on high-altitude diseases (23), the
meta-analyzed data on PAP in Andean high-altitude dwellers
would tend to be even lower than those in Tibetans (21.9 2.2
vs. 24.4 1.5 mmHg). There is clearly a need for additional
larger studies of PAP in the general population of Tibetan and
Ethiopian high-altitude dwellers. Interestingly, the meta-ana-
lyzed data on arterial oxygen saturation demonstrate an effect
of ethnic background, because arterial oxygen saturation in

Downloaded from http://jap.physiology.org/ by 10.220.32.247 on November 25, 2017

Tibetans is considerably lower than in Andeans. Since it is
thought that arterial oxygen saturation is an important deter-
minant of PAP, this observation could suggest that Tibetans are
better capable of maintaining a normal PAP in the face of a low
arterial oxygen saturation than Andeans.
This meta-analysis also goes against the current belief that
pulmonary hypertension is common at high altitude and is in
proportion to decreased arterial oxygen saturation (32). This
concept was based on data obtained by heart catheterization
studies, and it appears possible that selection bias (sick subjects
are more likely to accept invasive procedures) may have led to
this conclusion. It needs to be noted here, however, that similar
to sea-level natives, severe high-altitude pulmonary hyperten-
sion may occur in a small proportion (probably not more than
Fig. 4. Predicted distributions of mean systolic pulmonary artery pressure (A)
1%, because, of the 834 high-altitude dwellers included in the
and arterial oxygen saturation (B) in high- and low-altitude dwellers. meta-analyzed studies, 1% had a PAP 27.1 mmHg) of
healthy high-altitude natives (30) and then become a cause of
high-altitude right heart failure. Finally, the present data call
the average resting systolic PAP of 25 mmHg found in the for an urgent revision of the consensus document definition of
present meta-analysis represents the normal pulmonary artery high-altitude pulmonary hypertension (mean PAP 30 mmHg,
pressure in the general population for the altitude range where systolic PAP 50 mmHg) (23) and a consensus on how to
the very vast majority of human high-altitude populations are account for the effects of increased hematocrit on this variable.
living worldwide. With regard to the low-altitude studies, we did not include in
We used meta-regression models to identify factors that may our meta-analysis a very large study on reference intervals of
have influenced PAP in high-altitude dwellers. The mean age PAP among echocardiographically normal subjects, because
of the study population did not appear to have a significant the study was based on patients rather than the general popu-
effect on PAP. In contrast, a study in children (mean age 9.5 lation (28). Of note, however, the mean systolic PAP (18.3
yr) living at high altitude (43), and a very large study at low mmHg) reported in this study is almost identical to the combined
altitude encompassing participants between 1 and 89 yr (28) estimate (18.4 mmHg) of the present meta-analysis. Finally, it is
found a significant trend for PAP to increase with age. In the unlikely that different ethnicities reported between the low- and
latter study, the age-associated increase of PAP at low altitude high-altitude cohorts undermined the conclusion of the meta-
was mainly evident in participants 50 yr or older but much less analysis, since at low altitude PAP has been reported normal and
pronounced in those 50 yr. It is possible that we may have comparable with Caucasians in Andean (32), Amhara (18), and
missed a weak association between age and PAP, because in all Tibetan (33), (Scherrer U, Rexhaj E, Sartori C, Rimoldi SF,
but one study meta-analyzed here the mean age was often unpublished observations) high-altitude populations.
markedly 50 yr, and we did not have access to the individual To conclude, this is the first systematic review and meta-
participants data, limiting our analysis of the association of analysis of studies reporting echocardiographic estimations of
PAP with age to aggregate-level (ecological) analyses. We did PAP and measurements of arterial oxygen saturation in healthy
not detect a significant relationship between daytime arterial people of the general population living at high or low altitude.
oxygen saturation and PAP. Sleep-disordered breathing is not The analysis shows that for altitudes between 3,600 and 4,300
infrequent at high altitude, and recently, an inverse relationship m at which the very large majority of high-altitude populations
between nocturnal oxygen saturation and daytime systolic PAP are living worldwide, mean systolic PAP in the general popu-
in Andean high-altitude dwellers has been reported (37). It lation is roughly 7 mmHg higher than at low altitude. More-

J Appl Physiol doi:10.1152/japplphysiol.00394.2016 www.jappl.org

1158 Pulmonary Artery Pressure at High Altitude Soria et al.

over, pulmonary hypertension appears to be very rare among 8. Claessen G, La Gerche A, Voigt JU, Dymarkowski S, Schnell F, Petit
apparently healthy high-altitude inhabitants. The results from T, Willems R, Claus P, Delcroix M, Heidbuchel H. Accuracy of
echocardiography to evaluate pulmonary vascular and RV function during
this meta-analysis are expected to provide useful reference exercise. JACC Cardiovasc Imaging 9: 532543, 2016. doi:10.1016/
values for these variables in the general population, and j.jcmg.2015.06.018.
marked deviations from these values, such as those in meta- 9. DAlto M, Romeo E, Argiento P, DAndrea A, Vanderpool R, Correra
analyzed studies (6) and (42), should render suspicion of a A, Bossone E, Sarubbi B, Calabr R, Russo MG, Naeije R. Accuracy
potential systematic error. More research in distinct ethnic and precision of echocardiography versus right heart catheterization for
groups, particularly Tibetans and Africans, living at high and the assessment of pulmonary hypertension. Int J Cardiol 168: 4058 4062,
2013. doi:10.1016/j.ijcard.2013.07.005.
low altitudes is needed to provide information on how evolu- 10. Eichstaedt CA, Anto T, Pagani L, Cardona A, Kivisild T, Mormina
tion may have resulted in different high-altitude phenotypes in M. The Andean adaptive toolkit to counteract high altitude maladaptation:
humans. genome-wide and phenotypic analysis of the Collas. PLoS One 9: e93314,
2014. doi:10.1371/journal.pone.0093314.
ACKNOWLEDGMENTS 11. Faoro V, Huez S, Vanderpool R, Groepenhoff H, de Bisschop C,
Martinot JB, Lamotte M, Pavelescu A, Gunard H, Naeije R. Pulmo-
We thank Drs. Yves Allemann, Emrush Rexhaj, David Cerny, Marcel
Zwahlen, Beatrice Minder, Doris Kopp for helpful comments and guidance. nary circulation and gas exchange at exercise in Sherpas at high altitude.
J Appl Physiol (1985) 116: 919 926, 2014. doi:10.1152/japplphysiol.
GRANTS 00236.2013.
12. Galderisi M, Henein MY, Dhooge J, Sicari R, Badano LP, Zamorano
The authors acknowledge funding from Bundesbehrden der Schweizeri- JL, Roelandt JR; European Association of Echocardiography. Rec-

Downloaded from http://jap.physiology.org/ by 10.220.32.247 on November 25, 2017

schen Eidgenossenschaft (Federal Authorities of the Swiss Confederation). R. ommendations of the European Association of Echocardiography: how to
Soria was the recipient of a Swiss Federal-Excellence Scholarship for foreign use echo-Doppler in clinical trials: different modalities for different pur-
scholars. poses. Eur J Echocardiogr 12: 339 353, 2011. doi:10.1093/ejechocard/
jer051.
DISCLOSURES 13. Groepenhoff H, Overbeek MJ, Mul M, van der Plas M, Argiento P,
All authors have read and approved the manuscript. This material has not Villafuerte FC, Beloka S, Faoro V, Macarlupu JL, Guenard H, de
been reported previously except as an abstract and is not under consideration Bisschop C, Martinot JB, Vanderpool R, Penaloza D, Naeije R.
for publication elsewhere. No conflicts of interest, financial or otherwise, are Exercise pathophysiology in patients with chronic mountain sickness
declared by the author(s). exercise in chronic mountain sickness. Chest 142: 877884, 2012. doi:
10.1378/chest.11-2845.
AUTHOR CONTRIBUTIONS 14. Grnig E, Mereles D, Hildebrandt W, Swenson ER, Kbler W,
Kuecherer H, Brtsch P. Stress Doppler echocardiography for identifi-
R.S., N.B., and S.F.R. performed experiments; R.S., M.E., U.S., N.B., and
cation of susceptibility to high altitude pulmonary edema. J Am Coll
S.F.R. analyzed data; R.S., M.E., U.S., N.B., and S.F.R. interpreted results of
Cardiol 35: 980 987, 2000. doi:10.1016/S0735-1097(99)00633-6.
experiments; R.S. prepared figures; R.S. drafted manuscript; R.S., M.E., U.S.,
15. Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of
N.B., and S.F.R. approved final version of manuscript; M.E., U.S., N.B., and
S.F.R. conception and design of research; M.E., U.S., N.B., and S.F.R. edited Interventions, version 4.2.6. The Cochrane Collaboration, 2006.
and revised manuscript. 16. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analy-
sis. Stat Med 21: 1539 1558, 2002. doi:10.1002/sim.1186.
REFERENCES 17. Hoit BD, Dalton ND, Erzurum SC, Laskowski D, Strohl KP, Beall
CM. Nitric oxide and cardiopulmonary hemodynamics in Tibetan high-
1. Aessopos A, Farmakis D, Taktikou H, Loukopoulos D. Doppler- landers. J Appl Physiol (1985) 99: 1796 1801, 2005. doi:10.1152/
determined peak systolic tricuspid pressure gradient in persons with japplphysiol.00205.2005.
normal pulmonary function and tricuspid regurgitation. J Am Soc Echo- 18. Hoit BD, Dalton ND, Gebremedhin A, Janocha A, Zimmerman PA,
cardiogr 13: 645649, 2000. doi:10.1067/mje.2000.104535. Zimmerman AM, Strohl KP, Erzurum SC, Beall CM. Elevated pul-
2. Allemann Y, Sartori C, Lepori M, Pierre S, Mlot C, Naeije R, monary artery pressure among Amhara highlanders in Ethiopia. Am J Hum
Scherrer U, Maggiorini M. Echocardiographic and invasive measure- Biol 23: 168 176, 2011. doi:10.1002/ajhb.21130.
ments of pulmonary artery pressure correlate closely at high altitude. Am 19. Huez S, Faoro V, Gunard H, Martinot JB, Naeije R. Echocardio-
J Physiol Heart Circ Physiol 279: H2013H2016, 2000. graphic and tissue Doppler imaging of cardiac adaptation to high altitude
3. Allemann Y, Stuber T, de Marchi SF, Rexhaj E, Sartori C, Scherrer in native highlanders versus acclimatized lowlanders. Am J Cardiol 103:
U, Rimoldi SF. Pulmonary artery pressure and cardiac function in chil- 16051609, 2009. doi:10.1016/j.amjcard.2009.02.006.
dren and adolescents after rapid ascent to 3,450 m. Am J Physiol Heart 20. Huez S, Roufosse F, Vachiry JL, Pavelescu A, Derumeaux G, Wau-
Circ Physiol 302: H2646 H2653, 2012. doi:10.1152/ajpheart.00053. trecht JC, Cogan E, Naeije R. Isolated right ventricular dysfunction in
2012. systemic sclerosis: latent pulmonary hypertension? Eur Respir J 30:
4. Antezana AMAG, Antezana G, Aparicio O, Noriega I, Velarde FL, 928 936, 2007. doi:10.1183/09031936.00025607.
Richalet JP. Pulmonary hypertension in high-altitude chronic hypoxia:
21. Jayet PY, Rimoldi SF, Stuber T, Salmn CS, Hutter D, Rexhaj E,
response to nifedipine. Eur Respir J 12: 11811185, 1998. doi:10.1183/
Thalmann S, Schwab M, Turini P, Sartori-Cucchia C, Nicod P,
09031936.98.12051181.
Villena M, Allemann Y, Scherrer U, Sartori C. Pulmonary and systemic
5. Baillie JK, Thompson AA, Irving JB, Bates MG, Sutherland AI,
vascular dysfunction in young offspring of mothers with preeclampsia.
Macnee W, Maxwell SR, Webb DJ. Oral antioxidant supplementation
does not prevent acute mountain sickness: double blind, randomized Circulation 122: 488 494, 2010. doi:10.1161/CIRCULATIONAHA.
placebo-controlled trial. QJM 102: 341348, 2009. doi:10.1093/qjmed/ 110.941203.
hcp026. 22. Kriemler S, Jansen C, Linka A, Kessel-Schaefer A, Zehnder M,
6. Bates MG, Thompson AA, Baillie JK, Sutherland AI, Irving JB, Schrmann T, Kohler M, Bloch K, Brunner-La Rocca HP. Higher
Hirani N, Webb DJ. Sildenafil citrate for the prevention of high altitude pulmonary artery pressure in children than in adults upon fast ascent to
hypoxic pulmonary hypertension: double blind, randomized, placebo- high altitude. Eur Respir J 32: 664 669, 2008. doi:10.1183/
controlled trial. High Alt Med Biol 12: 207214, 2011. doi:10.1089/ 09031936.00166407.
ham.2011.0007. 23. Len-Velarde F, Maggiorini M, Reeves JT, Aldashev A, Asmus I,
7. Bruno RM, Cogo A, Ghiadoni L, Duo E, Pomidori L, Sharma R, Bernardi L, Ge RL, Hackett P, Kobayashi T, Moore LG, Penaloza D,
Thapa GB, Basnyat B, Bartesaghi M, Picano E, Sicari R, Taddei S, Richalet JP, Roach R, Wu T, Vargas E, Zubieta-Castillo G, Zubieta-
Pratali L. Cardiovascular function in healthy Himalayan high-altitude dwell- Calleja G. Consensus statement on chronic and subacute high altitude
ers. Atherosclerosis 236: 4753, 2014. doi:10.1016/j.atherosclerosis. diseases. High Alt Med Biol 6: 147157, 2005. doi:10.1089/ham.2005.
2014.06.017. 6.147.

J Appl Physiol doi:10.1152/japplphysiol.00394.2016 www.jappl.org

 Pulmonary Artery Pressure at High Altitude Soria et al. 1159
24. Len-Velarde F, Villafuerte FC, Richalet JP. Chronic mountain sick- tissue Doppler imaging study. JACC Cardiovasc Imaging 6: 12871297,
ness and the heart. Prog Cardiovasc Dis 52: 540 549, 2010. doi:10.1016/ 2013. doi:10.1016/j.jcmg.2013.08.007.
j.pcad.2010.02.012. 35. Pratali L, Rimoldi SF, Rexhaj E, Hutter D, Faita F, Salmn CS,
25. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gtzsche PC, Ioannidis Villena M, Sicari R, Picano E, Allemann Y, Scherrer U, Sartori C.
JP, Clarke M, Devereaux PJ, Kleijnen J, Moher D. The PRISMA Exercise induces rapid interstitial lung water accumulation in patients with
statement for reporting systematic reviews and meta-analyses of studies chronic mountain sickness. Chest 141: 953958, 2012. doi:10.1378/
that evaluate health care interventions: explanation and elaboration. PLoS chest.11-0084.
Med 6: e1000100, 2009. doi:10.1371/journal.pmed.1000100. 36. Reichenberger F, Kohstall MG, Seeger T, Olschewski H, Grimminger
26. Lindqvist P, Henein M, Kazzam E. Right ventricular outflow-tract F, Seeger W, Ghofrani HA. Effect of sildenafil on hypoxia-induced
fractional shortening: an applicable measure of right ventricular systolic changes in pulmonary circulation and right ventricular function. Respir
function. Eur J Echocardiogr 4: 29 35, 2003. doi:10.1053/euje.4.1.29. Physiol Neurobiol 159: 196 201, 2007. doi:10.1016/j.resp.2007.07.005.
27. Lorenzo FR, Huff C, Myllymki M, Olenchock B, Swierczek S, Tashi 37. Rexhaj E, Rimoldi SF, Pratali L, Brenner R, Andries D, Soria R,
T, Gordeuk V, Wuren T, Ri-Li G, McClain DA, Khan TM, Koul PA, Salinas C, Villena M, Romero C, Allemann Y, Lovis A, Heinzer R,
Guchhait P, Salama ME, Xing J, Semenza GL, Liberzon E, Wilson A, Sartori C, Scherrer U. Sleep-disordered breathing and vascular function
Simonson TS, Jorde LB, Kaelin WG Jr, Koivunen P, Prchal JT. A in patients with chronic mountain sickness and healthy high-altitude
genetic mechanism for Tibetan high-altitude adaptation. Nat Genet 46: dwellers. Chest 149: 991998, 2016. doi:10.1378/chest.15-1450.
951956, 2014. doi:10.1038/ng.3067. 38. Ricart A, Maristany J, Fort N, Leal C, Pags T, Viscor G. Effects of
28. McQuillan BM, Picard MH, Leavitt M, Weyman AE. Clinical corre- sildenafil on the human response to acute hypoxia and exercise. High Alt
lates and reference intervals for pulmonary artery systolic pressure among Med Biol 6: 4349, 2005. doi:10.1089/ham.2005.6.43.
echocardiographically normal subjects. Circulation 104: 27972802, 39. Richalet JP, Rivera-Ch M, Maignan M, Privat C, Pham I, Macarlupu
JL, Petitjean O, Len-Velarde F. Acetazolamide for Monges disease:
2001. doi:10.1161/hc4801.100076.
efficiency and tolerance of 6-month treatment. Am J Respir Crit Care Med

Downloaded from http://jap.physiology.org/ by 10.220.32.247 on November 25, 2017

29. Modesti PA, Vanni S, Morabito M, Modesti A, Marchetta M, Gam-
177: 1370 1376, 2008. doi:10.1164/rccm.200802-196OC.
beri T, Sofi F, Savia G, Mancia G, Gensini GF, Parati G. Role of
40. Rose G. Sick individuals and sick populations. Int J Epidemiol 14: 3238,
endothelin-1 in exposure to high altitude: Acute Mountain Sickness and
1985. doi:10.1093/ije/14.1.32.
Endothelin-1 (ACME-1) study. Circulation 114: 1410 1416, 2006. doi:
41. Schwab M, Jayet PY, Stuber T, Salinas CE, Bloch J, Spielvogel H,
10.1161/CIRCULATIONAHA.105.605527. Villena M, Allemann Y, Sartori C, Scherrer U. Pulmonary-artery pres-
30. Naeije R. RV is doing well at high altitudes!Always? JACC Cardiovasc sure and exhaled nitric oxide in Bolivian and Caucasian high altitude dwellers.
Imaging 6: 1298 1300, 2013. doi:10.1016/j.jcmg.2013.09.011. High Alt Med Biol 9: 295299, 2008. doi:10.1089/ham.2008.1057.
31. Naeije R, Huez S, Lamotte M, Retailleau K, Neupane S, Abramowicz 42. Senn O, Clarenbach CF, Fischler M, Thalmann R, Brunner-La Rocca
D, Faoro V. Pulmonary artery pressure limits exercise capacity at high H, Egger P, Maggiorini M, Bloch KE. Do changes in lung function
altitude. Eur Respir J 36: 1049 1055, 2010. doi:10.1183/09031936. predict high-altitude pulmonary edema at an early stage? Med Sci Sports
00024410. Exerc 38: 15651570, 2006. doi:10.1249/01.mss.0000229102.37359.af.
32. Penaloza D, Arias-Stella J. The heart and pulmonary circulation at high 43. Stuber T, Sartori C, Salmn CS, Hutter D, Thalmann S, Turini P,
altitudes: healthy highlanders and chronic mountain sickness. Circulation Jayet PY, Schwab M, Sartori-Cucchia C, Villena M, Scherrer U,
115: 11321146, 2007. doi:10.1161/CIRCULATIONAHA.106.624544. Allemann Y. Respiratory nitric oxide and pulmonary artery pressure in
33. Petousi N, Croft QP, Cavalleri GL, Cheng HY, Formenti F, Ishida K, children of Aymara and European ancestry at high altitude. Chest 134:
Lunn D, McCormack M, Shianna KV, Talbot NP, Ratcliffe PJ, 996 1000, 2008. doi:10.1378/chest.08-0854.
Robbins PA. Tibetans living at sea level have a hyporesponsive 44. Stuber T, Sartori C, Schwab M, Jayet PY, Rimoldi SF, Garcin S,
hypoxia-inducible factor system and blunted physiological responses to Thalmann S, Spielvogel H, Salmn CS, Villena M, Scherrer U, Alle-
hypoxia. J Appl Physiol (1985) 116: 893904, 2014. doi:10.1152/ mann Y. Exaggerated pulmonary hypertension during mild exercise in
japplphysiol.00535.2013. chronic mountain sickness. Chest 137: 388 392, 2010. doi:10.1378/
34. Pratali L, Allemann Y, Rimoldi SF, Faita F, Hutter D, Rexhaj E, chest.09-1355.
Brenner R, Bailey DM, Sartori C, Salmon CS, Villena M, Scherrer U, 45. Stuber T, Scherrer U. Circulatory adaptation to long-term high altitude
Picano E, Sicari R. RV contractility and exercise-induced pulmonary exposure in Aymaras and Caucasians. Prog Cardiovasc Dis 52: 534 539,
hypertension in chronic mountain sickness: a stress echocardiographic and 2010. doi:10.1016/j.pcad.2010.02.009.

J Appl Physiol doi:10.1152/japplphysiol.00394.2016 www.jappl.org