Factors affecting running economy in trained distance runners.

Abstract

Running economy (RE) is typically defined as the energy demand for a given velocity of submaximal running, and is determined by measuring the steady-state consumption of oxygen (VO2) and the respiratory exchange ratio. Taking body mass (BM) into consideration, runners with good RE use less energy and therefore less oxygen than runners with poor RE at the same velocity. There is a strong association between RE and distance running performance, with RE being a better predictor of performance than maximal oxygen uptake (VO2max) in elite runners who have a similar VO2max). RE is traditionally measured by running on a treadmill in standard laboratory conditions, and, although this is not the same as overground running, it gives a good indication of how economical a runner is and how RE changes over time. In order to determine whether changes in RE are real or not, careful standardisation of footwear, time of test and nutritional status are required to limit typical error of measurement. Under controlled conditions, RE is a stable test capable of detecting relatively small changes elicited by training or other interventions. When tracking RE between or within groups it is important to account for BM. As VO2 during submaximal exercise does not, in general, increase linearly with BM, reporting RE with respect to the 0.75 power of BM has been recommended. A number of physiological and biomechanical factors appear to influence RE in highly trained or elite runners. These include metabolic adaptations within the muscle such as increased mitochondria and oxidative enzymes, the ability of the muscles to store and release elastic energy by increasing the stiffness of the muscles, and more efficient mechanics leading to less energy wasted on braking forces and excessive vertical oscillation. Interventions to improve RE are constantly sought after by athletes, coaches and sport scientists. Two interventions that have received recent widespread attention are strength training and altitude training. Strength training allows the muscles to utilise more elastic energy and reduce the amount of energy wasted in braking forces. Altitude exposure enhances discrete metabolic aspects of skeletal muscle, which facilitate more efficient use of oxygen. The importance of RE to successful distance running is well established, and future research should focus on identifying methods to improve RE. Interventions that are easily incorporated into an athlete's training are desirable.

Full text

Page 1

Sports Med 2004; 34 (7): 465-485
0112-1642/04/0007-0465/$31.00/0
REVIEW ARTICLE
 2004 Adis Data Information BV. All rights reserved.
Factors Affecting Running Economy
in Trained Distance Runners
Philo U. Saunders,1,2 David B. Pyne,1 Richard D. Telford3 and John A. Hawley2
1
2
Department of Physiology, Australian Institute of Sport, Belconnen, ACT, Australia
Exercise Metabolism Group, Faculty of Medical Sciences, RMIT University, Bundoora,
Victoria, Australia
School of Physiotherapy and Exercise Science, Griffith University, Gold Coast,
Queensland, Australia
3
Contents
Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
1. Measurement of Running Economy (RE). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
1.1 Treadmill RE Compared with Outdoor Running. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
1.2 Reliability of RE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468
1.3 Correcting RE for Body Mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468
2. RE and Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
3. Physiological Factors Affecting RE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470
4. Biomechanical Factors Affecting RE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
4.1 Anthropometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
4.2 Kinematics and Kinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
4.3 Flexibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 474
4.4 Ground Reaction Forces. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
5. Interventions to Improve RE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477
5.1 Strength Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477
5.2 Altitude Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
5.3 Training in the Heat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480
6. Conclusions and Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480
Running economy (RE) is typically defined as the energy demand for a given
velocity of submaximal running, and is determined by measuring the steady-state
consumption of oxygen (˙VO2) and the respiratory exchange ratio. Taking body
mass (BM) into consideration, runners with good RE use less energy and therefore
less oxygen than runners with poor RE at the same velocity. There is a strong
association between RE and distance running performance, with RE being a better
predictor of performance than maximal oxygen uptake (˙VO2max) in elite runners
who have a similar ˙VO2max.
RE is traditionally measured by running on a treadmill in standard laboratory
conditions, and, although this is not the same as overground running, it gives a
good indication of how economical a runner is and how RE changes over time. In
order to determine whether changes in RE are real or not, careful standardisation
of footwear, time of test and nutritional status are required to limit typical error of
measurement. Under controlled conditions, RE is a stable test capable of detecting
Abstract

Page 2

466
Saunders et al.
relatively small changes elicited by training or other interventions. When tracking
RE between or within groups it is important to account for BM. As ˙VO2 during
submaximal exercise does not, in general, increase linearly with BM, reporting
RE with respect to the 0.75 power of BM has been recommended.
A number of physiological and biomechanical factors appear to influence RE
in highly trained or elite runners. These include metabolic adaptations within the
muscle such as increased mitochondria and oxidative enzymes, the ability of the
muscles to store and release elastic energy by increasing the stiffness of the
muscles, and more efficient mechanics leading to less energy wasted on braking
forces and excessive vertical oscillation.
Interventions to improve RE are constantly sought after by athletes, coaches
and sport scientists. Two interventions that have received recent widespread
attention are strength training and altitude training. Strength training allows the
muscles to utilise more elastic energy and reduce the amount of energy wasted in
braking forces. Altitude exposure enhances discrete metabolic aspects of skeletal
muscle, which facilitate more efficient use of oxygen.
The importance of RE to successful distance running is well established, and
future research should focus on identifying methods to improve RE. Interventions
that are easily incorporated into an athlete’s training are desirable.
The ability to metabolise energy aerobically is a
prerequisite for superior endurance performance.[1-3]
In competitive distance running, successful per-
formance has been correlated to an athlete’s maxi-
mal oxygen uptake (˙VO2max).[1,3-5] Performance in
endurance events is directly influenced by altera-
tions in the availability of oxygen, carbohydrate and
fat, and the density of muscle mitochondria.[6]
˙VO2max is influenced by a variety of factors includ-
ing muscle capillary density, haemoglobin mass,
stroke volume, aerobic enzyme activity and muscle
fibre type composition.[6] Although a high ˙VO2max
is required for distance running, other physiological
and performance factors are important in determin-
ing endurance capacity.[4] These factors depend on
the race distance and include the percentage of
˙VO2max a runner can sustain without accumulating
lactic acid, the ability to utilise fat as a fuel at high
work rates and thereby ‘spare’ carbohydrate and
running at race pace with relatively low energy
expenditure (i.e. good running economy [RE]). The
velocity associated with attainment of ˙VO2max
(v˙VO2max) and the velocity at the onset of blood
lactate accumulation are good indicators of distance
running performance.[7]
Efficient utilisation of available energy facilitates
optimum performance in any endurance running
event. Efficiency refers to the ratio of work done to
energy expended.[8] RE is represented by the energy
expenditure and expressed as the submaximal ˙VO2
at a given running velocity.[4,9-11] The energy cost of
running reflects the sum of both aerobic and anaer-
obic metabolism, and the aerobic demand (measured
by the ˙VO2 in L/min) at a given speed does not
necessarily account for the total energy cost of run-
ning, which is measured in joules or kilojoules of
work done.[8] Runners with good RE use less oxygen
than runners with poor RE at the same steady-state
speed.[12] Figure 1 illustrates two international cali-
bre 10km runners measured in our laboratory: both
runners had a similar ˙VO2max, with the more effi-
cient runner (better RE) having a 10km time of 1
minute faster than the less efficient runner. The
steady-state condition is verified by the maintenance
of blood lactate concentration (La) at baseline
levels[13] and a respiratory exchange ratio (RER)
<1.[4] RE can vary among runners with a similar
˙VO2max by as much as 30%.[8] In elite or near-elite
runners with a similar ˙VO2max, RE is a better predic-
tor of performance than ˙VO2max.[5,14] Accordingly, it
 2004 Adis Data Information BV. All rights reserved. Sports Med 2004; 34 (7)

Page 3

Running Economy467
tively eliminated during indoor running; however,
transferring treadmill data to overground running
requires caution.[8,17] Pugh[18] estimated that 8% of
the total energy cost of middle-distance track run-
ning (5000m) is expended overcoming air resis-
tance. Another study estimated the amount of ener-
gy required to overcome air resistance was 4% for
middle-distance runners and 2% for marathon run-
ners.[19] When a tailwind velocity is equal to running
velocity, overground ˙VO2 was equivalent to tread-
mill ˙VO2.[17] Differences between overground run-
ning and treadmill running are more likely to be
observed as speed increases and the effect of air
resistance becomes more pronounced.[8] Hagerman
et al.[20] reported lower submaximal ˙VO2 values at
an altitude where the air is less dense than at sea
level. Using a 14.5 km/h headwind in order to simu-
late outdoor conditions, Costill and Fox[21] reported
a 15% difference in submaximal ˙VO2 between con-
trol conditions (no wind) and a simulated headwind.
It is clear that running on the treadmill is not the
same as running over ground, where wind resistance
affects ˙VO2. Furthermore, the technique of running
on a treadmill is different to running over ground
where the hamstrings are used to a greater extent to
produce propulsive forces. However, we can be
confident that RE measured on a treadmill is highly
correlated to RE over ground. It is reasonable, then,
to assume that interventions affecting RE on the
treadmill will similarly affect RE over ground. From
work in our laboratory, we have determined that
reliable measures of RE need to be obtained at
speeds eliciting ≤85% of ˙VO2max in highly trained
distance runners.
Recent technological advances have allowed
measurement of overground RE using portable oxy-
gen analysers. The K4 Cosmed analyser (Rome,
Italy) described by Hausswirth et al.[22] is a light-
weight, accurate, telemetric system that enables
measurement of energy requirements during both
submaximal and maximal exercise in the laboratory
or field. The K4 system allows continuous recording
of ˙VO2 during incremental progressive field tests to
accurately determine an athlete’s ventilatory charac-
teristics. The validity of the K4 Cosmed telemetric
Speed (km/h)
14 1618
Max
30
40
50
60
70
80
90
Subject 1
Subject 2
VO2 (mL/kg/min)
·
Fig. 1. Comparison of oxygen uptake (˙VO2) [mL/kg/min] in two
international calibre 10km runners, one with good economy (sub-
ject 1) and the other with poor economy (subject 2) [Saunders et al.
unpublished data, 2003]. Max = maximum.
follows that substantial improvements in RE could
facilitate improved performance in distance runners.
In summary, the relationship between RE and per-
formance is well documented, with many indepen-
dent reports demonstrating a strong relationship be-
tween RE and distance running perform-
ance.[1,4,5,10,15,16]
The purposes of this review are to examine the
validity and reliability of currently used tests for
measuring RE, examine research relating to physio-
logical and biomechanical factors which influence
RE, describe interventions that have attempted to
improve RE, and discuss potential areas for future
research directions in this field.
1. Measurement of Running
Economy (RE)
1.1 Treadmill RE Compared with
Outdoor Running
Measures of RE have typically been determined
in the laboratory by having the athlete run on a
motorised treadmill. This practice partially over-
comes many of the difficulties in obtaining reliable
metabolic data in the field (i.e. during training and
competition).[14] Air and wind resistance are effec-
 2004 Adis Data Information BV. All rights reserved. Sports Med 2004; 34 (7)

Page 4

468
Saunders et al.
system was demonstrated against a metabolic mea-
surement cart (CPX, Medical Graphics, Saint-Paul,
Minnesota, USA) during both submaximal and max-
imal exercise, with no difference observed between
the two oxygen analysis systems.[22] Recent research
has utilised the K4 portable oxygen analyser to
measure ˙VO2 in various intervention studies on
moderate to highly trained distance runners.[23-26] It
appears that overground RE can be measured in a
natural field setting with the K4 Cosmed telemetric
system and similar devices, although careful atten-
tion must be made to ensure post-testing results are
not influenced by changes in environmental condi-
tions.
vidual RE in 17 male runners following 30–60 min-
utes of treadmill familiarisation at the same time of
the day, in the same pair of shoes and in a non-
fatigued state.[32] There was a high day-to-day corre-
lation in RE (r = 0.95) with a mean coefficient of
variation (CV) of 1.3%. Pereira et al.[34] reported a
CV of 1.5% for RE in five trained male runners by
careful control of extrinsic factors such as time of
testing, diet, footwear and relative workload.
Pereira and Freedson[33] note that previous stud-
ies investigating intra-individual variation in RE
have not compared intra-individual variability be-
tween runners differing in training level. These in-
vestigators used seven highly trained males
(˙VO2max; 69.1 mL/kg/min) and eight moderately
trained males (˙VO2max 58.3 mL/kg/min) with test-
ing being carried out for 3 weeks at ~88% of the
velocity associated with the individual lactate
threshold (LT). Time of day, day of the week, diet
and footwear were controlled within each subject
across the three tests. CVs of 1.8% for the highly
trained group and 2.0% for the moderately trained
group (average 1.9%) were reported.[33] After ac-
counting for technical error, biological variation ac-
counted for ~94% of the intra-individual variation in
RE. The results suggest that workloads below LT
may permit more stable measures of RE to be ob-
tained. Brisswalter and Legros[28] demonstrated that
RE, respiratory measures and stride rate were stable
measures for assessing energy cost associated with
running in elite middle-distance runners. The au-
thors reported a variation of 4.7% in RE in ten elite
800m runners (˙VO2max, 68 mL/kg/min; mean 800m
time, 1:49 minutes). In a study of elite French dis-
tance runners, subjects were tested three times over
a 12-month period to determine the stability of RE.
RE was stable over the 12-month period despite an
improvement in ˙VO2max. The authors concluded
that in elite distance runners, RE is a difficult param-
eter to improve.[35]
1.2 Reliability of RE
Consideration of the typical intra-individual vari-
ation in RE is essential when investigating the effec-
tiveness of interventions aimed at modifying RE. As
has been previously noted,[14] small sample sizes
and omission of the typical error (TE) restricts the
degree to which meaningful conclusions on the im-
pact an intervention has on RE can be drawn. In
order to interpret the practical significance of vari-
ous interventions aimed at improving RE, a state-
ment of the test-retest reliability or TE should be
provided. In research settings, a rigorous experi-
mental design is necessary to control confounding
variables and to permit a valid determination of the
impact of interventions on RE. Daniels et al.[27]
observed an 11% variation in the stability of RE in
ten trained males running at 16 km/h, even after
controlling for variability associated with footwear
and test equipment. Well controlled reliability stud-
ies measuring RE show intra-individual variations
between 1.5–5%,[28-34] indicating that test-retest in-
tra-individual results are relatively stable.
Factors such as treadmill running experience,
footwear, time of day of testing, prior training activ-
ity and nutritional status may affect intra-individual
variation in RE.[29] Morgan[29] investigated 16 male
subjects who completed two 10-minute RE tests at
the same time of the day within a 4-day period,
wearing the same pair of shoes. The intra-individual
RE was 1.6%. Another study examined intra-indi-
1.3 Correcting RE for Body Mass
RE can be expressed as a ratio of a runners’ ˙VO2
(L/min) divided by their body mass (BM) in kilo-
grams.[36] However, when comparing individuals or
 2004 Adis Data Information BV. All rights reserved. Sports Med 2004; 34 (7)

Page 5

Running Economy469
groups who differ in BM this expression may induce
error because submaximal ˙VO2 during running does
not increase proportionately to BM.[37] In animals,
the oxygen cost of running does not increase propor-
tionately to BM[38] and in humans, the ˙VO2 per
kilogram of BM is higher in children than
adults.[8,39-50] Bergh et al.[37] suggests that, in
humans, the higher submaximal ˙VO2 observed in
children relates to differences in body size and not
merely growth and maturation. Sjodin and
Svedenhag[51] concur with this, and suggest that the
improved RE measured in relation to BM, observed
in adolescent boys during growth, may largely be
attributable to its measurement of ˙VO2 per kilogram
of BM.
Theory associated with elastic components of
muscle and connective tissue estimates ˙VO2 to be
proportional to the 0.75 power of BM.[52] Bergh et
al.[37] suggested that submaximal ˙VO2 and ˙VO2max
measures during running are better related to
BM–0.66 or BM–0.75 than BM–1. Furthermore, sever-
al studies[40,53,54] have shown an inverse relationship
between BM and submaximal ˙VO2/kg, providing
further support that ˙VO2 reported as mL/kg/min
may provide misleading comparative results.[37,51]
Another factor affecting RE is the pattern of
distribution of mass in the body. Carrying mass
distally increases the aerobic demand of running to a
greater extent than carrying mass closer to the centre
of mass.[55-58] Aerobic demand is increased by 1%
for every extra kilogram carried on the trunk, how-
ever, when the mass was carried in the shoes, aero-
bic demand increased by 10% for every additional
kilogram.[58] Jones et al.[56] found an average in-
crease in ˙VO2 of 4.5% per kilogram of load carried
on the feet when running at 12 km/h. Another study
investigated the effect of carrying mass on either the
thighs or feet and reported a 7% per kilogram in-
crease in ˙VO2 when the mass was carried on the
thigh compared with 14% per kilogram increase in
˙VO2 when carried on the feet.[57] The cushioning of
shoes also affects RE, with an approximate 2.8%
energy saving realised for treadmill running in well
cushioned shoes compared with poorly cushioned
shoes of similar mass.[59] The authors attributed this
finding to the extra muscular effort required to pro-
vide cushioning if the shoe itself does not provide
adequate shock absorption.
2. RE and Performance
The relationship between RE and performance is
well documented. Early research comparing elite
American distance runners (˙VO2max 79 mL/kg/min)
with good distance runners (˙VO2max; 69.2 mL/kg/
min), indicated that the elite runners had better RE
than good runners. When expressed as a percentage
of ˙VO2max this difference in RE was magnified,
with the elite runners working at a lower percentage
of their ˙VO2max.[16] Di Prampero et al.[15] stated that
a 5% increase in RE induced an approximately 3.8%
increase in distance running performance. As an
example of the relationship between RE and per-
formance, a case study of American mile record
holder Steve Scott, reported that during a 6-month
period of training, Scott improved his ˙VO2max by
3.8% (74.4 to 77.2 mL/kg/min).[8] During the same
period there was a 6.6% improvement in RE (48.5 to
45.3 mL/kg/min) at a running velocity of 16 km/h.
The combined improvement of an increased
˙VO2max and a better RE reduced the relative intensi-
ty of running at 16 km/h by 10.0% (65.1% to 58.6%
of ˙VO2max) and was associated with improved per-
formance during this period.[10]
Svedenhag and Sjodin[60] observed variations in
RE and performance in elite distance runners
(˙VO2max 75 mL/kg/min) who undertook alternating
sessions of slow distance, uphill and interval train-
ing over a 22-month period. Athletes significantly
reduced their ˙VO2 at 15 and 20 km/h accompanied
by enhanced performances over 5000m. However,
not all studies have demonstrated a significant rela-
tionship between RE and performance. Williams
and Cavanagh[54] failed to identify a significant rela-
tionship between RE at 13 km/h and 10km perform-
ance (~35 minutes) in a group of 16 runners. The
percentage of slow-twitch muscle fibres and the
runners ˙VO2max correlated best with 10km perform-
ance. Conley and Krahenbuhl[4] showed that RE was
a good predictor of performance in runners of com-
parable ability. In that study, 12 highly trained male
 2004 Adis Data Information BV. All rights reserved. Sports Med 2004; 34 (7)

Page 6

470
Saunders et al.
Performance in
distance runners
Environment
Altitude
Heat
Resistance
Training
Plyometrics
Training phase
Speed, volume,
intervals, hills, etc.
Running economy
Physiology
Adolescent
development
Metabolic
factors
Influence of
different running
speeds
Biomechanics
Elastic stored
energy
Mechanical
factors
Ground reaction
force
Flexibility
Anthropometry
Bodyweight
and composition
Muscle stiffness,
tendon length
Limb
morphology
Maximal oxygen
uptake (VO2max)
·
Fig. 2. Factors affecting running economy.
distance runners (˙VO2max ~72 mL/min and 10km
performance ~32 minutes) were tested 3–6 days
after they had competed in a 10km race. There were
significant correlations between submaximal ˙VO2
and performance at running speeds of 14, 16 and 18
km/h. Some 65% of the variation in race perform-
ance could be attributed to differences in RE, with
the more economical runners performing the best.
The more economical runners were able to run at a
lower percentage of their ˙VO2max, resulting in lower
La at a given speed. The latter factor is closely
associated with the pace a runner is able to maintain
for races >15 minutes.[61] These data also provide
evidence that RE at slower speeds are useful in
predicting performance in races at faster speeds.
Weston et al.[62] investigated the RE and perform-
ance of eight African (Kenyan) and eight Caucasian
distance runners. The Kenyan runners had similar
10km race performance to the Caucasian group,
despite having a 13% lower ˙VO2max. The RE of the
Kenyan runners was 5% better than the Caucasian
group, and when RE was normalised to BM–0.66 the
Kenyans had an 8% better RE. The Kenyan runners
also completed the 10km at a higher percentage of
their ˙VO2max but with similar La as the Caucasian
runners.
These studies indicate that improving athletes RE
is related to improvements in distance running per-
formance. RE is likely to be influenced by a number
of factors (figure 2) and any intervention (training,
altitude, heat) that can reduce the oxygen cost over a
range of running velocities will conceivably lead to
enhanced performance.
3. Physiological Factors Affecting RE
Fluctuations in physiological factors such as core
temperature (CTemp), heart rate (HR), ventilation
(VE) and La, may be associated with changes in RE
during competition.[14,63-65] Thomas et al.[12] investi-
gated the effect of a simulated 5km race on RE, VE,
CTemp, La and HR. RE was determined using a
constant treadmill speed eliciting 80–85% of the
athletes ˙VO2max. RE decreased significantly and
VE, CTemp, La and HR all increased significantly
from the beginning to the end of the 5km run. The
increased VE was the only factor that correlated
moderately with the decrease in RE (r = 0.64; p <
0.05), indicating a greater oxygen cost was associat-
ed with the increase in VE. A higher CTemp increases
 2004 Adis Data Information BV. All rights reserved. Sports Med 2004; 34 (7)

Page 7

Running Economy471
˙VO2 at a given speed.[66-69] Increases in the metabol-
ic cost from augmented circulation, VE and sweating
are the major factors that increase submaximal ˙VO2
and decrease RE.[68] In contrast, Rowell et al.[70]
stated that the mechanical efficiency of muscle in-
creases when CTemp is mildly elevated, reducing
˙VO2 by an amount equal to or greater than the
increase caused by changes in the cost of circulation,
VE and sweating. The composition of muscle fibres
also seems to influence RE. It has been suggested
that a higher percentage of slow-twitch muscle fib-
res is associated with better RE,[54,71,72] indicating
that metabolic activity or actual speed of contraction
of the muscle fibres may influence RE. Myocardial
˙VO2 also constitutes a significant fraction of whole
body ˙VO2 during exercise. Reductions in myocardi-
al ˙VO2 would result in improved RE from a more
efficient combination of HR and stroke volume (i.e.
a reduction in HR and increase in stroke volume).[73]
Little consensus exists on the effects of training
and RE, largely as a consequence of limitations in
existing experimental designs such as small sample
sizes, lack of multiple economy measures to account
for normal intra-individual variation and failure to
control for factors that influence RE (e.g. fatigue
level, state of training, treadmill experience and
footwear). Some,[10,74-81] but not all[5,42,82] studies
have reported improvements in RE after various
training interventions. The initial level of fitness of
the subjects is an important factor when considering
whether training alters RE.[42] Numerous studies
indicate that trained subjects are more economical
than untrained or less trained subjects[8,74,79,82-84] and
long-distance runners are more economical than
middle-distance runners.[8,16,84] The better RE in
long distance runners is largely attributable to a
lower vertical displacement of the runner’s centre of
mass during running probably related to neuromus-
cular adaptations induced by long, slow distance
training.[85] Endurance training leads to increases in
the morphology and functionality of skeletal muscle
mitochondria. An increase in the respiratory capaci-
ty of skeletal muscle permits trained runners to use
less oxygen per mitochondrial respiratory chain for
a given submaximal running speed. These responses
invoke improvements in RE, a smaller disturbance
in homeostasis and slower utilisation of muscle gly-
cogen in the working musculature.[86] Daniels and
Daniels[87] found that 800/1500m specialists were
more economical than marathon runners at veloci-
ties above 19 km/h, yet were less economical than
marathon runners at slower speeds. Without a mea-
sure of the anaerobic contribution to total metabo-
lism, it is difficult to conclude that the 800/1500m
specialists were more economical than the marathon
runners at the faster running velocities. Males were
more economical than females at common speeds
and relative intensities, but there was no difference
in RE between males and females at typical race
intensities for each sex. Another study was unable to
detect significant differences in RE between trained
male and female distance runners across four run-
ning velocities (12–16 km/h).[88]
Franch et al.[78] investigated the effects of three
types of intensive running training on RE in 36 male
recreational runners (˙VO2max ~55 mL/kg/min). Sub-
jects were assigned to either continuous-distance
training, long-repetition training (4–6 × 4 minutes
run with 3 minutes rest) or short-repetition training
(30–40 × 15 seconds run with 15 seconds rest)
groups, and trained three times a week for 6
weeks.[78] Runners undertaking continuous-distance
training and long-repetition training increased their
RE by approximately 3% while short-repetition
training had little effect on RE (0.9% change), sug-
gesting that longer training is the best way to im-
prove RE. Thomas et al.[65] suggests that those train-
ing in an effort to improve RE need to concentrate
on improving physiological characteristics such as
HR, VE, La and CTemp regulation in order to de-
crease the energy demand associated with these
parameters. Interval training may be beneficial to
RE by reducing HR, VE and La at higher running
speeds.[89]
4. Biomechanical Factors Affecting RE
Running involves the conversion of muscular
forces translocated through complex movement pat-
terns that utilise all the major muscle joints in the
body. High performance running is reliant on skill
 2004 Adis Data Information BV. All rights reserved.Sports Med 2004; 34 (7)

Page 8

472
Saunders et al.
Table I. Biomechanical factors related to better economy in runners[9]
Factor
Height
Ponderal index
Body fat
Leg morphology
Pelvis
Feet
Shoes
Stride length
Kinematics
Description for better running economy
Average or slightly smaller than average for males and slightly greater than average for females
High index and ectomorphic or mesomorphic physique
Low percentage
Mass distributed closer to the hip joint
Narrow
Smaller than average
Lightweight but well cushioned shoes
Freely chosen over considerable training time
Low vertical oscillation of body centre of mass
More acute knee angles during swing
Less range of motion but greater angular velocity of plantar flexion during toe-off
Arm motion that is not excessive
Faster rotation of shoulders in the transverse plane
Greater angular excursion of the hips and shoulders about the polar axis in the transverse plane
Low peak ground reaction forces
Effective exploitation of stored elastic energy
Comprehensive training background
Intermediate compliance
Kinetics
Elastic energy
Training
Running surface
and precise timing in which all movements have
purpose and function.[9] Clearly, changing aspects of
running mechanics that result in a runner using less
energy at any given speed is advantageous to per-
formance.[90,91] Biomechanical characteristics asso-
ciated with improved RE are shown in table I.[9] The
spring-mass model is an important factor associated
with RE, where the bounce of the body on the
ground is counteracted by the spring behaviour of
the support leg. During the eccentric phase of con-
tact, mechanical energy is stored in the muscles,
tendons and ligaments acting across joints. Recov-
ery during the concentric phase of the stored elastic
energy reduces the energy expenditure. An oscillat-
ing system is also characterised by a resonant fre-
quency. The resonant frequency is the frequency at
which a system freely vibrates after a mechanical
impulse.[92] RE was significantly correlated with
muscle stiffness (r = 0.80) and resonant frequency (r
= 0.79) of the propulsive leg, with stiffer muscles
operating at lower resonant frequencies eliciting the
best RE.[92] Several studies that have examined RE
and running mechanics after previously fatiguing
exercise and have reported little change in running
kinematics to explain decreases in RE.[93-96] In con-
trast, Hausswirth et al.[97] showed that RE was im-
paired during the last 45 minutes of a marathon run
on a treadmill, which was partly attributed to bi-
omechanical factors such as a greater forward lean
and a decrease in stride length. In a similar study,
investigating the effects of running a marathon on
RE, both submaximal ˙VO2 and RER increased dur-
ing, and 2 hours after, the marathon. The impaired
RE observed could not be completely explained by
any changes observed in the mechanics, and was
attributed to the increasing physiological stresses
(e.g. heat accumulation and increased reliance of fat
utilisation) associated with running a marathon.[94]
Thomas et al.[65] investigated the effects of a simu-
lated 5km race on RE and running mechanics of
trained female athletes. RE decreased during the
5km race, with athletes metabolising more oxygen
at the same intensity. The changes in RE observed in
this study were not caused by any alterations in the
mechanics of running, indicating that physiological
factors are more important in reducing RE. Taken
collectively, the weight of evidence from the ex-
isting literature suggests that if previously fatiguing
exercise is to reduce RE, it is likely to be through
physiological rather than biomechanical factors.
 2004 Adis Data Information BV. All rights reserved.Sports Med 2004; 34 (7)

Page 9

Running Economy473
4.1 Anthropometry
particular stride length/stride frequency for a given
running speed.[90]
Anthropometric characteristics such as height,
limb dimensions, body fat, as well as BM, have been
addressed as potential influences on RE. While leg
length contributes to angular inertia and the meta-
bolic cost of moving the legs during running,[9] there
seems little consensus on whether leg length is an
important factor in determining RE. In 31 male
distance runners with a 10km performance time of
~35 minutes, a large variation in RE was observed in
the absence of any differences associated with seg-
mental lengths and masses.[54] In contrast, there is
evidence that leg mass and distribution of mass may
influence RE. Williams and Cavanagh[54] reported a
modest inverse relationship between BM and sub-
maximal ˙VO2/kg (r = –0.52) and between maximal
thigh circumference and submaximal ˙VO2/kg (r =
–0.58), indicating that heavier than average runners
use less oxygen per kilogram of BM. Myers and
Steudel[58] hypothesised that a runner with a propor-
tionally smaller amount of BM concentrated in the
extremities, particularly the legs, would perform
less work moving their body segments during run-
ning, assuming that all other factors are unchanged
(e.g. speed, BM, running style).
The first studies comparing the biomechanical
characteristics of elite and good runners indicated
that elite runners had slightly less vertical oscilla-
tion, were more symmetrical, and had better RE.[102]
Williams and Cavanagh[103] found that better RE in
elite male distance runners was associated with a
more extended lower leg at foot strike, a lower
vertical force peak and a longer contact time. More
economical runners tend to exhibit less arm move-
ment, as measured by wrist excursion during the
stride.[54,104] Greater maximal plantar flexion veloc-
ity and greater horizontal heel velocity at foot con-
tact are also associated with better RE in elite male
distance runners.[103] While these authors demon-
strated links with various kinematic parameters and
RE, it would appear that further research is war-
ranted to determine if changing a runner’s kinemat-
ics induces an improvement in RE.
Recent research has comprehensively investigat-
ed biomechanical factors affecting RE.[105] ˙VO2 at
12–13 different running speeds was compared with
kinematic data and three-dimensional ground reac-
tion forces (GRF) simultaneously with telemetric
EMG recordings of selected leg muscles. Joint mo-
ments and power were calculated using two-dimen-
sional video analysis and the digitised segment coor-
dinates were transferred to a computer system. The
biomechanical parameters examined (angular dis-
placements between the ankle, knee and hip joints;
joint angular velocities) were not good predictors of
RE. However, force production during ground con-
tact, coupled with the activation of the leg extensors
during the pre-activity and braking phases and their
coordination with longer-lasting activation of the
hamstring muscles were of importance. The authors
pointed out that co-activation of the muscles around
the knee and ankle joints increases the joint stiff-
ness, which appears to be related to better RE. The
action of the hip extensors also becomes beneficial
in this respect during ground contact.[105] Refining
mechanical elements such as stride length and fre-
quency or the integration and timing of muscle
activity to utilise the storage and release of elastic
4.2 Kinematics and Kinetics
Early research suggested that well trained run-
ners running at 14 and 16 km/h were most economi-
cal at the runner’s self-selected stride length, com-
pared with other pre-determined stride lengths.[98]
More recent work has confirmed that the aerobic
demand of running at a given speed is lowest at a
self-selected stride length.[90] Submaximal ˙VO2 in-
creases curvilinearly as stride length is either length-
ened or shortened from that self-selected by the
runner.[90,98-101] Cavanagh and Williams[90] conclud-
ed that there is little need to dictate stride length for
well trained athletes since they display near optimal
stride length. They postulated two mechanisms for
this phenomenon. Firstly, runners naturally acquire
an optimal stride length and stride rate over time,
based on perceived exertion. Secondly, runners may
adapt physiologically through repeated training at a
 2004 Adis Data Information BV. All rights reserved. Sports Med 2004; 34 (7)

Page 10

474
Saunders et al.
energy more effectively may lead to improvements
in RE.[9]
Williams and Cavanagh[54] provided substantial
support for the notion that more economical runners
have identifiable kinetic patterns in their running
mechanics. They observed that ground-support time
and peak medial force correlated with submaximal
˙VO2 (r = 0.49 and 0.50, respectively). More eco-
nomical runners had lower first peaks in the vertical
component of the GRF, smaller antero-posterior and
vertical peak forces, and a more predominant rear-
foot striking pattern.[54] The authors suggested that
these characteristics affect muscular demands both
before and during support, with forefoot strikers
relying on musculature to assist with cushioning,
making them less economical. In contrast, rear foot
strikers tend to rely on footwear and skeletal struc-
tures to take the load and are more economical.[54]
Well cushioned shoes reduce oxygen cost by up to
2.8% over stiffer shoes of the same weight.[59,91]
However, it appears that there may be an individual-
ly optimal degree of cushioning, as shoes with a
‘spring rate’ that compliments the muscle-tendon
units contribute to the exploitation of stored elastic
energy.[9] Elastic energy stored during the eccentric
contractions of running substantially contributes to
propulsion via release during subsequent contrac-
tions.[106-109]
It has been estimated that the Achilles tendon and
tendons in the arch of the foot can store 35% and
17%, respectively, of the kinetic and potential ener-
gy gained and dissipated in a step while running at
moderate speeds.[110] Cavagna et al.[111] estimates
that ˙VO2 during running might be 30–40% higher
without contributions from elastic energy storage
and return. At higher speeds, elastic recovery of
energy prevails over the contractile machinery and
accounts for most of the work.[109,112] Elastic capaci-
tance is influenced by the rate and magnitude of
stretch, the level of activation and stiffness of the
muscle tendon unit, muscle length at completion of
the stretch and the time lag between completion of
the stretch and initiation of the succeeding concen-
tric contraction.[106,107,109] The major role of the mus-
cles during running is to modulate the stiffness of
the springs to maximise the exploitation of elastic
energy.[112-114]
4.3 Flexibility
Several studies contend that trunk and lower limb
flexibility affects RE.[115-117] Godges et al.[117] ob-
served that moderately trained athletic college stu-
dents increased their RE at all speeds (40, 60 and
80% ˙VO2max) with improved hip flexion and exten-
sion. Improved hip flexibility, myofascial balance,
and pelvic symmetry are thought to enhance neuro-
muscular balance and contraction, eliciting a lower
˙VO2 at submaximal workloads. These findings are
compatible with the general belief among runners
and coaches that improved flexibility is desirable for
increasing RE.[115]
In contrast, Gleim et al.[116] found that untrained
subjects who exhibited the lowest flexibility were
the most economical when running at speeds rang-
ing from 3–11 km/h. This finding was explained by
inflexibility in the transverse and frontal planes of
the trunk and hip regions of the body, stabilising the
pelvis at the time of foot impact with the ground.
This has the effect of reducing both excessive range
of motion and metabolically expensive stabilising
muscular activity.[116] Elastic energy storage and
return could be enhanced by having a tighter mus-
culo-tendinous system.[118-120] Tightness in the mus-
cles and tendons could increase elastic storage and
return of energy and reduce the submaximal ˙VO2
demand.
Craib et al.[115] examined the relationship be-
tween RE and selected trunk and lower limb flexi-
bility in well trained male distance runners. Inflexi-
bility in the hip and calf regions was associated with
better RE by minimising the need for muscle
stabilising activity and increasing the storage and
return of elastic energy. Another study found that
lower limb and trunk flexibility was negatively re-
lated to RE in international standard male distance
runners, with a significant relationship between the
sit-and-reach test score and submaximal ˙VO2 at 16
km/h.[121] Improved RE may reflect greater stability
of the pelvis, a reduced requirement for additional
muscular activity at foot strike, and a greater storage
 2004 Adis Data Information BV. All rights reserved.Sports Med 2004; 34 (7)

Page 11

Running Economy475
and return of elastic energy due to inflexibility of the
lower body.[121] A short and rapid stretch with a
short coupling time and a high force at the end of
pre-stretch increases musculo-tendon elasticity.[105]
Kyrolainen et al.[105] found that stiffer muscles
around the ankle and knee joints in the braking
phase of running increased force potentiation in the
push-off phase. Having stiffer, more inflexible mus-
cles in the legs and lower trunk could enhance RE
via increased energy from elastic storage and return,
which has no additional oxygen cost.
Taken collectively, the findings from these stud-
ies suggest that there is an optimal level of flexibility
whereby RE can benefit, although a certain degree
of muscle stiffness is also required to maximise
elastic energy storage and return in the trunk and
legs. Runners should not abandon stretching as part
of their training programmes, as a certain amount of
flexibility is also required for optimal stride length
at high running speeds.
In a well controlled study, Heise and Martin[123]
investigated the support requirements during foot
contact of 16 moderately trained male runners
(˙VO2max 62 mL/kg/min). Less economical runners
exhibited greater total and net vertical impulse, indi-
cating wasteful vertical motion. Correlations be-
tween total vertical impulse and ˙VO2, and net verti-
cal impulse and ˙VO2 were r = 0.62 and 0.60, respec-
tively. The combined influence of vertical GRF and
the time course of the force application explained
38% of the inter-individual variability in RE. Al-
though positive relationships were observed, other
GRF characteristics such as twisting, medial-lateral
or antero-posterior moments were not significantly
correlated with submaximal ˙VO2. Kyrolainen et
al.[105] found that GRF and the rate of force produc-
tion increased with increasing running speed. They
suggested that increasing the pre-landing and brak-
ing activity of the leg extensor muscles might pre-
vent unnecessary yielding of the runner during the
braking phase, helping them tolerate higher impact
loads. Pre-activation of these muscles is a preparato-
ry requirement for the enhancement of EMG activity
during the braking phase and for the time of muscu-
lar action with respect to the ground contact. Cen-
trally programmed pre-landing activity appears to
regulate the landing stiffness and compensates for
local muscular failure. Pre-activity increases the
sensitivity of the muscle spindle via enhanced alpha-
gamma co-activation potentiating stretch reflexes,
and enhancing musculo-tendon stiffness, with a re-
sulting improvement in RE.[105]
The requirement to support BM is a major meta-
bolic cost of running.[124] Vertical GRF is the major
determinant of the metabolic cost during run-
ning.[122,123,125] However, horizontal forces can sub-
stantially affect the metabolic cost of running, and
this is clearly observed when running on a windy
day.[124] Using a wind tunnel to apply a horizontal
impeding force, Pugh[126] showed that the metabolic
cost of running increased with the square of head-
wind velocity. Similarly, a harness to apply imped-
ing forces increased the metabolic cost of running
proportionally with an increase in external
work.[127-129] In a recent study, horizontal force was
4.4 Ground Reaction Forces
Fresh insight into the inter-individual variations
in RE has come from comparative biology. Kram
and Taylor[122] investigated the aerobic demand of
running, hopping and trotting in a variety of animal
species. They presented a simple inverse relation-
ship between aerobic demand and stance time inde-
pendent of an animal’s size, indicating that the ener-
gy cost of running is determined by the cost of
supporting an animal’s mass and the time course of
generating force.[122] GRF reflect the functional and
mechanical requirements during stance. During
ground contact, a runner activates muscles for the
purpose of stability and maintenance of forward
momentum. Excessive changes in momentum in the
vertical, antero-posterior and medial-lateral direc-
tions are wasteful in terms of metabolic energy
requirements. Linear impulse measures the change
in momentum and quantifies the time course of the
GRF. Quantifying the magnitude of support and
forces during ground contact may explain at least in
part the variability in RE among individuals of simi-
lar fitness.[123] Figure 3 depicts typical vertical and
horizontal GRF for three steps of one subject.
 2004 Adis Data Information BV. All rights reserved.Sports Med 2004; 34 (7)

Page 12

476
Saunders et al.
Speed (km/h)
0.20.00.60.40.81.0
−500
0
500
1000
1500
c
Force (N)
Impact
peak
Active
peak
−500
0
500
1000
1500
b
Force (N)
−500
0
500
1000
1500
a
Force (N)
Fv
Fh
Fig. 3. Typical vertical (Fv) and horizontal (Fh) ground reaction forces for three steps of a single subject: for unloaded, 0% bodyweight
applied horizontal force (AHF) control condition (a), with an impeding force of 6% bodyweight AHF (b), and with an aiding force of 15%
bodyweight AHF (c) [reproduced from Chang and Kram,[124] with permission].
altered to both impede and assist runners using a
pulley and rubber harness. At the two extreme con-
ditions, a 33% reduction in metabolic cost with a
15% assisting force and a 30% increase in metabolic
cost with a 6% impeding force were observed.[124]
The authors concluded that generating horizontal
force is metabolically more expensive per unit of
force than horizontal braking force during steady-
state running. It appears that the net resultant force
generated on the ground affects net muscle moments
 2004 Adis Data Information BV. All rights reserved.Sports Med 2004; 34 (7)

Page 13

Running Economy477
at each joint, as well as the force of each muscle
crossing the joint. Therefore, it may not be appropri-
ate to consider vertical and horizontal GRF as inde-
pendent determinants of metabolic cost.[124]
triathletes (˙VO2max 69 mL/kg/min).[137] In this
study, 14 weeks of a HWT intervention elicited in
the endurance/strength group a significantly lower
(11%) submaximal ˙VO2 compared with the endur-
ance-only group, and a marginal enhancement in RE
in the endurance/strength group compared with their
pre-test.
A specific type of strength training, known as
explosive-strength or plyometric training, invokes
specific neural adaptations such as an increased
activation of the motor units, with less muscle hy-
pertrophy than typical heavy-resistance strength
training.[130,138,139] Plyometric training enhances the
muscles’ ability to generate power by exaggerating
the stretch-shorten cycle, using activities such as
bounding, jumping and hopping.[140] Plyometric
training also has the potential to increase the stiff-
ness of the muscle-tendon system, which allows the
body to store and utilise elastic energy more effec-
tively.[141] Both these adaptations from plyometric
training could conceivably improve RE by generat-
ing more force from the muscles without a propor-
tionate increase in metabolic energy requirement.
Paavolainen et al.[80] indicated that 9 weeks of ex-
plosive-strength training improved RE (8%) and
5km performance (3%) with no changes in ˙VO2max
in moderately trained runners. This group also mea-
sured neuromuscular characteristics using a 20m
sprint test, the distance covered in five alternate
forward leg jumps and the corresponding contact
times (shorter times being better), as well as vertical
and horizontal forces measured on a force plate
during a constant-speed 200m run. The experimen-
tal group improved in all of these tests compared
with the control group.[80] These findings indicate
that explosive-strength training can improve RE and
performance as a consequence of enhanced neuro-
muscular function. Similarly, recent studies have
shown improvements in RE and performance after 6
weeks of plyometric training in moderately trained
subjects with no change in ˙VO2max,[140,141] with the
former study showing a 6% improvement in RE
across three running speeds and a 3% increase in
3km run performance. The study by Spurrs et al.[141]
demonstrated improvement in muscle-tendon stiff-
5. Interventions to Improve RE
RE is influenced by many physiological and bi-
omechanical variables; however, little research ex-
ists with regard to improving RE and endurance
performance by manipulation of these factors.[73]
Endurance training coupled with various other train-
ing methods has been shown to improve RE in
untrained and moderately trained subjects, with
trained runners having a better RE than their un-
trained or less trained counterparts.[8,74,79,82-84] Most
studies demonstrating improvements in RE as a
result of training have used untrained or moderately
trained subjects, and improvement in fitness is a
natural adaptation from endurance training. In high-
ly trained runners who already possess a well devel-
oped RE through years of endurance training, fur-
ther improvements in RE are seemingly difficult to
obtain. Three areas that have potential to improve
RE are strength training, altitude exposure and train-
ing in a warm to hot environment.
5.1 Strength Training
Endurance athletes must be able to sustain a high
average running velocity for the duration of a race.
This emphasises the role of neuromuscular charac-
teristics in voluntary and reflex neural activation,
muscle force and elasticity, running mechanics, and
the anaerobic capacities in elite endurance run-
ners.[80,130] The use of strength training is one inter-
vention thought to improve RE. Strength training
can improve anaerobic characteristics such as the
ability to produce high La and the production of
short contact times and fast forces.[131,132] Heavy
resistance training improves the endurance perform-
ance of untrained subjects[133-135] and RE of moder-
ately trained female distance runners without con-
comitant changes in ˙VO2max.[136] Recent work has
shown that a combination of heavy-weight training
(HWT) and endurance training improved running
performance and enhanced RE in well trained
 2004 Adis Data Information BV. All rights reserved. Sports Med 2004; 34 (7)

Page 14

478
Saunders et al.
ness and rate of force development during a seated
calf-raise test, a finding which also supports the
theory that improved RE from plyometrics is attrib-
utable to increased muscular power, and greater
storage and return of elastic energy. To date, there is
little research investigating the effects of plyometric
training on elite (˙VO2max >70 mL/kg/min) distance
runners.
livery and utilisation,[147,158,168,170] mechanisms that
potentially could improve an athlete’s RE. Howev-
er, to date little research has been undertaken on the
effects of altitude exposure on RE in highly trained
distance runners. Three investigations have reported
no change in submaximal ˙VO2 after a period of
altitude exposure,[145,156,164] while others have
demonstrated improved RE after various stints of
hypoxic exposure.[155,171,172] Similarly, after a period
of altitude acclimatisation, sea-level ˙VO2 during
submaximal cycling is either reduced[151,153,154,157] or
unchanged.[146,160] A tentative interpretation of these
findings is that altitude exposure for runners and
cyclists has no detrimental effects on economy and
there is good evidence to suggest that it may lead to
improvements.
In a study undertaken at sea level, RE of high-
altitude residents was compared with that of sea-
level residents. ˙VO2 was significantly lower in the
highlanders, indicating greater economy.[154] Green
et al.[153] examined the effects of a 21-day expedition
to altitude on submaximal cycling economy. Exper-
ienced mountain climbers were recruited for this
study and were based at 2160m, ascending to
heights of 6194m. Three days after the expedition,
subjects had a significantly lower ˙VO2 during a
40-minute submaximal cycling test (20 minutes at
60% ˙VO2max and 20 minutes at 75% ˙VO2max) than
before the expedition. With resting ˙VO2 unchanged
pre- and post-acclimatisation, the authors concluded
that the altitude exposure was responsible for the
improved economy, indicating net efficiency in-
creased by ~21% post-altitude acclimatisation, inde-
pendent of the power output. A similar study of
mountain climbers showed an ~8% reduction in
˙VO2 during steady-state, two-legged kicking exer-
cise, 3 days after a 21-day mountain climbing expe-
dition to 6194m.[157] These studies involved cycling
and two-legged kicking as the mode of exercise to
measure economy. Evidence to suggest that altitude
exposure improves economy of highly trained ath-
letes was demonstrated by Gore et al.[151] in a study
where six highly trained triathletes (˙VO2max 73 mL/
kg/min) improved cycling efficiency after 23 nights
sleeping at a simulated altitude of 3000m, compared
5.2 Altitude Exposure
The effect of altitude training on endurance per-
formance has been researched extensively.[142-168]
There is a widespread belief in the athletic commu-
nity that altitude training can enhance sea-level ath-
letic performance.[149,156,166] The mechanisms for
these improvements are not clear, but may include
haematological changes (i.e. increased red cell
mass)[148,156] and local muscular adaptations (such as
improved skeletal muscle buffer capacity).[151] The
traditional approach to altitude training involves
athletes living and training at a moderate
(1500–3000m) natural altitude. A more recent ap-
proach is for athletes to live/sleep at altitude and
train near sea level, the so-called live-high train-low
(LHTL) method.[156] Because the geography of
many countries does not readily permit LHTL, a
further refinement involves athletes living at simu-
lated altitude under normobaric conditions and
training at, or close to, sea level.[161] In recent years,
endurance athletes have utilised several new devices
and modalities to complement the LHTL approach.
These modalities include: normobaric hypoxia via
nitrogen dilution, which allows athletes to undertake
LHTL; supplemental oxygen to simulate normoxic
or hyperoxic conditions during exercise/sleep at nat-
ural altitude; and hypoxic sleeping devices which
permit athletes to sleep low and train high (LLTH).
Intermittent hypoxic exposure is another method
involving brief periods of hypoxic exposure via a
hypobaric chamber or inhalation of a hypoxic gas
mixture to stimulate erythropoietin production. Data
to support these claims are minimal and inconclu-
sive.[169]
Altitude acclimatisation results in both central
and peripheral adaptations that improve oxygen de-
 2004 Adis Data Information BV. All rights reserved.Sports Med 2004; 34 (7)

Page 15

Running Economy 479
with seven control athletes (˙VO2max 73 mL/kg/min)
who trained the same but slept at normal altitude
(~600m). In the LHTL group, overall submaximal
cycling efficiency improved significantly by 1%.[151]
Although this group was highly trained and mainly
consisted of triathletes who undertook a lot of run-
ning training, cycling was still the mode of exercise
used to determine improved economy, and further
research is required to determine whether altitude
training improves RE in highly trained runners.
of the excitation and contraction process to perform
work at lower energy costs.[153] Roberts et al.[160]
observed that 4300m altitude acclimatisation for 21
days decreased the reliance on fat as a fuel during
rest and cycling at 50% ˙VO2max. The authors sug-
gested that the shift towards increased dependence
on glucose metabolism and away from reliance on
fatty acid consumption under conditions of acute
and chronic hypoxia, is advantageous because glu-
cose is a more efficient fuel in terms of generating
ATP per mole of oxygen. Another suggested im-
provement in efficiency is the reduced energy re-
quirement of one or more processes involved in
excitation and contraction of the working muscles as
a result of metabolic adaptations from altitude accli-
matisation.[153] The reduction in by-product accumu-
lation, such as adenosine diphosphate (ADP), inor-
ganic phosphate and hydrogen that occur after alti-
tude acclimatisation, increases the amount of free
energy released from ATP hydrolysis and depresses
the need to maintain hydrolysis rates at pre-accli-
matised levels.[153,173]
Currently, three studies have demonstrated im-
proved economy in highly trained runners.[155,171,172]
Katayama and colleagues[171] have demonstrated on
two occasions that intermittent hypoxic exposure
improves RE in well trained runners. The first study
reported that simulated hypoxic exposure using in-
termittent hypobaria of 4500m 3 hours per day for
14 consecutive days improved RE and performance
in well trained runners (˙VO2max 68 mL/kg/min).
Altitude exposure improved RE by 2.6% (14 km/h)
and 3.3% (16 km/h), improved 3000m run time by
1% and time to exhaustion on the treadmill by 2.7%.
The improvement in RE accounted for 37% of the
improvement observed in the 3000m time trial.[155]
More recently, it was demonstrated that 3 hours per
day for 2 weeks of intermittent exposure to
normobaric hypoxia (12.3% oxygen) improved RE
by 2.6% (14 km/h) and 2.9% (16 km/h) in well
trained runners (˙VO2max 68 mL/kg/min). The im-
proved RE was accompanied by a decreased HR
(3.3% and 3.9% at 14 and 16 km/h, respectively)
and a trend towards improved 3000m run time
(1.3%, p = 0.06).[171] Another recent study demon-
strated that 20 days of sleeping at simulated altitude
(2000–3100m) and training near sea level (600m)
improved (3.3%, p = 0.005) RE in elite distance
runners (˙VO2max 73 mL/kg/min) in the absence of
any changes in cardiorespiratory measures or red
cell mass.[172]
Saltin et al.[174] investigated the physiological
characteristics of Kenyan and Scandinavian runners.
The Kenyan runners lived and trained at altitude
while the Scandinavian runners lived and trained at
sea level. Kenyan runners did not accumulate La
during running until near very high or peak exercise
intensities, and had much lower La both at altitude
and sea level at high relative exercise intensities.
Similarly, Weston et al.[175] reported Kenyan runners
had higher resistance to fatigue when running at the
same percentage of peak treadmill velocity than
Caucasian runners, despite similar ˙VO2max values in
the two groups. Whilst these studies of runners
native to high altitude do not necessarily indicate the
effect of training at altitude, it has been reported that
exercise after altitude training results in reduced La
production at submaximal exercise, with lower
blood and muscle La being reported.[146,152,176] On
this basis, altitude training allows athletes to main-
tain a given exercise intensity with lower accumula-
tion of La during post-acclimatised sea-level exer-
cise. One of the mechanisms for lower plasma La
accumulation is an increase in skeletal muscle oxi-
Mechanisms that have been suggested to improve
economy after altitude exposure include: decreased
cost of VE, a shift towards a greater glycolytic
involvement of adenosine triphosphate (ATP) re-
generation, greater carbohydrate utilisation for oxi-
dative phosphorylation and/or an increased ability
 2004 Adis Data Information BV. All rights reserved. Sports Med 2004; 34 (7)