Energetic Cost of Running in Track and Treadmill

Carlo M. Biancardi

, Leonardo Lagos-Hausheer

1,2 b

, Germán Pequera

1,3 c

, Enzo Castroman

Federico Cazot

, Enzo Martinez

and Renata L. Bona

LIBiAM, Department of Biological Sciences, CENUR Litoral Norte, Universidad de la República, Paysandú, Uruguay

Movement Physiology Research Laboratory, Department of Kinesiology, Faculty of Medicine, University of Concepcion,

Concepcion, Chile

Ingeniería Biológica, CENUR Litoral Norte, Universidad de la República, Paysandú, Uruguay

Keywords: Metabolic Power, Running Economy, Step Frequency, GPS, Auditory Feedback.

Abstract: The metabolic power and cost of running per unit distance on a track have been estimated and compared with

data collected indoor, in a laboratory on a treadmill. Oxygen uptake have been collected using a portable

device, while speed was regulated by auditory feedback (metronome) and verified using GPS. Speed

fluctuations remained within an acceptable range. Metabolic power increased linearly with speed, with a slope

significantly lower on the track than on the treadmill (p = 0.017). However, statistical comparisons at the

same speed did not yield significant differences between the two conditions. The average cost of transport

was slightly, but not significantly, lower on the track (4.20 J/kg/m) than on the treadmill (4.35 J/kg/m), and it

remained nearly independent of speed over a wide range. Nevertheless, in the lower and higher speed ranges

on the track, the cost of transport tended to increase. A similar non-linear trend was observed in the cost of

transport in relation to step frequency, with the minimum values falling within a range of 160 to 180 steps per

minute. These preliminary results are encouraging and warrant further research to explore the differences

between running on a treadmill and on a track.

1 INTRODUCTION

Motorized treadmills are widespread tools in research

laboratories specialized in exercise physiology and

biomechanics of locomotion. They are also widely

used on a professional level in rehabilitation and

sport, by runners, coaches and practitioners. One

question that arises pertains to the reliability of

information collected in the laboratory when it is

applied to real outdoor conditions (Jones and Doust,

1996). Lindsay et al. (2014) demonstrated that

quantifying gait parameters during treadmill running

can yield different results compared to overground

settings.

While it is not a novel topic, the energetic cost of

human running is worthy of further investigation,

https://orcid.org/0000-0002-5566-3958

https://orcid.org/0000-0003-2588-1548

https://orcid.org/0000-0002-2696-1630

https://orcid.org/0009-0000-6843-7282

https://orcid.org/0009-0009-1174-7023

https://orcid.org/0009-0000-5436-8757

https://orcid.org/0000-0003-4343-7336

especially under racing-like conditions, such as those

found on an athletic track. The metabolic power and

the metabolic cost of running per unit distance are

determined by analyzing oxygen uptake in relation to

speed. (Schmidt-Nielsen, 1972).

The speed control in track is a challenge that have

been met in different ways: with a human guide

(Jones and Doust, 1996; Tam et al., 2012), with a laser

or light guide (Minetti et al., 2013; Pind et al., 2019),

with an auditory feedback (Lagos et al., 2023).

Minetti et al. (2013) demonstrated that smooth

fluctuations of the running speed does not affect the

metabolic cost, allowing for less strict control of

speed.

The effect of air resistance, which is absent while

running “in place” on a treadmill, have been

Biancardi, C., Lagos-Hausheer, L., Pequera, G., Castroman, E., Cazot, F., Martinez, E. and Bona, R.

Energetic Cost of Running in Track and Treadmill.

DOI: 10.5220/0012202300003587

In Proceedings of the 11th International Conference on Sport Sciences Research and Technology Support (icSPORTS 2023), pages 173-178

ISBN: 978-989-758-673-6; ISSN: 2184-3201

173

addressed by Pugh (1970; 1971), that produced

equations to estimate the extra-cost to win air

resistance in track. Other researchers followed the

suggestion of Jones and Doust (1996) of adding a 1%

gradient to the treadmill trials, in order to compensate

the lack of air resistance on treadmill, but with

controversial results (Mooses et al., 2015; Pind et al.,

2019).

Initial comparisons between treadmill and track

results were conducted using Douglas bags for gas

analysis (Pugh, 1970; Basset et al., 1985; Jones and

Doust, 1996). However, the use of these cumbersome

tools adversely affected running performance and

yielded questionable results (Mooses et al., 2015).

More recently, with the availability of portable

metabolic devices, other research projects have

tackled this question by analyzing the metabolic cost

of running outdoors (Tam et al., 2012), or by

comparing the cost between treadmill and track

settings (Mooses et al., 2015; Pind et al., 2019). These

recent studies have indicated better running economy

(lower metabolic cost) on a track compared to

treadmill conditions. It is important to note that both

Mooses et al. (2015) and Pind et al. (2019) involved

endurance runners as participants. The findings of

Tam et al. (2012) were consistent, and they were also

obtained through an analysis of elite marathon

runners.

The objective of this ongoing project was to

analyse the energetic cost of running on a track using

a portable metabolic device, GPS, and auditory

feedback for speed control. The study involved a

diverse sample of athletes from various sports

disciplines and aimed to compare the results with data

collected indoors on a treadmill.

2 METHODS

A total of 56 healthy men participate in this research,

they were measured and weighed in the lab, prior to

the trials (Age 28 ± 10 years; Weight 74.9 ± 11.2 kg;

Height 177 ± 7 cm). All participant were runners with

weekly volume of 5 to 25 km, with at least 1 year of

continuous practice.

2.1 Data Collection

Part of the treadmill metabolic data used in this

research have been collected by the authors during a

period from 2016 and 2022, and published in previous

works and/or available in public repositories (Lagos

et al., 2022; Lagos et al., 2023; Pequera et al., 2020;

Pequera et al., 2023). New treadmill (3 subjects) and

track (22 subjects) metabolic data have been collected

in 2023 at the Biomechanics and Movement Analysis

Research Laboratory (LIBiAM) of the Universidad

de la República in Paysandú, (Uruguay) at a

controlled temperature of 22°C, and at the athletic

track in the Polideportivo Paysandú (certified by the

International Association of Athletics Federations), at

an average temperature of 22 ± 2 °C with almost no

wind. All procedures were in accordance with the

latest version of the Declaration of Helsinki (2013).

The study protocol was approved by the Ethics

Committee of the CENUR Litoral Norte—

Universidad de la República (Exp. #311170-000921-

19).

Metabolic data were collected breath by breath by

a wearable metabolic system (K5, Cosmed, Italy).

Reference resting values of each participant were

assessed by a first record of 5 minutes in a quiet

orthostatic position. Trials, performed according to

the protocol described below, lasted 5 minutes each,

but only the last minute of each trial, when a steady

state oxygen flow was reached, was considered for

energetic analyses.

Cosmed K5 includes an integrated GPS (position

accuracy within 2.5 m and speed accuracy within 0.1

m/s) (De Blois et al., 2021).

2.2 Treadmill Protocol

After a session of familiarization with the treadmill,

they realised between three and six running trials on

a treadmill (T2100, General Electric, USA) at

controlled constant speeds, in a range between 1.67

and 3.61 m/s. Further details in Pequera et al. (2023).

2.3 Track Protocol

The preparation was performed directly in the track

(200 m from the lab location). Data collection was

performed during mild uruguayan autumn days, with

comfortable temperatures and wind almost absent. A

10 m speed trap was positioned along the straight

stretch, were the performance was recorded with a

portable device mounted on a tripod for afterwords

control.

In the first trial, participants were asked to

perform a 5-minutes run, maintaining a constant pace

at their comfortable running speed, in the lane 6 of the

track. The step frequency maintained during the trial

was measured thrice, and the average value was

recorded as the preferred step frequency (Psf, beats

per minute) and was used to compute the rules for the

following trials.

icSPORTS 2023 - 11th International Conference on Sport Sciences Research and Technology Support

174

During the second and third trials, participants

were asked to wear a headset connected to a

smartphone. Auditive feedbacks of a metronome

performing a beat frequency were provided

(Metronome.com, click sound, ¼ time), and runners

were asked to adapt their step frequency to the sound

(one heel strike every click, minding that two steps

corresponds to one stride). During the second trials

the beat frequency was 15% slower than the Psf,

while during the third trial it was 15% faster than the

Psf. A resting period of 3-5 minutes was observed

between two trials.

2.4 Data Analysis

The net oxygen uptake was computed by subtracting

the average resting value from the average oxygen

flow rate (𝑉

)

measured during the last minute of

each trial. Respiratory quotient (RQ), the ratio

between CO

and O

flow rates, was also averaged

during the last minute of the trial, and used to convert

mlO

to Joules (di Prampero 2015). The net metabolic

power (MetP; W/kg) so obtained was divided by the

forward speed (m/s) in order to achieve the running

economy, or cost of transport (CoT; J/kg/m), the

metabolic energy needed to move one unit mass one

unit distance (Schmidt Nielsen, 1972).

When running outdoors we need to consider an

extra-cost, which would account for the energy

required to overcome the air resistance, absent during

indoor treadmill exercises (Jones and Doust, 1996).

The resulted values of oxygen uptake in track were

corrected using the equation (1) provided by Pugh

(1971), assuming a wind speed equal to the forward

speed (calm or absent wind):

Δ𝑉

= 0.00354

(1)

where Δ𝑉

is the fraction of 𝑉

necessary to win

the air resistance, in L/min; A

is the body surface

projected area, which was assumed constant at the

value of 0.436 m

(Pugh, 1970); and v is the wind

speed in m/s.

Videos collected during the track protocol were

analysed to check the forward speed and the observed

step frequency (Osf).

ANOVA and Student t-test, or the equivalent non-

parametric Kruskall-Wallis and Mann-Whitney U-

test were applied, depending on the results of

normality test. Alpha was set to 0.05, effect size was

expressed as Cohen’s d. Linear or quadratic

regression was applied to fit the data. Magnitude of

association (Pearson’s r) and coefficient of

determination (r

) were showed.

3 RESULTS AND DISCUSSION

3.1 Speed in Track

GPS speed was compared with the speed obtained by

video analyses, with a difference < 2%, within the

speed accuracy declared by Cosmed.

The participants were able to maintain an almost

constant speed during the track trials, at least after a

first part of “speed adaptation”, which occurred

during the first minute of each trial. The 5 minutes

duration of each exercise protects our data from

possible negative effects of the first adaptation

stretch, as the steady-state of oxygen flow rate during

running is attained within 2 minutes at constant speed

(Carter et al., 2000).

The energetic cost of running is not affected by

cycles of acceleration/deceleration (Minetti et al.,

2013). However, our purpose was to compare track

and treadmill under similar conditions. The speed in

track was recorded by GPS simultaneously with

physiological parameters, therefore breath-by-breath

and not at a constant rate. We computed the standard

deviation (SD) of the speed along the period used for

the energetic cost calculation. We obtained an

average SD of 0.13 m/s during the first trial, at

comfortable speed, and an average SD of 0.11 m/s

during the trials supported by auditory feedback,

which helped to maintain a constant pace (Lagos et

al., 2023).

3.2 Metabolic Power

In both conditions the MetP linearly increased with

speed (Figure 1). Both regressions lines were

statistically significants (p < 0.001). On the treadmill

the slope resulted 4.56, r = 0.87, r

= 0.75. On the

track the slope resulted 3.65, r = 0.85, r

= 0.72. Both,

magnitude of association and coefficient of

determination indicate a strong relationship (Thomas

and Nelson, 2001). The slope difference was

statistically significant (F = 5.77; p = 0.017).

The trend of the treadmill results is in agreement

with previous works, where the slope of the MetP vs.

speed relationship was > 4 (Tam et al., 2012; Kipp et

al., 2018; Pind et al., 2019), while in an analysis of

half-marathon and marathon runners the MetP

increased with a slope of about 3 with respect to speed

(Di Prampero et al., 1986). The slope of the

regression in track was slightly less than that obtained

in previous works with a similar protocol (Tam et al.,

2012; Pind et al., 2019).

Energetic Cost of Running in Track and Treadmill

175

Figure 1: Metabolic power vs. speed. Grey circles: treadmill

experiments; Black circles: track experiments. Regression

lines are displayed accordingly.

3.3 Cost of Transport

The CoT was obtained by dividing the MetP for the

forward speed. Therefore, assuming that the intercept

of MetP at rest should be nearly zero, we would

expect an almost constant (speed independent) CoT

in a range from 3.5 to 4.5 J/kg/m. The overall mean

CoT in treadmill (4.35 ± 0.55) was not significantly

different from the overall mean CoT in track (4.20 ±

0.45): U Mann Whitney = 3325, p = 0.09, d = -0.16.

When the number of observations allowed to compare

treadmill and track at the same speed, no significant

difference were found in both CoT and MetP (Table

1). Differently from our results, Mooses et al. (2015)

and Pind et al. (2019) found a significantly lower CoT

overground than on a treadmill. However, they

analysed high level endurance runners on a narrower

range of speeds.

Table 1: Results of a t-test between treadmill and track re-

sults at different speeds. Speeds in m/s; df = degrees of free-

dom; d = Cohen’s d (effect size).

Forward

speed

CoT t-test MetP t-test

2.64 (df = 5)

p = 0.173; d =1.22 p = 0.162; d =1.25

2.78 (df = 9)

p = 0.100; d =-1.12 p = 0.141; d=-0.98

3.06 (df = 20)

p = 0.998; d =-0.01 p = 0.997; d =0.01

3.19 (df = 7)

p = 0.162; d =-1.05 p = 0.197; d=-0.96

3.61 (df = 8)

p = 0.358; d =-0.62 p = 0.308; d=-0.69

The speed-independent behaviour of the cost of

transport (CoT) has been documented in various

studies (Kram and Taylor, 1990; Minetti et al., 2013;

Arellano and Kram, 2014; Pavei et al., 2015; Pequera

et al., 2023). However, there is some evidence

suggesting that the cost of running may not be entirely

independent of running speed, particularly among

elite runners and at speeds beyond the average range

(Batliner et al., 2018). Our results regarding CoT vs.

speed are summarized in Figure 2. On the treadmill,

the coefficient of determination for the linear model

was r² = 0.11. On the track, where a wider range of

speeds was achieved, a non-linear pattern appears to

emerge, with higher CoT values observed at very low

or very high forward speeds (Degree 2 polynomial: r²

= 0.24).

Figure 2: Cost of transport vs. speed. Grey circles: treadmill

experiments; Black circles: track experiments. Best fit

curves are displayed accordingly.

3.4 CoT and Step Frequency

In figure 3 the CoT was plotted versus the step

frequency (Spm: steps per minute). Both treadmill

and track distributions displayed a quadratic fit line:

treadmill r

= 0.14; track r

= 0.34. The minimum of

both trend lines corresponds to a range of step

frequencies between 160 and 180 Spm.

Figure 3: Cost of transport vs. step frequency. Grey circles:

treadmill experiments; Black circles: track experiments.

Best fit curves are displayed accordingly.

These results align with the preferred and optimal

step frequencies identified by Snyder and Farley

(2011) and Lieberman et al. (2015), which were

approximately 170 steps per minute (Spm) in both

cases. In running, variations in speed are primarily

attributed to changes in stride length rather than stride

icSPORTS 2023 - 11th International Conference on Sport Sciences Research and Technology Support

176

(or step) frequency (Cavanagh and Kram, 1989;

Lieberman et al., 2015). It is important to note that the

cost of transport (CoT) is considered one of the

determinants of the optimal step frequency, although

mechanical variables such as peak forces and torques

also play a role.

4 CONCLUSIONS

Modern equipment and technology enable us to

measure the energetic cost of locomotion using

experimental setups that closely mimic racing-like

conditions. Although our research is ongoing and not

yet finalized, our preliminary results have revealed

significant differences in the rate of metabolic power

increase with respect to speed between running on a

track and running on a treadmill. Additionally, we

observed a slightly better running economy on the

track compared to the treadmill, although the cost of

transport (CoT) did not exhibit a significant

difference. Furthermore, our findings provided

insights into an optimal range of step frequencies that

appear to minimize CoT.

ACKNOWLEDGEMENTS

We thanks the director of the sport facility that

include the athletic track, Mateo Arbiza. We would

like to thanks the anonymous reviewers for their

useful suggestions.

REFERENCES

Arellano, C. J., & Kram, R. (2014). Partitioning the Meta-

bolic Cost of Human Running: A Task-by-Task Ap-

proach. Integrative and Comparative Biology, 54(6),

1084.

Bassett Jr, D. R., Giese, M. D., Nagle, F. J., Ward, A., Raab,

D. M., & Balke, B. (1985). Aerobic requirements of

overground versus treadmill running. Medicine and

Science in Sports and Exercise, 17(4), 477-481.

Batliner, M. E., Kipp, S., Grabowski, A. M., Kram, R., &

Byrnes, W. C. (2018). Does metabolic rate increase lin-

early with running speed in all distance runners?. Sports

medicine international open, 2(01), E1-E8.

Carter, H., Jones, A. M., Barstow, T. J., Burnley, M., Wil-

liams, C. A., & Doust, J. H. (2000). Oxygen uptake ki-

netics in treadmill running and cycle ergometry: a com-

parison. Journal of applied physiology, 89(3), 899-907.

Cavanagh, P. R. and Kram, R. (1989). Stride length in dis-

tance running: velocity,body dimensions, and added

mass effects. Med. Sci. Sports Exerc. 21, 467-479.

De Blois, J. P., White, L. E., & Barreira, T. V. (2021). Re-

liability and validity of the COSMED K5 portable met-

abolic system during walking. European Journal of Ap-

plied Physiology, 121, 209-217.

Jones, A. M., & Doust, J. H. (1996). A 1% treadmill grade

most accurately reflects the energetic cost of outdoor

running. Journal of sports sciences, 14(4), 321-327.

Kipp, S., Grabowski, A. M., & Kram, R. (2018). What de-

termines the metabolic cost of human running across a

wide range of velocities? Journal of Experimental Biol-

ogy, 221(18), jeb184218.

Kram, R., & Taylor, C. R. (1990). Energetics of running: a

new perspective. Nature, 346(6281), 265-267.

Lagos-Hausheer., L, Biancardi, CM & Bona, RL. (2022).

Cardiopulmonary exercise breath-by-breath data during

locomotion at variable speed in 28 healthy young sub-

ject [Data set]. Zenodo. https://doi.org/10.5281/

zenodo.6403278

Lagos-Hausheer., L, Bona, RL & Biancardi, CM. (2023).

Effectiveness of a Variable Speed Control Based on

Auditory Feedback: Is It Possible? International Jour-

nal of Sports Physiology and Performance, (Ahead of

Print) https://doi.org/10.1123/ijspp.2022-0465

Lieberman, D. E., Warrener, A. G., Wang, J., & Castillo, E.

R. (2015). Effects of stride frequency and foot position

at landing on braking force, hip torque, impact peak

force and the metabolic cost of running in humans.

Journal of Experimental Biology, 218(21), 3406-3414.

Lindsay, T. R., Noakes, T. D., & McGregor, S. J. (2014).

Effect of treadmill versus overground running on the

structure of variability of stride timing. Perceptual and

motor skills, 118(2), 331-346.

Minetti, A. E., Gaudino, P., Seminati, E., & Cazzola, D.

(2013). The cost of transport of human running is not

affected, as in walking, by wide acceleration/decelera-

tion cycles. Journal of Applied Physiology, 114(4),

498-503.

Mooses, M., Tippi, B., Mooses, K., Durussel, J., & Mäestu,

J. (2015). Better economy in field running than on the

treadmill: evidence from high-level distance runners.

Biology of sport, 32(2), 155-159.

Pavei, G., Biancardi, C. M., & Minetti, A. E. (2015). Skip-

ping vs. running as the bipedal gait of choice in hypo-

gravity. Journal of Applied Physiology, 119(1), 93-100.

Pequera, G., Bona, R., Bonezi, A., Polero, P., Biancardi,

C.M. (2020), “Cardiopulmonary exercise breath-by-

breath data during locomotion in 6 healthy young sub-

jects”, Mendeley Data, V1, doi: 10.17632/vmwrtj29kr.1

Pequera, G., Yelós, V., & Biancardi, C. M. (2023). Reduc-

ing cost of transport in asymmetrical gaits: lessons from

unilateral skipping. European Journal of Applied Phys-

iology

, 123(3), 623-631.

Pind, R., Mooses, K., Suvi, S., Purge, P., Viru, M., Pehme,

A., Kaasik, P. & Mooses, M. (2019). Better economy

on indoor track compared to treadmill running with 1%

inclination. Research Quarterly for Exercise and Sport,

90(4), 470-478.

di Prampero, P. E. (2015). La locomozione umana su terra,

in acqua,in aria. Milano : Edi-Ermes.

Energetic Cost of Running in Track and Treadmill

177

di Prampero, P. E., Atchou, G., Brückner, J. C., & Moia, C.

(1986). The energetics of endurance running. European

journal of applied physiology and occupational

physiology, 55(3), 259-266.

Pugh, L. G. C. E. (1970). Oxygen intake in track and

treadmill running with observations on the effect of air

resistance. The Journal of physiology, 207(3), 823-835.

Pugh, L. G. E. (1971). The influence of wind resistance in

running and walking and the mechanical efficiency of

work against horizontal or vertical forces. The Journal

of physiology, 213(2), 255-276.

Schmidt-Nielsen, K. (1972). Locomotion: energy cost of

swimming, flying, and running. Science 177(4045):

222–228.

Snyder, K. L., & Farley, C. T. (2011). Energetically optimal

stride frequency in running: the effects of incline and

decline. Journal of Experimental Biology, 214(12),

2089-2095.

Tam, E., Rossi, H., Moia, C., Berardelli, C., Rosa, G.,

Capelli, C., & Ferretti, G. (2012). Energetics of running

in top-level marathon runners from Kenya. European

journal of applied physiology, 112(11), 3797-3806.

Thomas, J. R., & Nelson, J. K. (2001). Research methods

in physical activity. Champaign, IL: Human Kinetics.

icSPORTS 2023 - 11th International Conference on Sport Sciences Research and Technology Support

178