Você está na página 1de 8

This article was downloaded by: [Thibault Lussiana]

On: 15 July 2015, At: 22:05


Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place,
London, SW1P 1WG

Journal of Sports Sciences


Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/rjsp20

Reliability and validity of the Myotest® for measuring


running stride kinematics
a ab c d
Cyrille Gindre , Thibault Lussiana , Kim Hebert-Losier & Jean-Benoit Morin
a
Volodalen Compagny, Research and Development department, Chaveria, France
b
Research unit EA4660, Culture Sport Health Society and Exercise Performance Health
Innovation Platform, Franche-Comté University, Besançon, France
c
National Sports Institute of Malaysia, Kompleks Sukan Negara, Butik Jalil, Peti Surat 7102,
Kuala Lumpur, Malaysia
d
Research Unit EA4338, Laboratory of Exercise Physiology, CHU Bellevue – Medicine of Sport
and Myology, University of Lyon, Saint-Etienne, France
Published online: 15 Jul 2015.
Click for updates

To cite this article: Cyrille Gindre, Thibault Lussiana, Kim Hebert-Losier & Jean-Benoit Morin (2015): Reliability and validity
of the Myotest® for measuring running stride kinematics, Journal of Sports Sciences, DOI: 10.1080/02640414.2015.1068436

To link to this article: http://dx.doi.org/10.1080/02640414.2015.1068436

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
JOURNAL OF SPORTS SCIENCES, 2015
http://dx.doi.org/10.1080/02640414.2015.1068436

Reliability and validity of the Myotest® for measuring running stride kinematics
Cyrille Gindre1, Thibault Lussiana1,2, Kim Hebert-Losier3 and Jean-Benoit Morin4
1
Volodalen Compagny, Research and Development department, Chaveria, France; 2Research unit EA4660, Culture Sport Health Society and
Exercise Performance Health Innovation Platform, Franche-Comté University, Besançon, France; 3National Sports Institute of Malaysia, Kompleks
Sukan Negara, Butik Jalil, Peti Surat 7102, Kuala Lumpur, Malaysia; 4Research Unit EA4338, Laboratory of Exercise Physiology, CHU Bellevue –
Medicine of Sport and Myology, University of Lyon, Saint-Etienne, France

ABSTRACT ARTICLE HISTORY


Accelerometer-based systems are often used to quantify human movement. This study’s aim was to Accepted 28 June 2015
assess the reliability and validity of the Myotest® accelerometer-based system for measuring running
KEYWORDS
stride kinematics. Twenty habitual runners ran two 60 m trials at 12, 15, 18 and 21 km·h−1. Contact time, Running; biomechanics;
aerial time and step frequency parameters from six consecutive running steps of each trial were technology; concurrent
extracted using Myotest® data. Between-trial reproducibility of measures was determined by comparing validity; reliability
kinematic parameters from the two runs performed at the same speed. Myotest® measures were
compared against photocell-based (Optojump Next®) and high-frequency video data to establish
Downloaded by [Thibault Lussiana] at 22:05 15 July 2015

concurrent validity. The Myotest®-derived parameters were highly reproducible between trials at all
running speeds (intra-class correlation coefficient (ICC): 0.886 to 0.974). Compared to the photo-cell and
high-speed video-based measures, the mean contact times from the Myotest® were 34% shorter and
aerial times were 64% longer. Only step frequency was comparable between systems and demon-
strated high between-system correlation (ICC ≥ 0.857). The Myotest® is a practical portable device that is
reliable for measuring contact time, aerial time and step frequency during running. In terms of validity,
it provides accurate step frequency measures but underestimates contact time and overestimates aerial
time compared to photocell- and optical-based systems.

Introduction measurements and their concurrent validity against gold stan-


dard methods, such as force platforms and high-frequency video
In running, the temporal aspects of the stride cycle (i.e. con-
cameras, are well established (Castagna et al., 2013; Glatthorn
tact time, aerial time and stride frequency) are often of interest
et al., 2011). One advantage of the Optojump Next® is the pos-
because these relate to running efficiency and performance.
sibility to serially connect several single-metre bars to measure
For instance, contact time has been related to both maximal
consecutive running cycles. Although this provides a better over-
running speed (r = −0.52, P < 0.01) and running economy
all representation of individuals’ stride kinematics given the
(r = 0.49, P < 0.05) in endurance runners (Nummela, Keränen,
redundancy of cycles, the set-up of such a long photocell-
& Mikkelsson, 2007). In sprinters, contact time and stride fre-
based measurement area is rather time-consuming and expen-
quency are suggested to be the most important kinematic
sive, limiting its practical use in a coaching environment.
parameters for generating differences between elite athletes
In contrast, the Myotest® (Myotest SA, Sion, Switzerland) is
(Coh, Milanović, & Kampmiller, 2001), with early studies by
a lightweight portable accelerometer that provides real-time
Mero and Komi (1986) indicating that maximal speeds are
kinematic feedback and can capture numerous movement
achieved primarily by increasing stride frequency rather than
cycles. Although reportedly valid for measuring force and
stride length. More recently, Weyand, Sandell, Prime, and
power production during squat and bench press exercises
Bundle (2010) concluded that the rates at which the lower
(Comstock et al., 2011), and accelerations during high-pull
extremities can apply forces in stance limit maximal running
manoeuvres (Sato, Smith, & Sands, 2009), its validity for mea-
speeds, with faster runners applying greater mass-specific
suring biomechanical parameters during squat, countermove-
ground-contact forces over a shorter period of time (Weyand
ment and repeated jumps has been questioned (Casartelli,
et al., 2010). As such, it appears important for coaches and
Müller, & Maffiuletti, 2010; Castagna et al., 2013; Houel, Dinu,
athletes to have access to portable user-friendly devices that
Faury, & Seyfried, 2011). For instance, during countermove-
are able to accurately quantify stride parameters.
ment jumps, the Myotest® overestimates aerial time by 7.2%
In recent years, light-based optical technologies such as the
(Castagna et al., 2013) and vertical jump height by approxi-
Optojump Next® (Microgate, Bolzano, Italy) have become a clini-
mately 30% (Casartelli et al., 2010) in comparison to the
cally accessible and accurate means for computing ground-con-
Optojump Next®.
tact and aerial times during jumping, hopping and running
The between-system discrepancies are likely due to the
(Bosquet, Berryman, & Dupuy, 2009; Debaere, Jonkers, &
computational methods applied. The Optojump Next®
Delecluse, 2013). The reproducibility of photocell-based

CONTACT Thibault Lussiana thibault.lussiana@gmail.com University of Franche Comté, 31 chemin de l’Epitaphe, 25000 Besançon, France.
© 2015 Taylor & Francis
2 C. GINDRE ET AL.

considers the contact time as the time during which the foot is with runners wearing their habitual running clothing and
in contact with the ground and the aerial time as the time shoes. The weight and height of participants on the day of
elapsed from take-off to landing, as used with force platforms testing were recorded and entered into the Myotest® device.
(Bosco, Tarkka, & Komi, 1982). In contrast, the Myotest® defines After a 5 min running warm-up at a self-selected speed, parti-
the contact time as the time during which the vertical force is cipants were familiarised with the testing protocol, running
greater than body weight and the aerial time as the time 60 m once at the four selected experimental speeds, which
during which the vertical force is below body weight, named enabled participants to become accustomed to the test
the effective contact and flight times by Cavagna, Franzetti, speeds.
Heglund, and Willems (1988). To delineate these effective For testing, participants were required to run 60 m from
times, the vertical force is calculated based on Newton’s sec- stand-still twice at 12, 15, 18 and 21 km·h−1 separated by a 3
ond law using the body weight of individuals and the vertical min rest period during which participants were permitted to
acceleration data from the Myotest® device as inputs. The walk. The order of testing was held constant across partici-
duration of the contact and aerial times are determined pants to allow a gradual build-up in running speed during the
directly from the vertical acceleration signal processing, as experimental protocol. During trials, participants were not
described by Flaction, Quievre, and Morin (2013). paced or provided with real-time feedback regarding their
Despite its mediocre validity and divergent results from running speeds. However, to monitor speed in quasi real
gold standard assessment methods, the high reliability of the time, a timing-gate system (Racetime2, MicroGate, Timing
Myotest® system for measuring jump and strength training and Sport, Bolzano, Italy) was used and paired photoelectric
parameters makes it a practical tool for intra- and inter- sensors were placed at the 20 and 40 m mark of the 60 m trial.
individual comparisons. More specifically, when measures are A running trial was repeated after 3 min of rest when partici-
Downloaded by [Thibault Lussiana] at 22:05 15 July 2015

taken at various time points of an athlete’s development, pants ran at a speed of ±0.5 km·h−1 of the requested one, with
these can be used to identify gains from training programmes participants being told how much faster or slower than the
(Cook, Beaven, & Kilduff, 2013) or, on the contrary, detriments target speed they had ran so they could adjust their speed for
from maladaptive responses to training (Kraemer et al., 2004). the subsequent trial. No participant expressed residual fatigue
Despite the array of studies examining the Myotest® device, from preceding procedures.
none has yet reported the reproducibility or validity of the
running stride parameters it acquires. Determining the psy-
Equipment
chometric properties of the Myotest® for calculating running
stride characteristics could assist current users to interpret During experimentation, the Myotest®, the Optojump Next®
their data, facilitate between-study comparison and aid poten- and two high-speed video cameras (Casio High Speed EXILIM
tial users to estimate the cost–benefit of purchase of this EX-FH25®, CASIO Europe GmbH, Norderstedt, Germany)
software for use in their practical environment. As such, our sampled data simultaneously at 500, 1000 and 300 Hz, respec-
aims were to assess the reproducibility of contact time, aerial tively. The Myotest® device was fixed onto a neoprene belt,
time and step frequency parameters acquired during running fastened around the waist of participants, and placed level
using the Myotest® to determine its reliability and assess its with the navel following the manufacturer’s recommenda-
concurrent validity against a photocell- and a video-based tions. The Myotest® was triggered at the start of each 60 m
system. Considering the prior studies addressing its psycho- run by participants pressing the start button.
metric properties (Bernard et al., 2012; Casartelli et al., 2010; The Optojump Next® system was placed in the middle of
Comstock et al., 2011; Houel et al., 2011; Nuzzo, Anning, & the 60 m runway to sample data from 20 to 40 m by connect-
Scharfenberg, 2011; Ruben, Saffel, McCrory, Cormie, & Haff, ing several single-metre bars in series. The Optojump Next®
2011), we hypothesised that the Myotest® would be reliable therefore measured 20 m of data, with the transmitting and
for measuring stride kinematics, but would overestimate aerial receiving bars positioned 2 m apart.
time. For high-speed video analysis, one camera was positioned
on a 0.5 m high tripod at the 30 m mark perpendicular to the
running direction, 4 m to the side of participants, to capture
Methods
sagittal plane motion. A second camera was positioned on a 1
Twenty habitual male runners (age 31.6 ± 9.2 years, height m high tripod 10 m from the end of the measuring area, in
178 ± 5.4 cm, body mass 72.5 ± 9.8 kg) who ran on average front of participants, to capture frontal plane motion. This
4.0 ± 2.1 h·wk−1 and had been for the last 10.1 ± 6.8 years configuration allowed six running steps to be measured for
volunteered to participate. After a detailed explanation of the all participants in the mid-part of each 60 m run. The same six
study, participants provided their written informed consent. steps from all three systems were analysed using the video
The research protocol was approved by the local ethics com- camera as basis for synchronisation.
mittee and adhered to the latest amendments of the
Declaration of Helsinki.
Parameters
Contact time (ms), aerial time (ms) and step frequency
Procedures
(step·min−1) were computed and exported from the Myotest®
The experiment was held in an indoor facility on a flat cement and Optojump Next® software, with step frequency being the
surface to maintain test conditions stable across participants, inverse of the duration of the summation of contact and aerial
JOURNAL OF SPORTS SCIENCES 3

times. The data from the six defined running steps were sub- On the basis of commonly used thresholds, the relative
sequently isolated. The same step parameters were computed reliability and validity of measures were considered poor,
from the video data using Dartfish Pro Analysis Software v.5.5 fair and good when the corresponding ICC values were
(Dartfish Company, Fribourg, Switzerland). The mean value <0.4, 0.4–0.75 and ≥0.75 (Portney & Watkins, 2009; Rosner,
from the six steps recorded from each individual and measur- 2010). The absolute reliability and validity of measures were
ing device was used for subsequent analyses. considered adequate when the corresponding CV values
were equal to or lower than 10% (Hébert-Losier & Beaven,
2014; Holsgaard Larsen, Caserotti, Puggaard, & Aagaard,
Statistical analysis 2007; Stokes, 1985). All statistical analyses were performed
using SigmaStat for Windows 3.5 (Systat Software Inc., San
Descriptive statistics were computed to summarise the data
Jose, CA, USA) and Microsoft Excel 2010 (Microsoft Corp.,
from the 20 runners. All data were normally distributed on
Redmont, WA, USA) with a level of significance set at
the basis of Kolmogorov–Smirnov test. Student’s t-test for
α ≤ 0.05.
paired samples was employed to detect systematic bias
between trials performed at the same speed (Atkinson &
Nevill, 1998). To assess the relative between-trial reproduci-
bility of measurements, intra-class correlation coefficients Results
(ICCs) were computed. These coefficients were computed as
Reliability
[ICC = 1 – (SEM/SD)2], where SEM is the standard error in
measurement and SD is the mean between-participant SD of The stride parameters calculated for each running trial and
Downloaded by [Thibault Lussiana] at 22:05 15 July 2015

the two trials obtained by weighing the variances on the the mean between-trial difference are reported in Table I for
basis of their degrees of freedom (Hopkins, 2007). The SEM the three systems. The between-trial differences in contact
was computed as [SEM = SD (between-trial difference in time, aerial time and step frequency across running speeds
p
measures) / 2]. were 2.0 ± 0.6 ms, 1.1 ± 0.9 ms and 1.7 ± 1.7 step·min−1
To quantify the absolute reliability, the between-trial differ- for the Myotest®; 3.8 ± 1.9 ms, 1.4 ± 1.1 ms and
ence in means was computed (in raw units and %) and the 1.3 ± 0.4 step·min−1 for the Optojump Next®; and
SEM defined above was expressed as a coefficient of variation 2.1 ± 0.3 ms, 2.1 ± 1.9 ms and 0.9 ± 0.5 step·min−1 for the
(CV) on the basis of [CV = (SEM/mean)/100%], where the mean high-speed video cameras.
takes into consideration all 20 participants and both trials at The ICC and CV values specific to the reliability of each
the same speed (Atkinson & Nevill, 1998; Hopkins, 2000). The stride parameter at the different test speeds are reported in
SEM was expressed as a CV considering that the interest is not Table II for the three systems. The means of the ICCs were
in the uncertainty of a single mean, but rather that of an effect 0.932 ± 0.057 (range 0.82–0.99), 0.950 ± 0.034 (range 0.882–
involving comparisons of two means (Hopkins, Marshall, 0.985) and 0.908 ± 0.094 (range 0.718–0.986) for the
Batterham, & Hanin, 2009). Myotest®, Optojump Next® and high-speed video cameras,
Concurrent validity was assessed by comparing the mean respectively. Their corresponding mean CV values were
of trials performed at the same speed between two systems. 5.6 ± 1.9 (range 3.9–9.9), 6.8 ± 3.6 (range 3.9–14.3) and
Similar statistical measures than those used to assess reliability 6.1 ± 2.7% (range 3.8–10.5). Overall, all three systems
were employed for concurrent validity. More precisely, we demonstrated good relative and adequate absolute reliabil-
computed Student’s t-test for paired samples (systematic ity for all stride parameters, except for aerial times at speeds
bias), ICC values (relative validity), between-system differences of 12–18 km·h−1 for the Optojump Next® and 18 km·h−1 for
in means (absolute validity in raw units and %) and CVs the high-speed video cameras, where absolute reliability
(absolute validity in %). was inadequate (Table II).

Table I. Running stride parameters calculated using the Myotest®, Optojump Next® and high-speed video camera systems during 60 m running trials.
Myotest® Optojump Next® High-speed cameras
Parameter (unit) Trial 1 Trial 2 Δ (%) Trial 1 Trial 2 Δ (%) Trial 1 Trial 2 Δ (%)
Contact time (ms)
12 km·h−1 165 ± 16 167 ± 15 0.7 265 ± 19 272 ± 16 2.4 251 ± 19 253 ± 16 0.8
15 km·h−1 155 ± 15 153 ± 15 −1.2 238 ± 15 236 ± 16 −1.2 224 ± 14 222 ± 12 −0.8
18 km·h−1 137 ± 16 134 ± 16 −1.9 209 ± 15 207 ± 12 −1.0 199 ± 13 197 ± 11 −1.3
21 km·h−1 115 ± 17 117 ± 16 1.9 181 ± 16 184 ± 15 2.0 172 ± 13 174 ± 12 1.2
Aerial time (ms)
12 km·h−1 204 ± 14 205 ± 14 0.7 102 ± 20 99 ± 21 −2.4 117 ± 17 117 ± 18 −0.3
15 km·h−1 205 ± 14 202 ± 14 −1.1 120 ± 19 119 ± 21 −0.5 135 ± 18 133 ± 19 −1.4
−1
18 km·h 205 ± 15 205 ± 16 0.1 131 ± 18 131 ± 19 0.2 141 ± 17 145 ± 22 3.3
−1
21 km·h 202 ± 16 203 ± 15 0.3 136 ± 17 134 ± 18 −1.6 144 ± 18 143 ± 18 −1.1
Step frequency (step·min−1)
12 km·h−1 164 ± 9 163 ± 11 −0.1 164 ± 10 162 ± 9 −1.1 163 ± 9 163 ± 9 −0.4
15 km·h−1 166 ± 9 170 ± 10 −2.0 168 ± 10 170 ± 9 0.9 168 ± 9 169 ± 9 0.9
−1
18 km·h 176 ± 9 178 ± 11 1.6 177 ± 10 178 ± 10 0.5 177 ± 9 176 ± 12 −0.3
−1
21 km·h 188 ± 12 188 ± 10 −0.2 190 ± 13 189 ± 11 −0.5 190 ± 13 190 ± 11 −0.4
Note: The mean ± SD for each trial and test speed, and the difference between trials (Δ, in %), are reported.
4 C. GINDRE ET AL.

Table II. The relative (intra-class correlation coefficient, ICC) and absolute (coef- Table IV. The relative (intra-class correlation coefficient, ICC) and absolute
ficient of variation, CV) reproducibility of running stride parameters calculated (coefficient of variation, CV) concurrent validity of running stride parameters
using the Myotest®, Optojump Next® and high-speed video camera systems calculated using the Myotest® against the Optojump Next® and high-speed
during 60 m running trials. video camera systems during 60 m running trials.
Myotest® Optojump Next® High-speed cameras Optojump Next® High-speed cameras
Parameter (unit) ICC CV (%) ICC CV (%) ICC CV (%) Parameter (unit) ICC CV (%) ICC CV (%)
Contact time (ms) Contact time (ms)
12 km·h−1 0.993 6.5 0.903 4.6 0.963 4.8 12 km·h−1 0.632 17.7 0.715 15.9
15 km·h−1 0.877 6.7 0.882 4.5 0.860 4.0 15 km·h−1 0.674 16.2 0.470 14.1
18 km·h−1 0.946 8.3 0.927 4.5 0.828 4.3 18 km·h−1 0.752 16.4 0.629 14.7
21 km·h−1 0.973 9.9 0.949 6.1 0.974 5.0 21 km·h−1 0.822 17.9 0.741 16.0
Aerial time (ms) Aerial time (ms)
12 km·h−1 0.936 4.6 0.963 14.3 0.967 10.5 12 km·h−1 0.672 24.7 0.689 20.5
15 km·h−1 0.951 4.8 0.970 11.7 0.986 9.7 15 km·h−1 0.717 19.8 0.655 16.3
18 km·h−1 0.983 5.2 0.981 10.0 0.718 9.7 18 km·h−1 0.781 17.1 0.656 14.8
21 km·h−1 0.978 5.2 0.982 9.1 0.981 8.7 21 km·h−1 0.817 15.3 0.771 13.7
Step frequency (step·min−1) Step frequency (step·min−1)
12 km·h−1 0.944 4.4 0.921 4.1 0.953 3.8 12 km·h−1 0.863 4.1 0.889 3.9
15 km·h−1 0.820 3.9 0.978 3.9 0.940 3.8 15 km·h−1 0.941 3.9 0.946 3.7
18 km·h−1 0.835 4.1 0.985 4.0 0.753 4.5 18 km·h−1 0.925 3.8 0.844 3.8
−1 −1
21 km·h 0.942 4.1 0.956 4.5 0.970 4.4 21 km·h 0.865 4.6 0.860 4.6
Note: ICC < 0.75 and CV > 10% are italicised and represent fair relative Note: ICC < 0.75 and CV > 10% are italicised and represent fair relative
reproducibility and less than adequate absolute reproducibility of measures, reproducibility and less than adequate absolute reproducibility of measures,
respectively. respectively.
Downloaded by [Thibault Lussiana] at 22:05 15 July 2015

Concurrent validity overall fair for contact and aerial parameters against the
The mean of both trials is presented in Table III for each Optojump Next® (0.733 ± 0.071) and high-speed video cam-
system, as are the differences between the Myotest® and the eras (0.666 ± 0.092). The corresponding relative (CV) concur-
two reference systems. The Myotest® recorded significantly rent validity measures were less than adequate (18.1 ± 3.0 and
shorter contact time (range −57 to −102 ms) and longer aerial 15.8 ± 2.1%). On the other hand, the step frequency values
time (range 58–104 ms) compared to the Optojump Next® and from the Myotest® demonstrated good absolute and adequate
high-speed video cameras (Table II), with no significant differ- relative concurrent validity against the Optojump Next® (ICC
ences observed in terms of step frequency (range −0.6 to 0.899 ± 0.040, CV 4.1 ± 0.4%) and high-speed video cameras
1.8 ms). In contrast, stride parameters recorded using the (ICC 0.885 ± 0.05, CV 4.0 ± 0.4%). In contrast, the Optojump
Optojump Next® and high-speed video cameras showed no Next® and high-speed video cameras exhibited overall good
significant differences (all parameters, P > 0.05). absolute concurrent validity with ICCs of 0.901 ± 0.069 (range
The ICC and CV values describing the concurrent validity of 0.738–0.983) and adequate relative concurrent validity with
running stride parameters calculated using the Myotest® CVs of 6.7 ± 3.0% (range 3.9–12.8%).
against the two reference systems are reported in Table IV.
The absolute (ICC) concurrent validity of the Myotest® was
Discussion
In agreement with our hypotheses, the Myotest® was reliable
Table III. Running stride parameters calculated using the Myotest®, Optojump for measuring stride kinematics, but overestimated aerial time
Next® and high-speed video camera systems during 60 m running trials.
compared to photocell- and high-speed video camera sys-
High- Myotest® vs Myotest® vs
Parameter Optojump speed Optojump high-speed tems. This overestimation in aerial time was observed with
(unit) Myotest® Next® cameras Next® cameras an underestimation in contact time. Hence, the Myotest® can
Contact time (ms) Δ (%) be considered as a reliable system for computing stride para-
12 km·h−1 166 ± 15 268 ± 17 252 ± 17 38* 34* meters during overground running at speeds ranging from 12
15 km·h−1 154 ± 15 237 ± 15 223 ± 13 35* 31*
18 km·h−1 135 ± 16 208 ± 13 198 ± 12 35* 32* to 21 km·h−1. However, the Myotest® did not demonstrate
21 km·h−1 116 ± 16 182 ± 16 173 ± 12 36* 33* good concurrent validity for aerial and contact time measures,
Aerial time (ms) Δ (%) warranting caution against the comparisons of results
12 km·h−1 205 ± 13 101 ± 20 117 ± 17 −103* −75* between photocell-, video- and accelerometer-based systems.
15 km·h−1 204 ± 14 119 ± 20 133 ± 18 −71* −52*
18 km·h −1
205 ± 15 131 ± 18 143 ± 20 −57* −43* Only step frequency parameters derived from the Myotest®
−1
21 km·h 202 ± 15 135 ± 17 144 ± 18 −50* −41* proved to be highly reliable and valid against our reference
Step frequency (step·min−1) Δ (%) systems.
12 km·h−1 163 ± 10 163 ± 9 163 ± 9 −0.1 −0.2 The use of several statistical parameters is recommended
15 km·h−1 168 ± 9 169 ± 9 168 ± 9 0.5 0.2
18 km·h −1
177 ± 10 178 ± 10 176 ± 11 0.3 −0.3 for quantifying the reliability of measures (Atkinson & Nevill,
21 km·h−1 188 ± 11 190 ± 12 190 ± 12 0.8 0.9 1998). In this study, all statistical indicators implied high relia-
Notes: The mean ± SD for both trials combined at each test speed for each bility of stride characteristics derived from the three different
system, and the differences between the Myotest® and the other two systems systems, except for aerial times at speeds of 12–18 km·h−1
(Δ, in %), are reported.
*P < 0.05, significant difference between the Myotest® and the Optojump Next® using the Optojump Next® and at 18 km·h−1 using the high-
or high-speed video cameras using paired t-tests. speed video cameras. Other than measurement noise, the
JOURNAL OF SPORTS SCIENCES 5

slight non-significant differences in running speeds observed but low validity of the Myotest® for measuring contact and
between trials ran at the same speed (e.g. 14.7 ± 0.7 vs. aerial times, implementing correction factors to achieve valid
15.0 ± 0.6 km·h−1, respectively) and individual variations in absolute measures and to facilitate cross-study comparisons of
running biomechanics likely contributed to the imperfect relia- results may be proposed. Linear regression analyses on our
bility of measures observed in all three systems. The subtle data suggest using the following equations to obtain contact
differences in running speed and technique between trials are (xc) and aerial (xa) times from the Myotest® that are compar-
normal variations expected in any testing situation, hence the able to those from the Optojump Next® (1.274 xc + 42.9 and
importance of establishing the reproducibility of test measures 1.964 xa + 75.2) and high-speed video cameras (1.151 xc + 47.2
to determine which differences exceed typical variations in and 1.814 xa + 42.6) when individuals run at speeds ranging
performance. from 12 to 21 km·h−1.
For the Myotest® specifically, the between-trial ICCs were all We also observed that, despite our attempts to tightly fix
above 0.820, and CVs below 10%. Our indicators of reliability the Myotest® device around participants, undesirable move-
of accelerometer-derived parameters are in line with those ments of the accelerometer were unavoidable. Since the accel-
reported for quantifying vertical jump heights based on aerial erometer was attached over clothing and around soft tissues,
times (ICCs 0.74–0.99, CVs 4.3–6.4%) (Casartelli et al., 2010; notably the abdomen, aberrant movement artefacts could not
Choukou, Laffaye, & Taiar, 2014) and walking gait on firm be eliminated completely. Furthermore, although positioned
surfaces (ICCs 0.76–0.97) (Cole et al., 2014; Henriksen, Lund, to approximate the position of the centre of mass of indivi-
Moe-Nilssen, Bliddal, & Danneskiod-Samsøe, 2004). Together, duals, the location of the Myotest® during running is not
these results tend to highlight that neither the positioning of mechanically equivalent to the actual position of the centre
the device (i.e. anterior to participants for running and lateral of mass. These limitations associated with the external posi-
Downloaded by [Thibault Lussiana] at 22:05 15 July 2015

for jumping) nor the direction or complexity of the movement tioning of the accelerometer during a high-impact movement
(i.e. horizontal bounding gait vs. vertical explosive jumps) such as running are of particular concern in individuals with
substantially affect the reproducibility of accelerometer- high waist circumferences. Indeed, with increments in adipose
derived parameters. The confirmation of the reproducibility tissue or thick clothing, the measurement quality from accel-
of stride parameters during running gait using the Myotest® erometers may decrease, as can be observed in Figure 1. Non-
allows practitioners to be confident in their ability to track vertical motions of the device during running primarily caused
these particular measures over time using this device, by tilting and rotation of the pelvis are another potent source
whereby changes in stride parameters captured using the of measurement error. The highlighted differences between
Myotest® would reflect actual changes in running stride the portable accelerometer and the other two fixed reference
characteristics. systems might also have contributed to the between-system
That said, the concurrent validity of running stride para- differences observed in aerial and contact times. Further
meters against the Optojump Next® and high-speed video refinement of the filtering algorithms within the Myotest®
cameras was only fair. On average, the Myotest® underesti- software is needed to increase the robustness of this device
mated contact time by 34% and overestimated aerial time by to different body sizes and movement artefacts.
64%. These results most likely reflect the between-system One limitation of the present study was the use of photo-
difference in computational methods. The Myotest® device cell- and video-based systems as gold standard reference for
divides the step cycle into two parts based on when the establishing concurrent validity instead of ground reaction
vertical force is lower than and greater than body weight to force measuring instruments, such as a force platform. As
define contact and aerial times, respectively, named the effec- we do not possess such equipment in our laboratory, the
tive contact and aerial times by Cavagna et al. (1988). On the use of the Optojump Next® and video cameras were consid-
other hand, the traditional method of dividing the step cycle ered as adequate proxy systems given their demonstrated
according to foot-ground contact is used by the Optojump good validity in comparison to force platforms (Castagna
Next® and high-speed video cameras. That these latter two et al., 2013; Glatthorn et al., 2011). Furthermore, the
systems relied on similar methods for defining contact and Optojump Next® and video cameras are more practical and
aerial times explains their higher concurrent validity and portable for recording several consecutive steps than force
between-system agreement. In contrast, since step frequency platforms imbedded into the ground in series where partici-
directly depends on step duration (i.e. the inverse of the pants often have to adjust stride lengths and target force
contact time plus the aerial time), the differences between platforms to obtain clearly defined foot contact data. As
the three systems were negated, leading to very close agree- such, our measures with the Optojump Next® and video
ment between the Myotest® measures and those from the two cameras are considered to be more eco-valid. Another con-
other devices (ICCs ~0.90 and CVs ~4%). Clinicians and scien- sideration is that our test–retest runs were performed within a
tists must be aware of these between-system discrepancies, given session rather than on different days. A smaller time
particularly when comparing results from different studies, frame between test sessions usually leads to greater reprodu-
labs or clinics, and acquiring new equipment for the purpose cibility of performance measures (Ditroilo, Forte, McKeown,
of quantifying running stride kinematics. Only step frequency Boreham, & De Vito, 2011). As such, our current reliability
can be reliably contrasted between photocell-, video- and statistics might not generalise to runs performed several
accelerometer-based systems. Considering the high reliability days apart.
6 C. GINDRE ET AL.

Figure 1. For a participant who had a high waist circumference, the Myotest® device misclassified a flight phase into a ground contact phase (shaded area) because
Downloaded by [Thibault Lussiana] at 22:05 15 July 2015

of accentuated accelerations caused by soft tissues.

Conclusion Bosco, C., Tarkka, I., & Komi, P. V. (1982). Effect of elastic energy and
myoelectrical potentiation of triceps surae during stretch-shorten-
On the basis of traditional thresholds, the relative (i.e. ICC) and ing cycle exercise. International Journal of Sports Medicine, 3(3),
absolute (i.e. CV) reliability of contact time, aerial time and step 137–140.
frequency parameters derived using the Myotest® device were Bosquet, L., Berryman, N., & Dupuy, O. (2009). A comparison of 2 optical
timing systems designed to measure flight time and contact time
classified as adequate for running assessments. These findings during jumping and hopping. Journal of Strength and Conditioning
suggest that the Myotest® is practically useful for monitoring Research, 23(9), 2660–2665.
individuals and quantifying changes in functional performance Casartelli, N., Müller, R., & Maffiuletti, N. A. (2010). Validity and reliability of
over time. However, the concurrent validity of the Myotest® the Myotest accelerometric system for the assessment of vertical jump
against optical-based reference systems (i.e. Optojump Next® height. Journal of Strength and Conditioning Research, 24(11), 3186–
3193.
and video cameras) was only fair for contact and aerial times. Castagna, C., Ganzetti, M., Ditroilo, M., Giovannelli, M., Rocchetti, A., &
Clinicians and scientists must consider these between-system dis- Manzi, V. (2013). Concurrent validity of vertical jump performance
crepancies when comparing findings from various studies, labs or assessment systems. Journal of Strength and Conditioning Research,
clinics. Only step frequency can be reliably contrasted between 27(3), 761–768.
the Myotest® and photocell- or video-based systems, showing Cavagna, G. A., Franzetti, P., Heglund, N. C., & Willems, P. (1988). The
determinants of the step frequency in running, trotting and hop-
both adequate reliability and concurrent validity of measures. ping in man and other vertebrates. The Journal of Physiology, 399,
81–92.
Choukou, M. A., Laffaye, G., & Taiar, R. (2014). Reliability and validity of an
Acknowledgements accele-rometric system for assessing vertical jumping performance.
We gratefully acknowledge the Swiss Olympic Medical Centre of the Biology of Sport, 31(1), 55–62.
Clinique Romande de Rehabiliation SuvaCare in Sion (Switzerland) for its Coh, M., Milanović, D., & Kampmiller, T. (2001). Morphologic and
involvement in the experimental set-up. We also thank Dr Laurent Mourot kinematic characteristics of elite sprinters. Collegium Antropologicum,
for his critical reading of this paper. 25(2), 605–610.
Cole, M. H., Van Den Hoorn, W., Kavanagh, J. K., Morrison, S., Hodges,
P. W., Smeathers, J. E., & Kerr, G. K. (2014). Concurrent validity of
accelerations measured using a tri-axial inertial measurement unit
Disclosure statement
while walking on firm, compliant and uneven surfaces. PloS One,
No potential conflict of interest was reported by the authors. 9(5), e98395.
Comstock, B. A., Solomon-Hill, G., Flanagan, S. D., Earp, J. E., Luk, H. Y.,
Dobbins, K. A., & Kraemer, W. J. (2011). Validity of the Myotest® in
References measuring force and power production in the squat and bench press.
Journal of Strength and Conditioning Research, 25(8), 2293–2297.
Atkinson, G., & Nevill, A. M. (1998). Statistical methods for assessing Cook, C. J., Beaven, C. M., & Kilduff, L. P. (2013). Three weeks of eccentric
measurement error (reliability) in variables relevant to sports medicine. training combined with overspeed exercises enhances power and run-
Sports Medicine (Auckland, N.Z.), 26(4), 217–238. ning speed performance gains in trained athletes. Journal of Strength
Bernard, P. L., Amato, M., Degache, F., Edouard, P., Ramdani, S., Blain, H., & and Conditioning Research, 27(5), 1280–1286.
Codine, P. (2012). Reproducibility of the time to peak torque and the Debaere, S., Jonkers, I., & Delecluse, C. (2013). The contribution of step
joint angle at peak torque on knee of young sportsmen on the iso- characteristics to sprint running performance in high-level male and
kinetic dynamometer. Annals of Physical and Rehabilitation Medicine, female athletes. Journal of Strength and Conditioning Research, 27(1),
55(4), 241–251. 116–124.
JOURNAL OF SPORTS SCIENCES 7

Ditroilo, M., Forte, R., McKeown, D., Boreham, C., & De Vito, G. (2011). Intra- performance and hormonal concentrations over a big ten soccer sea-
and inter-session reliability of vertical jump performance in healthy son in starters and nonstarters. Journal of Strength and Conditioning
middle-aged and older men and women. Journal of Sports Sciences, Research, 18(1), 121–128.
29(15), 1675–1682. Mero, A., & Komi, P. V. (1986). Force-, EMG-, and elasticity-velocity relation-
Flaction, P., Quievre, J., & Morin, J. B. (2013). U.S. Patent n° US 2013/ ships at submaximal, maximal and supramaximal running speeds in
0190658. Washington, DC: U.S. Patent and Tradematk Office. sprinters. European Journal of Applied Physiology and Occupational
Glatthorn, J. F., Gouge, S., Nussbaumer, S., Stauffacher, S., Impellizzeri, F. Physiology, 55(5), 553–561.
M., & Maffiuletti, N. A. (2011). Validity and reliability of optojump Nummela, A., Keränen, T., & Mikkelsson, L. O. (2007). Factors related to top
photoelectric cells for estimating vertical jump height. Journal of running speed and economy. International Journal of Sports Medicine,
Strength and Conditioning Research, 25(2), 556–560. 28(8), 655–661.
Hébert-Losier, K., & Beaven, C. M. (2014). The MARS for squat, counter- Nuzzo, J. L., Anning, J. H., & Scharfenberg, J. M. (2011). The reliability of
movement, and standing long jump performance analyses: Are mea- three devices used for measuring vertical jump height. Journal of
sures reproducible? Journal of Strength and Conditioning Research, 28 Strength and Conditioning Research, 25(9), 2580–2590.
(7), 1849–1857. Portney, L. G., & Watkins, M. P. (2009). Foundations of clinical research:
Henriksen, M., Lund, H., Moe-Nilssen, R., Bliddal, H., & Danneskiod-Samsøe, Applications to practice (3rd ed.). Upper Saddle River, NJ: Pearson/
B. (2004). Test-retest reliability of trunk accelerometric gait analysis. Gait Prentice Hall.
& Posture, 19(3), 288–297. Rosner, B. (2010). Fundamentals of biostatistics (7th ed.). Boston, MA:
Holsgaard Larsen, A., Caserotti, P., Puggaard, L., & Aagaard, P. (2007). Cengage Learning.
Reproducibility and relationship of single-joint strength vs multi-joint Ruben, R. M., Saffel, H., McCrory, J. L., Cormie, P., & Haff, G. G. (2011).
strength and power in aging individuals. Scandinavian Journal of Comparison of accelerometer based vertical jump assessments to a
Medicine & Science in Sports, 17(1), 43–53. linear position transducer plus force plate system. The Journal of
Hopkins, W. G. (2000). Measures of reliability in sports medicine and Strength & Conditioning Research, 25, S37.
science. Sports Medicine (Auckland, N.Z.), 30(1), 1–15. Sato, K., Smith, S. L., & Sands, W. A. (2009). Validation of an acceler-
Downloaded by [Thibault Lussiana] at 22:05 15 July 2015

Hopkins, W. G. (2007). Understanding statistics by using spreadsheets to ometer for measuring sport performance. Journal of Strength
generate and analyze samples. Sportscience, 11, 23–36. and Conditioning Research, 23(1), 341–347. doi:10.1519/
Hopkins, W. G., Marshall, S. W., Batterham, A. M., & Hanin, J. (2009). JSC.0b013e3181876a01
Progressive statistics for studies in sports medicine and exercise Stokes, M. (1985). Reliability and repeatability of methods for measuring
science. Medicine & Science in Sports & Exercise, 41(1), 3–13. muscle in physiotherapy. Physiotherapy Theory and Practice, 1(2), 71–76.
Houel, N., Dinu, D., Faury, A., & Seyfried, D. (2011). Accuracy and reliability doi:10.3109/09593988509163853
of the Myotest pro system to evaluate a squat jump. Procedia Weyand, P. G., Sandell, R. F., Prime, D. N. L., & Bundle, M. W. (2010). The
Engineering, 13, 434–438. biological limits to running speed are imposed from the ground up.
Kraemer, W. J., French, D. N., Paxton, N. J., Häkkinen, K., Volek, J. S., Journal of Applied Physiology (Bethesda, Md.: 1985), 108(4), 950–961.
Sebastianelli, W. J., & Knuttgen, H. G. (2004). Changes in exercise doi:10.1152/japplphysiol.00947.2009

Você também pode gostar