ArticlePDF Available

A 1% treadmill grade most accurately reflects the energetic cost of outdoor running

Authors:

Abstract and Figures

When running indoors on a treadmill, the lack of air resistance results in a lower energy cost compared with running outdoors at the same velocity. A slight incline of the treadmill gradient can be used to increase the energy cost in compensation. The aim of this study was to determine the treadmill gradient that most accurately reflects the energy cost of outdoor running. Nine trained male runners, thoroughly habituated to treadmill running, ran for 6 min at six different velocities (2.92, 3.33, 3.75, 4.17, 4.58 and 5.0 m s-1) with 6 min recovery between runs. This routine was repeated six times, five times on a treadmill set at different grades (0%, 0%, 1%, 2%, 3%) and once outdoors along a level road. Duplicate collections of expired air were taken during the final 2 min of each run to determine oxygen consumption. The repeatability of the methodology was confirmed by high correlations (r = 0.99) and non-significant differences between the duplicate expired air collections and between the repeated runs at 0% grade. The relationship between oxygen uptake (VO2) and velocity for each grade was highly linear (r > 0.99). At the two lowest velocities, VO2 during road running was not significantly different from treadmill running at 0% or 1% grade, but was significantly less than 2% and 3% grade. For 3.75 m s-1, the VO2 during road running was significantly different from treadmill running at 0%, 2% and 3% grades but not from 1% grade. For 4.17 and 4.58 m s-1, the VO2 during road running was not significantly different from that at 1% or 2% grade but was significantly greater than 0% grade and significantly less than 3% grade. At 5.0 m s-1, the VO2 for road running fell between the VO2 value for 1% and 2% grade treadmill running but was not significantly different from any of the treadmill grade conditions. This study demonstrates equality of the energetic cost of treadmill and outdoor running with the use of a 1% treadmill grade over a duration of approximately 5 min and at velocities between 2.92 and 5.0 m s-1.
Content may be subject to copyright.
A 1% treadmill grade m ost accurately re¯ ects the
energetic cost of outdoor running
A N D R E W M . JO N ES
Ó
and JO N AT H A N H . D O U ST *
Chelsea School Research Centre, U niversity of B righton, Gaudick R oa d, Eastbour ne B N 20 7SP, U K
Accepted 21 January 1996
When running indoors on a treadmill, the lack of air resistance results in a lower energy cost compared with
running outdoors at the sam e velocity. A slight incline of the treadmill gradient can be used to increase the
energy cost in compensation. The aim of this study was to determine the treadmill gradient that most
accurately re¯ ects the energy cost of outdoor running. Nine trained m ale runners, thoroughly habituated to
treadmill running, ran for 6 m in at six different velocities (2.92, 3.33, 3.75, 4.17, 4.58 and 5.0 m s
-1
) with
6 min recover y between runs. This routine was repeated six times, ® ve times on a treadmill set at different
grades (0% , 0%, 1% , 2%, 3% ) and once outdoors along a level road. Duplicate collections of expired air were
taken during the ® nal 2 min of each run to determine oxygen consumption. The repeatability of the
methodology was con® rm ed by high correlations (r = 0.99) and non-signi® cant differences between the
duplicate expired air collections and between the repeated runs at 0% grade. The relationship between oxygen
uptake (
Ç
VO
2
) and velocity for each grade was highly linear (r > 0.99). At the two lowest velocities,
Ç
VO
2
during
road running was not sign cantly different from treadm ill running at 0% or 1% grade, but was signi® cantly
less than 2% and 3% grade. For 3.75 m s
-1
, the
Ç
VO
2
during road running was sign cantly different from
treadmill running at 0% , 2% and 3% grades but not from 1% grade. For 4.17 and 4.58 m s
-1
, the
Ç
VO
2
during
road running was not signi® cantly different from that at 1% or 2% g rade but was sign cantly greater than 0%
grade and sign cantly less than 3% grade. At 5.0 m s
-1
, the
Ç
VO
2
for road ru nning fell between the
Ç
VO
2
value
for 1% and 2% grade treadmill running but was not sign cantly different from any of the treadmill grade
conditions. This study demonstrates equality of the energetic cost of treadmill and outdoor running with the
use of a 1% treadmill grade over a duration of ~ 5 min and at velocities between 2.92 and 5.0 m s
-1
.
Keywords: Fitness assessment, running economy, treadmill running.
In troduction
In recent years there has been a growing interest by
athletes and coaches in physiolog ical tests which track
changes in training status, predict current perform ance
capability and which may help guide prescription of
training prog ram m es. This has focused attention on the
validity of transferring information gained in the labor-
atory to the outdoor environment. For runners and
team -game players, the m otorized treadm ill is com-
monly used to im pose an exercise stress; however,
unlike outdoor running, there is no air resistance dur-
ing treadmill running. This has potential consequences
fo r both therm oregulation and the energy cost of over-
com ing air resistance. The m ovem ent of air over the
body aids convective heat loss an d can reduce exercise
heart rate, since the demand for peripheral blood ¯ ow
is reduced (Riggs
et al.
, 1981; W illiam s and Kilgour,
1993). To com pensate for this, electric fans are com-
monly used in the laboratory. M oving air from a fan
aids convective heat loss but the mass of m oving air is
far too sm all to impose any signi® cant force retarding
fo rward motion. In order to compensate for the lack of
air resistance, som e researchers have used sligh t
inclinations of the treadm ill, including 1.0% (H eck
et
al
., 1985) and 2.0% (Tegtbur
et al.
, 1993); however
neither of these studies rep orted data in support of
*Author to whom all correspondence should be addressed.
Ó
Present address
: E xercise Physiology L aboratory, The Sports Council
for Wales, Welsh Institute of Sport, Sophia Gardens, Cardiff CF1
9SW, UK.
0264-0414 /96 © 1996 E. & F.N. Spon
Jour nal of Sports Sciences
, 1996,
14
, 321-327
their chosen adjustm ent. O ther researchers have not
made any attem pt to account fo r the effect of air resist-
ance (Noakes
et al.
, 1990; Weltman
et al.
, 1990) or else
made no reference to the m atter (Hale
et al.
, 1988).
The aim of this study was to determ ine the treadm ill
gradient that m ost accurately re¯ ects the energy cost of
outdoor running over the range of velocities com monly
used fo r training and racing in a g roup of good stand-
ard runners. We hypothesized that oxygen uptak e
(
Ç
V
O
2
) wh en running on the level outdoors would be
higher at any given velocity than
Ç
V
O
2
when running on
a level treadm ill indoors and that the difference would
increase w ith increasing velocity.
M ethods
Subjects and exper imental plans
Nine m ale runners w ho were in regular training for dis-
tance running, and who were thorough ly fam iliar with
treadm ill ergom etry and laboratory procedures, volun-
teered to participate in the study. T he subjects provided
written informed consent and the experim ental proto-
col was approved by the local ethical com m ittee. The
su bjects’ m ean (±
S
.
D
.) age, body mass, sum of four
skinfolds (Durnin and Womersley, 1974) and
Ç
V
O
2
max
were 24.9 ± 5.2 years, 72.0 ± 3.3 kg, 24.9 ± 5.7 m m
and 65.1 ± 2.7 m l kg
-1
min
-1
, respectively.
The subjects were required to r un for 6 min at each
of six different velocities (2.92, 3.33, 3.75, 4.17, 4.58
and 5.0 m s
-1
, i.e. 10.5-18.0 km h
-1
) w ith 6 m in recov-
ery between runs. This routine was repeated six tim es,
® ve times on a treadmill set at different grades (0%,
0%, 1%, 2%, 3%) and once outdoors. The 0% tread-
mill g rade condition was applied tw ice to allow assess-
ment of the reliab ility of the test methodology. The tests
took place on six separate days and the six conditions
were applied in random order. Testing was completed
within 2 weeks for all subjects. The velocities were
selected to represent typical training and com petition
velocities for com petitive but non-elite m ale r unners.
On each occasion, the subjects reported to the labor-
atory following an overnight fast and each subject was
tested at the sam e tim e of day (06.00-08.30 h). Stren-
uous training was forbidden in the 24 h preceding each
test. The subjects w ore the sam e footwear and light-
weight running kit on each occasion, and performed
the sam e individual warm -up routine.
During the last 2 m in of each run,
Ç
V
O
2
was deter-
mined by indirect calorim etry. T he subjects breathed
through a Salford low-resistance valve and w ide-bore
Falconia tubing. Expired air was collected into D ouglas
bags over a period of 50 s. D uplicate collections were
made to assess the repeatability of the method and also
to establish w hether a steady-state had been reached.
Expired air was analysed fo r the concentrations of O
2
and CO
2
through a param agnetic transd ucer (Ser vo-
mex series 1100, Crowborough, England) or an infra-
red transducer (Servom ex m odel 1490) respectively,
previously calibrated using three-point standards. Gas
vo lum e was deter m ined using a dry gas m eter (H ar vard
Ltd, Edenbridge, England), previously calibrated using
a Tissot spirometer and checked against a precision 7
litre gas syringe (H ans Rudolph Inc., K ansas, USA).
All gas volumes were corrected to standard tem pera-
ture and dry pressure. Heart rate (HR ) was measured
telemetrically using a Polar Sport Tester heart rate
monitor (Polar Electro, K em pele, Finland).
Treadm ill procedures
The subjects ran on a slatted-b elt Woodway m odel
ELG 2 treadm ill (CardioKinetics, Salford, England).
The treadm ill speed was checked by tim ing belt revolu-
tions during the last m inute of each run fo r each sub-
ject. T he treadm ill speed was always within 0.03 m s
-1
of the desired speed. The treadm ill gradient was
checked by a spirit level and large p rotractor arranged
on a long ¯ at plank of wood. Through a screwthread
device, the spirit level could be adju sted to the exact
horizontal and the grade read from the protractor.
Outdoor procedures
For the outdoor condition, the subjects ran on a level
(0.2 m descent in 1.6 km) seafront promenade using
an out-and-back course. They were paced at the appro-
priate velocities by an exp erimenter who cycled at a
steady pace 2 m to the side of the subject. The cycle
was ® tted with a `bicycle com puter for continuous
monitoring of velocity and distance covered. The
bicycle computer was calibrated regularly using the
treadm ill to ensure consistency in velocity m easure-
ments between the indoor and outdoor conditions. The
total tim e taken to cover the distance gave an independ-
ent con® rm ation of the average velocity of running.
Actual velocities for each individual were used in the
data analysis and these were w ithin ± 0.06 m s
-1
of the
desired velocity. Shortly after reaching the designated
turning point fo r each run, the su bjects were ® tted with
a nose clip and m outhpiece. A fter 60 s to allow habit-
uation,
Ç
V
O
2
was determ ined by the sam e method used
in the treadm ill runs. The D ouglas bag was suppor ted
by another exp erimenter running ab out 1 m to the side
and slightly behind the subject. T his m ethod ensured
minim al interference to the air movem ent exp erienced
by the subject. W ind speed was continually m onitored
using an anemom eter. O utdoor r unning was only per-
fo rm ed if the wind speed was less than 2.0 m s
-1
;
322
Jones and Doust
usually wind sp eed was m uch less than this and aver-
aged 0.2 ± 0.5 m s
-1
. The laboratory and outside tem -
peratures were sim ilar an d ranged b etween 14¡ and
18¡ C .
Statistical analyses
Repeatability data were analysed with Pearson
product-m om ent correlation coe cients and paired
t
-
tests. The
Ç
V
O
2
and HR at different running velocities
and grades were analysed with analysis of variance with
the site of signi® cant (
P
< 0.05) differences deter-
mined using Tukey s
post-hoc
com parison.
Results
Attainment of a steady-state
It is of major im por tance to this study that measure-
ments were made in the steady-state and were reliable.
The
Ç
V
O
2
was m easured tw ice during the last 2 m in of
each run. T he two collections were highly correlated
(
r
= 0.99). For all collections (over all velocities and
grades both indoors and outdoors), the
Ç
V
O
2
from the
® rst D ouglas bag (43.8 ± 10.3 ml kg
-1
min
-1
) was not
signi® cantly different to the
Ç
V
O
2
from the second
Douglas bag (44.0 ± 10.5 m l kg
-1
min
-1
). The mean of
the two collections was used in the data analysis. Test-
retest reliability was assessed at 0% grade. T he data
were highly correlated (
r
= 0.99) and were not signi® -
cantly different between the two tests. Heart rate dur-
ing test 2 was lower than HR during test 1 and this
approached signi® cance (
P
= 0.053). H owever, the
mean difference in H R between test and retest (1.6
beats min
-1
) was trivially small.
Oxygen consumption
Table 1 shows the oxygen consumption at each velocity
for each condition. T he
Ç
V
O
2
during road r unning was
not signi® cantly different from that at 1% treadm ill
grade at any velocity. T he trend was for road running to
require a sim ilar oxygen cost to 0% or 1% g rade at the
slower velocities, while at higher velocities the oxygen
cost of ro ad running was closer to the oxygen cost of
treadm ill running at 1% and 2% grade. At 5.0 m s
-1
,
the
Ç
V
O
2
during road running was not signi® cantly dif-
fe rent from the
Ç
V
O
2
at any treadmill grade.
The relationship between
Ç
V
O
2
and running velocity
(
v
) was highly linear. Regression equations for the vari-
ous treadm ill grades and for road running were as fol-
low s (
Ç
V
O
2
in m l kg
-1
min
-1
;
v
in m s
-1
):
0%
Ç
V
O
2
= 13.7
v
- 11.10 (
r
= 0.99,
S
.
E
. = 0.46)
1%
Ç
V
O
2
= 13.5
v
- 8.50 (
r
= 0.99,
S
.
E
. = 1.12)
2%
Ç
V
O
2
= 14.1
v
- 8.23 (
r
= 0.99,
S
.
E
. = 0.47)
3%
Ç
V
O
2
= 14.0
v
- 5.64 (
r
= 0.99,
S
.
E
. = 0.48)
Road
Ç
V
O
2
= 14.4
v
- 11.90 (
r
= 0.99,
S
.
E
. = 0.65)
Figure 1 sh ows the relationship between
Ç
V
O
2
and
velocity for each condition with the reg ression lines
shown. H eart rate followed a sim ilar pattern of change
to that show n by oxygen uptake (Table 2). R unning on
a level road elicited a hear t rate ~ 3-4 beats min
-1
higher than that found during running on a ¯ at
treadm ill.
Discussio n
The results dem onstrate the oxygen cost of running on
the level outdoors is g reater than w hen running on the
level indoor s. A 1% treadm ill grade was fo und to re¯ ect
most accurately the oxygen cost of running outdoors
and the oxygen cost at this grade was not signi® cantly
different to the oxygen cost of outdoor running fo r
velocities between 2.92 and 5.0 m s
-1
. These results
con® rm the ® nding of Davies (1980), who, using data
obtained from a lim ited number of subjects running on
a treadmill housed in a wind tunnel, reasoned that the
Table 1 M ean (±
S
.
D
.) oxygen consumption (ml kg
-1
min
-1
) at each velocity (m s
-1
) and in each experi-
mental condition
Experim ental condition
Running
velocity 0% grade 1% grade 2% grade 3% grade Level road
2.92 29.6 ± 1.2
c ,d
31.3 ± 1.4
c ,d
33.2 ± 1.7
a ,b ,d
35.6 ± 1.6
a ,b ,c ,e
31.1 ± 1.8
c ,d
3.33 34.6 ± 1.9
c ,d
36.4 ± 1.6
c ,d
38.4 ± 1.7
a ,b ,d ,e
40.9 ± 1.3
a ,b ,c ,e
35.7 ± 1.3
c ,d
3.75 39.0 ± 1.9
b, c ,d ,e
41.4 ± 1.2
a ,c , d
44.1 ± 1.9
a ,b ,d ,e
46.5 ± 1.7
a ,b ,c ,e
41.7 ± 1.3
a ,c , d
4.17 44.9 ± 1.9
b, c ,d ,e
48.3 ± 1.5
a ,d
50.3 ± 2.4
a ,d
53.0 ± 1.7
a ,b ,c ,e
48.1 ± 2.2
a ,d
4.58 51.0 ± 2.0
c ,d ,e
55.3 ± 2.3
a ,d
56.9 ± 2.3
a
59.3 ± 2.3
a ,b ,e
54.8 ± 2.7
a ,d
5.00 57.3 ± 2.7
c ,d
58.1 ± 2.6
d
61.9 ± 2.6
a
64.2 ± 2.7
a ,b
60.5 ± 3.0
Sign cantly different (
P
< 0.05) from:
a
0% grade,
b
1% grade,
c
2% g rade,
d
3% grade,
e
level road.
Treadmill g rade and the energetic cost of outdoor running
323
air resistance experienced during outdoor running on a
calm day was equivalent to running on a treadm ill with
1% slope. The present data also reinforce the work of
Heck
et al.
(1985), who m easured blood lactate but not
Ç
V
O
2
at a variety of treadm ill grades (0%-6% ) and dur-
ing outdoor running on a variety of surfaces. They con-
cluded that a treadm ill gradient set som ewhere
between 0% and 2% was necessary to allow application
of laboratory data to ® eld conditions.
Accurate evaluation of the energy demand for par-
ticular running velocities requires that exercise is per-
form ed in the steady-state. Barstow and M ole (1991)
showed that fo r intensities below the lactate threshold,
a steady-state for
Ç
V
O
2
is attained in less than 3 min,
while fo r m oderate-intensity work above the lactate
threshold the attainm ent of a steady-state is delayed. A
Ç
V
O
2
steady-state may never be attained during con-
stant load exercise at intensities greater than about
90%
Ç
V
O
2
max (Barstow and M ole, 1991). For the sub -
jects in the present exp erim ent, lactate threshold
(m easured in a parallel study by Jones and Doust,
1995) occurred at between 4.58 and 5.14 m s
-1
, so
som e of the subjects would have been exercising above
the lactate threshold for the 5 m s
-1
condition. H ow-
ever, despite the probability of elevated blood lactate
levels in som e subjects at 5.0 m s
-1
, the duplicate
Ç
V
O
2
determinations were not signi® cantly different and all
su bjects were able to com plete the 6 min of running.
For a m inority of subjects at 5 m s
-1
and at the higher
grades of 2% and 3%, oxygen uptake approached
Ç
V
O
2
max. This led to a larger standard error of the m ean
Ç
V
O
2
an d is the probable explanation for the insig-
ni® cant differences between conditions found at
5 m s
-1
.
While the
Ç
V
O
2
-velocity relationship for outdoor run-
ning is theoretically cur vilinear, with
Ç
V
O
2
increasing as
Figure 1 The relationship between oxygen consumption and running velocity on the road and at various treadmill grades.
Table 2 Mean (±
S
.
D
.) heart rate (beats min
-1
) at each velocity (m s
-1
) and in each experim ental con-
dition
Experim ental condition
Running
velocity 0% grade 1% grade 2% grade 3% grade Level road
2.92 116 ± 7
c ,d
119 ± 8
c ,d
126 ± 7
a , b,e
132 ± 6
a , b,e
119 ± 6
c ,d
3.33 124 ± 6
c ,d
128 ± 7
d
134 ± 7
a
141 ± 6
a , b,e
128 ± 7
d
3.75 137 ± 6
c ,d
140 ± 6
d
143 ± 6 153 ± 7
a , b,e
140 ± 8
d
4.17 149 ± 6
c ,d
153 ± 7
d
155 ± 7 164 ± 7
a , b,e
153 ± 7
d
4.58 160 ± 7
c ,d
164 ± 7
d
166 ± 9 174 ± 9
a , b
165 ± 7
5.00 171 ± 9
c ,d
173 ± 8
d
175 ± 9 186 ± 11
a ,b
174 ± 9
Sign cantly different (
P
< 0.05) from:
a
0% grade,
b
1% grade,
c
2% g rade,
d
3% grade,
e
level road.
324
Jones and Doust
a cubic function of velocity (Pugh, 1970, 1971), Leger
and M ercier (1984) reported that, at least for velocities
below 6.94 m s
-1
, the relationship can be adequately
described by a linear fu nction. Interestingly, however,
their regression equation fo r outdoor running was dis-
tinctly different from that used to describe level grade
treadm ill running , particularly for velocities exceeding
4.17 m s
-1
. Despite the non-linearity between
Ç
V
O
2
and
velocity described by Pugh (1970) for track running ,
Pugh also noted that the relationship could be ade-
quately described by a linear function between 2.22
and 6.11 m s
-1
, an observation subsequently con® rm ed
by M cM iken and D aniels (1976) and D avies (1981).
In the present study, a linear equation was found to ® t
the
Ç
V
O
2
-velocity relationship alm ost perfectly
(
r
2
= 99.7%), whereas a square function and a cubic
function ® tted less well (
r
2
= 98.3% and 95.7% ,
respectively). Extrapolation of the data to velocities in
excess of 5.0 m s
-1
should be m ade with caution since,
from both physical princip les and empirical studies
during human locomotion, the
Ç
V
O
2
-velocity relation-
ship takes on greater curvilinearity at higher velocities
(M cM iken and D aniels, 1976; D avies, 1980). N ever-
theless, the trend of the data, w ith the slope of the
regression for the road condition tending to be g reater
than that of any of the other regressions (see Fig. 1),
su ggest that at running velocities of ~ 5-6 m s
-1
, a
grade of 2% might best re¯ ect the
Ç
V
O
2
of horizontal
outdoor running, while at velocities around 2-3 m s
-1
,
a g rade of 0.5% is m ost appropriate. Although the sub-
jects in this study were trained m ales, there is no evi-
dence to suggest that the
Ç
V
O
2
-velocity relationship
differs between males and females (Falls and H um -
phreys, 1976; Daniels
et al
. 1986; Daniels and D aniels,
1992).
The regression equation for
Ç
V
O
2
on velocity during
horizontal treadmill running in the present study was
similar to other data on trained runners (Costill
et al.
,
1973; M cM iken and D aniels, 1976; Bransford and
Howley, 1977). H owever,
Ç
V
O
2
at any velocity was con-
sistently lower (i.e. the r unning economy was greater)
than has been reported previously for untrained sub-
jects (Shephard, 1969; Bransford and H ow ley, 1977;
Bassat
et al.
, 1985; AC SM , 1991), and the regression
equation for the present study provided
Ç
V
O
2
estim ates
that were considerably lower than those provided by
the ge neral equation calculated by Leger and M ercier
(1984) from pooled data of numerous early studies.
Evidence for better r unning economy in trained sub-
jects is compelling (Dan iels and D aniels, 1992; M or-
gan and Craib, 1992; M organ
et al
., 1994). H owever,
even in the present subjects, with high and relatively
hom ogeneous
Ç
V
O
2
max scores (coe cient of varia-
tion = 4%), large inter-individual differences in run-
ning economy were observed. For example, the
Ç
V
O
2
measured at 4.58 m s
-1
and 1% grade ranged from
49.3 to 57.0 m l kg
-1
min
-1
. These results underline the
potential error in the prediction of
Ç
V
O
2
max from per-
fo rmance tests such as the m ulti-stage shuttle (Ram s-
bottom
et al.
, 1988), and in perform ance prediction
from
Ç
V
O
2
max alone in groups with relatively similar
Ç
V
O
2
max scores (M organ
et al.
, 1989). O ther invest-
igators have reported w ide inter-individual differences
in running economy in groups of runners with homo-
geneous
Ç
V
O
2
max and have shown that running econ-
omy can discriminate perform ance in such groups
(C onley and Krahenbuhl, 1980; Powers
et al.
, 1983).
Other potential causes for differences between tread-
mill and outdoor running include the r unner gaining
som e energy from the motor-driven treadmill b elt,
changes in the pattern of locom otion due to differing
surfaces or to instability consequent on visual cues
ar riving from static rather than m oving su rroundings
(Van Ingen Schenau, 1980), and the degree of habitua-
tion to treadm ill running (C onley and Krahenbuhl,
1980). Van Ingen Schenau (1980) has questioned
whether there is any fu ndam ental difference between
the two types of running. H e dem onstrated mathemat-
ically that as long as the treadm ill belt velocity is con-
stant, then m echanically it is appropriate to use a
coordinate system that moves with the belt. With su ch a
system , no m echanical difference exists between tread-
mill and overground running. Although in some tread-
mills the belt sp eed can be unstable, the m odern
high-velocity m achines designed fo r precision scienti® c
work not only p rovide stable control of speed but also a
slatted surface that gives the runner very sim ilar pro-
prioceptive feedback to running overg round. It is po ss-
ible that some of the difference in
Ç
V
O
2
measured
between treadmill and outdoor running was a conse-
quence of the different elastic properties of the two
contact surfaces. Since the subjects wore the sam e run-
ning shoes on each occasion, 50% of the effect of sur-
face interactions was rem oved. The coef® cient of
restitution (
e
) of the treadm ill and the road was deter-
mined using a basketball (to most closely re¯ ect the
characteristics of running shoe rubber and in-sole air
bags). The ball-treadm ill (
e
= 0.62) and b all-road
(
e
= 0.64) interactions were similar and suggest the
results of our study were not affected by differences in
the elastic properties of the surfaces. For athletes who
are fully habituated to treadm ill running, as in the pre-
sent study, any effect on energy costs is likely therefore
to b e associated with the effects of air resistance. That
our data showed an effect propor tional to running
velocity supports our view that the causal factor fo r the
difference between indoor and outdoor running is the
extra work required to m ove through the air rather than
the result of m echanical factors.
Treadmill g rade and the energetic cost of outdoor running
325
In conclusion, this study is the ® rst to dem onstrate
no signi® cant difference between the
Ç
V
O
2
measured at
velocities between 2.92 and 5.0 m s
-1
during outdoor
running and the
Ç
V
O
2
measured at the sam e velocity
during indoor treadm ill running at 1% grade over a
period of around 5 m in. T his indicates that accurate
extrapolation of data generated by physiological assess-
ment of trained runners to conditions of outdoor road
running requires that the treadm ill b e set at a grade of
1%. Th e difference in the oxygen cost between outdoor
running and treadm ill running at 0% grade is relatively
sm all, rising from ~ 1.5 ml kg
-1
min
-1
at 2.92 m s
-1
to
~ 3 m l kg
-1
min
-1
at 5.0 m s
-1
. However, this differ-
ence equates to a difference between treadmill and out-
door running velocity of about 0.07 m s
-1
at the slowest
velocity, rising to about 0.28 m s
-1
at the highest velo-
city. E xpressed in terms of heart rate, the difference is
up to 8 beats m in
-1
. Such differences are of meaningful
magnitude in the prescription of athletic training pro-
gram mes. The use of appropriate treadmill grades to
com pensate for the lack of air resistance in the labor-
atory setting will help to im prove the precision with
which ap plications to outdoor training and competition
can be made.
Acknowledgem ents
The assistance of Rober t Harley, David Jam es and Danny
Wood in collecting the data is gratefully acknowledged.
References
American College of Sports M edicine (1991). G uideline s for
G raded Exercise Testing and Prescription. Philadelphia, PA:
Lea and Febiger.
Barstow, T. and Mole, P. (1991). Linear and non-linear char-
acteristics of oxygen uptake kinetics during heavy exercise.
Jou r nal of Applied Physiology, 71, 2099-2106.
Bassett, D., Giese, M ., Nagle, F., Ward, A., Raab, D. and
Balke, B. (1985). Aerobic requirements of overground ver-
sus treadm ill running. M edicine and Science in Sports and
Exercise, 17, 477-481.
Bransford, D. and Howley, R. (1977). Oxygen cost of run-
ning in trained and untrained men and women. M edicine
and Science in Sports and Exercise, 9 , 41-44.
Conley, D. and Krahenbuhl, G. (1980). Running economy
and distance running performance of highly trained ath-
letes. M edicine and S cience in S ports and Exercise, 12,
357-360.
Costill, D., Thomason, H. and Roberts, E. (1973). Fractional
utilisation of the aerobic capacity during distance running.
M edicine and Science in Sports and Exercise, 5, 248-252.
Daniels, J. and Daniels, N. (1992). Running economy of elite
male and elite female runners. M edicine and Science in
Sports and Exercise, 2 4, 483-489.
Daniels, J., Scardina, N., Hayes, J. and Foley, P. (1986). Elite
and subelite fem ale middle- and long-distance runners. In
Sp ort Elite Perfor m ers (edited by D. Landers), pp. 57-72.
Champaign, IL: Human Kinetics.
Davies, C .T.M . (1980). Effects of wind assistance and resist-
ance on the forward m otion of a runner. Jour nal of Applied
Physiolog y, 48, 702-709.
Davies, C.T.M. (1981). W ind resistance and assistance in
running. M edicine and Science in Sports and Exercise, 13,
199-213.
Durnin, J. and Womersley, J. (1974). Body fat assessed from
total body density and its estimation from skinfold thick-
ness: M easurements on 481 men and women aged from
16 to 72 years. B r itish Jou r nal of N utritio n, 32, 77-97.
Falls, D. and Humphreys, L. (1976). Energy cost of walking
and running in young women. M edicine and Science in
Sp orts and Exercis e, 8, 9-13.
Hale, T., Armstrong, N., Hardman, A., Jakeman, P., Sharp,
C. and Winter, E. (1988). Position State m ent on the Physio-
logical Assessm ent of the Elite Com petito r. Leeds: British
Association of Sports Sciences.
Heck, H., Mader, A., Hess, G., Mucke, S., M uller, R. and
Hollman, W. (1985). Justi® cation of the 4 m
M
lactate
threshold. Inter national Jour na l of Sp or ts Medicin e, 6,
117-130.
Jones, A.M . and D oust, J. (1995). The relationship of lactate
minimum velocity to 8 km running performance and com-
parison with other methods of determining `aerobic
threshold . Jour nal of S ports Science s, 13, 34.
Leger, L. and M ercier, D. (1984). Gross energy cost of hori-
zontal treadmill and track running. S ports Medicine, 1 ,
270-277.
McM iken, D. and D aniels, J. (1976). Aerobic requirements
and maximal aerobic power in treadmill and track run-
ning. Medicine and Science in Sp or ts and Exercise, 8,
14-17.
Morgan, D. and Craib, M . (1992). Physiological aspects of
running economy. Medicine a nd Science in Sports and Exer-
cise, 24, 456-461.
Morgan, D., Baldini, F.D., Martin, P.E. and Kohrt, W.M .
(1989). Ten kilometer performance and predicted velocity
at
Ç
VO
2
max among well-trained m ale runners. M edicine
and Science in Sports and E xercise, 21, 78-83.
Morgan, D., Craib, M., Krahenbuhl, G., Woodall, K.,
Jordan, S., Filarski, K., Burleson, C. and W illiams, T.
(1994). Daily variability in running economy among well-
trained male and female distance runners. Research Quar-
ter ly, 65, 72-77.
Noakes, T., Myburgh, K. and Schall, R. (1990). Peak tread-
mill velocity during the
Ç
VO
2
max test predicts running
perform ance. Jour na l of Sp orts Science s, 8, 35-45.
Powers, S., Dodd, S., Deason, R., Byrd, R. and McKnight, T.
(1983). Ventilatory threshold, running economy and dis-
tance running performance of trained athletes. Research
Q uarterly for Spor ts and E xercise, 54, 179-182.
326
Jones and Doust
Pu gh, L.G.C.E. (1970). Oxygen uptake in track and tread-
mill running with obser vations on the effect of air resist-
ance. Jour nal of Physiology, 207, 823-835.
Pu gh, L.G.C.E. (1971). T he in¯ uence of wind resistance in
running and walking and the m echanical ef® ciency of
work against horizontal or ver tical forces. Jour na l of Physi-
olog y, 213, 225-276.
Ramsbottom, R., Brewer, J. and William s, C. (1988). A pro-
gressive shuttle run test to estimate m axim al oxygen
uptake. B ritish Jour nal of Sports M edicine, 22, 141-144.
Riggs, C.E., Johnson, D.J., Konopka, B.J. and Kilgour, R.D.
(1981). Exercise hear t rate response to facial cooling.
European Jour nal of Applied P hysiology; 47 , 323-330.
Shephard, R. (1969). A nomogram to calculate the oxygen
cost of running at slow speeds. Jour nal of Sports M edicin e
and Physical F itness, 9, 10-16.
Tetgbur, U., Busse, M . and Braum ann, K. (1993). Estim a-
tion of individual equilibrium between lactate production
and catabolism during exercise. M edicine a nd Science in
Sp orts and Exercis e, 25, 620-627.
Van Ingen Schenau, G.J. (1980). Some fundam ental aspects
of the biomechanics of overground versus treadmill loco-
motion. M edicine and Science in Sports and Exercise, 12,
257-261.
Weltm an, A., Snead, D., Steim, P., Schurrer, R., Rutt, R. and
Weltmann, J. (1990). Reliability and validity of a continu-
ous incremental treadmill protocol for the determ ination
of lactate threshold, ® xed blood lactate concentrations and
Ç
VO
2
max. International Jour nal of Sports M edicine, 11,
26-32.
W illiams, P.A. and Kilgour, R.D. (1993). Cardiovascular res-
ponses to facial cooling during low and moderate intensity
exercise. European Jou r nal of A pplied Physiology, 67,
53-58.
Treadmill g rade and the energetic cost of outdoor running
327
... Historically, most previous studies have been performed on a treadmill. Due to the fact that an ideal treadmill (with perfect speed regulation) is an inertial frame, treadmill locomotion has the potential to be mechanically and energetically almost identical to overground locomotion, especially for lower speeds [10,13,33]. However, treadmill locomotion is different from overground locomotion in three important aspects. ...
Article
This study aimed to determine the net energy expenditure (EENET) required for overground walking and running 1200 m in a sample of healthy adolescent boys and girls. A secondary purpose was to describe the effect of body composition on energy expenditure (EE) of walking versus running. Twenty healthy adolescents (9 boys, 11 girls) aged 15.85 ± 2.80 years performed 2 field tests in regular outdoor conditions: overground walking (1.64 ± 0.17 m/s) and submaximal running (3.13 ± 0.42 m/s), at a self-selected steady pace. EE was measured via indirect calorimetry. Paired sample t-tests were used to determine if there were differences between walking and running conditions and mean percentage differences were estimated for various physiological parameters. Differences in EENET between conditions were performed for both genders using a two (condition) by two (gender) analysis of variance repeated measures design, with fat free mass as a covariate. Speed increased by 90.43% between the 2 conditions, while the different components of EE increased by almost 20%. Running elicited a significantly greater EENET than walking for both genders; however, boys’ and girls’ EE did not differ significantly. When EENET was adjusted for fat free mass, there was a statistically significant condition × fat free mass effect. The findings in this study indicate that both adolescent boys and girls expend more energy during running than walking, without being affected by body composition. Body mass and fat free mass significantly correlated with EE only during running. In addition, the trained participants of the study optimized locomotion to minimize EE.
... After the warmup period, the running velocity was increased by 1 km/h every minute until voluntary exhaustion [32]. The entire CPET lasted for 8-12 min and the treadmill grade was set at 1% to simulate the energy cost of outdoor running [33]. ...
Article
Full-text available
Background: Muscle soreness after a competition or a training session has been a concern of runners due to its harmful effect on performance. It is not known if stronger individuals present a lower level of muscle soreness after a strenuous physical effort. The aim of this study was to investigate whether the pre-race muscle strength or the V˙O2max level can predict muscle soreness 24, 48 and 72 h after a full marathon in men. Methods: Thirty-one marathon runners participated in this study (age, 40.8 ± 8.8 years old; weight, 74.3 ± 10.4 kg; height, 174.2 ± 7.6 cm; maximum oxygen uptake, V˙O2max, 57.7 ± 6.8 mL/kg/min). The isokinetic strength test for thigh muscles and the V˙O2max level was performed 15-30 days before the marathon and the participants were evaluated for the subjective feeling of soreness before, 24, 48 and 72 h after the marathon. Results: The participants presented more pain 24 h after the race (median = 3, IQR = 1) than before it (median = 0, IQR = 0) (p < 0.001), and the strength values for the knee extensor muscles were significantly associated with muscle soreness assessed 24 h after the race (p = 0.028), but not 48 (p = 0.990) or 72 h (p = 0.416) after the race. The V˙O2max level was not associated with the muscle pain level at any moment after the marathon. Conclusions: Marathon runners who presented higher muscular strength for the knee extensor muscles presented lower muscle soreness 24 h after the race, but not after 48 h or 72 h after the race. Therefore, the muscle soreness level 3 days after a marathon race does not depend on muscle strength.
... Participants were fitted with a heart rate (HR) belt (Polar H7, Kempele, Finland) and facemask for breath-by-breath sampling of pulmonary gas exchange (Metamax 3B, Cortex GMbH, Leipzig, Germany). The GXT was completed on a motorized treadmill (4Front, Woodway, Waukesha, USA) inclined at 1% (23). A custom protocol was used, which implemented ramp increments of 0.22 m.s -1 .min ...
... Each VO 2 was converted to the equivalent treadmill speed with the treadmill grade fixed at 1% using ACSM metabolic equations for all intensities and trials. A 1% grade was used because this has been shown to be most similar in energetic cost to outdoor running (16). During these trials, the subjects' breath-by-breath pulmonary gas exchange data were collected continuously to assess VO 2 and RER data. ...
Article
Full-text available
Fiddler R, Fiddler J, Ashton N, Lind E. The Effects of Beetroot Juice Supplementation on Oxygen Cost of Vigorous Intensity Aerobic Exercise in Trained Endurance Athletes. JEPonline 2018; 21(6):134-142. Nitric oxide (NO) regulates many physiological processes critical to exercise performance. Beetroot juice (BRJ) supplementation has been shown to increase plasma nitrate, increasing NO availability. The present study investigated the effects of BRJ supplementation during vigorous intensity aerobic exercise. Using a repeated measures crossover design with a 1-wk washout period, 11 Division III collegiate distance runners (mean ± SD: age = 20.3 ± 1.1 yr; VO 2 peak = 55.5 ± 8.1 mL·kg-1 ·min-1) consumed 120 mL·d-1 of BRJ or placebo (PL) for 4 consecutive days. On day 5 the subjects completed 5 min of treadmill running at 65%, 85%, and 100% of volume of oxygen uptake reserve (VO 2 R) separated by 2 min each. There were no statistically significant interactions between treatment and exercise intensity for VO 2 , HR, RER, and RPE. The main effect of treatment was statistically significant for VO 2 , where BRJ (2.43 ± 0.18 L·min-1) was lower compared to PL (2.49 ± 0.17 L·min-1), P=0.029. The main effect of treatment was not statistically significant for HR, P = 0.490; RER, P=0.462; or RPE, P=0.471. These results suggest that a 4-day treatment of 120 mL·d-1 of BRJ reduces VO 2 during vigorous intensity aerobic exercise in well-trained endurance athletes.
Article
Greater understanding of differences in technique between runners may allow more beneficial feedback related to improving performance and decreasing injury risk. The purpose of this study was to develop and test a support vector machine classifier, which could automatically differentiate running technique between experienced and novice participants using only wearable sensor data. Three-dimensional linear accelerations and angular velocities were collected from six wearable sensors secured to current common smart device locations. Cross-validation was used to test the classification accuracy of models trained with a variety of combinations of sensor locations, with participants running at different speeds. Average classification accuracies ranged from 71.3% to 98.4% across the sensor combinations and running speeds tested. Models trained with only a single sensor location still showed effective classification. With the models trained with only upper arm data achieving an average accuracy of 96.4% across all tested running speeds. A post-hoc comparison of biomechanical variables between the two subgroups showed significant differences in upper body biomechanics throughout the stride. Both the methodology used to perform the classifications and the biomechanical differences identified could prove useful when aiming to shift a novice runner's technique towards movement patterns more akin to those with greater experience.