Altitude training for sea level performance: a systematic review

Abstract

The aims of this systematic review are: 1) to collate and to critically
evaluate the empirical evidence sustaining the use of natural AT in athletes with the main goal of improving SL performance; and 2) to derive which of the different natural AT strategies is more efficient for enhancing SL performance when the athletes come back to SL training and competition. To achieve these goals, we systematically reviewed controlled and uncontrolled studies through the PubMed and SPORTDiscus databases. The studied participants were athletes from regional to elite level, the exposure of interest was natural AT, and the main outcome of interest was performance.

Full text

174 175
sport science curren t an d fu ture trends for performan ce op timization
SL-resident athletes who travel to and subsequently live and train at
moderate altitude. Despite being used by many elite swimmers and
coaches, there is no clear evidence that training at natural altitude en-
hances performance more than training at SL (6). In a milestone study
published in the mid-1990s, Levine & Stray-Gundersen provided evi-
dence that the “live high-train low” (Hi-Lo) strategy can improve 3000- to
5000-m running performance in collegiate/club runners (9). Later, this
approach was modified to limit the low-altitude training sessions to
only high-intensity workouts and was subsequently termed “live high-
train high and low” (Hi-HiLo). The improvement in running performance
was associated with increase in red cell mass, the subsequent increase
in maximal oxygen uptake (V O2max), the “high altitude eect”, and the
maintenance of high-intensity training velocities and oxygen flux to the
muscles “the low altitude eect” (18). This paradigm has been sustained
by later investigation in elite endurance athletes performing dierent
sports including running (19), orienteering (20), and cycling (21). How-
ever, these studies are diicult to compare with each other directly given
the many dierences in experimental design (12).
The aims of this systematic review are: 1) to collate and to critically
evaluate the empirical evidence sustaining the use of natural AT in ath-
letes with the main goal of improving SL performance; and 2) to derive
which of the dierent natural AT strategies is more eicient for enhanc-
ing SL performance when the athletes come back to SL training and
competition. To achieve these goals, we systematically reviewed con-
trolled and uncontrolled studies through the PubMed and SPORTDiscus
databases. The studied participants were athletes from regional to elite
level, the exposure of interest was natural AT, and the main outcome of
interest was performance.
Methods
Literature search
This systematic review followed the PRISMA statement guidelines
(22). To achieve this, a systematic literature search was conducted for
studies in any language indexed in the PubMed and SPORTDiscus da-
tabases (up to March 2017). This search was performed using the fol-
lowing selected keywords: ALL FIELDS, altitude training AND sport AND
performance, NOT simulated OR artificial OR normobaric, NOT review.
To manage the bibliographic references the EndNote (ver. X7) soware
was used.
Eligibility criteria and study selection
Altitude training for sea level performance: a
systematic review
Ferran A. Rodríguez and Sergi Àvila
Introduction
Altitude training (AT) at natural environment has been a matter of
extensive research for half a century and, despite some sceptical views
(1-3), it continues to play an important role in the preparation of elite
and sub-elite athletes in many countries (4, 5). Paradoxically, there is a
remarkable lack of controlled and adequately powered studies on natu-
ral AT in the scientific literature, particularly in elite athletes, and there is
no clear evidence that AT enhances performance more than training at
sea level (SL) (1, 5-7). The theoretical concept behind this practice is the
independent and combined eects of the physiological processes of ac-
climatization to chronic hypoxia and those derived from training under
the additional stress imposed by exercising in a hypoxic environment
(8). In accordance with some investigations, altitude acclimatization re-
sults in central and peripheral adaptations, i.e. augmented red cell vol-
ume, haemoglobin (Hb) mass and maximal oxygen uptake (V O2max) that
improve primarily systemic oxygen delivery (“erythropoietic paradigm”)
(9, 10), while others argue against this view and support the concept of
“nonhematological” adaptations such as improved muscle eiciency,
greater muscle buering and the ability to tolerate lactic acid produc-
tion (11, 12).
Conversely, the combination of intense training and hypoxia may
have a negative impact on athlete’s performance capacity and health
status, causing unfavourable eects such as acute mountain sick-
ness (13), immune suppression (14), iron-deficient erythropoiesis (15),
cathecolamine mediated glycogen depletion (16) and increased oxida-
tive stress and tissue damage (17), among others. Interestingly, a recent
meta-analysis concluded that AT performance gains could be related to
a placebo or nocebo eect (7).
There are dierent strategies to train at altitude. The classical ap-
proach (“live high-train high”, Hi-Hi), used since the late 1960s, involves
. INEFC-Barcelona Sport Sciences Research Group, Institut Nacional d’Educació
Física de Catalunya (INEFC), Universitat de Barcelona, Spain

176 177
sport science curren t an d fu ture trends for performan ce op timization
the intervention, living and training characteristics and eects on perfor-
mance measures. Performance data extraction was done from time tri-
als, race results, power output and total work capacity (cycle ergometry)
and time scores. Weight-adjusted V O2max values and selected haemato-
logical data (Hb mass, Hb blood concentration and haematocrit ratio)
were also extracted.
Results
Study characteristics
A total of 20 studies, 7 controlled and 13 uncontrolled, were identi-
fied for inclusion in the systematic review. These studies involved 439
athletes, from which 173 took part in the controlled studies and 266 in
the uncontrolled studies. 247 out of the total 439 participants were elite
athletes (56%), and the rest were from regional/interregional to subelite
level athletes (44%), a fact that can bring about large dierences in per-
formance level and range for improvement, making diicult to compare
investigation results. The intervention methodologies and type of de-
signed utilized were as follows: 14 Hi-Hi (5 controlled), 4 Hi-Lo (3 con-
trolled), and 3 Hi-HiLo (1 controlled
Classical altitude training (Hi-Hi) controlled studies
Only 5 controlled Hi-Hi studies were identified (table 1). Using a
cross-over design, Adams et al. reported no potentiating eect of hard
endurance training at 2300 m over equivalent SL training on 2-mile per-
formance time or V O2max in well-conditioned middle-distance runners
(23) The study by McLean et al. consisted on a Hi-Hi training camp involv-
ing a group of 21 Australian football players and 9 Lo-Lo controls. The
Hi-Hi group likely improved 2000-m time trial performance by 1.5% aer
altitude with very large individual variability (90%CI: -3.3–6.3%) and low
individual responsiveness (0.8%). This change was paralleled by a very
likely increase in Hb mass (2.8%) (24). Levine & Stray-Gundersen failed
to find any eect on performance aer 4 weeks of Hi-Hi intervention de-
spite an increase of V O2max (3.4%) and red cell mass (10%) (9). Burtscher
et al. studied two groups of amateur runners and found no group dif-
ferences in cycling total work capacity between the Hi-Hi group and the
Lo-Lo controls: 3 and 16 days aer the intervention the Lo-Lo group im-
proved 8% and 17% whereas the Hi-Hi group improved 0.3% and 8%,
respectively (25).
Only Rodríguez et al. showed a significant improvement in swimming
performance aer living and training at 2320 m during 3 or 4 weeks (e.g.,
3.1% and 3.4% in specific 100-m or 200-m time trial), but this change was
In order to be considered eligible for inclusion, studies had to meet
the following criteria: 1) participants were healthy adult competitive ath-
letes; 2) studies were controlled and uncontrolled; 3) altitude exposure
was natural (classic or terrestrial, not artificial or simulated); 4) primary
focus was SL performance (articles with no performance measures were
excluded); 5) original studies were used only (no reviews); and 6) studies
published in any language.
The flow chart of literature screening approach and study identifica-
tion is displayed in figure 1. From the 325 articles initially identified, we
removed 51 duplicated articles, 224 were excluded from title and ab-
stract information, and 30 articles were excluded aer reading the full
text for not meeting the eligibility criteria or due to impossibility to ac-
cess them. Finally, 20 studies were included in the qualitative review.
Figure 1. Flow chart of literature screening approach and study identification.
Data extraction
We developed a data extraction table classifying the type/level of
athletes, altitude strategy, sample size, study design, follow up during

178 179
sport science curren t an d fu ture trends for performan ce op timization
Noteworthy, in the study by Roels et al. the modest increase in 2000-m
trial (1.9%) was only significant when the swimmers lived and trained
at 1200 m of altitude, but not at 1850 m (29). Moreover, another study
showed a decrease in performance aer 21 to 27 days of Hi-Hi interven-
tion in four dierent groups (n = 97) of elite swimmers (31). The other 6
studies did not show significant changes in performance despite some
modest changes in V O2max (32, 33), Hb mass (31, 34), blood red cell mark-
ers (30, 35) or without haematological changes (36).
Hi-Lo and Hi-HiLo uncontrolled studies
The systematic review included 3 uncontrolled studies examining the
Hi-Lo strategy and 3 studies using the Hi-HiLo strategy (table 4). These
studies were conducted by mostly the same group of researchers and
used similar designs and methodologies. In a retrospective study using
the Hi-Lo approach, 8 out of 12 collegiate runners were classified as re-
sponders and improved their time in a 5000-m time trial by 3.6% whereas
the non-responder decreased their performance time by 1.3%. The bet-
ter times of the responders was paralleled by a non-significant increase
in V O2max (7.5%) (37). Coincident results were obtained by these authors
in a prospective study with elite distance runners (37). Also Stray-Gun-
dersen et al. reported modest but significant performance gains in 22
elite runners on a 3000-m time trial (1.1%) associated to a 3% increase in
V O 2max (18). To identify the optimal altitude for training using the Hi-HiLo
approach, Chapman et al. compared four groups of collegiate runners
living at four altitudes (from 1780 to 2800 m) and training at varying al-
titudes from 1250 to 3000 m. They found that only the middle altitudes
(i.e., 2084 and 2454 m) evoked significant gains in 3000-m time trial run-
ning performance (2.1 to 2.8%), associated to a 3% to 8% increase in
V O 2max (38). Similarly, Saugy et al. conducted a study with 13 well-trained
triathletes who lived at 220 m and trained at 1100-1200 m of altitude and
found an improvement in 3000-m time trial running performance (3.3%)
aer 3 weeks upon return to sea level, with no changes 1 and 7 days aer
the altitude training camp. V O2max also increased by 5.2% (39).
not significantly dierent from that experienced by the Lo-Lo control
group (3.7%) (26). Interestingly, two studies reporting increases in V O2max
did not find a concomitant improvement in performance compared with
SL controls. In summary, only one study actually provided evidence of
superior improvement of Hi-Hi altitude training compared with SL (24).
However, the magnitude of these changes seems lower than can be ex-
pected because of a SL training camp, and placebo or nocebo eects
cannot be ruled out.
Hi-Lo and Hi-HiLo controlled studies
Four controlled studies were identified (table 2). Levine & Stray-Gun-
dersen were the first to use the Hi-Lo strategy in their classical study cited
above (9), in which they assigned collegiate and club amateur runners to
Hi-Hi, Hi-Lo, and Lo-Lo (control) groups. They reported an improvement
on 5000-m running performance (1.3%) in the Hi-Lo runners three weeks
aer the training camp and attributed this to increased V O2max (5%) and
red cell mass (10%), according to the “erythropoietic paradigm”. In a
study by Dehnert et al. two groups of subelite triathletes followed a Lo-
Lo or Hi-Lo intervention for 2 weeks and found no eects on cycling or
treadmill running performance despite a 7% increase in V O2max and un-
changed Hb mass (27). In contrast, Wehrlin et al. (20) studied a group
of 10 elite orienteers using a Hi-Lo strategy for 24 days and comparing
them with 7 Lo-Lo cross country skiers, and reported an improvement
in 5000-m running performance (1.6%) in the Hi-Lo group, paralleled by
increased V O2max (4.1%) and red cell mass (5.3%), which are comparable
to previous results in runners (9). Finally, Rodríguez et al. conducted the
only controlled study using the Hi-HiLo strategy in which athletes live at
altitude and train at the same and a lower altitude (26). Four groups of
international level swimmers were compared: Hi-Hi for 3 and 4 weeks
(previously cited), and Hi-HiLo and Lo-Lo controls for 4 weeks. Although
all groups improved aer a well-controlled training camp, the Hi-HiLo
group of swimmers further improved 50-, 100- (sprinters) or 200- (non-
sprinters), and 400-m swimming performance (5.5%, 6.3% and 4.7%, re-
spectively) 2 to 4 weeks aer the training camp. However, this substan-
tial improvement in performance could not be attributed to changes in
V O 2max, Hb mass or swimming economy and, therefore, to the “erythro-
poietic paradigm”.
Classical altitude training (Hi-Hi) uncontrolled studies
Nine uncontrolled studies using the classical Hi-Hi strategy met the
inclusion criteria (table 3). Overall, only 2 studies showed some evidence
of beneficial eects on performance (28, 29) and 1 was uncertain (30).

180 181
sport science curren t an d fu ture trends for performan ce op timization
Study Subjects
(level) StrategyaSample
sizebDesigncFollow
updTraining
Eect s on performance
measurese
(D% from pre-values)
Other
changes
Altitude
(m)
Duration
(d)
Phase V O max Hbmass Otherf
Adams et
al. 
()
Runners
(middle-
distance)
Lo-Lo vs.
Hi-Hi
M CON
X-over
D- 



?
TT 2mi: 1.3% (n.s. vs. SL) ¯2.8%
? ?
Burt scher
et al. 
()
Runners
(amateur)
Lo-Lo vs.
Hi-Hi
M
M
CON
R
Pre-post
D, 

 ?CY TWC: 8% at D3, 17%
at D16
¯0.3% at D3,
8% at D16
(n.s. bet ween
groups)
10% at
D16
? ?
Levine
& Stray-
Gundersen

()
Runners
(collegiate/
club level)
Lo-Lo vs.
Hi-Hi
M, F
M, F
CON
R
Pre-post
D,,


-



? «
«
«
3 .4%*
RCM «
10%*
McLean et
al. 
()
Australian
footballers
(elit e)
Lo-Lo vs.
Hi-Hi
M
M
CON
Pre-post
NR
D, 
~


Pre-
Season
TT 2km: Post1: 1.5%*
(±4.8% 90%CL)
Post2: triv ial
changes vs. Post1
?Hi-Hi:
Post
. %*
Post
«
?
Rodríguez
et al. 
()
Swimmers
(elite)
Lo-Lo vs.
Hi-Hi
Hi-Hi
F, M
F, M
 F,  M
CON
NR
Pre-post
D,,,,  or






In-
season
TT 50, 100/200, 400m:
3.2%, 3.7%, 1.6%*
3.4%, 3.1%, 0.6%*
3.7% , 3.4%, 3.3%*
«
«
«
«
3 .8 %*
.%*
Economy«
«
«
Table 1. Summary of CONTROLLED studies on natural altitude training using
the Hi-Hi strategy for sea level performance enhancement.
a Lo-Lo, live low–train low; Hi-Hi, live high–train high.
b F, females; M, males.
c CON, controlled trial; UN, uncontrolled trial; R, randomised; NR,
non-randomised; X-over, cross-over
d D#, testing post-intervention (day number).
e TT, time trial in specific sport; CY, cycling test; TWC, total work capacity; TM,
treadmill test.
f RCM, red cell mass; RCV, red cell volume.
↑, improvement/increase/benefit; <—>, no change; ↓, worsening/decrease/
harm; =, same as ab
*, significantly dierent from values measured before training or compared
to sea-level controls (p<0.05); n.s., non-significant dierence (p≥0.05); ?, un-
certain/not reported.

182 183
sport science curren t an d fu ture trends for performan ce op timization
Study Subjects
(level) StrategyaSample
sizebDesigncFollow
updTraining
Eect s on
performance
measurese
(D% from pre-values)
Other
changes
Altitude
(m)
Duration (d) Phase V O max Hbmass Otherf
Levine
& Stray-
Gundersen

()
Runners
(collegiate/
club level)
Lo-Lo vs.
Hi-Lo
M, F
M, F
CON
R
Pre-post
D,, 
L , T 


? «
TT 5km 1. 3%*
«
%*
RCM «
.%*
Dehnert et
al. 
()
Tri at hle te s
(subelite)
Lo-Lo vs.
Hi-Lo
M, F
M, F
CON
R
Pre-post
D, 
L , T 
 ? «
« Incremental CY ramp,
« TM tests, n.s. trend to
improve running time
«
%
«
«
?
?
Wehrlin

()
Orienteers
and cross
country
skiers
(elite)
Lo-Lo vs.
Hi-Lo
M, F
M, F
CON
NR
Pre-Post
D -
L , T -


Pre-
season
No measures in Lo-Lo
group
TT 5km 1.6%
?
.%
«
.%
Rodríguez
et al. 
()
Swimmers
(elite)
Lo-Lo vs.
Hi-HiLo
F, M
F, M
CON
NR
Pre-post
D,,,
,
 or 
L/T , T 


In-
season
TT 50, 100/200, 400m:
3 .2 %*, 3.7%*, 1. 6%*
. %*, . %*, . %*
(Hi-
HiLo > Lo-Lo*)
«
«
«
1. 3%
« Economy
« Economy
Table 2. Summary of CONTROLLED studies on natural altitude training using
the Hi-Lo or Hi-HiLo strategies for sea level performance enhancement.
a Lo-Lo, living low, training low; Hi-Lo, live high–train low; Hi-HiLo,
live high–train high and low; L, living; T, training.
b F, females; M, males. -over, crossover; R, randomised; NR, non-randomised.
d D#, testing post-intervention (day number).
e TT, time trial in specific sport; CY, cycling test; TM, treadmill test.
f RCM, red cell mass; RCV, red cell volume.
, improvement/increase/benefit; «, no change; ¯, worsening/decrease/harm;
=, same as above.
*, significantly dierent from values measured before training or compared
to sea-level controls (p<0.05); n.s., non-significant dierence (p≥0.05); ?,
uncertain/not reported.

184 185
sport science curren t an d fu ture trends for performan ce op timization
Study Subjects
(level) Strategya
Sample
sizeb
e s
DesigncFollow updTraining
Eect s on
performance
measures
(D% from pre -
values)e
Other changes
Alti tude (m) Duration (d) Phase V O max Hbmass Otherf
Faulkner et
al. 
()
Swimmers,
fit men
Hi-Hi M
M
UN
Pre-post
D  

?«Tethered
swimming
, ,  yd
« « Hb %*
Htc %*
Faulkner et
al. 
()
Runners
(subelite)
Hi-Hi M
M
M
UN
Pre-post
D to D 
-
-



? ? TT mi .%
? TT mi -.%
? TT mi . %
% % ?
Mizuno et al.

()
X-country skiers
(subelite)
Hi-Hi M UN
Pre-post
?



? TM running time
to exhaustion
%*
« ?  O deficit
% buer
capacity
Roels et al .
 ()
Swimmers
(elite)
Hi-Hi
Hi-Hi
M UN
X-over
D,,, 



? TT km: .%*
«
«
«
«
«
RBC: «
«
Schmitt e t
al.  ()
X-countr y skiers,
swimmers,
runners (elite)
Hi-Hi M UN
Pre-post
D,   Balanced training
load
«CY PPO .% .%* economy
%*
Hue et al.

()
Swimmers,
(regional/
interregional)
Hi-Hi F,  M UN
Pre-post
D,  Competitive period « ? ? ?
Siewierski
et al. 
()
Swimmers (elite) Hi-Hi M, F UN
Pre-post
Race
results
pre-post
  Competitive
period
? .% pts ? ? RBC: .%
Hb: .%
Htc: .%
Gough et al.
 ()
Swimmers
(elite)
Hi-Hi M, F UN
Pre-post
D,, ,  -  ? « ? %*
Wachsmuth
et al. 
()
Swimmers (elite) Hi-Hi
Hi-Hi
Hi-Hi
Hi-Hi
M, F
M, F
M, F
M, F
UN
Pre-post
All racing
results








Two years
preparation for
Olympic Games
D- ¯ pts
D- ¯ pts
D- ¯ pt s
? .%
. %
.%
.%
%M,
,%F
?
Table 3. Summary of UNCONTROLLED studies on natural altitude training us
a Lo-Lo, living low, training low; Hi-Hi, live high–train high; L, living; T, training.
b F, females; M, males.
c CON, controlled trial (vs. sea level); UN, uncontrolled trial (vs. sea level);
X-over, crossover; R, randomised; NR, non-randomised.
d D#, testing post-intervention (day number).
e TT, time trial in specific sport; CY, cycling test; PPO, peak power output; TM,
treadmill test; Re, responders; NRe, non-responders; pts, FINA score points.
f RCM, red cell mass; RCV, red cell volume.
, improvement/increase/benefit; «, no change; ¯, worsening/decrease/harm;
=, same as above.
*, significantly dierent from values measured before training or compared
to sea-level controls (p<0.05); n.s., non-significant dierence (p≥0.05); ?, un-
certain/not reported.

186 187
sport science curren t an d fu ture trends for performan ce op timization
Study Subjects
(level) StrategyaSample
sizebDesigncFollow
updTraining
Eect s on performance
measures
(D% from pre-values)e
Other
changes
Alti tude (m) Duration
(d)
Phase VOmax Hbmass Other
Chapman et al.

(retrospective)
()
Runners
(collegiate)
Hi-Lo F, M UN
Pre-
post
D L 
T -
 ? R (n=) TT km .*%
NR (n=) ?¯ TT km -.%
R . %*
NR <->
? Hb: R .*
NR .*
Htc: R .*
NR . *
Chapman et al.

(prospective)
()
Distance
runners
(elite)
Hi-HiLo M, M UN
Pre-
post
D L 
T -
 ? R (n=) TT km -.s*
NR (n=) « TT km
R 3.4 ml /
kg·mi n
NR <->
(n. s .)
Stray-
Gundersen et
al. 
()
Runners
(elite)
Hi-HiLo  F,  M UN
Pre-
post
D L 
T 
 Near
season’s
fitnes s peak
TT 3 km: 1.1%* 3% 1 g/dl
Chapman et al.

()
Distance
runners
(collegiate)
Hi-HiLo
Hi-HiLo
Hi-HiLo
Hi-HiLo
F, M
F, M
F, M
F, M
UN
Pre-
post
D, L 
L 
L 
L 
(T -)
 ? TT  km: «
. %*D,
.%*D
.%* D,
.%*D
<->
2% D1,
4%* D14
3%* D1,
5%* D14
%* D,
%* D
%* D,
%* D
?
?
?
?
RCM: .%
.%
.%
.%
Saugy et al.

()
Tri at hle te s
(well
trained)
Hi-Lo M UN
Pre-
post
D,,  L 
(T -)
 Competitive
season
TT  km: « at D, D
.%* at D
.% 1.8% PPO: .%
Table 4. Summary of UNCONTROLLED studies on natural altitude training us-
ing the Hi-Lo or Hi -HiLo strategies for sea level performance enhancement.
a Lo-Lo, living low, training low; Hi-Hi, live high–train high; L, living; T, training.
b F, females; M, males.
c CON, controlled trial (vs. sea level); UN, uncontrolled trial (vs. sea level); X-
over, crossover; R, randomised; NR, non-randomised.
d D#, testing post-intervention (day number).
e TT, time trial in specific sport; CY, cycling test; PPO, peak power output;
TM, treadmill test; Re, responders; NR, non-responders; FINA pts, FINA score
points.
f RCM, red cell mass; RCV, red cell volume.
, improvement/increase/benefit; «, no change; ¯, worsening/decrease/harm;
=, same as above.
*, significantly dierent from values measured before training or compared
to sea-level controls (p<0.05); n.s., non-significant dierence (p≥0.05); ?, un-
certain/not reported.

188 189
sport science curren t an d fu ture trends for performan ce op timization
design (26). This investigation showed that SL swimming performance
of elite swimmers in 100- (sprinters) or 200-m (non-sprinters) time trials
was not altered, or in some cases impaired immediately, but improved
significantly by ~3.1–3.7% aer 1 to 4 weeks of recovery following com-
pletion of a coach-prescribed training camp, whether it was conducted
at SL or at moderate altitude (2320 m). By including 2 weekly sessions
of high-intensity training at lower altitude (Hi-HiLo strategy) a greater
improvement in performance occurred 2 and 4 weeks aer the train-
ing camp (5.3% and 6.3%, respectively). Similarly, further gains in 400-
and 50-m freestyle time trial performance was noted 2 weeks (4.2% and
5.2%, respectively) and 4 weeks (4.7% and 5.5%, respectively) following
return to SL. In addition, this study shows that the delayed performance
improvements are not linked to changes in V O2max, oxygen kinetics or
Hb mass and hence cannot be attributed exclusively to an enhanced sys-
temic oxygen transport capacity.
Globally, these results are in line with estimations published in a meta-
analytic review by Bonetti & Hopkins, who concluded that “performance
changes in studies using the conventional Hi-Hi approach were unclear,
whereas changes using the terrestrial Hi-Lo strategy were considered
likely to be eective both for elite and subelite athletes (~4%), or a more
realistic 1.5% when performance was predicted from uncontrolled stud-
ies” (7). These estimations are in line with a recent review by Saunders
et al. (5) in which, by using a regression analysis of average performance
changes, it was estimated that a 3-week terrestrial AT camp would elicit
mean performance improvements of ~1.8% (Hi-Hi) and ~2.5% (Hi-Lo) (5).
Levine and Stray-Gundersen were the first to test the hypothesis
that acclimatization to moderate altitude (2500 m) plus training at low
altitude (1250 m) (Hi-Lo paradigm) improves SL performance in well-
trained runners more than in equivalent SL or Hi-Hi controls (9). They
also concluded that the correlation between the increase in V O2max and
the improvement in 5000-m time aer the field training camp argues
strongly that this is a key adaptation during altitude training and a nec-
essary mechanism for improving SL performance. Despite this, the per-
formance gain was only 1.3% from pre-training values. Notwithstanding,
this study changed the previous AT paradigm that was only focussed on
the classical Hi-Hi strategy.
Concerning the 9 uncontrolled Hi-Hi studies included in this review,
we can see a very similar picture as for the controlled studies, since only
two studies showed some beneficial eects on performance (28, 29), one
was uncertain (30), and 1 even showed impaired performance aer in a
Discussion
This systematic review, which aimed at assessing the empirical evi-
dence sustaining the use of AT in athletes with focus on SL performance
enhancement, does not appear suiciently robust to determine the ef-
ficacy and appropriate characteristics (duration, altitude and training
requirements) of an AT camp. Neither it can conclude which of the natu-
ral AT strategies is best for enhancing performance at SL. The reviewed
studies are diicult to compare with each other directly, given the many
dierences in experimental design, type of participants, outcome meas-
ures and methodology. Notwithstanding, there seems to be a certain
consensus—perhaps lacking compelling evidence to support it—that
when athletes are exposed to a high enough altitude, for a long enough
amount of time, and are able to preserve fitness by training hard under
both hypoxic and normoxic conditions, the majority may improve physi-
cal performance (40).
We have reviewed publications from 1967 to 2017, a 50-year period.
A total of 20 studies have been appraised, but only 7 published articles
(33%) had a controlled design. The remaining 14 studies (67%) were un-
controlled and provide low quality evidence since performance changes
can be attributed to training alone, training camp eect or placebo/
nocebo eect. Concerning the 5 controlled Hi-Hi studies, only the one
published by MacLean et al. (24) provided limited evidence of superior
improvement compared with SL controls, although the high interindi-
vidual variability (-3.3–6.3%) argues against a real AT eect and placebo/
nocebo eects cannot be ruled out.
In the last decade the Hi-Lo approach has gained interest over the
classical Hi-Hi strategy in the scientific literature and among many en-
durance athletes (4). In our review, the results of the 4 controlled Hi-Lo or
Hi-HiLo studies seem somewhat more convincing compared with those
using the classical Hi-Hi approach. In their milestone study, Levine &
Stray-Gundersen showed that the Hi-Lo strategy evoked an increase in
5000-m time trial performance (1.3%) in collegiate and club runners (9),
despite some researchers argue that this modest improvement could be
attributed to a placebo or nocebo eect, as another Hi-Hi group showed
the same improvement in V O2max and red cell mass without any change
in performance (1). Similar limitations can be attributed to the study by
Wehrlin et al. since the improved 5000-m running performance (1.6%)
lacked of concomitant performance measures in the control group (20).
A recent investigation was the first to show substantial performance im-
provements aer a terrestrial Hi-HiLo intervention using a controlled

190 191
sport science curren t an d fu ture trends for performan ce op timization
Biol. ;():-.
Rodríguez FA. Training at real and simulated altitude in swimming: too high ex-
pectations? In: Kjendlie P-L, Stallman RK, Cabri J, editors. Biomechanics
and Medicine in Swimming XI. Oslo: Norwegian School of Sport Science;
. p. -.
Armstrong LE, Johnson EC, Ganio MS, Judelson DA, Vingren JL, Kupchak BR,
et al. Eective body water and body mass changes during summer ultra-
endurance road cycling. J Sports Sci. ;():-.
Levine BD, Stray-Gundersen J. Dose-response of altitude training: how much
altitude is enough? Adv Exp Med Biol. ;:-. Epub //.
Levine BD, Stray-Gundersen J. “Living high-training low”: eect of moderate-
altitude acclimatization with low-altitude training on performance. J Appl
Physiol. ;():-.
Levine BD, Stray-Gundersen J. Point: positive eects of intermittent hypoxia
(live high:train low) on exercise performance are mediated primarily by
augmented red cell volume. J Appl Physiol. ;():-.
Gore CJ, Hopkins WG. Counterpoint: positive eects of intermittent hypoxia
(live high:train low) on exercise performance are not mediated primarily
by augmented red cell volume. J Appl Physiol. ;():-; discus-
sion -.
Gore CJ, Clark SA, Saunders PU. Nonhematological mechanisms of improved
sea-level performance aer hypoxic exposure. Med Sci Sports Exerc.
;():-.
Fukuda T, Maegawa T, Matsumoto A, Komatsu Y, Nakajima T, Nagai R, et al. Ef-
fects of acute hypoxia at moderate altitude on stroke volume and cardiac
output during exercise. Int Heart J. ;():-.
Mazzeo RS. Altitude, exercise and immune function. Exerc Immunol Rev.
;:-.
Stray-Gundersen J, Alexander C, Hochstein A, Lemos D, Levine BD. Failure of
red cell volume to increas e with altitude exposure in iron deficient runne rs.
Med Sci Spor t Exerc. ;():S.
Young AJ. Energy substrate utilization during exercise in extreme environ-
ments. Exerc Sport Sci Rev. ;:-.
McGinnis G, Kliszczewiscz B, Barberio M, Ballmann C, Peters B, Slivka D, et al.
Acute hypoxia and e xercise-induced bloo d oxidative stress. Int J Spo rt Nutr
Exerc Metab. ;():-.
Stray-Gundersen J, Chapman RF, Levine BD. “Living high-training low” altitude
training improves sea level performance in male and female elite runners.
J Appl Physiol (). ;():-.
Stray-Gundersen J, Levine BD. Live high, train low at natural altitude. Scand J
Med Sci Spor ts. ; Suppl :-.
Wehrlin JP, Zuest P, Hallen J, Marti B. Live high-train low for  days increases
hemoglobin mass and red cell volume in elite endurance athletes. J Appl
Physiol. ;():-.
Jones AM, Haramizu S, Ranchordas M, Burke L, Stear S, Castell LM. A-Z of nu-
tritional supplements: dietary supplements, sports nutrition foods and
ergogenic aids for health and performance--Part . Br J Sports Med.
;():-.
very large number of elite swimmers (31).
Among the uncontrolled Hi-Lo or Hi-HiLo investigations, two (37)
should be analysed with caution as the participants, collegiate or elite
runners, were categorized post-facto into ‘responders’ and ‘non-re-
sponders’, as they aimed at investigating the individual variation in re-
sponse to AT. The other 2 studies showed performance changes similar
to the controlled groups, although the risk of placebo or nocebo eects
is exacerbated. Interestingly, moderate altitudes of ~2100-2500 m were
identified as optimal for benefits in 3000-m running performance (41).
Conclusions
There are several limitations in every study design using terrestrial
AT, including the impossibility of blinding the intervention, limitations in
recruiting large numbers of participants, diiculties in group randomisa-
tion, control of placebo and nocebo eects, large variability in the re-
sponse, etc. These barriers make diicult the comparison of the exist-
ing studies and the design of new investigations that can meet the high
standards of scientific research.
Contrary to common expectations, the systematic review of 20 arti-
cles published along 50 years (1967 to 2017) shows that the quality of
the empirical evidence about using natural altitude training in competi-
tive athletes with the main goal of improving sea level performance is
far from being compelling. However, the available evidence supports the
concept that the Hi-Lo and Hi-HiLo strategies oer the best potential for
performance benefits, as at least two controlled studies provided sound
evidence of positive eects on performance in collegiate/club runners
and elite swimmers, respectively. Uncontrolled studies also support this
concept despite the lower quality of the evidence.
References
Lundby C, Millet GP, Calbet JA, Bartsch P, Subudhi AW. Does ‘altitude train-
ing’ increase exercise performance in elite athletes? Br J Sports Med.
;():-.
Jacobs RA. Con: Live high-train low does not improve sea-level performance
beyond that achieved with the equivalent living and training at sea level.
High Alt Med Biol. ;():-.
Hahn AG, Gore CJ, Martin DT, Ashenden MJ, Roberts AD, Logan PA. An evalua-
tion of the concept of living at moderate altitude and training at sea level.
Comp Biochem Physiol A Mol Integr Physiol. ;():-.
Wilber RL. Application of altitude/hypoxic training by elite athletes. Med Sci
Sports Exerc. ;():-.
Saunders PU, Pyne DB, Gore CJ. Endurance training at altitude. High Alt Med

192 193
sport science curren t an d fu ture trends for performan ce op timization
fining the “dose” of altitude training: how high to live for optimal sea level
performance enhancement. J Appl Physiol (). .
Saugy JJ, Schmitt L, Cejuela R, Faiss R, Hauser A, Wehrlin JP, et al. Comparison
of “live high-train low” in normobaric versus hypobaric hypoxia. PLoS One.
;():e.
Wilber RL. In troduction to altitude/ hypoxic training symposium . Med Sci Sport s
Exerc. ;():-.
Rehrer NJ, Brouns F, Beckers EJ, Frey WO, Villiger B, Riddoch CJ, et al. Physi-
ological changes and gastro-intestinal symptoms as a result of ultra-en-
durance running. Eur J Appl Physiol Occup Physiol. ;():-.
Liberati A, Altman DG, Tetzla J, Mulrow C, Gotzsche PC, Ioannidis JP, et al. The
PRISMA statement for reporting systematic reviews and meta-analyses of
studies that evaluate health care interventions: explanation and elabora-
tion. PLoS medicine. ;():e.
Adams WC, Bernauer EM, Dill DB, Bomar JB, Jr. Eects of equivalent sea-level
and altitude training on VOmax and running performance. J Appl Physiol.
;():-.
McLean BD, Gore CJ, Kemp J. Application of ‘Live Low-Tr ain High’ for Enhancing
Normoxic Exercise Performance in Team Sport Athletes. Sports Med. .
Epub //.
Burtscher M, Nachbauer W, Baumgartl P, Philadelphy M. Benefits of training at
moderate altitude versus sea level training in amateur runners. Eur J Appl
Physiol. ():-.
Rodriguez FA, Iglesias X, Feriche B, Calderon-Soto C, Chaverri D, Wachsmuth
NB, et al. Altitude training in elite swimmers for sea level performance (Al-
titude Project). Med Sci Spor ts Exerc. ;():-.
Dehnert C, Hutler M, Liu Y, Menold E, Netzer C, Schick R, et al. Erythropoiesis
and performance aer two weeks of living high and training low in well
trained triathletes. Int J Sports Med. ;():-.
Mizuno M, Juel C, Bro-Rasmussen T, Mygind E, Schibye B, Rasmussen B, et al.
Limb skeletal mus cle adaptation in athletes aer tra ining at altitude. J Appl
Physiol (). ;():-.
Roels B, Hellard P, Schmitt L, Robach P, Richalet JP, Millet GP. Is it more eec-
tive for highly trained swimmers to live and train at  m than at 
m in terms of performance and haematological benefits? Br J Sports Med.
;():e.
Siewierski M, Słomiński P, Białecki R, Adamczyk J. Athletic performance of
swimmers aer altitude training (, m above sea level) in view of their
blood morphology changes. Biol Sport. ;().
Wachsmuth NB, Volzke C, Prommer N, Schmidt-Trucksass A, Frese F, Spahl O,
et al. The eects of classic altitude training on hemoglobin mass in swim-
mers. Eur J Appl Physiol. ;():-.
Faulkner JA, Kollias J, Favour CB, Buskirk ER, Balke B. Maximum aerobic capac-
ity and running performance at altitude. J Appl Physiol. ;():-.
Schmitt L, Millet G, Robach P, Nicolet G, Brugniaux JV, Fouillot JP, et al. Influ-
ence of “living high-training low” on aerobic performance and economy of
work in elite athletes. Eur J Appl Physiol. ;():-.
Gough CE, Saunders PU, Fowlie J, Savage B, Pyne DB, Anson JM, et al. Influence
of altitude training modality on performance and total haemoglobin mass
in elite swimmers. Eur J Appl Physiol. ;():-. Epub //.
Faulkner JA, Daniels JT, Balke B. Eects of training at moderate altitude on
physical performance capacity. J Appl Physiol. ;():-.
Hue O, Antoine-Jonville S , Sara F. Th e eect of  days of training in tropic al envi-
ronment on per formance in neutral climate in swimmers. Int J Spor ts Med.
;():-.
Chapman RF, Stray-Gundersen J, Levine BD. Individual variation in response to
altitude training. J Appl Physiol. ;():-.
Chapman RF, Karlsen T, Resaland GK, Ge RL, Harber MP, Witkowski S, et al. De-