Effects of shoe type and shoe–pedal interface on the metabolic cost of bicycling

Abstract

Cyclists, coaches, and equipment manufacturers claim that cycling-specific shoes coupled with clipless pedals are ‘more efficient’. However, scientific evidence supporting or refuting these claims is lacking. We measured the metabolic cost of cycling at sub-maximal power outputs and tested the null hypothesis that there would be no differences between three different shoe-pedal combinations. Eleven healthy subjects participated (six males, four females, age 24.9 ± 6.84 yr, mass 69.98 ± 9.37 kg). We compared: (1) Nike Free 3.0 running shoes with flat rubber pedals, (2) Nike Free 3.0 running shoes with classic aluminium quill pedals, toe clips, and straps, and (3) the cyclists' own rigid-soled, cleated cycling shoes with corresponding clipless pedals. With the three different shoe-pedal conditions in random order, subjects completed three sequential 5-min cycling trials (50, 100, and 150W all at 90RPM) on a custom pan-loaded cycle ergometer equipped with a standard Monark flywheel. Subjects remained seated with both hands placed on the tops of the ergometer's racing-style handlebars. A 5-min rest period separated each of the three sets of trials. We analysed each subject's expired gases and from the rates of oxygen consumption and carbon dioxide production, we calculated metabolic power in watts (W). As hypothesized, there were no significant differences (p > 0.57) in the metabolic power consumed for pedaling at 50,100, and 150 W: Nike Free and flat pedals: 445.7, 619.0, and 817.9 W; Nike Free and quill pedals with toe clips: 428.7,600.7, and 818.0 W and cycling shoes with clipless pedals: 441.6, 612.3, and 806.4 W, respectively. Though cycling shoes may have comfort or safety benefits, they do not enhance efficiency.

Full text

Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=tfws20
Download by: [Rodger Kram] Date: 09 July 2016, At: 09:41
Footwear Science
ISSN: 1942-4280 (Print) 1942-4299 (Online) Journal homepage: http://www.tandfonline.com/loi/tfws20
Effects of shoe type and shoe–pedal interface on
the metabolic cost of bicycling
Asher H. Straw & Rodger Kram
To cite this article: Asher H. Straw & Rodger Kram (2016) Effects of shoe type and
shoe–pedal interface on the metabolic cost of bicycling, Footwear Science, 8:1, 19-22, DOI:
10.1080/19424280.2016.1140817
To link to this article: http://dx.doi.org/10.1080/19424280.2016.1140817
Published online: 08 Jul 2016.
Submit your article to this journal
View related articles
View Crossmark data

Effects of shoe type and shoepedal interface on the metabolic cost of bicycling
Asher H. Straw*and Rodger Kram
Department of Integrative Physiology, University of Colorado, Boulder, CO USA
(Received 8 September 2015; accepted 7 January 2016)
Cyclists, coaches, and equipment manufacturers claim that cycling-specific shoes coupled with clipless pedals are ‘more
efficient’. However, scientific evidence supporting or refuting these claims is lacking. We measured the metabolic cost of
cycling at sub-maximal power outputs and tested the null hypothesis that there would be no differences between three
different shoe-pedal combinations. Eleven healthy subjects participated (six males, four females, age 24.9 §6.84 yr, mass
69.98 §9.37 kg). We compared: (1) Nike Free 3.0 running shoes with flat rubber pedals, (2) Nike Free 3.0 running shoes
with classic aluminium quill pedals, toe clips, and straps, and (3) the cyclists’ own rigid-soled, cleated cycling shoes with
corresponding clipless pedals. With the three different shoe-pedal conditions in random order, subjects completed three
sequential 5-min cycling trials (50, 100, and 150W all at 90RPM) on a custom pan-loaded cycle ergometer equipped with a
standard Monark flywheel. Subjects remained seated with both hands placed on the tops of the ergometer’s racing-style
handlebars. A 5-min rest period separated each of the three sets of trials. We analysed each subject’s expired gases and
from the rates of oxygen consumption and carbon dioxide production, we calculated metabolic power in watts (W). As
hypothesized, there were no significant differences (p>0.57) in the metabolic power consumed for pedaling at 50,100,
and 150 W: Nike Free and flat pedals: 445.7, 619.0, and 817.9 W; Nike Free and quill pedals with toe clips: 428.7,600.7,
and 818.0 W and cycling shoes with clipless pedals: 441.6, 612.3, and 806.4 W, respectively. Though cycling shoes may
have comfort or safety benefits, they do not enhance efficiency.
Keywords: economy; efficiency; cycling; footwear; pedal
Introduction
Inventors, racers, cycling enthusiasts, and manufacturers
have long claimed that rigid-soled cycling shoes and
shoe-pedal attachments allow riders to both pull up during
the upstroke of the crank cycle and apply power more
directly to the pedals during the downstroke. Some claim
that these factors add ‘as much as 30 percent to …pedal-
ing efficiency’ (Sloane, 1995). Skeptical of such claims,
we investigated the effects of shoe type and shoepedal
interfaces on the metabolic cost of sub-maximal cycling.
Bicycles, pedals, and cycling shoes have a colourful
co-evolutionary history. In 1867, Pierre Michaux invented
the first bicycle with rotary cranks and pedals (Herlihy,
2004). This innovation greatly enhanced cycling effi-
ciency compared to velocipedes which were propelled by
the feet in contact with the ground (Minetti, Pinkerton, &
Zamparo, 2001). In 1890, Rankin invented the first toe
clip, which prevented the rider’s typically leather soled
shoes from sliding forward off the pedal (Rankin, 1890).
The makers of the Ramsey swinging pedal ca. 1898,
essentially an integrated toe clip pedal, even offered a full
monetary refund if the pedals ‘did not enable you to
ascend hills with 25% less energy’ (Hadland & Lessing,
2014). Numerous clipless pedal designs also proliferated
briefly, including twist-in (Hanson, 1895), magnetic
(Tudor, 1896), and even suction cup attachments (Harris,
1896). However, leather or wood soled cycling shoes with
slotted cleats, quill pedals, and toe clips prevailed and
dominated competitive cycling for most of the twentieth
century.
In the modern era, Cinelli reincarnated clipless pedal
designs in 1970 with their M71 system (Hadland & Less-
ing, 2014) and the first nylon-soled cycling shoes emerged
in 1979 (www.ciclista-america.com/about_sidi). Look
introduced their clipless pedal system in 1983 and it
quickly supplanted all previous shoe-pedal interfaces
(Bernard & Mercier, 1987). Bont developed the first
cycling shoes with carbon fibre soles in 1989 (www.bont.
com/items/aboutus/) and functionally similar carbon fiber
soles and clipless pedals from various manufacturers are
now commonplace.
Has this evolution in cycling shoe and shoepedal
interface technology spanning three centuries actually
improved the efficiency of cycling? Electromyographic
(EMG) studies have been contradictory. Jorge and Hull
(1986) reported that at a power output of 100 W, wearing
rigid-soled, cleated shoes with toe clips required less leg
*Corresponding author. Email: asher.straw@colorado.edu
Ó2016 Informa UK Limited, trading as Taylor & Francis Group
Footwear Science, 2016
Vol. 8, No. 1, 1922, http://dx.doi.org/10.1080/19424280.2016.1140817
Downloaded by [Rodger Kram] at 09:41 09 July 2016

muscle activity than soft-soled shoes. Subsequently, Cruz
and Bankoff (2001) compared cycling using tennis shoes
with quill pedals and toe clips vs. clipless pedals and
found less EMG activity in the hamstrings and gastrocne-
mius muscles with clipless pedals. Mornieux, Stapelfeldt,
Gollhofer, and Belli (2008) quantified EMG and pedal
forces for recreational and professional riders using flat
pedals (without toe clips) and clipless pedals, focusing on
the upstroke. In contrast to the previous studies, they con-
cluded that the shoe-pedal interface did not influence
EMG or pedaling mechanics during submaximal cycling.
Other investigations have focused on pedaling
mechanics. Lafortune’s 1978 MS thesis compared the
effect of different shoe-pedal combinations (without toe
clips) on pedal force effectiveness (ratio of force applied
perpendicular to crank/resultant force). Rubber-soled and
leather-soled shoes had similar effectiveness values
(within 5%). Davis (1981) reported greater force effective-
ness values for rigid-soled cycling shoes and cleats with
toe clips compared to tennis shoes. However, Korff,
Romer, Mayhew, and Martin (2007) demonstrated that
although different pedaling techniques alter force effec-
tiveness, there are no corresponding metabolic differences.
Further, we have been unable to locate any scientific
support for the notion that shoe-pedal interface affects
metabolic efficiency during cycling. Mornieux et al.
(2008) compared tennis shoes on flat pedals without toe
clips to cycling shoes with clipless pedals in both non-
cyclists and elite racers. They found no significant effects
of shoe-pedal type on the rates of oxygen consumption.
Similarly, Ostler, Betts, and Gore (2008) compared sub-
maximal rates of oxygen consumption for cycling with
tennis shoes on both flat pedals and classic quill-style rac-
ing pedals equipped with toe clips and straps. They found
no significant differences (p>0.39) for power outputs
ranging from 60 to 240 W, all at a cadence of 60RPM.
Given the lack of consensus in the scientific literature
as well as the lack of research regarding cycling shoes in
general, we tested the null hypothesis that there would be
no differences between different shoe and shoe-pedal
interface combinations in terms of the metabolic cost of
cycling at sub-maximal power outputs. We compared two
extreme shoe types highly flexible running shoes and
carbon-soled cycling shoes paired with three shoe-pedal
interfaces. Our experimental design allowed us to con-
sider two factors: the longitudinal bending stiffness of the
shoe sole and the presence/lack of shoe-pedal attachment.
Methods
Ten healthy, injury-free, experienced competitive and/or
recreational cyclists (six males, four females, age 24.9 §
6.84 yr, mass 69.98 §9.37 kg) participated after provid-
ing written informed consent as per the University of Col-
orado Boulder Institutional Review Board. In addition to
age (1845) yrs and good health, one important inclusion
criteria was self-reported cycling a minimum of 80 km or
3 hours per week. Participants reported riding an average
of 297.7 §79.7 km/week.
We asked the participants to fast for at least two hours
prior to testing. Warm-up was ad libitum and generally
involved light pedaling and stretching. For each condition,
participants completed sets of three, 5-min cycling trials
(50, 100, and 150 W all at 90 RPM). They rode a custom,
pan-loaded cycle ergometer (Nobilette, Longmont CO)
equipped with a standard Monark flywheel (9.53 kg,
0.51 m radius) using three different shoe and shoepedal
interface combinations (nine total trials). Subjects
remained seated with both hands on the tops of the ergo-
meter’s racing handlebars. We instructed them to maintain
a cadence of 90 RPM given visual feedback via a digital
cadence meter mounted on the handlebars. A 5-minute
rest period separated each of the three sets of trials and
allowed for changing shoes and pedal types. All trials
were completed during a single experimental session. We
randomized the order of the footwear-pedal conditions.
We compared three conditions: (1) Nike Free 3.0 run-
ning shoes with flat rubber pedals, (2) Nike Free 3.0 run-
ning shoes with classic aluminium quill pedals, toe clips,
and straps, and (3) the cyclists’ own rigid-soled, cleated
cycling shoes and corresponding clipless pedals (e.g.
Speedplay, Look, Shimano) (Figure 1). We chose the
Figure 1. Test pedals used in the study. Left: flat rubber pedal. Middle: quill pedal with toe clip and strap. Right: Speedplay clipless
pedal
20 A.H. Straw and R. Kram
Downloaded by [Rodger Kram] at 09:41 09 July 2016

Nike Free 3.0 running shoes because with their extreme
flexibility and cushioned midsole, they constitute the
opposite of rigid-soled cycling shoes.
We collected each participants’ expired gases and cal-
culated the STPD rates of oxygen consumption (V
̇
O2)
and carbon dioxide production (V
̇
CO2) using an open-
circuit expired-gas analysis system (TrueOne 2400;
ParvoMedics, Sandy, UT). Before each experiment, we
calibrated the gas analysers and pneumotach using refer-
ence gases and a calibrated 3-L syringe, respectively. We
averaged V
̇
O2, V
̇
CO2, and respiratory exchange ratio
(RER) for the last 2 minutes of each trial. We planned to
exclude any participants whose RER values exceeded 1.0,
but all values remained below 1.0. From the V
̇
O2 and
V
̇
CO2 measurements, we calculated metabolic power
using the Brockway equation (Brockway, 1987). We
report absolute rates of metabolic power (W), oxygen
consumption (L O2/min), and RER. We calculated delta
efficiency for each subject for the three shoe-pedal condi-
tions from the slope of the regression of mechanical
power vs. metabolic power, from 50150 W.
Frederick (1983) determined it possible to detect oxy-
gen consumption differences of 1.6% with a sample size
of 10 subjects assuming a C.V of 2.5%. In our previous
research, we have resolved smaller than 2% differences
between footwear conditions using a sample size of 14
(Franz, Wierzbinski, & Kram, 2012). Anticipating differ-
ences larger than 2%, we recruited and tested 11 subjects,
but had to discard the data for one subject due to an equip-
ment malfunction. Using R software (www.rstudio.com),
we ran two-way repeated measures analysis of variance
(ANOVAs) (3 shoe-pedal conditions £3 power condi-
tions) for metabolic power, oxygen consumption, and
RER. We also ran a one-way repeated measures ANOVA
for delta efficiency across shoe-pedal conditions. We set
statistical significance at p<0.05. All values are reported
as means §SE unless noted otherwise.
Results
Across power output, there were no significant main
effects of shoe-pedal condition on metabolic power con-
sumption (pD0.57), oxygen consumption (pD0.48), or
RER (pD0.36) (Table 1). As expected, there was a sig-
nificant main effect of mechanical power on metabolic
power consumption (pD2£10
¡16
). The interaction
effect between shoe-pedal conditions and mechanical
power was not significant (pD0.66). Delta efficiency
also was not significantly different for the three shoe-
pedal conditions (pD0.06).
Discussion
We found that when healthy, active male and female
cyclists pedal at mechanical power outputs ranging from
50 to 150 W on a cycle ergometer, shoepedal interface
conditions had no effect on metabolic cost. Thus, we retain
our null hypothesis. Our experimental design of three
shoepedal conditions allows us to conclude that neither
longitudinal bending stiffness of the shoe sole nor shoe-
pedal attachment affect the metabolic cost of cycling.
Our findings are consistent with Ostler et al. (2008)
who found no difference in V
̇
O2 when comparing flat
pedals to toe clip pedals both using ‘gym shoes’. Although
our study design differed slightly from Ostler et al. (2008)
(90 vs. 60 RPM and 50150 W vs. 60240 W), our
results were similar.
Some may question the competitive relevance of our
study given the modest power outputs we evaluated. How-
ever, we found no interaction effect of mechanical power
Table 1. Rates of oxygen consumption (V
̇
O
2
) in L/min, respiratory exchange ratio (RER), metabolic power output (P
met
) in watts and
delta efficiency (%) (SE). There were no main effects across shoe-pedal conditions for any of these variables. As expected, at greater
mechanical power outputs, all variables increased significantly.
Mechanical
power
Nike Free and
flat pedals
Nike Free and
quill pedals
w/toe clips
Cycling shoes and
clipless pedals
50 W V
̇
O
2
1.32 (0.05) 1.28 (0.04) 1.31 (0.04)
P
met
RER
445.73 (12.15)
0.78 (0.01)
428.72 (12.72)
0.79 (0.02)
441.68 (14.98)
0.77 (0.02)
100 W V
̇
O
2
1.82 (0.04) 1.80 (0.04) 1.81 (0.03)
P
met
RER
619.00 (14.99)
0.81 (0.01)
600.79 (13.78)
0.82 (0.01)
612.31 (13.77)
0.81 (0.01)
150 W V
̇
O
2
2.40 (0.04) 2.40 (0.05) 2.33 (0.05)
P
met
RER
817.94 (17.11)
0.84 (0.01)
818.00 (15.40)
0.85 (0.01)
806.48 (13.62)
0.84 (0.01)
Delta Efficiency 27.09 (0.79) 25.92 (0.78) 27.58 (0.68)
Footwear Science 21
Downloaded by [Rodger Kram] at 09:41 09 July 2016

output on the differences in metabolic power between
shoepedal conditions. Nonetheless, studies of higher
calibre competitive cyclists at greater power outputs, dur-
ing accelerations, and/or uphill cycling using different
shoe-pedal conditions could be worthwhile. Given the
anecdotal reports of cyclists that rigid-soled cycling shoes
reduce fatigue, comparisons between longer duration
trials may also be warranted.
Our data suggest that in sprint triathlons, running
shoes with toe clip pedals or even flat pedals may be
advantageous compared to cycling-specific shoes because
they could facilitate a quicker bike-run transition without
any energetic sacrifice. It is possible that, in cyclocross
races with long running portions, using running shoes
may enhance running performance without any metabolic
disadvantage during the cycling portions.
We recognize that many riders simply prefer cycling
shoes and clipless pedals for comfort and safety but our
data refute the claims of improved efficiency via rigid-
soled cycling shoes with clipless pedals. In summary, we
found that different shoepedal combinations had no sig-
nificant effect on the metabolic cost of cycling.
Disclosure statement
RK is a consultant to Nike Inc. and Fi’zi:k, a manufacturer of
cycling components and shoes.
References
Bernard, J., Mercier, M., (1987). U.S. patent No. 4,686,867.
Washington, DC: U.S. Patent and Trademark Office.
Brockway, J.M. (1987). Derivation of formulae used to calculate
energy expenditure in man. Human Nutrition Clinical Nutri-
tion, 41, 463471.
Cruz, C.F., & Bankoff, A.D. (2001). Electromyography in
cycling: Difference between clipless pedal and toe clip
pedal. Electromyography Clinical Neurophysiology, 41,
247252.
Davis, R.R. (1981). Measurement of pedal loading in bicycling:
II. Analysis and results. Journal of Biomechanics, 14,
857872.
Franz, J.R., Wierzbinski, C.M., & Kram, R. (2012). Metabolic
cost of running barefoot versus shod: is lighter better? Medi-
cine and Science in Sports and Exercise, 44, 15191525.
Hadland, T., & Lessing, H-E. (2014). Bicycle design : An illus-
trated history. Cambridge, MA: MIT Press.
Hanson, C.M. (1895). U.S. patent No. 550,409. Washington, DC:
U.S. Patent and Trademark Office.
Harris, E.T. (1896). U.S. patent No. 554,686. Washington, DC:
U.S. Patent and Trademark Office.
Herlihy, D.V. (2004). Bicycle : The history. New Haven, CT:
Yale University Press.
Korff, T., Romer, L.M., Mayhew, I., & Martin, J.C. (2007).
Effect of pedaling technique on mechanical effectiveness
and efficiency in cyclists. Medicine and Science in Sports
and Exercise, 39, 991995.
Lafortune, M.A. (1978). A biomechanical analysis of cycling
under various shoe-pedal interfaces (Unpublished master’s
thesis). Pennsylvania State University., Pennsylvania, PA.
Minetti, A.E., Pinkerton, J., & Zamparo, P. 2001. From bipedal-
ism to bicyclism: Evolution in energetics and biomechanics
of historic bicycles. Proceedings of the Royal Society B:
Biological Sciences, 268, 13511360.
Mornieux, G., Stapelfeldt, B., Gollhofer, A., & Belli A. (2008).
Effects of pedal type and pull-up action during cycling.
International Journal of Sports Medicine, 29, 817822.
Ostler, L.M., Betts, J.A., & Gore, C.J. (2008). Gross cycling effi-
ciency is not altered with and without toe-clips. Journal of
Sports Science, 26,4755.
Rankin, W.G. (1890). U.S. patent No. 425,697. Washington, DC:
U.S. Patent and Trademark Office.
Sloane, E.A. (1995). The complete book of bicycling: 25th Anni-
versery Ed. New York, NY: Fireside.
Tudor, H. (1896). U.S. patent No. 588,038. Washington, DC:
U.S. Patent and Trademark Office.
22 A.H. Straw and R. Kram
Downloaded by [Rodger Kram] at 09:41 09 July 2016