2015 Book SportsPerformance

Kazuyuki Kanosue Editor in Chief
Tomoyuki Nagami
Jun Tsuchiya Editors
Sports
Performance
Sports Performance
Kazuyuki Kanosue
Editor in Chief
Tomoyuki Nagami • Jun Tsuchiya

Editors
Sports Performance
Editor in Chief
Kazuyuki Kanosue
Faculty of Sport Sciences
Waseda University
Saitama, Japan
Editors
Tomoyuki Nagami Jun Tsuchiya
Faculty of Sport Sciences Faculty of Sport Sciences
Waseda University Waseda University
Saitama, Japan Saitama, Japan
ISBN 978-4-431-55314-4 ISBN 978-4-431-55315-1 (eBook)

DOI 10.1007/978-4-431-55315-1
Library of Congress Control Number: 2015942559
Springer Tokyo Heidelberg New York Dordrecht London

© Springer Japan 2015
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or
dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained
herein or for any errors or omissions that may have been made.
Printed on acid-free paper
Springer Japan KK is part of Springer Science+Business Media (www.springer.com)

Introduction to the Series
Waseda University of Japan has a tradition of producing great athletes amongst its
graduates, such as Mikio Oda, the first Japanese ever to win an Olympic gold medal.
Waseda University strongly supports coaching techniques that embody a practical
application of the knowledge gained from the fundamental research findings of sports
science. Waseda University also takes pride in providing athletes with medical care
that utilizes leading-edge sports medicine, and formulates management strategies that
combine all these elements. This approach has led to a strong tradition of sports-based
research—what we like to call “Waseda Sports”—which has enjoyed an unprece-
dented level of success. This tradition was enhanced by the Faculty of Sport Sciences
in Waseda University in 2009 when they initiated the Global COE (Center of
Excellence) Program, entitled “Sport Sciences for the Promotion of Active Life”.
The Global COE Program is under the aegis of the Japanese Ministry of Education,
Culture, Sports, Science, and Technology; this Ministry supports the development of
international centers of education and research excellence.
While life expectancy in Japan is the highest in the world, large-scale societal
changes here and elsewhere have led to an increase in health problems due to a
decrease in activity and physical fitness. In the aging population there has been a
deterioration of overall health, much of which can be attributed to inactivity and
excess body weight. It is especially troubling that similar problems are increasing
among children and are associated with severe physical and mental disabilities. The
international scope of the above problems provided the impetus for Waseda Uni-
versity to form the Global COE Program. This effort involved the construction of an
international hub of education and research specifically designed to develop and
encourage talented researchers to create sports programs that would contribute to an
active and vital lifestyle. The program emphasizes the development of specialist
knowledge in conjunction with a broad understanding and awareness of the diverse
world of sports. One of our goals was to focus not just on improving the individual
health of mind and body, but also to develop an understanding of the conditions
present in regions and societies that facilitate such improvements in the lifestyle of
individuals.
v
vi Introduction to the Series
The sports sciences have created and are extending an important body of
knowledge. It is critical that this information be utilized to produce an active,
two-way interaction between the investigators and the active participants of sport-
ing events. In order to provide a focus for developing this reciprocal intercommu-
nication, the Global COE program identified three strategic project themes:
(1) Active Children Project, (2) Active Elderly Project, and (3) Elite Athlete
Project. The COE Program was proactive in seeking out mature graduate students
who were returning to higher education after a period of work, thereby facilitating a
meaningful contribution to the formation of academic careers for specialists who
were active in the practical domain of sports. Many graduate students from abroad,
especially from Asian countries, joined the program and have contributed to our
goals via both the creation of academic knowledge and direct participation in the
sports relevant to their area of investigation.
The formal funding for the Global COE Program came to an end in March 2014,
but the projects initiated by the program and the activities of the graduates continue
unabated. The accomplishments made during the 5 years of the program have been
documented in a series of four books with the overall theme of “Sports Science and an
Active Life”. We are proud to present this substantial body of research in the following
series of books: Vol. 1: Sports Management and Sports Humanities (Kohei Kogiso,
Daichi Oshimi, Munehiko Harada, Eds.), Vol. 2: Physical Activity, Exercise, Seden-
tary Behavior, and Promoting Health (Satomi Oshima, Zhen-Bo Cao, Koichiro Oka,
Eds.), Vol. 3: Sports Performance (Tomoyuki Nagami, Jun Tsuchiya, Eds.), and Vol.
4: Sports Injuries and Prevention (Tetsuya Ogawa, Mako Fukano, Toru Fukubayashi,
Eds.). The series was written by the dedicated faculty members and young graduate
students and postdoctoral researchers under the guidance of investigators who took
part in the Global COE program. The series was also contributed to by leading
researchers around the world, most of whom belong to Waseda University’s research
institute or university partners. I appreciate their contributions as well as their partic-
ipation in the Global COE program. During the 5 years of the program, an interna-
tional network of individuals and universities doing active research in the area of
sports sciences has been established. I expect this network to grow wider and stronger
in the future and to contribute to the solution of many of the health problems that
plague modern societies. We will all continue to work hard to involve sports activities
in the solutions to these problems, and in the process, aid in advancing the sports
activities themselves.
Finally, I express my appreciation to the editors of each volume, who not only
did a fine job of organizing the volumes but also wrote chapters that were important
scientific contributions to the overall effort. We would also like to thank the Global
COE staff for their efficient work and the kind support they extended to the graduate
students. Drs. Larry Crawshaw and Candace S. O’Connor are thanked for their
enthusiastic editorial assistance.
Program Leader Kazuyuki Kanosue

Global COE “Sport Sciences for the
Promotion of Active Life”
Waseda University
Preface
This book focuses on sports performance. According to the Longman Dictionary of

Contemporary English, “performance” refers to “how well or badly a person,
company etc. does a particular job or activity”, and “high performance” describes
“cars, computers etc. that are able to go faster, do more work etc. than normal
ones”. We investigate the question of what high performance means in the context
of sports.
In the 100-m dash Usain Bolt is indubitably the fastest person in history, and
Javier Sotomayor, the world record holder in the high jump, has exhibited the
highest level of performance in this event. In these contests, the index of sports
performance is unitary; it is simply the time or the jumping/throwing distance.
What is it that allows such performers to achieve the fastest running time or the
highest jump? One of the topics covered in this book is an attempt to clarify some
of the unique motor skills and/or physical abilities that underlie such high
performances.
On the other hand, judging performance in many sports is much more complex.
While a pitcher in baseball may be a “flamethrower” who can hurl a pitch at over
100 mph, he might not necessarily be considered the best pitcher. In baseball, to be
the pitcher with the best performance, that player must be the most successful at
getting batters out. In addition to speed, this requires good control, a number of
different pitches, the ability to disguise the intended pitch, and a good sense of what
the batter intends to do. This is clearly a complex skill set, and not surprisingly it is
not totally understood how the best pitchers achieve success.
Likewise, although performances in artistic gymnastics or figure skating are
quantitatively assessed by judges, there are ambiguous factors such as “execution of
action” and “technical beauty” that are quite subjective. Thus the essential factors
that define the level of sports performance are different for each sport. Another
topic covered in this book is an examination of the ways to describe sports
performance in contents for which direct numerical quantification, such as time
or distance, is not extant.
vii
viii Preface
Sports performances, and particularly those of elite athletes, have not attracted
much attention in the scientific literature because of difficulty in generalizing the
findings. High sports performance, however, can be regarded as the “extreme” level
of performance of human movement, and the elite athletes’ accomplishments
reflect the current maximal physiological and biomechanical potential for humans.
Therefore, detailed knowledge of such performances will be very useful not only
for athletes but also for young children who are trying to perform better.
This book comprises a compilation of updated reviews on performance in
various sports, including both basic and applied research, and is divided into
three parts.
The central theme of Part I is the brain. Basic research on human locomotion,
motor imagery, and cognitive function are included in this part. In Part II, the focus
is on basic information involving high performance in sports, including the athletes’
physiology, genetics, nutrition, and biomechanics.
In Part III, entitled “Performance and Coaching in Various Sports”, the latest
findings involving skills and performance in individual sports are presented. These
performances are thoroughly described and, to the extent possible, explained
utilizing observations that involve applied biomechanics, coaching science, and
information technology. In the e-book version, videos and images are available,
which provide valuable information on movement in sports. We believe that this
book will awaken a deeper and more sophisticated interest in exceptional sports
performance, not only in specialists such as researchers, athletes, and coaches, but
also in laypeople who enjoy participating in and watching sports.
Saitama, Japan Tomoyuki Nagami

Jun Tsuchiya
Contents
Part I The Sporting Brain

1 Sports Performance and the Brain . . . . . . . . . . . . . . . . . . . . . . . . . 3
Hiroki Nakata
2 Brain Activity During Motor Imagery . . . . . . . . . . . . . . . . . . . . . . 13
Nobuaki Mizuguchi
3 Brain Oscillations and Athletic Performance . . . . . . . . . . . . . . . . . 25
Andreas Mierau, Thorben Hülsdünker, and Heiko K. Strüder
4 Intra- and Inter-person Coordinated Movements of Fingers
and Toes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Tetsuro Muraoka, Yuki Watanabe, and Kazuyuki Kanosue
5 Training Locomotor Function: From a Perspective
of the Underlying Neural Mechanisms . . . . . . . . . . . . . . . . . . . . . . 49
Tetsuya Ogawa and Kazuyuki Kanosue
6 On the Structure of Movement Preparation: Inferences
from Motor Schema Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Lu Xu, Werner Sommer, and Hiroaki Masaki
7 Muscle Relaxation and Sports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Kouki Kato and Kazuyuki Kanosue
8 Neural Mechanisms of Muscle Cramp . . . . . . . . . . . . . . . . . . . . . . 79
Kento Nakagawa, Naokazu Miyamoto, and Kazuyuki Kanosue
9 Task Difficulty Affects the Association Between Childhood
Fitness and Cognitive Flexibility . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Keita Kamijo and Hiroaki Masaki
ix
x Contents
Part II The Physiological Basis of Sports Performance

10 Genetic Polymorphisms Associated with Elite Athlete Status . . . . . 105
Eri Miyamoto-Mikami, Noriyuki Fuku, and Masashi Tanaka
11 Resting Energy Expenditure in Japanese Athletes -as Applied
to Dietary Management for Athletes- . . . . . . . . . . . . . . . . . . . . . . . 125
Motoko Taguchi and Satomi Oshima
12 Health Issues and Preventive Strategies for Heavy Athletes . . . . . . 139
Satomi Oshima and Motoko Taguchi
13 High Fat Diet and Endurance Exercise Performance . . . . . . . . . . . 151
Kazuhiko Higashida and Mitsuru Higuchi
14 Nonuniform Muscle Hypertrophy Along the Length Induced
by Resistance Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Taku Wakahara
15 Quantitative Profiles of the Quadriceps Femoris in Sport
Athletes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Ryoichi Ema and Yasuo Kawakami
16 Jump Performance Enhancement Induced
by Countermovement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Kuniaki Hirayama
17 Can a High-Intensity Contraction Be Enhanced
by a Conditioning Contraction? Insight from the Relationship
Between Shortening Velocity of Muscle Fibers and Postactivation
Potentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Atsuki Fukutani and Yasuo Kawakami
18 Is Graduated Pressure Profile an Essential Feature
for Compression Stockings to Reduce Fatigue Development
of the Plantar Flexors? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
Naokazu Miyamoto
19 Exercise in Space: Physical and Mental Benefit . . . . . . . . . . . . . . . 223
Stefan Schneider, Tobias Vogt, and Vera Abeln
Part III Performance and Coaching in Various Sports

20 Energetic Considerations in Cross-Country Skiing . . . . . . . . . . . . . 247
Walter Herzog, Anthony Killick, and Kevin R. Boldt
21 Biomechanical Analysis of V2 Skating in Cross-Country
Skiing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Zenya Fujita and Jun Tsuchiya
Contents xi
22 Limb Force Generation as a Limiting Factor for Maximum-Effort

Acceleration Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
Geng Luo and Darren J. Stefanyshyn
23 Optimal Technique, Variability and Control in Gymnastics . . . . . . 293
Michael Hiley and M.R. (Fred) Yeadon
24 Activity of the Trunk and Leg Musculature During
the Flutter Kick . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
Koji Kaneoka, Atsushi Imai, and Morimitsu Kohdate
25 Open Water Swimming Performance . . . . . . . . . . . . . . . . . . . . . . . 313
Reira Hara and Isao Muraoka
26 The Spin on a Baseball for Eight Different Pitches Thrown
by an Elite Professional Pitcher . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
Tomoyuki Nagami, Takatoshi Higuchi, and Kazuyuki Kanosue
27 Baseball Hitting Accuracy and Contributing Factors . . . . . . . . . . . 335
Takatoshi Higuchi, Tomoyuki Nagami, and Kazuyuki Kanosue
28 Automatic Tracking of Player Locations from Video Image
of Football Game . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
Masaaki Honda and Noriyuki Oosaka
29 Conceptualization of Coaching Process and Coaching
Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
Hiroyuki Horino
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
Part I
The Sporting Brain
Chapter 1
Sports Performance and the Brain
Hiroki Nakata
Abstract The relationship between sports performance and the brain was exam-
ined based on data obtained using non-invasive neurophysiological and neuroim-
aging methods. Background brain electrical activity has been recorded by
electroencephalography (EEG) during sports performance. EEG has previously
been used to investigate aiming sports such as billiards, darts, shooting, and golf.
This review mainly describes EEG data obtained during golf putting by experts and
non-experts. A focus was placed on neural substrates in the golfers’ brains as a
model of neural plasticity based on studies utilizing functional magnetic resonance
imaging (fMRI) and structural MRI. Several problems that should be addressed in
future studies in this field were also discussed.
Keywords Athlete • Golf • Aiming • Alpha • Theta
1.1 Introduction
Several studies have investigated the relationship between sports performance and
brain activity. This relationship is important in the field of sport sciences. One
approach has been to record brain activity during actual sports performance and
compare the results obtained in athletes with those in non-athletes (novices).
Several non-invasive recording methods are used to measure human brain activity.
Neurophysiological methods include electroencephalography (EEG), magnetoen-
cephalography (MEG), and transcranial magnetic stimulation (TMS). Neuroimag-
ing methods involve functional magnetic resonance imaging (fMRI), positron
emission tomography (PET), and near-infrared spectroscopy (NIRS).
Many studies have investigated the underlying neural mechanisms associated
with sports performance and motor control (for a review, see Yarrow et al. 2009).
However, neural activity has to be recorded during actual sporting activity and
exercise to clarify the relationship between sports performance and brain activity.
Therefore, many problems need to be resolved to obtain reliable data. For example,
it is physically impossible to record MEG and fMRI data during actual sports
H. Nakata (*)
Department of Human Sciences, Faculty of Letters, Nara Women’s University, Nara, Japan
e-mail: hiroki-nakata@cc.nara-wu.ac.jp
© Springer Japan 2015 3

K. Kanosue et al. (eds.), Sports Performance, DOI 10.1007/978-4-431-55315-1_1
4 H. Nakata
because participants can not move their heads during the recordings. Of course,
MEG and fMRI equipment was not designed to be used outside. After a thorough
literature search, one study attempted to record brain activity during dancing with
PET (Brown et al. 2006). In this study, tasks involved in the performance involved
simple bipedal dance movements on a laminated grid, and ten amateur dancers were
trained to be proficient in these dance steps in advance of the scanning session.
However, the subjects performed their tasks without moving their heads, which
cannot be considered to truly reflect real dancing.
An EEG offers high temporal resolution on the order of milliseconds and
measures the electrical voltage of the brain through electrodes placed on the scalp
in accordance with the standardized guidelines of the International 10–20 system
(Jasper 1958). However, EEG data is often contaminated by artifacts originating
from eye movements, skin, muscle, and the surrounding environment. In spite of
this, the number of potential applications is markedly larger with EEG than with the
other neurophysiological methods described earlier because the limitation for
spatially involuntary movements ismuch less restricted. Some research groups
recently attempted to record neural activity during sports performance by
using EEG.
In this review, EEG studies attempting to clarify brain activity during actual
sports performance were discussed with a focus on golf putting. Differences in the
underlying mechanisms of brain activity between athletes (expert golfers) and
non-athletes were then described. Finally, based on these findings, solutions to
several issues associated with recording brain activity were suggested for use in the
field of sports sciences.
1.2 Brain Activity During Sports Performance: Golf

Putting
EEG techniques can be used to detect psychological and physiological responses

during sport activities; however, these responses can only be measured in a limited
number of sports, which does not include open skill and team sports such as soccer,
tennis, basketball, volleyball, and baseball. Target sports executing closed skill
tasks, which only require simple body motions, have been used to date. For
example, the spectral power in background EEG has been examined during the
pre-shot period of shooting (Hatfield et al. 1984; Bird 1987; Hillman et al. 2000;
Loze et al. 2001; Kerick et al. 2001, 2004), archery (Salazar et al. 1990; Landers
et al. 1991, 1994), and dart-throwing (Radlo et al. 2002). These studies have been
listed in a review article by Hatfield and colleagues (2004). A Fast Fourier Trans-
form is used to analyze spectral power algorithm and this decomposes the electrical
signal into its frequency components; therefore, the amplitude of each designated
frequency can be measured and expressed as an absolute or relative power. Abso-
lute power represents the mean power in each frequency band selected, while
1 Sports Performance and the Brain 5
relative power represents the relationship between the power in the selected fre-
quency bands and the total power (Crews and Landers 1993). Alpha (about 8–
12 Hz) and beta (about 14–30 Hz) band oscillations generally decrease over
sensorimotor cortical areas during motor preparation and the execution of voluntary
self-paced movements. This phenomenon is termed event-related desynchro-
nization (ERD) (for a review, see Pfurtscheller and Lopes da Silva 1999).
In addition to target sports including shooting, archery, and dart-throwing, golf
putting has also been often used to investigate the relationship between sports
performance and the brain because the putting swing and motion are not very
dynamic and pure EEG signals can be recorded during the swing. Ability in golf
putting is associated with synchronizing sensory information in planning and
control for the appropriate motor response (Craig et al. 2000). To successfully
perform golf putting, the golfer has to consider the distance from the hole to the
ball, ball direction, putting force, and environmental conditions such as slope and
grain direction. Accordingly, the visual system must orient to and process the most
salient perceptual cues necessary to ascertain both distance and direction informa-
tion, while working memory matches stroke tempo with the requisite stroke force
(Mann et al. 2011).
Crews and Landers (1993) recorded EEG activity in 34 highly skilled golfers
with electrodes attached over the motor and temporal cortices during the 3 s prior to
golf putting. They analyzed three parameters from the EEG activity: slow shift,
40 Hz (gamma) band activity, and relative power spectrum. They found a decrease
in the three parameters in the left hemisphere, motor cortex activity as the players
prepared to putt, and during the last second preceding the putt, and increased alpha
activity in the right hemisphere correlated with decreased error.
Baumeister and colleagues (2008) compared theta (4.75–6.75 Hz), alpha-1 (7–
9.5 Hz), alpha-2 (9.75–12.5 Hz), and beta-1 (12.75–18.5 Hz) spectral powers in golf
putting between expert golfers and unskilled novices. Skill-dependent differences
were observed in frontal theta and parietal alpha-2 spectral power, which suggested
that golfers develop task solving strategies including focused attention and an
economy in parietal sensory information processing, resulting in a more successful
performance.
Babiloni and colleagues (2008) examined EEG spectral power in expert golfers
during putting. They reported that high-frequency alpha power (about 10–12 Hz)
was smaller in amplitude over the frontal midline, and also smaller in the arm and
hand region of the right primary sensorimotor area during successful putting than
during unsuccessful putting; the greater the reduction in alpha power, the smaller
the error in unsuccessful putts. The results of this study suggest that novices may
use alpha power to evaluate their performances.
Baumeister and colleagues (2010) compared brain activity during motor tasks
between real and virtual (the Nintendo Wii) golf putting in ten golfers. The score
and EEG activity were recorded continuously, with a significantly better score
being recorded in real putting. Theta spectral power at frontal electrodes was
significantly larger in real putting than in the virtual putting, and alpha-2 power at
the parietal electrodes was also significantly larger in real putting than in virtual
6 H. Nakata
putting. They suggested that putting performance and brain activity were influenced
by the choice of a real or virtual environment; the increase in frontal theta power
indicated more focused attention, and higher alpha-2 power was related to the
quantity of sensory information processing in real putting.
Babiloni and colleagues (2011) analyzed the coordination of cortical activity, as
reflected by functional coupling of alpha rhythms across cortical regions, in
12 expert golfers. They showed that intra-hemispheric low-frequency alpha coher-
ence (8–10 Hz) in bilateral parietal-frontal regions and parietal-central regions was
higher in amplitude in successful than in unsuccessful putts. The same phenomenon
was confirmed in intra-hemispheric high-frequency alpha coherence (10–12 Hz) in
bilateral parietal-frontal regions. These findings suggest that the intra-hemispheric
functional coupling of cortical alpha rhythms between the visuo-spatial parietal
area and other cortical areas is implicated in the fine motor control of a golfer’s
performance.
Mann and colleagues (2011) recorded movement-related cortical potentials
(MRCPs) and the quiet eye period (QE) to assess the potential mechanisms under-
lying the psychomotor skills that differentiate expert and near-expert performers.
MRCPs are recorded before self-initiated voluntary movement, and reflect move-
ment preparation processing not involving cognitive processing for an imperative
stimulus (reviewed in Shibasaki and Hallett 2006). These potentials begin with a
slow rising negativity, called the Bereitschaftspotential (BP), and progress to a
steeper, later negativity, which starts approximately 500 ms before movement
onset, and is called the negativity slope (NS’). According to Vickers and Adolphe
(1997), the QE refers to a gaze behavior observed immediately prior to movement
in aiming tasks, and a temporal period in which task-relevant environmental cues
are processed and motor plans are coordinated for the successful completion of an
upcoming task. Mann and colleagues (2011) categorized 20 golfers into two groups,
those with a low handicap (LH: experts) and those with a high handicap (HH:
non-experts). Differences were observed in QE duration and BP, with experts
exhibiting a prolonged QE duration and greater cortical activation in the right-
central region than those of the non-experts. A significant relationship between
cortical activation and QE duration was also noted. Their findings suggest a motor
programming/preparation function for the QE duration.
Reinecke and colleagues (2011) compared brain activity results from laboratory
and field conditions during golf putting in 12 university students. They reported that
a significant difference was only observed in theta power at the F4 electrode
between two conditions. The results of their study indicated the possibility to
extend the limitations of the EEG methodology in its application to sports and
exercise sciences.
Taking these studies into consideration, alpha band activity appears to be related
to sports performance, especially successful or unsuccessful trials. In other words,
the functional characteristics of athletes’ brains may be explained by the coupling
of alpha rhythms during actual sporting activities. In general, alpha rhythms reflect
functional modes of the basal forebrain, thalamus, and cortical loops that facilitate/
inhibit the transmission and retrieval of both sensorimotor and cognitive
information into the brain (Babiloni et al. 2011). However, there appears to be a
difference in the functional meaning between low-frequency (about 8–10 Hz) and
high-frequency (about 10–12 Hz) alpha band activities. Low-frequency alpha band
activity reflects arousal, attentive readiness, and effort (Klimesch 1999;
Pfurtscheller and Lopes da Silva 1999), while high-frequency alpha band activity
is related to the task-related oscillation of specific neural systems for sensorimotor
or semantic information (Klimesch 1999). Based on these findings, alpha band
activity in the brain may be one of the key factors determining actual sports
performance. On the other hand, theta band activity may be directly associated
with attention processing and working memory rather than motor control. Previous
studies suggest that the anterior cingulate cortex (ACC) may be the generator of
frontal theta band activity based on the findings obtained from EEG and MRI
(Gevins et al. 1997), a dipole source model (Onton et al. 2005), and LORETA
(Sauseng et al. 2007). The ACC was shown to be involved in a range of executive
functions such as processing information and decision making; however, most
investigators view this subcortical region as an important component of the
human attentional control system (Baumeister et al. 2010).
1.3 Golfers’ Brain
Several studies have focused on the characteristics of athletes’ brains, by comparing

their results with those obtained from non-athletes (novices). Neuroimaging studies
using fMRI and structural MRI have examined specific brain activities in golfers.
fMRI, which measures blood oxygenation level-dependent (BOLD) signals, has
been used not only as a tool for mapping brain activity, but also as a means of
studying the dynamics of neural networks by tracking fMRI response characteris-
tics across various spatial and temporal scales (Logothetis et al. 2001). As men-
tioned above, fMRI cannot be used to directly record brain activity during actual
sports performance; however, this technique can be used to demonstrate differences
in the underlying mechanisms of brain activity between experts and non- experts.
Ross and colleagues (2003) evaluated the brain regions of motor imagery during
the golf swing in six golfers with various handicaps with fMRI. The golfers showed
activation areas in the motor cortex, parietal cortex, frontal lobe, cerebellum, and
vermis.
Milton and colleagues (2007) investigated brain activity related to motor plan-
ning in expert golfers and non-athletes during their pre-shot routine. The authors
reported that activity was primarily observed in the superior parietal lobule, dorsal
lateral premotor area, and occipital area of the experts, whereas the posterior
cingulate, the amygdala–forebrain complex, and the basal ganglia were active in
the non-athletes. These results suggest that experts were able to focus on motor
planning that integrated visual information with motor commands, whereas novices
were more likely to plan the movement in the context of emotional and possibly
task-irrelevant considerations.
8 H. Nakata
Jäncke and colleagues (2009) investigated the relationship between a particular

level of proficiency in playing golf, as indicated by golf handicap level and specific
neuroanatomical features, by utilizing anatomical MRI methods including voxel-
based morphometry (VBM) of grey (GM) and white matter (WM) volumes and
fractional anisotropy (FA) measures of the fiber tracts. VBM involves a voxel-wise
comparison of the local concentration of gray and white matter between two groups
of subjects, and between two data points observed at different timing in the same
subjects (Ashburner and Friston 2000). Larger GM volumes were observed in
skilled golfers in a fronto-parietal network involving the premotor and parietal
areas. Skilled golfers had smaller WM volumes and FA values in the vicinity of the
corticospinal tract at the level of the internal and external capsule and in the parietal
operculum. However, no structural differences were observed within the skilled and
less-skilled golfer groups. These results suggest that there is not a linear relation-
ship between anatomical findings and the handicap level.
Bezzola and colleagues (2011) did not analyze differences in brain activity or
structure between experts and non-experts, but rather showed training-induced
neural adaptations in golf novices between the age of 40 and 60 years. After 40 h
of golf practice performed as a leisure activity with highly individual training
protocols, the volume of gray matter increased in a task-relevant cortical network
including sensorimotor regions and areas belonging to the dorsal stream. The
relationship between training intensity and structural changes was observed in the
parieto-occipital junction.
Bezzola and colleagues (2012) also performed a longitudinal study using fMRI
during a motor imagery task in order to explore the dynamic neuro-functional
changes induced by highly complex physical training. Eleven golf novices between
the age of 40 and 60 years practiced motor training as a leisure activity, and the golf
novices showed a decrease in hemodynamic responses in non-primary motor areas
during the mental rehearsal of a golf swing after the 40 h of golf practice. Their
findings support the hypothesis that an improvement in skill is associated with a
modified activation pattern in the associated neuronal network that can be identified
during the mental rehearsal of a practiced task.
Neurophysiology and neuroimaging studies have recently shown that the neural
activation of motor-related cortical areas in humans undergoing short-term training
is modulated during the acquisition and performance of new motor skills (reviewed
in Karni et al. 1998). In contrast, it may be useful to investigate differences in the
brain structures of athletes and non-athletes in order to maintain consistency with
one of the models used to examine neuroplasticity and long-term training. As
described above, previous fMRI and structural MRI studies that focused on differ-
ences in brain activity between golfers (experts) and novices (non-experts) repre-
sent one of the approaches that can be used to establish an ideal model to investigate
plastic changes in the human brain. In addition, these studies may also reveal the
neural basis related to actual sports performance.
1.4 Future Studies
In this section, I describe approaches to improve sports sciences by using neuro-

physiology and neuroimaging, and to clarify the relationship between sports per-
formance and brain activity in future studies.
First, I introduced the relationship between sports performance and brain activity
by describing the use of recording background EEG in aiming sports such as
billiards, darts, shooting, and golf. I mainly elaborated on the data of golf putting.
In aiming sports, the processing of critical visual information and the ability to self-
regulate cognitive and emotional activity are keys to the successful execution of
self-paced movement skills (Williams et al. 2002). However, complex tracking and
aiming skills are needed to accurately perform the movements required in many
sports. For example, when an infielder throws a ball to the first baseman in baseball,
he has to aim at the baseman to accurately throw the ball. A tennis player has to use
the QE on a tossed ball during the serve to produce a high-speed ball. Therefore, the
ability of athletes to aim should be analyzed in future studies with EEG and the QE
– not only in closed skill sports, but also in open skill and team sports such as
soccer, tennis, basketball, volleyball, and baseball.
Previous studies reported differences in brain activity between athletes and
non-athletes. However, little is known about how long non-athletes (novices)
should train in order to acquire similar brain activity to that in athletes, and thus
achieve higher levels of performance. Although many studies have used cross-
sectional (transversal) methods, longitudinal studies are needed to quantify the
training effect over a longer period. Several studies demonstrated that musicians’
brains develop according to acquired factors that operate by virtue of
neuroplasticity rather than according to congenital factors (Altenmüller
et al. 2006). Therefore, many of the findings on athletes’ brains described in this
review may relate to acquired, rather than congenital factors. Longitudinal studies
and comparisons, including those between beginners and young players are needed
in the future. This issue is also relevant to neurorehabilitation prograns in terms of
the amount of time and practice required to regain motor skills.
Investigating brain activation during actual sports performance has remained a
major challenge for sports sciences and neuroscience. Wireless and mobile EEG
systems have recently been developed as a new technique. This method may allow
for a better understanding of motor-related brain activity and cognitive function not
only in sporting activities, but also daily life and clinical applications. With mobile
EEG, researchers now have access to a tool that can help address these issues
(Kranczioch et al. 2014; Wong et al. 2014). In addition, combined studies with
wireless EEG and electromyograms (EMGs) may be more useful for investigating
the relationship between actual sports performance and brain activity. Since there
are numerous studies examining the coherence between EEG and EMG at the
laboratory level (reviewed in Mima and Hallett 1999; Brown 2000), this analysis
should be developed after confirming stable recordings with wireless and
mobile EEG.
10 H. Nakata
NIRS could also be useful for the non-invasive investigation of brain activity
during voluntary motion and exercise, based on cerebral hemodynamic responses
with temporal resolution and good spatial resolution (1–2 cm) (for a review, see
Perrey 2008). Several studies have already reported brain activity during exercise in
healthy young participants (González-Alonso et al. 2004; Subudhi et al. 2007;
Timinkul et al. 2008). Thus, NIRS can be used to compare neural activity between
athletes and non-athletes.
In this review, the characteristics of athletes’ brains were described based on
research findings using neurophysiology and neuroimaging. Previous studies exam-
ining the relationship between sports performance and the brain were introduced
with a focus on golf putting, the neural substrate in golfers, and differences between
experts and non-experts. These methods provide valuable evidence in the ongoing
effort to develop a better understanding of motor control in humans.
References
Altenmüller E, Wiesendanger M, Kesselring J (eds) (2006) Music. Motor control and the brain.
Oxford University Press, Oxford
Ashburner J, Friston KJ (2000) Voxel-based morphometry–the methods. Neuroimage 11:805–821
Babiloni C, Del Percio C, Iacoboni M, Infarinato F, Lizio R, Marzano N, Crespi G, Dass u F,
Pirritano M, Gallamini M, Eusebi F (2008) Golf putt outcomes are predicted by sensorimotor
cerebral EEG rhythms. J Physiol 586:131–139
Babiloni C, Infarinato F, Marzano N, Iacoboni M, Dass u F, Soricelli A, Rossini PM, Limatola C,
Del Percio C (2011) Intra-hemispheric functional coupling of alpha rhythms is related to
golfer’s performance: a coherence EEG study. Int J Psychophysiol 82:260–268
Baumeister J, Reinecke K, Liesen H, Weiss M (2008) Cortical activity of skilled performance in a
complex sports related motor task. Eur J Appl Physiol 104:625–631
Baumeister J, Reinecke K, Cordes M, Lerch C, Weiss M (2010) Brain activity in goal-directed
movements in a real compared to a virtual environment using the Nintendo Wii. Neurosci Lett
481:47–50
Bezzola L, Mérillat S, Gaser C, Jäncke L (2011) Training-induced neural plasticity in golf novices.
J Neurosci 31:12444–12448
Bezzola L, Mérillat S, Jäncke L (2012) The effect of leisure activity golf practice on motor
imagery: an fMRI study in middle adulthood. Front Hum Neurosci 6:67
Bird EI (1987) Psychophysiological processes during rifle shooting. Int J Sport Psychol 18:9–18
Brown P (2000) Cortical drives to human muscle: the Piper and related rhythms. Prog Neurobiol
60:97–108
Brown S, Martinez MJ, Parsons LM (2006) The neural basis of human dance. Cereb Cortex
16:1157–1167
Craig CM, Delay D, Grealy MA, Lee DN (2000) Guiding the swing in golf putting. Nature
405:295–296
Crews DJ, Landers DM (1993) Electroencephalographic measures of attentional patterns prior to
the golf putt. Med Sci Sports Exerc 25:116–126
Gevins A, Smith ME, McEvoy L, Yu D (1997) High-resolution EEG mapping of cortical
activation related to working memory: effects of task difficulty, type of processing, and
practice. Cereb Cortex 4:374–385
González-Alonso J, Dalsgaard MK, Osada T, Volianitis S, Dawson EA, Yoshiga CC, Secher NH
(2004) Brain and central haemodynamics and oxygenation during maximal exercise in
humans. J Physiol Lond 557:331–342
Hatfield BD, Landers DM, Ray WJ (1984) Cognitive processes during self-paced motor perfor-
mance: an electroencephalographic profile of skilled marksmen. J Sport Psychol 6:42–59
Hatfield BD, Haufler AJ, Hung TM, Spalding TW (2004) Electroencephalographic studies of
skilled psychomotor performance. J Clin Neurophysiol 21:144–156
Hillman CH, Apparies RJ, Janelle CM, Hatfield BD (2000) An electrocortical comparison of
executed and rejected shots in skilled marksmen. Biol Psychol 52:71–83
Jäncke L, Koeneke S, Hoppe A, Rominger C, Hänggi J (2009) The architecture of the golfer’s
brain. PLoS One 4:e4785
Jasper HH (1958) The ten-twenty electrode system of the International Federation. Electroen-
cephalogr Clin Neurophysiol 10:371–375
Karni A, Meyer G, Rey-Hipolito C, Jezzard P, Adams MM, Turner R, Ungerleider LG (1998) The
acquisition of skilled motor performance: fast and slow experience-driven changes in primary
motor cortex. Proc Natl Acad Sci USA 95:861–868
Kerick SE, McDowell K, Hung T, Santa Maria DL, Spalding TW, Hatfield BD (2001) The role of
the left temporal region under the cognitive motor demands of shooting in skilled marksmen.
Biol Psychol 58:263–277
Kerick SE, Douglass L, Hatfield BD (2004) Cerebral cortical adaptations associated with
visuomotor practice. Med Sci Sports Exerc 36:118–129
Klimesch W (1999) EEG alpha and theta oscillations reflect cognitive and memory performance: a
review and analysis. Brain Res Brain Res Rev 29:169–195
Kranczioch C, Zich C, Schierholz I, Sterr A (2014) Mobile EEG and its potential to promote the
theory and application of imagery-based motor rehabilitation. Int J Psychophysiol 91:10–15
Landers DM, Petruzzello SJ, Salazar W, Kubitz KA, Gannon TL, Han M (1991) The influence of
electrocortical biofeedback on performance in pre-elite archers. Med Sci Sports Exerc
23:123–129
Landers DM, Han MW, Salazar W, Petruzzello SJ, Kubitz KA, Gannon TL (1994) Effects of
learning on electroencephalographic and electrocardiographic patterns in novice archers. Int J
Sport Psychol 25:313–330
Logothetis NK, Pauls J, Augath M, Trinath T, Oeltermann A (2001) Neurophysiological investi-
gation of the basis of the fMRI signal. Nature 412:150–157
Loze GM, Collins D, Holmes PS (2001) Pre-shot EEG alpha-power reactivity during expert
air-pistol shooting: a comparison of best and worst shots. J Sports Sci 19:727–733
Mann DT, Coombes SA, Mousseau MB, Janelle CM (2011) Quiet eye and the Bereitschaft-
spotential: visuomotor mechanisms of expert motor performance. Cogn Process 12:223–234
Milton J, Solodkin A, Hlustı́k P, Small SL (2007) The mind of expert motor performance is cool
and focused. Neuroimage 35:804–813
Mima T, Hallett M (1999) Corticomuscular coherence: a review. J Clin Neurophysiol 16:501–511
Onton J, Delorme A, Makeig S (2005) Frontal midline EEG dynamics during working memory.
Neuroimage 27:341–356
Perrey S (2008) Non-invasive NIR spectroscopy of human brain function during exercise.
Methods 45:289–299
Pfurtscheller G, Lopes da Silva FH (1999) Event-related EEG/MEG synchronization and
desynchronization: basic principles. Clin Neurophysiol 110:1842–1857
Radlo SJ, Steinberg GM, Singer RN, Barba DA, Melnikov A (2002) The influence of an
attentional focus strategy on alpha brain wave activity, heart rate, and dart-throwing. Int J
Sports Psychol 33:205–217
Reinecke K, Cordes M, Lerch C, Koutsandréou F, Schubert M, Weiss M, Baumeister J (2011)
From lab to field conditions: a pilot study on EEG methodology in applied sports sciences.
Appl Psychophysiol Biofeedback 36:265–271
12 H. Nakata
Ross JS, Tkach J, Ruggieri PM, Lieber M, Lapresto E (2003) The mind’s eye: functional MR
imaging evaluation of golf motor imagery. AJNR Am J Neuroradiol 24:1036–1044
Salazar W, Landers DM, Petruzzello SJ, Han MW, Crews DJ, Kubitz KA (1990) Hemispheric
asymmetry, cardiac response, and performance in elite archers. Res Q Exerc Sport 61:351–359
Sauseng P, Hoppe J, Klimesch W, Gerloff C, Hummel FC (2007) Dissociation of sustained
attention from central executive functions: local activity and interregional connectivity in the
theta range. Eur J Neurosci 2:587–593
Shibasaki H, Hallett M (2006) What is the Bereitschaftspotential? Clin Neurophysiol
117:2341–2356
Subudhi AW, Dimmen AC, Roach RC (2007) Effects of acute hypoxia on cerebral and muscle
oxygenation during incremental exercise. J Appl Physiol 103:177–183
Timinkul A, Kato M, Omori T, Deocaris CC, Ito A, Kizuka T, Sakairi Y, Nishijima T, Asada T,
Soya H (2008) Enhancing effect of cerebral blood volume by mild exercise in healthy young
men: a near-infrared spectroscopy study. Neurosci Res 61:242–248
Vickers JN, Adolphe RM (1997) Gaze behaviour during a ball tracking and aiming skill. Int J
Sports Vision 4:18–27
Williams AM, Singer RN, Frehlich SG (2002) Quiet eye duration, expertise, and task complexity
in near and far aiming tasks. J Mot Behav 34:197–207
Wong SW, Chan RH, Mak JN (2014) Spectral modulation of frontal EEG during motor skill
acquisition: a mobile EEG study. Int J Psychophysiol 91:16–21
Yarrow K, Brown P, Krakauer JW (2009) Inside the brain of an elite athlete: the neural processes
that support high achievement in sports. Nat Rev Neurosci 10:585–596
Chapter 2
Brain Activity During Motor Imagery
Nobuaki Mizuguchi
Abstract Motor imagery practice is useful for the acquisition of motor skills.
Understanding the neural mechanisms underlying motor imagery is important not
only for effective motor imagery practice but also for understanding the basic
mechanisms involved with motor control. It is well documented that brain activity
during motor imagery is similar to that which occurs during normal motor execu-
tion. This similarity supports the finding that motor skills can be acquired via motor
imagery training. In this chapter, I will summarize available information on the
brain activity that occurs during motor imagery.
Keywords Motor imagery • Mental practice • Brain imaging • Functional

magnetic resonance imaging • Corticospinal excitability
2.1 Introduction
Motor imagery is defined as the mental execution of a movement without any overt
movement or muscle activation (Lotze and Halsband 2006). Motor imagery has
been typically separated into two types “first person perspective imagery” and
“third person perspective imagery”. Brain activities are different for these two
types of imagery (Decety and Lindgren 1991; Munzert et al. 2008; Holmes and
Calmels 2008). Motor imagery practice is useful for athletes, patients and healthy
people in the acquisition of not only motor skills, but also muscle strength (Feltz
and Landers 1983; Yue and Cole 1992; Driskell et al. 1994; Murphy 1994; Pascual-
Leone et al. 1995; Stevens and Stoykov 2003; Mulder et al. 2004; Allami
et al. 2008; Holmes and Calmels 2008; Gentili et al. 2010; Guillot et al. 2010;
Mizuguchi et al. 2012a). These positive effects would be related to the fact that
neural substrate during motor imagery sheres motor execution.
N. Mizuguchi (*)
Faculty of Sport Sciences, Waseda University, Saitama, Japan
e-mail: mizuguchi@aoni.waseda.jp

14 N. Mizuguchi
2.2 Measurement of Brain Activity
The motor imagery is based on brain activity in many regions and along numerous
pathways. Techniques utilized to study imagery include functional magnetic reso-
nance imaging (fMRI), positron emission tomography (PET), transcranial magnetic
stimulation (TMS), magnetoencephalography (MEG), and electroencephalography
(EEG).
2.2.1 fMRI and PET Studies
Numerous studies on imagery have utilized fMRI and PET. As previously men-
tioned, brain activity during motor imagery largely overlaps that which occurs
during normal motor execution (Hanakawa et al. 2003; Lacourse et al. 2005;
Lotze and Halsband 2006; Imazu et al. 2007; Hanakawa et al. 2008; Guillot
et al. 2009;,Mizuguchi et al. 2013a). Areas activated during motor imagery include
the supplementary motor area (SMA), the premotor cortex (PM), the primary motor
cortex (M1), the parietal cortex, basal ganglia and the cerebellum (Decety
et al. 1994; Lotze et al. 1999; Naito et al. 2002; Stippich et al. 2002; Hanakawa
et al. 2003, 2008; Kuhtz-Buschbeck et al. 2003; Lacourse et al. 2005; Michelon
et al. 2006; Lotze and Halsband 2006; Szameitat et al. 2007a, b; Higuchi et al. 2007;
Imazu et al. 2007; Munzert et al. 2008; Chen et al. 2009; Guillot et al. 2009; Lorey
et al. 2010; Mizuguchi et al. 2013a). The most consistent findings involved activa-
tion of the SMA, PM, parietal cortex and cerebellum. These regions certainly play
major roles not only in normal motor execution but also in motor imagery. How-
ever, altered neural activity in many brain regions such as the precentral gyrus and
cerebellun is lower during motor imagery than during the activation of actual
movements (Hanakawa et al. 2003; Michelon et al. 2006). On the other hand, the
superior parietal cortex is activated more during motor imagery than motor execu-
tion (Hanakawa et al. 2003). Motor imagery has also been implicated in the
activation of other brain region, such as the dorsolateral prefrontal cortex
(DLPFC), the putamen and the insula (Hanakawa et al. 2003; Szameitat
et al. 2007a, b; Munzert et al. 2008; Guillot et al. 2009). However, the extent of
activation of these areas differs across studies. These differences might be related to
type of tasks imaged and subjects utilized.
2.2.2 TMS Studies
TMS has proven to be a valuable technique. TMS is performed by passing a high-

current pulse through a magnetic coil placed on the scalp. Motor evoked potentials
(MEPs) can be recorded from muscles activated by a single suprathreshold TMS
2 Brain Activity During Motor Imagery 15
pulse delivered to the contralateral primary motor cortex. The magnitude of the
MEPs depends on the level of excitability of the corticospinal tract (Rothwell 1991;
Hallett 2000; Reis et al. 2008).
Corticospinal excitability during motor imagery leads to increases above the
normal resting excitability level. Kasai and colleagues (1997) demonstrated that the
MEPs amplitude of the flexor carpi radialis muscle dramatically increased during
imagery of wrist flexion. Subsequent studies have confirmed the phenomenon
(Yahagi and Kasai 1998; Hashimoto and Rothwell 1999; Fadiga et al. 1999; Li
et al. 2004a). However, the increase in excitability is smaller than that which occurs
during actual movement activation (Li 2007). Corticospinal excitability during
motor imagery of a movement follows the same temporal changes in muscle
activities that occur during the actual execution of the movement (Hashimoto and
Rothwell 1999). In addition, this enhancement of corticospinal excitability is
related to the imagined force level (Mizuguchi et al. 2013b). This is another
example of how, during motor imagery, the brain shows an activity pattern very
similar to that which occurs during the elicitation of an actual execution.
Abbruzzese and colleagues (1999) as well as Liepert and Neveling (2009) observed
that during motor imagery, intracortical inhibition was decreased while
intracortical facilitation remained unchanged. This finding suggests that the
enhancement of corticospinal excitability that occurrs during motor imagery is
caused by disinhibition (Takemi et al. 2013).
The difference in corticospinal excitability across individuals corresponds
closely to differences in the vividness of their motor imagery as assessed by
questionnaire. That is, subjects who can imagine a movement more vividly would
also show an enhanced corticospinal excitability during motor imagery as com-
pared to subjects with low imagery ability (Fourkas et al. 2008; Lebon et al. 2012;
Williams et al. 2012).
2.2.3 MEG and EEG Studies
It is well known that internal and external events result in a change in many
frequency bands of ongoing EEG signals. Depending on the nature of the alteration
of the waveform, this is referred to as either event-related desynchronization (ERD)
or event-related synchronization (ERS) (see review by Pfurtscheller and Lopes da
Silva 1999). During motor imagery (kinesthetic imagery) of hand movement, ERD
of 8–13 Hz (alpha bands) and 14–30 Hz (beta bands) in the contralateral hemi-
sphere was observed (Nam et al. 2011). The ERD recorded from the MI associated
with downregulation of intracortical inhibition during motor imagery (Matsumoto
et al. 2010; Takemi et al. 2013).
Event-related potentials (ERP) have also been used in the investigation of motor
imagery (Naito and Matsumura 1994; Romero et al. 2000). ERP, which are
generally averaged EEG waveforms with respect to each event, reflect time-locked
changes in brain activity (Picton et al. 2000). Naito and Matsumura (1994)
16 N. Mizuguchi
measured a negative potential recorded at FCz in response to a visual cue with

motor execution, motor imagery or No-go response. They demonstrated that the
amplitude of the negative component was smaller in the imagery condition than in
the execution condition. By contrast, the amplitude in the No-go condition was
larger than that in the Go condition. Naito and Matsumuras’ findings suggest that
neural processes of involved with motor imagery differ from simple motor
suppression.
2.3 Role of Each Region
2.3.1 Supplementary Motor Area and Premotor Cortex
Most studies indicate that the SMA and PM are activated during motor imagery
(Decety et al. 1994; Lotze et al. 1999; Naito et al. 2002; Stippich et al. 2002;
Hanakawa et al. 2003, 2008; Kuhtz-Buschbeck et al. 2003; Lacourse et al. 2005;
Michelon et al. 2006; Lotze and Halsband 2006; Szameitat et al. 2007a, b; Higuchi
et al. 2007; Imazu et al. 2007; Munzert et al. 2008; Chen et al. 2009; Guillot
et al. 2009; Lorey et al. 2010; Mizuguchi et al. 2013a). Taken together, these
findings indicate the SMA and the PM are very likely to be essential parts of the
neuronal network involved with motor imagery. Indeed, activity in the SMA and
PM during imagery of the fingers, toes and tongue were activated in the area
corresponding to the each body part (Ehrsson et al. 2003). The same corresponding
areas were then activated during the actual execution. In addition, activity in the PM
correlated with the vividness of motor imagery as assessed by a questionnaire
utilizing a 7-point scale (Lorey et al. 2011).
By contrast, Kasess and colleagues (2008) investigated the effective connectiv-
ity during motor imagery utilizing fMRI and dynamic causal modeling analysis.
Their results suggest that the activity in the SMA suppresses the activity in the M1.
This conclusion indicates that the lack of activation in M1 during motor imagery
could be caused by suppression emanating from the SMA (Kasess et al. 2008). A
recent MEG study also reported that the activity in the SMA would inhibit the
activity in the M1 during motor imagery (Di Rienzo et al. 2014). Therefore, the
SMA apparently functions not only to generate motor representation but acts
suppress the M1 during motor imagery in order to inhibit muscle contraction.
2.3.2 Parietal Cortex
Sirigu and colleagues (1996) evaluated the contribution of the parietal cortex to
motor imagery by using mental chronometry in patients. They report that a patient
with lesions restricted to this area of the cortex was not able to imagine an action
accurately. According to Desmurget and colleagues (2009), stimulation of the

inferior parietal lobule produced intention to move or patients reported to have
moved, even in the absence of actual motor responses. Single-pulse TMS over the
parietal cortex (superior parietal cortex), which can temporarily deactivate neurons
in healthy subjects (virtual lesion) (Pascual-Leone et al. 2000), decreased the
accuracy of motor imagery (Fleming et al. 2010). In this experiment, to assess the
accuracy of motor imagery, subjects were asked to imagine a sequence of hand and
arm movements given by verbal instruction, and then asked to tell the investigator
the final position of the hand and arm. Subjects answered correctly if they were able
to imagine the movements as the instructions progressed. These results would
indicate that the parietal cortex plays an important role for the recognition of
imagined action.
Recently, cortico-cortical connectivity between parietal and bilateral primary
motor cortices was investigated during motor imagery by using a combination of
TMS and transcranial direct current stimulation (tDCS) (Feurra et al. 2011). During
motor imagery, anodal tDCS over the parietal cortex increased MEPs amplitude,
but only from the ipsilateral M1 and not from the contralateral M1. These results
suggest that a major role in motor imagery is restricted to the ipsilateral parieto-
motor circuitry.
2.3.3 Primary Motor Cortex
While some researchers found M1 activation during motor imagery (Porro

et al. 1996; Lotze et al. 1999; Chen et al. 2009; Guillot et al. 2009), others did
not (Decety et al. 1994; Naito et al. 2002; Kuhtz-Buschbeck et al. 2003; Higuchi
et al. 2007; Szameitat et al. 2007a). This discrepancy might be associated with such
factors as the degree of muscle activity, type of tasks, and subjects (Munzert
et al. 2009). Kuhtz-Buschbeck and colleagues (2003) investigated brain activity
and corticospinal excitability using fMRI and TMS during motor imagery. They
found a significant enhancement of corticospinal excitability with TMS, but not
significant activation in M1 when utilizing fMRI. These findings indicated a
possibility that sensitivity for the detection of neural activation, especially in the
M1, was higher for TMS than fMRI.
The effect of physical practice on motor performance is higher 30 min after
practice than right after practice (Debarnot et al. 2011). That is, skill acquisition
occurs not only during physical practice but continue after the practice is over. This
phenomenon is called “early consolidation” (Muellbacher et al. 2002). Early
consolidation was also observed 30 min after a motor imagery practice (Debarnot
et al. 2011). Inhibition of neuronal activity in M1 by a continuous theta-burst
stimulation (cTBS) immediately after motor imagery practice blocked the early
consolidation. These findings suggest that neuronal activity in M1 is involved in the
motor learning that accrues to imagery practice.
18 N. Mizuguchi
2.3.4 Spinal Cord
It has been reported that motor imagery did not modulate spinal excitability as
assessed by the H reflex (Abbruzzese et al. 1996; Aoyama and Kaneko 2011) or F
wave (Facchini et al. 2002). By contrast, the stretch reflex was augmented during
motor imagery (Li et al. 2004b; Aoyama and Kaneko 2011). Aoyama and Kaneko
(2011) clarified the effect on this potential by demonstrating that the amplitude of
the stretch reflex was increased during motor imagery of dorsiflexion and
plantarflexion, but the amplitude of H-reflex was not altered. Thus, while motor
imagery seems to increase the excitability of some spinal reflexes, others are
unaffected. At present, the nature of spinal cord involvement with motor imagery
is unclear.
2.4 Influence of Somatosensory and Visual Information

on Motor Imagery
Somatosensory input interacts with motor imagery (Naito et al. 2002; Thyrion and
Roll 2009). Naito and colleagues (2002) report that the kinesthetic illusion elicited
by tendon vibration (proprioceptive input) of the wrist extensor is enhanced by
motor imagery of wrist flexion. The actual hand posture also affects brain activity
during motor imagery (Vargas et al. 2004; Fourkas et al. 2006; Linag et al. 2007).
For example, Vargas and colleagues (2004) showed that corticospinal excitability
during imagery of finger-thumb opposition was larger when the actual posture was
the same as the imagined (congruent) hand than when it was different from the
imagined (incongruent) hand. This result suggests that proprioceptive inputs which
are congruent with imagined actions enhance corticospinal excitability during
motor imagery. However, the effect of posture on motor imagery seems to depend
on the type of motor imagery (kinesthetic or visual), because posture has been
found to affect corticospinal excitability during kinesthetic imagery, but not during
visual imagery (Fourkas et al. 2006).
In many sports various tools and objects such as balls, gloves, bats and rackets
are used. In the process of manipulating the objects, tactile information provides an
important input. In such motor imagery, tactile input generated by passively
touching an object was found to increase corticospinal excitability (Mizuguchi
et al. 2009, 2011, 2012b). This enhancement of corticospinal excitability was not
merely the effect of just holding something, in this case a ball, because corticospinal
excitability did not increase while holding the object without motor imagery.
Further, when the actual posture was different from the imagined action, tactile
input did not increase corticospinal excitability (Mizuguchi et al. 2011). These
results suggest that corticospinal excitability during imagery with an object is
affected by the actual touching of the object and is thus modulated via combination
of tactile and proprioceptive inputs.
Visual information also influences corticospinal excitability during motor imag-

ery. Sakamoto and colleagues (2009) demonstrated that corticospinal excitability
during simultaneous observation and imagery of elbow flexion was facilitated as
compared to the excitability that occurred during observation or imagery alone.
However, facilitation due to the combination of observation and imagery was not
obtained when the participants imagined an action that was out of phase with the
actual observation. Therefore, visual input that is congruent with an imagined
action is likely to increase corticospinal excitability during motor imagery. This
evidence suggests that somatosensory and visual inputs which temporally closely
resemble the actual execution increase the effectiveness of motor imagery. This
implies that appropriate input selection is critical in maximizing the effectiveness
of motor imagery in training or rehabilitation.
References
Abbruzzese G, Trompetto C, Schieppati M (1996) The excitability of the human motor cortex
increases during execution and mental imagination of sequential but not repetitive finger
movements. Exp Brain Res 111(3):465–472
Abbruzzese G, Assini A, Buccolieri A, Marchese R, Trompetto C (1999) Changes of intracortical
inhibition during motor imagery in human subjects. Neurosci Lett 263(2–3):113–116
Allami N, Paulignan Y, Brovelli A, Boussaoud D (2008) Visuo-motor learning with combination
of different rates of motor imagery and physical practice. Exp Brain Res 184(1):105–113
Aoyama T, Kaneko F (2011) The effect of motor imagery on gain modulation of the spinal reflex.
Brain Res 1372:41–48
Chen H, Yang Q, Liao W, Gong Q, Shen S (2009) Evaluation of the effective connectivity of
supplementary motor areas during motor imagery using Granger causality mapping.
Neuroimage 47(4):1844–1853
Debarnot U, Clerget E, Olivier E (2011) Role of the primary motor cortex in the early boost in
performance following mental imagery training. PLoS One 6(10):e26717
Decety J, Lindgren M (1991) Sensation of effort and duration of mentally executed actions. Scand
J Psychol 32(2):97–104
Decety J, Perani D, Jeannerod M, Bettinardi V, Tadary B, Woods R, Mazziotta JC, Fazio F (1994)
Mapping motor representations with positron emission tomography. Nature 371(6498):
600–602
Desmurget M, Reilly KT, Richard N, Szathmari A, Mottolese C, Sirigu A (2009) Movement
intention after parietal cortex stimulation in humans. Science 324:811–813
Di Rienzo F, Guillot A, Daligault S, Delpuech C, Rode G, Collet C (2014) Motor inhibition during
motor imagery: a MEG study with a quadriplegic patient. Neurocase 20(5):524–539
Driskell J, Copper C, Moral A (1994) Does mental practice enhance performance. J Appl Psychol
79:481–492
Ehrsson HH, Geyer S, Naito E (2003) Imagery of voluntary movement of fingers, toes, and tongue
activates corresponding body-part-specific motor representations. J Neurophysiol 90(5):
3304–3316
Facchini S, Muellbacher W, Battaglia F, Boroojerdi B, Hallett M (2002) Focal enhancement of
motor cortex excitability during motor imagery: a transcranial magnetic stimulation study.
Acta Neurol Scand 105(3):146–151
20 N. Mizuguchi
Fadiga L, Buccino G, Craighero L, Fogassi L, Gallese V, Pavesi G (1999) Corticospinal excit-

ability is specifically modulated by motor imagery: a magnetic stimulation study. Neuropsy-
chologia 37(2):147–158
Feltz DL, Landers DM (1983) The effects of mental practice on motor skill learning and
performance: a meta-analysis. J Sport Psychol 5:25–57
Feurra M, Bianco G, Polizzotto NR, Innocenti I, Rossi A, Rossi S (2011) Cortico-cortical
connectivity between right parietal and bilateral primary motor cortices during imagined and
observed actions: a combined TMS/tDCT study. Front Neural Circ 5: Article 10
Fleming MK, Stinear CM, Byblow WD (2010) Bilateral parietal cortex function during motor
imagery. Exp Brain Res 201(3):499–508
Fourkas AD, Ionta S, Aglioti SM (2006) Influence of imagined posture and imagery modality on
corticospinal excitability. Behav Brain Res 168(2):190–196
Fourkas AD, Bonavolonta V, Avenanti A, Aglioti SM (2008) Kinesthetic imagery and tool-
specific modulation of corticospinal representations in expert tennis players. Cereb Cortex
18(10):2382–2390
Gentili R, Han CE, Schweighofer N, Papaxanthis C (2010) Motor learning without doing: trial-by-
trial improvement in motor performance during mental training. J Neurophysiol 104(2):
774–783
Guillot A, Collet C, Nguyen VA, Malouin F, Richards C, Doyon J (2009) Brain activity during
visual versus kinesthetic imagery: an fMRI study. Hum Brain Mapp 30(7):2157–2172
Guillot A, Tolleron C, Collet C (2010) Does motor imagery enhance stretching and flexibility. J
Sports Sci 28(3):291–298
Hallett M (2000) Transcranial magnetic stimulation and the human brain. Nature 406(6792):
147–150
Hanakawa T, Immisch I, Toma K, Dimyan MA, Van Gelderen P, Hallett M (2003) Functional
properties of brain areas associated with motor execution and imagery. J Neurophysiol
89(2):989–1002
Hanakawa T, Dimyan MA, Hallett M (2008) Motor planning, imagery, and execution in the
distributed motor network: a time-course study with functional MRI. Cereb Cortex
18(12):2775–2788
Hashimoto R, Rothwell JC (1999) Dynamic changes in corticospinal excitability during motor
imagery. Exp Brain Res 125(1):75–81
Higuchi S, Imamizu H, Kawato M (2007) Cerebellar activity evoked by common tool-use
execution and imagery tasks: an fMRI study. Cortex 43(3):350–358
Holmes P, Calmels C (2008) A neuroscientific review of imagery and observation use in sport. J
Mot Behav 40(5):433–445
Imazu S, Sugio T, Tanaka S, Inui T (2007) Differences between actual and imagined usage of
chopsticks: an fMRI study. Cortex 43(3):301–307
Kasai T, Kawai S, Kawanishi M, Yahagi S (1997) Evidence for facilitation of motor evoked
potentials (MEPs) induced by motor imagery. Brain Res 744(1):147–150
Kasess CH, Windischberger C, Cunnington R, Lanzenberger R, Pezawas L, Moser E (2008) The
suppressive influence of SMA on M1 in motor imagery revealed by fMRI and dynamic causal
modeling. Neuroimage 40(2):828–837
Kuhtz-Buschbeck JP, Mahnkopf C, Holzknecht C, Siebner H, Ulmer S, Jansen O (2003) Effector-
independent representations of simple and complex imagined finger movements: a combined
fMRI and TMS study. Eur J Neurosci 18(12):3375–3387
Lacourse MG, Orr EL, Cramer SC, Cohen MJ (2005) Brain activation during execution and motor
imagery of novel and skilled sequential hand movements. Neuroimage 27(3):505–519
Lebon F, Byblow WD, Collet C, Guillot A, Stinear CM (2012) The modulation of motor cortex
excitability during motor imagery depends on imagery quality. Eur J Neurosci 35(2):323–331
Li S (2007) Movement-specific enhancement of corticospinal excitability at subthreshold levels
during motor imagery. Exp Brain Res 179(3):517–524
Li S, Latash ML, Zatsiorsky VM (2004a) Effects of motor imagery on finger force responses to
transcranial magnetic stimulation. Brain Res Cogn Brain Res 20(2):273–280
Li S, Kamper DG, Stevens JA, Rymer WZ (2004b) The effect of motor imagery on spinal
segmental excitability. J Neurosci 24(43):9674–9680
Liang N, Ni Z, Takahashi M, Murakami T, Yahagi S, Funase K, Kato T, Kasai T (2007) Effects of
motor imagery are dependent on motor strategies. Neuroreport 18(12):1241–1245
Liepert J, Neveling N (2009) Motor excitability during imagination and observation of foot
dorsiflexions. J Neural Transm 116(12):1613–1619
Lorey B, Pilgramm S, Walter B, Stark R, Munzert J, Zentgraf K (2010) Your mind’s hand: motor
imagery of pointing movements with different accuracy. Neuroimage 49(4):3239–3247
Lorey B, Pilgramm S, Bischoff M, Stark R, Vaitl D, Kindermann S, Munzert J, Zentgraf K (2011)
Activation of the parieto-premotor networks is associated with vivid motor imagery – a
parametric fMRI study. PLoS One 6(5):e20368
Lotze M, Halsband U (2006) Motor imagery. J Physiol Paris 99(4–6):386–395
Lotze M, Montoya P, Erb M, Hulsmann E, Flor H, Klose U, Birbaumer N, Grodd W (1999)
Activation of cortical and cerebellar motor areas during executed and imagined hand move-
ment: an fMRI study. J Cogn Neurosci 11(5):491–501
Matsumoto J, Fujiwara T, Takahashi O, Liu M, Kimura A, Ushiba J (2010) Modulation of mu
rhythm desynchronization during motor imagery by transcranial direct current stimulation. J
Neuroeng Rehabil 7:27
Michelon P, Vettel JM, Zacks JM (2006) Lateral somatotopic organization during imagined and
prepared movements. J Neurophysiol 95(2):811–822
Mizuguchi N, Sakamoto M, Muraoka T, Kanosue K (2009) Influence of touching an object on
corticospinal excitability during motor imagery. Exp Brain Res 196(4):529–535
Mizuguchi N, Sakamoto M, Muraoka T, Nakagawa K, Kanazawa S, Nakata H, Moriyama N,
Kanosue K (2011) The modulation of corticospinal excitability during motor imagery of
actions with objects. PLoS One 6(10):e26006
Mizuguchi N, Nakata H, Uchida Y, Kanosue K (2012a) Motor imagery and sport performance. J
PFSM 1:103–111
Mizuguchi N, Sakamoto M, Muraoka T, Moriyama N, Nakagawa K, Nakata H, Kanosue K
(2012b) Influence of somatosensory input on corticospinal excitability during motor imagery.
Neurosci Lett 514(1):127–130
Mizuguchi N, Nakata H, Hayashi T, Sakamoto M, Muraoka T, Uchida Y, Kanosue K (2013a)
Brain activity during motor imagery of an action with an object: a functional magnetic
resonance imaging study. Neurosci Res 76(3):150–155
Mizuguchi N, Umehara I, Nakata H, Kanosue K (2013b) Modulation of corticospinal excitability
dependent upon imagined force level. Exp Brain Res 230(2):243–249
Muellbacher W, Ziemann U, Wissel J, Dang N, Kofler M, Facchini S, Boroojerdi B, Poewe W,
Hallett M (2002) Early consolidation in human primary motor cortex. Nature 415(6872):
640–644
Mulder T, Zijistra S, Zijlstra W, Hochstenbach J (2004) The role of motor imagery in learning a
totally novel movement. Exp Brain Res 154(2):211–217
Munzert J, Zentgraf K, Stark R, Vaitl D (2008) Neural activation in cognitive motor processes:
comparing motor imagery and observation of gymnastic movements. Exp Brain Res 188(3):
437–444
Munzert J, Lorey B, Zentgraf K (2009) Cognitive motor processes: the role of motor imagery in
the study of motor representations. Brain Res Rev 60(2):306–326
Murphy SM (1994) Imagery interventions in sport. Med Sci Sports Exerc 26(4):486–494
Naito E, Matsumura M (1994) Movement-related slow potentials during motor imagery and motor
suppression in humans. Brain Res Cogn Brain Res 2(2):131–137
Naito E, Kochiyama T, Kitada R, Nakamura S, Matsumura M, Yonekura Y, Sadato N (2002)
Internally simulated movement sensations during motor imagery activate cortical motor areas
and the cerebellum. J Neurosci 22(9):3683–3691
22 N. Mizuguchi
Nam CS, Jeon Y, Kim YJ, Lee I, Park K (2011) Movement imagery-related lateralization of event-
related (de)synchronization (ERD/ERS): motor-imagery duration effects. Clin Neurophysiol
122(3):567–577
Pascual-Leone A, Nquyet D, Cohen LG, Brasil-Neto JP, Cammarota A, Hallett M (1995) Mod-
ulation of muscle responses evoked by transcranial magnetic stimulation during the acquisition
of new fine motor skills. J Neurophysiol 74(3):1037–1045
Pascual-Leone A, Walsh V, Rothwell JC (2000) Transcranial magnetic stimulation in cognitive
neuroscience – virtual lesion, chronometry, and functional connectivity. Curr Opin Neurobiol
10(2):232–237
Pfurtscheller G, Lopes da Silva FH (1999) Event-related EEG/MEG synchronization and
desynchronization: basic principles. Clin Neurophysiol 110(11):1842–1857
Picton TW, Bentin S, Berg P, Donchin E, Hillyard SA, Johnson R Jr, Miller GA, Ritter W,
Ruchkin DS, Rugg MD, Taylor MJ (2000) Guidelines for using human event-related potentials
to study cognition: recording standards and publication criteria. Psychophysiology 37(2):
127–152
Porro CA, Francescato MP, Cettolo V, Diamond ME, Baraldi P, Zuiani C, Bazzocchi M, di
Prampero PE (1996) Primary motor and sensory cortex activation during motor performance
and motor imagery: a functional magnetic resonance imaging study. J Neurosci 16(23):
7688–7698
Reis J, Swayne OB, Vandermeeren Y, Camus M, Dimyan MA, Harris-Love M, Perez MA,
Ragert P, Rothwell JC, Cohen LG (2008) Contribution of transcranial magnetic stimulation
to the understanding of cortical mechanisms involved in motor control. J Physiol 586(2):
325–351
Romero DH, Lacourse MG, Lawrence KE, Schandler S, Cohen MJ (2000) Event-related potentials
as a function of movement parameter variations during motor imagery and isometric action.
Behav Brain Res 117:83–96
Rothwell JC (1991) Physiological studies of electric and magnetic stimulation of the human brain.
Electroencephalogr Clin Neurophysiol Suppl 43:29–35
Sakamoto M, Muraoka T, Mizuguchi M, Kanosue K (2009) Combining observation and imagery
of an action enhances human corticospinal excitability. Neurosci Res 65(1):23–27
Sirigu A, Duhamel JR, Cohen L, Pillon B, Dubois B, Agid Y (1996) The mental representation of
hand movements after parietal cortex damage. Science 273(5264):1564–1568
Stevens JA, Stoykov ME (2003) Using motor imagery in the rehabilitation of hemiparesis. Arch
Phys Med Rehabil 84(7):1090–1092
Stippich C, Ochmann H, Sartor K (2002) Somatotopic mapping of the human primary sensori-
motor cortex during motor imagery and motor execution by functional magnetic resonance
imaging. Neurosci Lett 331(1):50–54
Szameitat AJ, Shen S, Sterr A (2007a) Effector-dependent activity in the left dorsal premotor
cortex in motor imagery. Eur J Neurosci 26(11):3303–3308
Szameitat AJ, Shen S, Sterr A (2007b) Motor imagery of complex everyday movements. An fMRI
study. Neuroimage 34(2):702–713
Takemi M, Masakado Y, Liu M, Ushiba J (2013) Event-related desynchronization reflects
downregulation of intracortical inhibition in human primary motor cortex. J Neurophysiol
110(5):1158–1166
Thyrion C, Roll JP (2009) Perceptual integration of illusory and imagined kinesthetic images. J
Neurosci 29(26):8483–8492
Vargas CD, Olivier E, Craighero L, Fadiga L, Duhamel JR, Sirigu A (2004) The influence of hand
posture on corticospinal excitability during motor imagery: a transcranial magnetic stimulation
study. Cereb Cortex 14(11):1200–1206
Williams J, Pearce AJ, Loporto M, Morris T, Holmes P (2012) The relationship between
corticospinal excitability during motor imagery and motor imagery ability. Behav Brain Res
226(2):369–375
Yahagi S, Kasai T (1998) Facilitation of motor evoked potentials (MEPs) in first dorsal
interosseous (FDI) muscle is dependent on different motor images. Electroencephagr Clin
Neurophysiol 109(5):409–417
Yue G, Cole KJ (1992) Strength increases from the motor program: comparison of training with
maximal voluntary and imagined muscle contractions. J Neurophysiol 67(5):1114–1123
Chapter 3
Brain Oscillations and Athletic Performance
Andreas Mierau, Thorben Hülsdünker, and Heiko K. Strüder
Abstract Large populations of synchronously active cortical neurons produce

oscillations which can be measured on the surface of the scalp using electroen-
cephalography (EEG). These cortical oscillations can be analysed in the frequency
domain. Power in different frequency bands (e.g. delta, theta, alpha, beta and
gamma) has been shown to correlate with specific perceptual, sensorimotor or
cognitive processes. Therefore, cortical oscillations can be used to better under-
stand how the cortex is involved during accomplishment of different tasks. This
chapter discusses research addressing the relationship between cortical oscillations
and sport performance in ecologically valid settings (Sect. 3.1). Since balance
ability has been suggested to play a crucial role for performance in various sports,
this chapter also reviews studies describing brain activity during balance control
(Sect. 3.2). For a more complete analysis, experiments where EEG neurofeedback
training was applied to enhance sport performance will also be taken into consid-
eration (Sect. 3.3). Finally, some concluding remarks and suggestions for future
research will be outlined (Sect. 3.4).
Keywords EEG • Cortex • Sport • Expert • Golf • Shooting
Athletes have the ability to achieve high levels of sensorimotor skills following
years of extensive training. The development of such expert performance levels is
associated with a substantial reorganisation of the central nervous system. A
growing number of studies suggest that the athletes’ brain is characterised by
specific adaptations that facilitate information processing relevant to the sport
(Nakata et al. 2010). Research in this field has important fundamental and applied
implications as it demonstrates the extent of brain plasticity and helps to advance
interventions to enhance performance. A good technique to study brain function in
human sport and exercise science research is electroencephalography (EEG). It can
be applied without risk in ecologically valid settings and provides excellent tem-
poral resolution. The raw EEG signal can be measured on the scalp and it contains
A. Mierau (*) • T. Hülsdünker • H.K. Strüder

Institute of Movement and Neurosciences, German Sport University Cologne, Cologne,
Germany
e-mail: mierau@dshs-koeln.de

26 A. Mierau et al.
both, amplitude and frequency information. When analysing data in the time
domain, event-related potentials (ERPs) can be investigated. These represent tem-
porally distinct changes in absolute electrical activity induced or evoked by external
stimuli or events. Since cortical activity is oscillatory in nature, the frequency
information embedded in the signal can be extracted via spectral analysis as Fourier
transformation and changes in spectral power can be determined. Typically such
analyses include the well-known oscillatory components; delta (~0.5–3 Hz), theta
(~4–7 Hz), alpha (~8–12 Hz), beta (~13–30 Hz) and gamma (>30 Hz). This chapter
focuses on research from our own laboratory, as well as other studies addressing
EEG spectral power associated with skilled sensorimotor performance in sports. As
balance ability is discussed as a potential key determinant for performance in many
sports (Hrysomallis 2011; Zemková 2013), this chapter also includes a section
dealing with brain activity during balance control. The penultimate part of the
chapter outlines research aiming to enhance athletic performance through the
altering of individual’s EEG; also known as neurofeedback training. The chapter
closes with some conclusions and suggestions for future research.
3.1 Cortical Oscillations and Sport Performance

in Ecologically Valid Settings
Probably the most extensively studied cortical rhythm with regard to sport perfor-
mance is alpha. Early EEG studies conducted in the 1980s and 1990s focused on
alpha power during the preparation of self-paced visuomotor tasks in skilled
marksman (Hatfield and Landers 1987) and archers (Salazar et al. 1990). A prom-
inent feature of these studies is a progressive increase of alpha power in the left
temporal hemisphere (scalp site T3), while alpha power in the right temporal
hemisphere (scalp site T4) remained unchanged closer towards the trigger pull/
arrow release (reviewed in Hatfield et al. 2004). As alpha power typically decreases
(desynchronises) with increased cortical information processing, an increase (syn-
chronisation) in alpha power has been suggested to reflect a reduction in cortical
activation (Andreassi 2007). Therefore, the above findings were interpreted to
indicate a deactivation of the left hemisphere during aiming. Studies examining
the functional differences between the two cortical hemispheres suggest that visuo-
spatial information is processed mainly in the right hemisphere, whereas the left
hemisphere is involved in verbal-analytic processing (e.g. Reeves 1983). Subse-
quently, it has been hypothesised that skilled aiming performance is characterised
by reduction of verbal-analytic processes (Hatfield et al. 2004). This hypothesis was
further supported by later research demonstrating that during shot preparation right-
handed expert marksmen exhibited greater alpha power in the left hemisphere,
when compared to right-handed novice shooters (Haufler et al. 2000). Also, alpha
power increased during preparation of the arrow release at the left temporal site, but
remained unchanged at the right temporal site following archery (Landers
3 Brain Oscillations and Athletic Performance 27
et al. 1994) and marksman (Kerick et al. 2004) training. Based on the well-
established stage structure of sensorimotor skill acquisition (Fitts and Posner
1967), it was suggested that the observed decrease in cortical activation with
increasing skill level, reflects a reduction of cognitive effort due to the automaticity
of information processing. This should result in minimising potential interference
with visuomotor processes during motor control and is consistent with the “neural
efficiency” hypothesis (Hatfield and Hillman 2001).
In a more recent study Del Percio et al. (2009) used a high density 56 channel
EEG array to analyse event-related desynchronisation/synchronisation in the alpha
frequency band. The study measured the last 3 s of aiming in right-handed expert
and novice pistol shooters. With reference to a baseline period, during the aiming
process novice shooters exhibited a larger alpha band desynchronisation over the
whole scalp, compared to that of expert shooters. In addition, a significant differ-
ence in cortical activation was found between high and low scores in experts, but
not in novices. Specifically, high-frequency alpha (10–12 Hz) desynchronised
during aiming in low scores, but synchronised in high scores at the left central
(contralateral) and right parietal (visuo-attentional) regions. The cumulative pattern
of these results was suggested to indicate that superior performance is related to a
global reduction of cortical activation, as a possible index of spatially selective
cortical processes and thus, neural efficiency (Del Percio et al. 2009). This view
however, is challenged by studies using other sensorimotor tasks. For example, it
has been shown that in expert golfers high-frequency alpha desynchronisation was
more pronounced during preparation of successful compared to unsuccessful golf
putts over the fronto-central midline (Fz and Cz), as well as over the arm and hand
region of the right sensorimotor area (C4). Moreover, desynchronisation of high-
frequency alpha was inversely related to the error of the unsuccessful putts
(i.e. distance from the hole) (Babiloni et al. 2007). Recently, Cooke (2013)
hypothesised that the different patterns of EEG alpha power that precede shooting
and golf putting may reflect different attentional demands for the tasks (Cooke
2013). For example, marksmen are required to allocate a substantial amount of
attentional resources to control posture and balance (Ball et al. 2003), whereas
postural stability has been shown to be unrelated to golf putting performance
(Babiloni et al. 2007). Beyond that, golf putting may require cortical excitation to
execute the putting movement, whereas shooting may require a substantial amount
of cortical inhibition to minimize body sway and aim point fluctuations. As a
consequence, sensorimotor alpha power should decrease during putting but
increase during shooting. The latter is suggested because alpha synchronisation
has been reported to play a major role for the implementation of inhibitory control
(Hummel et al. 2002; Klimesch 2012). In line with this, it has been shown that a
motor evoked potential was diminished when alpha power immediately preceded
the magnetic pulse of transcranial magnetic stimulation (TMS) was high, and vice
versa (Sauseng et al. 2009). This effect was only found for local EEG alpha activity,
at sites overlying the cortical motor areas to which the TMS pulses were applied.
These data provide evidence that the magnitude of motor cortical excitability is
determined by the amount of topographically specific alpha oscillations in the
28 A. Mierau et al.
sensorimotor cortex. Therefore, alpha synchronisation over the contralateral sen-

sorimotor arm/hand region during aiming in high scores when compared to low
scores, as found by Del Percio et al. (2009), may indicate motor inhibition.
Consistent with this, in a recent study we found reduced gamma power at the
contralateral sensorimotor arm/hand region (C3) in expert shooters when compared
to novices, as well as in high compared to low shooting scores. Moreover, senso-
rimotor gamma power correlated positively with aim point fluctuations and nega-
tively with shooting score (Mierau et al. submitted for publication). This is in good
agreement with the observation that increasing movement amplitude (and velocity)
is associated with increased contralateral sensorimotor cortex gamma synchronisa-
tion during finger movements (Muthukumaraswamy 2010). Gamma band synchro-
nisation has been suggested to have a functional role in movement control probably
through the integration and processing of afferent feedback (Szurhaj et al. 2005).
Therefore, reduced aim point fluctuations, as typically found during high shooting
scores, will produce less afferent signals and thus, less gamma activity.
It has been shown that attentional and working memory processes are usually
accompanied by an increase in frontal midline theta oscillations (Fmθ) (Sauseng
et al. 2009; Gevins et al. 1997; Nakashima and Sato 1993; Onton et al. 2005).
Functional magnetic resonance imaging (fMRI) revealed that attentional focus in
working memory is associated with activation of the prefrontal cortical regions, the
anterior cingulate cortex (ACC) and superior parietal areas (Osaka et al. 2007).
Consistent with this, it has been suggested that the generators of Fmθ are most
likely located in the in the medial prefrontal areas and/or the ACC (Gevins
et al. 1997; Asada et al. 1999). Doppelmayr et al. (2008) studied the time course
of Fmθ during the aiming period in expert and novice rifle shooters. It was found
that the time course during the aiming period was significantly different between
groups. Specifically, Fmθ gradually increased as trigger pull approached in experts,
but not in novices. Source analyses (LORETA) indicated a significantly stronger
theta activity for experts strictly located at the anterior cingulate area and the medial
frontal cortex. These results were interpreted to reflect the ability of experts to
increase attentional focus right until the moment of trigger pull, whereas novices do
not have this ability (Doppelmayr et al. 2008). These findings are in agreement with
a previous study reporting higher power in a fraction of the theta range (6–7 Hz)
during aiming in marksmen compared to novice shooters (Haufler et al. 2000). In
addition, Baumeister et al. (2008) reported increased theta power during golf
putting in experienced golfers compared to novices (Baumeister et al. 2008). In a
recent study, Kao et al. (2013) used a within-subject design and compared theta
power between the best and worst golf putts in skilled golfers. Their analyses
revealed that theta power significantly increased during the last 3 s before back-
swing. More importantly however, the Fmθ power was significantly reduced when
comparing successful and unsuccessful putts. From these results, it was suggested
that the appearance of Fmθ before backswing is a basic component of skilled
performance, probably representing successful engagement of top-down sustained
attention. However, higher Fmθ during preparation of unsuccessful compared to
successful putts was interpreted to reflect an excessive amount of attentional
engagement; this may be associated with elevated volitional control and thus, less
automatic performance (Kao et al. 2013).
3.2 Brain Activity During Balance Control
Recent literature reviews suggest that balance ability is one of the limiting factors
for performance in a range of sports (Hrysomallis 2011; Zemková 2013). However,
the current understanding of supraspinal mechanisms underlying balance control is
rather poor. Different attempts have evolved in neuroscience research to identify
brain areas that are related to balance control. Ouchi et al. (1999) conducted a study
using a mobile gantry positron emission tomography (PET) system to investigate
the neural substrates of postural control. They found increased activity in the
cerebellar anterior vermis, as well as the posterior lobe lateral cortex ipsilateral to
the weight-bearing side during standing on one foot when compared to a supine
position. Furthermore, standing in tandem was accompanied by activation within
the visual association cortex, the anterior and posterior vermis, as well as within the
midbrain (Ouchi et al. 1999). Additionally, when using fMRI, successful recogni-
tion of unstable postures induced activation of distinct areas of the brain. These
include bilateral parietal cortex, anterior cingulate cortex and bilateral cerebellum
(Slobounov et al. 2006). Unfortunately, in the case of fMRI, experimental protocols
are restricted to imaginary tasks or action observation while actual corrective
movement execution in the context of transient or persistent disruptions of balance
is not possible. In addition, both PET and fMRI are characterised by a relatively low
temporal resolution and brain activity is only indirectly measured by assessing the
energy metabolism or the blood oxygen level dependent contrast. Therefore, these
techniques are not capable of identifying rapid changes in balance-related brain
dynamics. When balance is disrupted, postural adjustments have to be initiated
within milliseconds to maintain balance and to avoid falling. Afferent sensory
signals were detected in the primary sensory cortex after only 40–50 ms following
perturbation (Jacobs and Horak 2007). It has been suggested that muscular output
executed in a time frame of up to 85 ms following perturbation is primarily
regulated by spinal loops, while efferent responses above 85 ms could well be
influenced by cortical structures via direct monosynaptic projections (Taube 2006).
Such neural dynamics are probably best studied with techniques allowing high
temporal resolution such as EEG or MEG. In addition, EEG can be used in
ecologically valid settings. Consequently, there is an increasing amount of studies
using EEG during transient balance tasks (Jacobs et al. 2008; Mochizuki et al. 2008;
Slobounov et al. 2005, 2008, 2009, 2013). When balance was disrupted
unpredictably, negative potentials (N1) occurred with an onset latency of about
90 ms in frontal, central and parietal regions of the cortex. These potentials were
most pronounced over the central midline electrodes overlying the sensorimotor
leg/foot region (FCz, Cz) but were also recorded at frontal (Fz) and parietal
(Pz) sites (Adkin et al. 2006). Furthermore, research on EEG spectral analysis
30 A. Mierau et al.
revealed increased power in the gamma frequency band (>30 Hz), which was most
pronounced over central sites (C3, Cz, C4), when corrective movements were
initiated to resume a stable position (Slobounov et al. 2008). Finally, when contin-
uous theta burst stimulation (cTBS) is applied to the primary motor cortex, com-
pensatory reactions following perturbation are distorted (Bolton et al. 2012). These
studies clearly indicate the involvement of primary motor areas during balance
control. However, when balance is perturbed an “error detection system” is
required that might be defined as the sensory part of balance control to identify
postural instability. There is growing evidence in the literature suggesting that the
ACC may serve as such an error detection system or may at least be one component
of it (Jacobs and Horak 2007; Slobounov et al. 2009, 2013; Adkin et al. 2006;
Gehring and Knight 2000; Krigolson and Holroyd 2007). Krigolson and Holroyd
(2007) identified error-related negativity (ERN) potentials in frontal regions of the
cortex when errors were made during a continuous tracking task (Krigolson and
Holroyd 2007). In addition, Gehring and Knight (2000) reported ERN potentials in
the ACC following erroneous responses in a visual reaction task while these
potentials were absent or attenuated in amplitude during correct trials. Interestingly,
subjects with lateral prefrontal damage exhibited no differences in ERN potentials
between erroneous and correct trials. The authors interpreted these results to
indicate an interaction between the prefrontal cortex and the ACC in action
monitoring and guiding of compensatory actions (Gehring and Knight 2000).
Using spectral analysis and a source localisation approach (LORETA), Slobounov
et al. (2013) reported increased theta power in the ACC during postural control
demands. It is assumed that the ACC, amongst others, monitors postural stability
and reacts (as indicated by the ERN) when balance is disrupted (Jacobs and Horak
2007; Slobounov et al. 2009, 2013). This error detection system might be driven by
cortical theta oscillations while the motor part of balance is controlled by gamma
oscillations in the primary motor cortex. In fact, there is evidence in the literature
that theta and gamma oscillations are functionally interrelated during motor control
(Perfetti et al. 2011).
To this point there is less research on dynamic balance tasks characterised by a
continuous disturbance of balance induced by an altered base of support, surface
stability or visual input. However, such tasks are widely used for balance and
stability training to improve athletic performance as well as during rehabilitation
from injury (Muehlbauer et al. 2012; Zech et al. 2010). In a recent experiment
conducted in our own laboratory (Hülsdünker et al. in preparation), we investigated
balance tasks varying in difficulty by altering base of support and surface stability.
It was hypothesised that power in the theta frequency band, as a physiological
indicator of error detection and action monitoring, should be increased in a task
with high compared to low balance demands. We found a pronounced increase in
theta power in the frontal, midline central and midline parietal cortex when subjects
were standing on an oscillating platform using their non-dominant leg, compared to
a bipedal stance on a solid surface (see Fig. 3.1). This result indicates that contin-
uous balance control involves an extended action monitoring network that inte-
grates sensory and motor information.
Fig. 3.1 Cortical mapping (view from the top) of electrical power in the theta (4–7 Hz) frequency
band during balance tasks displayed for bipedal stance on a solid surface (left) and unipedal stance
on the non-dominant leg on an oscillating platform (right). The color scale indicates
ln-transformed electrical power values (μV2) over the scalp calculated by spline interpolation
3.3 EEG Neurofeedback Training to Enhance Athletic

Performance
The basic principle of cortical oscillations-based neurofeedback training (NFT) is

to visually and/or acoustically feed back cortical oscillations, which are typically
recorded by EEG. NFT has previously been applied in a range of clinical and
non-clinical conditions. For example, it has been reported that NFT is an effective
tool to reduce the symptomatology in attention deficit hyperactivity disorder (Arns
et al. 2009), the frequency of seizures in epileptic patients (Tan et al. 2009) and the
number of awakenings in insomniac patients (Schabus et al. 2013). Furthermore,
NFT has been shown to enhance music, acting and dance performance (Gruzelier
2013a), as well as microsurgical skills (Ros et al. 2009). There are also a few studies
where NFT was applied to enhance athletic performance. Landers et al. (1991)
examined whether EEG biofeedback training for approximately 60 min could
improve archery performance, as well as self-reported measures of concentration
and self-confidence. For this purpose, 24 experienced pre-elite archers were ran-
domly assigned to one of three treatment conditions: (a) correct feedback (i.e.,
reduced left temporal activation), (b) incorrect feedback (i.e., reduced right tempo-
ral activation) and (c) control without NFT. The rationale for the NFT protocol was
derived from previous studies suggesting that reduced activation of the left relative
32 A. Mierau et al.
to the right hemisphere is associated with superior performance (see Sect. 3.1).
Results indicated that only the correct feedback group significantly improved
performance, while no changes were observed in the control group. In the incorrect
feedback group, shooting accuracy was even worse after treatment (Landers
et al. 1991). In a more recent study, Rostami et al. (2012) investigated the effect
of fifteen 60-min sessions NFT on performance in expert rifle shooters. Participants
in the intervention group received two different NFT protocols within each session.
One protocol focused on simultaneous enhancement of the sensorimotor rhythm
(SMR; 13–15 Hz) and inhibition of high beta activity (20–30 Hz). The other
focused on modulation of the alpha (8–12 Hz)/theta (4–8 Hz) ratio combined
with inhibition of high beta activity (20–30 Hz). These (or similar) protocols
have been successfully applied in numerous NFT studies outside of the sporting
arena and therefore, they represent rather standard NFT protocols. Compared to an
untreated control group, shooting performance significantly improved in the NFT
group (Rostami et al. 2012). Arns et al. (2008) studied a new method for golf
performance enhancement employing personalised event-locked EEG
neurofeedback during putting. Six amateur golfers received three real-life NFT
sessions. Each session consisted of four series of 80 putts. However, feedback was
provided only in the second and fourth series whereas series one and three served as
a control condition. Target frequency bands and amplitudes for the individualised
NFT were derived from a prior assessment comparing successful and unsuccessful
putts. The overall percentage of successful putts was significantly larger in those
series with feedback compared to the control (no feedback) series. The average
improvement in performance with feedback on the personalised EEG profile was
about 25 % (Arns et al. 2008). The results of this study may suggest that individual
rather than generalised EEG profiles should be used for NFT. However, the subjects
were amateur and not expert golfers. Their task-related EEG profiles may not only
be different from those of experts but also, they may show larger variability/less
stability. Therefore, it could be hypothesised that individualised NFT protocols may
be particularly useful in less skilled athletes, whereas for experts there might be a
higher chance to identify common EEG features of peak performance. Consistent
with this, a preliminary study conducted by Muangjaroen and Wongsawat (2012)
indicates that high-frequency alpha power at electrode C4, theta power at Fz, as
well as theta and high alpha power at Pz are sufficient to calculate an index to
predict successful golf putting Muangjaroen and Wongsawat (2012). However,
caution is recommended before suggesting a particular NFT protocol for a partic-
ular sport and level of expertise. As indicated by the study of Landers et al. (1991),
incorrect feedback (i.e. wrong NFT protocol) may even deteriorate performance.
3.4 Concluding Remarks and Suggestions for Future

Research
The results presented in this chapter indicate that superior sport performance is
associated with changes in cortical oscillation patterns. However, these changes
depend on the specific demands of the sport. A good example of this is the
relationship between high-frequency alpha power and performance in golf versus
shooting. High-frequency alpha power at electrodes overlying sensorimotor cortical
areas was reduced during successful compared to unsuccessful putting in expert
golfers (Babiloni et al. 2007) but increased during high compared to low shooting
scores in expert pistol shooters (Del Percio et al. 2009). Alpha desynchronisation
prior to putting could be interpreted to reflect an increase in cortical activation in
order to facilitate movement execution (Pfurtscheller et al. 1996), whereas alpha
synchronisation during aiming prior to the shot may indicate inhibitory processes
(Hummel et al. 2002) to avoid aim point fluctuations. Future experiments may
manipulate conditions (e.g. distance from the target, target size, time to complete
the task, dual tasking) to gain a better understanding of the processes underlying
changed EEG profiles during superior performance. Furthermore, more longitudi-
nal studies are needed and tasks other than golf putting and shooting should be
studied. It is also recommended to consider multimodal approaches for a more
comprehensive view.
Skilled athletic performance often requires a substantial amount of action
monitoring and adequate error detection on a subsecond timescale. It is suggested
this is processed in the theta frequency band. Consistent with this, studies revealed
that theta oscillations are linked to balance control (Slobounov et al. 2009, 2013), as
well as performance in golf (Baumeister et al. 2008; Kao et al. 2013) and pistol
shooting (Doppelmayr et al. 2008). Future research on balance control may identify
characteristic changes in brain activity that are related to the difficulty of balance
tasks and to balance performance. Furthermore, adaptations to different balance
training programs on a behavioral and cortical level may help to better identify
optimal balance exercises to improve performance in sports.
To this point of time, only few studies are published that applied EEG
neurofeedback to enhance athletic performance. These studies revealed that NFT
has the potential to improve the performance of archers (Landers et al. 1991),
golfers (Arns et al. 2008) or shooters (Rostami et al. 2012). Beyond that, NFT has
been associated with positive effects on cognitive and affective outcome measures
(Gruzelier 2013b), creativity (Gruzelier 2013a), fine motor skills (Ros et al. 2009),
reaction time (Doppelmayr and Weber 2011), as well as sleep quality and memory
(Schabus et al. 2013). These results suggest that NFT could be a useful tool to
directly or indirectly (e.g. via improved sleep) enhance athletic performance.
34 A. Mierau et al.
References
Adkin AL, Quant S, Maki BE, McIlroy WE (2006) Cortical responses associated with predictable
and unpredictable compensatory balance reactions. Exp Brain Res 172:85–93
Andreassi JL (2007) Psychophysiology: human behavior and physiological response, 5th edn.
Lawrence Erlbaum, Mahwah
Arns M, Kleinnijenhuis M, Fallahpour K, Breteler R (2008) Golf performance enhancement and
real-life neurofeedback training using personalized event-locked EEG profiles. J Neurother
11:11–18
Arns M, de Ridder S, Strehl U, Breteler M, Coenen A (2009) Efficacy of neurofeedback treatment
in ADHD: the effects on inattention, impulsivity and hyperactivity: a meta-analysis. Clin EEG
Neurosci 40:180–189
Asada H, Fukuda Y, Tsunoda S, Yamaguchi M, Tonoike M (1999) Frontal midline theta rhythms
reflect alternative activation of prefrontal cortex and anterior cingulate cortex in humans.
Neurosci Lett 274:29–32
Babiloni C, Del Percio C, Iacoboni M, Infarinato F, Lizio R, Marzano N et al (2007) Golf putt
outcomes are predicted by sensorimotor cerebral EEG rhythms. J Physiol 586:131–139
Ball KA, Best RJ, Wrigley TV (2003) Inter- and intra-individual analysis in elite sport: pistol
shooting. J Appl Biomech 19:28–38
Baumeister J, Reinecke K, Liesen H, Weiss M (2008) Cortical activity of skilled performance in a
complex sports related motor task. Eur J Appl Physiol 104:625–631
Bolton DAE, Williams L, Staines W, McIlroy WE (2012) Contribution of primary motor cortex to
compensatory balance reactions. BMC Neurosci 13:102
Cooke A (2013) Readying the head and steadying the heart: a review of cortical and cardiac studies
of preparation for action in sport. Int Rev Sport Exerc Psychol 6:122–138
Del Percio C, Babiloni C, Bertollo M, Marzano N, Iacoboni M, Infarinato F et al (2009) Visuo-
attentional and sensorimotor alpha rhythms are related to visuo-motor performance in athletes.
Hum Brain Mapp 30:3527–3540
Doppelmayr M, Weber E (2011) Effects of SMR and theta/beta neurofeedback on reaction times,
spatial abilities, and creativity. J Neurother 15:115–129
Doppelmayr M, Finkenzeller T, Sauseng P (2008) Frontal midline theta in the pre-shot phase of
rifle shooting: differences between experts and novices. Neuropsychologia 46:1463–1467
Fitts PM, Posner MI (1967) Human performance, Basic concepts in psychology series. Psychol-
ogy: a natural science. Brooks/Cole Pub. Co, Belmont
Gehring WJ, Knight RT (2000) Prefrontal – cingulate interactions in action monitoring. Nat
Neurosci 3:516–520
Gevins A, Smith ME, McEvoy L, Yu D (1997) High-resolution EEG mapping of cortical
activation related to working memory: effects of task difficulty, type of processing, and
practice. Cereb Cortex 7:374–385
Gruzelier JH (2013a) EEG-neurofeedback for optimising performance. II: creativity, the
performing arts and ecological validity. Neurosci Biobehav Rev 44:142–158
Gruzelier JH (2013b) EEG-neurofeedback for optimising performance. I: a review of cognitive
and affective outcome in healthy participants. Neurosci Biobehav Rev 44:124–141
Hatfield BD, Hillman C (2001) The psychophysiology of sport: a mechanistic understanding of the
psychology of superior performance. In: Singer RN (ed) Handbook of sport psychology, John
Wiley. New York, NY. pp 362–388
Hatfield BD, Landers DM (1987) Psychophysiology in exercise and sport research: an overview.
Exerc Sport Sci Rev 15:351–387
Hatfield BD, Haufler AJ, Hung T, Spalding TW (2004) Electroencephalographic studies of skilled
psychomotor performance. J Clin Neurophysiol 21:144–156
Haufler AJ, Spalding TW, Santa Maria DL, Hatfield BD (2000) Neuro-cognitive activity during a
self-paced visuospatial task: comparative EEG profiles in marksmen and novice shooters. Biol
Psychol 53:131–160
Hrysomallis C (2011) Balance ability and athletic performance. Sports Med 41:221–232
Hummel F, Andres F, Altenmüller E, Dichgans J, Gerloff C (2002) Inhibitory control of acquired
motor programmes in the human brain. Brain 125:404–420
Jacobs JV, Horak FB (2007) Cortical control of postural responses. J Neural Transm
114:1339–1348
Jacobs JV, Fujiwara K, Tomita H, Furune N, Kunita K, Horak FB (2008) Changes in the activity of
the cerebral cortex relate to postural response modification when warned of a perturbation. Clin
Neurophysiol 119:1431–1442
Kao S, Huang C, Hung T (2013) Frontal midline theta is a specific indicator of optimal attentional
engagement during skilled putting performance. J Sport Exerc Psychol 35:470–478
Kerick SE, Douglass LW, Hatfield BD (2004) Cerebral cortical adaptations associated with
visuomotor practice. Med Sci Sports Exerc 36:118–129
Klimesch W (2012) Alpha-band oscillations, attention, and controlled access to stored informa-
tion. Trends Cogn Sci 16:606–617
Krigolson OE, Holroyd CB (2007) Predictive information and error processing: the role of medial-
frontal cortex during motor control. Psychophysiology 44:586–595
Landers DM, Petruzzello SJ, Salazar W, Crews DJ, Kubitz KA, Gannon TL et al (1991) The
influence of electrocortical biofeedback on performance in pre-elite archers. Med Sci Sports
Exerc 23:123–129
Landers DM, Han M, Salazar W, Petruzzello SJ, Kubitz KA, Gannon TL (1994) Effects of
learning on electroencephalographic and electrocardiographic patterns in novice archers. Int
J Sport Psychol 25:313–330
Mochizuki G, Sibley K, Esposito J, Camilleri J, McIlroy W (2008) Cortical responses associated
with the preparation and reaction to full-body perturbations to upright stability. Clin
Muangjaroen P, Wongsawat Y (2012) Real-time index for predicting successful golf putting
motion using multichannel EEG. Conf Proc IEEE Eng Med Biol Soc 2012:4796–4799.
doi:10.1109/EMBC.2012.634703
Muehlbauer T, Roth R, Bopp M, Granacher U (2012) An exercise sequence for progression in
balance training. J Strength Cond Res 26:568–574
Muthukumaraswamy SD (2010) Functional properties of human primary motor cortex gamma
oscillations. J Neurophysiol 104:2873–2885
Nakashima K, Sato H (1993) Relationship between frontal midline theta activity in EEG and
concentration. J Hum Ergol (Tokyo) 22:63–67
Nakata H, Yoshie M, Miura A, Kudo K (2010) Characteristics of the athletes’ brain: evidence from
neurophysiology and neuroimaging. Brain Res Rev 62:197–211
Onton J, Delorme A, Makeig S (2005) Frontal midline EEG dynamics during working memory.
Osaka M, Komori M, Morishita M, Osaka N (2007) Neural bases of focusing attention in working
memory: an fMRI study based on group differences. Cogn Affect Behav Neurosci 7:130–139
Ouchi Y, Okada H, Yoshikawa E, Nobezawa S, Futatsubashi M (1999) Brain activation during
maintenance of standing postures in humans. Brain 122(Pt 2):329–338
Perfetti B, Moisello C, Landsness EC, Kvint S, Lanzafame S, Onofrj M et al (2011) Modulation of
gamma and theta spectral amplitude and phase synchronization is associated with the devel-
opment of visuo-motor learning. J Neurosci 31:14810–14819
Pfurtscheller G, Stancák A, Neuper C (1996) Event-related synchronization (ERS) in the alpha
band–an electrophysiological correlate of cortical idling: a review. Int J Psychophysiol
24:39–46
Reeves WH (1983) Right cerebral hemispheric function: behavioral correlates. Int J Neurosci
18:227–230
Ros T, Moseley MJ, Bloom PA, Benjamin L, Parkinson LA, Gruzelier JH (2009) Optimizing
microsurgical skills with EEG neurofeedback. BMC Neurosci 10:87
36 A. Mierau et al.
Rostami R, Sadeghi H, Karami KA, Abadi MN, Salamati P (2012) The effects of neurofeedback
on the improvement of rifle shooters’ performance. J Neurother 16:264–269
Salazar W, Landers DM, Petruzzello SJ, Han M, Crews DJ, Kubitz KA (1990) Hemispheric
asymmetry, cardiac response, and performance in elite archers. Res Q Exerc Sport 61:351–359
Sauseng P, Klimesch W, Gerloff C, Hummel F (2009) Spontaneous locally restricted EEG alpha
activity determines cortical excitability in the motor cortex. Neuropsychologia 47:284–288
Schabus M, Heib DP, Lechinger J, Griessenberger H, Klimesch W, Pawlizki A et al (2013)
Enhancing sleep quality and memory in insomnia using instrumental sensorimotor rhythm
conditioning. Biol Psychol 95:126–134
Slobounov S, Hallett M, Stanhope S, Shibasaki H (2005) Role of cerebral cortex in human postural
control: an EEG study. Clin Neurophysiol 116:315–323
Slobounov S, Wu T, Hallett M (2006) Neural basis subserving the detection of postural instability:
an fMRI study. Motor Control 10:69–89
Slobounov S, Hallett M, Cao C, Newell K (2008) Modulation of cortical activity as a result of
voluntary postural sway direction: an EEG study. Neurosci Lett 442:309–313
Slobounov S, Cao C, Jaiswal N, Newell KM (2009) Neural basis of postural instability identified
by VTC and EEG. Exp Brain Res 199:1–16
Slobounov SM, Teel E, Newell KM (2013) Modulation of cortical activity in response to visually
induced postural perturbation: combined VR and EEG study. Neurosci Lett 547:6–9
Szurhaj W, Bourriez J, Kahane P, Chauvel P, Mauguière F, Derambure P (2005) Intracerebral
study of gamma rhythm reactivity in the sensorimotor cortex. Eur J Neurosci 21:1223–1235
Tan G, Thornby J, Hammond DC, Strehl U, Canady B, Arnemann K et al (2009) Meta-analysis of
EEG biofeedback in treating epilepsy. Clin EEG Neurosci 40:173–179
Taube W (2006) Direct corticospinal pathways contribute to neuromuscular control of perturbed
stance. J Appl Physiol 101:420–429
Zech A, Hübscher M, Vogt L, Banzer W, Hänsel F, Pfeifer K (2010) Balance training for
neuromuscular control and performance enhancement: a systematic review. J Athl Train
45:392–403
Zemková E (2013) Sport-specific balance. Sports Med 44(5):579–590
Chapter 4
Intra- and Inter-person Coordinated
Movements of Fingers and Toes
Tetsuro Muraoka, Yuki Watanabe, and Kazuyuki Kanosue
Abstract When performing intra- or inter-person coordination of cyclical move-

ments of two joints, there is a perceptual-cognitive constraint based on spatial
information for performing such coordination. However, bimanual coordination
of the index finger flexion-extension is an exception in which the constraint occurs
based on motoric information, which might be peculiar to digits. In order to
investigate whether intra- and inter-person coordination of fingers and toes were
constrained based on spatial or motoric information, coordinated movements were
performed in one of two modes, flexing fingers concomitant with either toe flexion
or toe extension, with the forearm either in the pronated or supinated position. In
intra-person coordination, both the relative direction of movement and activation
coupling influenced the stability of coordination to a similar extent. In inter-person
coordination, the coordination with alternate activation of the corresponding mus-
cles of fingers and toes in the opposite direction was less stable as compared to other
coordination modes. These findings suggest that both spatial and motoric informa-
tion are utilized in intra-person coordination of the fingers and toes, whereas either
spatial or motoric information is utilized in inter-person coordination to meet a
specific requirement for each task.
Keywords Coordination • Interlimb • Digits
T. Muraoka (*)
College of Economics, Nihon University, Tokyo, Japan
Consolidated Research Institute for Advanced Science and Medical Care, Waseda University,
Tokyo, Japan
e-mail: muraoka.tetsurou@nihon-u.ac.jp
Y. Watanabe
School of Sport Sciences, Waseda University, Saitama, Japan
K. Kanosue

38 T. Muraoka et al.
4.1 Introduction
Rhythmic interlimb coordination is constrained by perceptual-cognitive, musculo-

skeletal and neural factors (Swinnen 2002; Swinnen and Wenderoth 2004).
Mechsner and colleagues claim that the perceptual-cognitive factor is a predomi-
nant constraint in interlimb coordination (Mechsner et al. 2001; Mechsner 2004). In
support, many studies have shown that the stability of interlimb coordination
depends on the relative direction of the moving limbs (i.e., same or opposite
directions on the sagittal plane, or symmetrical or parallel directions on the hori-
zontal plane) independent of the muscles employed in the coordination (Baldissera
et al. 1982; Carson et al. 1995; Salesse et al. 2005). In addition, a difficult interlimb
coordination, such as a coordination pattern with a 90 relative phase (Φ; a phase
difference between movements) between the arms (Kovacs et al. 2009) or a
bimanual coordination with polyrhythms (Mechsner et al. 2001; Kovacs
et al. 2010) can be performed stably by using appropriate feedback that makes it
possible to perform interlimb coordination as an integrated goal at the perceptual-
cognitive level.
Results from experiments using coordination between moving a visual stimulus
and the unimanual swing of a pendulum (Hajnal et al. 2009) or unimanual manip-
ulation of a joystick (Wilson et al. 2005a, b) suggest that the relative phase is
utilized as crucial information in visuo-motor coordination. In inter-person and
visuo-motor coordination, anti-phase (opposite directional; Φ ¼ 180 ) coordination
often abruptly changes to in-phase (same directional; Φ ¼ 0 ) coordination with an
increase in frequency. This change to in-phase is followed by variability in Φ.
In-phase coordination does not spontaneously change to anti-phase coordination
with a decrease in frequency. Since these characteristics can be observed in intra-
person coordination as well as inter-person and visuo-motor coordination, infor-
mation of relative phase may also be utilized in intra-person coordination.
The relative phase of a moving visual stimulus or human movement seems to be
calculated based on spatial information in the external world (spatial relative phase)
because constraint in inter- (Temprado et al. 2003; Temprado and Laurent 2004;
Fine et al. 2013) and intra-person (Baldissera et al. 1982; Carson et al. 1995; Salesse
et al. 2005) coordination occurs mainly based on spatial information rather than
motoric (anatomical) information. However, bimanual coordination of index finger
flexion-extension is an exception in which constraint occurs based on motoric
information (Riek et al. 1992). When doing bimanual index finger flexion-extension
cyclically, flexion-flexion coupling is more stable compared to flexion-extension
coupling irrespective of relative direction between moving index fingers in the
external world. That is, the relative phase of finger movement is calculated based on
motoric information in an internal world (motoric relative phase). Does the con-
straint based on motoric information occur only in bimanual coordination of index
finger flexion-extension? If such constraint is related to the joint function of finger
flexion-extension (i.e., grasping), it may also occur in coordination with toe use
because the toes also have the ability to grasp. In the present chapter we report a
4 Intra- and Inter-person Coordinated Movements of Fingers and Toes 39
study in which we investigate whether intra- and inter-person coordination between

toes and fingers are constrained based on motoric information.
4.2 Methods
4.2.1 Subjects
Experiments 1 and 2 utilized the same group of ten normal adults (two females and
eight males), ranging in age from 22 to 23 years. They were fully informed about
the purpose of the study and informed consent was obtained from all participants.
The experiments were conducted in accordance with the Declaration of Helsinki.
4.2.2 Apparatus
Subjects sat in a chair with the right forearm supported in a horizontal position on
an armrest. The right foot was put on an incline board with toes in the air. Angular
displacements of the middle phalanx of the index finger relative to the dorsum of
the hand and those of the distal phalanx of the second toe relative to the dorsum of
the foot were measured at 1 kHz using electrical goniometers (SG150 and SG110,
Biometrics, UK). The joint angular signals were stored on a computer via an AD
converter (PowerLab16/30, ADInstruments, Australia), and subsequently were
low-pass filtered with a cutoff frequency of 10 Hz.
4.2.3 Procedure
In experiment 1, subjects performed cyclical extension-flexion movements of the

right fingers (digits 2–5) and the right toes (digits 1–5) in the sagittal plane while
watching the movements and hearing white noise through earphones. The four
conditions are illustrated in Fig. 4.1: simultaneous activation of the corresponding
muscles of fingers and toes (i.e., flexors or extensors) with hand in the pronated
position (task 1), alternate activation of the corresponding muscles of fingers and
toes with hand in the pronated position (task 2), simultaneous activation with the
hand in the supinated position (task 3), and alternate activation with the hand in the
supinated position (task 4). In this situation “simultaneous activation” refers to
finger flexion coupled with toe flexion, whereas “alternate activation” refers
to finger flexion coupled with toe extension. In tasks 1 and 4, the joints of the
fingers and toes rotated in the “same-direction”, whereas they rotated in the
Fig. 4.1 Tasks involving coordinated movements of the right fingers and the right toes. The
fingers and the toes simultaneously flex or extend in Simul-Same and Simul-Opp, whereas finger
flexion (viz. extension) is coupled with toe extension (viz. flexion) in Alter-Opp and Alter-Same.
Fingers and toes rotate in the same direction in Simul-Same and Alter-Same, whereas they rotate in
the opposite direction in Alter-Opp and Simul-Opp
“opposite-direction” in tasks 2 and 3. Thus, tasks 1, 2, 3, and 4 were referred to as

Simul-Same, Alter-Opp, Simul-Opp, and Alter-Same, respectively (Fig. 4.1).
In the experimental session, subjects were initially instructed to perform the four
tasks several times; each task involved five cycles. This served as a practice session.
They were then instructed to perform the tasks at their preferred frequency, but to
perform all tasks at the same frequency. We requested that the subjects synchronize
the timing of finger peak extension or flexion to toe peak extension or flexion as
precisely as possible. The subjects were permitted to choose their preferred range of
motion (ROM) of fingers and toes for each task as long as the ROMs differed less
than 30 % among the tasks. This was done to minimize the influence of the
difference in attentional cost among the tasks in order to maintain a given ROM
on the stability of the coordinated movements. After the practice session, the
subjects were instructed to perform one of the tasks for 25 cycles, in random
order within each cluster of the four tasks. Five clusters were repeated for each
subject. Subjects were instructed to not actively resist pattern change and to
establish a comfortable pattern. There was no pacing signal such as an auditory
metronome during trials.
In experiment 2, ten subjects were initially grouped into five pairs in the order of
their average values of movement frequency in experiment 1. Then, across subjects,
the pairs of the right hands and feet were made and the subjects performed tasks
1, 2, 3, and 4. Each pair of subjects was seated in a chair, side by side. The subject
seated on the right seat put his or her right foot on the incline board and moved their
right toes, and the subject seated on the left seat put his or her right forearm on the
armrest and moved their right fingers. The subjects were asked to coordinate their
periodic movements of toes or fingers to perform tasks 1, 2, 3, or 4 at their preferred
frequency. As in experiment 1, the subjects first did a practice session and also
decided on movement frequency and ROM in the same way as in experiment 1. The
subjects then switched appendages, so the subjects using their right toes used their
right fingers and the subjects using their right fingers used their right toes. The tasks
were run again. The results of the two combinations were analyzed separately.
4.2.4 Calculation of Relative Phase Between Toes

and Fingers Movements
The period of each movement cycle (extension + flexion) of the fingers was com-
pared to the corresponding period of the movement cycle of the toes (extension
+ flexion in the tasks 1 and 3, or flexion + extension in the tasks 2 and 4). Φ was then
calculated as the difference in time between the mid-point of the two periods and
expressed in degrees of angular displacement, taking the period of finger movement
as a reference. In the present study, the movement cycle of the toes was differently
defined for tasks 1 and 3, and for tasks 2 and 4, such that the required Φ was 0 in all
tasks. When Φ was less than 90 , we considered the cycle to be successfully
performed. When Φ was more than 90 , we considered a loss of stability to have
occurred. The standard deviation of Φ (SDΦ) for each trial was calculated using Φ
of the first 20 cycles after the intended coordination started or Φ from the beginning
of the intended coordination before the occurrence of a loss of stability. The SDΦ
was used to represent coordination stability. Trials in which a loss of stability
occurred within the first seven cycles after the initiation of the intended coordina-
tion were excluded from statistical analysis for the SDΦ calculations. The number
of excluded trials was nine in experiment 1 (4.5 % of the total number of trials) and
3 in experiment 2 (1.5 % of the total number of trials).
Movement frequency of the fingers was calculated for each cycle; then mean
movement frequency for each trial was calculated using the movement frequency
data obtained from the same cycles utilized for the calculation of SDΦ.
4.2.5 Statistics
Differences in SDΦ among the tasks were tested by a two-way ANOVA with
repeated measures design (2 2, movement coupling movement direction). The
movement coupling condition consisted of two levels: simultaneous flexion of
fingers and toes versus alternate flexion of fingers and toes. The movement direction
condition consisted of two levels: fingers and toes rotated in the same-direction
versus fingers and toes rotated in the opposite-direction. When there was an
interaction between movement coupling movement direction, the Bonferoni test
was used to test the difference in SDΦ among the four conditions. Values are
presented as means SD. Statistical difference was set at a level of p < 0.05.
4.3 Results
In the intra-person coordination of experiment 1, there were significant main effects

of both movement coupling (F(1,9) ¼ 15.1, p < 0.005) and movement direction (F
(1,9) ¼ 28.6, p < 0.0005) on SDΦ, whereas an interaction between activation
coupling movement direction was not significant (F(1,9) ¼ 0.3, p > 0.05)
(Fig. 4.2). SDΦ for each task was in accordance with the number of trials in
which a loss of stability occurred (Simul-Same ¼ 2 % (1 trial), Alter-Opp ¼ 30 %
(15 trials), Simul-Opp ¼ 10 % (five trials), and Alter-Same ¼ 12 % (six trials),
50 trials for each task).
In the inter-person coordination of experiment 2, there was a significant main
effect of movement coupling (F(1,9) ¼ 49.3, p < 0.0001) on SDΦ, though a main
effect of movement direction was not significant (F(1,9) ¼ 4.8, p ¼ 0.056). There
was a significant interaction between activation coupling movement direction (F
(1,9) ¼ 10.2, p < 0.05) on SDΦ. The Bonferoni test showed that SDΦ in Alter-Opp
was significantly greater than that in Simul-Same (p < 0.01), Simul-Opp
Fig. 4.2 Standard

deviation of the relative
phase in intra-person
coordination. There were
significant main effects of
both activation coupling
(simultaneous versus
alternate activation of
corresponding muscles of
fingers and toes) and
movement direction (same-
versus opposite-directions).
Vertical bars represent
SD. N ¼ 10
(p < 0.001), and Alter-Same (p < 0.05), and that there was no significant difference
in SDΦ among Simul-Same, Simul-Opp, and Alter-Same (Fig. 4.3). Similar to the
case in the intra-person coordination in experiment 1, SDΦ for each task was in
accordance with the number of trials in which a loss of stability occurred (Simul-
Same ¼ 2 % (one trial), Alter-Opp ¼ 8 % (four trials), Simul-Opp ¼ 4 % (two
trials), and Alter-Same ¼ 2 % (one trial), 50 trials for each task), though total
number of a loss of stability in the inter-person coordination (eight trials) was
smaller than that in the intra-person coordination (27 trials).
The mean movement frequency in Simul-Same, Alter-Opp, Simul-Opp, and
Alter-Same in experiment 1 were 1.15 0.24 Hz, 1.16 0.23 Hz, 1.15 0.23 Hz,
and 1.15 0.23 Hz, respectively. The mean movement frequencies in Simul-Same,
Alter-Opp, Simul-Opp, and Alter-Same in experiment 2 were 1.32 0.08 Hz,
1.26 0.09 Hz, 1.31 0.09 Hz, and 1.28 0.09 Hz, respectively. Similar values
were obtained for the different tasks in each experiment.
4.4 Discussion
The fundamental goal of this study was to investigate whether intra- and inter-
person coordination of fingers and toes were constrained based on spatial or motoric
information. We measured stability of the coordinated movements of fingers and
Fig. 4.3 Standard deviation of the relative phase in inter-person coordination. SD of relative
phase was significantly greater in Alter-Opp compared to Simul-Same, Simul-Opp, and Alter-
Same. Vertical bars represent SD. N ¼ 10
toes in two modes of coordination with two different forearm positions in order to
dissociate influences of spatial and motoric information. The results showed that
both relative direction of the movement and activation coupling influenced the
stability of coordination to a similar extent in intra-person coordination, and that the
coordination with alternate activation of corresponding muscles of fingers and toes
in the opposite direction was less stable compared to other coordination modes in
inter-person coordination. Consequently, we suggest that both spatial and motoric
information are utilized in intra-person coordination of fingers and toes, whereas
either spatial or motoric information are utilized in inter-person coordination to
meet a specific requirement for each task.
4.4.1 Intra-personal Coordination of Fingers and Toes
The present study showed that both relative direction of the movement and activa-
tion coupling influenced the stability of coordination of fingers and toes to a similar
extent. It has been shown that bimanual coordination of index finger flexion-
extension is predominantly constrained based on motoric relative phase (Riek
et al. 1992) and that other coordinated movements of two limbs are predominantly
constrained based on spatial relative phase (Baldissera et al. 1982; Carson

et al. 1995; Mechsner et al. 2001; Li et al. 2004; Salesse et al. 2005; Muraoka
et al. 2013). However, the present study showed that intra-person coordination
between toes and fingers was constrained based on both motoric and spatial relative
phases independently, which did not agree with previous studies.
The fact that constraint based on motoric relative phase only occurred in
coordination between digits flexion/extension in different limbs may be because
digit movements are inseparable from functional meaning (i.e., grasping) at the
perceptual-cognitive level. On the other hand, the joint movements often utilized in
previous studies on intra-person coordination such as flexion/extension of the wrist,
elbow, ankle, and knee are not related to specific meaningful behavior. In interlimb
coordination of meaningful behavioral patterns, simultaneous execution of mean-
ingful behavior (e.g., simultaneous grasping) could utilize common neural pro-
cesses for the movements of two limbs and thus would result in more stable
coordination compared to alternate execution of meaningful behavioral patterns.
On the other hand, when coordinating two limbs movements that normally operate
without a specific functional meaning, in terms of the spatial aspects of movements,
movements of two limbs in a same direction could utilize a common neural process
and thus result in more stable coordination as compared to movements in an
opposite direction. We rarely grasp objects using toes in our daily life and our
toes are so short that their ability to grasp is limited, and thus the relationship
between movement and specific functional meaning might have become less strong
for human toe movement. In support of this, the somatic sensibilities for the toes are
less than for those of the fingers in humans, while those for the toes and the fingers
are almost the same in monkeys, based on studies of evoked potentials in the
thalamus and cortex (Kandel 2000). A decline in the toes’ grasping ability in
humans might have resulted in the coordination of the fingers and the toes being
constrained based on the spatial relative phase as well as the motoric relative phase.
Regarding the hypothesis that tight perceptual-cognitive coupling between digit
movements and specific functional meaning results in constraint based on the
motoric relative phase, the reference frame utilized for digit movements should
be taken into consideration. The goal of movement of limbs (except digits) is often
to move them in the external world, and the movement is, at least partly, encoded in
the external world in the higher motor areas of the brain (Kakei et al. 2001). On the
other hand, when movement of proximal joints localizes digits in the external
world, movement of the digits could be encoded in a reference frame attached to
the hand, not in a reference frame attached to the external world. In support of this,
the space surrounding the hands is encoded in a hand-centered reference frame in
the premotor and posterior parietal cortices (Brozzoli et al. 2012), and thus the digit
movements might be encoded in a hand-centered reference frame. If so, simulta-
neous flexion of fingers and toes could be considered not only as simultaneous
execution of perceptual-cognitive meaningful behavior but also as a simultaneous
execution of same directional movement in the hand/foot centered reference frame.
We speculate that constraint based on the spatial relative phase in the present study
reflects the extent of neural populations encoding toe movement in a reference
frame attached to the external world which increased as the result of a decline, over
evolutionary time, in the toes’ grasping ability.
Although some studies using coordination of flexion/extension in both wrists
also showed that both motoric and spatial relative phases independently influenced
coordination stability to the same extent (Temprado et al. 2003; Temprado and
Swinnen 2005), Salesse et al. (2005) showed that constraint based on the motoric
relative phase was minor one. In addition, in the learning of the intermediate
relative phase in the coordination of flexion/extension of both wrists, transfer of
learning occurred based on spatial relative phase, not on motoric relative phase
(Temprado and Swinnen 2005). Further study is required to investigate the con-
straint of coordination of flexion/extension for both wrists.
4.4.2 Inter-personal Coordination of Fingers and Toes
The present result showed that there was a significant interaction between the
relative direction of movement and activation coupling, and that inter-person
coordination of fingers and toes was stable when motoric and/or spatial relative
phases were in-phase compared to when both motoric and spatial relative phases
were anti-phase. Previous studies showed that inter-person coordination was stable
when spatial relative phase was in-phase compared to anti-phase (Schmidt
et al. 1990, 1998; Amazeen et al. 1995; Temprado et al. 2003; Temprado and
Laurent 2004; Richardson et al. 2008; Coey et al. 2011), which is not in line with
the present findings.
In inter-person coordination, visual information contributes to form the coordi-
nation state. Previous studies on inter-person coordination utilized movements
including manipulating a pendulum (Amazeen et al. 1995; Schmidt et al. 1998;
Richardson et al. 2008; Coey et al. 2011), manipulating a joystick (Temprado
et al. 2003; Temprado and Laurent 2004), knee flexion/extension (Schmidt
et al. 1990), and elbow flexion/extension for one subject with shoulder internal/
external rotation for the other (Kilner et al. 2003; Fine et al. 2013). Visual
information derived from these inter-person coordination to form the coordination
state was considered to be the spatial relative phase (Temprado et al. 2003;
Temprado and Laurent 2004; Fine et al. 2013), and was thought not to be related
to meaningful perceptual-cognitive behavior. In addition, in coordination between
limb movement and a moving visual stimulus, in-phase (i.e. same direction in the
external world) was more stable compared to anti-phase in terms of spatial relative
phase (Hajnal et al. 2009). On the other hand, movements of both the fingers and the
toes are naturally related to specific perceptual-cognitive behavior (i.e., grasping),
and thus the motoric as well as the spatial relative phase could be used as visual
information. Given that both the motoric and the spatial relative phases of the other
person’s finger/toe movement were utilized as information, similar results would be
obtained both in intra- and inter-person coordination of fingers and toes. However,
the present study clearly showed that the results obtained in inter-person
coordination were different from those in intra-person coordination. If we used only

the motoric relative phase as information, inter-person coordination would be
stabilized as in-phase coordination based on the motoric relative phase, which
was inconsistent with the present result. This can be applied to the case in which
we used only the spatial relative phase. On the other hand, if it were possible to
derive useful information from the other person’s finger/toe movement according to
the coordination mode, Simul-Opp and Alter-Same could be performed as in-phase
in terms of motoric and spatial relative phase, respectively. Alter-Opp was anti-
phase in terms of both motoric and spatial relative phase, and thus coordination
would be less stable. These situations are consistent with the present results.
Therefore it was suggested that other person’s finger/toe movement could be
recognized either in motoric or spatial relative phase, to meet a specific requirement
for each task (i.e., simultaneous grasping or same directional movement).
Acknowledgement The authors thank Dr. Larry Crawshaw for English editing. This work was
partly supported by the “Establishment of Consolidated Research Institute for Advanced Science
and Medical Care” Project, MEXT and by JSPS KAKENHI Grant Number 21700660.
References
Amazeen PG, Schmidt RC, Turvey MT (1995) Frequency detuning of the phase entrainment
dynamics of visually coupled rhythmic movements. Biol Cybern 72:511–518
Baldissera F, Cavallari P, Civaschi P (1982) Preferential coupling between voluntary movements
of ipsilateral limbs. Neurosci Lett 34:95–100
Brozzoli C, Gentile G, Ehrsson HH (2012) That’s near my hand! Parietal and premotor coding of
hand-centered space contributes to localization and self-attribution of the hand. J Neurosci
32:14573–14582
Carson RG, Goodman D, Kelso JAS, Elliott D (1995) Phase-transitions and critical fluctuations in
rhythmic coordination of Ipsilateral hand and foot. J Motor Behav 27:211–224
Coey C, Varlet M, Schmidt RC, Richardson MJ (2011) Effects of movement stability and
congruency on the emergence of spontaneous interpersonal coordination. Exp Brain Res
211:483–493
Fine JM, Gibbons CT, Amazeen EL (2013) Congruency effects in interpersonal coordination.
JExPH 39:1541–1556
Hajnal A, Richardson MJ, Harrison SJ, Schmidt RC (2009) Location but not amount of stimulus
occlusion influences the stability of visuo-motor coordination. Exp Brain Res 199:89–93
Kakei S, Hoffman DS, Strick PL (2001) Direction of action is represented in the ventral premotor
cortex. Nat Neurosci 4:1020–1025
Kandel ER (2000) From nerve cells to cognition. In: Kandel ER, Schwartz JH, Jessell TM (eds)
Principles of neural science, 4th edn. McGraw-Hill companies, New York, pp 381–403
Kilner JM, Paulignan Y, Blakemore SJ (2003) An interference effect of observed biological
movement on action. Curr Biol 13:522–525
Kovacs AJ, Buchanan JJ, Shea CH (2009) Bimanual 1:1 with 90 degrees continuous relative
phase: difficult or easy! Exp Brain Res 193:129–136
Kovacs AJ, Buchanan JJ, Shea CH (2010) Impossible is nothing: 5:3 and 4:3 multi-frequency
bimanual coordination. Exp Brain Res 201:249–259
Li Y, Levin O, Carson RG, Swinnen SP (2004) Bimanual coordination: constraints imposed by the
relative timing of homologous muscle activation. Exp Brain Res 156:27–38
Mechsner F (2004) A psychological approach to human voluntary movement. J Mot Behav

36:355–370
Mechsner F, Kerzel D, Knoblich G, Prinz W (2001) Perceptual basis of bimanual coordination.
Nature 414:69–73
Muraoka T, Ishida Y, Obu T, Crawshaw L, Kanosue K (2013) Ipsilateral wrist-ankle movements
in the sagittal plane encoded in extrinsic reference frame. Neurosci Res 75:289–294
Richardson MJ, Lopresti-Goodman S, Mancini M, Kay B, Schmidt RC (2008) Comparing the
attractor strength of intra- and interpersonal interlimb coordination using cross-recurrence
analysis. Neurosci Lett 438:340–345
Riek S, Carson RG, Byblow WD (1992) Spatial and muscular dependencies in bimanual coordi-
nation. J Hum Mov Stud 23:251–265
Salesse R, Oullier O, Temprado JJ (2005) Plane of motion mediates the coalition of constraints in
rhythmic bimanual coordination. J Mot Behav 37:454–464
Schmidt RC, Carello C, Turvey MT (1990) Phase transitions and critical fluctuations in the visual
coordination of rhythmic movements between people. J Exp Psychol Hum Percept Perform
16:227–247
Schmidt RC, Bienvenu M, Fitzpatrick PA, Amazeen PG (1998) A comparison of intra- and
interpersonal interlimb coordination: coordination breakdowns and coupling strength. J Exp
Psychol Hum Percept Perform 24:884–900
Swinnen SP (2002) Intermanual coordination: from behavioural principles to neural-network
interactions. Nat Rev Neurosci 3:348–359
Swinnen SP, Wenderoth N (2004) Two hands, one brain: cognitive neuroscience of bimanual skill.
Trends Cogn Sci 8:18–25
Temprado JJ, Laurent M (2004) Attentional load associated with performing and stabilizing a
between-persons coordination of rhythmic limb movements. Acta Psychol (Amst) 115:1–16
Temprado JJ, Swinnen SP (2005) Dynamics of learning and transfer of muscular and spatial
relative phase in bimanual coordination: evidence for abstract directional codes. Exp Brain Res
160:180–188
Temprado JJ, Swinnen SP, Carson RG, Tourment A, Laurent M (2003) Interaction of directional,
neuromuscular and egocentric constraints on the stability of preferred bimanual coordination
patterns. Hum Mov Sci 22:339–363
Wilson AD, Collins DR, Bingham GP (2005a) Perceptual coupling in rhythmic movement
coordination: stable perception leads to stable action. Exp Brain Res 164:517–528
Wilson AD, Collins DR, Bingham GP (2005b) Human movement coordination implicates relative
direction as the information for relative phase. Exp Brain Res 165:351–361
Chapter 5
Training Locomotor Function: From
a Perspective of the Underlying Neural
Mechanisms
Tetsuya Ogawa and Kazuyuki Kanosue
Abstract In a variety of sports activities and in our daily lives, we utilize locomo-
tory movements such as walking and running. It is well understood that maintaining
and improving their function can be of major significance in the acquisition of a
better sports performances and a more fulfilling life. To facilitate appropriate
changes in performance, it is essential to know the basic mechanisms underlying
them. In the case of the basic locomotory movements, their neuronal control
mechanisms are predominantly automatic and quite different from those that
underlie voluntarily-induced movements. A number of studies in the last several
decades have described the characteristic features and responsible mechanisms in
both animals and humans. On the basis of the knowledge obtained in these studies,
this chapter will review the recently acquired knowledge to elucidate the neural
mechanisms underlying execution of locomotion movements and provide informa-
tion for construction of possible intervention for improvement in their performance.
Keywords Locomotion-related neural mechanism • Automaticity • Plasticity •

Specificity
5.1 Introduction
Walking to and from school every day. Running a marathon race. Dashing to catch a
fly ball in a baseball game. Riding a bicycle for a trip. Regardless of the difference in
the type of movement or the context of execution, these movements are all locomotory
movements. They are utilized with the common aim of moving in space and share the
common features of precisely-timed movements in two multi-segmented lower limbs
and joints in each limb. A moments reflection reminds us that we do not have to pay
close attention to the detailed movement of the related joints and muscles while
T. Ogawa (*) • K. Kanosue

e-mail: t.ogawa@aoni.waseda.jp

50 T. Ogawa and K. Kanosue
executing these movements. Locomotory movements are initiated largely automati-

cally by the central nervous system (CNS), in which specialized neural mechanisms in
the spinal cord and the brain stem are known to underlie the automaticity (for review,
see Dietz 2002). A number of studies involving animals and humans have done much
to further knowledge about locomotory mechanisms in the last several decades.
On the other hand, the CNS has the potential to subserve plastic changes after
participation in physical activities over the long term (for review, see Zehr 2006;
Wolpaw 2007). This functional reorganization takes place due to repetitive use of
particular neural networks. One of the requirements for developing a superior motor
performance involves facilitating the functional reorganization by repetitive use.
Good training programs address this aspect of motor learning, and to do this in
the most efficient way it is necessary to understand how locomotion related neural
mechanism function, their basic characteristics, and how these features can be
modified and reorganized. In this chapter the authors will initially cover the general
features of locomotory movements. They will then provide information on the
neural mechanisms responsible for generating highly coordinated movements and
also offer ideas for facilitating functional reorganization of the basic mechanisms.
5.2 General Features
Among the common features of locomotory movements, probably the most char-
acteristic involve the highly-coordinated, stereotypical movement patterns and the
repeatable, consecutive movement cycles. Figure 5.1 portrays muscle activity in the
major lower limb muscles and ground reaction forces bilaterally in a human subject
while walking at a comfortable speed. As clearly seen in the electrical activity
pattern for each muscle, the waveforms are very similar in shape (both in timing and
relative amplitude) across the stride cycles and across homologous muscles in the
contralateral limb. Moreover, the presence or absence of electrical activity is very
specific depending on the muscle’s particular function and the regulation occurs in a
reciprocal manner. For example, activity in the dorsiflexor tibialis anterior (TA) and
the plantarflexor medial gastrocnemius (MG) occurs asynchronously and they
appear to be regulated so as not to interfere with each other. With respect to the
gait phases, activity in the TA muscles is evident only at the beginning of the stance
phase upon foot contact and during the swing phase (to ensure foot clearance).
Conversely, activity in the MG muscles is found mostly during the stance phases to
overcome loading and generate propulsive forces. The regular alteration in muscle
activity of these and many other muscles creates the net repetitive flexion and
extension movements in the involved joints. The consequence is a force exerted on
the ground (seen in the figure as the vertical component of the ground reaction
force), and finally a displacement of the center of body mass.
The emergence of these motor patterns takes place automatically and does not
necessarily accompany one’s intention. Indeed, we can walk or run while talking
with a friend. When dashing to catch a baseball, our attention is directed toward the
ball or the forthcoming action after catching the ball rather than on the details of the
5 Training Locomotor Function: From a Perspective of the Underlying Neural. . . 51
Fig. 5.1 Electromyographic (EMG) activities in lower limb muscles (TA tibialis anterior, MG
medial gastrocnemius, RF rectus femoris, BF biceps femoris) in a human subject while walking
comfortably. Calibration bars represent 100 μV and 500 N (vertical), and 1 second (horizontal),
respectively
ongoing running movement itself. How, then, are these highly coordinated move-
ments possible with little or no conscious awareness? We next describe the neural
mechanisms responsible for generating the motor output based on basic studies both
in animals and humans.
5.3 Central Pattern Generators: CPGs
5.3.1 Overview and Direct Evidences from Animal Studies
It is well known that human neonates exhibit stepping-like movements in their

lower extremities when the center of body mass is moved forward by external
support (Yang et al. 1998). Cats that have undergone surgical decerebration show
stepping-like hindlimb movements in accordance with the treadmill movement
underneath them (Barbeau and Rossignol 1987). These behavioral responses
clearly indicate that the initiation of locomotory behavior does not require a
contribution of the higher structures in the brain. Rather, as much research has
shown, there exist functional networks in the spinal cord that are specialized for
executing locomotory movements and play a significant role in generating rhythmic
motor patterns. These functional networks are referred to as “central pattern
generators” (CPGs) and consist of sets of interneurons in the spinal cord.
The original concept of spinal CPGs was probably first addressed in a study by
Sherrington (1910), and was closely followed by work by Brown (1911) who, using
a combination of low spinal transection and de-afferentation in a cat preparation,
demonstrated the existence of a movement control circuit in the lumbar region of
the spinal cord. In the last several decades, such neural mechanisms have been
extensively and systematically studied in both invertebrates (for example, locust
flight, Wilson 1961) and vertebrates such as swimming in the lamprey (Grillner
et al. 1981; Wallén and Williams 1984) and walking in the cat (Grillner and
Zangger 1979). In one example, Wallén and Williams (1984) obtained bilateral
electromyographic recordings from two different segments in the spinal cord of an
intact lamprey. The results, as expected, showed rhythmic activity, with a slight
time lag between the segments and alternating patterns with a lack of co-activation
contralaterally. More surprising and intriguing was the result that similar neural
activity was evident even under fictive locomotion with an isolated spinal cord.
That is, the rhythmic motor patterns were maintained without any external input. In
earlier work, patterned activities had even been identified in an isolated in vitro
spinal cord utilizing pharmacological intervention (Cohen and Wallén 1980). From
these results and others, CPGs were understood to be neural networks being capable
of the formation of rhythmic patterns, even in absence of external inputs such as
those from the brain or the periphery (Dietz 2002).
5.3.2 Possible Existence of CPGs in Humans
Despite the considerable body of evidence for the existence of CPGs in animals, the
possibility of its existence in humans has been a long-disputed question. The
presence of CPGs in humans has become evident only after studies by Calancie
et al. (1994) and subsequently by Dimitrijevic et al. (1998) in spinal-cord injured
patients. With a chronic, complete severing of the spinal-cord (C5 at the highest and
T8 at the lowest in the injury site) with patients lying in a supine posture, continuous
epidural stimulation was delivered to the lumbosacral region with a constant
frequency (Dimitrijevic et al. 1998). The result clearly showed that, despite the
trained stimulation, phasic locomotor-like EMG activities as low as 4 Hz (in the
stepping frequency) were evident when the stimulation was delivered at a certain
frequency (mostly between 25 and 50 Hz). The results were later supported by
Gerasimenko et al. (2010) using less invasive experimental procedures. Electro-
magnetic stimulation applied to the spinal cord in the lower thoracic region
(T11-T12) from the surface of the skin resulted in the emergence of involuntary
stepping-like movements in the lower limb among non-injured subjects while lying
in the decubitus position and relatively free from gravitational forces. The move-
ments were further enhanced by the addition of vibration to the leg muscles
(mimicking proprioceptive input). These results combined suggest the existence
of CPGs in humans, just as was found in the animal models utilized in the earlier
studies. From these studies it is also clear that, minimally, CPGs in humans are
located in the lower region of the spinal cord.
Other studies have addressed the location of CPGs in the spinal cord of humans.
Dietz et al. (1999) compared locomotory function in chronic spinal-cord injured
patients with different injury sites to that of normal subjects. Deficits in locomotory
function took place in a phased manner depending on the injury level rather than a
major, stepwise change at a particular level (Dietz et al. 1999). CPGs were therefore
shown to be located over a wide range in the spinal cord (up to the cervical level)
and not necessarily localized in the lower regions. The fact that CPGs are located as
high as the cervical level may be linked to recent results showing that upper limb
movements during locomotion are also under the control of CPGs (for review, see
Zehr and Duysens 2004). This higher location was demonstrated on the basis of
phase dependent modulation and reflex reversal in lower limb cutaneous reflex
responses during arm cycling (Zehr and Kido 2001).
5.4 Influence of Supraspinal and Afferent Input
Execution of locomotion movements is known to be affected by both supraspinal

inputs and inputs from the periphery. Schubert et al. (1997) investigated a possible
contribution of the corticospinal tract using transcranial magnetic stimulation
(TMS). By applying the magnetic stimulation to the motor cortex under different
phases of normal walking, they revealed that responses in the ankle flexor TA and
extensor MG muscles were modulated in a phase-dependent manner, independently
of the background EMG level. Moreover, the degree of facilitation relative to the
amplitude of the evoked potentials during tonic contraction of the muscles was
larger in the TA than in the MG muscle, thus showing a difference in the contri-
bution of corticospinal pathways to flexors and extensors. A recent study further
revealed that phase-dependent modulation of corticospinal excitability took place
even under passive stepping induced by use of a robotic driven-gait orthosis
(Kamibayashi et al. 2009). Since the latter result was not observed during the
same passive stepping task if it was executed in air, the authors concluded that
load-afferent input played a significant role in the phase-related modulation of
corticospinal excitability.
A similar phase-related modulation was observed in the excitability of spinally-
mediated reflex pathways. Nakajima et al. (2008) demonstrated that the cutaneous
reflex in the TA muscle was modulated in the same way in a normal gait as when
walking passively. Despite a constant background EMG level, the response was
prominently enhanced only under particular phases where the activity of the
particular muscle was essential for normal walking such as at the end of the stance
phase and at the beginning of the swing phases. The modulation however, was not
evident when the stepping movement was executed in the air (thus with no loading).
Using the monosynaptic H-reflex, Kamibayashi et al. (2010) showed similar phase-
dependent modulation. These last two studies demonstrate that afferent input is also
important in shaping locomotion related motor output.
5.5 Plasticity in the CNS and Its Specificity
In addition to the relatively automatic mechanisms involved in the control of

locomotory movements, the mammalian CNS also has the potential to undergo
plastic reorganization in response to particular stimuli. This reorganization
involved both anatomical and functional alterations (Raineteau and Schwab
2001). Examples of the former are lesions of neural pathways and their
reconnection through sprouting while the latter is demonstrated by functional
changes in pre-exiting pathways which occur and include changes in the efficacy
of synaptic transmission and level of excitability of the various pathways (for
review, see Zehr 2006; Wolpaw 2007). This plastic property of the CNS is an
important part of what is addressed by training procedures that are designed to
improve motor performance.
It has been demonstrated that functional alterations can take place even in the
simplest behavior of the CNS, spinally mediated monosynaptic reflexes. Operant
conditioning in monkeys (Wolpaw et al. 1986), rats (Chen and Wolpaw 1995), mice
(Carp et al. 2006), and recently in humans (Thompson et al. 2009) provides clear
evidence of this alteration. Subjects underwent long term training (several weeks to
months) requiring them to either increase or decrease their reflex amplitude.
Rewards of juice or money were provided on the basis of their success. Weeks or
months after the initiation of the training, the reflex responses were successfully
enhanced or depressed in accordance with the requirement for gaining a reward.
Since the responses were those obtained under a constant background EMG level, it
is considered that some functional reorganization took place within the neural
system.
Along with the above mentioned examples, in humans particularly, evidence for
functional alternations in the CNS have been shown by cross-sectional comparisons
among populations with different physical backgrounds. For example, a compari-
son of the amplitudes of the electrically induced Hoffmann (H-) reflex between elite
ballet dancers and ordinary well-trained athletes clearly showed a suppressed
response in the ballet dancers (Nielsen et al. 1993). Similarly, another study
demonstrated that the amplitude of the H-reflex in comparison to that of sedentary
subjects was smaller in sprinters and greater in endurance runners (Maffiuletti
et al. 2001). These results clearly show that not only the amount but also the type
of physical activity has an influence on the functional alteration that is made. Thus,
functional reorganization is very context specific.
5.6 Specificity in Locomotion-Related Neural Mechanisms
5.6.1 Direct Evidence from Animal Studies
Given the specificity of functional alterations in particular neural pathways, a

question arising next upon construction of locomotion training is the possible
specificity in the neural mechanisms underlying locomotion movements. Recent
studies focusing on swimming behavior in larval zebrafish demonstrated specificity
in the spinal interneurons that were recruited depending on swimming modes with
different frequencies (McLean et al. 2007, 2008). For reference, it was previously
demonstrated that zebrafish have mainly two different modes (one with low fre-
quency to routinely swim and another with higher frequencies for quick move-
ments) (Budick and O’Malley 2000). McLean et al. (2008) showed that recruitment
of the spinal interneurons occurred depending on specific swimming frequencies
while that in motoneurons took place only incrementally following classic size
principle. That is, there was certain independence in the neural mechanisms
responsible for different swimming modes. From a phylogenetic perspective, it is
highly possible that the mechanisms underlying the basic function such as locomo-
tion are shared in humans as well.
5.6.2 Among Different Locomotion Modes in Humans
Given the results described above for lower animals, a reasonable question is
whether similar independence in the neural mechanisms underlying different loco-
motion modes exists in humans as well. From a phylogenetic perspective, it is
possible that basic functions such as locomotory control are shared across classes.
The authors recently addressed this possibility by focusing on two major modes of
locomotion in humans, walking and running (Ogawa et al. 2012). Obviously, both
walking and running are executed by movements in the same joints, and both
involve a repetition of patterns of flexion and extension in the hip, knee, and
ankle. Muscles responsible for the joint movements are therefore common, despite
obvious differences in the way they are used in the two locomotory modes
(Prilutsky and Gregor 2001; Sasaki and Neptune 2006; Cappellini et al. 2006).
The fact that both systems use the same muscles indicates that the neural mecha-
nisms, at least at the level of the motoneurons (the final common pathway), must be
shared to some degree across the two locomotory modes.
Since the neurons that provide input to the motoneurons in humans cannot be
directly accessed in humans, the authors instead utilized a behavioral experiment on
the basis of motor adaptation paradigms which involved walking and running on a
split-belt treadmill. In has been known that human walking is flexible enough to
meet demand in any given circumstance. In fact, humans can even walk in an
environment as novel as a split-belt treadmill where two belts (one underneith each
limb) run at different velocities. Moreover, continued walking under the novel
two belt environment eventually leads to an acquisition of a “novel movement
pattern” that is particularly appropriate for walking on the two belts which are
moving at different speeds. Upon return to the normal belt condition, where the
two belts are moving at the same speed, significant aftereffect are seen which
involve asymmetrical movement patterns. These patterns were those acquired
through previsouly walking on the two speed treadmill (Reisman et al. 2005).
Provided that similar aftereffect can be obtained in running as well (Ogawa
et al. 2012), the focus of the interest was whther the movement patterns acquired
in one mode (walk or run) are shared between the other. The resuits are described
below.
In contrast to the emergence of pronimently asymmetrical movement patterns in
walking after adapting to walk on the asymmetrically driven treadmill, the subjects
could walk normally (thus, with minimal emergence of asymmetrical movement
patterns) after adapting to run. Conversely, the subjects could run normally after
adaptation in walking in spite of the emergence of asymmetrical movement patterns
after adaptation in walking. Furthermore, even though the subjects could walk as
normal after adaptation in running as in above, beginning to run after centain period
of time in walking resulted in abrupt emergence of asymmetrical movement
patterns. Similarly, while running as normal after adapting to walk on the asym-
metry, beginning to run resulted in asymmetrical movement patterns. To summa-
rize, the asymmetrical movement patterns acquired in one mode transfered to
another only partially and were maintained independently of the subsequent exe-
cution in another. The results therefore edvocate the possibility that walking and
running, despite the use of duplicated joints and muscles, are controlled by different
functional networks being capable of the respective modes (Ogawa et al. 2012).
Along with the discussion on walking and running, some other recent studies
have addressed specificity in neural networks among other locomotion contexts
such as the direction and the velocity of walking (Choi and Bastian 2007;
Vasudevan and Bastian 2010). These results combined suggest that the locomotion
related neural mechanism is not a general module that are shared among different
situation demand, but is more specialized depending on the respective context.
5.7 Summary
The rhythmic and highly stereotypical motor patterns in locomotion movements are
generated through the specialized neural mechanisms in the CNS and controlled
largely automatically. The CNS can undergo functional reorganization through
repetitive use of the neural networks. The functional reorganization however,
takes place very specifically depending on given types or contexts of the physical
activity. Neural mechanisms responsible for locomotion movements have also been
described as having certain task and context dependency. Given these backgrounds,
construction of adequate intervention program for acquisition of locomotion
performance should account for the “automaticity” and the “specificity”. Lack in
these perspectives may perhaps, leads to interference with generation of physio-
logical movement and inefficiency of the training.
References
Barbeau H, Rossignol S (1987) Recovery of locomotion after chronic spinalization in the adult cat.
Brown TG (1911) The intrinsic factors in the act of progression in the mammal. Proc R Soc Lond B
84:308–319
Budick SA, O’Malley DM (2000) Locomotor repertoire of the larval zebrafish: swimming, turning
and prey capture. J Exp Biol 203:2565–2579
Calancie B, Needham-Shropshire B, Jacobs P, Willer K, Zych G, Green BA (1994) Involuntary
stepping after chronic spinal cord injury. Evidence for a central rhythm generator for locomo-
tion in man. Brain 117:1143–1159
Cappellini G, Ivanenko YP, Poppele RE, Lacquaniti F (2006) Motor patterns in human walking
and running. J Neurophysiol 95:3426–3437
Carp JS, Tennissen AM, Chen XY, Wolpaw JR (2006) H-reflex operant conditioning in mice. J
Chen XY, Wolpaw JR (1995) Operant conditioning of H-reflex in freely moving rats. J
Choi JT, Bastian AJ (2007) Adaptation reveals independent control networks for human walking.
Nat Neurosci 10:1055–1062
Cohen AH, Wallén P (1980) The neuronal correlate of locomotion in fish. “Fictive swimming”
induced in an in vitro preparation of the lamprey spinal cord. Exp Brain Res 41:11–18
Dietz V (2002) Proprioception and locomotor disorders. Nat Rev Neurosci 3:781–790
Dietz V, Nakazawa K, Wirz M, Erni T (1999) Level of spinal cord lesion determines locomotor
activity in spinal man. Exp Brain Res 128:405–409
Dimitrijevic MR, Gerasimenko Y, Pinter MM (1998) Evidence for a spinal central pattern
generator in humans. Ann N Y Acad Sci 860:360–376
Gerasimenko Y, Gorodnichev R, Machueva E, Pivovarova E, Semyenov D, Savochin A, Roy RR,
Edgerton VR (2010) Novel and direct access to the human locomotor spinal circuitry. J
Neurosci 30:3700–3708
Grillner S, Zangger P (1979) On the central generation of locomotion in the low spinal cat. Exp
Grillner S, McClellan A, Perret C (1981) Entrainment of the spinal pattern generators for
swimming by mechano-sensitive elements in the lamprey spinal cord in vitro. Brain Res
217:380–386
Kamibayashi K, Nakajima T, Takahashi M, Akai M, Nakazawa K (2009) Facilitation of
corticospinal excitability in the tibialis anterior muscle during robot-assisted passive stepping
in humans. Eur J Neurosci 30:100–109
Kamibayashi K, Nakajima T, Fujita M, Takahashi M, Ogawa T, Akai M, Nakazawa K (2010)
Effect of sensory inputs on the soleus H-reflex amplitude during robotic passive stepping in
humans. Exp Brain Res 202:385–395
Maffiuletti NA, Martin A, Babault N, Pensini M, Lucas B, Schieppati M (2001) Electrical and
mechanical H(max)-to-M(max) ratio in power- and endurance-trained athletes. J Appl Physiol
90:3–9
McLean DL, Fan J, Higashijima S, Hale ME, Fetcho JR (2007) A topographic map of recruitment
in spinal cord. Nature 446:71–75
McLean DL, Masino MA, Koh IY, Lindquist WB, Fetcho JR (2008) Continuous shifts in the active
set of spinal interneurons during changes in locomotor speed. Nat Neurosci 11:1419–1429
Nakajima T, Kamibayashi K, Takahashi M, Komiyama T, Akai M, Nakazawa K (2008) Load-
related modulation of cutaneous reflexes in the tibialis anterior muscle during passive walking
in humans. Eur J Neurosci 27:1566–1576
Nielsen J, Crone C, Hultborn H (1993) H-reflexes are smaller in dancers from The Royal Danish
Ballet than in well-trained athletes. Eur J Appl Physiol Occup Physiol 66:116–121
Ogawa T, Kawashima N, Ogata T, Nakazawa K (2012) Limited transfer of newly acquired
movement patterns across walking and running in humans. PLoS One 7:e46349
Prilutsky BI, Gregor RJ (2001) Swing- and support-related muscle actions differently trigger
human walk-run and run walk transitions. J Exp Biol 204:2277–2287
Raineteau O, Schwab ME (2001) Plasticity of motor systems after incomplete spinal cord injury.
Nat Rev Neurosci 2:263–273
Reisman DS, Block HJ, Bastian AJ (2005) Interlimb coordination during locomotion: what can be
adapted and stored? J Neurophysiol 94:2403–2415
Sasaki K, Neptune RR (2006) Differences in muscle function during walking and running at the
same speed. J Biomech 39:2005–2013
Schubert M, Curt A, Jensen L, Dietz V (1997) Corticospinal input in human gait: modulation of
magnetically evoked motor responses. Exp Brain Res 115:234–246
Sherrington CS (1910) Flexion-reflex of the limb, crossed extension-reflex, and reflex stepping and
standing. J Physiol 40:28–121
Thompson AK, Chen XY, Wolpaw JR (2009) Acquisition of a simple motor skill: task-dependent
adaptation plus long-term change in the human soleus H-reflex. J Neurosci 29:5784–5792
Vasudevan EV, Bastian AJ (2010) Split-belt treadmill adaptation shows different functional
networks for fast and slow human walking. J Neurophysiol 103:183–191
Wallén P, Williams TL (1984) Fictive locomotion in the lamprey spinal cord in vitro compared
with swimming in the intact and spinal animal. J Physiol 347:225–239
Wilson DM (1961) The central nervous control of flight in a locust. J Exp BioI 38:471–499
Wolpaw JR (2007) Spinal cord plasticity in acquisition and maintenance of motor skills. Acta
Physiol 189:155–169
Wolpaw JR, O’Keefe JA, Noonan PA, Sanders MG (1986) Adaptive plasticity in primate spinal
stretch reflex: persistence. J Neurophysiol 55:272–279
Yang JF, Stephens MJ, Vishram R (1998) Infant stepping: a method to study the sensory control of
human walking. J Physiol 507:927–937
Zehr EP (2006) Training-induced adaptive plasticity in human somatosensory reflex pathways. J
Appl Physiol 101:1783–1794
Zehr EP, Duysens J (2004) Regulation of arm and leg movement during human locomotion.
Neuroscientist 10:347–361
Zehr EP, Kido A (2001) Neural control of rhythmic, cyclical human arm movement: task
dependency, nerve specificity and phase modulation of cutaneous reflexes. J Physiol
537:1033–1045
Chapter 6
On the Structure of Movement Preparation:
Inferences from Motor Schema Theory
Lu Xu, Werner Sommer, and Hiroaki Masaki
Abstract In the first part of this chapter we review the schema theory of motor
control (Schmidt 1975). We emphasize its two main components, generalized
motor programs (GMP) and parameterization. Accumulated evidence suggests
the independence of GMP and parameterization. However, there is no consensus
whether or not they proceed in a fixed order with the assembly of GMPs first,
followed by parameterization. These questions could be studied in more detail by
using event-related brain potentials. We review this approach in the second part.
Keywords Motor control • Motor schema • Generalized motor program •

Parameterization
6.1 Motor Control Theories
6.1.1 Two Types of Control Systems
Motor control is essential in daily life as well as in sports activities. It involves

commands from central and peripheral neural systems to muscles and joints and
may be achieved by open-loop and/or closed-loop systems (for reviews, see Magill
and Anderson 2007). Both types of systems include a movement control center,
movement commands, and movement effectors. In open-loop control system, all
information is contained in the commands sent from the control center to the
effectors and output activity cannot be changed once the commands have been
L. Xu
Graduate School of Sport Sciences, Waseda University, Saitama, Japan
Japan Society for the Promotion of Science, Tokyo, Japan
e-mail: luxupsy@163.com
W. Sommer
Department of Psychology, Humboldt University, Berlin, Germany
e-mail: werner.sommer@cms.hu-berlin.de
H. Masaki (*)
e-mail: masaki@waseda.jp

60 L. Xu et al.
sent. In contrast, in a closed-loop control system, feedback containing both extero-

ceptive and proprioceptive information is sent to the movement control center. The
control center can modify the commands based on the feedback. The updated
commands are then sent to the effectors.
There are two main differences between open- and closed-loop control systems.
First, feedback is included in the closed-loop control system sent from the effector
to the movement control center, but not in the open-loop control system; second, in
the closed-loop control system, feedback information exists in updated commands
sent from the movement control center to the movement effectors. Basically,
whether the planned movement relies on an open-loop or the closed-loop system
depends on duration of the movement. Ballistic movements, for example, are too
rapid for the possibility of modification by proprioceptive feedback.
6.1.2 Adams’ Closed-Loop Theory
Focused on simple, self-paced, limb positioning movements, Adams formulated his

closed-loop theory (Adams 1971). In Adams’ theory, movement is initiated by a
processed named memory trace, which contains the necessary information about
the motor commands. Once all necessary information is collected in the memory
trace, commands will be sent to initiate the movement without involvement of any
feedback. The performance is then evaluated with respect to a reference called
perceptual trace which is derived from memories of earlier movements. This
process compares feedback information about the actual movement with the goal
movement. If any discrepancy is detected, motor commands will be modified and
sent to the effectors until a match is achieved upon which the movement will be
stopped.
6.1.3 Motor Schema Theory
Schmidt (1975) pointed out that Adams’ theory has difficulty explaining two
problems. The storage problem concerns the limited memory storage of various
movements; it would be very difficult to store representations of all possible
movements. The novelty problem relates to the difficulty of accounting for the
production of novel movements, which are not yet stored in the perceptual or
memory trace. To solve this inconsistency, Schmidt proposed his schema theory
for discrete actions. A schema is defined as a set of rules which act as a basis for
decisions. A schema is formed by related information gathered from previous
experiences. For example, after seeing lots of cars in real life as well as on TV
and in books, you will likely develop a set of rules which form your concept of
“car”. This set of rules constitutes a schema. Using this schema, one can recognize
if an object is a car, even if they have never seen this specific one before.
6 On the Structure of Movement Preparation: Inferences from Motor Schema Theory 61
According to Schmidt’s schema theory, only a limited number of generalized

motor programs (GMP) concerning fixed movement structures are stored in mem-
ory. The performer has to apply certain movement parameters to an abstract GMP
before a specific movement can be executed. This process is governed by the motor
response schema, which has two components. One component is the recall schema,
which has the responsibility of applying specified response characteristics to a
certain GMP and executing a particular action. The second component is the
recognition schema, which monitors performance and modifies movement com-
mands based on feedback.
According to Schmidt’s schema theory, movement features may be fixed or
variable. Fixed features are held to be associated with invariant spatiotemporal
motor patterns including sequencing of actions and relative timing. In contrast,
examples of variable parameters include overall duration and overall force. It
should be noted that some studies (Shea and Wulf 2005; Leuthold and Jentzsch
2011; plus many others) refer to relative force as a GMP, however, Schmidt (2003)
has pointed out that relative force is not invariant. Rather, relative force is a
parameter that scales muscles. Variable features allow the adaptation of individual
movements to fit the requirements of any given situation.
A typical example to illustrate motor schema involves each person’s written
signature, which remains relatively constant whether it is large or small, or written
by hand or foot (Schmidt and Lee 2013). The spatial and temporal characteristics of
a written signature are GMPs, whereas the size of the handwriting and the effector
used are parameters that need to be assembled.
6.2 Studies on Motor Schema Theory
6.2.1 The Independence of GMP and Parameterization
Previous studies have shown that some factors that are beneficial for the learning of
GMPs can have negative effects on parameterization learning. This supports the
independence of the two processes (Wulf and Schmidt 1989; Wulf et al. 1993,
1994; Whitacre and Shea 2000; Shea and Wulf 2005). For example, Wulf and Lee
(1993) asked participants to sequentially press four buttons at a fixed ratio of
intervals (relative timing, i.e., 1: 2: 1.5) in different goal movement time (overall
timing, i.e., task A: 200–400–300 ms; task B: 250–500–450 ms; task C: 300–600–
450 ms, respectively.). They separated overall errors by calculating differences
between actual and target movement time, and differences between actual and
target intervals. The results suggested that learning of relative timing benefited
from random practice, whereas blocked practice improved the learning of absolute
timing. The dissociation of errors in relative and absolute timing agrees with the
concept that GMP and parameterization learning are independent.
62 L. Xu et al.
6.2.2 The Ordering of GMP and Parameterization
Schmidt and Lee (2013) assert that following stimulus identification, the GMP is
retrieved from long-term memory in the response selection stage. Parameterization
is held to occur during the following motor programming stage. However, there is
conflicting evidence regarding the ordering of GMP and parameterization.
For example, Lee, Elliott, and Carnahan (1987) adopted Rosenbaum’s precue
paradigm (Rosenbaum 1980, 1983), in which a precue (S1) provides either partial,
full, or no information about the upcoming movement. After a certain interval, the
so-called foreperiod, the imperative stimulus (S2) provides full information about
the required movement. To the extent that participants utilize the precue during the
foreperiod to at least partially prepare for the subsequent movement, reaction time
(RT) decreases with increasing amounts of precue information. Lee and colleagues
presented precues which specified the type of upcoming actions (GMP), for exam-
ple, lifting or rotating objects versus no information on movement distance (param-
eters; far or near versus no information). The results showed that RTs were the same
regardless of the order in which the information about actions or movements was
provided in advance. Thus participants did not benefit from advance information
about the GMP required. Lee et al. (1987) concluded that their results supported a
variable order of GMP specification and parameterization rather than a fixed order,
as was specified for the GMP of the first model.
Wright, Black, Park, and Shea (2001) manipulated both relative timing and
overall duration in a sequential key tapping task. According to the Schema theory
(Schmidt 1975), the relative timing represents GMP and the overall duration is
considered as a parameter. Wright et al. (2001) provided their participants with
either no, partial (relative timing or overall duration), or full precue information.
They found shorter RTs when partial information about relative timing was
presented rather than when information of overall duration was provided. These
findings supported Schmidt’s assertion that parameterization occurs after prepara-
tion of GMP.
Leuthold and Jentzsch (2011) pointed out that in the experiment of Wright et al.,
both digits and letters served as stimuli. Thus the time demands for encoding the
stimuli might be involved in RT, which made the results ambiguous. Leuthold and
Jentzsch used the same task as Wright and coworkers, but redesigned the stimuli to
exclude potential effects brought about by stimulus identification. In the latter
experiment by Leuthold and Jentzsch, the RTs showed no difference between the
two partially precued conditions (relative timing versus overall duration). This
finding is at variance with a fixed order of GMP and parameterization.
6.2.3 Tools to Investigate Movement Preparation
Event-related potentials (ERPs) are extensively used in the study of human infor-
mation processing. ERPs are extracted from an electroencephalogram, usually by
averaging. ERPs consist of distinct components with functional specificity,
allowing the investigator to track psychologically meaningful processes with a
high temporal resolution. Here we review several movement-related ERP compo-
nents, which can serve as useful indices in the investigation of movement
preparation.
The contingent negative variation (CNV) (Walter et al. 1964) is a slow,
negative-going wave. It is usually observed in paradigms where a warning stimulus
(S1) is followed by an imperative stimulus (S2) that requires a response. Järviletho
and Frühstorfer (1970) first reported a frontally-distributed early component asso-
ciated with an orienting response to S1, and a centrally-distributed late component
associated with response preparation (see also, Loveless and Sanford 1974). Later
studies have reported larger late CNVs for stronger force exertions (Low and
McSherry 1968), more rapid force increments (Van Boxtel et al. 1993), and faster
responses that emphasize response speed (e.g., Rohrbaugh et al. 1976). This
suggests relationships between the late CNV and certain aspects of motor control.
Another slow wave in the ERP, which precedes self-paced movements, is
referred to as the readiness potential (RP) (Vaughan et al. 1968). It emerges up to
2 s prior to a voluntary movement. This potential includes a gradual negativity,
Bereitschaftspotential (BP) (Kornhuber and Deecke 1965), and a steeper negative
slope (NS’). The RP is considered to be included, at least partly, in the late CNV,
and the distribution of the later CNV is more complex than the RP (for details, see
Brunia 2003).
From the RP, a useful ERP component can be extracted. This is the so-called
lateralized readiness potential (LRP). It is obtained by subtracting the RP recorded
over electrode sites ipsilateral to the responding hand from the RP recorded over
contralateral sites. The LRP is considered to represent the activation of effector-
specific response-related processes (Coles 1989; Miller et al. 1992). Thus, as soon
as the LRP deviates from zero in a negative voltage direction, the response hand
required for the task is activated. More specifically, it has been shown that the LRP
starts after response hand selection and at the beginning of motor programming
(Masaki et al. 2004).
6.2.4 Neural Correlates of GMP and Parameterization
Most studies concerning the GMP and parameterization have utilized behavioral
data. However, by recording ERPs as noted above, we can reveal the time course of
different processes during movement preparation. Only a few studies have tested
the schema theory in association with neural processing.
64 L. Xu et al.
Masaki, Takasawa, and Yamazaki (1998) examined the effects of GMP and
parameterization on the RP. Participants were asked to exert different levels of
target force successively in three conditions. In the single target condition, they
repeatedly produced a single target force (13 N) on every trial. In the multiple target
condition, they produced one of three target forces (5, 13, or 21 N). These forces
were alternated pseudo-randomly in every trial. According to the schema theory,
these two conditions were involved with parameterization. In the third condition
(the tracking condition), different GMPs were involved in the three different tasks.
In addition to target force production (13 N) the participants were asked to perform
two different tracking tasks. Although only the 13 N target trials were compared
among the three conditions, larger RP amplitudes were found for the tracking and
multiple target tasks than for the single target task. Hence the RP amplitude is
associated with both GMP and parameterization and may be used for online-
investigation of these processes.
In their study mentioned earlier, Leuthold and Jentzsch (2011) asked participants
to press the same key successively with required intervals (relative timing) and
overall duration in a precue paradigm. Precues provided either no, partial, or full
information about the forthcoming movement. The authors hypothesized that if
GMP was always prepared in advance of parameterization, the participants would
benefit from precue about GMP. Results showed larger CNVs over the left central
regions with an increasing amount of information provided by the precue. However,
a foreperiod LRP was only observed in the full precue condition. The absence of a
foreperiod LRP in both partial precue conditions was seen to be at variance with a
fixed order of GMP and parameterization. This supports their conclusions from the
RT results.
6.3 Conclusion
Motor schema theory plays an important role in motor learning and control. It
suggests how to produce particular actions, including novel ones, without an
extensive burden on memory circuits. Most evidence on the function of motor
schema supports the existence and independence of the GMP and parameterization.
However, the sequence of GMP and parameterization during movement prepara-
tion remains unclear, as do the neural mechanisms that underlie their functioning.
More evidence from psychophysiological and brain imaging studies will shed light
on the neural mechanisms that are involved in motor programming.
Acknowledgement This study was supported by JSPS Research Fellowships for Young Scien-
tists, and GCOE scholarship for international student to L. Xu, and a Grant-in-Aid for Scientific
Research (C) 24530925 from the JSPS to H. Masaki.
References
Adams JA (1971) A closed-loop theory of motor learning. J Mot Behav 3(2):111–150

Brunia CH (2003) CNV and SPN: indices of anticipatory behavior. In: The Bereitschaftspotential.
Springer, New York, pp 207–227
Coles MG (1989) Modern mind-brain reading: psychophysiology, physiology, and cognition.
Psychophysiology 26(3):251–269
Järvilehto T, Fruhstorfer H (1970) Differentiation between slow cortical potentials associated with
motor and mental acts in man. Exp Brain Res 11(3):309–317
Kornhuber HH, Deecke L (1965) Hirnpotentialänderungen bei Willkürbewegungen und passiven
Bewegungen des Menschen: Bereitschaftspotential und reafferente Potentiale. Pflüger’s
Archiv für die gesamte Physiologie des Menschen und der Tiere 284(1):1–17
Lee TD, Elliott D, Carnahan H (1987) The preparation of actions and parameters of action: a fixed
or variable process? Acta Psychol (Amst) 66(1):83–102
Leuthold H, Jentzsch I (2011) Are temporal response features prepared in fixed order? Inferences
from movement-related potentials. Psychophysiology 48(5):633–644
Loveless NE, Sanford AJ (1974) Effects of age on the contingent negative variation and prepa-
ratory set in a reaction-time task. J Gerontol 29(1):52–63
Low MD, McSherry JW (1968) Further observations of psychological factors involved in CNV
genesis. Electroencephalogr Clin Neurophysiol 25(3):203–207
Magill RA, Anderson DI (2007) Motor learning and control: concepts and applications. McGraw-
Hill, New York
Masaki H, Takasawa N, Yamazaki K (1998) The effect of motor programming and the parameter-
applying-processes on the readiness potential. Jpn J Physiol Psychol Psychophysiol 16(2):
85–91
Masaki H, Wild-wall N, Sangals J, Sommer W (2004) The functional locus of the lateralized
readiness potential. Psychophysiology 41(2):220–230
Miller J, Riehle A, Requin J (1992) Effects of preliminary perceptual output on neuronal activity
of the primary motor cortex. J Exp Psychol Hum Percept Perform 18(4):1121–1138
Rohrbaugh JW, Syndulko K, Lindsley DB (1976) Brain wave components of the contingent
negative variation in humans. Science 191(4231):1055–1057
Rosenbaum DA (1980) Human movement initiation: specification of arm, direction, and extent. J
Exp Psychol Gen 109(4):444–474
Rosenbaum DA (1983) The movement precuing technique: assumptions, applications, and exten-
sions. Adv Psychol 12:231–274
Schmidt RA (1975) A schema theory of discrete motor skill learning. Psychol Rev 82(4):225–260
Schmidt RA (2003) Motor schema theory after 27 years: reflections and implications for a new
theory. Res Q Exerc Sport 74(4):366–375
Schmidt R, Lee T (2013) Motor learning and performance, 5E with web study guide: from
principles to application. Human Kinetics, Champaign, pp 107–108
Shea CH, Wulf G (2005) Schema theory: a critical appraisal and reevaluation. J Mot Behav
37(2):85–102
Van Boxtel GJ, Van den Boogaart B, Brunia CH (1993) The contingent negative variation in a
choice reaction time task. J Psychophysiol 7(1):11–23
Vaughan HG Jr, Costa LD, Ritter W (1968) Topography of the human motor potential. Electroen-
cephalogr Clin Neurophysiol 25(1):1–10
Walter W, Cooper R, Aldridge VJ, McCallum WC, Winter AL (1964) Contingent negative
variation: an electric sign of sensori-motor association and expectancy in the human brain.
Nature 203:380–384
Whitacre CA, Shea CH (2000) Performance and learning of generalized motor programs: relative
(GMP) and absolute (parameter) errors. J Mot Behav 32(2):163–175
Wright DL, Black C, Park JH, Shea CH (2001) Planning and executing simple movements:
contributions of relative-time and overall-duration specification. J Mot Behav 33(3):273–285
66 L. Xu et al.
Wulf G, Lee TD (1993) Contextual interference in movements of the same class: differential
effects on program and parameter learning. J Mot Behav 25(4):254–263
Wulf G, Schmidt RA (1989) The learning of generalized motor programs: reducing the relative
frequency of knowledge of results enhances memory. J Exp Psychol Learn Mem Cogn
15(4):748–757
Wulf G, Schmidt RA, Deubel H (1993) Reduced feedback frequency enhances generalized motor
program learning but not parameterization learning. J Exp Psychol Learn Mem Cogn 19(5):
1134–1150
Wulf G, Lee TD, Schmidt RA (1994) Reducing knowledge of results about relative versus absolute
timing: differential effects on learning. J Mot Behav 26(4):362–369
Chapter 7
Muscle Relaxation and Sports
Kouki Kato and Kazuyuki Kanosue
Abstract Muscle relaxation is important in both daily life and sports. In addition,
movement disorders such as Parkinson’s disease and Dystonia are often character-
ized by deficits of muscle relaxation. Activation of some brain areas has been
demonstrated during muscle relaxation, suggesting that muscle relaxation might
be an “active process”, and not just the end of contraction. In the sports field,
beginners or players under stress often find relaxation of particular muscles is
difficult. However, the mechanisms underlying muscle relaxation during sports is
not well understood.
Keywords Inhibition • Motor control • No-go task
7.1 Muscle Relaxation
Our daily life involves many motor activities, all of which require a fine control of
muscle contraction and relaxation. Although control over muscle contraction has
been given much attention, this addresses only half of the story. Muscle relaxation
is equally important. The mechanisms underlying precise control over reductions in
muscle contraction intensity also merit serious consideration. In the fields of sports
and music, inappropriate muscle contraction and relaxation are often observed.
However, a concrete knowledge base that would allow for the development of
solutions to such problems is unavailable. In addition, this knowledge base could
also be utilized to solve problems associated with neurological maladies such as
Parkinsonism and Dystonia in which both involuntary muscle contractions and
atonia are common (Grasso et al. 1996; Buccolieri et al. 2004a).
K. Kato (*)
e-mail: kouki0104@gmail.com
K. Kanosue

68 K. Kato and K. Kanosue
The term “muscle relaxation” is widely used, but with many different meanings.
In the sports field, muscle relaxation generally refers to a state in which the athlete
doesn’t produce any unnecessary contractions. In some cases, relaxation is used to
mean simply the end of contraction. In other situations, the word is used to describe
a state where no contraction is present (i.e. the resting condition). In this review, we
will utilize the latter definition for muscle relaxation.
Serious attention to the study of “muscle relaxation” began at the very end of the
twentieth century (Terada et al. 1995; Toma et al. 1999). Until then, muscle
relaxation was simply seen as the end of contraction. Toma et al. (1999) launched
a detailed study of relaxation utilizing functional magnetic resonance imagery
(fMRI). Their study indicated that activity in the primary motor cortex
(M1) “increased” during voluntary muscle relaxation as well as during muscle
contraction. This highlighted the notion that muscle relaxation is an “active pro-
cess” requiring cortical activation (Toma et al. 1999, 2000; Motawar et al. 2012).
Since then, brain activity during muscle relaxation has been investigated with many
different techniques, including transcranial magnetic stimulation (TMS), fMRI and
electroencephalogram (EEG). We discuss these studies in next section.
7.2 Disorder Associated with Muscle Relaxation
Deficits in muscle relaxation are involved in the pathophysiology of movement

disorders that occur in such maladies as Parkinson’s disease, dystonia, and stroke.
These diseases all involve complex and different constellations of symptoms that
involve relaxation to some degree. While the overt behavior varies, it can include
bradykinesia, rigidity, sustained muscle contractions, muscle spasms, and tremors.
7.2.1 Parkinson’s Disease
Parkinson’s disease is a degenerative disorder of the central nervous system. It is

well known that atrophy of the dopaminergic cells in the substantia nigra is a
primary cause of the motor symptoms seen in Parkinson’s disease. Early in the
course of the disease, movement-related symptoms are the most obvious diagnostic
characteristic, such as shakiness of body parts, rigidity of joints, and a slowness of
movement. Recently, it has been demonstrated that the behavior of relaxation tasks
changes concurrently with the development of the typical motor symptoms of
Parkinson’s disease (Neely et al. 2013). Relaxation can be quantified by mea-
suring the reaction time of relaxation, which is determined by monitoring the
force change from a set value to zero. This duration progressively becomes longer
in patients with Parkinson’s disease as compared to normals (Jordan et al. 1992;
Wing 1988). Furthermore, longer electromyographic (EMG) reaction times of
relaxation are associated with higher bradykinesia scores in Parkinson’s patients,
7 Muscle Relaxation and Sports 69
even for those who are medicated (Grasso et al. 1996). A separate study, utilizing
the H-reflex technique, has shown that the shorter reaction time of relaxation in
healthy individuals can be at least partially attributed to a presynaptic inhibitory
action in the spinal cord (Schieppati et al. 1986). Since presynaptic inhibition is
reduced in patients with Parkinson’s disease, and is more prominent on the side
with the symptoms, a deficit in presynaptic inhibition could be one of the reasons
why the reaction time for relaxation is longer for the Parkinson’s patients (Lelli
et al. 1991).
7.2.2 Dystonia
Dystonia is defined as a syndrome which involves sustained involuntary muscle

contractions, frequent twisting and repetitive movements, and/or abnormal postures
(Fahn 1988). It has been suggested that dystonia might be a result of a functional
disturbance of the basal ganglia (Berardelli et al. 1998). EMG studies show that
patients with dystonia are slower than normal when they try to carry out desired
movements. Abnormal co-contraction of agonist – antagonist muscles is well
documented in dystonia patients, and this may partially explain the observed
insufficient muscle relaxation of antagonistic muscles (van der Kamp et al. 1989;
Berardelli et al. 1996). EMG bursts are usually prolonged when such patients carry
out simple rapid movements (van der Kamp et al. 1989; Berardelli et al. 1996).
Furthermore, Buccolieri and colleagues (2004a) note that although the reaction
time of simple muscle contraction is not different between patients and healthy
individuals, the reaction time of muscle relaxation is prolonged in patients with
dystonia (Buccolieri et al. 2004a). It has been demonstrated that in arm-reaching
movements, the motion is slower and more variable in dystonia patients as com-
pared with normal, and that the deceleration phase is abnormally prolonged
(Inzelberg et al. 1995). Based on these studies, it is clear that patients with dystonia
tend to exhibit prolonged motion, especially during relaxation. Although the mech-
anisms of the delayed reaction time of muscle relaxation in dystonia are not been
well understood, they might be caused by deficits in the cortical area and the spinal
cord. One electroencephalogram (EEG) study showed that the amplitudes of
movement-related cortical potentials (MRCPs), which appear prior to self-initiated
voluntary movement and reflect movement preparation processing (Shibasaki and
Hallett 2006), were smaller before relaxation than those before contraction in
healthy normal participants (Rothwell et al. 1998). MRCPs decreased more during
voluntary muscle relaxation in patients of dystonia as compared to normals
(Yazawa et al. 1999). Moreover, reduced intracortical inhibition was shown in
dystonic patients studied with the paired pulse transcranial magnetic stimulation
(TMS) technique (Ridding et al. 1995; Chen et al. 1997). Other studies have
suggested that dystonia may be caused by defective function of presynaptic inhib-
itory mechanisms in the spinal cord (Priori et al. 1995; Lorenzano et al. 2000).
In sum, the above evidence indicates that both cortical and spinal inhibitory mech-
anism are likely involved in the decline of motor function seen in dystonic patients.
7.2.3 Stroke
Stroke is caused by a disturbance in the blood supply to the brain. Patients are often
inflicted with permanent damage which can produce chronic motor deficits and
diminished capacity to manipulate objects in the paretic hand (Gray et al. 1990;
Nakayama et al. 1994; Parker et al. 1986). These deficits may relate not only to
muscle contraction (Cruz et al. 2005; Kamper and Rymer 2001), but also to
relaxation in the paretic limb (e.g. release of an object) (Nowak et al. 2003, 2007;
Seo et al. 2009). A delay in relaxation from the paretic hand grip as well as
insufficient release of grip force was observed during grip-and-lift tasks in stroke
patients (Nowak et al. 2003, 2007). Seo et al. (2009) demonstrated that delays in
grip initiation (contraction) and termination (relaxation) were greatest for the
paretic hand (1.9 and 5.0 s), followed by the nonparetic hand (0.5 and 1.6 s), and
smallest for the hands of normal controls (0.2 and 0.4 s). These data show that
muscle alterations in patients are prolonged more in the relaxation phase than in the
contraction phase.
As described above, the distinctive symptoms of movement disorders seen in
neurological patients are often partly due to problems in muscle relaxation. How-
ever, the neural mechanisms involved in muscle relaxation are complex, and as yet
incompletely understood.
7.3 Neural Mechanisms of Muscle Relaxation
7.3.1 TMS Studies
The level of corticospinal excitability can be assessed from the amplitude of the
motor evoked potentials (MEPs). The MEPs decrease during relaxation of the
muscle as compared to the resting condition (Buccolieri et al. 2004b; Begum
et al. 2005; Motawar et al. 2012). However, the cause of this lower level of
corticospinal activity remains unclear. One possible mechanism is that muscle
relaxation may be mediated by increased intracortical inhibition (Buccolieri
et al. 2004b; Motawar et al. 2012), leading to the lowering or withdrawal of
corticospinal output (Kamper et al. 2003; Rothwell et al. 1998) (Fig. 7.1). Alterna-
tively (or possibly in addition), muscle relaxation may be mediated by the activa-
tion of spinal inhibitory interneurons with augmented inputs from the cortex
(Schieppati and Crenna 1984; Schieppati et al. 1986; Begum et al. 2005).
Fig. 7.1 Schematic

Buccolieri et al., 2005 Begum et al., 2005
diagram of possible
mechanisms for muscle Motawar et al., 2012
relaxation. One possible
mechanism is shown on the Cortical
Cortical level
left side: Increases in pyramidal
intracortical inhibition Intracortical
inhibitory neuron
would contribute to
decreased cortical interneuron
excitability in the target
muscle. On the right side is
another possible corticospinal
mechanism: A different neuron
corticospinal pathway from
Spinal level
the one that provides an
excitatory influence on
spinal motoneurons
contributes to muscle spinal inhibitory
relaxation spinal interneuron
motoneuron
muscle
relaxation
TMS can be used not only to investigate the excitability of the corticospinal
pathway, but also to evaluate the excitability of a variety of intracortical circuits.
One such circuit involves the GABAergic connections in the motor cortex (short
interval intracortical inhibition: SICI) (Kujirai et al. 1993). Reynolds and Ashby
(1999) showed that SICI decreases during contraction, and hence the cortical
excitability of the target muscle is increased. As for muscle relaxation, the SICI
increased during the termination of contraction, which could contribute to
decreased cortical excitability in the target muscle (Fig. 7.1 left; Buccolieri
et al. 2004b: Motawar et al. 2012). On the other hand, Begum et al. (2005) found
that SICI decreased prior to relaxation (thus increasing disinhibition). These authors
postulated that spinal inhibitory interneurons are activated due to a decrease of
SICI, and hence assist in relaxation of the target muscles (Fig. 7.1 right). After the
above statement was published, a debate ensued as to whether muscle relaxation is
executed via the cortex or spinal mechanisms. Recently, Motawar et al. (2012)
indicated that muscle relaxation accompanies enhancement of SICI, and then
induces a gradual increase of intracortical inhibition with progression of the
relaxation phase. Although the latter mechanism is still being investigated, at the
very least it is certain that muscle relaxation involves an increase in intracortical
inhibition.
7.3.2 fMRI Studies
To elucidate the mechanism of muscle relaxation, brain activity was assessed using
fMRI in several studies (Toma et al. 1999; Spraker et al. 2009). Some studies
demonstrated an active role for the brain in muscle relaxation. For example, Toma
et al. (1999) found that increases in the blood oxygenation level-dependent (BOLD)
signal in M1 were observed during muscle relaxation as well as during muscle
contraction. Areas outside M1, such as the supplementary motor area (SMA) and
pre-supplementary motor area (pre-SMA), were also activated during muscle
relaxation. Toma et al. (1999) compared the BOLD signals in the SMA and the
pre-SMA and found that the activation volume was greater for muscle relaxation
than for muscle contraction (Toma et al. 1999). Since the SMA is directly
connected to M1 (Dum and Strick 1992), the inhibitory command might be derived
from the SMA. Moreover, Spraker et al. (2009) demonstrated that the right dorso-
lateral prefrontal cortex (DLPFC) had a greater activity during muscle relaxation of
the right hand as compared with muscle contraction in this hand. Previously, it was
demonstrated that BOLD activation in the DLPFC increased during a Go/No-go
task (Garavan et al. 1999; Nakata et al. 2008). The Go/No-go task has been widely
utilized to investigate the inhibitory processes of motor control (Waldvogel
et al. 2000; Nakata et al. 2006). In the Go/No-go task, subjects respond to one
cue (the Go stimulus), and they are required to not respond to another cue (the
No-go stimulus). It is possible that a population of prefrontal cortical neurons on the
ipsilateral side may be associated with an inhibitory mechanism during the No-go
task involved muscle relaxation.
7.3.3 EEG Studies
Movement-related cortical potentials (MRCP) during muscle relaxation have

enabled a comparison between the cortical mechanisms of muscle contraction
and relaxation (Terada et al. 1995; Rothwell et al. 1998; Yazawa et al. 1999).
Readiness potentials, termed Bereitschaftspotential (BP), occur in the SMA (Terada
et al. 1995, 1999; Yazawa et al. 1998), pre-SMA (Yazawa et al. 1998) and M1
before voluntary muscle relaxation. The BP that is generated in M1 and which is
associated with isotonic muscle relaxation tasks using gravity (passive flexion) has
been shown to be remarkably similar to that associated with muscle contraction
(Terada et al. 1995; Yazawa et al. 1999). However, Rothwell et al. (1998) suggest
that the BP related to M1 is reduced in amplitude for muscle relaxation for
non-movement (isometric) tasks as compared to those involved with contraction.
Finally, Pope et al. (2007) directly compared muscle contraction and relaxation for
both isotonic and isometric tasks and verified that the BP in M1 is greater for
muscle contraction than relaxation during isometric tasks, but is similar for isotonic
tasks. Therefore, they suggest that isotonic tasks require an active process for
relaxation in the motor cortex, as compared with isometric tasks (Rothwell

et al. 1998; Pope et al. 2007). Pope et al. also noted that the greater BP seen with
isotonic relaxation tasks is related to afferent proprioceptive feedback from the
muscle spindle. Thus, the neural mechanism of muscle relaxation may change
depending upon whether an isotonic task or isometric task is involved.
7.4 Effect of Muscle Relaxation on the Remote Parts

of the Body
Many studies have investigated how muscle contraction influences neural mecha-
nisms of the other limb (remote effect; Baldissera et al. 2002; Tazoe et al. 2007).
Because muscle relaxation is an “active process” requiring cortical activation,
muscle relaxation as well as contraction might have remote effects However, to
date the mechanisms of muscle relaxation that pertain during multi-limb, complex
motions have been only minimally examined. Recently, Kato et al. (2014) demon-
strated that muscle relaxation in one limb suppressed muscle activity in the other
ipsilateral limb (Fig. 7.2). In the experiment, the participants were instructed to
execute simultaneous relaxation and contraction in the ipsilateral hand and foot.
Although the subjects tried to separately relax and contract their hand and foot, the
EMG activity of contracted muscle in one limb became weakened when it was
executed with relaxation in the other limb as compared with when the contraction
was made alone. Therefore, muscle relaxation in one limb seems to influence
muscle activity of the other (ipsilateral) limb.
7.5 Muscle Relaxation and Sports
Not only relaxation as the cessation of contraction, but also relaxation as a state in
which there is no unnecessary contraction are important to consider for sports
performance. Among the sports players and musicians, it is generally acknowl-
edged that proper muscle relaxation is absolutely necessary in the selective utili-
zation of multiple body parts when smooth movements are required. When we try to
improve a multi-limb skill that is necessary for a particular sport, we are often
frustrated with the unwanted contraction of muscles which are inappropriate for a
desired action. Indeed, novice sports and music players often show unintended
over-contraction of relatively superfluous muscles and insufficient contraction of
necessary muscles (Lay et al. 2002; Sakurai and Ohtsuki 2000; Fujii et al. 2009;
Yoshie et al. 2009).
Unskilled badminton players showed continuous, unnecessary contraction of the
triceps brachii when they swang a racket, whereas skilled players showed a minimal
amount of unnecessary contraction (Sakurai and Ohtsuki 2000). Furthermore, a
20°
wrist angle
Hand Contraction
50µV
extensor
EMG
−100 0 100 200 300 400 500
wrist angle 20°

With Foot Relaxation
50µV
Hand Contraction
extensor
EMG
−100 0 100 200 300 400 500

ankle angle
dorsiflexor
EMG
−100 0 100 200 300 400 500
(ms)
Fig. 7.2 A typical example of EMG activity (hand extensor and foot dorsiflexor) and joint angle
(wrist and ankle) for a hand contraction task (top) and hand contraction with foot relaxation task
(bottom) in a participant (Kato et al. 2014). Time 0 denotes the beep. The grey arrows indicate the
prolonged reaction time and suppressed EMG activity
decrease in unnecessary contraction of the triceps brachii was observed following

6 days of training of the unskilled players. Fujii et al. (2009) recorded activities in
agonist and antagonist muscles during tapping movements on a drum. In novice
drummers, a relatively large amount of activity in the antagonist muscles was
observed during contraction of the agonist muscles in the forearm. On the other
hand, expert drummers were able to suppress activity in the antagonist muscles
(i.e. relaxation). The suppression was even more pronounced in the world’s fastest
drummer (Fujii et al. 2009). As for pianists, it was observed that execution of
muscle relaxation is difficult in situations that produce anxiety or tension (Yoshie
et al. 2009). For example, muscle activity in the biceps brachii and upper trapezius
while playing the piano in a competition increased compared to what was recorded
during a rehearsal. Moreover, strong co-contraction of agonist and antagonist
muscles was observed in a concert (Yoshie et al. 2009). Strong co-contraction of
antagonistic muscles deteriorates physiological efficiency and thus produces mus-
cle fatigue (Lay et al. 2002). Not surprisingly, Yoshie et al. also reported that
performance quality, relative to the rehearsal, decreased in the concert.
7.6 Future Studies on Muscle Relaxation and Sports
In summary, muscle relaxation requires a characteristic brain activation pattern just

as does muscle contraction. In the applied field of body movements, moreover,
although muscle relaxation is clearly essential for skilled movements, it is difficult
for novice players to execute relaxation of the appropriate muscles, especially in
situations that make them nervous. In most situations, skilled-players are able to
simultaneously use the required muscles with an appropriate force level along with
simultaneous relaxation (no or minimal contraction) of unnecessary muscles. On
the other hand, when a beginner or unskilled-player tries to perform a complex
motor skill, they are often frustrated by the contraction of muscles in body parts that
should remain relaxed during the desired action. On such occasions, coaches are
often heard to say “Relax more!” to the player. However, to simply relax the
unintentionally contracted muscles is not so easy as coaches think. First, relaxation
of one particular muscle during contraction of another muscle is hard to execute
(e.g. badminton, drum and piano players; see previous chapter). Muscle contraction
in one limb tends to induce activation of circuits controlling other limbs (Baldissera
et al. 2002; Tazoe et al. 2007). Second, as noted above, when we focus on an
inappropriate contraction and try to suppress it, the desired muscle contraction is
likely to also be suppressed (Kato et al. 2014). This inhibitory mechanism may be
associated with the simultaneous relaxation and contraction seen during active
sports.
In previous studies on muscle relaxation, only mechanisms involved with mus-
cle relaxation of a particular target muscle were analyzed. In the future, it will be
important to clarify the reason why proper muscle relaxation, in the sports field, is
so difficult. This problem needs to be investigated from many different
perspectives.
References
Baldissera F, Borroni P, Cavallari P, Cerri G (2002) Excitability changes in human corticospinal

projections to forearm muscles during voluntary movement of ipsilateral foot. J Physiol
539:903–911
Begum T, Mima T, Oga T, Hara H, Satow T, Ikeda A, Nagamine T, Fukuyama H, Shibasaki H

(2005) Cortical mechanisms of unilateral voluntary motor inhibition in humans. Neurosci Res
53:428–435
Berardelli A, Hallett M, Rothwell JC, Agostino R, Manfredi M, Thompson PD, Marsden CD
(1996) Single-joint rapid arm movements in normal participants and in patients with motor
disorders. Brain 119:661–674
Berardelli A, Rothwell JC, Hallett M, Thompson PD, Manfredi M, Marsden CD (1998) The
pathophysiology of primary dystonia. Review. Brain 121:1195–1212
Buccolieri A, Avanzino L, Marinelli L, Trompetto C, Marchese R, Abbruzzese G (2004a) Muscle
relaxation is impaired in dystonia: a reaction time study. Mov Disord 19:681–687
Buccolieri A, Abbruzzese G, Rothwell JC (2004b) Relaxation from a voluntary contraction is
preceded by increased excitability of motor cortical inhibitory circuits. J Physiol 558:685–695
Chen R, Wassermann EM, Ca~ nos M, Hallett M (1997) Impaired inhibition in writer’s cramp
during voluntary muscle activation. Neurology 49:1054–1059
Cruz EG, Waldinger HC, Kamper DG (2005) Kinetic and kinematic workspaces of the index
finger following stroke. Brain 128:1112–1121
Dum RP, Strick PL (1992) Medial wall motor areas and skeletomotor control. Curr Opin
Neurobiol 2:836–839
Fahn S (1988) Concept and classification of dystonia. Adv Neurol 50:1–8
Fujii S, Kudo K, Ohtsuki T, Oda S (2009) Tapping performance and underlying wrist muscle
activity of non-drummers, drummers, and the world’s fastest drummer. Neurosci Lett
459:769–773
Garavan H, Ross TJ, Stein EA (1999) Right hemispheric dominance of inhibitory control: an
event-related functional MRI study. Proc Natl Acad Sci U S A 96:8301–8306
Grasso M, Mazzini L, Schieppati M (1996) Muscle relaxation in Parkinson’s disease: a reaction
time study. Mov Disord 11:411–420
Gray CS, French JM, Bates D, Cartlidge NE, James OF, Venables G (1990) Motor recovery
following acute stroke. Age Ageing 19:179–184
Inzelberg R, Flash T, Schechtman E, Korczyn AD (1995) Kinematic properties of upper limb
trajectories in idiopathic torsion dystonia. J Neurol Neurosurg Psychiatry 58:312–319
Jordan N, Sagar HJ, Cooper JA (1992) A component analysis of the generation and release of
isometric force in Parkinson’s disease. J Neurol Neurosurg Psychiatry 55:572–576
Kamper DG, Rymer WZ (2001) Impairment of voluntary control of finger motion following
stroke: role of inappropriate muscle coactivation. Muscle Nerve 24:673–681
Kamper DG, Harvey RL, Suresh S, Rymer WZ (2003) Relative contributions of neural mecha-
nisms versus muscle mechanics in promoting finger extension deficits following stroke. Muscle
Nerve 28:309–318
Kato K, Muraoka T, Higuchi T, Mizuguchi N, Kanosue K (2014) Interaction between simulta-
neous contraction and relaxation in different limbs. Exp Brain Res 232:181–189
Kujirai T, Caramia MD, Rothwell JC, Day BL, Thompson PD, Ferbert A, Wroe S, Asselman P,
Marsden CD (1993) Corticocortical inhibition in human motor cortex. J Physiol 471:501–519
Lay BS, Sparrow WA, Hughes KM, O’Dwyer NJ (2002) Practice effects on coordination and
control, metabolic energy expenditure, and muscle activation. Hum Mov Sci 21:807–830
Lelli S, Panizza M, Hallett M (1991) Spinal cord inhibitory mechanisms in Parkinson’s disease.
Neurology 41:553–556
Lorenzano C, Priori MDA, Curra A, Gilio F, Manfredi M, Berardelli A (2000) Impaired EMG
inhibition elicited by tendon stimulation in dystonia. Neurology 55:1789–1793
Motawar B, Hur P, Stinear J, Seo NJ (2012) Contribution of intracortical inhibition in voluntary
muscle relaxation. Exp Brain Res 221:299–308
Nakata H, Inui K, Wasaka T, Tamura Y, Akatsuka K, Kida T, Kakigi R (2006) Higher anticipated
force required a stronger inhibitory process in go/nogo tasks. Clin Neurophysiol 117:1669–
1676
Nakata H, Sakamoto K, Ferretti A, Gianni Perrucci M, Del Gratta C, Kakigi R, Luca Romani G
(2008) Somato-motor inhibitory processing in humans: an event-related functional MRI study.
Nakayama H, Jorgensen HS, Raaschou HO, Olsen TS (1994) The influence of age on stroke
outcome. Stroke 25:808–813
Neely KA, Planetta PJ, Prodoehl J, Corcos DM, Comella CL, Goetz CG, Shannon KL,
Vaillancourt DE (2013) Force control deficits in individuals with Parkinson’s disease, multiple
systems atrophy, and progressive supranuclear palsy. PLoS One 8, e58403
Nowak DA, Hermsd€orfer J, Topka H (2003) Deficits of predictive grip force control during object
manipulation in acute stroke. J Neurol 250:850–860
Nowak DA, Grefkes C, Dafotakis M, Küst J, Karbe H, Fink GR (2007) Dexterity is impaired at
both hands following unilateral subcortical middle cerebral artery stroke. Eur J Neurosci
25:3173–3184
Parker VM, Wade DT, Langton Hewer R (1986) Loss of arm function after stroke: measurement,
frequency, and recovery. Int Rehabil Med 8:69–73
Pope PA, Holton A, Hassan S, Kourtis D, Praamstra P (2007) Cortical control of muscle
relaxation: a lateralized readiness potential (LRP) investigation. Clin Neurophysiol
118:1044–1052
Priori A, Berardelli A, Mercuri B, Manfredi M (1995) Physiological effects produced by botuli-
num toxin treatment of upper limb dystonia. Changes in reciprocal inhibition between forearm
muscles. Brain 118:801–807
Reynolds C, Ashby P (1999) Inhibition in the human motor cortex is reduced just before a
voluntary contraction. Neurology 53:730–735
Ridding MC, Sheean G, Rothwell JC, Inzelberg R, Kujirai T (1995) Changes in the balance
between motor cortical excitation and inhibition in focal, task specific dystonia. J Neurol
Neurosurg Psychiatry 59:493–498
Rothwell JC, Higuchi K, Obeso JA (1998) The offset cortical potential: an electrical correlate of
movement inhibition in man. Mov Disord 13:330–335
Sakurai S, Ohtsuki T (2000) Muscle activity and accuracy of performance of the smash stroke in
badminton with reference to skill and practice. J Sports Sci 18:901–914
Schieppati M, Crenna P (1984) From activity to rest: gating of excitatory autogenetic afferences
from the relaxing muscle in man. Exp Brain Res 56:448–457
Schieppati M, Nardone A, Musazzi M (1986) Modulation of the Hoffmann reflex by rapid muscle
contraction or release. Hum Neurobiol 5:59–66
Seo NJ, Rymer WZ, Kamper DG (2009) Delays in grip initiation and termination in persons with
stroke: effects of arm support and active muscle stretch exercise. J Neurophysiol
101:3108–3115
Shibasaki H, Hallett M (2006) What is the Bereitschaftspotential? Clin Neurophysiol
117:2341–2356
Spraker MB, Corcos DM, Vaillancourt DE (2009) Vaillancourt, cortical and subcortical mecha-
nisms for precisely controlled force generation and force relaxation. Cereb Cortex
19:2640–2650
Tazoe T, Endoh T, Nakajima T, Sakamoto M, Komiyama T (2007) Disinhibition of upper limb
motor area by voluntary contraction of the lower limb muscle. Exp Brain Res 177:419–430
Terada K, Ikeda A, Nagamine T, Shibasaki H (1995) Movement-related cortical potentials
associated with voluntary relaxation. Electroencephalogr Clin Neurophysiol 95:333–345
Terada K, Ikeda A, Yazawa S, Nagamine T, Shibasaki H (1999) Movement-related cortical
potentials associated with voluntary relaxation of foot muscles. Clin Neurophysiol
110:397–403
Toma K, Honda M, Hanakawa T, Okada T, Fukuyama H, Ikeda A, Nishizawa S, Konishi J,
Shibasaki H (1999) Activities of the primary and supplementary motor areas increase in
preparation and execution of voluntary muscle relaxation: an event-related fMRI study. J
Neurosci 19:3527–3534
Toma K, Nagamine T, Yazawa S, Terada K, Ikeda A, Honda M, Oga T, Shibasaki H (2000)

Desynchronization and synchronization of central 20 Hz rhythms associated with voluntary
muscle relaxation: a magnetoencephalographic study. Exp Brain Res 134:417–425
van der Kamp W, Berardelli A, Rothwell JC, Thompson PD, Day BL, Marsden CD (1989) Rapid
elbow movements in patients with torsion dystonia. J Neurol Neurosurg Psychiatry
52:1043–1049
Waldvogel D, van Gelderen P, Muellbacher W, Ziemann U, Immisch I, Hallett M (2000) The
relative metabolic demand of inhibition and excitation. Nature 406:995–998
Wing AM (1988) A comparison of the rate of pinch grip force increases and decreases in
parkinsonian bradykinesia. Neuropsychologia 26:479–482
Yazawa S, Ikeda A, Kunieda T, Mima T, Nagamine T, Ohara S, Terada K, Taki W, Kimura J,
Shibasaki H (1998) Human supplementary motor area is active in preparation for both
voluntary muscle relaxation and contraction: subdural recording of Bereitschaftspotential.
Neurosci Lett 244:145–148
Yazawa S, Ikeda A, Kaji R, Terada K, Nagamine T, Toma K, Kubori T, Kimura J, Shibasaki H
(1999) Abnormal cortical processing of voluntary muscle relaxation in patients with focal hand
dystonia studied by movement-related potentials. Brain 122:1357–1366
Yoshie M, Kudo K, Murakoshi T, Ohtsuki T (2009) Music performance anxiety in skilled pianists:
effects of social-evaluative performance situation on participative, autonomic, and electro-
myographic reactions. Exp Brain Res 199:117–126
Chapter 8
Neural Mechanisms of Muscle Cramp
Kento Nakagawa, Naokazu Miyamoto, and Kazuyuki Kanosue
Abstract Although muscle cramps are a common problem for many athletes, the
underlying mechanisms are still poorly understood. In this chapter, we review
the proposed causes of muscle cramps. Dehydration and electrolyte disturbance
are generally considered to be causes of muscle cramps, but this is unlikely.
Alternatively, either neural activity in the spinal cord or in the peripheral could
be the cause of cramps. Definitive evidence is scarce and controversy abounds.
Keywords Central nervous system • Peripheral • Dehydration • Electrolyte

disturbance • Muscle fatigue
8.1 Introduction
A muscle cramp is defined as a sudden, involuntary, spasmodic and painful

contraction of skeletal muscle (Schwellnus et al. 1997; Minetto et al. 2013). It
often occurs during or after various sports (e.g. triathlon (Sulzer et al. 2005) and
marathons (Schwellnus 2007)), and it is a well known clinical problem that athletes
try hard to avoid. However, the mechanisms underlying muscle cramps are still
poorly understood. Therefore, there is no evidence-based way to efficiently prevent
muscle cramps although it is widely known that stretching a “cramping” muscle can
effectively stop the cramp (Norris et al. 1957; Bertolasi et al. 1993). In this chapter
we evaluate evidence for the various theories that purport to explain the cause of
muscle cramps. We concentrate on mechanisms involving dehydration and elec-
trolyte disturbances as well as neural factors.
K. Nakagawa (*)
e-mail: nakagawa@aoni.waseda.jp
N. Miyamoto
National Institute of Fitness and Sports in Kanoya, Kagoshima, Japan
K. Kanosue

80 K. Nakagawa et al.
8.2 How to Induce Muscle Cramps?
A major problem in the research of muscle cramps is that it is difficult to induce

muscle cramps experimentally, because they normally occur unpredictably and
spontaneously. Anecdotal studies have compared people who got cramps with
those who did not in sports events such as long distance races (Maughan 1986;
Schwellnus 2007) and triathlons (Sulzer et al. 2005). Although the merit of this
method would be the availability of a large number of subjects, it is almost
impossible to control the occurrence, the intensity of the cramps and the chemical
or fatigue state of the cramping muscle.
In the laboratory, researchers have predominantly utilized two methods to
induce muscle cramps: (1) strong voluntary contraction of target muscles, and
(2) electric/magnetic stimulation of nerves or muscles. The former is the classical
approach and is widely available since it does not required a special apparatus to
induce cramps (Khan and Burne 2007; Nakagawa et al. 2013; Norris et al. 1957;
Roeleveld et al. 2000; Ross and Thomas 1995; Ohno and Nosaka 2004; Goodman
and Zwetsloot 2013). Although maximal contraction is required of subjects in most
studies, the target muscle differs across studies. The electrical stimulation approach
has also been utilized widely. The site of stimulation varies depending on the study.
High-frequency electrical stimulation of the tibial nerve innervating calf muscles or
of intrinsic muscles of foot has often been used to induce muscle cramps (Bertolasi
et al. 1993; Lambert 1968; Obi et al. 1993). This method enables the investigator to
quantify the stimulation threshold for the induction of muscle cramps as “threshold
frequency” with high reliability (Miller and Knight 2007; Stone et al. 2003), which
is defined as the minimum frequency of simulation that can provoke muscle cramps
(Miller and Knight 2007, 2009; Stone et al. 2003, 2010). The stimulation intensity
in these studies is generally very high (e.g. 30 % higher intensity than that needed to
evoke a maximal M response at the frequencies of about 10 Hz (Bertolasi
et al. 1993)). In addition, since it has been shown that the threshold frequency
correlates with the incidence of muscle cramps (Miller and Knight 2009), it has
been proposed that the index can be used to evaluate individuals’ risk of cramp.
However, although applying electrical stimulation to a peripheral nerve is a useful
way to maintain constant conditions, it is generally accompanied by strong pain.
Therefore, electrical stimulation is often applied directly to the muscle motor point.
This produces less pain, and has been developed as an alternative to peripheral
nerve stimulation and has been utilized (Minetto and Botter 2009; Minetto
et al. 2008, 2009a, b, 2011). A study with the motor point stimulation demonstrated
that the threshold frequency and pain in intrinsic foot muscle were lower than those
in leg muscle. The authors concluded that the intrinsic foot muscle is the most
suitable for the study of muscle cramps (Minetto and Botter 2009). It has been
found that the high frequency magnetic stimulation to peripheral nerve can provoke
muscle cramps, which are less painful than those produced by electrical stimulation
(Caress et al. 2000). Moreover, other methods based on activating the afferent
nerves have been performed. For example, applying tapping/vibration to the tendon
8 Neural Mechanisms of Muscle Cramp 81
with an electromagnetic hammer, electrical stimulation to the Ia afferent

(Baldissera et al. 1994), or nociceptive stimulation to myofascial trigger points
(Ge et al. 2008) also evoke muscle cramps.
8.3 Do Dehydration and Electrolyte Disturbance Cause

Muscle Cramps?
Dehydration and electrolyte disturbances are popularly thought to cause muscle

cramps, and a large number of studies have evaluated this hypothesis. There is an
early anecdotal observation that cramps often occur in workers toiling in hot and
humid conditions (Talbott and Michelsen 1933). It is typically held that depletion of
sodium with significant dehydration by sweating induces the contraction of the
extracellular fluid compartment which results in a disruption of ion balance. A loss
of interstitial volume can cause a mechanical deformation of nerve endings, and
increase the surrounding neurotransmitter concentration (Bergeron 2003, 2008;
Schwellnus 2009), which could produce muscle cramps. However, there is no
scientific evidence to support these explanations. Studies supporting the above
hypothesis relied heavily on anecdotal reports and involved few properly controlled
experiments (Ohno and Nosaka 2004). Furthermore, the hypothesis cannot explain
why muscle cramps also often occur under cool conditions (Miller et al. 2010;
Armstrong et al. 2007).
Indeed, many studies show no association between muscle cramps and dehy-
dration or electrolyte disturbance. Maughan (1986) investigated biochemical indi-
ces of runners with and without cramps in a full marathon race. Serum electrolyte
concentrations such as sodium and potassium were not different between the two
groups after the race. A similar study on runners in an ultra-marathon race (56 km)
produced the same results (Schwellnus et al. 2004). Other studies also found no
association between serum electrolytes and muscle cramp in a triathlon race (Sulzer
et al. 2005; Schwellnus et al. 2011). That is, there was no significant difference in
serum electrolyte concentrations and hemoglobin concentration between cramping
and non-cramping groups after the race. In addition to these studies in competition
races, a laboratory experiment indicated that muscle cramp occurred both with and
without dehydration and electrolyte loss with fatiguing exercise (Jung et al. 2005).
Braulick et al. (2013) indicated that the threshold frequency for inducing a muscle
cramp with the high frequency electrical stimulation to peripheral nerve did not
differ significantly between the euhydrated and hypohydrated conditions. Addition-
ally, it is also unlikely that magnesium is effective for the treatment of muscle
cramp (Garrison et al. 2012). The above evidence makes it abundantly clear that
muscle cramps are not, or are only very minimally, associated with dehydration and
electrolyte disturbance.
8.4 Muscle Fatigue and Cramps
Alternatively, the theory that “muscle fatigue” can cause muscle cramps has been
raised (Schwellnus 2009; Schwellnus et al. 1997). Muscle cramps often occur in
events, such as long distance races that would be expected to induce muscle fatigue
(Maughan 1986; Schwellnus et al. 2004; Sulzer et al. 2005). Thus, it is likely that
the cause of muscle cramp is not dehydration or electrolyte disturbance but muscle
fatigue. Indeed, Jung et al. (2005) demonstrated that supplementation of carbohy-
drate hindered the occurrence of muscle cramps during fatiguing exercise, perhaps
owing to energy from the carbohydrate that delayed the occurrence of muscle
fatigue. One view holds that neuromuscular control is altered by muscle fatigue.
This could lead to an increase in the activity of muscle spindles and a decrease in
the activity of Golgi tendon organs, which together would result in sustained motor
neuron activity, and subsequently muscle cramps (Schwellnus 2009). However, the
causal relationship has never been validated. Additionally, it has been reported that
the threshold frequency for inducing cramps is higher under fatigue than under
non-fatigue conditions (Stone et al. 2010). This seems to contradict the above
proposed fatigue theory. Moreover, the fact that cramps can occur without fatigue
in most laboratory experiments also suggest that muscle fatigue is not a necessary
condition for the occurrence of cramps. Most importantly, studies proposing a
muscle fatigue mechanism have not quantified “muscle fatigue”. The consensus
has now shifted, and it is generally considered that “altered neuromuscular control”
triggered by muscle fatigue is an important cause of muscle cramps rather than
simply muscle fatigue itself (Armstrong and Cross 2013; Schwellnus 2009;
Schwellnus et al. 1997). In succeeding sections, we review investigations the neural
mechanisms involved with muscle cramps.
8.5 Neural Mechanisms of Muscle Cramp
There have been many debates on the nature of the neural origins of muscle cramps
(Miller and Layzer 2005; Minetto et al. 2013; Layzer 1994; Jansen et al. 1990).
There are mainly two theories (Fig. 8.1). The first theory is the “central origin”
theory, which holds that muscle cramps originate from the central nervous system
(CNS), particularly the spinal cord. The second theory is the “peripheral origin”
theory, which holds that muscle cramps are the result of events in the peripheral
level, such as motor nerve fibers, nerve branch terminals or muscle fibers.
spinal cord
Ib afferent neuron
Ia afferent neuron
muscle
α-motor neuron
peripheral origin central origin

Fig. 8.1 Scheme depicting the two major mechanisms purported to underlie the development of
muscle cramps
8.5.1 Central Origin Theory
Norris et al. (1957) proposed the central origin theory after performing experiments
involving electromyography (EMG). Their study showed that synchronized dis-
charges between different motor units were observed in cramping muscle. More-
over, voluntary contraction of the homologous muscle in the contralateral limb
increased cramp discharge, while voluntary contraction of the ipsilateral antagonist
muscle reduced cramp discharge. The authors concluded that the neural activity
during muscle cramp originates from the spinal cord. Although this observation was
anecdotal, the following studies have observed similar phenomenon in laboratory
experiments.
Several studies have recorded motor unit activities during cramps (Minetto
et al. 2009b, 2011; Norris et al. 1957; Ross and Thomas 1995). Shapes of the
motor unit potentials (Ross and Thomas 1995; Minetto et al. 2011), and firing rates
(Minetto et al. 2009b) during muscle cramps resemble those seen in normal
voluntary contractions. However, the variability of motor unit firing rates was
larger during cramp than during voluntary contractions (Minetto et al. 2009b,
2011; Ross and Thomas 1995). The high variability of motor unit discharge is
speculated to be caused because the afferent inputs to the motor neurons generate
synaptic noise (Merletti et al. 2011; Minetto et al. 2009b). However, since vari-
ability of motor unit discharge during cramp without afferent contribution by
peripheral nerve block is greater than that during normal cramps (Minetto
et al. 2011), the above mentioned explanation seems precarious. Thus more sub-
stantial evidence is required to implicate spinal involvement in cramp production.
Other studies have produced more direct evidence for spinal involvement in
cramp production by investigating the spinal reflex (e.g. Mills et al. 1982;
Baldissera et al. 1994; Khan and Burne 2007). In a case report Mills et al. (1982)
noted that transcutaneous nerve stimulation over the calf muscle (by which afferent
signals affect spinal neural activity) could abort cramps in a patient who suffered
from severe muscle cramps. From this result they inferred that the mechanism of
muscle cramps must involve the spinal cord. However, since only one patient was
involved, the results cannot be generalized. Alternatively, recent studies report that
nociceptive stimulation by a bolus injection of glutamate into latent myofacial
trigger points induces cramps (Ge et al. 2008). In addition, increased nociceptive
muscle afferent activity induced by injection of hypertonic saline decreased the
threshold frequency of electrically elicited muscle cramps (Serrao et al. 2007).
However, since these techniques may affect not only afferent activity but also the
muscle itself, it does not conclusively indicate CNS involvement. Other studies
have clarified the relationship between the kinesthetic afferent input and muscle
cramps. Baldissera et al. (1994) showed that electrical stimulation to the Ia afferent
nerve with low intensity triggered a cramp in the soleus muscle in three subjects
who suffered from muscle cramps. In addition, the finding that taps and continuous
vibration to the Achilles tendon also could induce the cramp (Baldissera et al. 1994)
indicates that the afferent signals may also influence cramps. Moreover, Khan and
Burne (2007) observed that EMG activity during cramp of calf muscle was
inhibited by the electrical stimulation applied to the Achilles tendon. The intensity
and timing of the disappearance of the EMG activity were similar to those of
voluntary contractions seen in the same muscle at similar background EMG levels.
Thus the authors suggested that the same reflex pathway was involved in the
inhibition of both voluntary contractions and muscle cramps. Additionally, Ross
and Thomas (1995) indicated that the tonic vibration reflex was depressed or absent
after muscle cramps, while it never changed after voluntary contractions. This
suggests that muscle cramps may inhibit the function of the spinal reflex. In
contrast, prolonged enhancement of the H-reflex after cramping was observed,
but it was not observed after voluntary contractions (Ross 1976). Because the
H-reflex reflects the excitability of the motor neuron pool in the spinal cord, these
two studies seems to be in conflict; one suggests inhibition and the other suggests
the enhancement of the spinal neural activity. However, the fact that a muscle
cramp has prolonged overall effects on neural excitability of the spinal cord is very
likely.
Although many studies have proposed a central origin theory, the proposed
mechanisms purporting to explain muscle cramps are different. Ideas put forth
includes hyperexcitability of motor neurons with presynaptic inputs produced by a
positive feedback loop between afferent nerves and motor neurons (Ross and
Thomas 1995), bistability of the motor neuron membrane (Baldissera et al. 1994),
and dysfunction of interneurons via GABA (Obi et al. 1993). When looked at
closely, these ideas must be considered suppositions, since they are not supported
by concrete evidences.
8.5.2 Peripheral Origin Theory
The first study that supported the peripheral origin theory utilized anesthesia to
produce a peripheral nerve block (Lambert 1968). The nerve block shut off the
contribution of supraspinal neural activity. Thus, efferent signals from the motor
neuron pool could not reach the muscles, and afferent signals from sensory recep-
tors were also unable to reach the spinal cord. It was thus possible to investigate the
role of the nerves distal to the blocked point. Lambert (1968) observed that, in
healthy subjects, repetitive electrical stimulation of the peripheral nerve distal to the
block could induce a muscle cramp. This result strongly suggested that muscle
cramps could occur without a supraspinal contribution, and cramps likely originate
in the periphery, probably in the intramuscular nerve terminal. In later work,
Bertolasi et al. (1993) replicated the induction of muscle cramps during a peripheral
nerve block. In addition, they indicated that no muscle cramp was induced without a
shortening of the muscle even when electrical stimulation was delivered, and also
found that stretching to the muscle could interrupt cramps even after the nerve
block in normal subjects. These data suggest that muscle length strongly influences
muscle cramps, and that muscle cramps probably originate from the periphery, in
particular intramuscular branches, rather than from the CNS. It seems that the
peripheral origin theory was mainly based on the two above-mentioned studies.
Although the two studies (Bertolasi et al. 1993; Lambert 1968) appear to provide
concrete evidences, other some studies have provided differing results. In 1993 Obi
et al. showed that high-frequency electrical stimulation to the peripheral nerve
distal to the blocked portion did not induce a muscle cramp. They additionally
indicated that diazepam or baclofen, a GABA receptor agonist, prevented the
induction of cramps by electrical stimulation. They speculated that abnormal
activity of GABAergic interneurons in the spinal cord were involved in the mech-
anisms underlying cramps. However, this was an anecdotal report. There were no
controls, and the study involved only two patients, both of whom had a motor
neuron disease. A more convincing study was performed by Minetto et al. (2011),
who provided evidence against the peripheral origin theory with experiments
utilizing peripheral nerve block. They studied eight normal subjects, and were
able to induce muscle cramps by electrically stimulating the muscle motor point
with or without nerve block. They investigated the difference in characteristics of
surface EMGs and motor unit potentials between cramps electrically induced under
the two conditions. The results indicated that the threshold frequency of electrical
stimulation for eliciting cramps was greater in the blocked condition than in the
non-blocked condition. In addition, the duration and EMG amplitude of muscle
cramps in the blocked condition were noticeably smaller. Cramps in the blocked
condition showed a higher rate of motor unit discharge as well as irregular dis-
charge patterns. The authors concluded that the CNS is involved in both the origin
and sustenance of muscle cramps, rather than peripheral mechanisms.
In addition, Roeleveld et al. (2000) examined the detailed characteristics of
muscle activity by multi-channel surface EMG recordings on the triceps surae
during muscle cramps. They observed that involuntary EMG activity (i.e., muscle
cramps) induced by maximal voluntary contraction (MVC) initially occurred over a
small area and gradually spread over a larger region. Moreover, the strongly-
activated area moved from one area to another, and then the intensity and area of
the cramp decreased and disappeared. The results from these local EMG recordings
indicate that muscle cramp might originate from close to, or even at, the level of the
muscle fiber.
Recently, we investigated the involvement of peripheral mechanisms in muscle
cramps in nine healthy subjects who were able to volitionally evoke muscle cramps
of the abductor halluces (AH) by voluntary contraction (Nakagawa et al. 2013).
High-intensity electrical stimulation to the tibial nerve, which induced a maximal
M-wave (Mmax), was applied during the muscle cramp as well as during voluntary
contraction of the abductor halluces. This was done to evaluate peripheral involve-
ment, since the amplitude of Mmax is not affected by spinal activity. Subjects first
voluntarily elicited a maximal voluntary contraction (MVC) of the target muscle in
order to induce a cramp. Once the cramp occurred, the subjects were instructed to
cease the volitional input and remained relaxed until the cramp diminished natu-
rally. Throughout the trial, electrical stimulation to the tibial nerve was applied
every 3 s, and the evoked EMG activity recorded. The onset of the cramp was
defined as the moment when the subjects declared “I’m cramping” and the offset as
the moment when the EMG burst stopped. The results indicated that the amplitude
of Mmax decreased or disappeared during a muscle cramp, but not during voluntary
contraction task (Fig. 8.2). We suspect that the decrease of Mmax during MVC that
was seen before the occurrence of the cramp (as defined), happened during the time
when the cramp had actually begun but before conscious awareness of this occur-
rence was attained by the subject. A significant negative correlation between Mmax
amplitude and intensity of background EMG was observed during the muscle
cramp task. The larger the background EMG, the greater the decrease in the
amplitude of Mmax, although Mmax did not change during the voluntary contrac-
tion period (Fig. 8.3). Notably, the amplitude of Mmax in the soleus obtained
simultaneously with the M-wave in the AH was not changed during the AH muscle
cramp (Fig. 8.2). Overall, the results strongly suggest that the abnormal discharge
seen during the muscle cramp occurred distal to the site of stimulation. However,
this result does not directly exclude the possibility of spinal reflex involvement.
Problems and Foresight
As we noted, the origin of muscle cramps has been largely seen as either in the
spinal cord or the periphery. However, higher brain activity may also be involved in
the genesis of muscle cramps. For example, muscle cramps occur on occasion in top
ranking athletes even before the start of competition. Since such maladies often
happen in big matches with a large media attendance, it is generally held that in
these cases, mental stress triggers the muscle cramps. This makes good sense, and
thus brain activity must be included as an important influence on the induction of
Fig. 8.2 (a) Modulation of maximal M-waves during muscle cramp and (b) voluntary contrac-
tion. The abductor hallucis (AH) is the cramping muscle, and the soleus is the control muscle that
is not cramping. The onset of the cramp was defined as the moment when the subjects declared “I
am cramping” and the offset as the moment when the EMG burst stopped
muscle cramps, either directly or indirectly. There is little doubt that it would be
difficult indeed to explain such phenomena as emanating via with spinal or periph-
eral mechanisms. However, while logically possible, at the current time it would be
AH
Amplitude of Mmax (% control)
100 Voluntary contraction
Muscle cramp
80
60
y = -0.71x + 82.49
40
r2 = 0.34
20 p < 0.01
0
0 20 40 60 80 100
Background EMG (% MVC)
Fig. 8.3 Relationship between the amplitude of maximal M-waves (Mmax) and intensity of the
background electromyographic (EMG) activity during muscle cramps (red circles) and voluntary
contraction (blue diamonds) of the abductor hallucis (AH). A significant negative correlation was
found during muscle cramping
extremely difficult to establish the mechanism by which the brain contributes to

stress induced cramps.
The current difficulty of separating the contribution of CNS and peripheral
mechanisms may be due to the fact that both the CNS and the periphery play
roles in the initiation and prolongation of muscle cramps (Minetto et al. 2009a;
Merletti et al. 2011). Very likely, the locus responsible for the generation and
prolongation of muscle cramps will be found to differ in dissimilar situations.
Indeed, involuntary discharge can occur without a spinal contribution, whereas
discharge duration and intensity becomes larger with spinal involvement (Minetto
et al. 2011). That may indicate that peripheral mechanisms produce cramps while
spinal or higher brain mechanisms play a role in cramp prolongation. Future
research will be needed if we are to clarify the detailed mechanisms that produce
muscle cramps. Such an understanding will be critical for developing a strategy to
prevent or minimize the occurrence of muscle cramps and will be a very welcome,
overdue advancement for those who suffer from frequent muscle cramps.
References
Armstrong S, Cross T (2013) Exercise-associated muscle cramps. Med Today 14:62–65

Armstrong LE, Casa DJ, Millard-Stafford M, Moran DS, Pyne SW, Roberts WO (2007) Exertional
heat illness during training and competition. Med Sci Sports Exerc 39:556–572
Baldissera F, Cavallari P, Dworzak F (1994) Motor neuron ‘bistability’ a pathogenetic mechanism
for cramps and myokymia. Brain 117:929–939
Bergeron MF (2003) Heat cramps: fluid and electrolyte challenges during tennis in the heat. J Sci
Med Sport 6:19–27
Bergeron MF (2008) Muscle cramps during exercise-is it fatigue or electrolyte deficit? Curr Sports
Med Rep 7:S50–S55
Bertolasi L, de Grandis D, Bongiovanni LG, Zanette GP, Gasperini M (1993) The influence of
muscular lengthening on cramps. Ann Neurol 33:176–180
Braulick KW, Miller KC, Albrecht JM, Tucker JM, Deal JE (2013) Significant and serious
dehydration does not affect skeletal muscle cramp threshold frequency. Br J Sports Med
47:710–714
Caress JB, Bastings EP, Hammond GL, Walker FO (2000) A novel method of inducing muscle
cramps using repetitive magnetic stimulation. Muscle Nerve 23:126–128
Garrison SR, Allan GM, Sekhon RK, Musini VM, Khan KM (2012) Magnesium for skeletal
muscle cramps. Cochrane Database Syst Rev 9, CD009402
Ge HY, Zhang Y, Boudreau S, Yue SW, Arendt-Nielsen L (2008) Induction of muscle cramps by
nociceptive stimulation of latent myofascial trigger points. Exp Brain Res 187:623–629
Goodman A, Zwetsloot KA (2013) Voluntary inducement of triceps surae muscle cramping. Int J
Athlet Ther Train 18:40–43
Jansen PH, Joosten EM, Vingerhoets HM (1990) Muscle cramp: main theories as to aetiology. Eur
Arch Psychiatry Neurol Sci 239:337–342
Jung AP, Bishop PA, Al-Nawwas A, Dale RB (2005) Influence of hydration and electrolyte
supplementation on incidence and time to onset of exercise-associated muscle cramps. J Athl
Train 40:71–75
Khan SI, Burne JA (2007) Reflex inhibition of normal cramp following electrical stimulation of
the muscle tendon. J Neurophysiol 98:1102–1107
Lambert EH (1968) Electromyography in amyotrophic lateral sclerosis. In: Motor neuron diseases
on amyotrophic lateral sclerosis and related disorders. Grune and Stratton, New York, pp
135–153
Layzer RB (1994) The origin of muscle fasciculations and cramps. Muscle Nerve 17:1243–1249
Maughan RJ (1986) Exercise-induced muscle cramp: a prospective biochemical study in marathon
runners. J Sports Sci 4:31–34
Merletti R, Botter A, Lanfranco F, Minetto MA (2011) Spinal involvement and muscle cramps in
electrically elicited muscle contractions. Artif Organs 35:221–225
Miller KC, Knight KL (2007) Pain and soreness associated with a percutaneous electrical
stimulation muscle cramping protocol. Muscle Nerve 36:711–714
Miller KC, Knight KL (2009) Electrical stimulation cramp threshold frequency correlates well
with the occurrence of skeletal muscle cramps. Muscle Nerve 39:364–368
Miller TM, Layzer RB (2005) Muscle cramps. Muscle Nerve 32:431–442
Miller KC, Stone MS, Huxel KC, Edwards JE (2010) Exercise-associated muscle cramps: causes,
treatment, and prevention. Sports Health 2:279–283
Mills KR, Newham DJ, Edwads RH (1982) Severe muscle cramps relieved by transcutaneous
nerve stimulation: a case report. J Neurol Neurosur Psychiatry 45:539–542
Minetto MA, Botter A (2009) Elicitability of muscle cramps in different leg and foot muscles.
Muscle Nerve 40:535–544
Minetto MA, Botter A, Ravenni R, Merletti R, de Grandis D (2008) Reliability of a novel
neurostimulation method to study involuntary muscle phenomena. Muscle Nerve 37:90–100
Minetto MA, Botter A, de Grandis D, Merletti R (2009a) Time and frequency domain analysis of
surface myoelectric signals during electrically-elicited cramps. Neurophysiol Clin 39:15–25
Minetto MA, Holobar A, Botter A, Farina D (2009b) Discharge properties of motor units of the
abductor hallucis muscle during cramp contractions. J Neurophysiol 102:1890–1901
Minetto MA, Holobar A, Botter A, Ravenni R, Farina D (2011) Mechanisms of cramp contrac-
tions: peripheral or central generation? J Physiol 589:5759–5773
Minetto MA, Holobar A, Botter A, Farina D (2013) Origin and development of muscle cramps.
Exerc Sport Sci Rev 41:3–10
Nakagawa K, Miyamoto N, Murakami Y, Kanosue K (2013) Decrease in amplitude of M-wave

during muscle cramp. Tokyo J Phys Educ 4:11–16 (in Japanese)
Norris FH Jr, Gasteiger EL, Chatfield PO (1957) An electromyographic study of induced and
spontaneous muscle cramps. Electroencephalogr Clin Neurophysiol 9:139–147
Obi T, Mizoguchi K, Matsuoka H, Takatsu M, Nishimura Y (1993) Muscle cramp as the result of
impaired GABA function–an electrophysiological and pharmacological observation. Muscle
Nerve 16:1228–1231
Ohno M, Nosaka K (2004) Effect of muscle fatigue and dehydration on exercise induced muscle
cramp (EIMC). Jpn J Phys Fitness Sports Med 53:131–140 (in Japanese)
Roeleveld K, van Engelen BG, Stegeman DF (2000) Possible mechanisms of muscle cramp from
temporal and spatial surface EMG characteristics. J Appl Physiol 88:1698–1706
Ross BH (1976) Muscle cramp and the Hoffmann reflex. In: Proceedings of the 20th world
congress sports medicine, Carlton, pp 67–70
Ross BH, Thomas CK (1995) Human motor unit activity during induced muscle cramp. Brain
113:983–993
Schwellnus MP (2007) Muscle cramping in the marathon: aetiology and risk factors. Sports Med
37:364–367
Schwellnus MP (2009) Cause of exercise associated muscle cramps (EAMC)–altered neuromus-
cular control, dehydration or electrolyte depletion? Br J Sports Med 43:401–408
Schwellnus MP, Derman EW, Noakes TD (1997) Aetiology of skeletal muscle ‘cramps’ during
exercise: a novel hypothesis. J Sports Sci 15:277–285
Schwellnus MP, Nicol J, Laubscher R, Noakes TD (2004) Serum electrolyte concentrations and
hydration status are not associated with exercise associated muscle cramping (EAMC) in
distance runners. Br J Sports Med 38:488–492
Schwellnus MP, Drew N, Collins M (2011) Increased running speed and previous cramps rather
than dehydration or serum sodium changes predict exercise-associated muscle cramping: a
prospective cohort study in 210 Ironman triathletes. Br J Sports Med 45:650–656
Serrao M, Arendt-Nielsen L, Ge HY, Pierelli F, Sandrini G, Farina D (2007) Experimental muscle
pain decreases the frequency threshold of electrically elicited muscle cramps. Exp Brain Res
182:301–308
Stone MB, Edwards JE, Babington JP, Ingersoll CD, Palmieri RM (2003) Reliability of an
electrical method to induce muscle cramp. Muscle Nerve 27:122–123
Stone MB, Edwards JE, Huxel KC, Cordova ML, Ingersoll CD, Babington JP (2010) Threshold
frequency of an electrically induced cramp increases following a repeated, localized fatiguing
exercise. J Sports Sci 28:399–405
Sulzer NU, Schwellnus MP, Noakes TD (2005) Serum electrolytes in Ironman triathletes with
exercise-associated muscle cramping. Med Sci Sports Exerc 37:1081–1085
Talbott JH, Michelsen J (1933) Heat cramps. A clinical and chemical study. J Clin Invest
12:533–549
Chapter 9
Task Difficulty Affects the Association
Between Childhood Fitness and Cognitive
Flexibility
Keita Kamijo and Hiroaki Masaki
Abstract In this chapter we provide additional insight into the association between
aerobic fitness and cognitive flexibility in preadolescent children. We compared
lower-fit and higher-fit children’s task performance during the Trail Making Test
(TMT). The TMT consists of two parts. TMT-A requires participants to draw a line
connecting 25 randomly positioned numbers in numerical order, while TMT-B asks
them to draw a line connecting numbers and letters in an alternating order. That is,
the TMT-B necessitates a greater cognitive flexibility than the TMT-A. Analyses
revealed that higher-fit children exhibited superior task performance relative to
lower-fit children for the TMT-A, corroborating previous findings indicating the
general nature of the relationship between fitness and cognition in preadolescent
children. In contrast, task performance did not differ between the fitness groups for
the TMT-B. Thus, the positive association between aerobic fitness and cognitive
flexibility disappeared, likely because the cognitive task was too difficult for
preadolescent children. This result indicates that task difficulty affects the associ-
ation between fitness and cognitive control during childhood.
Keywords Children • Fitness • Cognitive flexibility • Cognitive control • Task

difficulty
9.1 Introduction
For more than a decade, a growing body of research has suggested that greater
physical activity and aerobic fitness are associated with superior cognitive function
across the lifespan. In studies of older adult populations, it has been well
documented that the positive association of fitness and cognitive function is selec-
tively and disproportionately greater for tasks or task components requiring exten-
sive amounts of cognitive control. Cognitive control refers to “the ability to
orchestrate thought and action in accord with internal goals” (Miller and Cohen
K. Kamijo (*) • H. Masaki

e-mail: k-kamijo@aoni.waseda.jp

92 K. Kamijo and H. Masaki
2001, p. 167). Three core proccesses of cognitive control have been suggested:
inhibition (i.e., the ability to ignore distraction and maintain focus), working
memory (i.e., the ability to temporarily retain and manipulate information), and
cognitive flexibility (i.e., the ability to alter goal-directed behavior according to
changes in environmental demands; Diamond 2006). It has been well established
that the prefrontal cortex, which exhibits disproportionately larger age-related
degradation (Raz et al. 1997) and protracted maturation (Gogtay et al. 2004),
plays an important role in cognitive control (Miller and Cohen 2001).
Kramer et al. (1999) indicated that aerobic fitness training selectively improved
performance on task conditions requiring greater amounts of cognitive control such
as inhibition and cognitive flexibility in older adults. Colcombe and Kramer (2003)
conducted a meta-analysis of randomized controlled trials that examined the effects
of aerobic fitness training on cognitive function in older adults. Although they
found general benefits of fitness training regardless of the type of cognitive task, the
improvements were disproportionately greater for task types requiring greater
amounts of cognitive control, supporting the above observation. Further, Hillman
et al. (2006b) examined the relationship among age, physical activity level, and
inhibition by comparing younger and older adults’ task performance during a
modified flanker task (Eriksen and Eriksen 1974) and corroborated the previous
findings (Colcombe and Kramer 2003; Kramer et al. 1999). The flanker task
requires participants to respond based on a centrally presented target and to ignore
flanking stimuli, which are task irrelevant. A spatial flanker task consists of
congruent trials, in which flanking arrows face in the same direction as the target
arrow (i.e., <<<<< or >>>>>), and incongruent trials, in which flanking arrows
face in the opposite direction from the target arrow (i.e., > > <> > or < <> < <).
Incongruent trials require greater amounts of cognitive control to inhibit the
interference caused by the flanking stimuli relative to congruent trials. Hillman
et al. (2006b) indicated that higher physical activity levels were associated with a
greater response accuracy across congruent and incongruent trials for older adults,
but not for younger adults. Further, they found that a reduced interference effect
(i.e., the difference in response accuracy between congruent and incongruent trails)
was associated with greater participation in physical activity in older adults,
confirming the selective nature of the relationship between fitness/physical activity
and cognitive function. Thus, in older adult populations, there is a consensus that
the association between fitness and cognitive function is selectively and dispropor-
tionately greater for higher-order cognitive functions (i.e., those involving cogni-
tive control).
Given that there is growing concern that children in industrialized nations have
become more sedentary and less fit (Department of Health and Human Services and
Education 2000; Tomkinson et al. 2012), recent research has begun to examine the
above described relationship during childhood. Thus Hillman et al. (2009) investi-
gated the relationship between aerobic fitness and inhibitory control using the
modified flanker task in preadolescent children. Their results indicated that
higher-fit children exhibited a greater response accuracy than their lower-fit peers
across congruent and incongruent trials. Buck et al. (2008) also indicated a general,
9 Task Difficulty Affects the Association Between Childhood Fitness and. . . 93
rather than selective, relationship between aerobic fitness and cognition in pread-
olescent children. They employed the Stroop task, which requires variable amounts
of cognitive control. The Stroop task consists of three conditions, with two control
conditions (word reading and color naming) and one incongruent condition in
which participants are asked to name the color of the ink that was mismatched to
the word (e.g., GREEN printed in blue ink). The incongruent condition requires
greater amounts of cognitive control to inhibit a prepotent but inappropriate
response (i.e., word reading) and instead activate a less frequent response (i.e.,
color naming). Buck et al. (2008) showed that greater aerobic fitness was related to
superior task performance across all conditions of the Stroop task.
In contrast, Pontifex et al. (2011) observed a selective relationship between
fitness and cognition in preadolescent children that was based on cognitive control
demands. They investigated the association between aerobic fitness and cognitive
flexibility through manipulation of stimulus – response compatibility during a
modified flanker task. The compatible condition asked the participants to press a
button consonant with the direction of the central target, as in the above studies
(Hillman et al. 2006b, 2009), whereas in the incompatible condition, participants
pressed a button that opposed the direction of the central target. Given that the
incompatible, relative to the compatible, condition necessitates greater flexibility in
the modulation of cognitive control (Friedman et al. 2009), this manipulation
allows for an assessment of cognitive flexibility. Pontifex et al. (2011) indicated
that higher-fit children exhibited comparable response accuracy between the com-
patibility conditions, whereas lower-fit children could not maintain their response
accuracy in the incompatible condition (i.e., lower response accuracy for the
incompatible relative to the compatible condition). These results suggest that
greater aerobic fitness is associated with superior cognitive flexibility during
preadolescent childhood and imply that the association is selectively and dispro-
portionately greater for higher-order cognitive functions even in preadolescent
children. As mentioned earlier, this association is also the case in older adults.
However, this appears to be inconsistent with the other developmental studies
indicating the general relationship between fitness and cognition in children
(Buck et al. 2008; Hillman et al. 2009). It is noteworthy that the positive association
between aerobic fitness and task performance (i.e., response accuracy) was
observed across congruent and incongruent trials in Pontifex et al. (2011), which
replicated the findings of Hillman et al. (2009). Thus, it is considered that Pontifex
et al. (2011) indicated both the selective and general nature of the relationship
between fitness and cognition during childhood.
One possible explanation of the inconsistent findings (i.e., general versus selec-
tive nature of the relationship) is that the relationship during childhood may differ
based on specific components within the overall rubric of cognitive control.
Although the core processes of cognitive control (i.e., inhibition, working memory,
and cognitive flexibility) are interrelated, these processes are distinct and distin-
guishable (Best and Miller 2010; Miyake et al. 2000). Based on the previous
findings, it is speculated that the relationship between aerobic fitness and cognitive
function during childhood is selectively and disproportionately greater for tasks
requiring greater demands upon cognitive flexibility, but not upon inhibitory
control demands.
Most studies investigating the relation of aerobic fitness to child cognition have
focused on inhibition. Yet, evidence regarding the association between childhood
fitness and cognitive flexibility has remained scarce. Thus, the present study was
designed to provide additional insight into the association between aerobic fitness
and cognitive flexibility during childhood by using the Trail Making Test (TMT).
This test is a traditional measure of cognitive flexibility, and consists of two parts, A
and B. TMT-A requires participants to draw a line connecting 25 randomly posi-
tioned numbers in numerical order (i.e., 1-2-3-. . .), while TMT-B asks them to draw
a line connecting numbers (e.g., 1–13) and letters (e.g., A-L) in an alternating order
(i.e., 1-A-2-B-3-. . .). That is, the TMT-B necessitates greater cognitive flexibility
than does TMT-A (Arbuthnott and Frank 2000; Kortte et al. 2002). Accordingly, we
predicted that higher-fit children would exhibit a superior task performance relative
to their lower-fit counterparts, and that this difference would be disproportionately
greater for TMT-B relative to TMT-A.
9.2 Methods
9.2.1 Participants
Fifty-three preadolescent children between 8 and 12 years of age participated in this

study. Given that childhood obesity has been associated with cognitive flexibility
(Kamijo et al. 2014), an extremely obese participant ( 95th BMI percentile; Kato
et al. 2011) was excluded from the analysis. Fitness group assignments (i.e., lower-
fit and higher-fit) were based on whether a child’s 20-m shuttle run test score fell
below the 30th percentile or above the 70th percentile according to normative data
provided by the Ministry of Education, Culture, Sports, Science and Technology
(MEXT 2012), Japan. Given that moderate-fit participants can blur differences in
fitness-related cognitive function between groups (Hillman et al. 2012), children
who did not qualify as lower-fit or higher-fit were excluded from further analysis.
Finally, we obtained 19 lower-fit and 19 higher-fit children for the study. Addition-
ally, to verify whether the manipulation of cognitive flexibility demands based on
the task conditions would be successful, 30 undergraduate students were recruited
from Waseda University. Prior to testing, all participants reported being free of
neurological diseases or physical disabilities and indicated normal or corrected-to-
normal vision. All participants provided written informed consent, or written
informed consent was provided by the participants’ legal guardian, in accordance
with the Ethics Committee on Human Research of Waseda University.
9.2.2 Procedure
After providing informed consent, participants had their height and weight mea-
sured. Standing height and weight measurements were completed with participants
wearing light-weight clothing and no shoes. Height and weight were measured
using a Tanita WB-3000 digital scale (Tanita Corp., Tokyo, Japan). BMI was
calculated by dividing body mass (kg) by height (m) squared (kg/m2). Concur-
rently, the participants’ legal guardians completed the Attention Deficit Hyperac-
tivity Disorder Rating Scale IV (DuPaul et al. 1998) and the Physical Activity
Readiness Questionnaire (Thomas et al. 1992) to screen for any previous health
issues that might be exacerbated by exercise. Further, given that socioeconomic
status has been associated with cognitive control (Mezzacappa 2004) and fitness
(Freitas et al. 2007), maternal educational attainment – quantified on a scale from
1 indicating that they did not complete high-school to 5 indicating earning an
advanced degree – was assessed as a proxy for socioeconomic status. Participants
then completed the TMT. Participants were given instructions, afforded the oppor-
tunity to ask questions, and were able to practice the task prior to the start of testing.
The 20-m shuttle run test was conducted on a different day.
9.2.3 Trail Making Test
In TMT-A the participants were asked to draw a line connecting circles with the
numbers 1–25 in numerical order. Following completion of the TMT-A, partici-
pants completed TMT-B, in which they were instructed to draw a line connecting
circles with the numbers 1–13 and 12 hiragana (Japanese cursive syllabary) letters
by alternating between the numbers and letters in numerical and Japanese alpha-
betical order. Participants were told to complete each part of the TMT as accurately
and quickly as possible. When an error was made, the examiner pointed it out
immediately and participants were required to return to the circle where the error
occurred and continue. The recorded score was the time in seconds required to
complete each TMT, with a higher score indicating poorer performance.
9.2.4 Fitness Assessment
The 20-m shuttle run test was performed according to Leger et al. (1988). Partic-
ipants were required to run back and forth between two lines 20 m apart at a
specified pace. The initial speed was set at 8.5 km/h with the speed increasing by
0.5 km/h every minute. A pre-recorded CD was used to set the pace to run laps. The
test was continued until the participant stopped due to fatigue or could not reach the
end lines concurrent with the audio signals on two consecutive occasions. The
finished lap number was recorded. To exclude age- and sex-related differences, in
this study, the percentile score was calculated as an index of aerobic fitness based
on normative data provided by the MEXT (2012).
9.2.5 Statistical Analysis
The TMT score was analyzed using a 3 (Group: lower-fit children, higher-fit
children, young adults) Condition (TMT-A, B) multivariate analysis of variance.
Post hoc analyses were conducted using the Scheffé test. All statistical analyses
were conducted using a significance level of p ¼ .05.
9.3 Results
9.3.1 Participant Characteristics
Participant demographic and fitness data are provided in Table 9.1. Fitness com-
parisons using a between-participants t-test (i.e., lower-fit children versus higher-fit
children) indicated that higher-fit children ran more laps during the 20-m shuttle run
test than lower-fit children, t(36) ¼ 9.1, p < .001, confirming the aerobic fitness
groupings. The other variables did not differ between the child groups, t(36) 1.0,
p .25.
9.3.2 Cognitive Performance
Analysis revealed main effects for Group, F(2, 65) ¼ 59.2, p < .001, η2p ¼ .65, and
Condition, F(1, 65) ¼ 98.1, p < .001, η2p ¼ .60, which were qualified by a
Group Condition interaction, F(2, 65) ¼ 9.9, p < .001, η2p ¼ .23. Figure 9.1 illus-
trates the Group Condition interaction. Scheffé post hoc analyses for the TMT-A
revealed a higher score (i.e., poorer performance) for lower-fit children relative to
higher-fit children, p ¼ .02, and young adults, p < .001, and a higher score for
higher-fit children relative to young adults, p < .001. For the TMT-B, both lower-
fit and higher-fit children had a higher score relative to young adults, p < .001, but
no difference was observed between lower-fit and higher-fit children, p ¼ .99.
Table 9.1 Mean (SD) values for participant demographics and fitness data
Measure Lower-fit children Higher-fit children Young adults
No. of participants 19 (9 females) 19 (10 females) 30 (14 females)
Age (years) 10.6 (1.1) 10.4 (0.8) 21.5 (1.1)
20-m shuttle run (#laps)* 31.6 (10.0) 69.8 (15.2) –
20-m shuttle run (%ile)* 17.6 (7.9) 86.8 (9.8) –
Body mass index (kg/m2) 17.1 (2.0) 16.4 (1.4) 22.2 (1.9)
Maternal education 2.8 (0.8) 2.9 (0.9) –
ADHD 7.7 (6.7) 7.4 (6.1) –
Notes: Maternal education – educational attainment was quantified on a scale from 1 indicating
that they did not complete high-school to 5 indicating earning an advanced degree. ADHD – scores
on the Attention Deficit Hyperactivity Disorder Rating Scale IV. Significant difference, unpaired t-
test between the lower-fit and higher-fit groups, *p < .05
Fig. 9.1 Mean (SE) score Lower-fit children Young adults

for each group and Higher-fit children
condition
100
*
*
80
*
60 *
Time (sec)
*
40
20
0
TMT-A TMT-B
9.4 Discussion
This study examined the association between aerobic fitness and cognitive flexi-
bility, which is one aspect of cognitive control, in preadolescent children using the
TMT. The main finding indicated that higher-fit children exhibited superior task
performance relative to their lower-fit peers on the TMT-A, which requires minimal
cognitive control. This result corroborates previous findings indicating the general
nature of the relationship between fitness and cognition in preadolescent children
(Buck et al. 2008; Hillman et al. 2009). In contrast, task performance did not differ
between lower-fit and higher-fit children for the TMT-B. Given that prior research
indicated that differences in task performance between lower-fit and higher-fit
children (i.e., superior task performance for higher-fit children) was disproportion-
ately larger for a task condition requiring greater cognitive flexibility (Pontifex
et al. 2011), we hypothesized that differences in task performance between groups
would be disproportionately greater for the TMT-B relative to the TMT-A. The
current finding did not support our hypothesis.
One possible explanation for the unexpected finding might be that the TMT-B
was too difficult for both lower-fit and higher-fit children. This possibility is
supported by previous studies examining the relationship between fitness/physical
activity and cognitive flexibility using task-switching tasks in young adults. A
typical task-switching task consists of pure and mixed task conditions. The pure
task condition includes a single rule set (i.e., AAAAAA. . . or BBBBBB. . .) and
asks participants to respond based on the stimuli presented. The mixed-task condi-
tion consists of two or more tasks and requires participants to change rapidly
between different tasks (e.g., AABBAA. . . or ABABAB. . .). The mixed task
condition requires greater amounts of cognitive control such as cognitive flexibility,
working memory, and inhibition (for a review, see Monsell 2003). Hillman
et al. (2006a) used the task-switching task to examine the relationship between
physical activity level and cognitive control in younger and older adults, and
indicated that physically active individuals exhibited a shorter reaction time
(RT) relative to inactive individuals irrespective of age. Kamijo and Takeda
(2010) showed, using the task-switching task, that physically active young adults
had a shorter RT for the mixed task condition relative to their inactive peers,
whereas no such difference was observed for the pure task condition. These results
suggest that physical activity level is selectively associated with task performance
for task conditions requiring greater cognitive flexibility. Thus, it is likely that a
higher level of physical activity is associated with superior cognitive flexibility
even during young adulthood.
In contrast, Scisco et al. (2008) observed no relationship between aerobic fitness
and task performance during the task-switching task in young adults. One possible
source of this discrepancy relates to differences in task difficulty among these
studies. In Hillman et al. (2006a) and Kamijo and Takeda (2010), single digit
numbers (digits 1–4 and 6–9) were used for the task-switching task. Specifically,
participants were required to judge whether the number was odd or even, or whether
the number was greater or lesser than 5. In Scisco et al. (2008), participants were
asked to judge whether double digit numbers (digits 11–99) were odd or even,
whether the numbers were greater or lesser than 50, whether the sum of two digits
was odd or even, or whether the sum of two digits was greater or lesser than 10.
Task-switching tasks including four tasks as well as calculation (Scisco et al. 2008)
must undoubtedly be more difficult than those involving only two tasks (Hillman
et al. 2006a; Kamijo and Takeda 2010). Thus, it is speculated that the relationship
between fitness/physical activity and cognitive control might blur if the cognitive
task is too difficult, even though the task requires variable amounts of cognitive
control based on task conditions.
Previous research, which investigated the relation between aerobic fitness and
cognitive function in younger and older adults using task difficulty manipulation
during an oddball task, supports this speculation (Pontifex et al. 2009). Specifically,
the oddball task required participants to respond to rare target stimuli (i.e., oddballs)
and withhold their response to frequent nontarget stimuli. The target stimuli were
5.5 cm diameter white circles in both easy and difficult task conditions, whereas the
diameter of the nontarget stimuli (i.e., white circles) were 3.0 cm and 5.0 cm in the
easy and difficult condition, respectively. Results indicated that higher-fit individ-
uals exhibited shorter RTs relative to their age-matched lower-fit counterparts for
the easy, but not for the difficult, task condition. Thus, it is likely that the positive
association between aerobic fitness and cognitive function is affected by task
difficulty. Taken together, it is plausible that, in this study, task performance did
not differ between the child groups for the TMT-B since this task condition was too
difficult for preadolescent children.
It is noteworthy that differences in task performance between children and
young adults were larger for the TMT-B (about 40 ms) relative to the TMT-A
(about 20–30 ms). Given that there is no doubt that young adults have superior
cognitive flexibility compared to children, the differences in task performance as a
function of age and the task condition indicate that cognitive control demands were
successfully manipulated between the TMT-A and TMT-B in this study. Nonethe-
less, a positive association between aerobic fitness and cognitive function was
selectively observed for the TMT-A requiring minimal cognitive control. Given
that accumulating evidence has demonstrated that the association is selectively and
disproportionately greater for higher-order cognitive functions (i.e., cognitive con-
trol) not only during adulthood (e.g., Colcombe and Kramer 2003; Hillman
et al. 2006a; Kamijo and Takeda 2010; Kramer et al. 1999) but also during
childhood (e.g., Kamijo et al. 2011; Pontifex et al. 2011), the result of no associ-
ation between fitness and task performance on the TMT-B might not be due to its
cognitive control demands, but due to its task difficulty for the current participants
(i.e., preadolescent children) as discussed above.
In sum, aerobic fitness was associated with task performance for the TMT-A, but
not for the TMT-B in preadolescent children. These findings support the general
nature of the relationship between fitness and cognition during childhood and
suggest that this relationship can be underestimated or even disappear if cognitive
tasks are too difficult. Thus, in conclusion, we propose that researchers should
select cognitive tasks with caution when examining the association between fitness
and cognition. This caution should include not only aspects involved with cognitive
control demands, but also aspects dealing with task difficulty.
References
Arbuthnott K, Frank J (2000) Trail making test, part B as a measure of executive control:
validation using a set-switching paradigm. J Clin Exp Neuropsychol 22:518–528. doi:10.
1076/1380-3395(200008)22:4;1-0;FT518
Best JR, Miller PH (2010) A developmental perspective on executive function. Child Dev
81:1641–1660. doi:10.1111/j.1467-8624.2010.01499.x
Buck SM, Hillman CH, Castelli DM (2008) The relation of aerobic fitness to stroop task
performance in preadolescent children. Med Sci Sports Exerc 40:166–172. doi:10.1249/mss.
0b013e318159b035
Colcombe S, Kramer AF (2003) Fitness effects on the cognitive function of older adults: a meta-
analytic study. Psychol Sci 14:125–130. doi:10.1111/1467-9280.t01-1-01430
Department of Health and Human Services, Department of Education (2000) Promoting better
health for young people through physical activity and sports. A report to the President from the
Secretary of Health and Human Services and the Secretary of Education. Centers for Disease
Control, Silver Spring
Diamond A (2006) The early development of executive functions. In: Bialystok E, Craik FIM
(eds) Lifespan cognition: mechanisms of change. Oxford University Press, New York, pp
70–95
DuPaul GJ, Power TJ, Anastopoulos AD, Reid R (1998) ADHD rating scale – IV: checklists,
norms, and clinical interpretation. Guilford Press, New York
Eriksen BA, Eriksen CW (1974) Effects of noise letters upon the identification of a target letter in a
nonsearch task. Percept Psychophys 16:143–149. doi:10.3758/BF03203267
Freitas D, Maia J, Beunen G, Claessens A, Thomis M, Marques A, Crespo M, Lefevre J (2007)
Socio-economic status, growth, physical activity and fitness: the Madeira growth study. Ann
Hum Biol 34:107–122. doi:10.1080/03014460601080983
Friedman D, Nessler D, Cycowicz YM, Horton C (2009) Development of and change in cognitive
control: a comparison of children, young adults, and older adults. Cogn Affect Behav Neurosci
9:91–102. doi:10.3758/CABN.9.1.91
Gogtay N, Giedd JN, Lusk L, Hayashi KM, Greenstein D, Vaituzis AC, Nugent TF 3rd, Herman
DH, Clasen LS, Toga AW, Rapoport JL, Thompson PM (2004) Dynamic mapping of human
cortical development during childhood through early adulthood. Proc Natl Acad Sci U S A
101:8174–8179. doi:10.1073/pnas.0402680101
Hillman CH, Kramer AF, Belopolsky AV, Smith DP (2006a) A cross-sectional examination of age
and physical activity on performance and event-related brain potentials in a task switching
paradigm. Int J Psychophysiol 59:30–39. doi:10.1016/j.ijpsycho.2005.04.009
Hillman CH, Motl RW, Pontifex MB, Posthuma D, Stubbe JH, Boomsma DI, de Geus EJ (2006b)
Physical activity and cognitive function in a cross-section of younger and older community-
dwelling individuals. Health Psychol 25:678–687. doi:10.1037/0278-6133.25.6.678
Hillman CH, Buck SM, Themanson JR, Pontifex MB, Castelli DM (2009) Aerobic fitness and
cognitive development: event-related brain potential and task performance indices of executive
control in preadolescent children. Dev Psychol 45:114–129. doi:10.1037/a0014437
Hillman CH, Kamijo K, Pontifex MB (2012) The relation of ERP indices of exercise to brain
health and cognition. In: Boecker H, Hillman CH, Scheef L, Strèuder HK (eds) Functional
neuroimaging in exercise and sport sciences. Springer, New York, pp 419–446. doi:10.1007/
978-1-4614-3293-7_18
Kamijo K, Takeda Y (2010) Regular physical activity improves executive function during task
switching in young adults. Int J Psychophysiol 75:304–311. doi:10.1016/j.ijpsycho.2010.01.
002
Kamijo K, Pontifex MB, O’Leary KC, Scudder MR, Wu CT, Castelli DM, Hillman CH (2011) The
effects of an afterschool physical activity program on working memory in preadolescent
children. Dev Sci 14:1046–1058. doi:10.1111/j.1467-7687.2011.01054.x
Kamijo K, Pontifex MB, Khan NA, Raine LB, Scudder MR, Drollette ES, Evans EM, Castelli DM,
Hillman CH (2014) The negative association of childhood obesity to cognitive control of
action monitoring. Cereb Cortex 24:654–662. doi:10.1093/cercor/bhs349
Kato N, Takimoto H, Sudo N (2011) The cubic functions for spline smoothed L, S and M values
for BMI reference data of Japanese children. Clin Pediatr Endocrinol 20:47–49. doi:10.1297/
cpe.20.47
Kortte KB, Horner MD, Windham WK (2002) The trail making test, part B: cognitive flexibility or
ability to maintain set? Appl Neuropsychol 9:106–109. doi:10.1207/S15324826AN0902_5
Kramer AF, Hahn S, Cohen NJ, Banich MT, McAuley E, Harrison CR, Chason J, Vakil E,
Bardell L, Boileau RA, Colcombe A (1999) Ageing, fitness and neurocognitive function.
Nature 400:418–419. doi:10.1038/22682
Leger LA, Mercier D, Gadoury C, Lambert J (1988) The multistage 20 metre shuttle run test for
aerobic fitness. J Sports Sci 6:93–101. doi:10.1080/02640418808729800
Mezzacappa E (2004) Alerting, orienting, and executive attention: developmental properties and
sociodemographic correlates in an epidemiological sample of young, urban children. Child
Dev 75:1373–1386. doi:10.1111/j.1467-8624.2004.00746.x
Miller EK, Cohen JD (2001) An integrative theory of prefrontal cortex function. Annu Rev
Neurosci 24:167–202. doi:10.1146/annurev.neuro.24.1.167
Ministry of Education, Culture, Sports, Science and Technology (2012) Statistical abstract 2012,
3rd edn. Physical Education and Sports. http://www.e-stat.go.jp/SG1/estat/NewList.do?
tid¼000001016672. Accessed 24 Dec 2013
Miyake A, Friedman NP, Emerson MJ, Witzki AH, Howerter A, Wager TD (2000) The unity and
diversity of executive functions and their contributions to complex “Frontal Lobe” tasks: a
latent variable analysis. Cogn Psychol 41:49–100. doi:10.1006/cogp.1999.0734
Monsell S (2003) Task switching. Trends Cogn Sci 7:134–140. doi:10.1016/S1364-6613(03)
00028-7
Pontifex MB, Hillman CH, Polich J (2009) Age, physical fitness, and attention: P3a and P3b.
Psychophysiology 46:379–387. doi:10.1111/j.1469-8986.2008.00782.x
Pontifex MB, Raine LB, Johnson CR, Chaddock L, Voss MW, Cohen NJ, Kramer AF, Hillman CH
(2011) Cardiorespiratory fitness and the flexible modulation of cognitive control in preadoles-
cent children. J Cogn Neurosci 23:1332–1345. doi:10.1162/jocn.2010.21528
Raz N, Gunning FM, Head D, Dupuis JH, McQuain J, Briggs SD, Loken WJ, Thornton AE, Acker
JD (1997) Selective aging of the human cerebral cortex observed in vivo: differential vulner-
ability of the prefrontal gray matter. Cereb Cortex 7:268–282. doi:10.1093/cercor/7.3.268
Scisco JL, Leynes PA, Kang J (2008) Cardiovascular fitness and executive control during task-
switching: an ERP study. Int J Psychophysiol 69:52–60. doi:10.1016/j.ijpsycho.2008.02.009
Thomas S, Reading J, Shephard RJ (1992) Revision of the physical activity readiness question-
naire (PAR-Q). Can J Sport Sci 17:338–345
Tomkinson GR, Macfarlane D, Noi S, Kim DY, Wang Z, Hong R (2012) Temporal changes in
long-distance running performance of Asian children between 1964 and 2009. Sports Med
42:267–279. doi:10.2165/11599160-000000000-00000
Part II
The Physiological Basis of Sports
Performance
Chapter 10
Genetic Polymorphisms Associated with Elite
Athlete Status
Eri Miyamoto-Mikami, Noriyuki Fuku, and Masashi Tanaka
Abstract Variance in human physiological traits is understood to be the result of

both environmental and genetic factors. Indeed, it has been reported that 66 % of
the variance in athletic status is explained by genetic factors. Elucidating the
substantial genetic contribution to elite athletic performance would enhance our
understanding of the limits to human performance and improve methods for
detecting sports talent. To date, a number of studies have attempted to identify
genetic polymorphisms associated with athletic performance, and the number of
genetic polymorphisms associated with physical performance-related phenotypes
are increasing every year. In this chapter, we introduce some of these studies
focusing on mitochondrial DNA polymorphisms and the two most studied nuclear
DNA polymorphisms, namely angiotensin I-converting enzyme (ACE) and
α-actinin-3 (ACTN3) polymorphisms.
Keywords Elite athlete • Genetic polymorphism • Mitochondrial DNA • ACE •

ACTN3
10.1 Introduction
Elite/world-class athletic status is the result of a complex interaction of physiolog-

ical, psychological, and socio-cultural factors. Variance in human physiological
traits is understood to be the result of both environmental and genetic factors, which
interact with each other (Brutsaert and Parra 2006). Tucker and Collins (Tucker and
E. Miyamoto-Mikami
Ritsumeikan University, Shiga, Japan
e-mail: eri-m.d925@y.fuji.waseda.jp
N. Fuku (*)
Juntendo University, Chiba, Japan
e-mail: noriyuki.fuku@nifty.com
M. Tanaka
Tokyo Metropolitan Institute of Gerontology, Tokyo, Japan
e-mail: mtanaka@tmig.or.jp

106 E. Miyamoto-Mikami et al.
Collins 2012) proposed a theoretical model which holds that a threshold of indi-
vidual athletic performance is determined by genetic factors, and the genetic
potential is realized by optimal training (Tucker and Collins 2012). Given this
theoretical model, contribution of genetic factors to elite athletic performance
should be considerable. Indeed, it has been reported that significant genetic com-
ponent contributes to the variance in athlete status (De Moor et al. 2007). Eluci-
dating the substantial genetic contribution to elite athletic performance would
enhance our understanding of the limits to human performance and improve
methods to detect sporting talent.
Since a polymorphism of the angiotensin I-converting enzyme (ACE) gene was
firstly reported to have an impact on human physical performance in 1998 (Mont-
gomery et al. 1998), a number of studies have attempted to identify genetic poly-
morphisms associated with athletic performance. In 2009, over 200 nuclear and
mitochondrial DNA (mtDNA) polymorphisms associated with physical perfor-
mance traits have been summarised as the ‘The Human Gene Map for Performance
and Health-related Fitness Phenotypes’ (Bray et al. 2009), and the number of
genetic polymorphisms associated with physical performance-related phenotypes
are increasing every year (Perusse et al. 2013). In a review of 133 studies involving
athletic cohorts published during the period 1997–2012, 59 genetic markers were
reported to be associated with endurance performance, 20 with power/strength
related performance, while 25 % of these markers were positively associated with
performance in at least two studies (Ahmetov and Fedotovskaya 2012). In this
chapter, we introduce some of these studies by focusing on mtDNA polymorphisms
and most studied two nuclear DNA polymorphisms, namely ACE and α-actinin-3
(ACTN3) polymorphisms. We include some recent studies from our laboratory.
10.2 Mitochondrial DNA Polymorphisms
10.2.1 Mitochondrial Function and Mitochondrial Genetics
Almost all eukaryotic cells contain mitochondria, which have a primary purpose of
producing ATP through respiration. The majority of ATP consumed in skeletal
muscle during exercise is supplied by glycolysis and mitochondrial oxidative
phosphorylation. Mitochondrial oxidative phosphorylation is a highly-efficient
energy production system, providing 36 molecules of ATP per glucose molecule,
in contrast to the 2 ATP molecules produced by glycolysis. Mitochondria contain
their own circular DNA, namely mtDNA. The human mtDNA is 16,569-bp in
length and is inherited almost exclusively through the maternal lineage. The
mtDNA is comprised of 37 genes essential for mitochondrial function [13 proteins
of mitochondrial oxidative phosphorylation system (OXPHOS), 2 ribosomal RNAs
(rRNAs), and 22 transfer RNAs (tRNAs)] (Anderson et al. 1981). Therefore, it is
10 Genetic Polymorphisms Associated with Elite Athlete Status 107
possible that nucleotide variations in mtDNA influence the capacity/efficiency of

ATP production by mitochondria.
10.2.2 Mitochondrial Haplogroup
A ‘haplogroup’ is a group of similar haplotypes which are combinations of alleles at

multiple loci transmitted together on the same chromosome. As mentioned above,
mtDNA is inherited maternally, and therefore nucleotide substitutions that have
occurred in the maternal ancestry are directly transmitted to their progeny. Conse-
quently, people sharing a common maternal ancestor possess a characteristic cluster
of tightly linked mtDNA polymorphisms. The combination of mtDNA polymor-
phisms is called a ‘mitochondrial haplogroup’. Since polymorphisms existing on
mtDNA are linked with other mtDNA polymorphisms, it is difficult to detect a
causal polymorphism by analysis of each polymorphism. Therefore, in order to
examine the association between mtDNA polymorphisms and a phenotype, a
haplogroup analysis is required as a first step.
The distribution of mitochondrial haplogroups is continentally or ethnically
specific. Each of the African haplogroups (mainly, L0–L3) has deeper genetic
roots than those haplogroups in European and Asian populations, and haplogroup
L3 is proposed to be the ancestor of all non-African populations. Macrohaplogroups
N and M diverged from haplogroup L3, and European haplogroups (H, I, J, K, S, T,
U, V, W, etc.) belong to macrohaplogroup N (Torroni et al. 2000). On the other
hand, Asian haplogroups belong to both macrohaplogroups N and M
(haplogroups A, B, F, and N9a to macrohaplogroup N; haplogroups M7a, M7b,
M8, D, and G to macrohaplogroup M) (Tanaka et al. 2004). The geographic region-
specific variations of mitochondrial haplogroups are now known to have been
formed by natural selection, possibly to allow habitation in cold climatic environ-
ments (Mishmar et al. 2003; Ruiz-Pesini et al. 2004). These mtDNA variations are
suggested to influence mitochondrial function and bioenergetics (Mishmar
et al. 2003; Ruiz-Pesini et al. 2004; Wallace 2005).
10.3 Heritability of Physical Performance
The heritability is the proportion of the total phenotypic variance that can be
attributed to genetic differences among individuals, and the values of heritability
range from 0 to 100 %. A heritability close to zero indicates that genetic variance
does not contribute to differences in the phenotype between individuals, but rather
that observed phenotypic differences are primarily attributed to the differences in
environmental exposure. On the other hand, a heritability close to 100 % indicates
that genetic factors are the key determinant of the phenotypic variance.
For maximal leg extension power, the genetic contribution was estimated at
46 % in the twin study including 706 postmenopausal women: 227 pairs of
monozygous twins and 126 pairs of dizygous twins (Arden and Spector 1997).
For maximal isometric knee extensor strength, Silventoinen et al. (Silventoinen
et al. 2008) reported that genetic factor explained 50 % of strength variation in
154,970 brother pairs, 1,582 monozygous and 1,864 dizygous twin pairs, aged 16–
25 years. They also examined genetic co-variation between muscle strength at
different parts of body, and showed that part of genetic factors are common to
different strength indicators (i.e., elbow flexion and handgrip strength). Tiainen
et al. (Tiainen et al. 2005) have investigated whether knee extensor power and
maximal isometric knee extensor strength share a genetic component, and found
that knee extensor power and isometric knee extensor strength shared a genetic
component in common, which accounted for 32 % of the total variance in knee
extensor power and 48 % in isometric knee extensor strength. Collectively, the
estimated heritability of maximal leg extension power ranged from 32 to 46 %, and
heritability of maximal isometric knee extensor strength was found to be around
50 %.
In the context of endurance performance, Bouchard et al. (1986) have estimated
the size of the genetic effect for VO2max after controlling for the various sources of
bias based on a relatively large sample of 42 brothers, 66 dizygous twins and
106 monozygous twins, of 16–34 years of age. They found that the size of the
genetic effect in VO2max · kg1 body weight reached 40 %. On the other hand, a
familial study by the same laboratory reported that the maximal heritability of
VO2max adjusted for age, sex, and body mass reached about 50 % based on data
involving 429 individuals from 86 nuclear families, aged between 16 and 65 years
(Bouchard et al. 1998). However, due to the presence of a significant spouse
correlation of VO2max, which suggest a contribution of non-genetic factors to the
familial aggregation, the authors concluded that the genetic heritability of VO2max
was less than 50 %. Most recent twin studies have reported that heritability of
VO2max, adjusted for lean body mass, was 71 % based on 59 monozygous twins
and 92 dizygous twins, aged between 23 and 31 year (Mustelin et al. 2011). Thus,
estimations for the heritability of VO2max ranged from 40 to 71 %.
A twin study based on 793 monozygous and 1,000 dizygous twins have esti-
mated that the heritability of athletic status was 66 % (De Moor et al. 2007). This
proportion of a genetic factor determining athlete status seems to be reasonable in
terms of the heritability of physical performance such as muscle strength/power
(32–50 %) and VO2max (40–71 %).
Interestingly, the above mentioned literature on heritability of VO2max (Bou-
chard et al. 1998) indicated that the maternal influence was greater than the paternal
influence. In that study, maternal influence reached about 30 %, which was more
than half of total heritability of VO2max (50 %). Lesage et al. (1985) also found
aerobic capacity to be more strongly influenced by maternal rather than paternal
inheritance. Therefore, maternally inherited mtDNA is likely to contain candidate
genes influencing aerobic capacity and therefore potentially elite athlete status.
10.4 Mitochondrial Haplogroups and Elite Athlete Status
10.4.1 Mitochondrial Haplogroup and Elite African Athletes
In East African populations, which have been successful in international distance

running, the relationship of mitochondrial haplogroups with elite athlete status have
been examined. Scott et al. (2009) reported that mitochondrial haplogroups were
associated with elite Kenyan endurance athletic status, but not associated with elite
Ethiopian endurance athletic status (Scott et al. 2005). Elite Kenyan endurance
athletes displayed an excess of haplogroup L0 and a death of haplogroup L3
compared with the general Kenyan population. On the other hand, Deason
et al. (2012) reported the association between mitochondrial haplogroup and elite
athletic status in sprinters of sub-Saharan ancestry (i.e., Jamaican and African-
American sprinters). In that study, there was no significant difference in haplogroup
frequencies between elite Jamaican sprinters and Jamaican controls. However, elite
African-American sprinters showed a higher frequency of non-African mitochon-
drial haplogroups, which included all haplogroups not commonly found in
sub-Saharan Africa, as compared with African-American controls. This result
suggest that the maternal admixture may play a role in sprint performance.
10.4.2 Mitochondrial Haplogroup and Elite European

Athletes
The first study of Europeans regarding association of mitochondrial haplogroups

with elite athlete status was reported in 2005. Niemi and Majamaa (2005) deter-
mined mitochondrial haplogroup frequencies in Finnish elite endurance (n ¼ 52)
and sprint (n ¼ 89) athletes, and found that the frequencies of mitochondrial
haplogroups were significantly different between endurance and sprint athletes.
Elite endurance athletes showed lower frequencies of haplogroup K and haplogroup
J compared with elite sprint athletes and controls, where none of the elite endurance
athletes belonged to haplogroup K or subhaplogroup J2. This result was partially
consistent with the association between haplogroup J and low VO2max (Marcuello
et al. 2009). In the Spanish population, Castro et al. (2007) compared the frequency
distribution of mitochondrial haplogroups between healthy controls (n ¼ 250) and
elite endurance athletes including cyclists, long-distance runners, and long distance
rowers (n ¼ 95). They found that the frequency of haplogroup T was significantly
lower in elite endurance athletes as compared with controls.
10.4.3 Mitochondrial Haplogroup and Elite Asian Athletes
In Asian populations, we examined the associations between mitochondrial

haplogroups and elite Japanese athletic status (Mikami et al. 2011). The percent
frequency of mitochondrial haplogroups found in 139 Japanese Olympic athletes
from various sports was compared with that in 672 Japanese control data (CON)
from our Human Mitochondrial Genome Single Nucleotide Polymorphism Data-
base (http://mtsnp.tmig.or.jp/mtsnp/index.shtml) (Tanaka et al. 2004). The group of
139 athletes was comprised of 79 endurance/middle-power athletes (EMA) and
60 sprint/power athletes (SPA). From the analysis of hypervariable segment 1 poly-
morphisms and several protein-coding region polymorphisms on the mtDNA, the
subjects were classified into 12 major Japanese mitochondrial haplogroups (i.e., F,
B, A, N9a, N9b, M7a, M7b, M*, G2, G1, D5 or D4). When the frequency of each
haplogroup versus the sum of all others was compared between EMA or SPA and
CON, EMA displayed a greater proportion of haplogroup G1 (OR 2.52 [95 % CI
1.05–6.02], P ¼ 0.032; Fig. 10.1), with 8.9 % in EMA compared with 3.7 % in
CON. Additionally, the frequency of haplogroup B in EMA tended to be lower than
in CON (OR 0.45 [95 % CI 0.18–1.14], P ¼ 0.084, Fig. 10.1), though this difference
was not statistically significant. On the other hand, SPA showed an excess of
haplogroup F (OR 2.79 [95 % CI 1.28–6.07], P ¼ 0.007, Fig. 10.1), with 15 % in
SPA compared with 6.0 % in CON.
Interestingly, Okura et al. (2003) reported that m.15497G>A polymorphism
characterizing haplogroup G1 was associated with obesity-related phenotypes in
middle-aged individuals. Therefore, we hypothesized that haplogroup G1 is “a
thrifty genotype” because of tightly coupled OXPHOS (Fuku et al. 2013; Mikami
et al. 2011). Tightly coupled OXPHOS would be expected to decrease heat pro-
duction and result in a higher efficiency of ATP production. The improved effi-
ciency in ATP production could explain, at least in part, the association between
haplogroup G1 and endurance performance. On the other hand, this improved
efficiency in ATP production may predispose individuals to obesity if they become
sedentary later on in life. Further extensive studies are necessary to investigate this
hypothesis.
Sprint/power performance relies more on anaerobic glycolysis than mitochon-
drial OXPHOS. However, we found that mitochondrial haplogroup F was signifi-
cantly associated with elite sprint/power athletic status. Mitochondrial haplogroup
F is one of the major components of macrohaplogroup N (Tanaka et al. 2004). We
also found that macrohaplogroup N was significantly associated with stronger leg
extension power and higher vertical jump performance (Fuku et al. 2012). It is
possible that certain mtDNA polymorphisms may influence the regulation of ATP
production not only by the OXPHOS system in the mitochondria but also by the
glycolytic pathway in the cytosol. Indeed, Hwang et al. (2011) reported that hybrid
cells harboring haplogroups F and N9a exhibited significant differences in their
nuclear gene expression pattern; mitochondrial haplogroup F showed a decreased
gene expression of mitochondrial OXPHOS pathway and an increased gene
40
Endurance/middle-power (n = 79)
35 Control (n = 672)
30 Sprint/power (n = 60)
Frequency (%)
25
20
15 **
10 *
0
F B A N9a N9b M7a M7b M* G2 G1 D5 D4 others
Haplogroup
Fig. 10.1 Mitochondrial haplogroup distribution in endurance athletes sprint/power athletes, and
controls. *P ¼ 0.032 vs Control, *P ¼ 0.007 vs Control (Modified from Mikami 2011)
expression of the cytosolic glycolysis pathway compared with mitochondrial

haplogroup N9a. This observation can be regarded as a compensatory response
for decreased ATP production caused by a defective mitochondrial haplogroup,
resulting in an increased expression of the nuclear genes involved in glycolysis.
This phenomenon might explain, at least partly, the association between mitochon-
drial haplogroup F and elite SPA status.
In the Asian population, Kim et al. (2012) also investigated the association
between elite Korean athletic status and mitochondrial haplogroups. In that study,
they found that EMA had an excess of haplogroups M* and N9, but a dearth of
haplogroup B as compared with CON. On the other hand, the haplogroup distribu-
tion in SPA did not differ from CON. Although the dearth of haplogroup B in EMA
is consistent between elite Japanese (Mikami et al. 2011) and Korean (Kim
et al. 2012) athlete studies, other associations were not replicated in these studies.
Therefore, further replication studies and more detailed analysis (e.g. entire
mtDNA sequencing) are needed.
10.4.4 Comprehensive Analysis of mtDNA Variants in Elite

Japanese Athletes
As mentioned above, associations of different mitochondrial haplogroups with elite

athletic status has been reported in Europeans, Africans, and Asians, respectively.
However, the functional polymorphisms which are responsible for the previously
reported associations between haplogroups and elite athletic status have not been
identified. Each mitochondrial haplogroup is divided into several subhaplogroups
which are younger branches than the haplogroups in the mtDNA phylogenetic tree.
Young branches (subhaplogroups) in the mtDNA phylogenetic tree contain a higher
proportion of nonsynonymous substitutions in the protein-coding genes and sub-
stitutions in the rRNA and tRNA genes than old branches (haplogroups) (Elson
et al. 2004; Ruiz-Pesini and Wallace 2006), since influential variants have been
eliminated from the older branches of the tree by ‘purifying selection’
(in evolutionary terms). Thus, subhaplogroup-specific substitutions are more likely
to be associated with various health- and performance-related phenotypes. Then, in
order to identify the precise mtDNA polymorphisms which associate with elite
Japanese athletic status, we analyzed entire mtDNA sequences of 185 elite Japanese
athletes from various sports (Mikami et al. 2013). All athletes had represented
Japan at international competitions, and they were divided into 100 EMA and
85 SPA. The control group (CON) consisted of 672 Japanese individuals, whose
entire mtDNA sequences were registered in our Human Mitochondrial Genome
Single Nucleotide Polymorphism Database (http://mtsnp.tmig.or.jp/mtsnp/index.
shtml) (Tanaka et al. 2004). From the analysis of the entire mtDNA of 185 elite
Japanese athletes and 672 control subjects, we detected a total of 1,488 mtDNA
variants. Among these variants, a total of 311 variants were polymorphisms (minor
allele frequency > 1 % in CON), and the frequencies of these polymorphisms were
compared among the three groups. Consequently, we found that the EMA displayed
an excess of seven polymorphisms, including subhaplogroup D4e2- and
D4g-specific polymorphisms, as compared with CON (P < 0.05, Table 10.1),
whereas SPA displayed an excess of three polymorphisms and a dearth of nine
polymorphisms, including haplogroup G- and subhaplogroup G2a-specific poly-
morphisms, as compared with CON (P < 0.05, Table 10.1). However, none of these
polymorphisms differed significantly between groups after correcting for multiple
comparison (false discovery rate q-value > 0.05); a reflection most likely due to a
lack of sufficient statistical power. Therefore, replication studies are required to
confirm these associations between mtDNA variants and elite athletic performance.
10.5 ACTN3 and ACE Polymorphisms
Among all nuclear DNA polymorphisms, the most studied performance-associated

genes are the α-actinin-3 (ACTN3) and angiotensin I converting enzyme (ACE)
genes. ACTN3 gene seems to have the greatest impact on elite sprint/power athlete
status among previously published polymorphisms. The α-actinin-2 and -3 proteins
are localized to the Z-line in skeletal muscle, where these proteins help to anchor
actin filaments. The α-actinin-2 is expressed in all human skeletal muscle fibers,
whereas the α-actinin-3 is expressed only in fast-twitch muscle fibers (North and
Beggs 1996). Homozygosity for the common nonsense polymorphism R577X in
Table 10.1 Polymorphisms in the entire mtDNA with differences between groups (Modified from Mikami 2013)
10
CON EMA SPA

(n ¼ 672) (n ¼ 100) (n ¼ 85) EMA vs. CON SPA vs CON EMA vs. SPA
Amino Haplogroup/ OR OR OR
acid Minor Gene subhaplogroup P (95 % P (95 % P (95 %
Polymorphism change allele region specificity % (n) % (n) % (n) value CI) value CI) value CI)
Control region
m.16140 T>C C Control B5, B4c1b, 5.7 (38) 1.0 (1) 9.4 (8) 0.047 0.17 0.172 1.73 0.008 0.10
region M7a2 (0.02– (0.78– (0.10–
1.24) 3.85) 0.79)
m.16278C>T T Control B4d, D4g1, G2 8.3 (56) 10.0 (10) 0.0 (0) 0.578 1.22 0.006 0.00 0.003 INF
region (0.60– – –
2.48)
m.151C>T T Control D4a2a, D5c, 1.2 (8) 1.0 (1) 4.7 (4) 0.869 0.84 0.014 4.10 0.121 0.20
region B4d3, F1b1a1a1, (0.10– (1.21– (0.02–
M7b2b, Z1 6.78) 13.91) 1.87)
m.152 T>C C Control A, A1a, A1b, 18.5 28.0 (28) 24.7 0.025 1.72 0.167 1.45 0.613 1.19
region A2a, A3, A5c, (124) (21) (1.07– (0.85– (0.61–
B4b1b, B5b3, 2.77) 2.46) 2.29)
D4a, D4b1a1,
D4f, D4l1a,
Genetic Polymorphisms Associated with Elite Athlete Status
D5b1a, D5c, F1,

F1b1, F1c, F4a,
G2a1, G3a, G4a,
N1b, N9b,
M7a1a6, M7a1b,
M8a2, M12, Y1b,
Z
m.204 T>C C Control B4d3, B5, 4.9 (33) 2.0 (2) 10.7 (9) 0.192 0.40 0.029 2.32 0.013 0.17
region D4d1a, M7a1a, (0.09– (1.07– (0.04–
M7a1b, N9a2, Z4 1.67) 5.02) 0.81)
113
(continued)
Table 10.1 (continued)
114
CON EMA SPA

(n ¼ 672) (n ¼ 100) (n ¼ 85) EMA vs. CON SPA vs CON EMA vs. SPA
Amino Haplogroup/ OR OR OR
acid Minor Gene subhaplogroup P (95 % P (95 % P (95 %
Polymorphism change allele region specificity % (n) % (n) % (n) value CI) value CI) value CI)
m.514(CA)n (CA) Control A, B4a, B4c1b1, 40.6 27.0 (27) 40.0 0.009 1.85 0.912 1.03 0.061 0.55
5 region B4e, B5, C1, D4b, (273) (34) (1.16– (0.65– (0.30–
D4c, D5a2, F1, 2.95) 1.63) 1.03)
M7a1a, M7c,
M10a, N1b, Z5,
Z3
Poly–C stretch C7 Control C5, D4g1, F4b, 4.0 (27) 11.0 (11) 1.2 (1) 0.003 2.95 0.191 0.28 0.007 10.38
at m.568–573 region G4a, M10 (1.42– (0.04– (1.31–
6.16) 2.12) 82.17)
RNA-coding region
m.4343A>G G tRNA D4g 2.4 (16) 6.0 (6) 0.0 (0) 0.042 2.62 0.150 0.00 0.022 INF
Gln (1.00– – –
6.85)
m.5601C>T T tRNA G2a 4.8 (32) 3.0 (3) 0.0 (0) 0.429 0.62 0.040 0.00 0.107 INF
Ala (0.19– – –
2.06)
Protein-coding region
m.4833A>G Thr122Ala G ND2 G 8.6 (58) 11.0 (11) 2.4 (2) 0.438 1.31 0.044 0.26 0.022 5.13
(0.66– (0.06– (1.10–
2.59) 1.06) 23.83)
m.5108 T>C – C ND2 G, B4c2, M7a1b 9.4 (63) 12.0 (12) 2.4 (2) 0.408 1.32 0.029 0.23 0.013 5.66
(0.68– (0.06– (1.23–
2.54) 0.97) 26.05)
m.7600G>A – A COII G2a 4.8 (32) 3.0 (3) 0.0 (0) 0.429 0.62 0.040 0.00 0.107 INF
(0.19–
E. Miyamoto-Mikami et al.
– –
2.06)
10
m.9377A>G – G COIII G2a, D5b2 5.0 (34) 3.0 (3) 0.0 (0) 0.368 0.58 0.034 0.00 0.107 INF
(0.17– – –
1.93)
m.11215C>T – T ND4 D4e 4.8 (32) 10.0 (10) 3.5 (3) 0.031 2.22 0.610 0.73 0.086 3.04
(1.06– (0.22– (0.81–
4.67) 2.44) 11.42)
m.13104A>G – G ND5 D4g, D4k3 3.0 (20) 6.0 (6) 0.0 (0) 0.118 2.08 0.107 0.00 0.022 INF
(0.81– – –
5.31)
m.13563A>G – G ND5 G2 4.8 (32) 3.0 (3) 0.0 (0) 0.429 0.62 0.040 0.00 0.107 INF
(0.19– – –
2.06)
m.14200 T>C – C ND6 G2a 4.5 (30) 3.0 (3) 0.0 (0) 0.499 0.966 0.047 0.00 0.107 INF
(0.20– – –
2.21)
m.14569G>A – A ND6 G, B4b1b, N9a2c 9.5 (64) 13.0 (13) 2.4 (2) 0.279 1.42 0.027 0.23 0.008 6.20
(0.75– (0.06– (1.36–
2.68) 0.95) 28.32)
m.15314G>A Ala190Thr A Cytb D4a1a1 1.0 (7) 1.0 (1) 4.7 (4) 0.969 0.96 0.008 4.69 0.121 0.20
(0.12– (1.34– (0.02–
7.88) 16.37) 1.87)
Genetic Polymorphisms Associated with Elite Athlete Status
m.15518C>T – T Cytb D4g 2.4 (16) 6.0 (6) 0.0 (0) 0.042 2.62 0.150 0.00 0.022 INF
(1.00– – –
6.85)
m.15874A>G – G Cytb D4e2, A5 3.4 (23) 10.0 (10) 3.5 (3) 0.002 3.14 0.959 1.03 0.086 3.04
(1.45– (0.30– (0.81–
6.80) 3.51) 11.42)
Values in bold indicate significant differences between groups
ND NADH dehydrogenase, CO Cytochrome c oxidase, Cytb Cytochrome b, Gln glutamine, Ala alanine, Thr threonine, CON Controls, EMA Endurance/middle-
power athletes, SPA Sprint/power athletes, INF infinity
115
the ACTN3 gene results in a complete deficiency of α-actinin-3 in fast-twitch

muscle fibers. The R allele (RR + RX genotype) of this polymorphism was found
at a higher frequency in elite Australian sprint/power athletes relative to controls
(Yang et al. 2003) and this finding has been replicated in Finnish (Niemi and
Majamaa 2005), Greek (Papadimitriou et al. 2008), and Russian athletes
(Druzhevskaya et al. 2008). On the other hand, it was originally thought that the
XX genotype was advantageous to elite endurance performance (Yang et al. 2003).
Animal studies demonstrated that deficiency of α-actinin-3 results in a shift in
muscle metabolism towards the oxidative pathway and consequently greater run-
ning distance prior to exhaustion on a treadmill test (MacArthur et al. 2008;
MacArthur et al. 2007). However, there are inconsistent results regarding the
association between the XX genotype and elite endurance athletic status. A recent
meta-analysis of the published literature confirmed that the only association found
was between the ACTN3 RR + RX genotype and sprint/power athletic status in
Europeans (Alfred et al. 2011).
The insertion/deletion (I/D) polymorphism in the ACE gene was initially pro-
posed as an endurance performance-associated polymorphism (Montgomery
et al. 1998). The ACE gene product (i.e., angiotensin I converting enzyme) cata-
lyzes the conversion of angiotensin I to angiotensin II with action of elevating
arterial blood pressure. In addition, this enzyme is involved in glucose metabolism
through bradykinin, which induces muscle glucose uptake. The D allele in the ACE
gene is reported to be associated with higher plasma (Rigat et al. 1990) and tissue
(Danser et al. 1995) ACE activity than the I allele. So far, a number of studies have
examined the association between ACE I/D polymorphism and exercise perfor-
mance, and generally, the I allele is associated with endurance performance, and the
D allele with power phenotypes. However, associations in the opposite direction
were observed in the Asian population (Kim et al. 2010; Tobina et al. 2010).
10.5.1 ACTN3 Polymorphism in Elite Japanese Track

and Field Athletes
We investigated the association of the ACTN3 R577X polymorphism (Mikami et al.

2014) in elite Japanese track and field athletic status. 299 elite Japanese track and
field athletes and 649 Japanese nonathletic individuals (CON) were genotyped for
the ACTN3 R577X polymorphism. All athletes competed at the national or inter-
national level. They were divided into 134 SPA and 165 EMA based on event
duration and distance. When the ACTN3 R577X genotype frequency was compared
between the groups, the frequencies of the RR + RX genotype were significantly
higher in SPA than in CON (Fig. 10.2). Furthermore, sprinters with the RR + RX
genotype had faster personal best times for the 100-m event than those with the XX
genotype (Fig. 10.3a). Interestingly, all 7 male sprinters who had achieved the
London 2012 Olympic qualifying standard for the 100-m event (B standard:
10.24 s) possessed the RR or RX genotypes (Fig. 10.3a) and this was not the case
for the 400-m event (Fig. 10.3b). The ACTN3 R577X genotype accounted for
11.7 % of the variability in the personal best time for 100 m. On the other hand,
there was no significant association between the elite EMA status and the ACTN3
R577X genotype. These results indicate that the RR + RX genotype at amino acid
position 577 in the ACTN3 gene contributes to the advantage of elite sprint/power
performance in Asians as well as in Europeans. Furthermore, our results suggest
that the ACTN3 R577X polymorphism is associated with especially short sprinting
performance such as 100-m sprinting. In a previous study, it was reported that RR
genotype carriers showed a higher percentage of type IIx fibers than did XX
genotype carriers in the vastus lateralis. Also, RR genotype carriers showed higher
relative knee extension torques at a high velocity than did XX genotype carriers
(Vincent et al. 2007). These phenomena could explain, at least partly, the associ-
ation between the ACTN3 R577X polymorphism and short sprint performance.
Other published studies that investigated the relationship between ACTN3 and
athletic performance in Asians reported associations of ACTN3 R allele with elite
sprint/power athletic status (Chiu et al. 2011; Kikuchi et al. 2013) and with sprint/
power performance (Kikuchi et al. 2014; Shang et al. 2012). These results indicate
that the the ACTN3 R577X genotype is associated with elite sprint/power perfor-
mance not only in European athletes but also in Asian athletes.
10.5.2 ACTN3 and ACE Polymorphisms in Elite Caucasian

and East Asian Swimmers
In agreement with previous reports on European sprint/power athletes, we and

others confirmed the association between the ACTN3 R577X polymorphism and
elite sprint/power performance in Asians. However, the association between elite
sprint/power performance and the ACE I/D polymorphism is equivocal. In order to
explore further the associations of ACTN3 R577X and ACE I/D polymorphisms
with elite athletic status, we investigated whether such associations differed by
sports event or by ethnicity, by focusing on Caucasian and East Asian elite
swimmers (Wang et al. 2013). The Caucasian cohort was comprised of 200 elite
swimmers from European, Commonwealth, American, and Russian subcohorts.
The 200 swimmers were of world-class status or highly competitive in international
competitions, and were classified into 130 short/middle-distance (400 m) and
70 long-distance (>400 m) swimmers. Genotyping data from the ethnically
matched general population reported in previous studies were used as a control
group (CON) (Ahmetov et al. 2010; Lucia et al. 2006; Roth et al. 2008; Santiago
et al. 2010; Woods et al. 2001; Yang et al. 2003). The East Asian cohort was
comprised of 158 elite Japanese and 168 elite Taiwanese swimmers, and all had
participated in international competitions or national competitions. Since none of
these East Asian swimmers excelled at distances greater than 400 m, they were
Endurance/middle-power (n = 165)
70
Control (n = 649)
Sprint/power (n = 134)
60
Genotype frequency (%)
50
40
30
*
20
10
0
RR RX XX
Genotype
Fig. 10.2 Genotype frequencies of ACTN3 R577X polymorphism in elite track and field athletes
and controls. *R-dominant model: χ 2 ¼ 5.03, P ¼ 0.025 vs Control
a
*
b
11.0
50
100 m personal best time (s)
400 m personal best time (s)
10.8 49
10.6 48
10.4 47
10.2 46
10.0 45
9.8 44
RR RX XX RR RX XX
(n=6) (n=16) (n=6) (n=3) (n=12) (n=2)
Genotype Genotype
Fig. 10.3 Associations of ACTN3 R577X genotype with 100 m (a) and 400 m (b) personal best
time in male sprinters. *P ¼ 0.042 by t-test under the R-dominant model (Modified from Mikami
et al. 2014)
classified into 166 short-distance (<200 m) and 160 middle-distance (200 m)

swimmers. Controls for this group were 649 Japanese and 603 Taiwanese from the
general populations of these countries. For the ACTN3 R577X polymorphism,
although it was found that the RR + RX genotype tended to be over-represented
in the East Asian short-distance swimmers, there was no statistically significant
association in either the Caucasian or East Asian cohorts. For the ACE I/D poly-
morphism, the D allele was significantly associated with elite short/middle-distance
swimmer status in Caucasians (Fig. 10.4). On the other hand, the I allele was
significantly associated with elite short-distance swimmer status in East Asians
(Fig. 10.5). In both ethnic groups, there were no significant associations between
the ACE I/D polymorphism and elite long- or middle-distance status. In our studies,
associations between genotypes and elite athletic status differed by sports event
(i.e., track and field, and swimming), and by ethnicity (i.e., Caucasians and East
Asians). In terms of sports events, because determinants of sporting success are
sports dependent, it is reasonable to expect that different associations would be
observed between different sports events. For ethnicity, our study demonstrated that
opposite alleles of the ACE I/D polymorphism were associated with sprint swimmer
status in Caucasians and East Asians, respectively. Furthermore, two other studies
conducted on Asian athletes also showed an opposite tendency as compared with
Caucasians, where the D allele was associated with endurance performance (Tobina
et al. 2010), and the I allele was associated with power performance (Kim
et al. 2010). Although the different results between different ethnic groups may
be due to type I or type II errors, several other possibilities may explain the different
associations. One possible explanation is that, although the causative variants are
identical in Caucasians and East Asians, the haplotype blocks found in Caucasians
and East Asians are different in the environs of these polymorphisms, which result
in associations of opposite direction in the different ethnic groups. Additionally, it
may be that these polymorphisms affect the relevant physiology differently in
Caucasians and the East Asian population due to differences in environmental
factors such as training methods.
10.6 Prospect
In this chapter, we introduced mtDNA polymorphisms and two nuclear DNA

polymorphisms that are associated with elite athletic status, and also included
several of our recent reports. Our data suggested that the associations of genetic
polymorphisms with elite athletic status may not necessarily be the same between
different ethnicities. Most previous studies that examined the association between
genetic polymorphisms and physical performance were conducted on Caucasian
populations. It is definitely also important to accumulate evidence for associations
between genetic polymorphisms and physical performance in Asian populations.
Of course, the focused polymorphisms we reported on in this chapter are only a
small part of the genetic factors involved with elite athletic status. In order to
Long distance
60
Control
Short/middle distance
50
40
30
20
10
0
II ID DD
Genotype
Fig. 10.4 Genotype frequencies of ACE I/D polymorphism in elite Caucasian swimmers and
controls (Modified from Wang et al. 2013)
60
Middle distance
50 Control
Short distance
40
30
20
10
0
II ID DD
Genotype
Fig. 10.5 Genotype frequencies of ACE I/D polymorphism in elite East Asian swimmers and
controls (Modified from Wang 2013)
improve our understanding of the effects of genetic factors on elite athletic status, a
genome-wide approach will be necessary.
References
Ahmetov II, Fedotovskaya ON (2012) Sports genomics: current state of knowledge and future
directions. Cell Mol Exerc Physiol 1(1):e1
Ahmetov II, Druzhevskaya AM, Astratenkova IV, Popov DV, Vinogradova OL, Rogozkin VA
(2010) The ACTN3 R577X polymorphism in Russian endurance athletes. Br J Sports Med
44(9):649–652
Alfred T, Ben-Shlomo Y, Cooper R, Hardy R, Cooper C, Deary IJ, Gunnell D, Harris SE,
Kumari M, Martin RM, Moran CN, Pitsiladis YP, Ring SM, Sayer AA, Smith GD, Starr JM,
Kuh D, Day IN (2011) ACTN3 genotype, athletic status, and life course physical capability:
meta-analysis of the published literature and findings from nine studies. Hum Mutat 32(9):
1008–1018
Anderson S, Bankier AT, Barrell BG, de Bruijn MH, Coulson AR, Drouin J, Eperon IC, Nierlich
DP, Roe BA, Sanger F, Schreier PH, Smith AJ, Staden R, Young IG (1981) Sequence and
organization of the human mitochondrial genome. Nature 290(5806):457–465
Arden NK, Spector TD (1997) Genetic influences on muscle strength, lean body mass, and bone
mineral density: a twin study. J Bone Miner Res 12(12):2076–2081
Bouchard C, Lesage R, Lortie G, Simoneau JA, Hamel P, Boulay MR, Perusse L, Theriault G,
Leblanc C (1986) Aerobic performance in brothers, dizygotic and monozygotic twins. Med Sci
Sports Exerc 18(6):639–646
Bouchard C, Daw EW, Rice T, Perusse L, Gagnon J, Province MA, Leon AS, Rao DC, Skinner JS,
Wilmore JH (1998) Familial resemblance for VO2max in the sedentary state: the HERITAGE
family study. Med Sci Sports Exerc 30(2):252–258
Bray MS, Hagberg JM, Perusse L, Rankinen T, Roth SM, Wolfarth B, Bouchard C (2009) The
human gene map for performance and health-related fitness phenotypes: the 2006–2007
update. Med Sci Sports Exerc 41(1):35–73
Brutsaert TD, Parra EJ (2006) What makes a champion? Explaining variation in human athletic
performance. Respir Physiol Neurobiol 151(2–3):109–123
Castro MG, Terrados N, Reguero JR, Alvarez V, Coto E (2007) Mitochondrial haplogroup T is
negatively associated with the status of elite endurance athlete. Mitochondrion 7(5):354–357
Chiu LL, Wu YF, Tang MT, Yu HC, Hsieh LL, Hsieh SS (2011) ACTN3 genotype and swimming
performance in Taiwan. Int J Sports Med 32(6):476–480
Danser AH, Schalekamp MA, Bax WA, van den Brink AM, Saxena PR, Riegger GA, Schunkert H
(1995) Angiotensin-converting enzyme in the human heart. Effect of the deletion/insertion
polymorphism. Circulation 92(6):1387–1388
De Moor MH, Spector TD, Cherkas LF, Falchi M, Hottenga JJ, Boomsma DI, De Geus EJ (2007)
Genome-wide linkage scan for athlete status in 700 British female DZ twin pairs. Twin Res
Hum Genet 10(6):812–820
Deason M, Scott R, Irwin L, Macaulay V, Fuku N, Tanaka M, Irving R, Charlton V, Morrison E,
Austin K, Pitsiladis YP (2012) Importance of mitochondrial haplotypes and maternal lineage
in sprint performance among individuals of West African ancestry. Scand J Med Sci Sports
22(2):217–223
Druzhevskaya AM, Ahmetov II, Astratenkova IV, Rogozkin VA (2008) Association of the
ACTN3 R577X polymorphism with power athlete status in Russians. Eur J Appl Physiol
103(6):631–634
Elson JL, Turnbull DM, Howell N (2004) Comparative genomics and the evolution of human
mitochondrial DNA: assessing the effects of selection. Am J Hum Genet 74(2):229–238
Fuku N, Murakami H, Iemitsu M, Sanada K, Tanaka M, Miyachi M (2012) Mitochondrial

macrohaplogroup associated with muscle power in healthy adults. Int J Sports Med
33(5):410–414
Fuku N, Mikami E, Tanaka M (2013) Association of mitochondrial DNA polymorphisms and/or
haplogroups with elite Japanese athlete status. J Phys Fit Sports Med 2(1):17–29
Hwang S, Kwak SH, Bhak J, Kang HS, Lee YR, Koo BK, Park KS, Lee HK, Cho YM (2011) Gene
expression pattern in transmitochondrial cytoplasmic hybrid cells harboring type 2 diabetes-
associated mitochondrial DNA haplogroups. PLoS One 6(7):e22116
Kikuchi N, Ueda D, Min SK, Nakazato K, Igawa S (2013) The ACTN3 XX genotype’s under-
representation in Japanese elite wrestlers. Int J Sports Physiol Perform 8(1):57–61
Kikuchi N, Nakazato K, Min SK, Ueda D, Igawa S (2014) The Actn3 R577x polymorphism is
associated with muscle power in male Japanese athletes. J Strength Cond Res 28(7):1783–1789
Kim CH, Cho JY, Jeon JY, Koh YG, Kim YM, Kim HJ, Park M, Um HS, Kim C (2010) ACE DD
genotype is unfavorable to Korean short-term muscle power athletes. Int J Sports Med 31(1):
65–71
Kim KC, Cho HI, Kim W (2012) MtDNA haplogroups and elite Korean athlete status. Int J Sports
Med 33(1):76–80
Lesage R, Simoneau JA, Jobin J, Leblanc J, Bouchard C (1985) Familial resemblance in maximal
heart rate, blood lactate and aerobic power. Hum Hered 35(3):182–189
Lucia A, Gomez-Gallego F, Santiago C, Bandres F, Earnest C, Rabadan M, Alonso JM, Hoyos J,
Cordova A, Villa G, Foster C (2006) ACTN3 genotype in professional endurance cyclists. Int J
Sports Med 27(11):880–884
MacArthur DG, Seto JT, Raftery JM, Quinlan KG, Huttley GA, Hook JW, Lemckert FA, Kee AJ,
Edwards MR, Berman Y, Hardeman EC, Gunning PW, Easteal S, Yang N, North KN (2007)
Loss of ACTN3 gene function alters mouse muscle metabolism and shows evidence of positive
selection in humans. Nat Genet 39(10):1261–1265
MacArthur DG, Seto JT, Chan S, Quinlan KG, Raftery JM, Turner N, Nicholson MD, Kee AJ,
Hardeman EC, Gunning PW, Cooney GJ, Head SI, Yang N, North KN (2008) An Actn3
knockout mouse provides mechanistic insights into the association between alpha-actinin-3
deficiency and human athletic performance. Hum Mol Genet 17(8):1076–1086
Marcuello A, Martinez-Redondo D, Dahmani Y, Casajus JA, Ruiz-Pesini E, Montoya J, Lopez-
Perez MJ, Diez-Sanchez C (2009) Human mitochondrial variants influence on oxygen con-
sumption. Mitochondrion 9(1):27–30
Mikami E, Fuku N, Takahashi H, Ohiwa N, Scott R, Pitsiladis Y, Higuchi M, Kawahara T, Tanaka
M (2011) Mitochondrial haplogroups associated with elite Japanese athlete status. Br J Sports
Med 45(15):1179–1183
Mikami E, Fuku N, Kong QP, Takahashi H, Ohiwa N, Murakami H, Miyachi M, Higuchi M,
Tanaka M, Pitsiladis YP, Kawahara T (2013) Comprehensive analysis of common and rare
mitochondrial DNA variants in elite Japanese athletes: a case–control study. J Hum Genet
58(12):780–787
Mikami E, Fuku N, Murakami H, Tsuchie H, Takahashi H, Ohiwa N, Tanaka H, Pitsiladis YP,
Higuchi M, Miyachi M, Kawahara T, Tanaka M (2014) ACTN3 R577X genotype is associated
with sprinting in Elite Japanese athletes. Int J Sports Med 35(2):172–177
Mishmar D, Ruiz-Pesini E, Golik P, Macaulay V, Clark AG, Hosseini S, Brandon M, Easley K,
Chen E, Brown MD, Sukernik RI, Olckers A, Wallace DC (2003) Natural selection shaped
regional mtDNA variation in humans. Proc Natl Acad Sci U S A 100(1):171–176
Montgomery HE, Marshall R, Hemingway H, Myerson S, Clarkson P, Dollery C, Hayward M,
Holliman DE, Jubb M, World M, Thomas EL, Brynes AE, Saeed N, Barnard M, Bell JD,
Prasad K, Rayson M, Talmud PJ, Humphries SE (1998) Human gene for physical performance.
Nature 393(6682):221–222
Mustelin L, Latvala A, Pietilainen KH, Piirila P, Sovijarvi AR, Kujala UM, Rissanen A, Kaprio J
(2011) Associations between sports participation, cardiorespiratory fitness, and adiposity in
young adult twins. J Appl Physiol 110(3):681–686
Niemi AK, Majamaa K (2005) Mitochondrial DNA and ACTN3 genotypes in finnish elite
endurance and sprint athletes. Eur J Hum Genet 13(8):965–969
North KN, Beggs AH (1996) Deficiency of a skeletal muscle isoform of alpha-actinin (alpha-
actinin-3) in merosin-positive congenital muscular dystrophy. Neuromuscul Disord 6(4):
229–235
Okura T, Koda M, Ando F, Niino N, Tanaka M, Shimokata H (2003) Association of the
mitochondrial DNA 15497G/A polymorphism with obesity in a middle-aged and elderly
Japanese population. Hum Genet 113(5):432–436
Papadimitriou ID, Papadopoulos C, Kouvatsi A, Triantaphyllidis C (2008) The ACTN3 gene in
elite Greek track and field athletes. Int J Sports Med 29(4):352–355
Perusse L, Rankinen T, Hagberg JM, Loos RJ, Roth SM, Sarzynski MA, Wolfarth B, Bouchard C
(2013) Advances in exercise, fitness, and performance genomics in 2012. Med Sci Sports Exerc
45(5):824–831
Rigat B, Hubert C, Alhenc-Gelas F, Cambien F, Corvol P, Soubrier F (1990) An insertion/deletion
polymorphism in the angiotensin I-converting enzyme gene accounting for half the variance of
serum enzyme levels. J Clin Invest 86(4):1343–1346
Roth SM, Walsh S, Liu D, Metter EJ, Ferrucci L, Hurley BF (2008) The ACTN3 R577X nonsense
allele is under-represented in elite-level strength athletes. Eur J Hum Genet 16(3):391–394
Ruiz-Pesini E, Wallace DC (2006) Evidence for adaptive selection acting on the tRNA and rRNA
genes of human mitochondrial DNA. Hum Mutat 27(11):1072–1081
Ruiz-Pesini E, Mishmar D, Brandon M, Procaccio V, Wallace DC (2004) Effects of purifying and
adaptive selection on regional variation in human mtDNA. Science 303(5655):223–226
Santiago C, Rodriguez-Romo G, Gomez-Gallego F, Gonzalez-Freire M, Yvert T, Verde Z,
Naclerio F, Altmae S, Esteve-Lanao J, Ruiz JR, Lucia A (2010) Is there an association between
ACTN3 R577X polymorphism and muscle power phenotypes in young, non-athletic adults?
Scand J Med Sci Sports 20(5):771–778
Scott RA, Wilson RH, Goodwin WH, Moran CN, Georgiades E, Wolde B, Pitsiladis YP (2005)
Mitochondrial DNA lineages of elite Ethiopian athletes. Comp Biochem Physiol B Biochem
Mol Biol 140(3):497–503
Scott RA, Fuku N, Onywera VO, Boit M, Wilson RH, Tanaka M, W HG, Pitsiladis YP (2009)
Mitochondrial haplogroups associated with elite Kenyan athlete status. Med Sci Sports Exerc
41(1):123–128
Shang X, Zhang F, Zhang L, Huang C (2012) ACTN3 R577X polymorphism and performance
phenotypes in young Chinese male soldiers. J Sports Sci 30(3):255–260
Silventoinen K, Magnusson PK, Tynelius P, Kaprio J, Rasmussen F (2008) Heritability of body
size and muscle strength in young adulthood: a study of one million Swedish men. Genet
Epidemiol 32(4):341–349
Tanaka M, Cabrera VM, Gonzalez AM, Larruga JM, Takeyasu T, Fuku N, Guo LJ, Hirose R,
Fujita Y, Kurata M, Shinoda K, Umetsu K, Yamada Y, Oshida Y, Sato Y, Hattori N, Mizuno Y,
Arai Y, Hirose N, Ohta S, Ogawa O, Tanaka Y, Kawamori R, Shamoto-Nagai M,
Maruyama W, Shimokata H, Suzuki R, Shimodaira H (2004) Mitochondrial genome variation
in eastern Asia and the peopling of Japan. Genome Res 14(10A):1832–1850
Tiainen K, Sipila S, Alen M, Heikkinen E, Kaprio J, Koskenvuo M, Tolvanen A, Pajala S,
Rantanen T (2005) Shared genetic and environmental effects on strength and power in older
female twins. Med Sci Sports Exerc 37(1):72–78
Tobina T, Michishita R, Yamasawa F, Zhang B, Sasaki H, Tanaka H, Saku K, Kiyonaga A (2010)
Association between the angiotensin I-converting enzyme gene insertion/deletion polymor-
phism and endurance running speed in Japanese runners. J Physiol Sci 60(5):325–330
Torroni A, Richards M, Macaulay V, Forster P, Villems R, Norby S, Savontaus ML, Huoponen K,
Scozzari R, Bandelt HJ (2000) mtDNA haplogroups and frequency patterns in Europe. Am J
Hum Genet 66(3):1173–1177
Tucker R, Collins M (2012) What makes champions? A review of the relative contribution of
genes and training to sporting success. Br J Sports Med 46(8):555–561
Vincent B, De Bock K, Ramaekers M, Van den Eede E, Van Leemputte M, Hespel P, Thomis MA
(2007) ACTN3 (R577X) genotype is associated with fiber type distribution. Physiol Genomics
32(1):58–63
Wallace DC (2005) A mitochondrial paradigm of metabolic and degenerative diseases, aging, and
cancer: a dawn for evolutionary medicine. Annu Rev Genet 39:359–407
Wang G, Mikami E, Chiu LL, Perini ADE, Deason M, Fuku N, Miyachi M, Kaneoka K,
Murakami H, Tanaka M, Hsieh LL, Hsieh SS, Caporossi D, Pigozzi F, Hilley A, Lee R,
Galloway SD, Gulbin J, Rogozkin VA, Ahmetov II, Yang N, North KN, Ploutarhos S,
Montgomery HE, Bailey ME, Pitsiladis YP (2013) Association analysis of ACE and ACTN3
in elite Caucasian and East Asian swimmers. Med Sci Sports Exerc 45(5):892–900
Woods DR, Hickman M, Jamshidi Y, Brull D, Vassiliou V (2001) Elite swimmers and D allele of
the ACE I/D polymorphism. Hum Genet 108(3):230–232
Yang N, MacArthur DG, Gulbin JP, Hahn AG, Beggs AH, Easteal S, North K (2003) ACTN3
genotype is associated with human elite athletic performance. Am J Hum Genet 73(3):627–631
Chapter 11
Resting Energy Expenditure in Japanese
Athletes-as Applied to Dietary Management
for Athletes-
Motoko Taguchi and Satomi Oshima
Abstract Resting energy expenditure (REE) is known to be influenced by different

factors, such as body size, body composition (including mass of internal organs and
tissues), thyroid hormones, and menstrual cycle. However, the relationship between
these factors and REE has not been fully investigated nor reviewed in a way that is
optimally useful for the athletic population. Thus the purpose of the chapter is to
introduce and summarize studies which have investigated these relationships in
Japanese athletes. Fat-free mass (FFM) constitutes not just skeletal muscle, but also
internal organs with high metabolic rates. As FFM becomes larger, these tissues and
organs also become larger. Since athletes typically have a large FFM, FFM is the
major determinant of REE in athletes and consequently it is strongly recommended
that FFM be used to estimate REE for Japanese athletes. The daily energy require-
ment can be estimated based on the REE, and therefore REE is an important part of
a sound dietary management program for maximizing athletic performance. Nev-
ertheless, further studies are necessary to investigate such poorly understood factors
as genetic background and brown adipose tissue, as well as how these and other
elements influence the REE of Japanese athletes.
Keywords Resting energy expenditure • Japanese athletes • Body composition •

Dietary management
11.1 Introduction
The components of total energy expenditure are generally divided into three main
categories; resting energy expenditure (REE), thermic effect of food (TEF), and
thermic effect of physical activity (TEA). In the non-athletic population, REE
accounts for 60–75 % of the total daily energy expenditure (Elia 1992).
In 1991, Myerson et al. found that the REE of female athletes was significantly
lower in amenorrheic runners than in eumenorrheic runners. This was the first study
M. Taguchi (*) • S. Oshima

e-mail: mtaguchi@waseda.jp

126
Table 11.1 Reference of resting energy expenditure (REE) study in Japanese athletes
Gas REE
Primary collection (kcal/kgBW/ (kcal/kgFFM/
author Year Subjects, number Sex Weight (kg) FFM (kg) device (kcal/day) day) day)
Taguchi M 2001 16 runners Female 48.2 4.1 40.1 2.8 Douglas 1,246 121 26.0 3.1 29.1 3.8
bag
8 rowers Female 57.0 5.6 46.5 3.5 Douglas 1,351 170 23.9 3.9 30.6 3.5
bag
Takahashi 2008 42 athletes in various Female 58.9 5.8 47.2 4.2 Douglas 1,311 138 22.3 2.1 27.8 2.5
E sports bag
Taguchi M 2010 21 endurance athletes Female 54.2 7.0 43.1 3.9 Douglas 1,214 161 22.5 3.0 28.0 3.5
bag
40 athletes in strength and Female 56.1 7.2 45.2 4.0 Douglas 1,238 184 22.2 3.0 27.4 3.4
sprint-type sports bag
20 ball-game athletes Female 61.0 8.0 48.3 4.7 Douglas 1,322 207 21.9 3.0 27.6 3.5
bag
Taguchi M 2011 93 athletes in various Female 57.0 9.2 45.4 6.2 Douglas 1,258 202 22.1 2.2 27.7 2.6
sports bag
Taguchi M 2011 122 athletes in various Female 57.2 7.9 45.4 5.2 Douglas 1,242 190 Not available Not available
sports bag
Oshima S 2011 42 American football and Male 78.4 11.5 67.0 8.1 Douglas 1,856 225 23.8 1.8 27.8 1.9
15 handball players bag
Tatsuta W 2012 29 elite athletes in various Male 74.7 12.8 64.9 9.3 Douglas 1,896 235 25.7 3.2 29.4 3.0
sports bag
Oshima S 2013 37 American football Male 81.2 11.3 67.7 7.4 Douglas 1,869 230 23.2 2.0 27.7 1.9
players bag
M. Taguchi and S. Oshima
11 Resting Energy Expenditure in Japanese Athletes -as Applied to. . . 127
which focused directly on energy balance and weight control in the athletic
population.
Table 11.1 presents a list of REE studies on Japanese athletes. Few studies have
been done in this area of research.
Currently, in the area of sports nutrition the equation being used to estimate
energy requirements is based on the REE. Therefore, the REE has a large impact on
the dietary management programs being used with athletes. In this chapter we
briefly review the research utilized to estimate the REE of Japanese athletes. We
will include some of our recent work in this area. Later in the chapter we employ
these findings to construct a dietary management program for athletes.
11.2 Measurement of REE
In order to accurately measure REE, a number of conditions need to be met. These

conditions are generally similar in different studies, but the details may differ to
some degree. The following are the conditions utilized in our studies of REE: The
subjects are instructed to fast and refrain from exercise for 12 h before measure-
ments are taken. Athletes are asked to minimize any exertion prior to the laboratory
visit in the early morning. REE is measured by indirect calorimetry using a Douglas
bag. Athletes remain awake in the supine position at a comfortable room temper-
ature (20 C–25 C). After a 30 min habituation period with the mask, two samples
of expired gas are collected in Douglas bag for 10 min each and the mean value of
the two samples is used for analysis (see Fig. 11.1). Oxygen and carbon dioxide
concentrations are analyzed using a gas analyzer, and the volume of expired gas is
determined using a dry gas volume meter. Then, REE is calculated using Weir’s
equation (Weir 1949). Time variability in the REE of the subjects is less than 5.0 %
as measured by the mean coefficient of variance (Compher et al. 2006).
11.3 Factors That Influence REE for Japanese Athletes
11.3.1 Body Size and Composition
In Japanese non-athletic subjects, Ganpule et al. (2007) reported that body weight
(BW) accounted for 65–75 % of the variation in metabolic rates. Fat-free mass
(FFM), however, is a more valid predictor of REE than BW because there was a
higher association between FFM and REE than the association between BW and
REE. The data of Ganpule et al. demonstrated that FFM accounted for 84–89 % of
the variation in sleeping metabolic rate. Such a value is higher than other studies
had previously reported (Ravussin and Bogardus 1990; Toubro et al. 1996; Weyer
et al. 1999; Ganpule et al. 2007).
128 M. Taguchi and S. Oshima
Fig. 11.1 Depiction of the measurement of REE utilizing Douglas bags as the subject rests on
Japanese-style bedding
On the other hand, in athletes, body size and body composition vary widely
among different types of sport and different weight classes. For example, for sports
in which the BW must be moved over a long distance, such as distance running, a
lean physique with small body size can provide a competitive advantage. On the
other hand, heavyweight judo wrestlers and shot putters gain a great deal of
competitive advantage through having a large frame with a higher volume of
FFM. However, the relationship between the REE of athletes over a wide range
of body size and composition has not been investigated thoroughly in previous
studies.
Therefore, we performed an investigation in this area and found a significant
correlation (r ¼ 0.93, p < 0.01) between FFM and REE of Japanese male (n ¼ 60)
and female (n ¼ 205) athletes as shown in Fig. 11.2. Taguchi et al. (2010) reported
that the most powerful predictor of REE in female athletes by a multiple-regression
analysis was FFM (45 % contribution) for different types of sports. Oshima
et al. (2013) also reported that 69.7 % of variance of REE of male athletes could
be explained by FFM (n ¼ 37).
3000
2500
y = 27.0x + 30.6
r = 0.93, p<0.01
2000
REE (kcal/day)
1500
male
1000 female
500
0
20 30 40 50 60 70 80 90 100
FFM (kg)
Fig. 11.2 The relationship between FFM and REE in Japanese athletes
In order to better understand the relationship between REE and body composi-
tion for athletes of different body size, we looked more closely at body composi-
tion, and focused on the organ-tissue level, and especially on FFM. In this study the
subjects were 57 male athletes and 81 female athletes with a wide distribution of
body size. Body composition was measured by dual energy X-ray absorptiometry
(DXA), and FFM was calculated based on BW and % body fat. The subjects were
classified into three groups according to their BW : small-size athletes (<mean
BW–0.5SD), medium-size athletes (mean BW 0.5SD), and large-size athletes
(>mean BW + 0.5SD). The bone mineral content (BMC) of the whole body and
appendicular lean soft tissue (ALST) were also measured by DXA. Organ tissue
mass was calculated using the prediction model of Usui et al. (2012). Bone mass
(BM) was calculated as BMC times 1.85 (Heymsfield et al. 1990). Adipose tissue
mass (AT) was calculated as fat mass (FM) times 1.18 (Snyder et al. 1975;
Heymsfield et al. 2002). Skeletal muscle mass (SM) was calculated using the
prediction model of Kim et al. (2002). Finally, residual mass (RM) was calculated
as the difference between BW and the sum of the calculated BM, AT, and SM. This
can be express as follows:
80
BM AT SM RM
70 4.9*
(7.0%
weight of organ tissue (kg)
60 4.1
18.5*
(7.2%)
(26.6%)
50 3.7*
(7.6%*) 14.1
(24.8%)
9.9*
40 (20.3%*)
27.1*
30 21.5 (39.3%)
19.3* (37.8%)
(40.3%*)
20
15.3* 17.1 18.7*

10 (30.2%)
(31.9%*) (27.1%* )
0
S M L
RM, residual mass; SM, skeletal muscle; AT, adipose tissue; BM, bone mass. Significance was determined by
one-way ANOVA. * significantly different vs M (P<0.05), † significantly different vs S (P<0.05)
Fig. 11.3 Four organ tissue compartments expressed as weight and as their respective fractional
contributions to body weight
BMðkgÞ ¼ BMCðgÞ 1:85=1000,

AT ðkgÞ ¼ FMðkgÞ 1:18,
SMðkgÞ ¼ 1:13 ALST ðkgÞ 0:02 ageðyearsÞ þ 0:97,
RMðkgÞ ¼ BW ðBM þ AT þ SMÞ
The absolute and relative values of the four organ-tissue level compartments
increased significantly in accordance with body size as shown in Fig. 11.3. As
expected, absolute REE (kcal/day) increased significantly in accordance with body
size. However, if REE was adjusted by FFM (kcal/kg FFM/day), the REE/FFM was
not different regardless of BW.
FFM is not a single homogenous metabolic compartment (Gallagher et al. 1998;
Illner et al. 2000). RM and SM are high metabolic components. The metabolic rates
of these organs are 54 kcal/kg and 13 kcal/kg, respectively. Whereas, BM and AT
are low metabolic components (2.3 and 4.5 kcal/kg/day, respectively). Estimated
REE (REEe) from the specific metabolic rates of four major organ-tissue level
compositions was compared with measured REE (REEm). There was good agree-
ment between REEe and REEm (r ¼ 0.77, p < 0.01).
Taken together, these results, in athletes, show that REE increases in accordance
with increases in body size, and that there is a majorly influence via increases in
FFM. This increase seems not to be due to an elevation of organ tissue metabolic
rate. Together, the sum of skeletal muscle and internal organ volumes accounted for
over two thirds of the FFM in athletes and thus FFM is a major contributor to REE
for the athletic population.
11.3.2 Organ Tissue Mass
The FFM includes a number of different internal organs such as the brain, heart,
liver, and kidneys. Metabolic rate of these organs are 240, 440, 200, and 440 kcal/
kg/day, respectively (Elia 1992) and have exceptionally high metabolic rates when
compared to the other tissues. Using the magnetic resonance imaging (MRI)
technique, Midorikawa et al. (2007) reported that the metabolic rates per unit of
organs and tissue mass seem to be similar when comparing across Sumo wrestlers,
with a large organ tissue mass, and sedentary normal weight controls.
We also investigated the mass of high metabolic rate organs and their contribu-
tion to REE in athletes with different FFM levels. Thirty-seven male American
football players who participated in an intentional body weight gain program while
on a National Collegiate Athletic Association Division 1 team participated in the
study. The volumes of liver, brain, and kidneys were measured using MRI. Left
ventricular mass was measured using echocardiography. The results indicated that
the average relative contribution to REE from each tissue and organ was: skeletal
muscle 24.8 1.8 %, liver 18.7 2.1 %, brain 18.1 2.1 %, heart 7.3 1.2 %, and
kidneys 9.4 1.8 %. The average percentage of the sum of liver, brain, heart, and
kidneys to total REE was 53.5 3.9 %. The relative contributions of liver, heart and
kidney to REE were constant regardless of the FFM (Fig. 11.4) . On the other hand,
the relative rate of brain contribution to the measured REE became significantly
smaller when the FFM became larger (r ¼ 0.672, p < 0.001). Therefore, the
consistency of the REE/FFM ratio regardless of the FFM was suggested to be
largely due to the steady relative contribution of most of the internal organs on
REE.
11.3.3 Triiodothyronine (T3) Level
Triiodothyronine, also known as T3, is a thyroid hormone. It affects almost every

physiological process in the body, including growth and development, metabolism,
body temperature, and heart rate (Kim 2008). In our study, the T3 concentration
explained 8 % and 5.3 % of the variance of REE in male and female athletes,
respectively. T3 concentration was correlated moderately with REE, even when the
T3 plasma concentrations of the subjects were within the normal range.
11.3.4 Types of Exercise
It is well documented that there are a number of ways that exercise might indirectly
or directly influence REE in sedentary subjects. It is well documented that REE
increases for a period of time after strenuous exercise (Manore and Thompson
35
Skeletal muscle
REE from organ tissue mass / total REEm (%)
Liver
30
Brain
Kidneys
25
Heart
20
Brain
15 y = -0.183x + 30.422
r = -0.672, p<0.001
10
0
40 50 60 70 80 90 100
FFM (kg)
Fig. 11.4 The relative contribution of organ-tissue to REE
2000). However, athletes participate in training exercises almost every day for over
long periods of time. To clarify the effect of different types of sport, the REE of
female athletes (n ¼ 81) was compared across different types of sports. The athletes
were classified into three groups according to the type of sport (endurance, n ¼ 21;
power and strength, n ¼ 40; and ball games, n ¼ 20). The REE did not significantly
differ among the groups according to any assessment of metabolic energy. Thus, the
slopes and intercepts of the regression lines between REE and FFM were not
significantly different among the three groups. Therefore, it seems that types of
sport does not have a major impact on the REE of female Japanese athletes
(Taguchi et al. 2010).
11.3.5 Menstrual Cycle for Female Athletes
The difference between the REE during the follicular and luteal phases of
non-athletic females is about 100–300 kcal/day (Solomon et al. 1982; Webb
1986). In the athletic population, female athletes suffering with menstrual disorders
are quite common. Moreover, several studies report that the REE of amenorrheic
athletes is lower than that of eumenorrheic athletes(Myerson et al. 1991; Lebenstedt
et al. 1999).
In order to clarify the influence of menstrual cycles on REE, we evaluated

female collegiate athletes who had normal menstrual cycles. The menstrual cycles
of each subject were identified using an ovulation test kit (Do-test LH, ROHTO
Pharmaceutical Co., Ltd., Osaka, Japan) at basal body temperature for 4–5 months
over 3 successive cycles This information also aided in scheduling the experimental
tests necessary for each respective phase of the cycle. The follicular phase was
taken as 7–10 days before ovulation and the luteal phase as 4–9 days after
ovulation. In all subjects, measurements of REE and blood tests were taken at
each phase through one complete menstrual cycle. The results indicated that the
REE of each phase was not significantly different in terms of absolute value (kcal/
day), nor the REE adjusted by body weight or FFM. Progesterone (P4) is known to
influence luteinizing hormone pulsatility and increase body heat production. In this
study, we noted a low P4 concentration. This may have been caused by the
gynecological age (time since menarche) of the subjects, which was less than
10 years. The low P4 concentration may have been the cause of the unchanged
REE between the follicular and luteal phases. Therefore, the menstrual cycle phase
does not seems to influence the REE of Japanese female collegiate athletes
(unpublished data). However, further studies on the influence of menstrual cycles
for older female athletes needs to be investigated.
11.4 Equation for Estimating the REE of Japanese Athletes
Many equations have been developed for estimating the REE based on anthropo-
metric measurements, age, and sex (Cunningham 1980; Owen 1988; Ganpule
et al. 2007). These estimation equations can be helpful since it is not always
possible to measure REE. The Japanese National institute of Health and Nutrition
developed an equation for estimating REE of Japanese persons (Ganpule
et al. 2007). The equation was based on a body composition (FFM and fat mass)
which differed from the traditional equation, which uses body weight instead of
FFM. However, the accuracy and applicability of this equation to athletes needs to
be evaluated. Due to the difference in body composition between the non-athletic
population and athletes, which includes a greater FFM for the athletes, the Japan
Institute of Sports Sciences (JISS) presented a new equation for athletes. This
equation uses FFM and the invariant value of 28.5 kcal/kg FFM/day to estimate
REE (Koshimizu et al. 2005). Unfortunately, this equation was not derived using
actual measurements of REE for athletes.
Therefore, we examined the accuracy of several prediction equations of REE for
athletes. The study population consisted of 122 Japanese female athletes (age
20.2 1.3 years, height 162.5 6.3 cm, BW 57.4 7.7 kg, and FFM 45.5 5.1 kg)
(Taguchi et al. 2010). Measured REE (REEm) was assessed by indirect calorimetry
using Douglas bags, and predicted REE (REEp) was calculated from different
equations based on the athletes’ FFM. REEp from the equation of Taguchi
et al. {REE (kcal/day) ¼ 26.9 (kcal/kg FFM/day) x FFM (kg) + 36} and the equation
Table 11.2 Measured and predicted REE by each equations in Japanese female athletes
Correlation Paired
REE coefficient t-test TE
p p
Estimation equation (kcal/day) r value value (kcal)
Measured 1,242 190 — — — —
Predicted
JISS (2006) REE ¼ 28.5 FFM (kg) 1,296 146 0.711 0.001 0.001 112
Taguchi REE ¼ 26.9 FFM (kg) 1,259 138 0.711 0.001 0.160 106
et al. (2010) + 36
NIHN (2007) REE ¼ 0.0787 FFM (kg)- 1,294 110 0.700 0.001 0.001 114
0.0109 AGE
+ 0.0268 FM(kg)
0.3314 SEX + 2.3958
Cunningham REE ¼ 500 + 22 FFM (kg) 1,501 113 0.711 0.001 0.001 259
(1980)
Owen REE ¼ 23.6 FFM (kg) 1,259 121 0.711 0.001 0.161 106
et al. (1988)* + 186
REE desclived mean SD, FFM fat-free mass, FM fat mass, SEX male ¼ 1, female ¼ 2, SEE
Standard error of estimation, TE total error, JISS Japan Institute of Sports Sciences, NIHN National
Institute of Health and Nutrition
*Energy conversion factor: 1 kJ ¼ 4.184 kcal
of Owen et al. were not significantly different when compared to REEm, (Owen
1988) (Table 11.2). However, predicted values from the equation of JISS, National
Institute of Health and Nutrition, and Cunningham were significantly different from
the measured value (Cunningham 1980; Koshimizu et al. 2005; Ganpule
et al. 2007). Estimation error and total error were smaller in the equation of Taguchi
et al. and Owen et al. However, the equation of Owen et al. had a larger estimation
error for subjects who had a body weight less than 42 kg or larger than 69 kg.
Furthermore, the newer equation of Taguchi et al. {REE (kcal/day) ¼ 27.5 (kcal/kg
FFM/day) FFM + 5} was developed from a larger pooled data set (n ¼ 205) of
female athletes. The standard error of estimation (SEE) of this equation was smaller
than the previous equation of Taguchi et al., and systematic errors were minimal.
Therefore, the newer equation of Taguchi et al. developed from the measured REE
of Japanese female athletes is useful tool for predicting REE (Taguchi et al. 2011).
We also investigated the applicability of the same equation from Taguchi et al. for
male athletes and found that the same equation can be used for both sexes to predict
REE for Japanese athletes (unpublished data).
11.5 Applying REE to Dietary Management
Energy balance is of primary concern for most athletes (Nattiv et al. 2007). The
advantageous effects of training can be reduced when energy intake is insufficient
to match that expended since both body fat and body protein can be used for energy.
Table 11.3 Physical activity level (PAL) categories depending on characteristics and seasons of
sports by JISS
Periodization
Categoly Off-training season On-training season
Endurance sports 1.75 2.50
Strength and power sports 1.75 2.00
Ball-game sports 1.75 2.00
Other 1.50 1.75
In addition, if energy intake is limited or restricted, intake levels of other essential

nutrients necessary for optimal sport performance will be compromised. Such
conditions may leads to an anemia, low bone mineral density, eating disorders, or
menstrual disorders (Loucks et al. 1998; Nattiv et al. 2007). Therefore, it is essential
to properly manage energy intake and match it to the energy requirements if athletes
are to attain an optimal performance.
For suitable nutrition management, estimated energy requirements should be
calculated using an equation based on REE and physical activity level (PAL)
according to the Dietary Reference Intakes for Japanese (Ministry of Health Labour
and Welfare 2010). The equation to estimate energy requirement (EER) for athletes
based on REE is as follows:
Estimated energy requirement ðkcal=dayÞ

¼ REE ðkcal=kg FFMÞ fat free mass ðkgÞ physical activity level
The physical activity level for athletes is higher than that of the non-athletic
population, and thus, accurately estimating REE is critical if one is to correctly
predict the energy requirement. This is a significant matter if athletes are to reach
their maximal potential. JISS has estimated PAL categories according to each
sport’s characteristics and the seasons in which it is played. These estimations are
shown in Table 11.3. Below is an example depicting the EER for primary dietary
management of a specific athlete.
Male, 22 years old, Event: football
Height: 175 cm, Bodyweight: 72 kg, Body fat: 10 %, FFM: 64.8 kg
REEðkcal=dayÞ ¼ 27:5 kcal=kg FFM=day 64:8 kg ¼ 1, 780 kcal=day
EERðkcal=dayÞ ¼ 1, 780 kcal=day 2:00 ¼ 3, 560 kcal=day
In conclusion, the most powerful predictor of REE in Japanese athletes was FFM.
Thus, FFM is a useful measure to utilize for dietary management of athletes.
Further study investigating the influences of genetic factors, brown adipose tissue,
and energy balance on REE will provide yet another collection of valuable data and
supplement what is currently available to guide dietary management for athletes.
References
Compher C, Frankenfield D, Keim N, Roth-Yousey L, Grp EAW (2006) Best practice methods to
apply to measurement of resting metabolic rate in adults: a systematic review. J Am Diet Assoc
106(6):881–903. doi:10.1016/j.jada.2006.02.009
Cunningham JJ (1980) A reanalysis of the factors influencing basal metabolic rate in normal
adults. Am J Clin Nutr 33(11):2372–2374
Elia M (1992) Organ and tissue contribution to metabolic rate. In: Kinney J, Tucker H (eds)
Energy metabolism. Tissue determinants and cellular corollaries. Raven Press, New York, pp
61–80
Gallagher D, Belmonte D, Deurenberg P, Wang ZM, Krasnow N, Pi-Sunyer FX, Heymsfield SB
(1998) Organ-tissue mass measurement allows modeling of REE and metabolically active
tissue mass. Am J Physiol-Endocrinol Metab 275(2):E249–E258
Ganpule AA, Tanaka S, Ishikawa-Takata K, Tabata I (2007) Interindividual variability in sleeping
metabolic rate in Japanese subjects. Eur J Clin Nutr 61(11):1256–1261. doi:10.1038/sj.ejcn.
1602645
Heymsfield SB, Smith R, Aulet M, Bensen B, Lichtman S, Wang J, Pierson RN Jr (1990)
Appendicular skeletal muscle mass: measurement by dual-photon absorptiometry. Am J Clin
Nutr 52(2):214–218
Heymsfield SB, Gallagher D, Kotler DP, Wang ZM, Allison DB, Heshka S (2002) Body-size
dependence of resting energy expenditure can be attributed to nonenergetic homogeneity of
fat-free mass. Am J Physiol-Endocrinol Metab 282(1):E132–E138
Illner K, Brinkmann G, Heller M, Bosy-Westphal A, Muller MJ (2000) Metabolically active
components of fat free mass and resting energy expenditure in nonobese adults. Am J Physiol-
Endocrinol Metab 278(2):E308–E315
Kim B (2008) Thyroid hormone as a determinant of energy expenditure and the basal metabolic
rate. Thyroid : Off J Am Thyroid Assoc 18(2):141–144. doi:10.1089/thy.2007.0266
Kim J, Wang ZM, Heymsfield SB, Baumgartner RN, Gallagher D (2002) Total-body skeletal
muscle mass: estimation by a new dual-energy X-ray absorptiometry method. Am J Clin Nutr
76(2):378–383
Koshimizu T, Yanagisawa K, Higuchi M (2005) Estimating energy requirement for athletes. J
Train Sci Exerc Sport 17(4):245–250
Lebenstedt M, Platte P, Pirke KM (1999) Reduced resting metabolic rate in athletes with menstrual
disorders. Med Sci Sports Exerc 31(9):1250–1256
Loucks AB, Verdun M, Heath EM (1998) Low energy availability, not stress of exercise, alters LH
pulsatility in exercising women. J Appl Physiol (1985) 84(1):37–46
Manore M, Thompson J (2000) Energy requirements of the athlete: assessment and evidence of
energy efficiency. In: Burke L, Deakin V (eds) Clinical sports nutrition, 2nd edn. McGraw Hill,
Boston, pp 124–145
Midorikawa T, Kondo M, Beekley MD, Koizumi K, Abe T (2007) High REE in sumo wrestlers
attributed to large organ-tissue mass. Med Sci Sports Exerc 39(4):688–693. doi:10.1249/mss.
0b013e31802f58f6
Ministry of Health Labour and Welfare (2010) Dietary reference intakes for Japanese. Daiichi
Shupan Publishing, Tokyo
Myerson M, Gutin B, Warren MP, May MT, Contento I, Lee M, Pi-Sunyer FX, Pierson RN Jr,
Brooks-Gunn J (1991) Resting metabolic rate and energy balance in amenorrheic and
eumenorrheic runners. Med Sci Sports Exerc 23(1):15–22
Nattiv A, Loucks AB, Manore MM, Sanborn CF, Sundgot-Borgen J, Warren MP (2007) American
college of sports medicine position stand. The female athlete triad. Med Sci Sports Exerc
39(10):1867–1882. doi:10.1249/mss.0b013e318149f111
Oshima S, Miyauchi S, Asaka M, Kawano H, Taguchi M, Torii S, Higuchi M (2013) Relative
contribution of organs other than brain to resting energy expenditure is consistent among male
power athletes. J Nutr Sci Vitaminol 59(3):224–231
Owen OE (1988) Resting metabolic requirements of men and women. Mayo Clin proc Mayo Clin
63(5):503–510
Ravussin E, Bogardus C (1990) Energy expenditure in the obese: is there a thrifty gene?
Infusionstherapie 17(2):108–112
Snyder WS, Cook MJ, Nasset ES, Karhausen LR, Howells GP, Tipton IH (1975) Report of the task
group of reference Man. Pergamon Press, Oxford
Solomon SJ, Kurzer MS, Calloway DH (1982) Menstrual cycle and basal metabolic rate in
women. Am J Clin Nutr 36(4):611–616
Taguchi M, Tatsuta W, Higuchi M (2010) Resting energy expenditure of female athletes in
different types of sport. Jpn J Nutr Diet 68(5):289–297. doi:10.5264/eiyogakuzashi.68.289
Taguchi M, Ishikawa-Takata K, Tatsuta W, Katsuragi C, Usui C, Sakamoto S, Higuchi M (2011)
Resting energy expenditure can be assessed by Fat-free mass in female athletes regardless of
body size. J Nutr Sci Vitaminol 57(1):22–29
Toubro S, Sorensen TIA, Ronn B, Christensen NJ, Astrup A (1996) Twenty-four-hour energy
expenditure: The role of body composition, thyroid status, sympathetic activity, and family
membership. J Clin Endocrinol Metab 81(7):2670–2674. doi:10.1210/jc.81.7.2670
Usui C, Taguchi M, Ishikawa-Takata K, Higuchi M (2012) The validity of body composition
measurement using dual energy X-Ray absorptiometry for estimating resting energy expendi-
ture. In: Maghraoui AE (ed) Dual energy X-Ray absorptiometry. InTech, Rijeka, Croatia,
pp 45–58
Webb P (1986) 24-hour energy expenditure and the menstrual cycle. Am J Clin Nutr 44(5):
614–619
Weir JBD (1949) New methods for calculating metabolic rate with special reference to protein
metabolism. J Physiol-Lond 109(1–2):1–9
Weyer C, Snitker S, Rising R, Bogardus C, Ravussin E (1999) Determinants of energy expenditure
and fuel utilization in man: effects of body composition, age, sex, ethnicity and glucose
tolerance in 916 subjects. Int j obes relat metab disord: J Int Assoc Study Obes 23(7):715–722
Chapter 12
Health Issues and Preventive Strategies
for Heavy Athletes
Satomi Oshima and Motoko Taguchi
Abstract Some athletes are required to increase their body weight, since it directly
influences their performance. Rugby, Judo, American football, Weight lifting, Field
throwing, Bodybuilding, and Sumo are sports where heavy athletes are particularly
successful. In fact, for the majority of athletes in these sports, their body mass index
(BMI) exceeds 25 kg/m2. Unfortunately, since it is difficult to increase body weight
as fat-free mass (FFM), accumulation of unintentional body fat is often observed in
heavy athletes, especially among non-elite players. In fact, linemen in American
football have a high prevalence of metabolic syndrome and/or insulin resistance
(Borchers et al. 2009). Overfeeding and the ensuing creation of a positive energy
balance are essential for increasing body weight. However, overeating increases the
risk for developing visceral fat accumulation and becoming insulin resistant if done
without an adequate dietary plan. While developing health issues from overfeeding
is dependent upon one’s genetic complement to some degree, dietary composition
and food choices for overfeeding significantly influence the magnitude of visceral
fat accumulation as well as the extent of plasma insulin response. Screening and
monitoring which involve the periodic measure of body composition as well as
biochemical assessments may help to prevent cardiometabolic risks for heavy
athletes. These precautions will help ensure a sound physical condition for their
playing days and also for their health during the remainder of their lives (Haskins
et al. 2011).
Keywords Heavy athlete • Weight gain • Overfeeding • Body fat • Insulin

resistance
12.1 Introduction
When athletes commit to an intentional body weight gain through overfeeding and
resistance training, the goal for the majority of athletes is to increase strength and
power by increasing muscle mass (Berglund et al. 2011). However, it is difficult to
S. Oshima (*) • M. Taguchi

e-mail: satomioshima@fuji.waseda.jp

140 S. Oshima and M. Taguchi
limit increases in body weight to just muscle mass or fat-free mass (FFM). Thus, the
players often increase fat mass as well (Grandjean 1999). If there is a continuous
increase in fat mass, it may lead to an over accumulation of fat in the abdominal
region as visceral fat which is known to have a detrimental effect on cardiovascular
and metabolic health (Bjorntorp 1991). Unfortunately, there is a high prevalence of
overweight and obese American football players on college teams (Mathews and
Wagner 2008). Moreover, a consequence of inappropriate weight gain is serious
health problems. These include an increased risk of metabolic syndrome and a
higher incidence of cardiovascular problems that lead to a high mortality rate after
retirement. In this chapter we will confirm the prevalence of cardiometabolic issues
for heavy athletes, explain the risks of overfeeding, and to propose preventive
strategies for athletes desiring to gain weight.
12.2 Body Composition of Heavy Athletes
It is difficult to define what constitutes a heavy athlete, since different sports, and
positions within a sport, have dissimilar ideal body weights. For the purposes of this
chapter, we will define heavy athletes as those who have a body mass index (BMI)
of 25 kg/m2 or higher.
The linemen of American football largely perform short anaerobic exercise,
normally with a goal of stopping opponents from moving forward on the field.
Having a relatively large body mass generally aids in their performance. Unfortu-
nately, linemen are well known to have a high prevalence of obesity. The size of
American football players increases from generation to generation (Anzell
et al. 2013). In recent years, both offensive and defensive lineman of National
Collegiate Athletic Association Division 1 football players typically have a body
mass of over 110 kg, 30 kg/m2 for BMI, and 22 % for percent body fat (Mathews
and Wagner 2008). Among professional American football players of about a
decade ago (National Football League: NFL), the average offensive lineman
weighed 140 kg, had a BMI of 37.1 kg/m2, and had a fat content of 25.1 %. The
average defensive lineman weighed 126.8 kg, had a BMI of 34.6 kg/m2, and had a
fat content of 18.5 % body fat (Kraemer et al. 2005). The body weight of Japanese
division 1 university players of American football was significantly smaller than the
American division 1 university players for all positions (Iguchi et al. 2011). How-
ever, Iguchi et al. reported that offensive linemen still had a body weight of more
than 106 kg with a body fat content of 19.4 % (Iguchi et al. 2011). This is similar to
our finding of a 19.2 % body fat percentage in the offensive linemen that we
collected data from (unpublished data).
Body mass for Rugby players has also significantly increased over the last three
decades (Olds 2001). The forwards position is known to require a larger body size
than the backs position, and thus forwards players have a significantly higher
percentage of body fat as compared to that of the backs players. In 1998 New
South Wales Super 12 rugby players, the body weight of front row forwards was
12 Health Issues and Preventive Strategies for Heavy Athletes 141
112.8 5.7 kg and for the remainder of the forwards was 108 5.3 kg. On the other
hand, the backs weighed 89.0 6.8 kg (Deacres-Manning 1998). The body mass
tend to be heavier at higher levels of competition for forwards. On the other hand, at
the higher levels of play, body fat percentage decreases in both forwards and backs.
This indicates that while elite Rugby players are required to have a large body mass,
the constituents of the large mass are what differentiate elite vs non-elite players
(Duthie et al. 2003). Mean values for 11 college division 1 forward players in Japan
were: body weight 97.4 kg and body fat percentage 17 % (unpublished data).
Sumo wrestlers are famous for their extremely large body mass. According to
previous anthropometric reports (1898–1991) on 664 professional Sumo wrestlers
in Japan, the range of body weight was 71.0–262.0 kg (mean: 115.8 kg) and the
BMI range was 25.0–74.9 kg/m2 (mean: 36.6 kg/m2). Even collegiate Sumo
wrestlers are reported to have an average BMI of 40.0 kg/m2 (Yamauchi
et al. 2004).
Judo competitions are currently divided by seven weight categories for both
male (<60 kg, 66 kg, 73 kg, 81 kg, 90 kg, 100 kg and >100 kg) and female judo
competitors (<48 kg, 52 kg, 57 kg, 63 kg, 70 kg, 78 kg and >78 kg). World and
Olympic level elite male judo athletes usually have less than 10 % body fat except
for the heavy weight class (>78 kg for female, >100 kg for male) (Koury
et al. 2005; Franchini et al. 2011). The body fat percentage typically ranges from
about 20 % or less in female elite Judo athletes (Japanese Olympic Committee
2001; Callister et al. 1990, 1991; Little 1991; Sertic et al. 2006; Koury et al. 2007).
On the other hand, data on university heavy weight class Judo athletes for male
(+95 kg) and for female (+78 kg) in Japan indicate body fat percentages that ranged
between 18.5–42.1 % (Iida et al. 1997) and 21.3–38.9 % (Murata et al. 2013),
respectively. One study found a significant difference in body fat between the best
ranked judo athletes and lower ranked athletes (Callister et al. 1991). Thus heavy
collegiate athletes and/or such athletes competing at lower levels are more at risk
for an over accumulation of body fat.
In addition to American football linemen, forward players of Rugby, Sumo
wrestlers, Judo wrestlers in the heavy-weight categories, Wrestlers and Weight
lifters in heavy-weight categories, and Field throwers are all often found in the
heavy athlete category (Japanese Olympic Committee 2001).
12.3 Cardiometabolic Risk for Heavy Athletes
12.3.1 During Active Players
Whether it is intentional or unintentional, body fat tends to increase with weight

gain. Unfortunately, in many cases it leads to an excessive body fat accumulation
which increases the risk of cardiovascular and metabolic issues even for active
athletes. For example, the metabolic syndrome and cardiovascular disease risks
seen in American football players are held to be due to the high prevalence of
overweight and obesity in American football players at all competitive levels
(Mathews and Wagner 2008; Wilkerson et al. 2010). Borchers et al. (2009) reported
that 21 % of players overall and 68 % of the linemen had insulin resistance, and that
all of the linemen had metabolic syndrome. Another study focused only on linemen
of divisions I, II, and III in the national collegiate athletic association and found that
about 34 out of every 70 linemen were identified as having metabolic syndrome
(Buell et al. 2008). A study on 504 American National Football League (NFL)
players (including all positions) revealed that most cardiovascular risk factors were
within the normal ranges (Allen et al. 2010). However, NFL linemen still had a high
prevalence of metabolic syndrome compared to the general US population (Selden
et al. 2009) and they are particularly at high risk for hypertension (Allen et al. 2010;
Tucker et al. 2009). Similar heath issues for heavy athletes are also reported in
Japan and China. Even though the Japanese collegiate American football players
are smaller than their counterparts in the USA, our unpublished data showed a high
prevalence of insulin resistance (as assessed by homeostatic model assessment for
insulin resistance (HOMA-IR)) among the upper division college level players.
Additionally, Chinese young professional athletes that competed in the heavy-
weight classes of Weightlifting, Judo, and Wrestling, as well as heavy Track and
Field players, also demonstrated a significantly increased risk of cardiometabolic
disease as compared with those at all other weight categories (Guo et al. 2013).
Unlike heavy males, there are only a few studies focused on heavy female athletes.
Murata et al. (2013) reported that female Judo heavy weight class players had a high
prevalence (9 out of 12 players) of insulin resistance as evaluated by HOMA-IR.
Therefore, it seems that worldwide, not just heavy male athletes, but also heavy
female athletes face an increased risk of developing cardiometabolic issues.
12.3.2 After Retirement
Former elite athletes are generally less prone to develop cardiovascular and meta-
bolic disease after they retire from active sports participation (Batista and Soares
2013). Heavy athletes are protected to some degree against fat accumulation and
related cardiometabolic risk factors while they participate in habitual high intensity
exercise. However, what happens when such heavy athletes stop exercising after
retiring from their sport?
There is a report based on 3,420 retired NFL players from the National Institute
for Occupational Safety and Health in the USA. Baron and Rinsky (1994) report
that for offensive and defensive linemen, the risk of dying from heart disease is
52 % greater than that of the general population and three times higher than that for
players of other football positions. The same report showed that the largest body
size category (64 % of all linemen) had a six times greater risk of developing heart
disease as compared to those of normal size. This, high mortality rate for former
NFL linemen is due to the high prevalence of metabolic syndrome, which is
comprised of a higher risk for obesity, hyperlipidemia, and impaired glucose

metabolism (Miller et al. 2008). The mean body weight and percentage body fat
of retired NFL linemen (n ¼ 164) was 127.3 kg and 31.4 % compared to that of
104.0 kg with 27.4 % for retired NFL non-linemen (n ¼ 346). Additionally, an
abnormally large left ventricular mass and left arterial size due to the strength and
resistance training regime as well as having a large body size may partially explain
the high cardiovascular mortality rate for retired linemen (Abernethy et al. 2003).
Another study was based on 664 Sumo wrestlers who belonged to the top division
of professional Sumo during the period from 1898 to 1991 (Hoshi and Inaba 1995).
This study reported that life expectancy was shorter in Sumo wrestlers as compared
to the average Japanese male’s life expectancy. In addition, Sumo wrestlers also
had a higher risk for becoming diabetic (Sugimoto 1963) as well as a higher risk for
dying from cardiovascular disease (Hoshi and Inaba 1995).
As it has been made quite clear in the previous data, heavy athletes are threat-
ened by an over accumulation of body fat through weight gain. Cardiovascular and
metabolic issues become particularly severe when these athletes retire from their
sport and stop training. Since former heavy athletes tend to maintain a large body
weight, they have a high risk for developing cardiovascular or metabolic diseases
which significantly shorten their lives. In the future it is important to conduct a
longitudinal study with the goal of following body composition changes and
cardiometabolic risk factors of the heavy athletes after they retire. Such a study
would be invaluable in the effort to establish an effective program to ameliorate
their increased disease risk.
12.4 Influence of Overfeeding
Athletes who practice intentional weight gain of necessity have to create a positive
energy balance through overfeeding. Our data show that after 1 year of overfeeding
for intentional weight gain, nineteen 1st year Japanese collegiate players of Amer-
ican football increased their cross-sectional area of visceral fat from 33.1 cm2 to
70.5 cm2 on average (unpublished data) and one of the players actually tripled his
visceral fat cross-sectional area (Fig. 12.1). This makes clear the danger of
overfeeding.
While many studies have investigated the influence of overfeeding, for ethical
reasons most of these studies were based on a short-term overfeeding intervention.
Thus, there are few studies involving a long-term overfeeding intervention. In the
1960s and on through the 1980s, several long-term overfeeding studies were
conducted and revealed that body weight gain was associated with decreased
glucose tolerance, the development of hyperinsulinemia, and increased insulin
resistance (Sims et al. 1968; Olefsky et al. 1975; Welle et al. 1986). Later, in the
1990s, in order to investigate not only the influence of long-term overfeeding but
also to evaluate genetic influences, Bouchard et al. (1990) conducted an overfeed-
ing study on monozygotic twin pairs. In the experiment, 12 pairs of young adult
Before After 1 year
Before After 1 year

Body weight (kg) 80.0 95.4
Body fat (%) 15.1 20.2
FFM (kg) 67.9 76.2
Waist circumference (cm) 83.2 97.3
Subcutaneous fat area (cm2) 171.1 195.3
Visceral fat area (cm2) 40.2 124.4
Fig. 12.1 Example of abdominal image (umbilical region) of first year college American football
player before and after 1 year of intentional weight gain
male monozygotic twins were overfed by 1,000 kcal per day (50 % CHO, 35 % fat,
and 15 % protein), 6 days a week, for a total of 84 days during a 100-day period.
Significant individual differences were observed on the degree of body weight gain
(4.3 kg to 13.3 kg). There was roughly three times more variance across pairs than
within pairs (p < 0.05) with respect to body weight, percentage of fat, fat mass, and
subcutaneous fat. Moreover, the variance across pairs for the amount of abdominal
visceral fat was six times greater than within pairs. Additionally, the increase in
total body fat was correlated with gains in abdominal subcutaneous fat but not with
the changes in visceral fat. These findings suggest that a person’s genotype is an
important determinant of adaptation to a sustained energy surplus (Bouchard
et al. 1990). This same research group also reported (Oppert et al. 1995) that insulin
(but not glucose) secretion increased both at fasting and after 100 days of over-
feeding as measured by the oral-glucose tolerance test. Thus, it was suggested that
long-term overfeeding puts athletes at risk of becoming insulin resistant. Further-
more, persons who showed an increased insulin production after overfeeding also
gained more subcutaneous fat. However, the increased insulin production did not
influence the gain in visceral fat (Oppert et al. 1995). These investigators also found
that there is more variance between pairs than within pairs for the change in insulin
production during the fasting state after 100 days of overfeeding.
It seems that there are significant genetic influences on body composition and
metabolic changes induced by overfeeding. However, most of these studies on these
genetic effects were conducted on the same participant group, so the findings may
not apply to other ethnic group with different dietary habit and genetic back-
grounds. As one example, Japanese are known to reduce the function of insulin
production and develop diabetes at a lower BMI (23.1 kg/m2) than Caucasians (Ex:
29.4 kg/m2 in United Kingdom) (Sone et al. 2003). This difference may be caused
by either a different genetic background and/or a different life style. For example,
the typical Japanese diet involves more carbohydrate and less fat. Thus, it is quite
important to perform similar long-term investigations on persons of different
ethnicities and with different genetic backgrounds.
12.5 Preventive Measure of Cardiometabolic Risk

for Heavy Athletes
12.5.1 Diet
The Australian Institute of Sports has recommended increasing the nutritional

intake by 500–1,000 kcal per day for athletes who want to increase their muscle
mass (Australian Institute of Sport 2009). In general, the diet should be balanced,
not only with an adequate mix of protein, fat, and carbohydrate (PFC), but also
including the minerals and vitamins that are necessary for efficient body building
(Rodriguez et al. 2009). In reality however, it is always a big challenge to control
the energy balance. Nagasawa et al. (2013) conducted a weight gain program that
involved dietary intervention. They showed that with a proper diet, FFM increases
without significant increases in body fat mass are possible. Their study utilized
Japanese college rowers who, for 12 weeks, received a 1,000 kcal/day positive
energy supply with a nutritional component generally modeled according to the
guidelines of the American Dietetic Association, the Dietitians of Canada, and the
American College of Sports Medicine (Rodriguez et al. 2009). The subjects gained
3.8 kg (1.2 kg body fat and 2.6 kg lean body mass) with no change in fasted plasma
cholesterol, triglycerides, or glucose (Nagasawa et al. 2013). The subjects’ diet had
a PFC balance of 13.2 %, 28.8 %, and 58 %, which is similar to the ideal PFC
composition recommended for the general Japanese population (Ministry of Health,
Labour and Welfare of Japan 2012). Thus, even for athletes who are trying to build
muscle, the recommended PFC diet composition does not need to be different from
a typical diet. All of the meals were typical Japanese style home-made dishes
prepared daily by the dietitian. Moreover, the Japanese diet is characterized by a
wide variety of available foods and a versatility of cooking methods which makes it
relatively easy to provide the necessary energy and nutrients sources, and match
them with each individual athlete’s needs and preferences.
Food choice has an influence on the metabolic health of athletes practicing
overfeeding. One study overfed subjects by 850–900 kcal with either candy or
peanuts for 2 weeks. The study found that only the candy group had a significant
increase in body weight; this increase was mostly due to fat accumulation. The
candy group also developed an unfavorable plasma lipid profile and showed signs
of insulin resistance. These untoward effects were not seen in the peanuts group
(Claesson et al. 2009). Candy is largely composed of sucrose, a short-chain

carbohydrate. It is thus easily digested and absorbed. On the other hand, while
peanuts are high in protein and fat, the fat is “good fat”, that is, mostly monoun-
saturated and polyunsaturated. Therefore, the food choices of athletes during
overfeeding have a major effect on which body compartments gain weight, and
how the weight gain affects metabolic health. Therefore, in order to achieve an
efficient and healthful increase in body weight, it is important to meet with a
certified dietitian to seek professional advice on a proper diet for overfeeding.
Utilizing a dietary assessment with a dietary record is an ideal tool to evaluate
dietary content and dietary habits. This can provide information from which
athletes can receive accurate and effective advice that meets their needs.
12.5.2 Body Composition Assessment
Regular monitoring of body composition is also very important for successful

weight gain. Such monitoring can help the athlete and supervisory personnel decide
whether the current weight gain regimen is adequate. Body mass in relation to
height is the most basic anthropometrical measurement, and can easily be used in
the field. Body mass index (BMI), calculated by dividing an individual’s body mass
(in kg) by the square of their height (in m2), is used to assess weight status for
overweight and obesity in routine situations. However, BMI cannot distinguish
between mass of body fat or muscle mass and thus is not appropriate for use with
athletes in order to evaluate their body fat status (Nevill et al. 2006). Calculating fat
mass based on body fat percentage is a more suitable method for judging how fat an
athlete’s is. Measurements of FFM allow an accurate assessment of muscle mass
changes. Bioelectric impedance analysis (BIA) and skin-fold thickness can both be
used to measure body fat percentage. These two methods do not require expensive
equipment and are relatively easy to utilize. Therefore, they are suitable for
monitoring daily, weekly, or monthly body fat percentage changes in the field. A
BIA device with a tetrapolar or more electrode arrangement has a better accuracy
than does a BIA device with only a bipolar electrode arrangement (Bosy-Westphal
et al. 2008). For the more detailed or precise measurements, dual energy x-ray
absorptiometry (DXA), densitometry with air or water, and ultrasound are
recommended if these are available (Ackland et al. 2012). These latter methods
require a testing facility with elaborate testing equipment, along with trained
technicians and staff members. Thus, these methods may only be available a few
times a year. In order to quantify visceral fat, MRI and CT scan are ideal methods to
use, but these methods are only available at sophisticated laboratory facilities and
thus not generally accessible to athletes. Therefore, estimation equations of visceral
fat by abdominal circumference or DXA are also available (Kaul et al. 2012;
Rankinen et al. 1999)
12.5.3 Biochemical Assessment
Even though it may be difficult to perform appropriate biochemical tests, perhaps

due to budgetary limitations, it is nevertheless highly recommended for high risk
athletes such as for athletes working on extreme weight gains and/or those who
need to maintain a large body mass (ex. body mass of >100 kg or BMI of 30 kg/m2
or above). Such biochemical assessments can help to screen athletes at risk, detect
athletes with improper dietary habits, and prevent irreversible cardiometabolic
illnesses or events (Haskins et al. 2011). The most important indicators in a lipid
profile are levels of triglyceride, total cholesterol, low-density lipoprotein choles-
terol, and high-density lipoprotein cholesterol. Insulin resistance, which tends to
occur during the early stages of metabolic disease, can be assessed by fasting blood
glucose and insulin via the HOMA-IR.
12.6 Conclusion
Due to a long period of overfeeding, many heavy athletes are at risk for body fat
accumulation. This can lead to the development of insulin resistance and an
increased risk of cardiovascular and metabolic disease. Moreover, when heavy
athletes retire and stop training, they quickly join the high cardiometabolic risk
population unless they lose weight. In fact, Sumo wrestlers and the linemen in
American football have a shorter longevity than the general population. Therefore,
it is very important for coaches and trainers to understand these health risks and the
consequences of overfeeding, and importantly, the proper approach for increasing
body weight. Diet composition is extremely important, and can determine whether
weight gain is healthy or detrimental to health. Thus, seeking advice from certified
dietitians and conducting careful dietary assessments are beneficial. Monitoring
body weight and composition are critical to achieving positive health and strength
in body building. Precise body compositional assessment and biochemical tests can
help detect unhealthy weight gain and thus avoid the development of irreversible
cardiovascular and metabolic maladies.
References
Abernethy WB, Choo JK, Hutter AM (2003) Echocardiographic characteristics of professional

football players. J Am Coll Cardiol 41(2):280–284. doi:10.1016/s0735-1097(02)02633-5
Ackland TR, Lohman TG, Sundgot-Borgen J, Maughan RJ, Meyer NL, Stewart AD, Muller W
(2012) Current status of body composition assessment in sport review and position statement
on behalf of the ad hoc research working group on body composition health and performance,
under the auspices of the I.O.C. Medical Commission. Sports Med 42(3):227–249
Allen TW, Vogel RA, Lincoln AE, Dunn RE, Tucker AM, Natl Football League Subcomm C
(2010) Body size, body composition, and cardiovascular disease risk factors in NFL players.
Phys Sportsmed 38(1):21–27
Anzell AR, Potteiger JA, Kraemer WJ, Otieno S (2013) Changes in height, body weight, and body
composition in American football players from 1942 to 2011. J Strength Cond Res 27(2):
277–284. doi:10.1519/JSC.0b013e31827f4c08
Australian Institute of Sport: Fact sheets, Increasing Muscle Mass. (2009). http://www.ausport.
gov.au/ais/nutrition/factsheets/body_size_and_shape/increasing_muscle_mass
Baron S, Rinsky R (1994) NIOSH mortality study of NFL football players: 1959–1988. 13
Batista C, Soares JM (2013) Are former elite athletes more protected against metabolic syndrome?
J Cardiol 61(5–6):440–445. doi:10.1016/j.jjcc.2013.01.011
Berglund L, Sundgot-Borgen J, Berglund B (2011) Adipositas athletica: a group of neglected
conditions associated with medical risks. Scand J Med Sci Sports 21(5):617–624. doi:10.1111/
j.1600-0838.2011.01322.x
Bjorntorp P (1991) Metabolic implications of body fat distribution. Diabetes Care 14(12):
1132–1143
Borchers JR, Clem KL, Habash DL, Nagaraja HN, Stokley LM, Best TM (2009) Metabolic
syndrome and insulin resistance in division 1 collegiate football players. Med Sci Sports
Exerc 41(12):2105–2110. doi:10.1249/MSS.0b013e3181abdfec
Bosy-Westphal A, Later W, Hitze B, Sato T, Kossel E, Gluer CC, Heller M, Muller MJ (2008)
Accuracy of bioelectrical impedance consumer devices for measurement of body composition
in comparison to whole body magnetic resonance imaging and dual X-ray absorptiometry.
Obes Facts 1(6):319–324. doi:10.1159/000176061
Bouchard C, Tremblay A, Despres JP, Nadeau A, Lupien PJ, Theriault G, Dussault J, Moorjani S,
Pinault S, Fournier G (1990) The response to long-term overfeeding in identical-twins. N Engl
J Med 322(21):1477–1482. doi:10.1056/nejm199005243222101
Buell JL, Calland D, Hanks F, Johnston B, Pester B, Sweeney R, Thorne R (2008) Presence of
metabolic syndrome in football linemen. J Athl Train 43(6):608–616. doi:10.4085/1062-6050-
43.6.608
Callister R, Callister RJ, Fleck SJ, Dudley GA (1990) Physiological and performance responses to
overtraining in elite judo athletes. Med Sci Sports Exerc 22(6):816–824
Callister R, Callister RJ, Staron RS, Fleck SJ, Tesch P, Dudley GA (1991) Physiological charac-
teristics of elite Judo athletes. Int J Sports Med 12(2):196–203. doi:10.1055/s-2007-1024667
Claesson AL, Holm G, Ernersson A, Lindstrom T, Nystrom FH (2009) Two weeks of overfeeding
with candy, but not peanuts, increases insulin levels and body weight. Scand J Clin Lab Invest
69(5):598–605. doi:10.1080/00365510902912754
Deacres-Manning S (1998) Anthropometry of the NSW rugby union Super 12 team. In: Australian
conference of science and medicine in sport, Adelaide, p. 94
Duthie G, Pyne D, Hooper S (2003) Applied physiology and game analysis of rugby union. Sports
Med 33(13):973–991. doi:10.2165/00007256-200333130-00003
Franchini E, Del Vecchio FB, Matsushigue KA, Artioli GG (2011) Physiological profiles of elite
Judo athletes. Sports Med 41(2):147–166. doi:10.2165/11538580-000000000-00000
Grandjean A (1999) Nutritional requirements to increase lean mass. Clin Sports Med 18(3):
623–632
Guo JJ, Zhang X, Wang L, Guo Y, Xie MH (2013) Prevalence of metabolic syndrome and its
components among Chinese professional athletes of strength sports with different body weight
categories. Plos One 8(11):7. doi:10.1371/journal.pone.0079758
Haskins S, Bernhardt DT, Koscik RL (2011) Screening for insulin resistance and cardiovascular
risk in collegiate football linemen. Clin J Sport Med 21(3):233–236. doi:10.1097/JSM.
0b013e31821a61f8
Hoshi A, Inaba Y (1995) Risk factors for mortality and mortality rate of Sumo wrestlers. Nihon
eiseigaku zasshi Jpn J Hyg 50(3):730–736
Iguchi J, Yamada Y, Ando S, Fujisawa Y, Hojo T, Nishimura K, Kuzuhara K, Yuasa Y, Ichihashi

N (2011) Physical and performance characteristics of Japanese division 1 collegiate football
players. J Strength Cond Res 25(12):3368–3377. doi:10.1519/JSC.0b013e318215fc19
Iida E, Nakajima T, Matsuura Y, Takeuchi M, Matsumoto D, Tanaka H, Komori F (1997) The
relationship between basic physical fitness and body fat in +95 kg category university Judo
athletes. Res J Budo 30(1):22–30
Japanese Olympic Committee (2001) The physical fitness test report of Japan national team; the
18th Olympic Winter Games (1998/Nagano) and the 27th Olympic Summer Games (2000/
Sydney)
Kaul S, Rothney MP, Peters DM, Wacker WK, Davis CE, Shapiro MD, Ergun DL (2012) Dual-
energy X-ray absorptiometry for quantification of visceral fat. Obesity (Silver Spring) 20(6):
1313–1318. doi:10.1038/oby.2011.393
Koury JC, De Oliveira CF, Portella ES, Oliveira AV, Donangelo CM (2005) Effect of the period of
resting in elite judo athletes – hematological indices and copper/zinc-dependent antioxidant
capacity. Biol Trace Elem Res 107(3):201–211. doi:10.1385/bter:107:3:201
Koury JC, Fernandes de Oliveira KJ, Lopes GC, de Oliveira AV Jr, Portella ES, de Moura EG,
Donangelo CM (2007) Plasma zinc, copper, leptin, and body composition are associated in
elite female judo athletes. Biol Trace Elem Res 115(1):23–30. doi:10.1385/bter:115:1:23
Kraemer WJ, Torine JC, Silvestre R, French DN, Ratamess NA, Spiering BA, Hatfield DL,
Vingren JL, Volek JS (2005) Body size and composition of national football league players.
J Strength Cond Res/Nat Strength Cond Assoc 19(3):485–489. doi:10.1519/18175.1
Little NG (1991) Physical performance attributes of junior and senior women, juvenile, junior, and
senior men judokas. J Sports Med Phys Fitness 31(4):510–520
Mathews EM, Wagner DR (2008) Prevalence of overweight and obesity in collegiate American
football players, by position. J Am Coll Health 57(1):33–37. doi:10.3200/jach. 57.1.33-38
Miller MA, Croft LB, Belanger AR, Romero-Corral A, Somers VK, Roberts AJ, Goldman ME
(2008) Prevalence of metabolic syndrome in retired national football league players. Am J
Cardiol 101(9):1281–1284. doi:10.1016/j.amjcard.2007.12.029
Ministry of Health, Labour and Welfare of Japan, Dietary Reference Intakes for Japanese, 2010.
(2012). http://www0.nih.go.jp/eiken/english/research/pdf/dris2010_eng.pdf
Murata H, Ishikawa-Takata K, Natsui H, Taguchi M (2013) Characteristics of body composition
and risks of metabolic syndrome in female judo athletes categorized in heavy weight classes. J
Jpn Soc Clin Sports Med 21(3):623–631
Nagasawa T, Mutara H, Muraoka Y, Natsui H, Taguchi M (2013) Study of new approaches to
increase the body weight of Japanese athletes. J Jpn Soc Clin Sports Med 21(2):422–430
Nevill AM, Stewart AD, Olds T, Holder R (2006) Relationship between adiposity and body size
reveals limitations of BMI. Am J Phys Anthropol 129(1):151–156. doi:10.1002/ajpa.20262
Olds T (2001) The evolution of physique in male rugby union players in the twentieth century. J
Sports Sci 19(4):253–262. doi:10.1080/026404101750158312
Olefsky J, Crapo PA, Ginsberg H, Reaven GM (1975) Metabolic effects of increased caloric intake
in man. Metab Clin Exp 24(4):495–503
Oppert JM, Nadeau A, Tremblay A, Despres JP, Theriault G, Deriaz O, Bouchard C (1995) Plasma
glucose, insulin, and glucagon before and after long-term overfeeding in identical twins. Metab
Clin Exp 44(1):96–105. doi:10.1016/0026-0495(95)90295-3
Rankinen T, Kim SY, Perusse L, Despres JP, Bouchard C (1999) The prediction of abdominal
visceral fat level from body composition and anthropometry: ROC analysis. Int J Obes 23(8):
801–809. doi:10.1038/sj.ijo.0800929
Rodriguez NR, Di Marco NM, Langley S (2009) American college of sports medicine position
stand. Nutrition and athletic performance. Med Sci Sports Exerc 41(3):709–731. doi:10.1249/
MSS.0b013e31890eb86
Selden MA, Helzberg JH, Waeckerle JF, Browne JE, Brewer JH, Monaco ME, Tang F, O’Keefe
JH (2009) Cardiometabolic abnormalities in current national football league players. Am J
Cardiol 103(7):969–971. doi:10.1016/j.amjcard.2008.12.046
Sertic H, Segedi I, Milanovic D (2006) Anthropological and fitness status of Croatian judoists.
Arch Budo 2:24–27
Sims EA, Goldman RF, Gluck CM, Horton ES, Kelleher PC, Rowe DW (1968) Experimental
obesity in man. Trans Assoc Am Physicians 81:153–170
Sone H, Ito H, Ohashi Y, Akanuma Y, Yamada N (2003) Obesity and type 2 diabetes in Japanese
patients. Lancet 361(9351):85. doi:10.1016/S0140-6736(03)12151-4
Sugimoto R (1963) Rikishi no Tounyoubyou hatsugen hindo ni kansuru tyousa seiseki. Jpn J Phys
Fit Sports Med 12(4):160–165
Tucker AM, Vogel RA, Lincoln AE, Dunn RE, Ahrensfield DC, Allen TW, Castle LW, Heyer RA,
Pellman EJ, Strollo PJ, Wilson PWF, Yates AP (2009) Prevalence of cardiovascular disease
risk factors among national football league players. J Am Med Assoc 301(20):2111–2119.
doi:10.1001/jama.2009.716
Welle SL, Seaton TB, Campbell RG (1986) Some metabolic effects of overeating in man. Am J
Clin Nutr 44(6):718–724
Wilkerson GB, Bullard JT, Bartal DW (2010) Identification of cardiometabolic risk among
collegiate football players. J Athl Train 45(1):67–74. doi:10.4085/1062-6050-45.1.67
Yamauchi T, Abe T, Midorikawa T, Kondo M (2004) Body composition and resting metabolic rate
of Japanese college Sumo wrestlers and non-athlete students: are Sumo wrestlers obese?
Anthropol Sci 112(2):179–185. doi:10.1537/ase.040210
Chapter 13
High Fat Diet and Endurance Exercise
Performance
Kazuhiko Higashida and Mitsuru Higuchi
Abstract The high carbohydrate diet has been a popular dietary intervention
among endurance athletes for many years, because the depletion of muscle glyco-
gen is associated with fatigue during endurance exercise. Therefore, numerous
studies regarding carbohydrate supplementation and exercise regimens have been
done and have demonstrated the association between muscle glycogen and perfor-
mance in endurance exercise. These results strongly suggest the importance of
maximizing muscle glycogen stores before the initiation of exercise to allow the
athlete to have an optimal endurance performance. On the other hand, high fat diet
intervention is not popular, although it has been reported that a high fat diet induces
an increase in mitochondrial biogenesis in skeletal muscle and enhances endurance
performance in animal studies. Therefore, the purpose of this article is to consider
the evidence that an increased in fat intake is beneficial for endurance performance.
Keywords High fat diet • Mitochondria • Skeletal muscle • Endurance

performance
13.1 Fat and Carbohydrate Metabolism During Endurance

Exercise
Fat and carbohydrate are the main fuels that are that are utilized by oxidative
processes to produce the ATP that is necessary for skeletal muscle function during
endurance exercise. Most of the fat in the body is stored in subcutaneous and
abdominal adipose tissue. A small amount of fat is also stored in skeletal muscle
as lipid droplets. Carbohydrate is stored in the body as glycogen, in the liver and
skeletal muscle. In skeletal muscle, glycogen stores are typically around 400 g, and
this value is increased up to 900 g when subject with a large muscle mass trains and
consumes a high carbohydrate diet (Jeukendrup et al. 1998).
Fat and carbohydrate are oxidized as a mixture, but the relative contribution of
these two substrates depends on exercise intensity, duration, aerobic capacity and
K. Higashida (*) • M. Higuchi

e-mail: khigashida@aoni.waseda.jp

152 K. Higashida and M. Higuchi
diet. During endurance exercise, lipolysis in adipose tissue is increased after the
onset of exercise by adrenergic stimulation of adipose-tissue triglyceride lipase and
hormone-sensitive lipase. Degradation of triglyceride results in the formation of
glycerol and fatty acids, which are released into the blood from the adipose tissue.
The free fatty acids bind to albumin and are carried to the skeletal muscles to be
oxidized. The free fatty acids must be dissociated from albumin in order to enter
muscle cells. Although it has been believed that fatty acid uptake into muscle cells
occurs via “passive diffusion” across the plasma membrane, recent evidence sug-
gests that the uptake of fatty acids from the blood into the muscle cells occurs via a
carrier-mediated diffusion process. Once the fatty acids enter the skeletal muscle
cells, they can be either converted to fatty acyl-CoA or incorporated into a triglyc-
eride molecule. Acyl-CoA is transported into the mitochondria by carnitine
palmitoyl transferase. In the mitochondria, the acyl-CoA enters the beta oxidation
process, and eventually produces ATP via the electron transport chain.
During low intensity exercise, plasma free fatty acids are a main energy source
for skeletal muscle, and fat oxidation increases when the exercise intensity
increases from low to moderate intensities. At higher intensities, fat oxidation is
inhibited and muscle glycogen becomes the most important substrate (Romijn
et al. 1993).
13.2 Effect of Short-Term High-Fat Diet Feeding

on Endurance Exercise Performance
Muscle glycogen is essential for prolonged, moderate-intensity exercise, and

exhaustion can result from the depletion of muscle glycogen (Bergstrom
et al. 1967). Because dietary intervention has been used to stimulate fat oxidation
by elevating free fatty acid concentrations in blood in an attempt to reduce glycogen
utilization and to thus improve performance, so-called “fat loading” is not a new
dietary intervention. Over the years, evidence has accumulated regarding interven-
tion studies in which dietary fat intake has been manipulated over a short period.
Zuntz (1901) reported that the adaptation to a fat rich diet led to respiratory
exchange ratio values during mild exercise that suggested an almost exclusive
oxidation of fat. However, most of the subsequent studies on short term high fat
diet adaptation showed that such diets resulted in a lower endurance performance
than when a high carbohydrate diet was consumed (Bergstrom et al. 1967; Martin
et al. 1978; Pitsiladis and Maughan 1999). This evidence clearly demonstrated that
short term adaptation to a high fat diet was not beneficial to endurance performance.
Because short term high fat intervention (3 days to 1 week) is detrimental to
endurance performance, researchers changed their focus to an investigation of the
effect of consuming a high fat diet or fat supplement just before an endurance
performance exercise bout. In the study of Pitsiladis et al. (1999), plasma free fatty
acid concentrations were elevated by consuming a fatty meal before exercise and an
13 High Fat Diet and Endurance Exercise Performance 153
intravenous administration of heparin 30 min before exercise. Heparin was used to

activate lipoprotein lipase and the hydrolysis of plasma triglyceride. The combina-
tion of heparin injection and a fatty meal resulted in raising the free fatty acid
concentration in plasma, and the time to exhaustion increased from 118.2 min on
the high carbohydrate diet trial to 127.9 min on the high fat diet trial. However, as
the authors mentioned, results from the study cannot be utilized by athletes, because
catheterization and infusion are illegal in sports competitions. Okano et al. (1996)
demonstrated that a high fat meal consumed 4 h before cycling exercise increased
the plasma free fatty acid concentration as compared to a high carbohydrate meal.
However, endurance capacity did not significantly differ between the high carbo-
hydrate meal group (128 min) and high fat meal (122 min) trials. Therefore, raising
blood free fatty acid level by consuming a high fat diet before exercise bout does
not seem to be an ergogenic aid for endurance performance.
13.3 Effect of Long Term High Fat Diet Feeding

on Endurance Exercise Performance
As mentioned above, short term high fat diet intervention is detrimental to endur-
ance exercise performance. We will now evaluate the effect of long term high fat
diet intervention on muscle adaptation and endurance performance. Miller
et al. (1984) reported that a 5-week high fat diet intervention improved endurance
exercise capacity in rats (Control diet: 35.5 3.1, High fat diet: 47.1 3.6 min)
(p < 0.05). In that study, the endurance capacity of rats was increased by 33 % when
compared with a normal diet, though the muscle glycogen content before the
endurance capacity test was higher in rats fed a normal diet (Control diet:
54.2 2.7, High fat diet: 40.2 3.1 μmol glycosyl units/g wet weight) (p < 0.05).
The improved endurance has been attributed to increased oxidative enzyme activ-
ities, such as citrate synthase (CS) and β-hydroxy-acyl CoA dehydrogenase (HAD),
and the decreased degradation of muscle glycogen during exercise. These adapta-
tions are similar to those observed after endurance exercise training. Interestingly,
the enzymatic adaptations seen following endurance exercise training are also
enhanced by a high fat diet. Simi et al. reported that the effect of a high fat diet
and endurance training on CS and HAD activity are additive (Simi et al. 1991). The
cumulative effect of training and a high fat diet on mitochondrial enzyme activities
thus resulted in further improvement of endurance performance. These results
suggest the possibility that endurance training and a high fat diet induce an increase
in mitochondrial biogenesis in skeletal muscle though different mechanisms.
However, the effect of long term high fat diets on endurance performance in
humans is less clear than the results of the animal studies. Phinney et al. (1983)
reported that consuming a high fat diet for 4–6 weeks resulted in increased fat
oxidation during exercise and the maintenance of endurance capacity. Another
study by Helge et al. (1998) showed that adaptation to a high fat diet for 4 weeks
maintained endurance capacity at 69 % of VO2 max, whereas a significant decrease

was observed after a 7-week high fat diet period. These results indicate the
possibility that a prolonged low carbohydrate diet has a detrimental effect on
endurance exercise performance in humans. However, it is remarkable that endur-
ance capacity was not impaired in all subjects, even though muscle glycogen
content before exercise was reduced by about 50 % and fat oxidation during
exercise was significantly increased. In addition, Fisher et al. (1983) demonstrated
that carnitine palmitoyl transferase I activity in skeletal muscle increased after a
4-week eucaloric ketogenic diet intervention. These results suggest that the adap-
tation to long term high fat diet observed in rat skeletal muscle did occur in the
skeletal muscle of humans. Therefore, since human skeletal muscle also adapts to a
high fat diet, and a high fat diet intervention, with some modification, may be useful
for nutritional intervention for endurance athletes. In order for this to be realized,
the precise molecular mechanisms which underlie the metabolic adaptations to a
high fat diet in skeletal muscle must be determined.
13.4 Mechanisms in Which High Fat Diet-Induced

an Increase in Mitochondrial Biogenesis in Skeletal
Muscle
The increases in mitochondrial enzyme activities in response to a high fat diet

contribute to increases in fat oxidation during exercise. As a consequence, there is a
simultaneous decrease in carbohydrate utilization due to glycogen sparing. How-
ever, the molecular mechanisms by which a high fat diet-induces an increase in
mitochondrial biogenesis are not known. For example, citrate synthase (CS) and
β-hydroxy-acyl CoA dehydrogenase (HAD) activities are well known to be induced
by a high fat diet, whereas the regulation of gene transcription of these enzymes is
regulated by different transcriptional factors. In addition, the mitochondrion has its
own genome. Thus, mitochondrial biogenesis requires the orchestrated expression
of the genes encoded in the mitochondrial genome and the nuclear genes encoding
mitochondrial proteins. In 1998, Spiegelman’s groups reported the mechanism
underlying how the signals generated by adaptive stimuli result in a coordinated
increase in expression of the many genes encoding mitochondrial proteins. They
named the involved protein peroxisome proliferator-activated receptor gamma
coactivator-1 alpha (PGC-1α), and this protein is now well-known as a master
regulator of adaptive mitochondrial biogenesis in skeletal muscle. Shortly after the
discovery of PGC-1α, it was reported that endurance exercise induces an increase in
PGC-1α protein content in rat skeletal muscle (Terada and Tabata 2004), and it is
now widely recognized that PGC-1α plays an important role in exercise-induced
increase in mitochondrial biogenesis of skeletal muscle. This evidence led
researchers to investigate that whether PGC-1α was involved in high fat diet-
induced mitochondrial biogenesis in skeletal muscle. In 2008, Hancock
13 High Fat Diet and Endurance Exercise Performance 155
et al. (2008) reported that consumption of a high fat diet for 4-weeks induced an
increase in PGC-1α protein content by about two-fold in skeletal muscle as com-
pared to that of rats fed a normal chow diet. Interestingly, these investigators
showed that consuming a high fat diet for 4-weeks induced an increase in
PGC-1α protein content, but not PGC-1α mRNA at the expression level. These
results suggest that the consumption of a high fat diet increases PGC-1α protein
content through an increase in the protein stability of PGC-1α. It is well known that
post-translational modification of proteins influences the protein’s stability, such as
by phosphorylation or acetylation. In regard to PGC-1α, phosphorylation by p38
MAP kinase increases PGC-1α protein stability (Puigserver et al. 2001), and
acetylation/de-acetylation modifies the transcriptional activity of PGC-1α (Rodgers
et al. 2008). The molecular mechanisms underlying protein stability changes of
PGC-1α in response to a high fat diet need to be investigated in future studies.
13.5 Summary
The possibility that a high fat diet can improve endurance exercise performance by
increasing fat oxidation and sparing muscle glycogen is intriguing. Long term
consumption of a high fat diet induces an increase in mitochondrial biogenesis,
and in rodents, concomitantly elevates endurance exercise performance, Evidence
for such an effect in humans is problematical. To clarify the importance of dietary
fat and carbohydrate content for human endurance performance, well-controlled
experiments are required. Particularly, studies that manipulate dietary fat content
for more than 4 weeks, and in which the consequent adaptations of skeletal muscle
are monitored, will be required. Long term, high fat diets are generally associated
with the development of obesity and cardiovascular disease and are thus seen to be
unhealthy. However, this effect may reflect the nature of the fats consumed rather
than the total amount of fat, and the quality of the fats consumed should be
addressed in future studies.
References
Bergstrom J, Hermansen L, Hultman E, Saltin B (1967) Diet, muscle glycogen, and physical
performance. Acta Physiol Scand 71:140–150
Fisher EC, Evans WJ, Phinney SD, Blackburn GL, Bistrian BR, Young VR (1983) Changes in
skeletal muscle metabolism induced by a eucaloric ketogenic diet. In: Knuttgen HG, Vogel JA,
Poortmans J (eds) Biochemistry of exercise. Human Kinetics Publishers Inc, Champaign, pp
497–507
Hancock CR, Han DH, Chen M, Terada S, Yasuda T, Wright DC, Holloszy JO (2008) High-fat
diets cause insulin resistance despite an increase in muscle mitochondria. Proc Natl Acad Sci
U S A 105(22):7815–7820
Helge JW, Wulff B, Kiens B (1998) Impact of a fat-rich diet on endurance in man: role of the
dietary period. Med Sci Sports Exerc 3:456–461
Jeukendrup AE, Thielen JJ, Wagenmakers AJ, Brouns F, Saris WH (1998) Effect of medium-chain
triacylglycerol and carbohydrate ingestion during exercise on substrate utilization and subse-
quent cycling performance. Am J Clin Nutr 67(3):397–404
Martin B, Robinson S, Robertshaw D (1978) Influence of diet on leg uptake of glucose during
heavy exercise. Am J Clin Nutr 1:62–67
Okano G, Sato Y, Takumi Y, Sugawara M (1996) Effect of 4h preexercise high carbohydrate and
high fat meal ingestion on endurance performance and metabolism. Int J Sports Med 17(7):
530–534
Phinney SD, Bistrian BR, Evans WJ, Gervino E, Blackburn GL (1983) The human metabolic
response to chronic ketosis without caloric restriction: preservation of submaximal exercise
capability with reduced carbohydrate oxidation. Metabolism 8:769–776
Pitsiladis YP, Maughan RJ (1999) The effects of exercise and diet manipulation on the capacity
to perform prolonged exercise in the heat and in the cold in trained humans. J Physiol
517(Pt 3):919–930
Pitsiladis YP, Smith I, Maughan RJ (1999) Increased fat availability enhances the capacity of
trained individuals to perform prolonged exercise. Med Sci Sports Exerc 31(11):1570–1579
Puigserver P, Rhee J, Lin J, Wu Z, Yoon JC, Zhang CY, Krauss S, Mootha VK, Lowell BB,
Spiegelman BM (2001) Cytokine stimulation of energy expenditure through p38 MAP kinase
activation of PPARgamma coactivator-1. Mol Cell 5:971–982
Rodgers JT, Lerin C, Gerhart-Hines Z, Puigserver P (2008) Metabolic adaptations through the
PGC-1 alpha and SIRT1 pathways. FEBS Lett 582(1):46–53
Romijn JA, Coyle EF, Sidossis LS, Gastaldelli A, Horowitz JF, Endert E, Wolfe RR (1993)
Regulation of endogenous fat and carbohydrate metabolism in relation to exercise intensity and
duration. Am J Physiol 265(3 Pt 1):E380–E391
Simi B, Sempore B, Mayet MH, Favier RJ (1991) Additive effects of training and high-fat diet on
energy metabolism during exercise. J Appl Physiol (1985) 71(1):197–203
Terada S, Tabata I (2004) Effects of acute bouts of running and swimming exercise on
PGC-1alpha protein expression in rat epitrochlearis and soleus muscle. Am J Physiol
Endocrinol Metab 286(2):E208–E216
Zuntz N (1901) Ueber die Bedeutung der verschiedenen Nährstoffe als Erzeuger der Muskelkraft.
Pflugers Arch 83:557–571
Chapter 14
Nonuniform Muscle Hypertrophy Along
the Length Induced by Resistance Training
Taku Wakahara
Abstract Resistance training induces an increase in skeletal muscle size (hyper-

trophy). This hypertrophy can occur nonuniformly along the length of a muscle or
muscle group. This review summarizes studies examining training-induced hyper-
trophy in multiple regions of human skeletal muscles. Results showed highly
variable hypertrophic patterns along the length of muscle(s), possibly due to
disparities in training programs and subject characteristics. Although differences
in muscle activation and contractile protein synthesis are associated with such
nonuniform hypertrophic change, the underlying mechanisms remain unclear.
Future investigations are needed to reveal the mechanisms and to design resistance
training programs that consider region-specific muscle hypertrophy.
Keywords CSA • Muscle activation • Protein synthesis
14.1 Introduction
Muscle size is a major determinant of its strength (Ikai and Fukunaga 1968;
Maughan et al. 1983; Fukunaga et al. 2001), and muscle size increases in response
to resistance training (muscle hypertrophy). Many studies have quantified muscle
hypertrophy induced by specific training programs, and these findings have been
summarized in several review articles (Kraemer and Ratamess 2004; Wernbom
et al. 2007; Ratamess et al. 2009; Schoenfeld 2010). Although many of these studies
evaluated the extent of muscle hypertrophy based on muscle thickness or anatom-
ical cross-sectional area (CSA) in a specific region, it has been demonstrated that
training-induced hypertrophy can occur nonuniformly along the length of a muscle
and muscle group (Narici et al. 1989). In addition, the magnitude of increase in a
single-slice CSA taken at the belly of a muscle group is different from that in the
entire volume of the group (Roman et al. 1993). These findings require a reconsid-
eration of muscle hypertrophy in terms of regional differences along the muscle
T. Wakahara (*)
Faculty of Health and Sports Science, Doshisha University, Kyoto, Japan
e-mail: twakahar@mail.doshisha.ac.jp

158 T. Wakahara
length. The goal of this article is to provide an overview of nonuniform hypertrophy

in human muscle along the longitudinal axis following resistance training.
Because of the wide variation among studies, it is first desirable to standardize
the region used to determine muscle size. Researchers have assessed muscle size at
an absolute distance from a reference point (e.g. 10 cm proximal to the elbow joint)
or a relative distance of segment length (SL, e.g. 50 % of femur length) or muscle
length. When SL was used, its definition (the reference point of the proximal and/or
distal ends) differed among studies. However, the relative distance of SL was
frequently used to describe the measurement region. Fortunately, disparities in
the reference point did not have a substantial impact on the main findings. There-
fore, when possible, the measurement region is expressed as the relative distance of
SL from the proximal (0 %) to distal ends (100 %) in this article.
14.2 Resistance Training-Induced Muscle Hypertrophy

Along the Muscle Length
14.2.1 Hypertrophic Changes Along the Length

of Quadriceps Femoris (QF)
Resistance training can result in nonuniform hypertrophy in various muscles or

muscle groups. Krotkiewski et al. (1979) provided an early report of this phenom-
enon. In this report, muscle thickness of the QF was measured in four different
regions before and after 5 weeks of isokinetic knee extension training. Significant
increases in muscle thickness were found at the lateral portion of 50 %SL and the
anterior portion of 67 %SL, but not at the medial or anterior portions of 50 %SL.
Following this report, Narici et al. (1989) investigated changes in CSA of the whole
QF and of each component of QF in seven regions along the length after 60 days of
isokinetic knee extension training. Thereafter, many studies examined changes in
multiple regions of QF CSA, and the results are summarized in Table 14.1. Studies
are not included in the table, if the data were not supported by statistical analysis
(Tracy et al. 1999) or if CSA included muscles other than the QF (Bloomquist
et al. 2013). The hypertrophic pattern along the length of QF is quite inconsistent.
Some studies reported preferential hypertrophy in the proximal (Smith and Ruth-
erford 1995; Hisaeda et al. 1996 (group a)); middle (Ahtiainen et al. 2003; Petersen
et al. 2011), or distal (Hisaeda et al. 1996 (group b); Seger et al. 1998 (group b))
region, but others failed to find a significant regional difference (Schott et al. 1995;
Higbie et al. 1996; Blazevich et al. 2007; Seynnes et al. 2007). Regional differences
in muscle hypertrophy vary according to program variables of the training (Hisaeda
et al. 1996; Seger et al. 1998; Häkkinen et al. 2003). For example, Hisaeda
et al. (1996) reported that low-intensity, high-volume knee extension training
significantly increased QF CSA at 30 %SL, but not at 50 % or 70 %SL, whereas
high-intensity, low-volume training increased CSA at 70 %SL, but not at 30 % or
14 Nonuniform Muscle Hypertrophy Along the Length Induced by Resistance Training 159
50 %SL. Hence, one reason for the inconsistent findings may be different combi-
nations of training intensity and volume. Seger et al. (1998) showed that eccentric
knee extension training significantly increased QF CSA at 12 cm distally from 50 %
SL, but not at 50 %SL. On the other hand, there was no significant change in CSA
after concentric knee extension training in either region. These results suggest that
the contraction type affects nonuniform muscle hypertrophy. However, the effect of
contraction type is unlikely to be strong, as Smith and Rutherford (1995) and Higbie
et al. (1996) demonstrated similar hypertrophic changes in the QF following
concentric or eccentric knee extension training. Even with the same training
program, hypertrophic changes in the QF along its length differ slightly among
subject groups (Häkkinen et al. 2002; Melnyk et al. 2009). Melnyk et al. (2009)
examined the effects of dynamic knee extension training on QF CSA at 30 %, 50 %,
and 70 %SL in four subject groups [young (20–30 years) and older (65–75 years)
men and women]. Young men and older men and women significantly increased QF
CSA in all three regions. On the other hand, young women significantly increased
CSA only at 50 % and 70 %SL. Häkkinen and colleagues studied changes in QF
CSA in various groups of subjects following similar training programs (Häkkinen
et al. 2001, 2002, 2003; Ahtiainen et al. 2003). The results indicated that variable
hypertrophic change along the muscle length could not be explained only by age or
gender. A resistance training-only group (16 men, 37 5 years; Häkkinen
et al. 2003) significantly increased QF CSA at 20–67 %SL, but not at 73 %SL
after 21 weeks of knee extension and leg press training. In contrast, a similar
training program resulted in significant increases in QF CSA at 40–67 %SL, but
not at 20–33 %SL in a separate group (8 men, 34 4 years; Ahtiainen et al. 2003).
In addition to the studies listed in Table 14.1, Kanehisa et al. (2003) investigated
the changes in QF CSA in junior weight lifters (seven boys). Eighteen months after
the baseline measurement, QF CSA increased significantly at 70 % SL, but not at
30 % or 50 %SL. Similarly, Kanehisa et al. (2006) examined changes in QF CSA at
30 %, 50 %, and 70 %SL in teenage tennis players (six boys and six girls) over a
2-year period. In boys, QF CSA increased significantly in all three regions, whereas
in girls, QF CSA increased significantly only at 30 % and 50 %SL. Unfortunately,
these studies (Kanehisa et al. 2003, 2006) did not include untrained control groups.
Therefore, it is unclear whether the nonuniform changes in QF CSA were due to
competitive training, natural growth, or both.
14.2.2 Hypertrophic Changes Along the Length of Each

Component of the QF
Table 14.2 shows reported regional hypertrophic changes in each component of QF

along the length of the muscle. The pattern of hypertrophy along the length varies
greatly in each of the vasti [vastus intermedius (VI), vastus lateralis (VL), and
vastus medialis (VM)]. As with QF CSA, this variation may be partly attributed to
160 T. Wakahara
Table 14.1 Studies on training-induced regional changes in QF CSA

Subjects Reps sets
Authors (mean Exercise (or time) Duration
(year) age) modality intensity frequency Results
Narici 4 M (28) KE (ISOK) 10 6 60 days [CSA]
et al. (1989) maximal 4 days/week 20, 30, 40, 50, 60,
70 %SL: ", 80 %
SL: n.s.
Schott 1 M, KE (ISOM) a. 3 s 10 4 14 weeks [CSA]
et al. (1995) 6 W (23) 70 %MVC
b. 30 s 4 1 3 days/week a. 25, 75 %SL: n.s.
70 %MVC
b. 25, 75 %SL: "
Smith and 5 M (21) a. KE (CON) 10 4 20 weeks [CSA]
Rutherford 5 W (20) b. KE (ECC) 10RM 3 days/week a. 25 %SL: ", 75 %
(1995) SL: n.s.
b. 25 %SL: ",
75 %SL: n.s.
Higbie a. 16 a. KE (CON, 10 3 10 weeks [CSA]
et al. (1996) W (20) ISOK)
b. 19 b. KE (ECC, maximal 3 days/week a. 20, 30, 40, 50,
W (20) ISOK) 60, 70, 80 %SL: "
b. 20, 30, 40, 50,
60, 70, 80 %SL: "
Hisaeda a. 5 KE (CON/ECC) a. 15 5–6 15– 8 weeks [CSA]
et al. (1996) W (20) 20RM
b. 6 b. 5 8–9 4– 3 days/week a. 30 %SL: ",
W(20) 5RM 50, 70 %SL: n.s.
b. 70 %SL: ",
30, 50 %SL: n.s.
Narici 7 M (29) KE (CON/ECC) 8 6 6 months [CSA]
et al. (1996) 80 %1RM 3.5 days/ 30, 40, 50,
week 60, 70 %SL: "
[Relative
increase in CSA]
30 %SL > 60 %SL
70 %SL > 40 %
SL > 50, 60 %SL
Seger a. 5 a. KE (CON, 10 4 10 weeks [CSA]
et al. (1998) M (24) ISOK)
b. 5 b. KE (ECC, maximal 3 days/week a. 50 %SL, Distala:
M (25) ISOK) n.s.
b. Distal: ",
50 %SL: n.s.
Häkkinen 10 W (64) KE (CON/ECC) 5–20 3–6 21 weeks [CSA]
et al. (2001) LP (CON/ECC) 40–80 %1RM 2 days/week 20, 27, 33, 40, 47,
53, 60, 67, 73 %
SL: ", 80 %SL:
n.s.
(continued)

Subjects Reps sets
Häkkinen a. 11 W KE (CON/ECC) 5–20 3–6 21 weeks [CSA]
et al. (2002) with FM
(39)
b. 10 LP (CON/ECC) 40–80 %1RM 2 days/week a. 20, 27, 33,
healthy 40, 47, 53, 60,
W (37) 67 %SL: ",
73, 80 %SL: n.s.
b. 20, 27, 33,
40, 47, 53, 60,
67, 73, 80 %SL: "
Häkkinen a. 16 a. KE a. 3–15 3–6 21 weeks [CSA]
et al. (2003) M (38) (CON/ECC) LP 50–80 %1RM
(CON/ECC)
b. 11 b. KE b. [RT] 3– a. 2 days/ a. 20, 27, 33,
M (37) (CON/ECC) 15 3–6 50– week 40, 47, 53, 60,
LP (CON/ECC) 80 %1RM 67 %SL: ", 73 %
Walking [ET] 30–90 min SL: n.s.
Cycling
b. 4 days/ b. 20, 27, 33,
week (2 for 40, 47, 53, 60,
RT and 2 for 67, 73 %SL: "
ET)
Ahtiainen 8 M (34) KE (CON/ECC) 3–15 3–6 21 weeks [CSA]
et al. (2003) LP (CON/ECC) 50–80 %1RM 2 days/week 40, 47, 53,
60, 67 %SL: ",
20, 27, 33 %SL:
n.s.
Blazevich 12 M a. KE (CON, 6 4–6 10 weeks [Relative increase
et al. (2007) (24), ISOK) in CSA]
12 W (21) b. KE (ECC, maximal 3 days/week 25 %SL vs. 75 %
ISOK) SL: n.s.
Seynnes 5 M, 2 KE (CON/ECC) 74 5 weeks [CSA]
et al. (2007) W (20) maximal 3 days/week 50, 75 %SL: "
Melnyk a. 11 [CSA]
et al. (2009) M (25)
a. 30, 50, 70 %
SL: "
b. 10 KE (CON/ECC) 5–20 5 9 weeks b. 50, 70 %SL: ",
W (26) 30 %SL: n.s.
c. 11 50 %1RM– 3 days/week c. 30, 50, 70 %
M (69) 5RM SL: "
d. 11 d. 30, 50, 70 %
W (68) SL: "
(continued)
162 T. Wakahara

Subjects Reps sets
Hudelmaier a. 16 W a. KE a. 6–12 1–3 12 weeks [CSA]
et al. (2010) (CON/ECC), 60–80 %1RM
LP(CON/ECC),
SQ(CON/ECC)
b. 19 b. Cycling b. 40 min 55– 3 days/week a. 10, 20, 30,
W (51) 85 %HRmax 40, 50, 60 %SL: ",
70, 80, 90 %SL:
n.s.
b. 10, 20, 30,
40, 50, 60, 70,
80 %SL: ", 90 %
SL: n.s.
Petersen a. 5 M, [CSA]
et al. (2011) 7 W (62)
b. 4 M, KE(CON/ECC), 8–15 4–5 12 weeks a. Middle: ",
7 W (62) LP(CON/ECC) Distal: n.s.
c. 5 M, 8–15RM 3 days/week b. Middle: ",
7 W (63)b Distal: n.s.
c. Middle: ",
Distal: n.s.c
M men, W women, FM fibromyalgia, CON Concentric training, ECC Eccentric training, CON/
ECC Combination of concentric and eccentric training, ISOM Isometric training, ISOK Isokinetic
training, KE Knee extension, LP Leg press, SQ: Squat, RM: Repetition maximum, HRmax Heart
rate max, RT Resistance training, ET Endurance training, CSA Cross-sectional area, SL Segment
length, " Significant increase, n.s. Not significant
a
Distal region corresponded to 12 cm distally from 50 %SL
b
Subjects were patients of knee osteoarthritis administered either glucosamine (a), ibuprofen (b) or
placebo (c)
c
Distal and middle regions corresponded to 10 and 20 cm above the lateral tibial plateau,
respectively
disparities in combinations of training intensity and volume, and in subject char-

acteristics. However, even with similar program variables of training in similar
subjects, Narici et al. (1989) and Housh et al. (1992) observed different hypertro-
phic patterns along the length of the vasti. Regarding this discrepancy, Housh
et al. (1992) suggested that there might have been a difference in training status
between the dominant (Narici et al. 1989) and non-dominant (Housh et al. 1992)
thighs at the beginning of the study. As for the rectus femoris (RF), Ema
et al. (2013) reported that the relative increase in CSA was significantly greater at
50 %SL than at 30 %SL. This is consistent with other studies (Housh et al. 1992;
Narici et al. 1996; Blazevich et al. 2007), which showed a tendency toward a greater
relative increase in RF CSA in the distal region, although statistical analysis
between regions was not reported.
Table 14.2 Studies on the training-induced regional changes in CSA of each component of
the QF
Subjects
Authors (mean Exercise Reps sets Duration
Narici 4 M (28) KE (ISOK) 10 6 60 days [VI: CSA]
et al. (1989) maximal 4 days/ 30, 40, 60, 70 %SL: ",
week 20, 50, 80 %SL: n.s.
[VL: CSA]
40 %SL: ", 20, 30,
50, 60, 70, 80 %SL:
n.s.
[VM: CSA]
20, 30, 50, 60, 80 %
SL: ", 40, 70 %SL:
n.s.
[RF: CSA]
20, 40, 50 %SL: ",
30, 60, 70 %SL: n.s.
Housh 13 M (25) KE (CON, 10 6 8 weeks [VI, VL: CSA]
et al. (1992) ISOK)
maximal 3 days/ 50 %SL: ", 33, 67 %
week SL: n.s.
[VM: CSA]
33, 50, 67 %SL: n.s.
[RF: CSA]
33, 50, 67 %SL: "
Narici 7 M (29) KE (CON/ECC) 86 6 months [VI, VL, VM, RF:
et al. (1996) CSA]
80 %1RM 3.5 days/ 30, 40, 50, 60, 70 %
week SL: "a
Housh 9 M (24) KE (CON) 6 3–5 8 weeks [VI, VL, VM, RF:
et al. (1998) CSA]
80 %1RM 3 days/ 33, 50, 67 %SL: "b
week
Häkkinen 10 W KE (CON/ECC) 5–20 3–6 21 weeks [VI: CSA]
et al. (2001) (64)
LP (CON/ECC) 40–80 % 2 days/ 20, 27, 40, 47, 53,
1RM week 60, 67 %SL: "
33, 73, 80 %SL: n.s.
[VL: CSA]
20, 27, 33, 40, 47,
53 %SL: "
60, 67, 73 %SL: n.s.
(continued)
164 T. Wakahara

Subjects
[VM: CSA]
47, 53, 60, 67, 73,
80 %SL: "
20, 27, 33, 40 %SL:
n.s.
[RF: CSA]
40 %SL: "
20, 27, 33, 47, 53,
60, 67 %SL: n.s.
Reeves 4 M, 5 W KE (CON/ECC) ~10 2 14 weeks [VL: CSA]
et al. (2004) (74)
LP (CON/ECC) ~60–80 % 3 days/ 12, 15, 18, 21, 24,
5RM week 27 cm from PI: "
3, 6, 9, 30 cm from
PI: n.s.
Blazevich 12 M a. KE (CON, 6 4–6 10 weeks [VI: Relative increase
et al. (2007) (24), ISOK) in CSA]
12 W b. KE (ECC, maximal 3 days/ 25 %SL < 75 %SL
(21) ISOK) week
[VL, VM, RF: Rela-
tive increase in CSA]
25 %SL vs. 75 %SL:
n.s.
Seynnes 5 M, 2 W KE (CON/ECC) 74 5 weeks [VI: CSA]
et al. (2007) (20)
maximal 3 days/ 75 %SL: ",
week 50 %SL:n.s.
[VL, VM: CSA]
50, 75 %SL: "
Ema [VI: CSA] 45, 65 %
et al. (2013) SL: "
[VI: Relative increase
in CSA]
45 %SL vs. 65 %SL:
n.s.
[VL: CSA] 45, 65 %
SL: "
[VL: Relative
increase in CSA]
(continued)

Subjects
11 M (27) KE (CON/ECC) 85 12 weeks 45 %SL < 65 %SL
80 %1RM 3 days/ [VM: CSA] 65, 85 %
week SL: "
[VM: Relative
increase in CSA]
65 %SL vs. 85 %SL:
n.s.
[RF: CSA] 30, 50 %
SL: "
[RF: Relative
increase in CSA]
30 %SL < 50 %SL
McMahon a. 6 M, KE(CON/ECC), 8–10 3–4 [VL: CSA]
et al. (2014) 4 W (19) LP(CON/ECC),
b. 5 M, SQ(CON/ECC)c a. 80 % 8 weeks a. 25, 50, 75 %SL: "
6 W (21) 1RM
b. 55 % 3 days/ b. 25, 50, 75 %SL: "
1RM week
M Men, W Women, CON Concentric training, CON/ECC Combination of concentric and eccentric
training, ISOK Isokinetic training, KE Knee extension, LP Leg press, SQ Squat, RM Repetition
maximum, CSA Cross-sectional area, SL Segment length, " Significant increase, n.s. Not signif-
icant, PI The proximal insertion
a
Significant differences in relative change in CSA were found among regions, but details were not
described
b
CSA increased significantly without the interaction involving muscle (VI, VL, VM, and RF) or
region (33, 50, 67 %SL)
c
The range of motion of the knee joint was restricted to from 50 to 0 (a) and 90 –40 (b)

of the Hamstrings
A few studies evaluated training-induced regional changes in CSA of the ham-

strings (Housh et al. 1992; Hudelmaier et al. 2010; Bloomquist et al. 2013).
Hudelmaier et al. (2010) studied changes in hamstrings CSA in middle-aged
women following resistance training involving various exercise modalities for the
lower extremity. The CSA was measured at 10 % intervals from 10 % to 90 %SL,
and increased significantly only from 20 % to 60 %SL. Bloomquist et al. (2013)
compared the effects of shallow (0–60 of knee joint flexion) or deep (0–120 of
166 T. Wakahara
knee joint flexion) squat training on hamstrings CSA in young men. In the deep
squat training group, hamstrings CSA increased significantly only in the second
most proximal region among six regions. In the shallow squat training group,
hamstrings CSA did not change in any region. Housh et al. (1992) determined
hypertrophic changes in CSA for each of the hamstrings (long head of biceps
femoris, semitendinosus, and semimembranosus) at 33 % (only long head of biceps
femoris and semitendinosus), 50 %, and 67 %SL induced by isokinetic training in
young men. The CSA of the long head of biceps femoris increased significantly
only at 50 %SL. In contrast, CSA of the semitendinosus increased significantly in
all three regions, and CSA of the semimembranosus did not change in either region.
Although some studies examined training-induced regional changes in the thick-
ness of the posterior thigh (Starkey et al. 1996; Abe et al. 2000a), it is difficult to
compare CSA and thickness, because the thickness of the posterior thigh includes
other muscles, such as the adductors, in addition to the hamstrings.

of the Triceps Brachii
A number of studies have evaluated training-induced increases in the triceps brachii

size along its length in young men. Some studies demonstrated regional differences
(Housh et al. 1992; Kawakami et al. 1995; Kanehisa et al. 2002; Wakahara
et al. 2012, 2013), but others did not (Matta et al. 2011; Ogasawara et al. 2013).
Kawakami et al. (1995) and Kanehisa et al. (2002) observed significant increases in
triceps brachii CSA in the middle regions, but not in the proximal or distal regions,
following dynamic (Kawakami et al. 1995) and isometric (Kanehisa et al. 2002)
elbow extension training. Housh et al. (1992) found a significant increase in triceps
brachii CSA at 55 % and 70 %SL, but not at 85 %SL, after isokinetic elbow
extension training. Wakahara et al. (2012) investigated changes in triceps brachii
CSA in 13 regions along its length following dynamic elbow extension training.
The CSA increased significantly in all 13 regions, and the relative increases in CSA
were significantly greater in the proximal to middle regions than in the distal
regions. This hypertrophic pattern along the length was quite different from the
finding of Wakahara et al. (2013), who studied the effect of resistance training with
a dumbbell press-type movement on triceps brachii CSA in five regions along the
length. The CSA increased significantly in all regions except for the most proximal
region, and relative increases in CSA were significantly greater in the three middle
regions than in the most proximal region. Because the subject populations were
similar, and because the program variables of resistance training were the same
except for the exercise modality [single-joint (lying triceps extension, Wakahara
et al. 2012) vs. multi-joint (dumbbell press-type movement, Wakahara et al. 2013)
exercises], the results indicate an impact of exercise modality on the hypertrophic
pattern along the length of the triceps brachii.
14.2.5 Hypertrophic Changes Along the Length of the Elbow

Flexors
Due to difficulty delineating the individual elbow flexor muscles, most researchers
have evaluated the size of the entire group. Housh et al. (1992) showed that, in
young men, CSA of elbow flexors increased significantly in all three regions (55 %,
70 %, and 85 %SL) after isokinetic elbow flexion training. Similar results were also
reported by Farthing and Chilibeck (2003), who studied the effects of contraction
type and speed of isokinetic training on the thickness of elbow flexors in three
regions (6 cm proximal to 67 %SL, 67 %SL, and 6 cm distal to 67 %SL) in young
men and women. Roman et al. (1993) examined the effect of dynamic resistance
training on elbow flexor CSA evaluated from consecutive slices of magnetic
resonance images in elderly men. Although a statistical analysis was not performed,
their results indicated that hypertrophy occurred mainly around the belly (middle to
distal regions). Yasuda et al. (2012) investigated changes in elbow flexor CSA in
two regions (50 %SL and 10 cm above the elbow joint) in young men following
low-intensity concentric or eccentric training with blood flow restriction. Concen-
tric training significantly increased CSA in both regions, whereas eccentric training
increased CSA only at 10 cm proximal to the elbow joint, which is around the belly
of the muscle group. Matta et al. (2011) measured the thickness of the biceps brachii
in three regions (50 %, 60 %, and 70 %SL) in young men before and after dynamic
resistance training. They observed a significant interaction between time and
region, indicating that training-induced changes in biceps brachii thickness were
greater at 50 %SL than at 70 %SL.

of the Pectoralis Major
Ogasawara et al. (2013) studied hypertrophic gains in the pectoralis major muscle
in young men following 24 weeks of periodic or continuous bench press training.
The change in pectoralis major CSA was not different among the three regions
(25 %, 50 %, or 75 %) in either training group.
Taken together, the hypertrophic pattern along the length is inconsistent for the
QF and for each component of the QF except for the RF, which shows preferential
hypertrophy in the distal region. For the other muscle(s), the number of studies is
too small to draw any conclusions about their hypertrophic pattern. Although
several studies suggest factors that influence region specificity of muscle hypertro-
phy, these factors do not explain all of the data.
168 T. Wakahara
14.3 Possible Causes of Nonuniform Muscle Hypertrophy
Differences in muscle activation and contractile protein synthesis have been pro-
posed as possible causes of different hypertrophic responses along the muscle
length (Narici et al. 1996). It was reported that muscle activation during a training
session differs among muscles belonging to the same group (Narici et al. 1996;
Escamilla et al. 1998). In addition, muscle activation can also vary among regions
within a muscle (Kinugasa et al. 2005; Segal and Song 2005; Miyamoto
et al. 2012). This regional difference in activation has been suggested to be related
to subdivisions of a muscle, known as neuromuscular compartments (subdivisions
within a muscle can be defined according to their architecture, innervation, and/or
histochemical composition, Segal et al. 1991). Although evidence for neuromus-
cular compartments in humans is limited to several muscles, such partitioning
might explain differential activation within a muscle. The association between
regional differences in muscle activation and hypertrophy was investigated by
Wakahara et al. (2012, 2013) using transverse relaxation time (T2) in magnetic
resonance imaging as an index of muscle activation. Wakahara et al. (2012) dem-
onstrated that muscle activation in a session of lying triceps extension was signif-
icantly greater in the proximal to middle regions of the triceps brachii than in the
distal region. Consistent with this, the relative increase in CSA following 12 weeks
of the training was also significantly greater in the proximal to middle regions than
in the distal region. In contrast, both muscle activation in a session of dumbbell
press-type movement and the hypertrophy induced by this training were signifi-
cantly greater in the middle to distal regions of the triceps brachii than in the
proximal region (Wakahara et al. 2013). These results suggest that nonuniform
muscle hypertrophy is, at least in part, caused by regional differences in muscle
activation. Regarding protein synthesis, its increase in response to mechanical
overload was shown to be dependent on muscle fiber types in mouse plantaris
(Goodman et al. 2012). Moreover, increases in fiber CSA after 10 days of overload
were also fiber type-dependent, with patterns similar to the increases in protein
synthesis (Goodman et al. 2012). In human muscles, hypertrophic changes in
muscle fiber CSA following resistance training differ between type I and II fibers
(Aagaard et al. 2001; Kuno et al. 1990). Because different muscles contain different
proportions of type I and II fibers (Johnson et al. 1973), fiber-type dependent protein
synthesis might play a role in the nonuniform hypertrophy within a muscle group.
Concerning the fiber type composition within a muscle, Elder et al. (1982) have
reported variable fiber type composition along the length of human vastus lateralis,
soleus, and each head of the biceps and triceps brachii, although Lexell et al. (1983)
found a similar fiber type composition along the length of the vastus lateralis.
Hence, regional differences in hypertrophy within a muscle might also be related
to fiber-type dependent protein synthesis. However, if different rates of protein
synthesis had been the major cause of nonuniform hypertrophy, the hypertrophic
pattern along the muscle length would have been specific to the target muscle,
irrespective of program variables of training. The highly variable hypertrophic
patterns reported in almost all muscles suggest that differences in protein synthesis
are not the major factor causing nonuniform muscle adaptation, although the
possibility of different protein synthesis rates within a muscle cannot be ruled out.
14.4 Functional Significance of Nonuniform Muscle

Hypertrophy
Muscle size has often been evaluated by an anatomical CSA or thickness deter-
mined in just one region. This approach was validated as an assessment of muscle
size in cross-sectional (Morse et al. 2007) and longitudinal (Popadic Gacesa
et al. 2011) studies. However, studies cited in this article showed inconsistent
hypertrophic patterns along the muscle length for almost all muscles examined,
indicating that a single-slice CSA or thickness can provide an inappropriate mea-
sure of muscle size. Actually, Roman et al. (1993) reported that the relative increase
in a single-slice CSA around the belly of elbow flexors (22.6 %) was different from
that of muscle volume (13.9 %). Therefore, it is preferable to determine muscle size
in multiple regions to adequately evaluate muscle hypertrophy.
The nonuniform hypertrophy of a muscle group may affect the joint angle-torque
(force) relationship. This relationship is comprised of the individual length-force
relationships of each component in the group. Hence, the different magnitude of
hypertrophy of these components could alter the shape of the angle-torque rela-
tionship as well as the optimum angle at which maximal joint torque is exerted. This
possibility was not substantiated in the report of Narici et al. (1996), who examined
changes in CSA of each component of the QF and the joint angle-torque relation-
ship for knee extension following 6 months of resistance training. Although there
was variable hypertrophy in each of QF, no shift in the optimum angle was found
for the angle-torque relationship. The authors stated that the results were due to the
fact that either no change in the length-force relationship of any of the components
occurred, or that the changes were not sufficiently large to modify the overall angle-
torque relationship. If the magnitude and/or pattern of hypertrophy is different from
that of Narici et al. (1996), there is still a possibility that the joint angle-torque
relationship is modified by nonuniform muscle hypertrophy.
Nonuniform muscle hypertrophy alters the distribution of muscle mass within a
segment, possibly affecting inertial properties such as center of mass and moment
of inertia of the segment. The distribution of muscle mass within a segment might
be associated with performance in certain sports. It has been reported that male
sprinters had significantly greater muscle thickness of the thigh anterior portion
than distance runners and untrained controls at 30 % and 50 %SL, but not at 70 %
SL (Abe et al. 2000b). Kumagai et al. (2000) investigated thigh muscle thickness in
male sprinters divided into two groups by their personal best time in a 100-m race
(S10: 10.00–10.90 s, S11: 11.00–11.70 s). The muscle thickness of S10 was
significantly greater than S11 in the proximal region (at 30 %SL of the anterior
170 T. Wakahara
portion and 50 %SL of the posterior portion), but not different in the distal region
(at 50 % or 70 %SL of the anterior portion, or 70 %SL of the posterior portion).
Furthermore, they reported that muscle thickness in the proximal region (at 30 %SL
of the anterior portion and at 50 %SL of the posterior portion) was significantly
correlated with 100-m sprint time. These findings suggest the importance of muscle
mass distribution in sprint performance.
14.5 Future Perspectives
This review demonstrates that resistance training-induced hypertrophic changes are

highly variable along the length of a muscle or muscle group. The mechanisms for
this nonuniform muscular adaptation remain unclear. Future studies should identify
the factors that determine the hypertrophic pattern along the muscle length. Find-
ings from such studies would contribute to the design of training programs that take
into account region-specific muscle hypertrophy.
References
Aagaard P, Andersen JL, Dyhre-Poulsen P, Leffers AM, Wagner A, Magnusson SP, Halkjaer-
Kristensen J, Simonsen EB (2001) A mechanism for increased contractile strength of human
pennate muscle in response to strength training: changes in muscle architecture. J Physiol
534:613–623
Abe T, DeHoyos DV, Pollock ML, Garzarella L (2000a) Time course for strength and muscle
thickness changes following upper and lower body resistance training in men and women. Eur J
Abe T, Kumagai K, Brechue WF (2000b) Fascicle length of leg muscles is greater in sprinters than
distance runners. Med Sci Sports Exerc 32:1125–1129
Ahtiainen JP, Pakarinen A, Alen M, Kraemer WJ, Häkkinen K (2003) Muscle hypertrophy,
hormonal adaptations and strength development during strength training in strength-trained
and untrained men. Eur J Appl Physiol 89:555–563
Blazevich AJ, Cannavan D, Coleman DR, Horne S (2007) Influence of concentric and eccentric
resistance training on architectural adaptation in human quadriceps muscles. J Appl Physiol
103:1565–1575
Bloomquist K, Langberg H, Karlsen S, Madsgaard S, Boesen M, Raastad T (2013) Effect of range
of motion in heavy load squatting on muscle and tendon adaptations. Eur J Appl Physiol
113:2133–2142. doi:10.1007/s00421-013-2642-7
Elder GC, Bradbury K, Roberts R (1982) Variability of fiber type distributions within human
muscles. J Appl Physiol 53:1473–1480
Ema R, Wakahara T, Miyamoto N, Kanehisa H, Kawakami Y (2013) Inhomogeneous architectural
changes of the quadriceps femoris induced by resistance training. Eur J Appl Physiol
113:2691–2703. doi:10.1007/s00421-013-2700-1
Escamilla RF, Fleisig GS, Zheng N, Barrentine SW, Wilk KE, Andrews JR (1998) Biomechanics
of the knee during closed kinetic chain and open kinetic chain exercises. Med Sci Sports Exerc
30:556–569
Farthing JP, Chilibeck PD (2003) The effects of eccentric and concentric training at different
velocities on muscle hypertrophy. Eur J Appl Physiol 89:578–586
Fukunaga T, Miyatani M, Tachi M, Kouzaki M, Kawakami Y, Kanehisa H (2001) Muscle volume
is a major determinant of joint torque in humans. Acta Physiol Scand 172:249–255
Goodman CA, Kotecki JA, Jacobs BL, Hornberger TA (2012) Muscle fiber type-dependent
differences in the regulation of protein synthesis. PLoS One 7:e37890. doi:10.1371/journal.
pone.0037890
Häkkinen K, Pakarinen A, Kraemer WJ, Häkkinen A, Valkeinen H, Alen M (2001) Selective
muscle hypertrophy, changes in EMG and force, and serum hormones during strength training
in older women. J Appl Physiol 91:569–580
Häkkinen K, Pakarinen A, Hannonen P, Häkkinen A, Airaksinen O, Valkeinen H, Alen M (2002)
Effects of strength training on muscle strength, cross-sectional area, maximal electromyo-
graphic activity, and serum hormones in premenopausal women with fibromyalgia. J
Rheumatol 29:1287–1295
Häkkinen K, Alen M, Kraemer WJ, Gorostiaga E, Izquierdo M, Rusko H, Mikkola J, Häkkinen A,
Valkeinen H, Kaarakainen E, Romu S, Erola V, Ahtiainen J, Paavolainen L (2003) Neuro-
muscular adaptations during concurrent strength and endurance training versus strength train-
ing. Eur J Appl Physiol 89:42–52
Higbie EJ, Cureton KJ, Warren GL 3rd, Prior BM (1996) Effects of concentric and eccentric
training on muscle strength, cross-sectional area, and neural activation. J Appl Physiol
81:2173–2181
Hisaeda H, Miyagawa K, Kuno S, Fukunaga T, Muraoka I (1996) Influence of two different modes
of resistance training in female subjects. Ergonomics 39:842–852
Housh DJ, Housh TJ, Johnson GO, Chu WK (1992) Hypertrophic response to unilateral concentric
isokinetic resistance training. J Appl Physiol 73:65–70
Housh DJ, Housh TJ, Weir JP, Weir LL, Evetovich TK, Donlin PE (1998) Effects of unilateral
concentric-only dynamic constant external resistance training on quadriceps femoris cross-
sectional area. J Strength Cond Res 12:185–191
Hudelmaier M, Wirth W, Himmer M, Ring-Dimitriou S, Sänger A, Eckstein F (2010) Effect of
exercise intervention on thigh muscle volume and anatomical cross-sectional areas – quanti-
tative assessment using MRI. Magn Reson Med 64:1713–1720. doi:10.1002/mrm.22550
Ikai M, Fukunaga T (1968) Calculation of muscle strength per unit cross-sectional area of human
muscle by means of ultrasonic measurement. Int Z Angew Physiol 26:26–32
Johnson MA, Polgar J, Weightman D, Appleton D (1973) Data on the distribution of fibre types in
thirty-six human muscles: an autopsy study. J Neurol Sci 18:111–129
Kanehisa H, Nagareda H, Kawakami Y, Akima H, Masani K, Kouzaki M, Fukunaga T (2002)
Effects of equivolume isometric training programs comprising medium or high resistance on
muscle size and strength. Eur J Appl Physiol 7:112–119
Kanehisa H, Funato K, Kuno S, Fukunaga T, Katsuta S (2003) Growth trend of the quadriceps
femoris muscle in junior Olympic weight lifters: an 18-month follow-up survey. Eur J Appl
Physiol 89:238–242
Kanehisa H, Kuno S, Katsuta S, Fukunaga T (2006) A 2-year follow-up study on muscle size and
dynamic strength in teenage tennis players. Scand J Med Sci Sports 16:93–101
Kawakami Y, Abe T, Kuno SY, Fukunaga T (1995) Training-induced changes in muscle archi-
tecture and specific tension. Eur J Appl Physiol Occup Physiol 72:37–43
Kinugasa R, Kawakami Y, Fukunaga T (2005) Muscle activation and its distribution within human
triceps surae muscles. J Appl Physiol 99:1149–1156
Kraemer WJ, Ratamess NA (2004) Fundamentals of resistance training: progression and exercise
prescription. Med Sci Sports Exerc 36:674–688
Krotkiewski M, Aniansson A, Grimby G, Bj€ orntorp P, Sj€
ostr€
om L (1979) The effect of unilateral
isokinetic strength training on local adipose and muscle tissue morphology, thickness, and
enzymes. Eur J Appl Physiol Occup Physiol 42:271–281
172 T. Wakahara
Kumagai K, Abe T, Brechue WF, Ryushi T, Takano S, Mizuno M (2000) Sprint performance is
related to muscle fascicle length in male 100-m sprinters. J Appl Physiol 88:811–816
Kuno S, Katsuta S, Akisada M, Anno I, Matsumoto K (1990) Effect of strength training on the
relationship between magnetic resonance relaxation time and muscle fibre composition. Eur J
Appl Physiol Occup Physiol 61:33–36
Lexell J, Henriksson-Larsén K, Sj€ostr€
om M (1983) Distribution of different fibre types in human
skeletal muscles. 2. A study of cross-sections of whole m. vastus lateralis. Acta Physiol Scand
117:115–122
Matta T, Simão R, de Salles BF, Spineti J, Oliveira LF (2011) Strength training’s chronic effects
on muscle architecture parameters of different arm sites. J Strength Cond Res 25:1711–1717.
doi:10.1519/JSC.0b013e3181dba162
Maughan RJ, Watson JS, Weir J (1983) Strength and cross-sectional area of human skeletal
muscle. J Physiol 338:37–49
McMahon G, Morse CI, Burden A, Winwood K, Onambélé GL (2014) Muscular adaptations and
insulin-like growth factor-I (IGF-I) responses to resistance training are stretch-mediated.
Muscle Nerve 49:108–119. doi:10.1002/mus.23884
Melnyk JA, Rogers MA, Hurley BF (2009) Effects of strength training and detraining on regional
muscle in young and older men and women. Eur J Appl Physiol 105:929–938. doi:10.1007/
s00421-008-0979-0
Miyamoto N, Wakahara T, Kawakami Y (2012) Task-dependent inhomogeneous muscle activities
within the bi-articular human rectus femoris muscle. PLoS One 7:e34269. doi:10.1371/journal.
pone.0034269
Morse CI, Degens H, Jones DA (2007) The validity of estimating quadriceps volume from single
MRI cross-sections in young men. Eur J Appl Physiol 100:267–274
Narici MV, Roi GS, Landoni L, Minetti AE, Cerretelli P (1989) Changes in force, cross-sectional
area and neural activation during strength training and detraining of the human quadriceps. Eur
J Appl Physiol Occup Physiol 59:310–319
Narici MV, Hoppeler H, Kayser B, Landoni L, Claassen H, Gavardi C, Conti M, Cerretelli P
(1996) Human quadriceps cross-sectional area, torque and neural activation during 6 months
strength training. Acta Physiol Scand 157:175–186
Ogasawara R, Yasuda T, Ishii N, Abe T (2013) Comparison of muscle hypertrophy following
6-month of continuous and periodic strength training. Eur J Appl Physiol 113:975–985. doi:10.
1007/s00421-012-2511-9
Petersen SG, Beyer N, Hansen M, Holm L, Aagaard P, Mackey AL, Kjaer M (2011) Nonsteroidal
anti-inflammatory drug or glucosamine reduced pain and improved muscle strength with
resistance training in a randomized controlled trial of knee osteoarthritis patients. Arch Phys
Med Rehabil 92:1185–1193. doi:10.1016/j.apmr.2011.03.009
Popadic Gacesa J, Dragnic NR, Prvulovic NM, Barak OF, Grujic N (2011) The validity of
estimating triceps brachii volume from single MRI cross-sectional area before and after
resistance training. J Sports Sci 29:635–641. doi:10.1080/02640414.2010.549498
Ratamess NA, Alvar BA, Evetoch TE, Housh TJ, Ben Kibler W, Kraemer WJ, Triplett NT (2009)
Progression models in resistance training for healthy adults. Med Sci Sports Exerc 41:687–708.
doi:10.1249/MSS.0b013e3181915670
Reeves ND, Narici MV, Maganaris CN (2004) Effect of resistance training on skeletal muscle-
specific force in elderly humans. J Appl Physiol 96:885–892
Roman WJ, Fleckenstein J, Stray-Gundersen J, Alway SE, Peshock R, Gonyea WJ (1993)
Adaptations in the elbow flexors of elderly males after heavy-resistance training. J Appl
Physiol 74:750–754
Schoenfeld BJ (2010) The mechanisms of muscle hypertrophy and their application to resistance
training. J Strength Cond Res 24:2857–2872. doi:10.1519/JSC.0b013e3181e840f3
Schott J, McCully K, Rutherford OM (1995) The role of metabolites in strength training. II. Short
versus long isometric contractions. Eur J Appl Physiol Occup Physiol 71:337–341
Segal RL, Song AW (2005) Nonuniform activity of human calf muscles during an exercise task.
Arch Phys Med Rehabil 86:2013–2017
Segal RL, Wolf SL, DeCamp MJ, Chopp MT, English AW (1991) Anatomical partitioning of three
multiarticular human muscles. Acta Anat 142:261–266
Seger JY, Arvidsson B, Thorstensson A (1998) Specific effects of eccentric and concentric training
on muscle strength and morphology in humans. Eur J Appl Physiol Occup Physiol 79:49–57
Seynnes OR, de Boer M, Narici MV (2007) Early skeletal muscle hypertrophy and architectural
changes in response to high-intensity resistance training. J Appl Physiol 102:368–373
Smith RC, Rutherford OM (1995) The role of metabolites in strength training. I. A comparison of
eccentric and concentric contractions. Eur J Appl Physiol Occup Physiol 71:332–336
Starkey DB, Pollock ML, Ishida Y, Welsch MA, Brechue WF, Graves JE, Feigenbaum MS (1996)
Effect of resistance training volume on strength and muscle thickness. Med Sci Sports Exerc
28:1311–1320
Tracy BL, Ivey FM, Hurlbut D, Martel GF, Lemmer JT, Siegel EL, Metter EJ, Fozard JL, Fleg JL,
Hurley BF (1999) Muscle quality. II. Effects of strength training in 65- to 75-yr-old men and
women. J Appl Physiol 86:195–201
Wakahara T, Miyamoto N, Sugisaki N, Murata K, Kanehisa H, Kawakami Y, Fukunaga T, Yanai
T (2012) Association between regional differences in muscle activation in one session of
resistance exercise and in muscle hypertrophy after resistance training. Eur J Appl Physiol
112:1569–1576. doi:10.1007/s00421-011-2121-y
Wakahara T, Fukutani A, Kawakami Y, Yanai T (2013) Nonuniform muscle hypertrophy: its
relation to muscle activation in training session. Med Sci Sports Exerc 45:2158–2165. doi:10.
1249/MSS.0b013e3182995349
Wernbom M, Augustsson J, Thomeé R (2007) The influence of frequency, intensity, volume and
mode of strength training on whole muscle cross-sectional area in humans. Sports Med
37:225–264
Yasuda T, Loenneke JP, Thiebaud RS, Abe T (2012) Effects of blood flow restricted low-intensity
concentric or eccentric training on muscle size and strength. PLoS One 7:e52843. doi:10.1371/
journal.pone.0052843
Chapter 15
Quantitative Profiles of the Quadriceps
Femoris in Sport Athletes
Ryoichi Ema and Yasuo Kawakami
Abstract This brief review summarizes available evidence on the quantitative

profiles of the quadriceps femoris in sports athletes. It is generally believed that
sports athletes show event-related morphological profiles in muscle size that are
specific to their competitive and training activities. Many previous analyses of
cross-sectional data have indicated that athletes have greater size of the quadriceps
femoris as compared to untrained controls. However, little is known about which
events have athletes with the greatest degree of quadriceps femoris. In addition,
longitudinal studies on the adaptation of the quadriceps are few, and there is a
paucity of quantitative evidence on adaptations of the quadriceps femoris induced
by regular training for competitive activities. Only a limited number of studies have
been carried out to examine the association between quadriceps femoris size and
sport performance, e.g., sprinters, weightlifters and cyclists, with the conclusions
controversial and inconsistent across studies. Furthermore, little is known regarding
inter-muscle and/or intra-muscle differences in the hypertrophic response, either in
cross-sectional or longitudinal studies. There is however strong evidence that
resistance training-induced changes in muscle size do not occur evenly among
the four muscles of the quadriceps femoris, nor along or across the same muscle.
Further studies will be needed if we are to develop an understanding of the
mechanisms leading to the many differences seen in the quadriceps femoris, both
within and across athletes.
Keywords Muscle volume • Anatomical cross-sectional area • Muscle thickness •

Magnetic resonance imaging • Ultrasonography
R. Ema
Research Fellow of Japan Society for the Promotion of Science, Tokyo, Japan
Y. Kawakami (*)
e-mail: ykawa@waseda.jp

176 R. Ema and Y. Kawakami
15.1 Introduction
Muscle architectural parameters, such as anatomical cross-sectional area (ACSA),

muscle thickness, fascicle length, and pennation angle, are determinants of muscle
function during human movements. Therefore, many studies have examined these
parameters in athletes (Abe et al. 2000; Kanehisa et al. 2003a; Kearns et al. 2000) as
well as in ordinary individuals (Blazevich et al. 2006; Ema et al. 2013b; Kawakami
et al. 1998). Above all, it is generally believed that sports athletes show event-
related morphological profiles of muscle size that are specific to their particular
competitive and training activities. Therefore, examination of the quantitative
characteristics of muscles in various athletes will help to clarify the relationship
between muscle size and sport-specific motions, and thus to aid in the design of
effective training programs for athletes and their coaches.
Muscle size has been evaluated in sport athletes as well as ordinary individuals
through tissue-visualizing techniques such as ultrasonography (Ikai and Fukunaga
1968), computed tomography (Schantz et al. 1983), and magnetic resonance
(MR) imaging (Narici et al. 1989). One of the most important muscles for sport
motions is the quadriceps femoris. Most human motions, including but not limited
to walking, running, jumping, and cycling necessarily activate the quadriceps
femoris. The quadriceps femoris is comprised of four muscles: the vastus lateralis,
vastus medialis, vastus intermedius, and rectus femoris. The three vasti are
monoarticular muscles, crossing the knee joint. The remaining muscle, the rectus
femoris, is a biarticular muscle crossing both the knee and hip joints. Hence, the
primary functions of the quadriceps femoris are knee extension and hip flexion. In
this chapter we summarize previous reports regarding the quantitative profiles of
the human quadriceps femoris in sports athletes (Table 15.1).
15.2 Cross-Sectional Study
Many studies have compared quantitative profiles of the quadriceps femoris across
different types of athletes as well as between athletes and untrained controls. In the
1980s, a number of studies measured quadriceps femoris size in athletes, such as
sprinters, marathon runners (Johansson et al. 1987; Maughan et al. 1983), and
bodybuilders (Schantz et al. 1983). Tsunoda et al. (1986) was the first to report
quadriceps femoris size in several kinds of sports and compared them to those of
untrained controls. They measured the ACSA of the quadriceps femoris at
mid-thigh by using ultrasonography on 89 male Japanese elite athletes and
14 untrained men. In their results, the greatest ACSA was observed in sumo
wrestlers, while those of sprinters and long distance runners were similar to those
of untrained controls. This study indicates that all competitive sport athletes do not
have hypertrophied quadriceps femoris. In addition, they also reported the relative
ACSA for each muscle to the total quadriceps femoris ACSA. Their data clearly
15 Quantitative Profiles of the Quadriceps Femoris in Sport Athletes 177
Table. 15.1 Studies regarding the quantitative profiles of the quadriceps femoris in athletes
Measurement
Study Subjects parameters Methods Results
Abe Sprinters MT of the vastus US Sprinters > Distance
et al. (2000) Distance lateralis runners
runners MT of the rectus Untrained
Untrained femoris + vastus controls
controls intermedius
Abe Sprinters MT of the vastus US Sprinters > Untrained
et al. (2001) Untrained lateralis controls
controls MT of the rectus
femoris + vastus
intermedius
Akima All Japan soc- ACSA of the MRI All Japan > Japan
et al. (1992) cer players total quadriceps league
Olympic femoris players
representatives ACSA for each
Japan league muscle
players
D’Antona Bodu builders Muscle volume MRI Body > Untrained
et al. (2006) Untrained of the total quad- builders controls
controls riceps femoris
ACSA of the
vastus lateralis
Ema Oarsmen Muscle volume MRI Oarsmen > Untrained
et al. (2014) Untrained for each muscle controls
controls ACSA for each Volume of the rectus femoris
muscle was similar
Funato Elite ACSA of the US No difference between the
et al. (2000) weightlifters total quadriceps two groups
College femoris
weightlifters
Hoshikawa Several kinds ACSA of the MRI Throwers and sumo wrestlers
et al. (2010) of athletes total quadriceps had greater ACSA than those
femoris of other athletes
Hug Road cyclists ACSA of the MRI Road > Sport
et al. (2006) Sport science total quadriceps cyclists Science
students femoris students
ACSA for each
muscle
Ikebukuro Weightlifters MT for each US Weightlifters > Untrained
et al. (2011) Sprinters muscle controls
Untrained Sprinters > Untrained
controls controls
(continued)
Table. 15.1 (continued)

Measurement
Study Subjects parameters Methods Results
Izquierdo Weightlifters ACSA of the US Weightlifters > Road
et al. (2004) Road cyclists total quadriceps cyclists
Untrained femoris Untrained
controls controls
Johansson Sprinters ACSA of the CT Sprinters > Marathon
et al. (1987) Marathon total quadriceps runners
runners femoris
ACSA of the
vastus lateralis
Kanda American Muscle volume MRI Line groups > Skill
et al. (2013) football for each muscle groups
players
Line groups
Skill groups
Kanehisa Weightlifters ACSA of the US Weightlifters > Untrained
et al. (1998a) Untrained total quadriceps controls
controls femoris
Kanehisa Weightlifters ACSA of the US Weightlifters > Wrestlers
et al. (1998b) Wrestlers total quadriceps
femoris
Kanehisa Soccer players MT of the vastus US Swimmers > Soccer
et al. (2003a) Swimmers lateralis players
Kubo Youth soccer ACSA of the MRI Professional > Youth
et al. (2010) players total quadriceps
Professional femoris
players
Kearns Sumo wrestlers MT of the vastus US Sumo > Untrained
et al. (2000) Untrained lateralis wrestlers controls
controls MT of the rectus
femoris + vastus
intermedius
Maughan Sprinters ACSA of the CT No difference between the
et al. (1983) Distance total quadriceps two groups
runners femoris
Tsunoda Several kinds ACSA for each US Sumo wrestlers had the
et al. (1986) of athletes muscle greatest ACSA
MT muscle thickness, ACSA anatomical cross-sectional area, US ultrasonography, CT computed
tomography, MRI magnetic resonance imaging
demonstrated event-related profiles for each muscle: for example, the percentage of
the rectus femoris to the total quadriceps femoris ACSA was higher in the soccer
players than in the volleyball players, long distance runners, oarsmen, and untrained
individuals. However, they reported the ACSA only at one region (mid-thigh level);
hence possible differences in hypertrophy among the four muscles at different
regions was not taken into account.
In addition to ultrasonography, after the 1990s, MR imaging was introduced to
evaluate the ACSA in many studies. This advance added a large field of view and
high resolution. Some showed the difference of the quadriceps femoris size among
the performance levels (Akima et al. 1992; Kubo et al. 2010) and among positions
in the same sport (Kanda et al. 2013). For example, Akima et al. (1992) compared
the ACSA of the quadriceps femoris at three (proximal, middle, distal) thigh levels
for different levels of competitive play. The clearest results were for the comparison
between the Japan national team players, a very elite group, and the Japan profes-
sional league players, a less elite group. The anatomical measurements indicated
that at all three levels of the quadriceps femoris, the ACSA was larger for the Japan
national team players. This study suggests that one of the important factors for good
performance in competitive soccer is the size of the quadriceps femoris, although
the underlying mechanisms for the greater ACSA in the Japan national team players
remain unclear. On the other hand, Kano et al. (1997) failed to show a relationship
between the ACSA of the quadriceps femoris and 100 m sprint time, suggesting that
at least for sprinters, an increase in the ACSA of the quadriceps femoris does not
lead to an improvement in sprint performance. Recent studies have supported this
result (Hoshikawa et al. 2006b; Sugisaki et al. 2011). However, some studies did
show a greater muscle thickness of the vastus lateralis (Abe et al. 2000, 2001) and
vastus medialis (Ikebukuro et al. 2011) in sprinters as compared to untrained
controls, with a significant association between sprint time and muscle thickness
of the vastus medialis and vastus intermedius relative to body mass1/3 (Ikebukuro
et al. 2011). Therefore, no consensus has been reached regarding whether sprinters
have an overall hypertrophied quadriceps femoris as compared to untrained
controls.
Details of the muscular profiles in cyclists are also controversial. Hug
et al. (2006) showed a greater ACSA of the quadriceps femoris in professional
road cyclists as compared to recreationally active students. On the other hand,
Izquierdo et al. (2004) reported a similar ACSA of the quadriceps femoris in
comparisons between top level amateur road cyclists and untrained controls. The
reasons for the inconsistency between these studies are unclear, but both studies
measured ACSA of the total quadriceps femoris at one region. However, the ACSA
of the total quadriceps femoris at one region does not represent the precise muscle
volume for the entire muscle (Morse et al. 2007), and hence a similar evaluation
may not have been done in both studies.
One of the athletic events in which athletes have remarkably hypertrophied

quadriceps femoris is weightlifting (Ikebukuro et al. 2011; Izquierdo et al. 2004;
Kanehisa et al. 1998a, b; Funato et al. 2000). For example, Kanehisa et al. (1998b)
showed that Olympic weightlifters had greater (about 50 %) ACSA of the total
quadriceps femoris than those of untrained controls. Moreover, recent investigation
suggested uneven hypertrophy among the four muscles of the quadriceps femoris in
weightlifters (Ikebukuro et al. 2011). The weightlifters had a greater muscle
thickness in the vastus lateralis, vastus medialis and vastus intermedius as com-
pared to those of untrained controls, but that of the rectus femoris was similar
between the two groups. Competitive weight lifting mainly consists of explosive
leg extensions (simultaneous extensions of knee and hip joints). Therefore,
it can be assumed that the quantitative profile in weightlifters was due to leg
extension exercises in their competitive and training activities. In a recent study
we examined quantitative muscle profiles in oarsmen and the data supported this
idea (Ema et al. 2014). We compared the muscle volume of the individual muscles
of the quadriceps femoris between experienced oarsmen and untrained controls.
It was shown that experienced oarsmen had about a 30 % greater volume of
the vastus lateralis, vastus medialis, vastus intermedius, but rectus femoris volume
was similar between the two groups. In addition to our cross-sectional studies,
previous functional studies (Chin et al. 2011; Escamilla et al. 1998, 2001) also
supported the above results. For example, Escamilla et al. (1998) determined
muscle activation level during squat and leg press exercises using electromyogra-
phy. The results indicated that the activity level in the vastus lateralis and vastus
medialis was 30–90 % greater than that in the rectus femoris. It was recently shown
that differences in muscle activation over synergistic muscles during resistance
training was associated with the differences in the magnitude of muscle hypertro-
phy (Wakahara et al. 2012). Given the above studies, it is likely that differences in
muscle activation during leg extension may lead to the quantitative profiles of the
quadriceps femoris in weightlifters. However, the underlying mechanisms for
the difference in muscle activation among the four muscles during leg extension
is not clear.
One interesting topic involves establishing the particular event in which athletes
have the greatest quadriceps femoris. It is very difficult to provide the answer for
this question, because only a few studies have compared the size of the quadriceps
femoris across different athletes (Hoshikawa et al. 2010; Tsunoda et al. 1986) and
because the physical characteristics, training habits and performance levels tend to
vary over the sports events and athletes. Hoshikawa et al. (2010) showed that
throwers and sumo wrestlers had a greater ACSA of the total quadriceps femoris
than those of volleyball players, soccer players, oarsmen, karate athletes, and
sprinters. These findings were generally consistent with the results of Tsunoda
et al. (1986). Kearns et al. (2000) showed a greater (about 30–55 %) muscle
thickness of the quadriceps femoris in sumo wrestlers as compared to those in
untrained controls. Weightlifters had about a 50 % greater ACSA for the total
quadriceps femoris than did untrained controls (Kanehisa et al. 1998b). Moreover,
it was shown that the vastus lateralis had a 55 % greater ACSA and the total
quadriceps femoris a 34 % greater volume in bodybuilders than in untrained
controls (D’Antona et al. 2006). In the future, studies should focus on the quanti-
tative profiles of the quadriceps femoris, and compare these profiles across different
types of athletes, including power athletes, to better understand the role of the
quadriceps femoris within and across different sports.
15.3 Longitudinal Study
Compared to cross-sectional studies, longitudinal studies are particularly valuable.

Unfortunately, longitudinal studies on alterations in the quantitative profiles of the
quadriceps femoris in athletes are few. Therefore, the influence of normal training
regimens of competitive sports activities on the size of the quadriceps femoris
remains poorly understood. Kanehisa et al. (2003b) measured ACSA of the total
quadriceps femoris twice, at an interval of 18 months, in junior Olympic weight
lifters. They found preferential hypertrophy in the distal regions. Similar results
were obtained in teenage tennis players (Kanehisa et al. 2006). Hoshikawa
et al. (2006a) measured changes in the ACSA of the quadriceps femoris in high
school volleyball players after 1 year and found an increase of the ACSA of the
quadriceps femoris. However, subjects in the above studies were junior athletes,
and the effect of growth may have influenced the results. On the other hand,
Rønnestad et al. (2010) showed that regular training of competitive cycling for
12 weeks did not induce an increase of the ACSA of the total quadriceps femoris in
well-trained cyclists. The above studies are inconclusive, but do indicate that cross-
sectional observations are not currently substantiated by longitudinal studies.
Moreover, no study has evaluated the component muscle volumes to evaluate the
influence of regular training of competitive sport activities on the individual
muscles of the quadriceps femoris. Uneven hypertrophy among the four muscles
of the quadriceps femoris does exist in some athletes (Ema et al. 2014): Oarsmen
have hypertrophied vasti but not rectus femoris compared to those of untrained
individuals. This suggests that leg extension training induces hypertrophy of the
monoarticular vasti but not the biarticular rectus femoris. On the other hand,
previous training intervention studies adopting knee extension training indicated
prominent hypertrophy of the rectus femoris compared to the vasti (Ema
et al. 2013a; Housh et al. 1992; Narici et al. 1996). Based on the above findings,
we can expect unique hypertrophic responses of the components of the quadriceps
femoris to resistance exercises, depending on the mono- or bi-articularity of the
joint movements involved. Athletic event-specific muscularity of sport athletes may
provide insights into such unique adaptability of individual skeletal muscles.
15.4 The Relationship Between Quadriceps Femoris Size

and Sport Performance
Do hypertrophied quadriceps femoris really contribute to better sport performance?

Regarding this point, some researchers say “yes”, others “no”. The ACSA of the
total quadriceps femoris was associated with Wingate peak power output in well-
trained cyclists (Rønnestad et al. 2010). In their results, in addition to the above
result, the increase in ACSA of the quadriceps femoris induced by resistance
training was also correlated with an increase in Wingate peak power out. This
suggests that hypertrophic adaptation of the quadriceps femoris can improve
cycling performance in well-trained cyclists. Similarly, for competitive weight
lifters, a significant relation has been observed between performance and muscle
thickness of the quadriceps femoris relative to body mass1/3 (Ikebukuro et al. 2011).
On the other hand, previous studies have suggested a negative contribution of the
hypertrophied quadriceps femoris to sprint performance (Hoshikawa et al. 2006b;
Sugisaki et al. 2011). Hoshikawa et al. (2006b) investigated the relation between
100 m race performance (mean velocity of the 100 m sprint running) and muscle
size in junior athletes by using a stepwise multiple-regression analysis. They found
that the ACSA of the quadriceps femoris at the mid-thigh had a negative regression
coefficient in the regression model. Moreover, although statistical significance was
not reached (P ¼ 0.061), a negative trend between the 30 m sprint performance and
the ACSA of the quadriceps femoris at the mid-thigh was also observed by Sugisaki
et al. (2011). On the other hand, Ikebukuro et al. (2011) showed a positive relation
between muscle thickness of the vastus medialis and vastus intermedius and sprint
performance. The difference in measurement parameters may be related to the
inconsistency seen above. To our knowledge, there is still no study that indicates
an association between quadriceps femoris hypertrophy induced by the regular
training regimens involved with competitive sport activities and the improvement
of actual sport performance. Furthermore, no study has produced direct evidence
regarding the relation between the muscular hypertrophy and sport-specific
motions. Future research is needed to clarify this point.
15.5 Conclusion
There is a paucity of studies that provide quantitative profiles of the quadriceps

femoris in athletes. Many studies have evaluated only the ACSA of the total
quadriceps femoris at one region or muscle thickness. This has led to inconsistency
over studies and made it difficult in compare findings across studies. Moreover,
very few studies have evaluated inter-muscle and intra-muscle differences for the
hypertrophic response and few valid longitudinal studies have been performed.
Nevertheless, there is adequate information to conclude that resistance training-
induced changes in muscle size do not occur evenly among the four muscles of the
quadriceps femoris, nor along or across a particular muscle. Still, additional longi-
tudinal studies are needed to substantiate the more numerous cross-sectional obser-
vations. Future investigation into the quantitative profiles of the muscles in various
sport athletes will provide new knowledge on how training-induced adaptations of
particular muscles can lead to better sport performance.
References
Abe T, Kumagai K, Brechue WF (2000) Fascicle length of leg muscles is greater in sprinters than
distance runners. Med Sci Sports Exerc 32:1125–1129
Abe T, Fukashiro S, Harada Y, Kawamoto K (2001) Relationship between sprint performance and
muscle fascicle length in female sprinters. J Physiol Anthropol Appl Human Sci 20:141–147
Akima H, Kuno S, Nishijima T, Maruyama T, Matsumoto M, Itai Y, Shimojo H, Katsuta S (1992)
Studies on muscle metabolism and cross-sectional area in the elite Japanese soccer players
using NMR. Jpn J Phys Fit Sport 41:368–375
Blazevich AJ, Gill ND, Zhou S (2006) Intra- and intermuscular variation in human quadriceps
femoris architecture assessed in vivo. J Anat 209:289–310
Chin LM, Kowalchuk JM, Barstow TJ, Kondo N, Amano T, Shiojiri T, Koga S (2011) The
relationship between muscle deoxygenation and activation in different muscles of the quadri-
ceps during cycle ramp exercise. J Appl Physiol 111:1259–1265
D’Antona G, Lanfranconi F, Pellegrino MA, Brocca L, Adami R, Rossi R, Moro G, Miotti D,
Canepari M, Bottinelli R (2006) Skeletal muscle hypertrophy and structure and function of
skeletal muscle fibres in male body builders. J Physiol 570:611–627
Ema R, Wakahara T, Miyamoto N, Kanehisa H, Kawakami Y (2013a) Inhomogeneous architec-
tural changes of the quadriceps femoris induced by resistance training. Eur J Appl Physiol
113:2691–2703
Ema R, Wakahara T, Mogi Y, Miyamoto N, Komatsu T, Kanehisa H, Kawakami Y (2013b) In
vivo measurement of human rectus femoris architecture by ultrasonography: validity and
applicability. Clin Physiol Funct Imaging 33:267–273
Ema R, Wakahara T, Kanehisa H, Kawakami Y (2014) Inferior muscularity of the rectus femoris
to vasti in varsity oarsmen. Int J Sports Med 35:293–297
Escamilla RF, Fleisig GS, Zheng N, Barrentine SW, Wilk KE, Andrews JR (1998) Biomechanics
of the knee during closed kinetic chain and open kinetic chain exercises. Med Sci Sports Exerc
30:556–569
Escamilla RF, Fleisig GS, Zheng N, Lander JE, Barrentine SW, Andrews JR, Bergemann BW,
Moorman CT 3rd (2001) Effects of technique variations on knee biomechanics during the squat
and leg press. Med Sci Sports Exerc 33:1552–1566
Funato K, Kanehisa H, Fukunaga T (2000) Differences in muscle cross-sectional area and strength
between elite senior and college Olympic weight lifters. J Sports Med Phys Fitness 40:312–318
Hoshikawa Y, Muramatsu M, Iida T, Uchiyama A, Nakajima Y, Kanehisa H, Fukunaga T (2006a)
Gender differences in yearly changes in the cross-sectional areas and dynamic torques of thigh
muscles in high school volleyball players. Int J Sport Health Sci 4:29–35
Hoshikawa Y, Muramatsu M, Iida T, Uchiyama A, Nakajima Y, Kanehisa H, Fukunaga T (2006b)
Influence of the psoas major and thigh muscularity on 100-m times in junior sprinters. Med Sci
Sports Exerc 38:2138–2143
Hoshikawa Y, Muramatsu M, Iida T, Uchiyama A, Nakajima Y, Kanehisa H (2010) Event-related
differences in the cross-sectional areas and torque generation capabilities of quadriceps
femoris and hamstrings in male high school athletes. J Physiol Anthropol 29:13–21
Housh DJ, Housh TJ, Johnson GO, Chu WK (1992) Hypertrophic response to unilateral concentric
isokinetic resistance training. J Appl Physiol 73:65–70
Hug F, Marqueste T, Le Fur Y, Cozzone PJ, Grélot L, Bendahan D (2006) Selective training-
induced thigh muscles hypertrophy in professional road cyclists. Eur J Appl Physiol
97:591–597
Ikai M, Fukunaga T (1968) Calculation of muscle strength per unit cross-sectional area of human
muscle by means of ultrasonic measurement. Int Z Angew Physiol 26:26–32
Ikebukuro T, Kubo K, Okada J, Yata H, Tsunoda N (2011) The relationship between muscle
thickness in the lower limbs and competition performance in weightlifters and sprinters. Jpn J
Phys Fit Sports Med 60:401–411
Izquierdo M, Ibá~nez J, Häkkinen K, Kraemer WJ, Ruesta M, Gorostiaga EM (2004) Maximal
strength and power, muscle mass, endurance and serum hormones in weightlifters and road
cyclists. J Sports Sci 22:465–478
Johansson C, Lorentzon R, Sj€ ostr€
om M, Fagerlund M, Fugl-Meyer AR (1987) Sprinters and
marathon runners. Does isokinetic knee extensor performance reflect muscle size and struc-
ture? Acta Physiol Scand 130:663–669
Kanda Y, Tanaka S, Teshima T, Tsunoda N (2013) Structural and functional characteristics of the
lower limb muscles in American football players. Tokyo J Phys Educ 4:38–45
Kanehisa H, Ikegawa S, Fukunaga T (1998a) Body composition and cross-sectional areas of limb
lean tissues in Olympic weight lifters. Scand J Med Sci Sports 8:271–278
Kanehisa H, Ikegawa S, Fukunaga T (1998b) Comparison of muscle cross-sectional areas between
weight lifters and wrestlers. Int J Sports Med 19:265–271
Kanehisa H, Muraoka Y, Kawakami Y, Fukunaga T (2003a) Fascicle arrangements of vastus
lateralis and gastrocnemius muscles in highly trained soccer players and swimmers of both
genders. Int J Sports Med 24:90–95
Kanehisa H, Funato K, Kuno S, Fukunaga T, Katsuta S (2003b) Growth trend of the quadriceps
femoris muscle in junior Olympic weight lifters: an 18-month follow-up survey. Eur J Appl
Physiol 89:238–242
Kanehisa H, Kuno S, Katsuta S, Fukunaga T (2006) A 2-year follow-up study on muscle size and
dynamic strength in teenage tennis players. Scand J Med Sci Sports 16:93–101
Kano Y, Takahashi H, Morioka Y, Akima H, Miyashita K, Kuno S, Katsuta S (1997) Relationship
between the morphological features of adductor muscles and sprinting performance in
sprinters. Jpn J Phys Educ 41:352–359
Kawakami Y, Ichinose Y, Fukunaga T (1998) Architectural and functional features of human
triceps surae muscles during contraction. J Appl Physiol 85:398–404
Kearns CF, Abe T, Brechue WF (2000) Muscle enlargement in sumo wrestlers includes increased
muscle fascicle length. Eur J Appl Physiol 83:289–296
Kubo T, Muramatsu M, Hoshikawa Y, Kanehisa H (2010) Profiles of trunk and thigh muscularity
in youth and professional soccer players. J Strength Cond Res 24:1472–1479
Maughan RJ, Watson JS, Weir J (1983) Relationships between muscle strength and muscle cross-
sectional area in male sprinters and endurance runners. Eur J Appl Physiol Occup Physiol
50:309–318
Morse CI, Degens H, Jones DA (2007) The validity of estimating quadriceps volume from single
MRI cross-sections in young men. Eur J Appl Physiol 100:267–274
Narici MV, Roi GS, Landoni L, Minetti AE, Cerretelli P (1989) Changes in force, cross-sectional
area and neural activation during strength training and detraining of the human quadriceps. Eur
J Appl Physiol Occup Physiol 59:310–319
Narici MV, Hoppeler H, Kayser B, Landoni L, Claassen H, Gavardi C, Conti M, Cerretelli P
(1996) Human quadriceps cross-sectional area, torque and neural activation during 6 months
strength training. Acta Physiol Scand 157:175–186
Rønnestad BR, Hansen EA, Raastad T (2010) Effect of heavy strength training on thigh muscle
cross-sectional area, performance determinants, and performance in well-trained cyclists. Eur J
Schantz P, Randall-Fox E, Hutchison W, Tydén A, Astrand PO (1983) Muscle fibre type distri-
bution, muscle cross-sectional area and maximal voluntary strength in humans. Acta Physiol
Scand 117:219–226
Sugisaki N, Kanehisa H, Tauchi K, Okazaki S, Iso S, Okada J (2011) The relationship between
30-m sprint running time and muscle cross-sectional areas of the psoas major and lower limb
muscles in male college short and middle distance runners. Int J Sport Health Sci 9:1–7
Tsunoda N, Kanehisa H, Fukunaga T, Konda M, Ikegawa S (1986) Characteristics of athletes in
muscle area of quadriceps femoris. Jpn J Phys Fit Sport 35:192–199
Wakahara T, Miyamoto N, Sugisaki N, Murata K, Kanehisa H, Kawakami Y, Fukunaga T, Yanai
T (2012) Association between regional differences in muscle activation in one session of
resistance exercise and in muscle hypertrophy after resistance training. Eur J Appl Physiol
112:1569–1576
Chapter 16
Jump Performance Enhancement Induced
by Countermovement
Kuniaki Hirayama
Abstract The execution of a countermovement prior to the main movement,

during which the agonist muscles experience a stretch–shortening cycle (shortening
after being lengthened) enhances the exercise performance outcome
(countermovement effect). During the stretch–shortening cycle, the fascicles and
tendons exhibit a unique behavior; much of the length changes of skeletal muscle
occur in tendons while fascicles contract almost isometrically. The
countermovement effect varies among individuals and changes through training
(or practice). The dynamics of muscle fascicles and tendons may underlie the
effects of interindividual differences and training on the countermovement effect.
Keywords Stretch–shortening cycle • Muscle • Tendon
16.1 Introduction
A preparatory movement performed in the opposite direction prior to the main

movement is called a countermovement. Many human movements (e.g., walking,
running, jumping, throwing, and batting) are accompanied by countermovements. It
is well known that a higher level of exercise performance (i.e., higher mechanical
output and exercise economy) can be accomplished if the movement is preceded by
a countermovement. In this chapter I investigate the increase in mechanical output
that occurs during jumping as a result of the countermovement effect and discuss
this effect in detail.
The extent of the countermovement effect shows inter-individual differences
and varies according to training. That is, the extent of the countermovement effect
differs between individual athletes and can be improved by practice or training.
Therefore, the extent of the countermovement effect can be considered as a
component closely related to exercise performance. The main purpose of this
chapter is to provide basic information to coaches and/or researchers by organizing
the information about the mechanisms underlying the countermovement effect,
K. Hirayama (*)
e-mail: hirayama@aoni.waseda.jp

188 K. Hirayama
factors affecting interindividual differences in the countermovement effect and the

mechanisms involved in training (or practicing) the countermovement effect.
16.2 Dynamics of Skeletal Muscle During Movement

with Countermovement
Human movements are combinations of movements of individual joints, which are

produced by the action of skeletal muscles. A skeletal muscle is stretched during
countermovement and shortens during the main movement. This skeletal muscle
action is known as the stretch–shortening cycle (SSC; Komi 2003). The mechanical
output of skeletal muscle is increased by the SSC. In instances where a skeletal
muscle shortens without stretching, the mechanical output is lower. This SSC
induced increase in the mechanical output of a skeletal muscle is the main mech-
anism underlying performance enhancement in single-joint and/or multi-joint
movements.
Skeletal muscles can be separated according to function into contractile and
elastic components; the latter consist of series and parallel components (Hill 1951)
(Fig. 16.1). Although these words do not refer to specific anatomical tissues, it is
commonly held that the contractile component and series elastic component are
mainly located in muscle fibers and tendons (Alexander and Bennet-Clark 1977),
respectively. Hence a skeletal muscle is often known as a muscle–tendon unit
(MTU), in order to emphasize the functional characteristics of muscle fibers and
tendons.
It has been confirmed that tendon length changes as a function of the force
applied to it (animals: Ker 1981; Woo et al. 1980—humans, in vivo: Fukashiro
et al. 1995; Ito et al. 1998) (Fig. 16.2). Accordingly, it is generally accepted that
tendons have elastic properties and play a significant role in storing and utilizing
elastic energy (Magnusson et al. 2008). When researchers first studied SSC, the
length changes in muscle fibers were not distinguished from those in MTUs
(Cavagna 1977). However, because little change was noted for muscle fiber length
in running cats (Hoffer et al. 1989), researchers started to distinguish the length
change between MTUs and muscle fibers. Research involving ultrasonography
revealed that stretch–shortening in the tendons accounts for that of the entire
MTUs, and that length change of the fascicles was smaller than that of the MTUs
(muscle–tendon interaction) during human movements such as walking (Fukunaga
et al. 2001; Lichtwark et al. 2007; Ishikawa et al. 2005), running (Ishikawa
et al. 2007; Lichtwark et al. 2007), and jumping (Kawakami et al. 2002; Kurokawa
et al. 2003; Sousa et al. 2007; Ishikawa and Komi 2004) (Fig. 16.3).
16 Jump Performance Enhancement Induced by Countermovement 189
Fig. 16.1 Mechanistic

model of skeletal muscle
function. CC, SEC, and
PEC represent contractile
component, series elastic
component, and parallel
elastic component,
respectively
Fig. 16.2 Force–length 2a 800

relationship of the human 90 deg
tendon (the tibialis anterior
muscle) in the 3 angles that 105 deg
were measured for the ankle 600
joint (Fukashiro et al. 1995) 120 deg
F (N)
400
200
0
0 5 10 15 20 25
l (mn)
16.3 Mechanisms of the Countermovement Effect
The following concepts have been proposed as mechanisms for the increase in
mechanical output produced by the SSC as compared with concentric contractions
where SSC is not present.
1. Stretch reflex
2. Potentiation of contractile component
3. Time available for active state development (preactivation)
4. Utilization of elastic energy
5. Muscle–tendon interaction
The stretch reflex is a muscle contraction in response to stretching; when muscle
fibers are rapidly lengthened, the muscle spindle detects the lengthening and
increases alpha motor neuron activity, resulting in muscle fiber contraction (Jones
and Watt 1971). Potentiation of the contractile component is force potentiation
derived from alteration in the properties of the contractile machinery during
eccentric contraction (Bosco et al. 1981; Herzog and Leonard 2000). During the
190 K. Hirayama
Fig. 16.3 Ankle joint angle, fascicle length of medial gastrocnemius muscle, reaction force at the
foot perpendicular to the force plate, and EMGs from the medial gastrocnemius muscle during
CMJ (A) and noCMJ (B). Both CMJ (A) and noCMJ (B) were performed with only the ankle joint.
In CMJ (A), fascicle length started to increase (passive lengthening) then remained constant for
most of the dorsiflexion phase, then began to decrease in the plantar flexion phase. In noCMJ (B),
the fascicle length kept decreasing throughout the movement (Kawakami et al. 2002)
braking phase (last half of the countermovement), muscles develop an active state if
time is available; this results in a higher active state of the muscle at the beginning
of the main movement (Bobbert et al. 1996; Bobbert and Casius 2005; Chapman
et al. 1985; Svantesson et al. 1994; Arakawa et al. 2010). Utilization of elastic
energy is implemented by a series of elastic components (Komi and Bosco 1978;
Komi 2003; Arakawa et al. 2010). Muscle–tendon interaction creates a favorable
situation for muscle fibers to exert force by decreasing muscle fiber length change
(which results in decreased contraction speed and adjusts the sarcomere length to
optimal) by utilizing the elastic behavior of the tendon (Kawakami and Fukunaga
2006).
Kawakami et al. (2002) reported that although gastrocnemius fascicles were not
lengthened during the braking phase of the countermovement jump (CMJ), which is
performed with the ankle joint alone, the mechanical output (e.g., mechanical
power and work) was increased compared with jumping without countermovement
(noCMJ) (Fig. 16.3). On the basis of this result, Kawakami and Fukunaga (2006)
pointed out that there are cases where the countermovement effect can be achieved
without the stretch reflex and potentiation of contractile components that occur
during forceful lengthening of the muscle fibers. Although Sugisaki et al. (2005)
observed a slight fascicle lengthening during the braking phase of a drop jump, they
also indicated that most of mechanical work of MTUs was assumed by tendons. In
addition, they reported that the mean amplitude of electromyographic activity
(mEMG) of the triceps surae muscle during a drop jump did not exceed that during
noCMJ. According to these findings, it would be reasonable to suggest that just
muscle–tendon interactions and the corresponding utilization of elastic energy, as
well as the time available for active state development, play significant roles as
mechanisms underlying the countermovement effect (at least for the gastrocnemius
muscle).
Through simulation studies of vertical jumps that involved the analysis of the
3 major joints (hip, knee, and ankle) of the lower extremity, Bobbert and Casius
(2005) reached the conclusion that the time available for active state development
of the contractile component is the only mechanism underlying the
countermovement effect. Conversely, Arakawa et al. (2010) conducted a simulation
study on a single MTU and pointed out that both active state development and
elastic energy contribute to the countermovement effect and that the degree of
contribution of these factors varies with length of the series elastic component
(a longer series elastic component makes the elastic energy contribution higher).
Bobbert and Casius (2005) utilized a model involving the lower extremity that
included muscles with relatively short tendons, such as the gluteus maximus
muscle. This likely explains why their results indicated that elastic energy did not
contribute to the countermovement effect. In addition, the degree of utilization of
elastic energy is affected by the intensity of the countermovement (Sugisaki
et al. 2004; Ishikawa and Komi 2004). With regard to CMJ from a standing
position, elastic energy may not play a significant role for the following reasons:
(1) The fall length of the center of gravity was relatively short, resulting in a lower
intensity of countermovement (than drop jumps). (2) The stretching load was
distributed over the MTUs of the lower extremity. When combined, the contribu-
tion of each mechanism of the countermovement effect varies with situations such
as tendon length of the intended MTUs and/or intensity of the countermovement.
16.4 Factors Related to Interindividual Differences

in the Countermovement Effect
Because mechanisms involved with the countermovement effect vary according to

the particular situation (i.e., characteristics of recruited MTUs and/or intensity of
countermovement), it is difficult to summarize the factors that relate to the
interindividual differences of the countermovement effect for every case. This
192 K. Hirayama
section thus focuses on the utilization of elastic energy and tendon elasticity which
is responsible for the utilization of elastic energy.
In an ultrasonography-based study, Kawakami et al. (2002) pointed out that
several interindividual differences were observed in fascicle behavior during CMJ
performed with a single joint. This finding implies that there must have been
interindividual differences in tendon behavior as well. Some studies have estimated
the degree of utilization of elastic energy by using an integrated EMG, which is
reported to highly correlate with oxygen consumption during jumping (Bosco
et al. 1987). Sugisaki et al. (2004) reported that there was a positive correlation
(r ¼ 0.84–0.86) between the extent of the countermovement effect and the contri-
bution of elastic energy. Ito and Saito (1989) indicated that gymnasts accustomed to
utilizing elastic energy can use a greater amount of elastic energy during a rebound
jump than can swimmers. Belli and Bosco (1992) reported that there is a strong
positive correlation (r2 ¼ 0.92) between rebound jump performance and amount of
elastic energy utilization measured through an analysis of expired gas. These
reports suggest that the amount of elastic energy utilization might have significant
impacts on interindividual differences in the countermovement effect.
Some research has been done to examine the relation between the
countermovement effect and the mechanical properties of tendons in order to
evaluate the likelihood that they are responsible for elastic energy utilization.
Kubo et al. (1999) reported that there is a negative correlation between the extent
of the countermovement effect achieved through multi-joint vertical jump and
tendon stiffness of the vastus lateralis muscle (subjects with stiffer tendons achieve
smaller countermovement effects). In contrast, Bojsen–Møller et al. (2005) showed
that there is no correlation between these parameters. Subjects with higher tendon
stiffness exhibited a higher rate of torque development, which resulted in a higher
jump (both noCMJ and CMJ) performance. Burgess et al. (2007) supported this
argument: Higher tendon stiffness enabled higher jumps. Kubo et al. (2007a)
reported a negative correlation between Achilles tendon stiffness and the
countermovement effect by employing a single-joint (ankle joint) jump exercise
that eliminated the influence of coordination on the jump. We (Hirayama
et al. 2010) employed single-joint (ankle joint) jump exercises (noCMJ and CMJ)
for the task and various athletes (sprinters, long distance runners and weightlifters)
as subjects, and examined the relationship between Achilles tendon stiffness and
jump performance as well as countermovement effect. Our results showed that
tendon stiffness neither correlated with jump performance nor the
countermovement effect. That is, our results did not support the findings of Kubo
et al. (1999/2007a), Bojsen–Møller et al. (2005), nor Burgess et al. (2007). Tendon
behavior during SSC exercise is influenced not only by tendon elasticity as mea-
sured under static conditions, but also neuromuscular activity and corresponding
muscle fibers’ behavior. Moreover, Huijing and Ettema (1988/89) report that the
force–length relationship of tendons varies with the manner of muscle contraction.
We conclude that the lack of accord in the results of research of tendon stiffness and
jump performance as well as the countermovement effect implies that
interindividual differences in neuromuscular activity may have a greater influence
on jump performance or the countermovement effect than tendon elasticity, at least

as measured under static conditions.
16.5 Improvements of the Countermovement Effect
To improve exercise performance by an effect on countermovement, plyometric

exercise (exercise with SSC of MTUs) is often employed in training programs
(Potach and Chu 2000; Radcliffe and Farentinos 1999). In a systematic review of
plyometric training, Markovic (2007) found that the training improved CMJ per-
formance to a greater degree (8.7 %) than for a noCMJ performance (4.7 %). This
implies that plyometric training increases the countermovement effect.
According to studies that compare the training effect of traditional resistance
exercises and plyometric exercises (or combined training utilizing both types of
exercises), the latter creates a greater CMJ performance improvement (Kubo
et al. 2007b; Toumi et al. 2004). Plyometric training increases joint stiffness during
the braking phase more than traditional resistance training, which is associated with
greater jump performance enhancement (Kubo et al. 2007a; Toumi et al. 2004).
Toumi et al. (2004) reported increased neuromuscular activity during the switching
phase from countermovement to the main movement as a reason for the increase in
joint stiffness and CMJ performance. Similar changes in neuromuscular activity
were also noted by Chimera et al. (2004), Häkkinen et al. (1990), and Kyr€oläinen
et al. (2005). Ishikawa et al. (2003) utilized ultrasonography and observed that even
if the subject jumped from the same height, muscle–tendon behavior during drop
jumps varied with the intended jump height (i.e., maximal or submaximal); the
greatest tendon shortening was observed when the subject jumped as high as
possible. The results of the above studies lead to the following hypothesis: Changes
in neuromuscular activity and the corresponding muscle–tendon behavior can be
engaged to achieve a greater CMJ performance enhancement after plyometric
training than after traditional resistance training.
We (Hirayama et al. 2012) examined neuromuscular activities and muscle–
tendon behavior during CMJ and noCMJ. These activities were performed with
the ankle joint, before and after a single practice session of plyometric (CMJ)
exercise. The results indicated that: (1) An earlier onset of neuromuscular activity
of the triceps surae (agonist) muscle occurred after the practice; (2) The muscle–
tendon interactions became more apparent after the practice (i.e., fascicle length-
ening and shortening decreased with increases in tendon length change, and short-
ening velocity of the fascicle during the main movement decreased) (Fig. 16.4); and
(3) CMJ performance (countermovement effect) improved after the practice (the
noCMJ performance was not improved). In most cases, the EMG onset of agonist
muscles followed that of the ground reaction force. This result suggests that the
ground reaction force started to increase with the passive lengthening of MTUs, and
muscle fibers subsequently began to actively contract. In the passive condition,
muscle fibers are more easily lengthened than tendon fibers (Herbert and Crosbie
194 K. Hirayama
Fig. 16.4 Time course of

length change in the fascicle
(closed circle) and tendon
(open circle) during CMJ 30
First
before ( first) and after
( fourth) a single practice 20 Fascicle
session of CMJ. The
fascicle length started to
10
increase then remained
almost constant in the
dorsiflexion (counter and 0 Tendon
braking) phase. Afterward,
in the first test trial (before -10
Length changes (mm)
the practice), the fascicle counter braking planter flexion

shortened during the phase phase phase
plantarflexion phase,
whereas the tendon 30
shortened a little. In Fourth
contrast, in the fourth test 20
trial (after the practice), the
fascicle maintained a
constant length, whereas the 10
tendon shortened
alternatively in the 0
plantarflexion phase
(Hirayama et al. 2012)
−10
0 0.1 0.2 0.3 0.4 0.5
Time (s)
1997). The decrease in time lag from the onset of the passive force to that of muscle
activation would decrease the magnitude of muscle fiber lengthening. This phe-
nomenon made the sarcomere length closer to its optimal length. In addition, earlier
muscle activation during the braking phase increased the time available for the
muscle to develop force. In the main movement, muscle fibers could have exerted a
higher force in conjunction with slower shortening by taking advantage of the
force–velocity relationship. These phenomena could be the mechanisms for CMJ
performance (countermovement effect) improvement after the plyometric exer-
cises. Our findings suggest that muscle–tendon behavior is neurally controlled
and that changes in neuromuscular activity and muscle–tendon behavior are
involved in jump performance (countermovement effect) improvement.
16.6 Final Remarks
A number of studies have examined the relationship between maximal strength or

power and CMJ performance. However, some studies failed to find a significant
relationship between these parameters (Young and Bilby 1993; Iossifidou
et al. 2005). Coordination among joints (Bobbert and van Ingen Schenau 1988)
might be considered as a reason for the discrepancy between interindividual
differences of physical resources and those of exercise performance. This chapter
provides evidence which suggests that interindividuality and trainability are evident
even in muscle and tendon coordination, and have an indispensable influence on the
exercise performance.
Long-term training also induces morphological and/or functional changes
in muscles and tendons (Jozsa and Kannus 1997; Knobloch, 2007; Foure
et al. 2010). Although Hirayama et al. (2012) did not give the subjects a specific
instruction (the instruction was just “jump as high as possible”), the subjects who
have various functional characteristics of MTUs adjusted their muscle–tendon
behavior to improve CMJ performance. Therefore, even if the morphological and
functional characteristics of MTUs were to change after long-term training,
muscle–tendon behavior could be adjusted to the SSC exercise under the condition
where training includes plyometric exercise.
There could well be cases where a poor physical condition precludes the use of
proper technique. For example, if one’s muscle strength is insufficient to maintain
constant length during the braking phase, tendon lengthening and subsequent
shortening could be decreased (MTUs work like a damper). This article has no
intention of denying the influence of the body’s physical capabilities on the
countermovement effect. Over long-term training, though, it can be expected that
changes in bodily capabilities and the acquisition of proper technique work together
and act in conjunction to improve the SSC exercise performance.
References
Alexander RM, Bennet-Clark HC (1977) Storage of elastic strain energy in muscle and other
tissues. Nature 265(5590):114–117
Arakawa H, Nagano A, Yoshioka S, Fukashiro S (2010) Interaction between elastic energy
utilization and active state development within the work enhancing mechanism during
countermovement. J Electromyogr Kinesiol 20(2):340–347
Belli A, Bosco C (1992) Influence of stretch-shortening cycle on mechanical behaviour of triceps
surae during hopping. Acta Physiol Scand 144(4):401–408
Bobbert MF, Casius LJ (2005) Is the effect of a countermovement on jump height due to active
state development? Med Sci Sports Exerc 37(3):440–446
Bobbert MF, van Ingen Schenau GJ (1988) Coordination in vertical jumping. J Biomech 21(3):
249–262
Bobbert MF, Gerritsen KG, Litjens MC, Van Soest AJ (1996) Why is countermovement jump
height greater than squat jump height? Med Sci Sports Exerc 28(11):1402–1412
Bojsen-Møller J, Magnusson SP, Rasmussen LR, Kjaer M, Aagaard P (2005) Muscle performance
during maximal isometric and dynamic contractions is influenced by the stiffness of the
tendinous structures. J Appl Physiol 99(3):986–994
Bosco C, Komi PV, Ito A (1981) Prestretch potentiation of human skeletal muscle during ballistic
movement. Acta Physiol Scand 111(2):135–140
196 K. Hirayama
Bosco C, Montanari G, Ribacchi R, Giovenali P, Latteri F, Iachelli G, Faina M, Colli R, Dal

Monte A, La Rosa M (1987) Relationship between the efficiency of muscular work during
jumping and the energetics of running. Eur J Appl Physiol Occup Physiol 56(2):138–143
Burgess KE, Connick MJ, Graham-Smith P, Pearson SJ (2007) Plyometric vs. isometric training
influences on tendon properties and muscle output. J Strength Cond Res 21(3):986–989
Cavagna GA (1977) Storage and utilization of elastic energy in skeletal muscle. Exerc Sport Sci
Rev 5:89–129
Chapman AE, Caldwell GE, Selbie WS (1985) Mechanical output following muscle stretch in
forearm supination against inertial loads. J Appl Physiol 59(1):78–86
Chimera NJ, Swanik KA, Swanik CB, Straub SJ (2004) Effects of plyometric training on muscle-
activation strategies and performance in female athletes. J Athl Train 39(1):24–31
Foure A, Nordez A, Cornu C (2010) Plyometric training effects on Achilles tendon stiffness and
dissipative properties. J Appl Physiol 109(3):849–854
Fukashiro S, Itoh M, Ichinose Y, Kawakami Y, Fukunaga T (1995) Ultrasonography gives directly
but noninvasively elastic characteristic of human tendon in vivo. Eur J Appl Physiol Occup
Physiol 71(6):555–557
Fukunaga T, Kubo K, Kawakami Y, Fukashiro S, Kanehisa H, Maganaris CN (2001) In vivo
behaviour of human muscle tendon during walking. Proc Biol Sci 268(1464):229–233
Häkkinen K, Pakarinen A, Kyrolainen H, Cheng S, Kim DH, Komi PV (1990) Neuromuscular
adaptations and serum hormones in females during prolonged power training. Int J Sports Med
11(2):91–98
Herbert RD, Crosbie J (1997) Rest length and compliance of nonimmobilised and immobilised
rabbit soleus muscle and tendon. Eur J Appl Physiol Occup Physiol 76:472–479
Herzog W, Leonard TR (2000) The history dependence of force production in mammalian skeletal
muscle following stretch-shortening and shortening-stretch cycles. J Biomech 33(5):531–542
Hill AV (1951) The mechanics of voluntary muscle. Lancet 24:947–951
Hirayama K, Sugisaki N, Kato E, Kanehisa H, Fukunaga T, Kawakami Y (2010) Influence of
tendon stiffness, muscle strength, and muscle activity on individual differences in mechanical
work enhancement by a counter movement. Jpn J Phys Educ Health Sport Sci 55(1):33–43
(in Japanese)
Hirayama K, Yanai T, Kanehisa H, Fukunaga T, Kawakami Y (2012) Neural modulation of
muscle-tendon control strategy after a single practice session. Med Sci Sports Exerc 44(8):
1512–1518
Hoffer JA, Caputi AA, Pose IE, Griffiths RI (1989) Roles of muscle activity and load on the
relationship between muscle spindle length and whole muscle length in the freely walking cat.
Prog Brain Res 80:75–85, discussion 57–60
Huijing PA, Ettema GJ (1988/89) Length-force characteristics of aponeurosis in passive muscle
and during isometric and slow dynamic contractions of rat gastrocnemius muscle. Acta
Morphol Neerl Scand 26(1):51–62
Iossifidou A, Baltzopoulos V, Giakas G (2005) Isokinetic knee extension and vertical jumping: are
they related? J Sports Sci 23(10):1121–1127
Ishikawa M, Komi PV (2004) Effects of different dropping intensities on fascicle and tendinous
tissue behavior during stretch-shortening cycle exercise. J Appl Physiol 96(3):848–852
Ishikawa M, Finni T, Komi PV (2003) Behaviour of vastus lateralis muscle-tendon during high
intensity SSC exercises in vivo. Acta Physiol Scand 178(3):205–213
Ishikawa M, Komi PV, Grey MJ, Lepola V, Bruggemann GP (2005) Muscle-tendon interaction
and elastic energy usage in human walking. J Appl Physiol 99(2):603–608
Ishikawa M, Pakaslahti J, Komi PV (2007) Medial gastrocnemius muscle behavior during human
running and walking. Gait Posture 25(3):380–384
Ito A, Saito M (1989) Rebound effect of triceps surae muscle: comparison between gymnasts and
swimmer. Research reports of sports medicine and science 31–36. Japan Sports Association,
Tokyo (in Japanese)
Ito M, Kawakami Y, Ichinose Y, Fukashiro S, Fukunaga T (1998) Nonisometric behavior of

fascicles during isometric contractions of a human muscle. J Appl Physiol 85(4):1230–1235
Jones GM, Watt DG (1971) Muscular control of landing from unexpected falls in man. J Physiol
219(3):729–737
J
ozsa L, Kannus P (1997) Functional and mechanical behavior of tendon: human tendons. Human
Kinetics, Champaign
Kawakami Y, Fukunaga T (2006) New insights into in vivo human skeletal muscle function. Exerc
Sport Sci Rev 34(1):16–21
Kawakami Y, Muraoka T, Ito S, Kanehisa H, Fukunaga T (2002) In vivo muscle fibre behaviour
during counter-movement exercise in humans reveals a significant role for tendon elasticity. J
Physiol 540(Pt 2):635–646
Ker RF (1981) Dynamic tensile properties of the plantaris tendon of sheep (Ovis aries). J Exp Biol
93:283–302
Knobloch K (2007) Eccentric rehabilitation exercise increases peritendinous type I collagen
synthesis in humans with Achilles tendinosis. Scand J Med Sci Sports 17(3):298–299
Komi P (2003) Stretch-shortening cycle. In: Komi P (ed) Strength and power in sport, 3rd edn.
Blackwell Scientific, Oxford, pp 184–202
Komi PV, Bosco C (1978) Utilization of stored elastic energy in leg extensor muscles by men and
women. Med Sci Sports 10(4):261–265
Kubo K, Kawakami Y, Fukunaga T (1999) Influence of elastic properties of tendon structures on
jump performance in humans. J Appl Physiol 87(6):2090–2096
Kubo K, Morimoto M, Komuro T, Tsunoda N, Kanehisa H, Fukunaga T (2007a) Influences of
tendon stiffness, joint stiffness, and electromyographic activity on jump performances using
single joint. Eur J Appl Physiol 99(3):235–243
Kubo K, Morimoto M, Komuro T, Yata H, Tsunoda N, Kanehisa H, Fukunaga T (2007b) Effects
of plyometric and weight training on muscle-tendon complex and jump performance. Med Sci
Sports Exerc 39(10):1801–1810
Kurokawa S, Fukunaga T, Nagano A, Fukashiro S (2003) Interaction between fascicles and
tendinous structures during counter movement jumping investigated in vivo. J Appl Physiol
95(6):2306–2314
Kyr€oläinen H, Avela J, McBride JM, Koskinen S, Andersen JL, Sipila S, Takala TE, Komi PV
(2005) Effects of power training on muscle structure and neuromuscular performance. Scand J
Med Sci Sports 15(1):58–64
Lichtwark GA, Bougoulias K, Wilson AM (2007) Muscle fascicle and series elastic element length
changes along the length of the human gastrocnemius during walking and running. J Biomech
40(1):157–164
Magnusson SP, Narici MV, Maganaris CN, Kjaer M (2008) Human tendon behaviour and
adaptation, in vivo. J Physiol 586(1):71–81
Markovic G (2007) Does plyometric training improve vertical jump height? A meta-analytical
review. Br J Sports Med 41(6):349–355, discussion 355
Potach D, Chu D (2000) Plyometric training. In: Baechle T, Earle R (eds) Essential of strength
training and conditioning. Human Kinetics, Champaign, pp 465–510
Radcliffe J, Farentinos R (1999) High-powered pliometrics. Human Kinetics, Tokyo
Sousa F, Ishikawa M, Vilas-Boas JP, Komi PV (2007) Intensity- and muscle-specific fascicle
behavior during human drop jumps. J Appl Physiol 102(1):382–389
Sugisaki N, Okada J, Kanehisa H, Fukunaga T (2004) Effect of elastic energy on the mechanical
work and power enhancement in counter movement exercise of ankle joint. Ergonomics
40(2):82–89 (in Japanese)
Sugisaki N, Kanehisa H, Kawakami Y, Fukunaga T (2005) Behavior of fascicle and tendinous
tissue of medial gastrocnemius muscle during rebound exercise of ankle joint. Int J Sport
Health Sci 3:100–109
Svantesson U, Grimby G, Thomee R (1994) Potentiation of concentric plantar flexion torque
following eccentric and isometric muscle actions. Acta Physiol Scand 152(3):287–293
198 K. Hirayama
Toumi H, Best TM, Martin A, Poumarat G (2004) Muscle plasticity after weight and combined
(weight + jump) training. Med Sci Sports Exerc 22(8):1580–1588
Woo SL, Ritter MA, Amiel D, Sanders TM, Gomez MA, Kuei SC, Garfin SR, Akeson WH (1980)
The biomechanical and biochemical properties of swine tendons–long term effects of exercise
on the digital extensors. Connect Tissue Res 7(3):177–183
Young WB, Bilby GE (1993) The effect of voluntary effort to influence speed of contraction on
strength, muscular power, and hypertrophy development. J Strength Cond Res 7(3):172
Chapter 17
Can a High-Intensity Contraction Be
Enhanced by a Conditioning Contraction?
Insight from the Relationship Between
Shortening Velocity of Muscle Fibers
and Postactivation Potentiation
Atsuki Fukutani and Yasuo Kawakami
Abstract The magnitude of twitch torque increases after a high-intensity contrac-

tion of the same muscle (conditioning contraction). This phenomenon is called
postactivation potentiation (PAP). Recently, it has been shown that the maximal
voluntary concentric torque or power attained during the maximal voluntary con-
centric contraction can be increased by a conditioning contraction, suggesting that
conditioning contractions are effective on not only twitch but also on maximal
voluntary contractions. In contrast, some studies have reported that a conditioning
contraction had no potentiation effect on subsequent electrically-evoked maximal
isometric force. This discrepancy among previous studies may be attributable to
differences in the mode of contraction (i.e., isometric, concentric or eccentric),
which can affect the extent of PAP. Therefore, the main purpose of this study was to
examine the influence of these aforementioned factors on the extent of PAP, and to
discuss the applications of a conditioning contraction to high-intensity contractions.
Keywords Myosin regulatory light chain phosphorylation • Actin-myosin

interaction • Ca2+ concentration • Firing frequency
17.1 Introduction
17.1.1 Preface
The contractile history of a muscle affects the force it generates. For example,
muscle fatigue induced by muscle contraction affects subsequent contractions,
A. Fukutani (*)
Research Organization of Science and Technology, Ritsumeikan University, Shiga, Japan
e-mail: atsukifukutani@gmail.com
Y. Kawakami

200 A. Fukutani and Y. Kawakami
decreasing the force-generating capability (Gandevia et al. 1995; Enoka and

Duchateau 2008). In contrast, postactivation potentiation (PAP) (Vandervoort
et al. 1983; Sale 2002), induced by a preceding muscle contraction, has the opposite
effect (Rassier and MacIntosh 2000; Behm et al. 2004). In concrete terms, twitch
torque elicited by electrical stimulation increases after a high-intensity contraction.
This phenomenon is called PAP, and a contraction that induces PAP is called a
conditioning contraction (Babault et al. 2008; Tillin and Bishop 2009).
Recent studies have shown that a conditioning contraction increases the maxi-
mal voluntary concentric torque or power attained during a maximal voluntary
concentric contraction (Baudry and Duchateau 2007; Miyamoto et al. 2011),
suggesting that conditioning contractions are effective for maximal voluntary
contractions as well as twitch contractions. However, some studies have reported
that a conditioning contraction had no potentiation effect on subsequent
electrically-evoked maximal isometric force (Vandenboom et al. 1993). This dis-
crepancy among studies may be attributable to differences in the contraction mode
used to examine the effect of a conditioning contraction. The contraction type
utilized for examining the effect of conditioning contraction was different among
previous studies (Vandenboom et al. 1993: isometric contraction, Baudry and
Duchateau 2007: concentric contraction), which may affect the extent of the
conditioning contraction-induced potentiation effect (Babault et al. 2008). There-
fore, the main purpose of this study was to examine the influence of contraction
type (i.e., shortening velocity of the muscle) on the extent of PAP and to discuss the
application of conditioning contraction to high-intensity contractions such as max-
imal voluntary concentric contractions.
17.1.2 Terminology
We have defined two terms below to aid in the understanding of this chapter.
1. Twitch tension response of a muscle induced by a conditioning contraction
Some studies investigating the phenomenon of PAP have used isolated muscle,
and as a result, have measured “force” as an index of the twitch tension response of
a muscle (Vandenboom et al. 1993; MacIntosh et al. 2008). Other studies have
utilized human joint performance and measured joint “torque” (force moment
arm) as a measure of muscle force (Hicks et al. 1991; Shima et al. 2006). In this
study, the conditioning contraction-induced increases in force or torque are treated
as comparable phenomena because the moment arm is not changed during the
conditioning contraction.
2. Contraction intensity
Contraction intensity is expressed differently depending on the measurement
used to control the contraction intensity. For example, when contraction intensity is
modulated using Ca2+, its expression is based on Ca2+ concentration (Stephenson
and Williams 1981; Galler and Rathmayer 1992). On the other hand, when
17 Can a High-Intensity Contraction Be Enhanced by a Conditioning Contraction?. . . 201
contraction intensity is modulated using electrical stimulation, its expression is

based on stimulation frequency (Thomas et al. 1991; Mrowczyński et al. 2006). In
this article, contraction intensity is defined as intensity relative to a maximal-
intensity contraction, which is elicited by sufficiently high Ca2+ concentration
and/or sufficiently high-frequency stimulation. For example, in a low-intensity
contraction, the Ca2+ concentration or stimulation frequency is not sufficient to
evoke maximal force/torque, whereas in a maximal-intensity contraction, no further
increase in force/torque is elicited, even if the Ca2+ concentration or stimulation
frequency increases. Generally, a maximal voluntary contraction is considered a
“maximal-intensity” contraction. However, the capacity of the human nervous
system to activate available motor neurons is limited in some cases (Clark and
Taylor 2011). In other words, the Ca2+ concentration or stimulation frequency is not
sufficient to evoke a “maximal-intensity” contraction in cases of a “maximal-
voluntary” contraction. Hence, “maximal-voluntary” contraction does not neces-
sarily indicate “maximal-intensity” contraction.
17.1.3 Mechanism of PAP
The underlying mechanism of PAP is related to myosin regulatory light chain

phosphorylation (MLCP). Previous studies simultaneously monitored time-
dependent changes in twitch force and the extent of MLCP, before and after
conditioning contraction (Moore and Stull 1984; Vandenboom et al. 1995), and
found that these parameters changed in a similar fashion. In addition, Zhi
et al. (2005) reported that substantial increases in twitch force after conditioning
contractions did not occur in myosin light chain kinase knockout mice. Therefore,
MLCP is considered a primary factor for increases in twitch force/torque induced
after a conditioning contraction. The details of the signaling cascade that increase
twitch force/torque following a conditioning contraction are as follows: Once
muscles are activated by a conditioning contraction, Ca2+ is released from the
sarcoplasmic reticulum, which activates calmodulin that, in turn, activates myosin
light chain kinase (Nairn and Perry 1979). Consequently, the myosin regulatory
light chain is phosphorylated by the activated myosin light chain kinase (Grange
et al. 1993). MLCP increases the sensitivity to Ca2+, which leads to an increase in
muscle force at a given Ca2+ concentration because of an increased number of
attached cross bridges at a certain moment due to an increase in the duration of a
strongly-bounding state (Manning and Stull 1982; Sweeney et al. 1993). Based on
this mechanism, the extent of the increase in conditioning contraction-induced
twitch force/torque should decay as the intensity of a contraction conducted before
and after conditioning contraction (defined as “test contraction” hereafter)
increases. This is because a larger number of cross bridges are attached to the
actin filament even without MLCP in a high-intensity test contraction than in a
low-intensity test contraction (Sweeney et al. 1993). Indeed, the extent of increase
in MLCP-induced (i.e., conditioning contraction-induced) twitch force was much
smaller when the Ca2+ concentration was higher (Persechini et al. 1985). This
concept had been described by MacIntosh (2010).
17.1.4 Influence of a Conditioning Contraction on the Force/

Torque Attained During a High-Intensity Contraction
As described above, the effect of MLCP is dependent upon the intensity of a test
contraction (MacIntosh and Willis 2000). Thus, it has been generally accepted that
the force/torque attained during a high-intensity test contraction does not increase
by the conditioning contraction. In addition, a conditioning contraction has no
effect on the maximal shortening velocity obtained during a test contraction (Stuart
et al. 1988). Based on these results, a conditioning contraction is considered to have
no effect on the two extremes of the force-velocity relationship (i.e., maximal
shortening velocity and maximal isometric force) (Sale 2002). However, because
a conditioning contraction increased the rate of force development obtained during
an isometric test contraction at any contraction intensity although the maximal
isometric force did not increase (Vandenboom et al. 1993), Sale (2002) suggested in
his review that conditioning contraction shifts the force-velocity relationship
upward and rightward. Indeed, Baudry and Duchateau (2007) examined whether
a conditioning contraction shifts the force-velocity relationship, and found that a
conditioning contraction shifted the force-velocity relationship upward and right-
ward. However, the mechanism for explaining this shift has not been elucidated.
17.1.5 Purpose
The main purpose of this chapter is to examine the influence of the shortening
velocity of the muscle on the extent of PAP, which may be a factor to explain the
discrepancy among previous studies. In addition, we also discuss whether the force/
torque attained during a high-intensity contraction can be increased by a condition-
ing contraction.
17.2 Influence of the Shortening Velocity of Muscle Fibers

During a Twitch Contraction
17.2.1 Introduction
Babault et al. (2008) found that the conditioning contraction-induced increase in

twitch torque was larger under passive shortening than under isometric or passive
lengthening conditions. Based on this finding, they suggested that a conditioning

contraction is more effective for concentric than eccentric contractions. However,
they did not directly measure the behavior of muscle fibers. Previous studies have
shown that joint angle changes do not necessarily match length changes of muscle
fibers because of interactions with tendons (Fukashiro et al. 1995; Ichinose
et al. 2000). Thus, to clarify the influence of the shortening velocity of muscle
fibers on PAP, direct measurement of the behavior of muscle fibers during twitch
contractions is necessary. Therefore, the purpose of this chapter is to examine the
influence of the shortening velocity of muscle fibers during twitch contractions on
PAP, by directly measuring fascicle behavior.
17.2.2 Methods
Subjects
Fifteen healthy subjects (age: 24.7 2.1 years, height: 1.73 0.04 m, body mass:
67.1 5.7 kg, mean standard deviation [SD]) were recruited for this study. Each
subject provided written informed consent prior to participation, and the Ethics
Committee on Human Research of Waseda University approved this study.
Experimental Design
Plantar flexors were adopted as target muscles for this investigation. To examine
the influence of the shortening velocity of muscle fibers on PAP, different shorten-
ing velocities of muscle fibers were implemented under isometric (Iso), passive
shortening (Pas-S), and passive lengthening (Pas-L) conditions (Fig. 17.1). During
Iso, a twitch contraction was elicited with the ankle joint angle fixed at 0 (anatom-
ical position). During Pas-S, a twitch contraction was elicited when the ankle joint
angle was changed from dorsiflexion to plantar flexion. During Pas-L, a twitch
contraction was elicited when the ankle joint angle was changed from plantar
flexion to dorsiflexion. In all trials, timing of the electrical stimulation was con-
trolled to obtain the peak torque during twitch contractions at 0 .
After a twitch contraction was elicited, maximal-voluntary isometric plantar
flexion was performed for six seconds as a conditioning contraction. The contrac-
tion intensity and the duration of the conditioning contraction were identical for the
three conditions. Ten seconds after completing the conditioning contraction, a
twitch contraction was elicited again to calculate the relative change in twitch
torque (i.e., the extent of PAP). Each condition was randomly performed, and the
conditions were provided at >10-min intervals to prevent any influence by the
conditioning contraction on subsequent trials.
Fig. 17.1 Typical examples of fascicle length changes during twitch contractions
Experimental Setup
The subjects were positioned on an isokinetic dynamometer (CON-TREX, CMV

AG, Switzerland) with the right knee and hip joints fully extended, and the right
foot fixed to the dynamometer’s footplate. The right thigh was fixed to the dyna-
mometer’s bench with a non-elastic strap. The center of rotation of the footplate
was visually aligned with the center of the subject’s ankle joint. The anatomical
position of the ankle joint was defined as 0 . The ankle joint range of motion during
passive movement was set from 20 (dorsiflexion) to 30 (plantar flexion). The
angular velocity of the passive movement was set at 120 /s during Pas-S and at
120 /s during Pas-L.
The posterior tibial nerve was stimulated percutaneously to evoke a twitch
contraction of the plantar flexors with a single stimulation pulse. A cathode
(11 mm diameter) was placed over the popliteal fossa, and an anode (40 50 mm)
was placed over the ventral aspect of the thigh near the patella. Single rectangular
pulses of 500 μs were delivered from a high-voltage stimulator (SEN-3301; Nihon
Kohden, Tokyo, Japan) with an isolator (SS-1963; Nihon Kohden). The stimulus
intensity for each subject was determined prior to the experiment by increasing the
voltage until the corresponding isometric twitch torque at 0 reached a plateau, and
an intensity of 20 % above the maximal voltage was adopted for all trials.
To control the timing of electrical stimulation during Pas-S and Pas-L, the ankle
joint angle was recorded with programming software at a sampling frequency of
4 kHz (Labview, National Instruments, Austin, TX, USA). The timing of electrical
stimulations was adjusted for each subject to match the occurrence of the peak
torque during twitch contractions. The peak torque during a twitch contraction was
confirmed to occur at 0 in all trials.
The muscle fiber’s length (i.e., fascicle length) of the medial gastrocnemius
(MG) was recorded during the elicited twitch contraction using an ultrasound
device (SSD-6500, Aloka, Tokyo, Japan) with a linear-probe array (7.5 MHz,
Aloka, Tokyo, Japan). The ultrasonographic image was recorded at a sampling
frequency of 96 Hz. The fascicle length was measured as the distance from the
intersection of the superficial aponeurosis and fascicle to the intersection of the
deep aponeurosis and fascicle using Image J software (National Institutes of Health,
MD, USA).
Measurements
The peak torque during twitch contractions was determined as the twitch torque.
The relative change of these values were calculated with the following equation:
Relative change (%) ¼ (value recorded after the conditioning contraction / value
recorded before the conditioning contraction) 100.
The fascicle shortening velocity during the twitch contraction elicited before the
conditioning contraction was calculated by dividing the fascicle shortening distance
from the onset of torque development to the occurrence of peak torque by the time
required (Fig. 17.1). In addition, the mean fascicle length over the same interval
was also calculated.
Statistics
Descriptive data are presented as mean SD. A one-way analysis of variance

(ANOVA) with repeated measures was used to test differences in the relative
change in twitch torque (i.e., the extent of PAP), fascicle shortening velocity, and
mean fascicle length under the three conditions (Iso, Pas-S and Pas-L). If there was
a significant main effect, a post-hoc test with a Bonferroni correction was used. The
significance was set at p < 0.05 level.
Fig. 17.2 The extent of postactivation potentiation (PAP) during the three conditions. Data are
shown as mean SD. Pas-L passive lengthening condition, Iso isometric condition, Pas-S passive
shortening condition. * Significant difference between conditions ( p < 0.05)
Fig. 17.3 Fascicle shortening velocity during the three conditions. Data are shown as mean SD.
Pas-L passive lengthening condition, Iso isometric condition, Pas-S passive shortening condition.
* Significant difference between conditions ( p < 0.05)
17.2.3 Results
A significant main effect was found for the extent of PAP between the three
experimental conditions (F value ¼ 50.501, p < 0.001, Fig. 17.2). Additional anal-
ysis showed that the extent of PAP during Iso was significantly smaller than that
during Pas-S, and significantly larger than that during Pas-L, respectively
( p < 0.05).
There was a significant main effect in the fascicle shortening velocity of the MG
during twitch contractions (F value ¼ 50.056, p < 0.001, Fig. 17.3). The fascicle
shortening velocity during Iso was significantly smaller and larger than during
Pas-S and Pas-L, respectively ( p < 0.05).
Typical fascicle behavior during twitch contractions is shown in Fig. 17.4. The
MG fascicle length during twitch contractions elicited before the conditioning
contraction was 52.6 7.0 mm in Pas-L, 58.8 6.6 mm in Iso, and 65.5 8.3 mm
in Pas-S. A significant main effect was found between conditions
(F value ¼ 39.128, p < 0.001), and post-hoc analysis showed that the mean fascicle
length was significantly shorter during Pas-L than Pas-S ( p < 0.05, Table 17.1).
Fig. 17.4 Exemplary fascicle behavior during twitch contractions. Black squares indicate fascicle
behavior during twitch contraction elicited before the conditioning contraction. White circles
indicate fascicle behavior during twitch contraction elicited after the conditioning contraction.
Pas-L passive lengthening condition, Iso isometric condition, Pas-S passive shortening condition
Table 17.1 Mean fascicle length of medial gastrocnemius during twitch contractions
Fascicle length (mm)
Pas-L* Iso Pas-S
52.6 7.0 58.8 6.6 65.5 8.3
Data are shown as mean SD
Pas-L passive lengthening condition, Iso isometric condition, Pas-S passive shortening condition
*Significant difference compared with Pas-S ( p < 0.05)
17.2.4 Discussion
The main purpose of this study was to examine the influence of the shortening
velocity of muscle fibers on PAP. Although the conditioning contraction was
similar, the extent of PAP was different among the three conditions. As the fascicle
shortening velocity during twitch contractions increased, the extent of PAP also
increased. These results indicate that the extent of PAP is related to shortening
velocity of muscle fibers during twitch contractions.
The extent of PAP was larger during Pas-S than Iso or Pas-L. These results are in
line with the report of Babault et al. (2008) where the extent of PAP was largest
during a passive shortening condition. These authors also suggested that eccentric
contractions, unlike concentric contractions, are not potentiated by a conditioning
contraction, based on the relationship between joint angular velocity and the extent
of PAP. However, the type of muscle contraction (i.e., isometric, concentric, or
eccentric) as estimated from joint angle changes, is not identical to the type of
muscle fibers contraction due to muscle-tendon interactions (Fukashiro et al. 1995;
Ichinose et al. 2000). In fact, even during Pas-L, the fascicle length decreased in the
present study (Fig. 17.1). With this result, we can assume that the extent of PAP is
larger when muscle fibers shorten at a faster velocity, as seen during Pas-S in the
current results. Thus the conclusion by Babault et al. (2008) that suggested the
extent of PAP was contraction-type dependent (smaller during eccentric than
concentric contractions), might be due to differences in shortening velocity of the
muscle fibers.
To understand why the extent of PAP was the largest during Pas-S, the mech-
anism of PAP needs to be considered. As described earlier, an increase in twitch
force/torque by a conditioning contraction is related to the MLCP (MacIntosh
et al. 2012). Once a myosin regulatory light chain is phosphorylated, actin-myosin
interactions are facilitated, causing an increase in twitch torque (Sweeney
et al. 1993). Thus, the effect of MLCP should be small when the actin-myosin
interaction is already sufficiently high without a conditioning contraction. Indeed,
electrically-evoked maximal isometric force has been shown not to increase after a
conditioning contraction (Vandenboom et al. 1993). However, if the actin-myosin
interaction differs among various shortening velocity conditions, the extent of PAP
should also be different. Piazzesi et al. (2007) reported that as the shortening
velocity of muscle fibers increased, the number of attached cross bridges decreased.
Therefore, facilitation of actin-myosin interactions induced by MLCP would have a
stronger effect when muscle fibers shorten faster, thereby causing a larger increase
in the twitch torque during the Pas-S condition. Taking this into account, the
maximal-voluntary concentric torque may be increased by a conditioning contrac-
tion if the shortening velocity of muscle fibers is large. Even during a maximal-
intensity (maximal stimulation frequency) contraction, muscle force decreases as
the shortening velocity of muscle fibers increases (Hill 1938), which is well known
as the force-velocity relationship. As described above, the mechanism underlying
this decrease in force is considered to be a decrease in the number of attached cross
bridges (Piazzesi et al. 2007). Therefore, a conditioning contraction (facilitating the
formation of attached cross bridges) can substantially enhance a subsequent
maximal-intensity contraction because many cross bridges would not attach to
actin filaments.
In addition, differences in the stimulation frequency required to induce a
maximal-intensity contraction among different joint angular velocity conditions
could also be a factor affecting the increase in maximal-voluntary concentric
torque. Using isolated muscle fibers, de Haan (1998) reported that the stimulation
frequency required for inducing a maximal-intensity contraction at a given con-
traction velocity was highest for fast shortening, followed by slow shortening and
then isometric contractions. On the other hand, the typical firing frequency during a
maximal voluntary concentric contraction (Harwood et al. 2011) is lower than the
minimum stimulation frequency required to electrically evoke a maximal-intensity
contraction at a high shortening velocity. Therefore, it is possible that a “maximal-
voluntary” concentric contraction, especially during fast shortening velocity con-
ditions, would be insufficient to induce a “maximal-intensity” concentric contrac-
tion. If the contraction intensity decreases, the effect of MLCP becomes prominent
(MacIntosh 2010). Therefore, shortening velocity-related decreases in relative

muscle activation levels would be one of factors for increasing the maximal
voluntary concentric torque (Miyamoto et al. 2011; Fukutani et al. 2012).
The extent of PAP is affected by muscle fiber length during a twitch contraction
(Vandervoort et al. 1983). Thus, it is possible that the PAP differences seen in the
current study could be caused by differences in muscle fiber length rather than the
shortening velocity difference. In the current study, although the peak torque
attained during all twitch contractions occurred when the ankle joint angle was
0 , the changes in fascicle length and joint angle were not necessarily identical
(Fukashiro et al. 1995; Ichinose et al. 2000). Indeed, the fascicle length during
Pas-L was shorter than during Pas-S. However, since the extent of PAP is known to
be larger when the muscle fiber length is shorter (Vandervoort et al. 1983), the
present PAP findings cannot be explained by fascicle length differences because the
extent of PAP was smaller when the fascicle length was shorter.
17.3 Conclusion and Future Perspectives
Shortening velocity of muscle fibers affects the increase in twitch torque, with
larger increases occurring during conditions of faster shortening velocity. This
concept can explain why the maximal-voluntary concentric torque is increased by
a conditioning contraction. However, to elucidate whether conditioning contraction
enhances “maximal-intensity” concentric contractions, further experiments that
adopt maximal-intensity concentric contractions evoked by maximal firing fre-
quency are needed. This will eliminate the possibility of an effect of shortening
velocity-related decreases in relative muscle activation level on increases in the
maximal-voluntary concentric torque.
References
Babault N, Maffiuletti NA, Pousson M (2008) Postactivation potentiation in human knee extensors
during dynamic passive movements. Med Sci Sports Exerc 40:735–743
Baudry S, Duchateau J (2007) Postactivation potentiation in a human muscle: effect on the load-
velocity relation of tetanic and voluntary shortening contractions. J Appl Physiol
103:1318–1325
Behm DG, Button DC, Barbour G, Butt JC, Young WB (2004) Conflicting effects of fatigue and
potentiation on voluntary force. J Strength Cond Res 18:365–372
Clark BC, Taylor JL (2011) Age-related changes in motor cortical properties and voluntary
activation of skeletal muscle. Curr Aging Sci 4:192–199
de Haan A (1998) The influence of stimulation frequency on force-velocity characteristics of in
situ rat medial gastrocnemius muscle. Exp Physiol 83:77–84
Enoka RM, Duchateau J (2008) Muscle fatigue: what, why and how it influences muscle function.
J Physiol 586:11–23
Fukashiro S, Itoh M, Ichinose Y, Kawakami Y, Fukunaga T (1995) Ultrasonography gives directly

but noninvasively elastic characteristic of human tendon in vivo. Eur J Appl Physiol Occup
Physiol 71:555–557
Fukutani A, Miyamoto N, Kanehisa H, Yanai T, Kawakami Y (2012) Influence of the intensity of a
conditioning contraction on the subsequent twitch torque and maximal voluntary concentric
torque. J Electromyogr Kinesiol 22:560–565
Galler S, Rathmayer W (1992) Shortening velocity and force/pCa relationship in skinned crab
muscle fibres of different types. Pflugers Arch 420:187–193
Gandevia SC, Enoka RM, McComas AJ, Stuart DG, Thomas CK (1995) Neurobiology of muscle
fatigue. Advances and issues. Adv Exp Med Biol 384:515–525
Grange RW, Vandenboom R, Houston ME (1993) Physiological significance of myosin phos-
phorylation in skeletal muscle. Can J Appl Physiol 18:229–242
Harwood B, Davidson AW, Rice CL (2011) Motor unit discharge rates of the anconeus muscle
during high-velocity elbow extensions. Exp Brain Res 208:103–113
Hicks AL, Cupido CM, Martin J, Dent J (1991) Twitch potentiation during fatiguing exercise in
the elderly: the effects of training. Eur J Appl Physiol Occup Physiol 63:278–281
Hill AV (1938) The heat of shortening and the dynamic constants of muscle. Proc R Soc Lond B
Biol Sci 126:136–195
Ichinose Y, Kawakami Y, Ito M, Kanehisa H, Fukunaga T (2000) In vivo estimation of contraction
velocity of human vastus lateralis muscle during “isokinetic” action. J Appl Physiol
88:851–856
MacIntosh BR (2010) Cellular and whole muscle studies of activity dependent potentiation. Adv
Exp Med Biol 682:315–342
MacIntosh BR, Willis JC (2000) Force-frequency relationship and potentiation in mammalian
skeletal muscle. J Appl Physiol 88:2088–2096
MacIntosh BR, Taub EC, Dormer GN, Tomaras EK (2008) Potentiation of isometric and isotonic
contractions during high-frequency stimulation. Pflugers Arch 456:449–458
MacIntosh BR, Robillard ME, Tomaras EK (2012) Should postactivation potentiation be the goal
of your warm-up? Appl Physiol Nutr Metab 37:546–550
Manning DR, Stull JT (1982) Myosin light chain phosphorylation-dephosphorylation in mamma-
lian skeletal muscle. Am J Physiol 242:C234–C241
Miyamoto N, Kanehisa H, Fukunaga T, Kawakami Y (2011) Effect of postactivation potentiation
on the maximal voluntary isokinetic concentric torque in humans. J Strength Cond Res
25:186–192
Moore RL, Stull JT (1984) Myosin light chain phosphorylation in fast and slow skeletal muscles in
situ. Am J Physiol 247:C462–C471
Mrowczyński W, Celichowski J, Krutki P (2006) Interspecies differences in the force-frequency
relationship of the medial gastrocnemius motor units. J Physiol Pharmacol 57:491–501
Nairn AC, Perry SV (1979) Calmodulin and myosin light-chain kinase of rabbit fast skeletal
muscle. Biochem J 179:89–97
Persechini A, Stull JT, Cooke R (1985) The effect of myosin phosphorylation on the contractile
properties of skinned rabbit skeletal muscle fibers. J Biol Chem 260:7951–7954
Piazzesi G, Reconditi M, Linari M, Lucii L, Bianco P, Brunello E, Decostre V, Stewart A, Gore
DB, Irving TC, Irving M, Lombardi V (2007) Skeletal muscle performance determined by
modulation of number of myosin motors rather than motor force or stroke size. Cell
131:784–795
Rassier DE, Macintosh BR (2000) Coexistence of potentiation and fatigue in skeletal muscle. Braz
J Med Biol Res 33:499–508
Sale DG (2002) Postactivation potentiation: role in human performance. Exerc Sport Sci Rev
30:138–143
Shima N, Rice CL, Ota Y, Yabe K (2006) The effect of postactivation potentiation on the
mechanomyogram. Eur J Appl Physiol 96:17–23
Stephenson DG, Williams DA (1981) Calcium-activated force responses in fast- and slow-twitch
skinned muscle fibres of the rat at different temperatures. J Physiol 317:281–302
Stuart DS, Lingley MD, Grange RW, Houston ME (1988) Myosin light chain phosphorylation and
contractile performance of human skeletal muscle. Can J Physiol Pharmacol 66:49–54
Sweeney HL, Bowman BF, Stull JT (1993) Myosin light chain phosphorylation in vertebrate
striated muscle: regulation and function. Am J Physiol 264:C1085–C1095
Thomas CK, Bigland-Richie B, Johansson RS (1991) Force-frequency relationships of human
thenar motor units. J Neurophysiol 65:1509–1516
Tillin NA, Bishop D (2009) Factors modulating post-activation potentiation and its effect on
performance of subsequent explosive activities. Sports Med 39:147–166
Vandenboom R, Grange RW, Houston ME (1993) Threshold for force potentiation associated with
skeletal myosin phosphorylation. Am J Physiol 265:C1456–C1462
Vandenboom R, Grange RW, Houston ME (1995) Myosin phosphorylation enhances rate of force
development in fast-twitch skeletal muscle. Am J Physiol 268:C596–C603
Vandervoort AA, Quinlan J, McComas AJ (1983) Twitch potentiation after voluntary contraction.
Exp Neurol 81:141–152
Zhi G, Ryder JW, Huang J, Ding P, Chen Y, Zhao Y, Kamm KE, Stull JT (2005) Myosin light
chain kinase and myosin phosphorylation effect frequency-dependent potentiation of skeletal
muscle contraction. Proc Natl Acad Sci 102:17519–17524
Chapter 18
Is Graduated Pressure Profile an Essential
Feature for Compression Stockings
to Reduce Fatigue Development
of the Plantar Flexors?
Naokazu Miyamoto
Abstract Most sports compression stockings have employed a graduated pressure

profile. This pressure profile is based on positive evidence from the medical
literature on patients. However, no consensus has been reached as to whether or
not the graduated pressure profile is an essential feature for reducing muscle fatigue
development in the lower legs. The purpose of the present chapter is to examine the
pressure profile of compression stockings on the degree to which muscle fatigue of
the plantar flexors develops during repetitive calf-raise exercise. Thirteen male
subjects completed 12 sets of 10 repetitions of calf-raise exercise in one control
(barefoot) and three compression stocking conditions with the following pressure
profiles; (1) graduated pressure profile (GRA), (2) uniform pressure distribution
(UNI), and (3) localized pressure at the calf region just over the gastrocnemius
muscle belly (LOC). During the exercise, ankle joint angle and electromyographic
signals from the medial gastrocnemius and soleus were recorded. The plantar
flexion angle significantly decreased with repetitions of the calf-raise under all
conditions. The plantar flexion angles were significantly smaller in CON than in the
GRA, UNI and LOC conditions, whereas there was no significant differences
among GRA, UNI and LOC conditions. No significant difference was observed in
muscle activities of the medial gastrocnemius and soleus during the exercise among
the four conditions. These findings suggest that peripheral muscle fatigue of the
plantar flexors was relieved when wearing the compression stockings used in the
present study. In addition, the results indicate that a graduated pressure profile is not
an essential feature for compression stockings to reduce development of muscle
fatigue during repetitive plantar flexion exercise.
Keywords Gastrocnemius • Soleus • Calf-raise exercise • Electromyography
N. Miyamoto (*)
National Institute of Fitness and Sports in Kanoya, Kagoshima, Japan
e-mail: miyamoto@nifs-k.ac.jp

214 N. Miyamoto
18.1 Introduction
Lower body elastic compression garments such as stockings and tights have been
widely used in sports such as running and cycling among athletes of all levels. It has
been reported that wearing a compression garment during exercise improves
running economy (Bringard et al. 2006) and decreases peripheral muscle fatigue
(Miyamoto et al. 2011). Increased peripheral circulation (e.g., venous return) is
thought to be a possible mechanism for the improved running economy and
decreased muscle fatigue by wearing a compression garment (Ibegbuna
et al. 2003; Maton et al. 2006).
Theoretically, the mechanism of action of compression garments is straightfor-
ward; the pressure provided by compression garments increases the intramuscular
pressure, and consequently the cross-sectional areas of the veins can be decreased
with the blood flow being accelerated. Thus, the use of compression garments can
decrease venous stasis at rest and facilitate venous return to the heart with the aid of
venous valves, during exercise by increasing the efficacy of the muscle pump
(Alimi et al. 1994; Raju et al. 1993). To better assist the muscle pump, most sports
compression garments which cover both lower limbs are designed to provide higher
pressure distally with gradually declining pressure proximally (Bovenschen
et al. 2013; MacRae et al. 2011; Miyamoto et al. 2011; Scanlan et al. 2008).
However, when wearing a compression garment with such a graduated pressure
profile, it is possible that the pressure intensity is not adequate at the proximal
region. Indeed, Miyamoto et al. have shown that muscle fatigue during repetitive
calf-raise exercise occurred only at the gastrocnemius, not at the soleus, and that
muscle fatigue of the gastrocnemius declined when using a graduated compression
stocking with a high pressure intensity (30 mmHg at the ankle, 21–25 mmHg at the
calf, and 10 mmHg below the knee), but not when using one with a low pressure
intensity (18, 12–14, and 7 mmHg at the same regions, respectively) (Miyamoto
et al. 2011). Taking these observations into account together with findings that the
gastrocnemius is located at the proximal part of the lower leg and is composed of a
higher percentage of fatigable type II muscle fibers compared with the soleus
(Edgerton et al. 1975; Johnson et al. 1973), one can expect that a graduated pressure
profile is not necessarily an essential feature for compression stockings to reduce
the development of muscle fatigue during exercise. In other words, compression
stockings with an adequate pressure intensity at the calf region, such that being
uniform or localized rather than graduated pressure profiles on the gastrocnemius,
may effectively reduce the muscle’s fatigue during exercise.
Therefore, the purpose of the present study was to examine the effect of pressure
profiles of compression stockings on the magnitude of muscle fatigue of the plantar
flexors induced by repetitive calf-raise exercise.
18 Is Graduated Pressure Profile an Essential Feature for Compression. . . 215
18.2 Methods
18.2.1 Subjects
Fourteen young male subjects (25.0 3.2 years, 174.1 4.2 cm, 67.8 4.9 kg;
mean SD) with no history of orthopedic or neuromuscular disorders participated
in this study. However, one subject was unable to complete all sessions, and thus the
final number of subjects used in the analyses was 13. The subjects were fully
informed of the procedures used and the possible risks as well as the purpose of
the study. The hypothesis of the present study was not explained to the subjects
until termination of the all measurements, to exclude any potential bias that might
affect the results (Goh et al. 2011; Miyamoto et al. 2011). Written informed consent
was obtained from all subjects prior to the investigation. The study was approved by
the local ethics committee on human research.
18.2.2 Experimental Protocol and Settings
Subjects participated in four separate conditions (three types of compression

stockings (ankle-to-knee length) and one control (CON) conditions). In the com-
pression stocking conditions, subjects put on one of three specially-made compres-
sion stockings with the following profile; (1) graduated pressure profile (GRA),
(2) uniform pressure distribution (UNI), and (3) localized pressure at the calf region
just over the gastrocnemius muscle belly (LOC). The pressure intensity at the ankle
and gastrocnemius muscle belly regions of the three compression stockings are
shown in Table 18.1, according to the manufacturer’s information (Toko INC.,
Japan). In the CON condition, the subjects were barefoot. They then performed the
required calf-raise exercise. For all four conditions, a repetitive one-legged calf-
raise exercise was used to induce fatigue of the plantar flexors. Subjects were
instructed to place the ball of the right foot on a wooden block (Fig. 18.1) and to
raise their heels (plantar flexion) from a maximally dorsiflexed position and imme-
diately lower them back to a start position at the sound of a metronome at 1 Hz
(Miyamoto et al. 2011). The subjects touched a bar very lightly, but did not grip it
nor lean against it. During the exercise, the ankle joint angle was measured by an
Table 18.1 Pressure Pressure intensity (mmHg)

intensity exerted by three
types of compression GRA UNI LOC
stockings Muscle belly of gastrocnemius 18 30 30
Ankle 30 30 18
GRA graduated pressure profile, LOC localized pressure profile at
the gastrocnemius region, UNI uniform pressure distribution
profile
216 N. Miyamoto
Fig. 18.1 Experimental

apparatus and positioning of
the subject
electronic goniometer (SG110/A, Biometrics, UK). The anatomical position of the

ankle joint was defined as 0 , with increasingly large numbers representing increas-
ing degrees of plantar flexion. The hip and knee joint were kept extended through-
out the exercise. The ankle joint angle during the exercise was displayed on a
computer monitor in front of the subject. The exercise consisted of a series of
12 sets of 10 repetitions with 30 s rest intervals between sets. The subjects’ rating of
perceived exertion (RPE) using a Borg 6–20 scale was recorded before and imme-
diately after each exercise set. Four conditions were carried out in a randomized
order during 4 different days spaced at least 7 days apart to allow sufficient rest. The
exercises were done at the same time of day to minimize circadian effects on
maximal force production.
Surface electromyographic (EMG) signals were recorded from the medial gas-
trocnemius and soleus by bipolar surface electrodes (3 mm diameter, 20 mm inter-
electrode distance). The reference electrode was placed over the left patella for all
EMG measurements. The electrode placement was preceded by abrasion of the skin
to reduce the source impedance to less than 3 kΩ. The EMG signals were band-pass-
filtered (5–1,000 Hz) and differentially amplified (gain: 1,000, model: MEG-6116,
Nihon-Kohden, Japan). The EMG and joint angle data were simultaneously and
continuously stored on the hard disk of a personal computer for later analysis using a
16 bit analogue-to-digital converter (PowerLab/16SP, ADInstrument, Australia)
with a sampling frequency of 2 kHz.
18.2.3 Data Analysis and Statistics
The maximal plantar flexion angle of the ankle joint was calculated for each
repetition. The root mean square values of EMG signals (RMS-EMG) were calcu-
lated separately in the shortening (plantar flexion) and lengthening (dorsiflexion)
phases of the exercise, which were determined from the ankle joint angle. For all
parameters, separate two-way analyses of variance (ANOVAs)
(Set Rep Condition) with repeated measures were conducted. When a signifi-
cant interaction was observed, additional two- and one-way ANOVAs with Dunnett
and Tukey post-hoc comparisons were performed. The significance level for all
comparisons was set at P < 0.05. The statistical analyses were performed by a
commercial statistical software package (SPSS Statistics 20, IBM Japan, Japan).
All data are expressed as mean and SD.
18.3 Results
For RPE data, a three-way ANOVA revealed that there was only a significant main
effect of Set and Rep (P < 0.05) and no significant interactions.
Figure 18.2 shows maximal plantar flexion angle during the calf-raise
exercise. According to the three-way ANOVA, a significant interaction of
Set Rep Condition was observed (P < 0.05). Further analyses demonstrated
that the plantar flexion angle was significantly decreased with repetitions of the
calf-raise in all conditions. The plantar flexion angles were significantly smaller
during the 3rd–12th sets in CON as compared to the GRA, UNI and LOC condi-
tions, whereas there were no significant differences among GRA, UNI and LOC
conditions.
For RMS-EMG of the medial gastrocnemius and soleus during the shortening
and lengthening phases of the calf-raise exercise, the three-way ANOVAs revealed
only a main effect of Set and Rep (P < 0.05) with no significant interactions.
Fig. 18.2 Changes in ankle joint angle during calf-raise exercise in control (CON: open circle),
compression stocking with graduated pressure profile (GRA: light gray square), compression
stocking with uniform pressure distribution (UNI: dark gray triangle), and compression stocking
with localized pressure at the calf region just over the gastrocnemius muscle belly (LOC: black
diamond) conditions
218 N. Miyamoto
18.4 Discussion
The aim of the present study was to examine the effect of the pressure profile of
compression stockings on muscle fatigue of the plantar flexors during repetitive
calf-raise exercise. The main findings of the present study was that the magnitude of
reduction of plantar flexion angle during the repetitive calf-raise exercise was
greater in the CON than in the GRA, UNI and LOC conditions whereas no
significant difference in RMS-EMG was observed among the conditions. These
results suggest that peripheral muscle fatigue of the plantar flexors was lessened
when wearing any of the compression stockings used in the present study. Thus, a
graduated pressure profile is not an essential feature for compression stockings in
order to reduce development of muscle fatigue during repetitive plantar flexion
exercise.
The effectiveness of compression garments has gained much interest in recent
years, although it remains unclear whether the use of compression garments such as
stockings have a positive effect on muscle fatigue during exercise (e.g., see review
by MacRae et al.). It is likely that one of the reasons for the non-consensus is the
difference in pressure intensities provided by different compression garments.
Information about a garment’s pressure intensity is essential to compare findings
across studies. However, pressure intensities of compression garments have often
gone unreported in previous studies. Among only a limited number of studies which
reported the pressure intensities, it was noted that compression stockings with a
pressure intensity of 30 mmHg at the ankle, 24 mmHg at the calf, and 14 mmHg at
the knee produced a greater mean venous flow velocity at rest than did those with a
lower pressure (18, 14, and 8 mmHg at the same positions, respectively) (Lawrence
and Kakkar 1980). Moreover, Miyamoto et al. have investigated the influence of
pressure intensities of graduated compression stockings on muscle fatigue of the
triceps surae induced by repetitive calf-raise exercise. They demonstrated that the
magnitude of reduced force-generating capacity of the muscle after the exercise
was smaller only when using a graduated compression stocking with an adequate
pressure (30 mmHg and 21 mmHg at the ankle and the muscle belly of the
gastrocnemius, respectively) (Miyamoto et al. 2011), which is similar to the GRA
of the present study. This is consistent with our finding that, after a 30 min
submaximal running exercise, the extent of muscle fatigue evaluated by
T2-weighted magnetic resonance imaging was smaller when wearing graduated
compression stockings with high pressure intensities (27 mmHg and 21 mmHg at
the ankle and the muscle belly of the gastrocnemius, respectively), but not when
wearing those with low pressure intensities (18 mmHg and 14 mmHg) (unpublished
data). These observations are consistent with the present finding that during the
calf-raise exercise the extent of muscle fatigue was smaller in the GRA conditions
as compared with those in the CON condition. Taken together, it is suggested that
even when using a graduated compression stocking, an adequate pressure intensity
is required to reduce development of muscle fatigue of the plantar flexors during
exercise.
From a historical viewpoint, graduated compression stockings originate from

medical applications. Previous studies have demonstrated increased venous blood
flow velocity, reduced venous pooling, and improved venous return in non-active,
resting post-operative patients (Lawrence and Kakkar 1980; Liu et al. 2008). Based
on these findings, most sports compression stockings have employed a graduated
pressure profile. As hypothesized, however, the present study revealed that such a
graduated pressure profile is not an essential feature for compression stockings to
reduce muscle fatigue development of the plantar flexors during repetitive calf-raise
exercise. In the case of the lower leg (especially plantar flexors), the gastrocnemius
composed of a relatively higher percentage of fatigable type II fibers is located at
the proximal part as compared with the soleus. During the repetitive muscle
contractions of plantar flexion, the magnitude of muscle fatigue was greater in the
medial gastrocnemius than in the soleus (Miyamoto et al. 2011). This indicates that
the proper pressure profiles obtained from clinical studies with resting patients are
not directly applicable to dynamic exercise in healthy subjects.
One of the possible mechanisms for the reduced muscle fatigue during the calf-
raise exercise is an improved peripheral circulation (Ibegbuna et al. 2003; Maton
et al. 2006). It has been reported that the decline of mean power frequency of the
EMG signal associated with muscle fatigue is smaller only when wearing graduated
compression stockings with an adequate pressure intensity (Miyamoto et al. 2011).
A shift of the power density spectrum of EMG signals toward lower frequency
bands is thought to be attributed to changes in the wave form of the individual
motor unit action potentials as well as a reduction in muscle fiber conduction
velocity. This could be caused by a lowered intramuscular pH (Béliveau
et al. 1992; Bouissou et al. 1989). Reduction of intramuscular pH has been reported
to impair muscle contractile properties such as maximal isometric force and
maximal shortening velocity (Rassier and Herzog 2002; Westerblad et al. 1997).
Taken together, it is suggested that the accumulation of metabolic bi-products such
as intramuscular H+ is smaller when using compression stockings with an adequate
pressure intensity at the calf region. This notion is also supported by a study using
T2-weighted magnetic resonance imaging technique (Miyamoto and Kawakami
2015). On the other hand, as Kraemer et al. mentioned (Kraemer et al. 2010), the
mechanism(s) by which compression garments mediate any positive effects should
be understood only in the context of whether the garment characteristics fit the
particular need of each individual. Additionally, the optimal pressure profile and
intensity for compression stockings were not clarified in the present study. There-
fore, further studies are required to reveal the detailed mechanisms responsible for
reduced muscle fatigue by wearing the compression stocking used in the present
study, and to systematically identify the optimal pressure profile and intensity.
Since prior knowledge of the assumed benefit of compression garments may
predispose subjects to believe that using compression garments can improve their
performance, the hypothesis was not explained to the subjects until termination of
all measurements in the present study. However, it is difficult to design a study
where the subject is completely blind to the experimental conditions, because
pressure is provided by compression garments to the body. Moreover, construction
220 N. Miyamoto
and pressure intensity of garments can create the potential for optimal skin contact,
which is vital for the kinesthetic sense. This sense mediates many of the garments
performance effects (Kraemer et al. 1996; Kraemer et al. 1998; Kraemer
et al. 2010). Thus, the CON condition in the present study might not serve as an
optimal control from the viewpoint of perceptual and psychological aspects. How-
ever, for the type of exercise performed, the stockings’ movement was minimal and
stayed in contact with the skin after the calf-raise exercise in all three stocking
conditions. Furthermore, there was no difference in RPE among the four conditions,
whereas muscle fatigue was significantly smaller in GRA, UNI, and LOC than in
CON. Therefore, perceptual and psychological effects are unlikely to have had a
significant impact on the physiological variables measured.
In conclusion, the present study showed that the magnitude of muscle fatigue of
the plantar flexors during repetitive calf-raise exercise was reduced when wearing
compression stockings. However, possibly because all stockings produced an
adequate pressure intensity at the calf region, the reduction in fatigue was similar
regardless of pressure profile of the compression stockings used. This suggests that
a graduated pressure profile is not an essential feature for compression stockings to
reduce development of muscle fatigue of the plantar flexors during exercise.
References
Alimi YS, Barthelemy P, Juhan C (1994) Venous pump of the calf: a study of venous and muscular
pressures. J Vasc Surg 20:728–735, doi:S0741521494001527
Béliveau L, Van Hoecke J, Garapon-Bar C, Gaillard E, Herry JP, Atlan G, Bouissou P (1992)
Myoelectrical and metabolic changes in muscle fatigue. Int J Sports Med 13(Suppl 1):S153–
S155
Bouissou P, Estrade PY, Goubel F, Guezennec CY, Serrurier B (1989) Surface EMG power
spectrum and intramuscular pH in human vastus lateralis muscle during dynamic exercise. J
Bovenschen HJ, Booij MT, van der Vleuten CJ (2013) Graduated compression stockings for
runners: friend, foe, or fake? J Athl Train 48:226–232. doi:10.4085/1062-6050-48.1.26
Bringard A, Perrey S, Belluye N (2006) Aerobic energy cost and sensation responses during
submaximal running exercise–positive effects of wearing compression tights. Int J Sports Med
27:373–378
Edgerton VR, Smith JL, Simpson DR (1975) Muscle fibre type populations of human leg muscles.
Histochem J 7:259–266
Goh SS, Laursen PB, Dascombe B, Nosaka K (2011) Effect of lower body compression garments
on submaximal and maximal running performance in cold (10 C) and hot (32 C) environ-
ments. Eur J Appl Physiol 111:819–826. doi:10.1007/s00421-010-1705-2
Ibegbuna V, Delis KT, Nicolaides AN, Aina O (2003) Effect of elastic compression stockings on
venous hemodynamics during walking. J Vasc Surg 37:420–425. doi:10.1067/mva.2003.104
Johnson MA, Polgar J, Weightman D, Appleton D (1973) Data on the distribution of fibre types in
thirty-six human muscles. An autopsy study. J Neurol Sci 18:111–129, doi:0022-510X(73)
90023-3
Kraemer WJ, Bush JA, Bauer JA, Triplett-McBride NT, Paxton NJ, Clemson A, Koziris LP,
Mangino LC, Fry AC, Newton RU (1996) Influence of compression garments on vertical jump
performance in NCAA division I volleyball players. J Strength Cond Res 10:180–183
Kraemer WJ, Bush JA, Newton RU, Duncan ND, Volek JS, Denegar CR, Canavan P, Johnston J,
Putukian M, Sebastianelli WJ (1998) Influence of a compression garment on repetitive power
output production before and after different types of muscle fatigue. Sports Med Train Rehabil
8:163–184
Kraemer WJ, Flanagan SD, Comstock BA, Fragala MS, Earp JE, Dunn-Lewis C, Ho JY, Thomas
GA, Solomon-Hill G, Penwell ZR, Powell MD, Wolf MR, Volek JS, Denegar CR, Maresh CM
(2010) Effects of a whole body compression garment on markers of recovery after a heavy
resistance workout in men and women. J Strength Cond Res 24:804–814. doi:10.1519/JSC.
0b013e3181d33025
Lawrence D, Kakkar VV (1980) Graduated, static, external compression of the lower limb: a
physiological assessment. Br J Surg 67:119–121
Liu R, Lao TT, Kwok YL, Li Y, Ying MT (2008) Effects of graduated compression stockings with
different pressure profiles on lower-limb venous structures and haemodynamics. Adv Ther
25:465–478. doi:10.1007/s12325-008-0058-2
MacRae BA, Cotter JD, Laing RM (2011) Compression garments and exercise: garment consid-
erations, physiology and performance. Sports Med 41:815–843. doi:10.2165/11591420-
000000000-00000
Maton B, Thiney G, Ouchene A, Flaud P, Barthelemy P (2006) Intramuscular pressure and surface
EMG in voluntary ankle dorsal flexion: influence of elastic compressive stockings. J
Electromyogr Kinesiol 16:291–302, doi:S1050-6411(05)00095-7
Miyamoto N, Kawakami Y (2015) No graduated pressure profile in compression stockings still
reduces muscle fatigue. Int J Sports Med 36:220–225
Miyamoto N, Hirata K, Mitsukawa N, Yanai T, Kawakami Y (2011) Effect of pressure intensity of
graduated elastic compression stocking on muscle fatigue following calf-raise exercise. J
Electromyogr Kinesiol 21:249–254, doi:S1050-6411(10)00132-X
Raju S, Fredericks R, Lishman P, Neglen P, Morano J (1993) Observations on the calf venous
pump mechanism: determinants of postexercise pressure. J Vasc Surg 17:459–469, doi:0741-
5214(93)90145-C
Rassier DE, Herzog W (2002) Effects of pH on the length-dependent twitch potentiation in skeletal
muscle. J Appl Physiol 92:1293–1299. doi:10.1152/japplphysiol.00912.2001
Scanlan AT, Dascombe BJ, Reaburn PR, Osborne M (2008) The effects of wearing lower-body
compression garments during endurance cycling. Int J Sports Physiol Perform 3:424–438
Westerblad H, Bruton JD, Lannergren J (1997) The effect of intracellular pH on contractile
function of intact, single fibres of mouse muscle declines with increasing temperature. J
Physiol 500:193–204
Chapter 19
Exercise in Space: Physical and Mental
Benefit
Stefan Schneider, Tobias Vogt, and Vera Abeln
Abstract In the last decades exercise in space has mainly been used to counteract
musculoskeletal and cardiovascular deconditioning. This is in accordance with the
mainstream approach of exercise science to preserve and restore physical health.
Today we know that exercise holds an additional dimension, affecting not only the
peripheral physiological system but also enhancing neurocognitve performance and
affective state. As a result medical interest in exercise as a countermeasure to mood
changes and neurocognitive function has recently emerged. Special attention was
given to alterations in brain cortical activity caused by exercise. These modifica-
tions are supposed to act as a multifunctional generator for the adaptation of mood,
vigilance and cognitive performance.
This chapter is dedicated to psycho-physiological effects of exercise in space.
The aim is to show that exercise in space means more than staying physically fit.
Exercise in space can also help to improve mood, enhance neurocognitive function
and therefore increase crew performance.
By reviewing a series of recent research activities from our own lab, this chapter
likes to draw attention to the fact that exercise can be regarded as a holistic
approach to regulate a multitude of psycho-physiological processes occurring in
space during long-term confinement. Of course it is easy to translate these findings
into everyday life. Living in microgravity, living in space is a very feasible
Electronic supplementary material The online version of this chapter (doi:10.1007/978-4-431-

55315-1_19) contains supplementary material, which is available to authorized users.
S. Schneider (*)
Center for Health and Integrative Physiology in Space & Institute of Movement and
Neurosciences, German Sport University Cologne, K€ oln, Germany
University of the Sunshine Coast, Maroochydore DC, QLD, Australia
e-mail: schneider@dshs-koeln.de
T. Vogt • V. Abeln
Center for Health and Integrative Physiology in Space & Institute of Movement and
Neurosciences, German Sport University Cologne, K€ oln, Germany

224 S. Schneider et al.
analogue for a sedentary life-style. Accordingly ideas and content of this chapter
will not only help to improve mission success and mission safety while living in
space, but might also contribute to the discussion about an active life style and its
relevance for socio-economic and health-political decisions of the upcoming years.
Keywords Mental health • Exercise • Spaceflight • Societal relevance • Brain

activity
19.1 Introduction
Long duration space flights have provided a considerable amount of scientific

research on human ability to function in extreme environments. Findings indicate
that long duration missions take a toll on the individual, both physiologically and
psychologically, as entry into a gravitationless environment and living in it is a
novel situation for humans, which the body may perceive as stress.
Within the last four decades, starting with the first manned space missions
(Halberg et al. 1970; Rummell et al. 1975) exercise has been proven to be an effective
method to counteract physiological deconditioning during long-term space flight.
Although it has been shown that a regular exercise program is not able to completely
avert deconditioning of the musculoskeletal system (Gopalakrishnan et al. 2010),
the decremental effects of living under zero-gravity conditions might be decelerated
by a regular exercise routine (Figs. 19.1, 19.2 and 19.3) (Video 19.1).
With regard to the positive effects of exercise on numerous physiological
systems, so far exercise science in space has widely neglected the benefit of
exercise on psychological parameters like mood and cognitive performance. This
is of major interest as the duration of space missions is significantly increasing.
Whereas early missions had a duration of several days only, currently the average
inhabitation on the International Space Station (ISS) is 6 months. Moreover
national and international space agencies are seriously considering a trip to Mars
(Fig. 19.4).
While exercise in space, from a physiological point seems insignificant for
mission success, its impact on psychological parameters during prolonged space
flight might be regarded as essential for mission success and safety on board.
19.2 The Physical Benefit of Exercise
The adaptation or deadaptation of physiological systems such as the musculoskel-

etal or cardiovascular system under missing gravity conditions is an extremely
interesting field of research and has provided us within the last four decades not
only with interesting but also important information, which are nowadays widely
19 Exercise in Space: Physical and Mental Benefit 225
Fig. 19.1 ESA Astronaut Christer Fuglesang exercising on a bicycle ergometer on Shuttle
middeck. Copyright NASA
Fig. 19.2 ESA Astronaut Andre Kuipers exercising in Node 3 of the International Space Station
(ISS). Copyright ESA
Fig. 19.3 ESA Astronaut Luca Parmitano biking in space. Copyright ESA
Fig. 19.4 Art work of ESAs exploration missions. Copyright ESA
used inter alia in the rehabilitation of patients suffering from the negative effects of
immobilisation, e.g. after surgery. Everyone agrees that an adequate exercise
program in space prevents from physiological deconditioning of muscles, bones
as well as the cardiovascular system, which helps astronauts and cosmonauts after
long duration spaceflight to quickly readapt to earth gravity (Fig. 19.5).
Nevertheless, regarding performance optimisation and workload while living in
zero gravity, exercise seems wasted time: A fundamental principle about the
adaptation of bones and muscles is that if being used, they will increase their
strength, while if not in use strength and stability will deteriorate. On earth, even
if we don’t exercise, our musculoskeletal system is adapted to the load of daily
living (walking, standing upright, carrying a box of water, etc.). The fact that there
is a decrease of muscle mass as well as a decrease of bone stability especially in the
lower extremities in space, simply is due to the fact that the musculoskeletal system
is not used adequately, or, to say it the other way round, is not required. A recent
paper by Gopalakrishnan et al. (2010) very nicely demonstrated that muscle mass
and strength for the upper extremities shows no major changes across a time frame
of approx. 6 months and also the strength losses in the ankle dorsiflexor group was
neglectable, due to the regular use of foot loops, which are used to align and move
the body against the resistance of inertial forces. This shows that muscles, that are
used and that are useful, do not show any loss of strength in space whereas those
that seem useless in space (e.g. calf muscles) degrade. The same thing is true for the
Fig. 19.5 ESA Astronaut Luca Parmitano during work out while preparing for space flight.
Copyright ESA
cardiovascular system. There is an adaptation when permanently exposed to

weightlessness, namely a decrease of heart rate similar to a supine position on
earth (Verheyden et al. 2010)1 – which seems not to be a problem unless returning
to earth gravity where orthostatic problems arise.
To conclude, keeping up muscle strength and volume by regular exercise is not a
key factor for performance and mission success during space missions. Neverthe-
less two reasons exist to further explore and understand the detrimental effect of
weightlessness on the musculoskeletal and the cardiovascular system and the role
exercise plays in counteracting this deadaptation: (1) the benefit for individuals
returning from long duration space flights experiencing a quicker rehabilitation and
1
One might even hypothesize, although longitudinal studies have not been performed yet, that
cardiovascular function might benefit from missing gravity as heart rate is permanently decreased
(Verheyden et al. 2010)
Fig. 19.6 ESA Astronaut Luca Parmitano back on earth after Expedition 36/37 (Soyuz
TMA-09M). Copyright ESA
(2) the possibility to translate these findings into a benefit for an aging society
helping to understand the underling physiological processes of degeneration and
help to address adequate countermeasure, especially exercise (Figs. 19.6).
Nevertheless we need to be aware of the fact, that exercise has a second
dimension, which so far has been widely disregarded in space science and also in
health sciences.
19.3 The Mental Benefit of Exercise
Living in space is characterized by a number of psycho-physiological stressors

(Fowler and Manzey 2000; Manzey 2000). While during short term space missions
and in the beginning of long term space flights, physiological stressors dominate,
with on-going duration the hostile environment, the limitations in the space habitat,
the tremendous mission workload and last not least the social situation on board
have been identified as main stressors. From studies of the last 20 years we know
that this multistressor environment has a negative impact on cognitive performance,
mental health and mood (Fowler and Manzey 2000; Kanas 1998) and therefore
affects mission success and safety (Palinkas 2001). Although the underlying neu-
rophysiological processes remain widely unclear, which is due to missing imaging
technologies in space, recent research let us assume that microgravity itself
Fig. 19.7 Multi-dimensional stressors during spaceflight are know to have a negative impact on
the musculoskeletal and cardiovascular but also on the cognitive and affective systems of space
travellers. In contrast exercise is known to positively affect the physiological as well as psycho-
logical systems and therefore might be identified as a holistic countermeasure guaranteeing the
maintenance of health during long-term spaceflight missions
(Schneider et al. 2008a; Cheron et al. 2006) but also the confinement within an
isolated environment (Schneider et al. 2010b) contribute to changes in brain
cortical function which might manifest in a decrease of cognitive performance,
mental health and mood (Fig. 19.7).
Whereas the positive effects of regular as well as acute exercise on cognitive
performance, mental health and mood have been substantially investigated through-
out the last two decades in earth-bound studies, it was only in the past few years that
research has indicated that these positive effects of exercise are caused by changes
in brain cortical function. Especially (pre-)frontal brain regions, which are well
known to play a major role in information as well as emotional processing, seem to
be affected by exercise. Two major theories have been developed to explain these
effects, the transient hypofrontality hypothesis by Dietrich (2006) and the dual-
mode theory by Ekkekakis and Acevedo (2006). Both theories assume that exercise
is accompanied by a redistribution of CNS activity to regions that are involved into
planning and executing motor commands (mainly the motor cortex, supplementary
motor areas and sensory cortex) and therefore regions not involved into the
demanding task of exercise (as fronto-temporal areas of the brain) show less
activity. If we regard the brain as a multiprocessor unit, than exercise is able to
reduce clock frequency in specific and essential processors (because calculating
capacity is used elsewhere) and this seems to reset basic procedures (e.g. cognitive
and emotional processing).
Within the next paragraphs we would like to review several studies performed
during the last years demonstrating the positive effects of exercise on mood and
cognitive performance, which will guide us to the assumption that exercise in space
might be an important tool for individual performance optimisation and mission
success.
19.4 Even a Short Bout of Exercise Is Able to Influence

Mood and Cognitive Performance
The Mars500 program, initiated by the European Space Agency (ESA) and the
Institute of Biomedical Problems (IBMP) in Moscow aimed to prepare for a human
spaceflight to Mars. In the years 2010/11 six crewmembers lived in a confined
environment, mimicking a space ship, located in the IBMP facilities in Moscow for
520 days. This includes a 485 days period of travelling to and from Mars as well as a
35 days period of Mars exploration. In order to prepare for this study, in the first half
of 2009 a 105 day isolation study was performed to evaluate procedures and
protocols. During this study we were able to perform several measurements of
electro-cortical activity using electroencephalogram (EEG) as well as a profile of
mood state (MoodMeter®) simultaneously before and after exercise (Figs. 19.8,
19.9, 19.10, 19.11, 19.12 and 19.13) (Video 19.2).
Cortical activity was defined by subdividing the raw EEG signal into its specific
frequency ranges alpha [8–13 Hz] and beta [13–35 Hz]. While alpha activity
generally reflects normal brain function, an increase in beta activity indicates
excitatory CNS activity (Lindsley 1960; Zuckermann 1991; Bonnet and Arand
2001).
Using the MoodMeter®, a tool that was specifically designed to evaluate short
term changes in mood (Schneider et al. 2008b, 2009a), we were able to evaluate the
effects of exercise on perceived physical state, perceived motivation and perceived
psychological state. Exercise consisted of a 10–12 min incremental exercise pro-
tocol starting with 30 W and increasing by 15 W every minute until 80 % of
individuals’ maximal intensity on a bike ergometer. Data was recorded 12 days
prior to exercise, on days 30, 64, 77 and 99 during the 105-day isolation as well as
8 days post isolation. A detailed description about methods and protocol can be
found in Schneider et al. (2010b).
Results, displayed in Table 19.1, showed a clear influence of exercise on brain
cortical function and perceived physical state. Two distinct processes seem note-
worthy: At first, we could observe a decrease of activity in both EEG frequency
ranges across the time frame of isolation. Whereas a general model of cortical
Fig. 19.8 Logo of the

German Sport Universities
contribution to the
MARS500 program
Utility module (EU-250)

gym Medical module (EU-100)
Simulator of the greenhouse
storage for resources habitable compartment
Martian surface fridge kitchen-dining-room
thermal chamber working places with
lavatory medical equlpment
lavatory
Habitable module (EU-150)

6 individual compartments
community room
main console
Simulator of the kitchen
landing Martian lavatory
ship (EU-50)
Fig. 19.9 The Mars500 isolation facility located in the Institute of Biomedical Problems (IBMP)
in Moscow. Copyright ESA
Fig. 19.10 Exercise within the MARS500 facilities. Copyright ESA
arousal would assume an increase in the beta frequency range to result in a decrease
of alpha activity, this was not the case here. With regard to the data (Table 19.1) it
seems that the prolonged isolation results in an overall decrease of brain cortical
activity, which was supposed to be a result of the confined situation with its specific
stressors (loneliness, social stress, boredom, sensory deprivation etc.). Second, at
any time point during the isolation, a small bout of exercise of about 10–12 min was
able to increase brain cortical function back to baseline, pre-isolation values.
In accordance with previous studies showing that exercise intensity and duration
(Ekkekakis and Acevedo 2006; Ekkekakis and Petruzzello 1999) as well as exercise
preferences (Schneider et al. 2009c) seem to be important for changes in brain
cortical activity, data provided here shows that even short bouts of exercise with
approximately 80 % maximal intensity seem to have a positive effect on brain
cortical activity.
Similar effects were found in all three dimensions of the MoodMeter®,
although only perceived physical state but not perceived motivational state and
psychological state reached significance (probably due to the low number of sub-
jects). Similar to brain cortical activity, a decrease throughout the 105 days of
isolation could be observed but already a small bout of exercise was able to
Fig. 19.11 Daily exercise routine during the MARS500 isolation study. Copyright ESA
improve perceived physical state, motivational state and psychological state

(Table 19.1). A subsequent correlation analysis, combining objective physiological
values and subjective perception, revealed a significant correlation between both,
brain cortical activity and mood values, showing the validity of such a psycho-
physiological approach.
Beside these changes in brain cortical activity and affective state, two further
studies from our lab let us assume that also an improvement in cognitive
Fig. 19.12 German participant Oliver Knickel during the MARS105 isolation study just before an
exercise intervention test. Copyright ESA
performance might be caused by exercise induced changes in brain cortical activity.

After 15 min exercise at moderate intensities (heart-rate 160–170 bpm) 10 school
kids showed an increase in alpha activity in the precuneus while beta activity in left
temporal areas, including the Wernicke area, which is involved into language
Fig. 19.13 Russian participant of the MASR500 isolation study during chest workout. Copyright
ESA
19
Table 19.1 Results of the MARS105 study

Perceived physical Perceived physical Perceived motivational Perceived motivational Perceived psychological Perceived psychological
state state state state state state
PREEXERCISE POSTEXERCISE PREEXERCISE POSTEXERCISE PREEXERCISE POSTEXERCISE
3.11+/ 1.67 3.12+/ 1.74 3.33+/ 0.93 3.83+/ 1.19a 2.79+/ 1.30 3.31+/ 1.05
4.14+/ 1.12 4.53+/ 0.90 3.77+/ 1.04 3.92+/ 0.66 3.81+/ 1.43 4.10+/ 1.05
4.14+/ 0.79 4.45+/ 0.92a 3.56+/ 0.43 3.77+/ 0.71 3.48+/ 0.99 3.73+/ 0.84
3.98+/ 0.81 4.41+/ 0.63a 3.27+/ 0.56 3.60+/ 0.97 3.48+/ 1.70 3.88+/ 1.05
4.14+/ 1.17 4.39+/ 1.11a 3.70+/ 1.52 3.81+/ 0.94 3.65+/ 1.38 4.02+/ 1.33
4.24+/ 0.93 4.50+/ 0.56 3.75+/ 0.94 3.85+/ 1.08 3.79+/ 0.89 4.17+/ 0.89
Alpha activity Alpha activity Beta1 activity Beta1 activity Beta2 activity Beta2 activity
[ln(μV2)] [ln(μV2)] [ln(μV2)] [ln(μV2)] [ln(μV2)] [ln(μV2)]
Exercise in Space: Physical and Mental Benefit
PREEXERCISE POSTEXERCISEb PREEXERCISE POSTEXERCISEb PREEXERCISE POSTEXERCISEb

2.32+/ 0.90 2.65+/ 0.81 4.88+/ 9.73 5.79+/ 10.62 1.43+/ 1.08 1.68+/ 1.03
2.17+/ 1.35 2.43+/ 1.17 3.64+/ 5.91 5.23+/ 9.65a 1.17+/ 0.97 1.79+/ 1.09a
1.90+/ 2.66 2.18+/ 2.08 2.43+/ 3.01 3.96+/ 5.48a 0.82+/ 1.91 1.37+/ 1.80a
1.81+/ 2.85 2.23+/ 1.69a 2.46+/ 4.63 2.66+/ 2.80 0.76+/ 2.27 1.10+/ 1.56a
2.40+/ 1.73 2.36+/ 1.83 4.46+/ 8.43 5.05+/ 9.91 1.26+/ 1.70 1.39+/ 1.50
2.22+/ 0.73 2.43+/ 0.92 4.09+/ 4.72 4.75+/ 6.58 1.38+/ 1.04 1.55+/ 0.95
Values represent means +/ 95 % confidence intervals
a
Indicate significant differences (LSD post-hoc: p < .05) between individual PRE and POSTEXERCISE sessions
b
Indicate significant differences (ANOVA: p < .05) comparing all sessions PREEXERCISE vs. POSTEXERCISE
237
Fig. 19.14 Participants of the school sport study during a game of catch
processing, was found to be reduced (Schneider et al. 2009d).2 The increase of

synchronised alpha activity (which is the dominating frequency in a relaxed
wakeful state) in the precuneus might be a sign of increased bodily relaxation.
These neurophysiological changes could explain well-known effects of physical
activities in school on the ability to concentrate and relax. With respect to a
resource theory (e.g. by Kahnemann (1993)) the decrease of high frequency activity
in the left temporal gyrus might be a sign of increased capacities available post
exercise. If we consider cognitive tasks to be closely related to speech processing,
especially in school, we can define a strong link between decreased activity in
Wernickes’ area and an increase in cognitive processing. Results of this study were
validated in a second study with 22 health orientated runners showing similar
results (Schneider et al. 2010a) (Fig. 19.14).
2
A localisation of brain cortical activity was made possible by low resolution brain electromag-
netic tomography, a software based solution offered as academic freeware by the KEY Institute for
Brain-Mind Research at the University of Zurich, Switzerland, (http://www.uzh.ch/keyinst/loreta),
that enables the three dimensional localisation of brain cortical activity by using standardised EEG
recordings. By subdividing the EEG in its standardised frequency bands alpha and beta it becomes
possible to see which part of the brain is active/inactive. Although the spatial resolution is limited
(in this study we used a minimum of 32 electrodes for localising brain activity), this approach
combines the simple and economic EEG recording with the possibility to record and localise brain
cortical activity even in experimental setups that do not allow for standardised brain imaging like
positron emission tomography (PET) or functional magnetic resonance imaging (fMRI), either due
to space and economic restrictions or methodical limitations as the application in weightlessness
(Schneider et al. 2008a).
19.5 Exercise Needs to be Individualised in Order

to Benefit from Its Impact on Mood and Cognitive
Performance
Something that has barely taken into account so far is the huge amount of exercise
(approx. 2 h/day) that would be necessary to (insufficiently) counteract musculo-
skeletal and cardiovascular deconditioning during long term spaceflight. This might
be experienced by individuals as very challenging and stressful during a space
mission with a tight schedule and high workload. As a result in the recent years
there have been many attempts to reduce the amount of time spend on exercising by
simultaneously maximizing its efficiency. For example a combination of resistive
exercise training and vibration training was proposed lately (Belavy et al. 2009; van
Duijnhoven et al. 2010). Although results so far remain inconsistent and do not
clearly prove an advantage of vibration training against traditional resistance
training (Mulder et al. 2009), at least the time-effectiveness of vibration augmented
resistance training must be highlighted. Reconsidering preceding thoughts about
the effects of exercise on mood and neurocognitive function, further research
activities should also take into account these factors. Especially as vibration
enhanced training will reduce training duration and therefore the impact of exercise
on cardiovascular regulation, which might be a constitutive principle of
neurocognitive amelioration (Ekkekakis and Acevedo 2006).
Another attempt that has been followed for several years now is the use of
artificial gravity, produced for example by a short arm human centrifuge (SAHC)
(Video 19.3).
Two centrifuges, fitting in a room not more than 40sqm have been developed
lately by ESA and are currently used to evaluate the effects of centrifugation on the
cardiovascular, musculoskeletal and central nervous system. As first results indi-
cate, artificial gravity results in a physical activation close to moderate aerobic
exercise. The nice thing about being “passively exercised” – at least at moderate
G-levels of 1–1.5G – is that the individual is in a supine or seated position and may
spend some time relaxing (reading a book, viewing a DVD, etc.) (Fig. 19.15).
One might also think about being centrifuged while sleeping. But although there
might be an impressive physiological effect, and artificial gravity might provide a
moderate exercise workout, we were able to show lately that the benefit on affective
state and mental performance fails to appear (Vogt et al. 2014) and therefore
artificial gravity might not replace exercise, at least not when taking into account
a holistic approach of exercise.
Nevertheless recent research has shown that a positive effect of exercise on
mental performance and affective state is dependant on two major principles: a
dose-response relationship (Ekkekakis and Petruzzello 1999) AND a positive bias
towards exercise (exercise preference hypothesis (Schneider et al. 2009c)). That
Fig. 19.15 The Short Arm

Human Centrifuge (SAHC)
at the German Aerospace
Center DLR in Cologne.
Copyright DLR
means in order to have a most beneficial effect on mood, cognitive performance and
stress reduction, we need to take into account not only the individual fitness level
and let the individual monotonically exercise at a given intensity (e.g. 80 %
VO2max) – we also need to be aware of the individuals preferences, which include
intensity and exercise mode.
In a recent study we let 21 subjects run at three different intensity levels of 50 %
and 80 % of VO2max as well as their preferred intensity. As described before, we
recorded brain cortical activity and affective state using the MoodMeter®. The
results clearly demonstrated that the most beneficial effects of exercise on mood
occurred while individuals exercised at a self-determined intensity (Schneider
et al. 2009b). In a subsequent study we were able to demonstrate that effects of
exercise on prefrontal cortex activity (Dietrich 2006; Ekkekakis and Acevedo 2006)
are most pronounced when individuals performed their preferred exercise routine
(running) rather than an unusual exercise routine (biking, arm-crank exercise)
(Schneider et al. 2009c). This might not make sense from a physiological point of
view, from a psychological point it does. Any exercise recommendation that is
solely orientated on maximising the physiological benefit of exercise is useless if
the individual, due to any kind of antipathy, stops exercising after a few weeks.
The most beneficial effect of exercise will be obtained by an exercise routine that is,
due to a predilection, performed regular and with empathy.3 Also the implementa-
tion of virtual realities might support the individuals’ motivation, e.g. jogging a
familiar home track by using a head mounted display.
19.6 Where Do We Go from Here? Implications for Future

Exercise Applications
At first there is the essential need for national as international space agencies to
define future aims of their “exercise in space” programs. As demonstrated, physical
exercise is a helpful tool to accelerate recovery to earth gravity after long duration
spaceflight. At this point research activities need to be intensified, not only showing
the amount of loss of muscle mass or bone stability during space flight, but
highlighting the impact of specific exercise programs on the timeline and efficiency
of rehabilitation processes. The same is true for cardiovascular function. How long
do orthostatic problems remain and can specific exercise programs influence this?
Research should also try to identify to what extend recovery processes after space
flight might be enhanced by a specific long term training program prior to space-
flight. Recent evidence let us assume that a physical active lifestyle does not only
prevent from musculoskeletal and cardiovascular deconditioning but its positive
effects are carried though times of immobilisation and accelerate recovery
processes.
If the agenda proposes to further investigate a holistic approach of exercise,
i.e. how exercise might enhance crew performance, mental health and mood, it
seems of utterly importance (1) to evaluate the effects of specific exercise programs
on brain cortical function and its impact on cognitive performance, mental health
and mood and (2) to have in mind that an individual dose-response relationship and
an individual exercise preference exists. Instead of torturing our astronauts and
cosmonauts through a variety of exercise approaches, there is a need to individu-
alise exercise programs including an individual bias.
3
To give an example: For long years a fundamental problem of physical education in school was
that the benefit of exercise was neither highlighted nor identified. Throughout the 1980s and 1990s
physical education concentrated on a merit principle. Just in the recent years, with increasing
health problems caused by a lack of physical activity (e.g. cardiovascular disease, diabetes), the
curriculum of physical education started to promote a physical active lifestyle that will be carried
into adulthood in order to prevent degenerative diseases caused by an absence of physical exercise
(Pediatrics 2000)
19.7 A Visionary Preview to 2110
Health, as defined by a very early definition of the World Health Organisation

(1946) is defined as a state of physical, mental and social well-being. Beside its
positive effects on physical and mental well-being, exercise is also assumed to
contribute to social well-being (Brembs 2009). Especially team sport is known to
contribute to the social dimension of exercise further impacting on both, physical
and mental well-being as team spirit keeps up commitment. Due to space limita-
tions (!) team sports so far have not been considered in space. But guided by the joy
and motivation going along with team sports and their positive impact on physio-
logical as well as psychological and sociological dimensions of health, we would
propose to develop a three dimensional version of soccer, football, handball or
basketball, played in multinational and multicultural teams. Obviously today space
cannot be provided for such activities but as soon as human mankind is going to
populate a permanent base on moon or Mars, gravity specific variations of well-
known sport games are most likely to be invented.
References
Belavy DL, Miokovic T, Armbrecht G, Rittweger J, Felsenberg D (2009) Resistive vibration

exercise reduces lower limb muscle atrophy during 56-day bed-rest. J Musculoskelet Neuronal
Interact 9(4):225–235
Bonnet MH, Arand DL (2001) Impact of activity and arousal upon spectral EEG parameters.
Physiol Behav 74(3):291–298
Brembs B (2009) The importance of being active. J Neurogenet 23(1–2):120–126
Cheron G, Leroy A, De Saedeleer C, Bengoetxea A, Lipshits M, Cebolla A, Servais L, Dan B,
Berthoz A, McIntyre J (2006) Effect of gravity on human spontaneous 10-Hz electroenceph-
alographic oscillations during the arrest reaction. Brain Res 1121(1):104–116
Dietrich A (2006) Transient hypofrontality as a mechanism for the psychological effects of
exercise. Psychiatry Res 145(1):79–83
Ekkekakis P, Acevedo EO (2006) Affective responses to acute exercise: toward a psychobiolog-
ical dose-response model. In: Acevedo EO, Ekkekakis P (eds) Psychobiology of physical
activity. Human Kinetics, Champaign, pp 91–109
Ekkekakis P, Petruzzello SJ (1999) Acute aerobic exercise and affect: current status, problems and
prospects regarding dose-response. Sports Med 28(5):337–374
Fowler B, Manzey D (2000) Summary of research issues in monitoring of mental and perceptual-
motor performance and stress in space. Aviat Space Environ Med 71(9 Suppl):A76–A77
Gopalakrishnan R, Genc KO, Rice AJ, Lee SM, Evans HJ, Maender CC, Ilaslan H, Cavanagh PR
(2010) Muscle volume, strength, endurance, and exercise loads during 6-month missions in
space. Aviat Space Environ Med 81(2):91–102
Halberg F, Vallbona C, Dietlein LF, Rummel JA, Berry CA, Pitts GC, Nunneley SA (1970)
Human circadian circulatory rhythms during weightlessness in extraterrestrial flight or bedrest
with and without exercise. Space Life Sci 2(1):18–32
Kahnemann D (ed) (1993) Attention and effort. Prentice-Hall, New Jersey
Kanas N (1998) Psychiatric issues affecting long duration space missions. Aviat Space Environ
Med 69(12):1211–1216
Lindsley DB (1960) Attention, consciousness, sleep and wakefulness. In: Magoun HW

(ed) Handbook of physiology. Sect 1 edn. Am Physiological Society, Washington, DC,
pp 1553–1593
Manzey D (2000) Monitoring of mental performance during spaceflight. Aviat Space Environ Med
71:A69–A75
Mulder ER, Horstman AM, Stegeman DF, de Haan A, Belavy DL, Miokovic T, Armbrecht G,
Felsenberg D, Gerrits KH (2009) Influence of vibration resistance training on knee extensor
and plantar flexor size, strength, and contractile speed characteristics after 60 days of bed rest.
J Appl Physiol 107(6):1789–1798
Palinkas LA (2001) Psychosocial issues in long-term space flight: overview. Gravit Space Biol
Bull 14(2):25–33
Pediatrics AAO (2000) Physical fitness and activity in schools. Pediatrics 105(5):1156–1157
Rummell JA, Sawin CF, Buderer MC, Mauldin DG, Michel EL (1975) Physiological response to
exercise after space flight–Apollo 14 through Apollo 17. Aviat Space Environ Med 46(5):
679–683
Schneider S, Askew CD, Abel T, Mierau A, Struder HK (2010a) Brain and exercise: a first
approach using electrotomography. Med Sci Sports Exerc 42(3):600–607
Schneider S, Askew CD, Brummer V, Kleinert J, Guardiera S, Abel T, Struder HK (2009a) The
effect of parabolic flight on perceived physical, motivational and psychological state in men
and women: correlation with neuroendocrine stress parameters and electrocortical activity.
Stress 12(4):336–349
Schneider S, Askew CD, Diehl J, Mierau A, Kleinert J, Abel T, Carnahan H, Struder HK (2009b)
EEG activity and mood in health orientated runners after different exercise intensities. Physiol
Behav 96(4–5):709–716
Schneider S, Brummer V, Abel T, Askew CD, Struder HK (2009c) Changes in brain cortical
activity measured by EEG are related to individual exercise preferences. Physiol Behav
98(4):447–452
Schneider S, Brummer V, Carnahan H, Dubrowski A, Askew CD, Struder HK (2008a) What
happens to the brain in weightlessness? A first approach by EEG tomography. Neuroimage
42(4):1316–1323
Schneider S, Brummer V, Carnahan H, Kleinert J, Piacentini MF, Meeusen R, Struder HK (2010b)
Exercise as a countermeasure to psycho-physiological deconditioning during long-term con-
finement. Behav Brain Res 211(2):208–214
Schneider S, Guardiera S, Kleinert J, Steinbacher A, Abel T, Carnahan H, Struder HK (2008b)
Centrifugal acceleration to 3Gz is related to increased release of stress hormones and decreased
mood in men and women. Stress 11(5):339–347
Schneider S, Vogt T, Frysch J, Guardiera P, Struder HK (2009d) School sport–a neurophysiolog-
ical approach. Neurosci Lett 467(2):131–134
van Duijnhoven NT, Thijssen DH, Green DJ, Felsenberg D, Belavy DL, Hopman MT (2010)
Resistive exercise versus resistive vibration exercise to counteract vascular adaptations to bed
rest. J Appl Physiol 108(1):28–33
Verheyden B, Liu J, Beckers F, Aubert AE (2010) Operational point of neural cardiovascular
regulation in humans up to 6 months in space. J Appl Physiol 108(3):646–654
Vogt T, Abeln V, Struder HK, Schneider S (2014) Artificial gravity exposure impairs exercise-
related neurophysiological benefits. Physiol Behav 123:156–161
Zuckermann M (1991) Psychobiology of personality. Cambridge University Press, New York,
Port Chester, Melbourne, Sydney
Part III
Performance and Coaching in Various
Sports
Chapter 20
Energetic Considerations in Cross-Country
Skiing
Walter Herzog, Anthony Killick, and Kevin R. Boldt
Abstract Cross-country skiing is a four-legged gait, and some gait patterns, such
as 2-skate skiing, are similar to those adopted by animals (galloping horse). Four-
legged animals change gait patterns with increasing speeds of locomotion, at least
in part, to minimize metabolic energy expenditure. For example, a horse will switch
from a walk, to a trot, and finally to a gallop as speed of locomotion increases.
Similarly, skate cross-country skiers will switch from a 2-skate gait to a 1-skate gait
with increasing speeds of locomotion, but then unlike any other animal, will revert
back to the previously rejected 2-skate gait pattern at very high speeds. We used
oxygen uptake measurements, force measurements in poles and skis, 3-dimensional
movement analysis and functional muscle properties to explain this result. We
found that propulsion in 1-skate skiing comes primarily from the arms/poles,
while propulsion comes primarily from the legs/skis in the 2-skate technique. We
also found that ground contact times for the skis are virtually independent of the
skiing speed while pole contact times decrease dramatically with increasing speeds.
Furthermore, propulsive forces from the arms dropped from skiing at 15 km/h to
skiing at 30 km/h while simultaneously requiring much more metabolic energy.
Finally, the cost of transport curves for 1-skate and 2-skate skiing intersected twice,
indicating better efficiency for the 2-skate technique at slow and very fast speeds,
and better efficiency for the 1-skate technique at intermediate to fast speeds.
Combined, these results suggest that arm/pole action is optimized at intermediate
speeds, thereby providing an advantage to the 1-skate technique which relies
primarily on arm propulsion, while arm/pole action is highly inefficient at very
high speeds, thus switching to the 2-skate technique which relies primarily on leg
propulsion, is good strategy.
Keywords Cost of transport • Efficiency • Optimal movement • Entrained

breathing • Skate technique • Classic technique • Functional muscle mechanics •
Force-velocity relationship • Power-velocity relationship • Force-length
relationship
W. Herzog (*) • A. Killick • K.R. Boldt

Human Performance Lab, University of Calgary, Calgary, Canada
e-mail: wherzog@ucalgary.ca

248 W. Herzog et al.
20.1 Introduction
Cross-country skiing is an ancient form of locomotion on skis across snow which is

thought to have originated prior to formal historical records (Dresbeck 1967). In the
nineteenth and early twentieth century, cross-country skiing was used as a way of
transportation and hunting, particularly in the Nordic countries of Europe. In 1924,
at the first winter Olympic Games, cross-country skiing was represented with a
50 km and an 18 km race (for men only). In the past few decades, cross-country
skiing has become a recreational sport of increasing popularity in “winter sport”
countries and is recognized for its whole body involvement and its high aerobic
endurance benefits.
Traditionally, competitions in cross-country skiing were performed using a set
of snow tracks in which skis moved parallel to each other, and arms and poles were
used alternately in a running type fashion. This technique of skiing is now referred
to as the classic technique and is characterized by a parallel gliding of the left and
right ski and the skis always pointing in the direction of the tracks. In the 1970s, the
Finnish skier Pauli Siitonen introduced the skating technique of cross-country
skiing into racing. In contrast to the classic technique, in the skating technique,
the skis are used in a “skating” V-like fashion: that is the skis are angled to each
other, and the direction of movement becomes partly sideways, rather than directly
along the path of the race course. In 1987, the skating technique was used for the
first time at a world championship, and then was also used for the first time at the
Olympic Games 1988 in Calgary, Canada.
Here, we will deal exclusively with the skating technique of cross-country
skiing, which in racing is also referred to as the “free technique”. In cross-country
ski skating, different gait patterns are used by athletes and accomplished recrea-
tional skiers. These gait patterns are classified based on the coordination of the arm
and pole actions with the leg and ski actions. At very slow speeds, typically on steep
uphill sections, recreational skiers use a single poling technique, similar to the
herringbone technique but with a small glide phase on each ski. This technique is
not used by racers, and will not be further considered here. At slow to intermediate
speeds, or when skiing in a relaxed manner or on an uphill section, skiers use the
so-called 2-skate technique (also referred to as “V1” or “V2 alternate” technique, or
offset technique when the poling action is slightly offset to the leg action) in which
skiers double pole on every other leg. Most skiers have a preferred side for double
poling and so will typically double pole either on the left or on the right leg. At
intermediate to fast speeds of skiing, skiers switch to the so-called 1-skate tech-
nique (also referred to as V2 or Wassberg technique) in which a double pole action
accompanies each step on the left and right leg. At fast to very fast (sprint) speeds,
skiers revert back to the 2-skate technique that was previously rejected for the
1-skate technique at intermediate to fast speeds. Finally, at very fast speeds, and
mostly on slight downhill sections, skate skiers use the so-called “free skating”
technique in which the poles are not used at all, and propulsion is exclusively
derived from the skating actions of the legs (Cross-country skiing 2014).
20 Energetic Considerations in Cross-Country Skiing 249
Fig. 20.1 Cost of transport (ml O2/m) for a horse walking, trotting and galloping at increasing
speeds of locomotion. Note that the transition from one gait pattern to the next occurs at speeds in
the vicinity of intersection between the cost of transport curves for gait patterns, indicating that
gait transitions might occur to minimize the energy required for locomotion at a given speed
(Adapted from Hoyt and Taylor 1981, with permission)
As mentioned above, cross-country skiing has been advocated as an excellent

aerobic training exercise because of its whole body engagement. Elite cross-
country ski racers therefore must be excellent endurance athletes with superior
oxygen uptake capacity. The highest ever recorded oxygen uptake capacity was
recorded for Bjørn Dæhlie, the most successful cross-country skier ever, with
12 Olympic medals (8 of them gold) and 17 (9 gold) world championship medals,
at an astonishing 96 ml/kg/min (Bjørn 2014). Because of the high demands on
oxygen uptake, it is reasonable to assume that cross-country skiers chose a gait
pattern while racing that minimizes oxygen uptake, similar to a horse that switches
from a walking gait, to a trot, and finally to a gallop with increasing speeds of
locomotion (Hoyt and Taylor 1981) (Fig. 20.1).
If ski racers indeed choose their gait pattern in skate skiing in a manner to
minimize the energy requirements for skiing at a particular speed, then it is
surprising that they choose the 2-skate technique (at slow to intermediate), and
then the 1-skate (at intermediate to fast) speeds, but then revert back to the 2-skate
technique for fast to very fast (sprinting) speeds. This is equivalent to saying that a
four legged animal, such as a horse, changes from a trot to a gallop and then back to
a trot for increasing speeds of running, something that would never happen.
Therefore, the question arises: why do cross-country skiers revert back to a tech-
nique (or gait pattern – 2-skate skiing) at very fast speeds, that was rejected at a
slower speed in favour of another gait pattern (the 1-skate technique)?
The purpose of this study was to answer this question. We hypothesized that the
reason for choosing the 2-skate over the 1-skate at slow and very fast speeds, but not
at intermediate speeds, was associated with the cost of transport. In other words, we
hypothesized that the 2-skate technique was more efficient at slow and very high
speeds of skiing compared to the 1-skate technique, while the reverse was true for
intermediate to fast speeds of skiing. A secondary question then became: is it
possible to explain why the 2-skate technique is more efficient at slow and very
high, but not at intermediate to fast speeds compared to the 1-skate technique? This,
we felt, was an intriguing question, as it is, to our knowledge, the only form of
locomotion, two-legged or four-legged, in which a gait pattern that was rejected at a
slow speed is re-introduced at a high speed.
Cross-country skiing is a four-legged gait with arms and legs contributing to
propulsion and the speed of locomotion. In four-legged gaits of dogs, horses and
rabbits, it has been observed that breathing is coordinated with the footfall patterns
(Ainsworth 1997; Attenburrow and Goss 1994; Bramble and Carrier 1983; Bramble
and Jenkins 1993). Specifically, in galloping animals, inhaling is associated with an
expansion of the chest cavity when the forelimbs are moved forward relative to the
body, while exhaling is associated with compression of the chest cavity as occurs
when the forelimbs are swung backwards relative to the body. 2-skate skiing has a
footfall pattern similar to a galloping horse, and the arms swing backwards and
forwards in unison in their double pole action. Therefore, it seems feasible to
assume, and it has been suggested, that inhaling and exhaling in 2-skate skiing is
directly tied to the arm movements (Faria 2008). It has been argued that this
“respiration coupling” in animals results in a reduced metabolic requirement for
locomotion, as the movement of the forelimbs in galloping animals causes natural
chest expansion and compression that assists the respiratory muscles and thus
reduces the metabolic cost of breathing (Ainsworth 1997; Bramble and Carrier
1983). However, whether this assertion is correct, and what magnitude this effect
might have, cannot be tested in animals, as animals cannot be asked to abandon the
natural respiration coupling pattern while galloping. However, in cross-country
skiers performing the 2-skate technique, it is easy to measure the oxygen uptake
when breathing occurs naturally with respiration coupled to the skiing motion, and
when skiers are asked to abandon the natural coupled breathing patterns. Thus, the
purpose of a second study described here was to test the hypothesis that respiration
coupling inferred a distinct energetic advantage to cross-country skiers using the
2-skate skiing technique.
20.2 Methods
In order to accomplish the aims of this study, it was necessary to measure the
kinematics and kinetics of cross-country skiing using the 1- and 2-skate techniques,
and measure the oxygen uptake required for these two gait patterns. It was also
necessary to make these measurements in a subset of skiers for the 2-skate
technique while using their normal respiration coupling patterns and while skiing at
the same speed and effort while abandoning the normal respiration coupling
pattern. All measurements were performed using roller skiing on a motor driven
skiing treadmill at the facilities of the National Centre for Excellence in Sport,
Calgary and Canmore.
Kinematic and Kinetic Measurements Kinematic measurements were
performed using two high speed video cameras, one placed with the optical axis
perpendicular to the sagittal plane of skiing, the other aligned with the optical axis
in the direction of skiing, filming the skiers from behind. Forces exerted by each
roller ski and each pole were obtained using instrumented skiing poles and roller
skis. Force transducers inserted into the shaft of the poles measured the forces along
the axis of the poles while compensating for bending forces at the pole. Forces in
the roller skis were measured by replacing the normal roller ski braces that connect
the ski part with the rollers with strain gauged braces that measured the forces
perpendicular to the roller skis in the vertical direction, and in the medial-lateral
(horizontal) plane relative to the roller skis. Anterior-posterior forces were not
measured as roller skis are essentially frictionless and effective propulsion in the
anterior-posterior direction is not possible in skate skiing, in contrast to classic
skiing where anterior-posterior forces are essential for effective striding.
Oxygen Uptake Measurements Oxygen uptake measurements were obtained
while skiers skied at steady-state, typically after 4–5 min into a specific experi-
mental condition (i.e. speed of skiing and slope) using a ParvoMedics TrueOne
2400 Metabolic Measurement System. Oxygen uptake was collected and expired
air was analyzed every thirty seconds. Expired air was converted to standard
temperature, pressure and dry (STPD) conditions, and analyzed to determine the
rate of oxygen consumption and metabolic energy consumed (WEIR 1949).
Protocols For comparison of the efficiency of the 1-skate and 2-skate techniques
across a range of speeds, skiers (n ¼ 8, young, active, elite level provincial and
national skiers) skied using the 1- and 2-skate techniques at speeds of 6–33 km/h at
increments of 3 km/h. They were asked to ski at each speed below the anaerobic
threshold for 3 min with data collection occurring once the metabolic steady-state
was achieved (typically after 2–2.5 min) (Solberg et al. 2005). For speeds above the
anaerobic threshold, skiers just skied for 1 min at each speed in a precisely timed
manner so that conditions for the 1- and 2-skate tests were identical, and thus
comparable.
In order to estimate the oxygen cost for just the upper body and poling action for
skate skiing using the 1- and 2-skate techniques, skiers performed an additional
series of tests with oxygen uptake measurements. In these tests, the skiers stood on
fixed skis and pulled on a cross-country arm ergometer at the frequency and
excursion of their own skiing at low, intermediate and high speeds (6, 15, and
30 km/h). They did this by viewing a video of their own skiing, and pretending to
follow the displayed skier in a perfectly matched manner. In order to obtain the
proper forces, the pole forces measured during the actual skiing test were fed back
to the skiers, and the resistance on the ergometer was adjusted until they satisfac-
torily matched the actual skiing forces. The legs were braced for this experiment so
that no leg propulsion was possible, ensuring that we measured exclusively the
oxygen cost of the arm and upper body motions at the target speeds.
In order to derive the functional force-velocity and power-velocity relationships
of the poling action for the 1- and 2-skate techniques, skiers were fixed on a roller
board on top of the skiing treadmill. The treadmill was then run below the fixed
skiers at speeds ranging from 6 to 42 km/h at increments of 6 km/h and skiers were
asked to perform double poling actions at maximal effort, controlled by the rhythm
of a metronome to give them the frequency of poling in the 1- and 2-skate
techniques. The forces for ten consecutive maximal effort poling actions at each
speed were measured, and the corresponding impulses calculated. From these force/
impulse-velocity relationships of the poling action, the corresponding power output
as a function of skiing speeds was calculated.
For comparison of the efficiency of skiing with and without respiration coupling
in the 2-skate technique, skiers (n ¼ 9, young, active, elite level provincial and
national skiers) were asked to ski at a pre-determined fast but sub-anaerobic
threshold (Solberg et al. 2005) speed for 5 min on the motor driven treadmill.
The speeds ranged from 18 to 22 km/h on a zero slope. Skiers performed the test
first using their normal breathing patterns and respiration coupling was confirmed
with the breath by breath analyzer. They then repeated this test twice more, first
with respiration coupling consciously abolished and breathing in a reverse coupled
manner: that is inhaling occurred during the push phase of the poles when the arms
swung backwards and compressed the chest cavity, and exhaling occurred in the
recovery phase, when the arms were swung forwards thereby expanding the chest
cavity, and second, in the control condition with breathing coupling re-established.
20.3 Results
As hypothesized, the cost of transport (or equivalently, the oxygen requirement for
a given speed of skiing) for the 1-skate and 2-skate techniques had two intersec-
tions, with the 2-skate technique being the more efficient gait pattern at the slow
(6, 9 km/h) and high speeds (>24 km/h), while the 1-skate technique was the more
efficient technique at the intermediate speeds (9–21 km/h) (Fig. 20.2). Out of the
eight skiers, four showed precisely this double intersection of the cost of transport
curves (Fig. 20.3), while the remaining four skiers, although not vastly different
from the first four, showed an approximation of those curves, but not the double
intersection.
Propulsion in the 1-skate technique came primarily from the arms and poles,
while propulsion was primarily derived from the legs and skis for the 2-skate
technique (Fig. 20.4). Interestingly, for both, the 1- and 2-skate techniques, the
propulsion contributed to skiing from the arms and poles did not increase between
skiing at speeds of 15 and 30 km/h, while, as one would expect, propulsion
Fig. 20.2 Mean (n ¼ 8) cost of transport values as a function of skiing speed for the 1-skate (black
dots) and the 2-skate (grey squares) skiing techniques. Note that the 2-skate technique was more
efficient (lower values) at the slowest (6, 9 km/h) and the highest (>27 km/h) speeds, while the
1-skate technique was metabolically more efficient at intermediate speeds (12–21 km/h)
Fig. 20.3 Cost of transport values as a function of skiing speed for a single skier. Note the
similarity of intersection of the 1-skate and 2-skate curves with the curves found for the mean of
the group
contributed by legs and skis increased with increasing speeds of skiing throughout
the tested speed range.
Oxygen uptake for the upper body and arm action increased with increasing
speeds of skiing (Fig. 20.5). For all speeds measured, the 1-skate technique required
more oxygen than the 2-skate technique.
The maximal impulse that can be provided by the arms and poles decreases with
increasing speeds of skiing, but is slightly higher for the 2-skate technique at the
Fig. 20.4 Impulse production by arms and legs in the 1-skate and 2-skate ski skate techniques.
Note that in the 1-skate technique (left), propulsion is derived primarily from the arms and poles,
while in the 2-skate technique (right), propulsion comes primarily from the leg and ski action.
Furthermore, note that propulsion from the skis increases with increasing speeds of skiing, while
there is a slight decrease for the propulsive impulses from the arms from the 15 to the 30 km/h
speeds
Fig. 20.5 Oxygen

consumption for the upper
arm and trunk action
exclusively for cross-
country skiing at increasing
speeds using the 1-skate and
2-skate technique. Oxygen
uptake increases with
increasing speeds of
locomotion and becomes
significantly higher for the
1-skate compared to the
2-skate technique at a speed
of skiing of 30 km/h
slow and high speeds, while it is slightly higher for the 1-skate technique at the
intermediate speeds of skiing (Fig. 20.6). The corresponding power-velocity rela-
tionship for the poling action is remarkably constant across speeds for the 2-skate
technique, but reaches a distinct maximum at about 18 km/h for the 1-skate
technique (Fig. 20.7). Double poling power output is higher for the 1-skate tech-
nique at intermediate speeds (18 and 24 km/h) and equal or lower for the slow and
very high speeds of skiing.
Respiration coupling was associated with an approximately 4 % lower oxygen
cost compared to uncoupled breathing (Fig. 20.8). When accounting for the slight
Fig. 20.6 Impulse provided by maximal pole action with increasing speeds of skiing for the
1-skate and 2-skate techniques. Note that the 1-skate and 2-skate impulses differ because of the
frequency of poling at a given speed which is always higher for the 1- compared to the 2-skate
technique. Note furthermore, that the 2-skate technique has higher impulses than the 1-skate
technique for the slowest (6 km/h) and highest speeds (>36 km/h) tested
Fig. 20.7 Power output by arms and poles as a function of skiing speed for the 1- and 2-skate
techniques of skate skiing. Note, the relatively flat power curve for the 2-skate and the more curved
power graph for the 1-skate technique, reaching a peak value at about 18 km/h. Note also, that the
power output is greater for the 1- skate compared to the 2-skate technique for speeds of 18 and
24 km/h while for speeds <18 km/h or greater than 24 km/h the 2-skate power values are greater
than the 1-skate values
shifts in the respiration exchange ratio, respiration coupling was also associated
with an approximately 5 % decrease in metabolic cost compared to the uncoupled
(or reverse) breathing pattern (result not shown). Careful analysis of the footfall
patterns and poling patterns and associated forces did not show any differences in
the kinematics or kinetics of 2-skate skiing using coupled or uncoupled breathing
(results not shown).
Fig. 20.8 Rate of oxygen consumption for skiing just below the anaerobic threshold using the
2-skate technique while breathing normally using respiration to locomotion coupling (control) and
while breathing in a reverse manner to the coupled breathing (Reverse Breathing). Note that
normal breathing is associated with a significantly lower oxygen demand suggesting that coupled
breathing is more efficient than un-coupled or reverse breathing
20.4 Discussion
The primary aim of this investigation was to determine why cross-country skiers
switched from the 2-skate to the 1-skate, and back to the 2-skate technique with
increasing speeds of skiing. We hypothesized that these gait changes might be
driven by the cost of transport. Specifically, we hypothesized that the 2-skate
technique might be metabolically more efficient at slow and very fast speeds of
skiing than the 1-skate technique, and that the 1-skate technique might be more
efficient than the 2-skate technique at intermediate to fast speeds. Particularly, we
were interested in why the 2-skate technique, after being rejected in favor of the
1-skate technique at intermediate to fast speeds, would be taken up again when
skiers go very fast. We likened this behavior to a horse that changes its gait pattern
from a trot to a gallop with increasing running speeds, and then, when running
speed is further increased, changes back from the gallop to the trotting gait, a
behavior that would never occur. So, why does this completely unique change in
gait pattern occur in cross-country skate skiers?
The average cost of transport was found to be lower at slow and very fast speeds
for 2-skate skiing and lower for intermediate to fast speeds of skiing for the 1-skate
technique (Figs. 20.2 and 20.3). However, the differences in the cost of transport
were relatively small, but nevertheless, were significantly different for the two
techniques at 12 and 15 km/h and at 24 km/h and faster. For the lowest speeds
tested (6 and 9 km/h) statistical differences between the 1- and 2-skate techniques
were not reached.
So, why do cross-country skiers revert back to a gait pattern (2-skate skiing) at
very high speeds when this pattern was abandoned at intermediate speeds for the
1-skate technique? A first clue to this puzzle might be revealed when realizing that
propulsion is primarily coming from the legs in the 2-skate technique, and from the
arms for the 1-skate technique (Fig. 20.4). A second observation is that for the 1-
and 2-skate techniques, propulsion from the arms does not increase with increasing
speeds of skiing beyond a speed of about 15 km/h (Fig. 20.4). Finally, despite a
decrease in the impulse by the arms and poles from 15 to 30 km/h skiing, the
oxygen required for performing less work at the higher speed increases substan-
tially (Fig. 20.5), particularly for the 1-skate technique. These observations indicate
that the arm and pole action becomes costly from a metabolic point of view, while
not resulting in increased propulsion for the skier. Therefore, the increased propul-
sion must come from the leg and ski action when skiing at speeds of greater than
15 km/h, and the leg and ski action must be less costly than the arm action, thus
favouring the 2-skate technique at very high speeds of skiing.
But why should the arm action become more inefficient with increasing speeds
of skiing than the ski action, particularly when speeds of skiing become high
(>24 km/h)? And why does the same difference in metabolic cost not exist in
other four-legged locomotion, such as a running horse? The primary mechanical
difference between a running horse and a cross-country skier is that in the horse,
once the hooves contact the ground, the hooves are fixed at zero speed relative to
the horse’s body that is traveling at, let’s say, 10 m/s relative to the ground and the
stationary hooves. Therefore, the ground reaction time of all legs for a horse is
limited by the geometry of the legs at ground contact and at take-off and the speed
of the centre of mass above the legs that are fixed to the ground. Thus, contact times
in a horse running decrease predictably with the speed of running (Fig. 20.9).
Fig. 20.9 Average hoove contact times for a horse walking, trotting and galloping at increasing
speeds of locomotion. Note that the contact times decrease with increasing speeds of locomotion,
as contact times are given by the geometry of the leg at first and last ground contact and the speed
of the centre of mass of the horse. Because of the given leg length in a horse, contact times must
decrease with increasing speeds of locomotion
Fig. 20.10 Mean (1SD) pole and ski contact times for cross-country skate skiers skiing at
increasing speeds. Note how the pole contact times decrease in a similar fashion to those observed
in a horse running, while the ski contact times are virtually unaffected by skiing speed. The reason
for this difference in contact time behaviour of poles and skis is that the poles are fixed to the
ground during ground contact, like a horse’s hooves or the feet of a runner, while the skis glide at
approximately the same speed as the centre of mass of the skier. Therefore, the pole contact times
of a skier are limited by geometry and speed, like the hooves of a horse, while the contact times of
the skis are theoretically independent of the geometry of legs and skis and the speed of skiing.
Therefore, the ski contact times can be (and are) much longer than the pole contact times, and this
discrepancy increases with increasing speeds of skiing, thus favouring the muscles of the legs over
those of the arms for providing propulsion at high speeds of skiing
The same is true for the arms and poles of a cross-country skier. Once the poles are
planted in the snow, they are fixed relative to the movement of the skier’s centre of
mass, and the contact time is limited (Fig. 20.10). Predictably, the times of pole
contact with the ground decrease with increasing speeds of skiing.
In contrast, the legs of a cross-country skier are not fixed to the ground once
ground contact is made. The skis glide at essentially the same speed as the centre of
mass of the skier, therefore, in theory, ground contact times could be infinite for any
speed of skiing. Of course, practically this is not possible. However, ground contact
times decrease much less with increasing speeds of skiing for the legs (skis)
compared to the arms (poles) (Fig. 20.10). In fact, ground contact times for skis
remain on average 850 ms for skiing at 32 km/h (~9 m/s), while the corresponding
pole contact times for that same speed is only about 200 ms. A horse galloping at
9 m/s has an average stance time of about 110–120 ms, and a human sprinter’s
contact time has dwindled to a mere 100 ms running at that speed. Therefore, a
cross-country skier, even when skiing at extremely fast speeds, still has plenty of
time to use the muscles of the legs to produce force and propulsion with the skis,
while propulsion through the poles and arms becomes severely limited at high
speeds of skiing because of the muscle’s force-velocity relationships (Hill 1938).
The question then remains, why do skiers prefer the 2-skate over the 1-skate
technique at slow speeds of skiing, and why is there a trend towards better
efficiency for the 2-skate technique at slow speeds. This is a difficult question to
answer, but likely has to do with the balancing and the stability on the skis at slow
speeds of skiing. In the 2-skate technique, balancing is typically not an issue for
even moderately competent skiers at slow speeds, while it is very difficult, and
often impossible to 1-skate ski at slow speeds for beginners and moderately
experienced skiers, because the weight transfer from one ski to the other, and
associated balancing, is much more difficult for the 1-skate than the 2-skate
technique. Therefore, we tentatively suggest that the 2-skate technique is more
efficient than the 1-skate technique because of the wasted energy when balancing
on the skis for the 1-skate technique at slow speeds of skiing. Since we used only
experienced racers for our study, we did not see significant differences in the cost of
transport between the 1- and 2-skate techniques at very slow speeds of skiing.
However, we believe that if we had used beginner or moderately competent skiers
only, where balancing and stability of skiing plays a major component in technique,
the 2-skate technique might have been significantly more efficient than the 1-skate
technique for slow speeds of skiing.
A second aim of this study was to test if respiration coupling, as observed in
galloping animals (Ainsworth 1997; Attenburrow and Goss 1994; Bramble and
Carrier 1983; Bramble and Jenkins 1993) and cross-country skiers (Faria 2008),
does indeed safe metabolic energy, as has been proposed. In order to address this
question, we asked experienced skiers to ski just below their anaerobic threshold
speed using the 2-skate technique while inhaling and exhaling with arm recovery
and arm propulsion, respectively (coupled breathing), and then reversing this
breathing pattern relative to the arm movements, for the uncoupled breathing
patterns. We found that coupled breathing was associated with an approximate
decrease in oxygen consumption of about 4 % compared to the reversed breathing
pattern (Fig. 20.8), illustrating that coupled breathing indeed offers a metabolic
advantage. Of course, one might argue that by using a “reverse” breathing pattern
and compare it to the coupled breathing pattern, we used the worst metabolic
scenario for testing this hypothesis: that is, the skiers were asked to inhale when
their chest cavity was compressed due to the arm action and to exhale when the
chest cavity was expanded. Skiers not using a coupled breathing pattern likely
would not fall into a reversed breathing pattern, and thus the metabolic inefficiency
of “random” breathing compared to coupled breathing might not be as dramatic as
what we found here. Nevertheless, our results suggest that breathing coupled with
the upper body and arm movements of cross-country skiing infers a metabolic
advantage, and thus should be used by racers in competitive situations.
20.5 Conclusions
We conclude from the results of this study, that in contrast to all other four-legged
animal movements, cross-country skiers revert to a gait pattern at high speeds of
skiing that was rejected at a slower speed of skiing in favour of another gait pattern.
We suggest that this is primarily caused by the increasingly inefficient action of the
arms and poles with increasing speeds of skiing, and the reliance on arm propulsion
when using the 1-skate technique. Furthermore, we conclude that breathing coupled
with the arm action in 2-skate skiing infers a distinct metabolic advantage com-
pared to uncoupled breathing. Therefore, we suggest that cross-country skiers
carefully monitor their breathing patterns in competitive situations.
References
Ainsworth DM (1997) Pulmonary-locomotary interactions in exercising dogs and horses. Respir

Physiol 110(2)/3:287–294
Attenburrow DP, Goss VA (1994) The mechanical coupling of lung ventilation to locomotion in
the horse. Med Eng Phys 16:188–192
Bjørn Dæhlie (2014) http://en.wikipedia.org/wiki/Bj%C3%B8rn_D%C3%A6hlie
Bramble DM, Carrier DR (1983) Running and breathing in mammals. Science 219:251–256
Bramble DM, Jenkins FA Jr (1993) Mammalian locomotor-respiratory integration: implications
for diaphragmatic and pulmonary design. Science 262:235–240
Cross-country skiing (2014) http://en.wikipedia.org/wiki/Cross-country_skiing
Dresbeck LJ (1967) The ski: its history and histography. Technol Cult 8(4):467–479
Faria IE (2008) Ventilatory response pattern of Nordic skiers during simulated poling. J Sports Sci
12(3):255–259
Hill AV (1938) The heat of shortening and the dynamic constants of muscle. Proc R Soc Lond
126:136–195
Hoyt DF, Richard TC (1981) Gait and the energetics of locomotion in horses. Nature
292(5820):239–240
Solberg G, Robstad B, Skjonsberg OH, Borchsenius F (2005) Respiratory gas exchange indices for
estimating the anaerobic threshold. J Sports Sci Med 4:29–36
WEIR JB (1949) New methods for calculating metabolic rate with special reference to protein
metabolism. J Physiol 109:1–9
Chapter 21
Biomechanical Analysis of V2 Skating
in Cross-Country Skiing
Zenya Fujita and Jun Tsuchiya
Abstract In this chapter, I focus on the motions and forces associated with an
increase in gliding speed. The results show that in V2 skating, increases in velocity
are accompanied by a flight phase in which the skis lift off the ground. The flight
phase occurs after the gliding phase, in which the skis are touching the ground, and
before the push-off motion exerted by the legs occurs. This study also showed that
the increase in speed during V2 skating was not accompanied by the flight phase in
female athletes. However, considering the changes in forces during gliding, the
potential that the flight phase can be generated exists. Further, the results also
implied that in V2 skating, the increase in gliding speed is achieved by exerting a
powerful force that is associated with the dynamic motions accompanying the flight
phase.
At the same time, in female athletes as well, the increase in gliding speed was
observed to be accompanied by the exertion of a dynamic force. However, the latter
was not strong enough to generate a flight phase. Therefore, in female athletes, the
exertion of a dynamic force accompanied by a flight phase is expected to be a
difficult skill to attain, owing to limiting factors that are different from those found
in male athletes. In other words, performing gliding motions with a flight phase has
the potential to exert a more dynamic force, and to increase the gliding speed. We
have verified this hypothesis by conducting training experiments. The results
showed that instructing and training female athletes to perform V2 skating with a
flight phase was an effective means of increasing the gliding speed.
Keywords Pole reaction force • Ski reaction force • Flight phase • Performance •
Female athlete
Z. Fujita (*)
Department of Sports Sciences, Japan Institute of Sports Sciences, Tokyo, Japan
e-mail: zenya.fujita@jpnsport.go.jp
J. Tsuchiya

262 Z. Fujita and J. Tsuchiya
21.1 Introduction
In V2 skating, which is used in cross-country skiing competitions, the athletes angle

the skis outward in an arrangement that resembles the letter “V.” Then, they glide by
using the propulsion force obtained through a pushing motion using both poles, as
well as through a push-off maneuver using the skis on either side consecutively. In a
previous study by Smith (1992), in which the reaction force of the skis was measured
during one cycle of V2 skating, it was shown that a bimodal force was active when
the athlete was in contact with the ground. In addition, St€oggl et al. (2008) measured
the reaction force of skis during V2 skating and observed a flight phase (the phase
when both the right and left skis completely lost contact with the surface of the snow)
intermediate to the glide phase. The gliding speed and time needed to complete one
cycle (cycle time) were different in these two studies. In the study conducted by
Smith (1992), the gliding speed was not recorded but the cycle time was 2 s.
However, in the study conducted by St€oggl et al. (2008), the gliding speed was
7.15 0.59 m/s, and the cycle time was 1.56 s. The cycles were thus completed
within a shorter period of time in the study conducted by St€oggl et al. (2008) than in
that conducted by Smith (1992). A shortening of the cycle time is equivalent to an
increase in pitch. The increase in pitch during V2 skating corresponds to an increase
in gliding speed (Bilodeau et al. 1992), and hence, the measurements can be assumed
to have been performed at a lower speed in the study conducted by Smith (1992) than
in that conducted by St€oggl et al. (2008). In other words, the flight phase can be
assumed to occur as a result of an increase in gliding speed. However, the factors that
activate the flight phase have not been studied to date.
Millet et al. (1998), St€oggl and Müller (2009), and Nilsson et al. (2004) reported
that when the pitch increased during V2 skating, corresponding to an increase in
speed, the stride increased significantly when the speed increased from 3.6 to 5.2 m/
s. However, it decreased when the speed increased from 5.2 to 6.2 m/s (Millet
et al. 1998). They also reported that during high-speed skating, for speeds between
7.0 and 8.8 m/s, the stride did not change corresponding to the increase in speed
(St€
oggl and Müller 2009), and that it remained unchanged irrespective of the speed
range (Nilsson et al. 2004). In other words, it could be stated that an increase in
speed was achieved principally through an increase in pitch. This finding suggested
that, in order to achieve high speeds, applying a strong force instantaneously was
better than applying a force continuously for a long period on the skis or poles. The
findings of Millet et al. (1998) and St€oggl et al. (2010) confirmed this conclusion.
Millet et al. (1998) conducted a study of the relationship between gliding speed and
the reaction force of the pole. They observed that an increase in speed was
accompanied by a decrease in the peak and average levels of the pole reaction
force, as well as a decrease in the pole pushing time. St€oggl et al. (2010) used roller
skis to measure the vertical component of the skis’ reaction force. They demon-
strated that an increase in leg strength was critical to an increase in gliding speed.
Concurrently, an increase in speed was shown to be accompanied by a decrease in
the push-off time. From the discussion above, it can be concluded that applying
forces on the skis and poles instantaneously is essential for achieving high speeds.
21 Biomechanical Analysis of V2 Skating in Cross-Country Skiing 263
In biomechanical research on V2 skating, the primary gliding method in cross-

country skiing competitions, the key problem is the lack of clarity on the following
issues: the contributions of the upper and lower limbs, the changes in force and
motion associated with an increase in gliding speed, and the differences in skills
associated with gender. Solving these issues is considered important for providing
solutions aimed at improving gliding skills, establishing workout trainings, and
improving performance.
21.2 Relationship Between Occurrence of Flight Phase

and Increase in Velocity During V2 Skating
21.2.1 Purpose
The purpose of this study is to establish whether or not there is a relationship between
the occurrence of the flight phase and the increase in velocity during V2 skating.
21.2.2 Methods
Subjects
Seven university male (age, 19.9 1.7 (mean standard deviation (SD)) years;
height, 172.1 4.3 cm; weight, 67.0 7.1 kg) elite cross-country skiers
volunteered to participate in the present study.
Experimental Procedures
In this study, the subjects were asked to participate in two types of trials requiring
different skiing velocities. Two velocity settings were selected: a medium-speed
setting, which is the competitive pace for a 10 km race; and a high-speed setting,
which is the competitive pace for a sprint race. The subjects were instructed to skate
while clearly keeping in mind these two velocities. The two skiing velocity settings
for medium- and high-speed skating were approximately 5.5 m/s and >6.0 m/s,
respectively.
Measurement of Motion Phases
The poles used in this study were designed for competitive use (Yoko Platinum
Power Grip, Karhu Sporting Goods Oy, Finland). The load was measured by
attaching a strain gage (N11-FA-5-1000-11, Showa Measuring Instruments
Co. Ltd., Japan) to the bottom of each pole grip. Two poles of lengths 150 and
155 cm were fabricated to suit the height of each subject, and the subjects were
asked to select their optimal length. Competition skis (Volcan skis, Karhu Sporting
Goods Oy, Finland) were used, and the load on the skis was measured by inserting
strain gages inside holes at the front and rear of the boots, which were bound to the
sliding surface of the skis. The holes were filled with restorative material. The skis
were designed for independent detection of the ski’s front and rear contact with the
ground. This design enabled the detection of contact even in case only one side of
the ski (front or rear) was in contact with the ground while skiing. The strain gage
signals were stored using a data logger (NR-600, Keyence Corporation, Japan) at a
sampling rate of 1,000 Hz. The phase times were calculated by measuring the ON
and OFF points from the obtained voltage waveforms. The conventional approach
with strain gages has been to convert voltage to force on the basis of calibration.
However, since there were situations where the skis deformed during skiing owing
to the unevenness of the snow surface, the converted values did not necessarily
reflect the force exerted by the body. Therefore, the waveform was used only for the
purpose of obtaining the timing of contact with and separation from the ground.
Photography of Motion
Videos were shot at 60 fields/s and a shutter speed of 1 ms, using two digital video
cameras (HVR-A1J, Sony Corporation, Japan), located on the side and front of a
custom-built course on the snow. The view angle was set to 15 4 2 m
(L W H), and calibration poles were placed at a total of 18 points (0.4 m
between marks). For this study, a stationary right-handed coordinate system was
established by setting the forward, left-to-right, and vertical directions as the y-, x-,
and z-axis, respectively. Prior to the experiment, the surface of the snow was
compacted using a snow vehicle.
Analysis
The video images captured using the two cameras were stored on a personal
computer, and 23 points on the subject’s body were digitized at a sampling rate
of 60 Hz using movement analysis software (Frame-DIAS II V3, DKH, Japan). The
digitized coordinate values were converted to actual lengths using a three-
dimensional (3-D) direct linear transformation method, and the 3-D coordinates
of each body part were determined. The optimal cut-off frequency (4.8–7.2 Hz) for
the calculated 3-D coordinates was determined using a residual analysis method,
and smoothing was performed using a fourth-order Butterworth low-pass digital
filter. To synchronize the data logger and the two cameras, a light emitting diode
was flashed in view of the two cameras by an external trigger at the start of data
collection. Voltage data obtained from the strain gage were normalized to 60 Hz,
and the movement data and sampling frequency were synchronized.
In this study, it was decided to adopt a left-right cycle within the video photog-
raphy zone for the skating data analysis. The skiing velocity of one cycle and the
center of mass in the vertical direction were derived using the calculated displace-
ment of the body’s center of mass. The body’s center of mass in the vertical
direction was normalized using the height of the subject.
The poling motion was categorized into two phases. The interval between pole
contact with the ground and the subsequent separation from it was defined as the
push phase and that between separation and subsequent contact as the recovery
phase. The definitions of the ski phases were based on the method reported by
Bilodeau et al. (1992). The interval between ski contact with the ground and the
start of the knee joint extension was defined as the glide phase, and that between the
start of the knee joint extension and separation from the ground was defined as the
kick phase. If the ski was observed to separate from the ground temporarily during
the glide phase, it was decided to define that period of separation as the flight phase.
In addition, it was decided to define the interval between ski separation and contact
with the ground as the recovery phase. The flight time was calculated for the right
leg during flight phase.
Statistics
Each set of data was expressed as its mean SD. To enable comparison between
trials, the start of a cycle was set to 0 % and the end to 100 %, and 101 sets of data
were normalized using cubic spline interpolation. Significant differences in the
skiing velocity between trials were determined using Student’s t-test.
21.2.3 Results
For the skiing velocity of one cycle, the high-speed trials (6.22 0.51 m/s)
exhibited values significantly higher than the medium-speed trials (5.55 0.39 m/
s) (t ¼ 4.580, p ¼ 0.004).
A phase wherein both the left and right skis completely separated from the
ground was observed only in the high-speed trials (Fig. 21.1). In the medium-speed
trial, the flight phase could not be observed in the case of all subjects; however, in
the high-speed trial, the flight phase was observed between 37.7–49.9 % for the
right ski and 0–2.5 % and 91.2–100 % for the left ski during one cycle (Fig. 21.2). In
the case of the right leg, the center of mass climbed vertically up to the middle of the
flight phase (18–43 %), reached its peak, and then descended. In the case of the left
leg, it climbed up to the start of the flight phase (68–93 %), reached its peak, and
then descended during the flight phase (Fig. 21.3). Neither significant interaction
nor primary effects were observed (F ¼ 1.206, p ¼ 0.262).
a
7
3
Voltage of ski censers (v)
0
b
7
6
Right ski
5
Left ski
4
0
0 10 20 30 40 50 60 70 80 90 100
Normalized time (0-100%)
Fig. 21.1 A typical example of ski-sensor voltage during (a) a medium-speed trial, and (b) a high
speed trial
21.2.4 Discussion
In this study, a flight phase was observed in the high-speed trials but not in the
medium-speed ones. This result implies that the occurrence of the flight phase was
related to V2 skating velocity.
The voltage in the middle of the glide phase during high-speed trials was 0 V
(Fig. 21.1), and thus, it was confirmed that the ski came completely off the ground.
Additionally, since the other ski and both poles were not in contact with the ground,
it is likely that the body was in a state of complete flight during this phase
(Fig. 21.2). Furthermore, a comparison of the skating motions at different velocities
showed that changes in the vertical displacement of the center of mass did not affect
the flight phase (Fig. 21.3). This fact suggest that, although the skis definitely
floated above the snow during the flight phase, the change in terms of differences
in kinematics was negligible.
a
8.1%( 0.8) 7.9%( 1.4)
Both pole push Push Push
46.5%( 5.9) 10.1%( 4.0)
Right ski contact Glide Kick
16.6%( 3.2) 8.4%( 5.9) 33.4%( 2.6)
Left ski contact Glide Kick Glide
b 60%
9.4%( 2.2) 8.8%( 2.2)
Both pole push Push Push
22.7%( 3.9) 12.1%( 5.3) 13.1%( 5.4) 11.2%( 4.6)
Right ski contact 2.5%( 2.7) Glide Flight Glide Kick
14.2%( 3.8) 9.3%( 4.1) 27.3%( 3.9) 8.8%( 5.0)
Left ski contact Glide Kick Glide Flight
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Normalized time (0-100%)
Fig. 21.2 Phase diagram of V2 skating for (a) a medium-speed trial and (b) a high-speed trial
Glide Kick Glide Medium

Left ski contact
Glide Kick Glide Flight High
Glide Kick Medium
Right ski contact High
Glide Flight Glide Kick
65
Vertical displacement (%Height)
60
55
50
Medium
45
High
40
0 10 20 30 40 50 60 70 80 90 100
Normarized time (0-100%)
Fig. 21.3 Mean values of the vertical displacement of the center of mass
The flight phase occurred over an extremely short period of time. Therefore, the
difficulty in specification, given the measurement precision in this research, is also
considered a factor in making it virtually impossible to clarify the mechanism of the
flight phase. The mean flight time was 0.18 s (Table 21.1). Assuming that the flight
Table 21.1 Data and results relating to cycle characteristics, upper body work, and leg work
Low Medium High
Cycle characteristics
Velocity (m/s) 3.50 0.21 5.09 0.24 6.31 0.55
Cycle rate (Hz) 0.85 0.07 1.05 0.11** 1.29 0.12***, a
Cycle length (m) 4.12 0.43 4.88 0.54* 4.92 0.50**
Upper body work
Poling time (s) 0.34 0.04 0.26 0.04** 0.22 0.02***, a
Swing time (s) 0.84 0.08 0.70 0.07** 0.56 0.07***, a
Poling peak force (N) 53.62 15.31 87.73 26.96* 138.21 72.64*
Poling mean force (N) 36.3 10.57 56.78 20.16* 79.8 44.32*
Leg work
Glide time (s) 0.86 0.09 0.64 0.10** 0.50 0.06***, a
Glide peak force (N) 680 40 695 44 743 41*
Glide mean force (N) 531 34 503 32 466 41*
Push-off time (s) 0.50 0.09 0.46 0.06 0.42 0.04
Push-off peak force (N) 699 36 863 53*** 991 77***, a
Push-off mean force (N) 526 23 573 23* 583 34***
Lowest force (N) 480 41 393 59** 251 96**, a
Poling peak force per push-off 7.7 2.1 10.1 3.0 13.7 6.1
peak force (%)
*Significant (P < 0.05), **high significant (P < 0.01) and ***highly significant (P < 0.001) dif-
ferent to low speed
a
Significant (P < 0.05) different to medium speed
phase is a parabolic motion of the center of mass, the jump height in the vertical
direction is only 3.7 cm, if it is estimated from the flight time. In other words, the
change in motion accompanying the occurrence of the flight phase is negligible.
This implies that it does not occur owing to a large jump in the forward direction, or
by lifting skis high off the snow surface. To analyze this small change will require
more refined kinematic analysis techniques, as well as mechanical analysis through
measurement of the ground reaction force.
21.3 Cycle Characteristics and Reaction Force Profiles

in V2 Skating Techniques of Female Cross-Country
Skiers
21.3.1 Purpose
The purpose of this study is to determine the cycle characteristics and the pole and
ski reaction forces of female cross-country skiers during V2 skating at low to high
velocities.
21.3.2 Methods
Subjects
Nine female elite cross-country skiers (mean SD: age, 20.2 1.0 years; height,
158.7 5.0 cm; weight, 56.5 4.3 kg) gave their informed consent for participa-
tion in this experiment. Their mean maximal oxygen consumption VO2max
(56.8 8.4 ml/kg/min) was measured during Nordic walking on a treadmill using
an ergospirometry system.
Pole and Ski Sensors
All subjects used racing poles (Yoko Platinum Power Grip, Karhu Sporting Goods
Oy, Finland). Specially made sensors were mounted on the right and left poles
(Fig. 21.4). In the test, special roller skis were fabricated using square pipe
aluminum (600 40 30 mm), bearing, and tire (610 mm, Marwe Oy, Hyvinkään
Kumi Oy). To measure the ski reaction force on the roller ski, strain gages
(N11-FA-5-1000-11, Showa Measuring Instruments Co. Ltd., Japan) were attached
to the front and rear of the right and left roller ski frames (Fig. 21.4).
Data Collection
Testing was conducted on a flat, straight outdoor asphalt track. After accelerating
over a 50 m section, the participant’s skating time over the next 20 m was measured
using two photo sensors (Wireless Sprint System, Brower Timing Systems, USA).
Strain
gauge
Strain gauge
Fig. 21.4 Settings for pole and roller ski sensors

Poling phase Swing phase

150
Pole
Pole and ski reaction force (N)
100
50
0
Glide phase Push-off phase Swing phase Roller ski
1000
800
600
400
200
0
0 500 1000 1500 2000
Time (s)
Fig. 21.5 Definitions of poling and ski phases during V2-skating
Each subject was instructed to ski at three different speeds (low: 3.5 m/s, medium:
5.0 m/s, and maximal effort).
Data Analysis
In this study, one cycle was defined as running from the start of the ground contact
of both poles to the start of their subsequent ground contact. The upper body work
was divided into the poling and swing phases used by the pole reaction forces. The
leg work was categorized into the glide, push-off, and swing phases used by the ski
reaction forces (Fig. 21.5). The zero level of the pole and ski reaction forces was
defined as the swing phase. The poling phase was the phase during which the pole
reaction force was greater than zero. The glide and push-off phases were deter-
mined based on the point of lowest value among the bimodal ski reaction forces.
The glide phase started with the placing of the roller ski on the ground, leading to a
first peak in the force curve. It ended at the point of minimum value of the ski force,
before it attained a second peak. The push-off phase started at the point of minimum
force and ended when the ski reaction force decreased to zero. The duration of the
poling, swing, glide, and push-off phases were calculated, and the bimodal peak
values during the glide and push-off phases were defined as the peak force during
each phase. The ratio of the peak force during the poling phase to the peak force
during the push-off phase was calculated. The lowest force was determined as the
minimum value between the glide and push-off peak forces.
Statistics
The following data were obtained from each trial: velocity, cycle rate, cycle length,
and peak and mean pole reaction forces during the poling phase; the lowest ski
reaction force during the glide phase; the peak and mean ski reaction forces during
the glide and push-off phases; and the poling, swing, glide, and push-off times. The
variables were reported as the mean SD. They were compared across velocities
using a one-way analysis of variance with repeated measures. When the F-values
were significant, individual comparisons were made using Bonferroni post hoc
analysis. Moreover, the correlation between the peak force during the push-off
phase and the lowest force were analyzed using Pearson’s product–moment corre-
lation analysis. Values of 0.05, 0.01, and 0.001 were accepted as the levels of
statistical significance.
21.3.3 Results
Cycle Characteristics
The cycle characteristics and kinetic parameters are listed in Table 21.1. The cycle
rate gradually increased, from low to high velocities, up to 1.29 Hz (all P < 0.05),
whereas the cycle length increased only from low to medium velocities (all
P < 0.05).
Arm Actions
The poling times decreased from low to medium velocities (P < 0.05), whereas the
swing times gradually decreased across velocities, from low to high (all P < 0.05).
The ratios of the poling and swing times, in each cycle, remained constant across
velocities. The poling peak and mean forces increased from low to medium
velocities (all P < 0.05).
Leg Actions
The typical reaction force curves of poles and skis during the low-, medium-, and
high-speed trials are shown in Fig. 21.6. The profiles of the ski reaction forces
varied with respect to velocity. The glide times gradually decreased across veloc-
ities, from low to high, whereas the push-off times remained unchanged. Ratios of
a
150
100 Right Pole
50 Left Pole
0
1000
800
600 Right Ski
400
200 Left Ski
0
b
150
100 Right Pole
Left Pole
Force (N)
50
0
1000
800
600 Right Ski
400 Left Ski
200
0
c
150
100 Right Pole
50 Left Pole
0
1000
800
600 Right Ski
400 Left Ski
200
0
0 1 2 3 4 5 6 7
Time (s)
Fig. 21.6 Typical examples of pole and ski reaction force profiles during (a) low-speed,
(b) medium-speed, and (c) high-speed trials
the glide and push-off times, in each cycle, decreased from low to high velocities
(all P < 0.05). The push-off peak force gradually increased from low to high
velocities (all P < 0.05), whereas the mean push-off force increased from low to
medium velocities (P < 0.05). The lowest forces gradually decreased from low to
high velocities (all P < 0.05). The ratio of the poling to the push-off peak forces
tended to increase from low to high velocities (all P < 0.1), whereas the ratio of the
poling to the push-off mean forces increased from low to medium velocities
(P < 0.05). A significant correlation was confirmed (P < 0.001, Fig. 21.6) between
the ski reaction force per unit of body weight during the push-off phase and the
lowest force applied to the skis per unit of body weight.
21.3.4 Discussion
The profile of the ski reaction force during V2 skating was bimodal. The bimodal
characteristics were attributed to the rise of the body in the glide phase and its fall,
immediately before the push-off phase. The depth of the valley in the bimodal
profile increased as the velocity increased. This phenomenon was similar to that
observed in the previous studies conducted on male skiers (Smith 1992; St€oggl
et al. 2008, 2010). However, a flight phase was not observed even at the maximum
velocity. St€oggl et al. (2008) observed that the lowest force becomes zero (the
“flight phase”, in which the ski separates from the snow surface) during the high-
speed (7.35 0.60 m/s) V2 skating of male skiers. Fujita et al. (2010) also reported
that a flight phase occurred with the high-speed V2 skating of male skiers.
Therefore, it was implied that the maximum velocity of female skiers
(6.31 0.55 m/s) in this study did not reach the velocity required for the occur-
rence of the flight phase. However, there was a negative correlation between the
peak and lowest forces (Fig. 21.6). These results implied that the decrease in the
lowest force was related to the occurrence of the flight phase and the increase in
velocity.
Perrey et al. (1998) examined electromyograms of lower limb muscles during
V2 skating of male skiers and showed that stretch-shortening cycles occurred at a
velocity of 5.3 m/s. The velocities in this study were comparable to those of Perrey
et al. Therefore, it was implied that the stretch-shortening cycle in this study
occurred during the push-off phase as well. Additionally, it was implied that the
lowest force was related to the stretch-shortening cycle. This would make an
interesting subject for future study.
With an increase in velocity, the poling forces increased, and the poling time
decreased. Lindinger et al. (2009) measured the poling forces generated by the
double-poling technique, with velocities in the range of 2.5–8.2 m/s, used by male
skiers. They showed that, in the range from 2.5 to 7.5 m/s, poling time decreased
and poling force increased with increases in velocity, and there were no changes in
the poling force in the range of 7.5–8.2 m/s. These results are consistent with those
of this study, conducted for the V2 skating technique. This similarity of results
implies that the increase in velocity was achieved using the same upper limb
mechanisms in both techniques.
21.4 Case Study of Technique for Increasing the Skiing

Velocity of Female Cross-Country Skiers
21.4.1 Purpose
The purpose of this study was to clarify the impact of the flight phase on compe-
tition performance during V2 skating. To do so, female athletes, whose motions
during V2 skating did not include the flight phase, were instructed to generate the
flight phase intentionally.
21.4.2 Methods
Subjects
The participants consisted of seven female cross-country ski athletes who had
previously won prizes in major competitions in Japan (age: 20 1 years; stature:
160.1 5.4 cm; body mass: 54.2 5.3 kg). The informed consent of the partici-
pants was obtained after the purpose of the study was explained in advance to them.
Data Collection
The poles and roller ski sensors used in this study were similar to those used in the
previous section. The experiments were conducted on a flat asphalt road within
Waseda University’s Tokorozawa campus. The time required for traversing the
course was measured by installing two photoelectric cells (Wireless Sprint System,
Brower Timing Systems, USA) separated by a 20-m measurement interval. For the
experiment, a 50-m acceleration interval was prepared ahead of the measurement
interval. First, each participant was instructed to perform skating trials with maxi-
mum effort (referred to below as the normal trial) 3–5 times. The force data revealed
the occurrence of the flight phase in the motions of two of the participants. Therefore,
they were excluded, and the study was conducted with five participants. In order to
generate the flight phase, the selected participants were instructed to jump immedi-
ately after the roller skis touched the ground, after which they were told to glide freely
for approximately 5–10 min. At that juncture, the participants were instructed to start
their pole-pushing motions as soon as they landed from the jump; they were
instructed to avoid slowing down during the period when they were excessively
focused on jumping. Later, they were instructed to perform a gliding trial with
maximum effort during the flight phase (referred to below as the flight trial) 3–5
times. The trials were performed with 5 min or more of rest in between to eliminate
the effects of fatigue on the motions. After the experiments, each participant submit-
ted their introspections regarding their experience during the flight trials.
Data Analysis
The series of motions using poles were categorized as the push and swing phases,
depending on the poles’ reaction forces. In addition, the series of motions using
roller skis were categorized as glide, flight, and push-off phases, depending on the
roller skis’ reaction forces. The push phase was designated as the time period that
began when the pole touched the ground and ended when it left the ground. The
beginning of the glide phase was designated as the time when the skis touched the
ground. The glide phase’s end was defined as the time when the reaction force curve
generated by the roller skis reached its minimum level, after the occurrence of the
first peak in a normal trial, and before the next peak. In the flight trial, the end was
specified as the time when the reaction force became zero as a result of the flight
phase. In the normal trial, the beginning of the push-off phase was denoted as the
time when the roller skis’ reaction force attained their minimal values. In a flight
trial, the end of the flight phase was considered as the trial’s beginning. The trial’s
end was considered the time when the reaction force became zero after the peak of
the second force curve.
Statistics
The following were calculated from each trial: the gliding speed during the two
trials; peak and mean values of the pole reaction force during the pitch, stride and
push phases; peak and mean values of the roller skis’ reaction force during the glide
and push-off phase; pole-pushing and gliding times; flight time (flight trial only),
and push-off time. The values were shown as mean SD, and the comparison
between the trials was conducted by using Student’s t-test. For the hazard ratio,
values less than 0.05 were considered significant, and values less than 0.1 were
considered as having a tendency toward significance.
21.4.3 Results
Table 21.2 lists the following: the gliding speeds in the two trials; stride, pitch, and
push time; glide and push-off times; peak and average values of the pole reaction
force during the push phase; roller ski reaction force during the push-off phase; and
flight time during the flight trial. The gliding speed during the flight trial was
significantly higher than that observed in the normal trial. In addition, compared
Table 21.2 Results of cycle characteristics and poling and ski reaction forces
Normal Flight P-value
Velocity (m/s) 5.95 0.22 6.3 0.18 0.001
Stride (m) 4.46 0.16 4.23 0.15 0.073
Pitch (Hz) 1.34 0.09 1.49 0.06 0.009
Poling time (s) 0.22 0.02 0.21 0.02 0.188
Glide time (s) 0.45 0.07 0.3 0.01 0.006
Flight time (s) – 0.14 0.01
Push off time (s) 0.42 0.05 0.35 0.04 0.041
Poling peak force (s) 111 8 141 12 0.002
Poling mean force (s) 75 10 87 9 0.022
Push off peak force (s) 920 164 979 159 <0.001
Push off mean force (s) 506 69 600 62 <0.001
N ¼ 5. All value shows mean SD
to the results of the normal trial, those of the flight trial showed a significantly
increased pitch, significantly reduced glide and push-off times, and a stride that
tended to decrease. In addition, compared to the findings of the normal trial, those
of the flight trial showed that the peak and mean values of the pole reaction force
during the push phase, as well as the peak and mean values of the ski reaction force
during the push-off phase, increased significantly.
In summary of the introspective thoughts provided by each participant after
performing the flight trial: three participants stated that they had a sensation of fear
when they were lifting off the ground, and four participants reported that they felt as
though they were going to lose balance.
21.4.4 Discussion
In this study, participants whose motions during V2 skating at maximum effort did
not include a flight phase, were trained for approximately 10 min to master the
motions aimed at generating the flight phase, and were instructed to glide again. As
a result, the flight phase occurred during skiing and the gliding speed increased
significantly. In the flight trials, the findings showed that the occurrence of the flight
phase led to an average increase of 0.14 s in flight time per cycle. However, the
glide and push-off times decreased by 0.15 and 0.07 s, respectively, resulting in an
increase in pitch. Concurrently, although the stride tended to decrease, the increase
in pitch was observed to compensate for the decrease in stride, and this resulted in
an increase in gliding speed. Millet et al. (1998) and St€oggl et al. (2010) showed that
in V2 skating, an increase in gliding speed was accompanied by a tendency of the
stride to remain unchanged or decrease, as well as an increase in pitch; and their
results were consistent with the findings of our study. In other words, the motions
performed during gliding, which generated the flight phase, could be used as an
appropriate method for increasing the sliding speed.
The ski reaction forces during V2 skating with a flight phase showed signifi-
cantly higher values than those observed during motions that did not generate the
flight phase. St€ oggl and Müller (2009) showed that, in V2 skating, an increase in
gliding speed was accompanied by an increase in the vertical component of the
roller skis’ reaction force. In other words, the increase in the ski reaction force
observed in this study can be considered to be an appropriate result of the effort to
increase the gliding speed. In addition, Sect. 3.3 of Chap. 3 showed that, in V2
skating, an increase in gliding speed was accompanied by an occurrence of the
flight phase. The results of this study imply that the motions that generated the flight
phase during gliding result in a decrease in push-off time. In addition, the exertion
of a strong force prompts an increase in gliding speed. Therefore, it can be stated
that, in the flight trial, the generation of the flight phase results in changes that make
the motions suitable for the exertion of a strong force; in other words, a strong force
exerted in a short period of time.
1200
Right pole
1000
Left pole
800
Force (N)
Right roller-ski
600
Left roller-ski
400
200
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Time (s)
1200
1000
800
Force (N)
600
400
200
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Time (s)
Fig. 21.7 Typical examples of pole and ski reaction force profiles during (a) a normal trial and (b)
a flight trial
The pole reaction force, during V2 skating with a flight phase, was significantly
stronger than that observed during V2 skating using motions that did not produce a
flight phase. The characteristics of the forces observed during the flight trial
(Fig. 21.7) show that the push phase started at the end of the flight phase. In other
words, it can be stated that, in a normal trial, the body’s mass is applied unilaterally
to a ski on one side when the push phase begins. On the other hand, in a flight trial,
the body mass that was applied to the roller skis during the normal trial was applied
to the poles. The force that was applied to the skis in normal trials was not oriented
in the traveling direction. However, by reason of being applied to the poles, it might
have turned toward the traveling direction, and this might have worked in favor of
the acceleration. Millet et al. (1998) showed that during V2 skating, an increase in
gliding speed was accompanied by an increase in the pole reaction force. In other
words, motions performed during gliding associated with a flight phase can be
considered to be effective in increasing the pole reaction force.
In this study, the participants were instructed to perform skiing with a flight
phase, and later were subjected to training for approximately 10 min. As a result,
each of their gliding motions was observed to be accompanied by a flight phase.
However, the flight phase was observed before the participants were instructed to
conduct a flight trial. This implied the existence of limiting factors that inhibited the
generation of the flight phase during skiing. In the findings from the introspections
with respect to the flight trial, three participants responded that they felt a sensation
of fear in regard to being lifted off the ground. In other words, the sensation of fear,
despite having the physical strength needed for generating a flight phase that lifts
the entire body off the ground, might be one of the technical limiting factors.
Acquisition of skills allowing for performing the flight phase in a stable manner
could be important for overcoming this limiting factor.
This study showed that for female athletes, whose skiing motions did not include
the flight phase, instruction in performing ski-gliding motions associated with a
flight phase was an excellent method for improving their competition performance
in a short period of time. Further, after all the participants were instructed to
perform ski gliding with a flight phase, four of them responded that they felt as
though they were going to lose balance. Since the measurement interval in this
study was a very short distance of 20 m, there was no significant loss of balance or
fall. However, in actual competitions, skiers need to glide over a distance of
approximately 1 km, even in short distance sprint competitions. When doing so,
performing V2 skiing, which involved the exertion of considerable force over a
long period of time, and accompanied by a flight phase, places a considerable strain
on the body. Even in physically strong and highly skilled first-class male V2 skating
athletes, gliding accompanied by a flight phase is not performed in all sections of
sprint competitions. It is generally used only during maneuvers and the final sprint.
This is because the athletes first envision their race strategy and select the best
gliding technique to achieve efficient, high-speed skiing. Next, they perform V2
skating with a flight phase, while giving a priority to the achievement of the desired
gliding speed. Using this gliding technique might also be preferable for female
athletes in circumstances where there is need to achieve a high gliding speed.
21.5 Conclusions
This study was focused on the motions and forces associated with an increase in
gliding speed. The results show that in V2 skating, the increase in running speed is
accompanied by a flight phase, in which the skis lift off the ground The flight phase
occurs after the gliding phase when the skis are touching the ground and before the
push-off motion exerted by the legs occurs. In addition, this study has shown that
the increase in speed during V2 skating was not accompanied by the flight phase in
female athletes. However, considering the changes in forces during gliding, the
potential that the flight phase can be generated exists. Further, the results also
implied that in V2 skating, the increase in gliding speed is achieved by exerting a
powerful force that is associated with the dynamic motions accompanying the flight
phase.
At the same time, in female athletes as well, the increase in gliding speed was
observed to be accompanied by an exertion of a dynamic force. However, the latter
was not strong enough to generate a flight phase. Therefore, in female athletes, the
exertion of a dynamic force accompanied by a flight phase is expected to be a
difficult skill, owing to limiting factors that are different from those found in male
athletes. In other words, performing gliding motions with a flight phase has the
potential to exert a more dynamic force, and to increase the gliding speed. We have
verified this hypothesis by conducting training experiments. The results showed that
instructing and training female athletes to perform V2 skating with a flight phase was
an effective means of increasing the gliding speed. In addition, the results implied
that in female athletes, the sensation of fear with regard to floating in the air, as well
as the sensation of loss of balance, inhibited the generation of a flight phase. This
acted as a limiting factor that restricted the increase in gliding speed. In other words,
for female athletes, going through a training program specifically aimed at generat-
ing the flight phase, and getting used to performing gliding motions accompanied by
a flight phase, are crucial for increasing their gliding speed.
References
Bilodeau B, Boulay MB, Roy B (1992) Propulsive and gliding phases in four cross-country skiing
techniques. Med Sci Sports Exerc 24:917–925
Fujita Z, Ishige Y, Yoshioka S, Tauchi K, Iso S, Fukashiro S, Tsuchiya J (2010) A relationship
between the occurrence of the flight phase and the increase in velocity during V2 skating. Int J
Sport Health Sci 8:113–120
Lindinger SJ, St€oggl T, Müller E, Holmberg H-C (2009) Control of speed during the double poling
technique performed by elite cross-country skiers. Med Sci Sports Exerc 41:210–220
Millet GY, Hoffman MD, Candau RB, Clifford PS (1998) Poling forces during roller skiing:
effects of technique and speed. Med Sci Sports Exerc 30:1645–1653
Nilsson J, Tveit P, Eikrehagen O (2004) Effects of speed on temporal patterns in classical style and
free style cross-country skiing. Sport Biomech 3:85–108
Perrey S, Millet GY, Candau R, Rouillon JD (1998) Stretch-shortening cycle in roller ski skating:
effect of technique. Int J Sports Med 19:513–520
Smith GA (1992) Biomechanical analysis of cross-country skiing techniques. Med Sci Sports
Exerc 24:1015–1022
St€
oggl TL, Müller E (2009) Kinematic determinants and physiological response of cross-country
skiing at maximal speed. Med Sci Sports Exerc 41:1476–1487
St€oggl T, Müller E, Lindinger S (2008) Biomechanical comparison of the double-push technique
and the conventional skate skiing technique in cross-country sprint skiing. J Sports Sci
26:1225–1233
St€oggl T, Kampel W, Müller E, Lindinger S (2010) Double-push skating versus V2 and V1 skating
on uphill terrain in cross-country skiing. Med Sci Sports Exerc 41:187–196
Chapter 22
Limb Force Generation as a Limiting Factor
for Maximum-Effort Acceleration
Performance
Geng Luo and Darren J. Stefanyshyn
Abstract Sprint acceleration is important in various sports. The current study

examined whether acceleration performance is limited by strength, more specifi-
cally the peak limb force generation. We tested this idea by altering the mechanical
demand for the supporting limb, and then determining if the peak limb force is
constantly operating at its maximum level across conditions. Thirteen athletes
performed maximum-effort sprint acceleration with and without an additional
mass equivalent to 17 % of their body mass. We found the athletes were capable
of generating a greater limb force when sprinting with the mass compared to the
control. This observation suggests that acceleration performance might not be
limited by the ultimate magnitude of limb force generation. It remains puzzling
why the athletes would not fully utilize their limb force for propulsion. Future
studies are needed to further explore other potential limiting factors for acceleration
performance, such as kinematic constraints for body lean.
Keywords Acceleration performance • Limiting factor • Strength • Sprint
22.1 Introduction
Performance of sprint acceleration is crucial in determining the outcome of various

sports. For example, time-motion analysis of elite soccer players revealed that the
number of high-intensity sprint bursts performed in each game is a key differential
character between the top-class and moderate players (Mohr et al. 2003). Despite
their importance, factors limiting maximum acceleration performance remain rel-
atively unexplored (Alexander 2002).
Acceleration performance is directly related to the force an athlete can apply to
the ground. During each ground contact, the orientation and magnitude of the
G. Luo • D.J. Stefanyshyn (*)

Human Performance Laboratory, University of Calgary, Calgary, Canada
e-mail: darren.stefanyshyn@ucalgary.ca

282 G. Luo and D.J. Stefanyshyn
ground reaction force, along with other external forces (e.g. gravity), determine the
instantaneous acceleration of that athlete’s body centre of mass. Over a given time
interval, acceleration performance can be assessed as the net gain of body momen-
tum in the desired direction. During level locomotion, a greater momentum gain
can be achieved by applying a larger horizontal propulsive force and/or an elon-
gated contact time. Between the two strategies, it was shown that athletes exerted a
much larger propulsive force within an even shortened contact when accelerating at
maximum versus submaximal effort (Kugler and Janshen 2010). Thus, a closer look
at factors influencing propulsive force may offer insights to the limiting factors for
acceleration performance.
One key biomechanical constraint for creating propulsive force could be the
maximum limb force generation. To create a greater propulsive force, the limb
force vector needs to be directed closer to the horizontal plane. While such vector
redirection occurs, the vertical force needs to remain sufficiently high for
supporting body weight. It is possible that during maximum-effort acceleration,
the limb force generation reaches its limit so further force redirection could no
longer be achieved. Earlier observations of the correlation between limb force
production and launching velocity have led to the proposal that muscle strength
strongly affects sprint start performance (Mero 1988). Experiments designed to
directly examine limb force as a limiting factor for acceleration performance are
scarce in the literature.
The purpose of the current study was to investigate whether athletes perform
full-effort acceleration with maximized limb force. We examined this idea by
introducing an experimental intervention of an additional mass to alter the mechan-
ical demand for the supporting limb. The logic was as follows. If maximum
acceleration performance is limited by limb force generation, as the demand for
weight support increases, we should observe an unchanged peak resultant limb
force, but changes in the force and body orientation for creating sufficient vertical
impulse. Specifically, it was hypothesized that, compared to control, accelerating
with an additional mass will lead to: (1) a constant peak resultant limb force, (2) an
increase in stance-average vertical force, (3) a decreased stance-average propulsive
force, (4) and a reduced body lean.
22.2 Methods
22.2.1 Participants
Thirteen male athletes participated in the current study (age: 27 6.0 years; body
mass: 74.2 5.8 kg; height: 174.5 4.8 cm). All subjects regularly participated in
recreational sports and had no injuries during the past year prior to the experiment.
Before each testing, written consent approved by the university ethics committee
was provided by the athletes.
22 Limb Force Generation as a Limiting Factor for Maximum-Effort Acceleration. . . 283
22.2.2 Set-up and Equipment
The athletes performed maximum-effort linear accelerations from standing starts

while kinematic and kinetic data were collected simultaneously. Spherical reflec-
tive markers (19 mm diameter) were placed at the sacrum, left and right anterior
superior iliac spine to represent the pelvis segment for estimating the location of
body centre of mass. Eight high-speed cameras (Eagle, Motion Analysis Corpora-
tion) were used to capture the 3D marker positions at 240 Hz within the collection
volume (width 1.5 m, length 3.5 m, height 2.3 m). One force platform (Z4852C,
Kistler Group) was used to record the ground reaction force (2,400 Hz) of the first
foot contact immediately after the start. The preferred starting foot and location was
chosen by the individuals so they could strike the force platform naturally. The
starting location was kept consistent across conditions for each athlete. To ensure
full-effort execution through the capture area, the athletes were instructed to keep
accelerating maximally for an additional 3 m beyond the collection volume before
slowing down.
A commercially available lifejacket (Fulcrum PFD, Mountain Equipment
Co-op) was modified to alter the athletes’ body mass while minimizing interference
to the movement execution (Fig. 22.1). Six pieces of diving lead were adhered to
the bottom of the lifejacket so that the additional mass could be located close to the
body centre of mass. The modified jacket contained a total mass of 12.4 kg,
equivalent to 17 % of the average body mass of the sampled population. This
magnitude was chosen through a pilot study to ensure sufficient increase in the
mechanical demand for the supporting limb without drastically interfering with the
movement execution.
Since available traction at the foot-ground interface can influence the execution
of linear acceleration (Alexander 2002), the same pair of testing footwear with
sufficient traction was used across subjects and conditions. Available traction
between the testing footwear and laboratory floor was quantified mechanically
with a previously validated protocol and its value (traction coefficient of 1.13)
exceeded the required amount (traction coefficient of 0.82) for maximum acceler-
ation performance (Luo and Stefanyshyn 2011). The footwear was cleaned between
trials, and the floor surface was cleaned between conditions.
22.2.3 Protocol
After a 20-min warm-up session, the athletes performed maximum-effort acceler-

ation with and without the additional mass. They started with four maximum-effort
acceleration sprints in the control condition. Then they performed eight trials in the
additional mass condition. After the additional mass trials, the athletes repeated
four trials in the control condition. To reduce the confounding effects of learning,
three practice trials were required before each condition change. To minimize the
Fig. 22.1 Photograph of

the experimental
intervention – a life jacket
with additional mass of
12.4 kg. The life jacket was
securely fastened to the
athletes and the additional
mass elements were located
at the bottom of the jacket
close to the body centre
of mass
influence of fatigue, a minimum of a 3-min resting period was given between trials.
To detect if there existed any learning and/or fatigue effects, the propulsive ground
reaction impulse was compared between the control trials performed before versus
after the additional mass condition using two-tailed paired t-tests (α ¼ 0.05). Two
athletes showed a significant reduction in propulsive impulse generation in the
second control trials and were thus excluded from further analyses.
22.2.4 Data Analysis
Raw ground reaction forces and marker positions were filtered with a fourth-order
recursive Butterworth low-pass filter at a cut-off frequency of 60 and 20 Hz,
respectively. Ninety-nine percent of the integrated power content of the original

signal was contained by the filtered data. Due to the nature of the movement task,
sagittal plane ground reaction force and body marker data were then analyzed.
Vertical ground reaction force was used to define stance intervals with an onset
threshold of 3 % body weight.
The sagittal plane resultant ground reaction force was used to represent limb
force. The peak resultant ground reaction force defines the maximum limb force
generation within each condition. Stance-average propulsive and vertical limb force
were determined and compared, as these vectors represent the overall force appli-
cation over stance.
Body lean was quantified using effective leg kinematics. The effective leg was
defined as the vector pointing from the centre of pressure to the body centre of mass.
Centre of pressure was calculated using force signals and force-plate specifications.
Body centre of mass was estimated using the three markers placed on the pelvis.
During quiet standing trials performed prior to the acceleration, the vertical location
of the estimated centre of mass was calculated as the average height of the markers;
the horizontal location of the estimated centre of mass was defined as the point
where the ground reaction force vector intersected with the horizontal plane at the
aforementioned height. Transformation matrices between the standing trial and
each movement frame were calculated using the pelvis markers. These transforma-
tion matrices were then applied to the standing centre of mass to calculate its
location during acceleration (S€oderkvist and Wedin 1993). Effective leg angle in
the sagittal plane was calculated over stance. Body lean was expressed as the
angular deviation from the vertical axis (0 : vertical; positive angles: leaning
forward; negative angles: leaning backward).
Two-tailed paired t-tests were used for the statistical analyses. Statistical signif-
icance level was set a priori at α ¼ 0.05.
22.3 Results
Peak resultant ground reaction force increased significantly from the control to the
additional mass condition (control: 1,502.9 152.9 N, additional mass:
1,572.5 143.0 N; p < 0.001; Fig. 22.2). Averaged over stance, the resultant
ground reaction force was observed to be significantly larger in the additional
mass condition compared to the control (Table 22.1).
The differences observed in resultant ground reaction force could largely be
attributed to the changes in vertical ground reaction force. Peak vertical ground
reaction force increased from 1,393.7 147.8 N in the control to 1,475.7 134.5 N
in the additional mass condition (p < 0.001; Fig. 22.3a), meanwhile no significant
difference was detected for the peak propulsive force (control: 621.8 57.3 N,
additional mass: 607.1 53.2 N; p ¼ 0.098; Fig. 22.3b). Averaged over stance the
vertical ground reaction force increased by 9.5 % while a 3.6 % decrease in net
Fig. 22.2 Sagittal plane resultant ground reaction forces exerted during maximum-effort accel-
eration in the control and additional mass conditions. Thick lines are the athlete sample average,
and the thin lines are the 1 standard deviation. When accelerating with the additional mass, the
athletes were capable of generating a peak limb force 5 % greater than control
Table 22.1 Variables derived from ground reaction force data. The listed values are the athlete
sample averages 1 standard deviation
Control Additional mass p-values
Average resultant force [N] 984.4 100.9 1,060.7 89.6 <0.001
Average vertical force [N] 900.9 95.3 986.3 86.6 <0.001
Average propulsive force [N] 362.9 43.2 349.9 37.8 0.045
Stance time [ms] 210.6 27.8 230.1 26.1 <0.001
Peak force instance [% Stance] 57.8 5.6 57.5 4.0 0.659
Fig. 22.3 Ground reaction forces in the vertical, Figure (a), and propulsive, Figure (b), directions.
Thick lines are the athlete sample average, and the thin lines are the 1 standard deviation. While
the vertical force varied significantly, the peak propulsive force remained unchanged between
conditions
Table 22.2 Kinematic parameters of the effective leg. Effective leg is defined as the vector
pointing from the centre of pressure to the body centre of mass. 0 represents upright, and positive
angles represent leaning forward. The listed values are the athlete sample averages 1 standard
deviation
Control Additional mass p-values
Average effective leg angle [ ] 19.5 1.7 17.8 1.1 0.005
Effective leg landing angle [ ] 0.0 3.1 1.6 3.0 0.016
Effective leg take-off angle [ ] 40.4 2.6 39.6 2.4 <0.001
propulsive force was observed in the additional mass condition compared to the
control (Table 22.1).
Over stance, the athletes adopted a more upright effective leg posture acceler-
ating with an additional mass compared to the control (Table 22.2).
22.4 Discussion
The purpose of the study was to examine whether limb force generation is at its
peak capacity during maximum-effort acceleration, thus limiting performance. We
attempted to address this question by implementing an additional mass to modify
the mechanical demand for the supporting limb; it was hypothesized that limb force
generation would remain operating at the same maximum magnitude across con-
ditions. The results indicate the athletes were indeed capable of generating a greater
amount of limb force in the additional mass condition compared to the control. This
observation suggests that acceleration performance is not limited by maximum limb
force generation.
Although not directly addressing the limb force limit hypothesis, some previous
studies made observations that indirectly support our interpretation – they found
that total limb force was not the performance differentiator for sprint acceleration.
Kugler and Janshen (2010) examined ground reaction forces and effective leg
kinematics of the first step acceleration from a standing start. Using net propulsive
ground reaction impulse as the performance measure, they divided the athletes into
higher versus lower performance groups for comparisons. They found no difference
in maximum limb force generation between the two groups. The better performers
produced a significantly larger peak propulsive force, but showed a trend of lower
peak vertical force. Their results suggested technique (i.e. the ability to orient force
toward the travelling direction), instead of strength, determines performance. Their
results of effective leg kinematics supported this point: the better performers leaned
more into the ground while generating the propulsive force. In a recent study by
Morin et al. (2011), correlations between stance-average limb forces and 4-s sprint
distance (performance measure) were examined. Similar to the study by Kugler and
Janshen (2010), it was found that stance-average total limb force did not signifi-
cantly correlate with performance, while the propulsive force and limb force
Fig. 22.4 Hypothetical scenarios of full-capacity limb force applications. Arrows represent the
force vectors. Dotted lines were used to discuss potential performance constraints (see text in
Sect. 22.4 for more details). In Figure (a), we posed the question why the athletes would not fully
exert the limb force while retaining the technique (i.e. force orientation). In Figure (b), we asked
why the athletes would not orient the full limb force forward enough that all the force gain is used
for propulsion
orientation did. The authors thus concluded that orientation, instead of the magni-
tude, of the total limb force seems to be the primary determinant of performance.
In the current study, the athletes were capable of generating a limb force of
larger magnitude in the additional mass condition, but did not do so while retaining
the force orientation. The increase in resultant limb force was oriented in a more
upright direction, instead of contributing to propulsion. Both the peak and stance-
average vertical forces increased significantly with the additional mass compared to
the control. Meanwhile, the peak propulsive force remained constant. Averaged
over stance, the net propulsive force was even reduced.
If the goal of the task is to maximize propulsive force and impulse, why did the
athletes not do so by fully applying the limb force while retaining the force orienta-
tion? Figure 22.4a illustrates this scenario using the stance-average force vectors
from the current study. Hereby we use the magnitude of the resultant ground reaction
force during the additional mass condition to define the limb force limit (the red arrow
and dotted line). The blue arrow represents the limb force during the control condition
and the dotted line depicts its orientation. The hypothetical green arrow represents
our ‘why-not’ question – a fully applied limb force with retained orientation. Could
the hypothetical force application lead to a negative consequence to performance?
As the control condition force vector extends to its full capacity, both the
propulsive force and vertical supporting force will increase. While an increase in
propulsive force is favored, increasing vertical impulse may have a negative
influence on performance by elongating flight.
During legged locomotion, athletes can only create propulsive impulse during
ground contact. Since the goal of acceleration is to accumulate travelling momen-
tum at the highest rate, it is likely that athletes would benefit from being almost in
constant contact with the ground while generating minimum vertical centre of mass
oscillation (i.e. operating at high duty factors essentially resembling a wheel).
Therefore an elongated flight can reduce the athletes’ potential to accumulate
propulsive impulse within a given time window. To illustrate this point, let us
assume the stance time remains constant between the control and full-capacity push
(blue vs. green vector in Fig. 22.4a) – which is rather conservative since the full-
capacity push would likely lead to an earlier take-off ceasing propulsive impulse
generation. Using the average ground reaction forces from the current experiment
(Table 22.1) and body weight, we can then calculate the net vertical impulses.
Knowing the body mass and vertical impulses we can derive the difference in
vertical launching velocity which in turn allows us to estimate flight time differ-
ences. Over stance, the difference in vertical impulse would lead to a 0.20 m/s
increase in vertical launching velocity for the full-capacity push, which elongates
flight time by 0.04 s, assuming the launching and landing height remains the same.
With this information we can compare the gain versus loss between the two
strategies. Throughout the initial ground contact, the larger propulsive force gen-
erated in the full-capacity case allows a 5.9 Ns propulsive impulse gain compared to
the control (difference in average propulsive force multiplied by stance time).
However, when the full-capacity case initiates the second ground contact, the
control case would have been in the second ground contact for 0.04 s and created
a propulsive impulse of 14.5 Ns (average control propulsive force multiplied by the
elongated flight time). Over such interval, exerting limb force at its full capacity
represents a net loss for generating propulsive impulse. Based on this logic, we
would expect the athletes to avoid an elongated flight by only creating sufficient
vertical impulse to, for example, reposition the opposite limb. Findings from a
previous sprint acceleration study indirectly support this speculation on a vertical
force constraint. Luo and Stefanyshyn (2011) examined the extent to which avail-
able traction can constrain acceleration performance. As available traction
increased from traction coefficient of 0.26–0.54, the athletes were able to lean
into the ground and generate a larger limb force. This was accomplished while the
stance-average vertical limb force remained operating at the same magnitude, as the
propulsive force increased by 64 %.
Assuming excessive vertical impulse can have a negative consequence for
acceleration performance, why would the athletes not fully apply the limb force
while keeping the vertical force at the control level? Figure 22.4b illustrates this
scenario. Here we assume: (1) the average resultant force generated during the
additional mass condition defines the limb force capacity (the red arrow and dotted
line), (2) and the vertical force during control defines the optimized level for flight
(blue arrow and dotted line). Then, the intersection between the two constraints
would define the hypothetical limb force (green arrow) in question. To explore this
question, we would likely need to consider another important control variable
during locomotion – angular momentum.
During contact, ground reaction force can create a moment about the body centre
of mass when its line of action does not pass through it; if not carefully controlled,
such moment can accumulate to an angular momentum large enough to perturb
balance (e.g. pitching forward). It has been shown that during walking, for example,
ground reaction force was carefully controlled to couple with the centre of mass
travel (Herr and Popovic 2008). Indeed, the fine control of body angular momentum
has been a major challenge in designing controllers for bipedal locomotion models
(de Lasa et al. 2010). For this reason, the limb force orientation needs to be coupled
to the effective leg orientation. Results from the current study further confirmed this
point. During the additional mass condition, as the athletes generated a more
upright force, the lean angle of the effective leg became more upright accordingly
(Table 22.2).
It is possible that during maximum-effort acceleration, the effective leg reaches its
lean angle constraint, thus limiting the athletes from directing limb force more
anteriorly. Given such a kinematic constraint, a more anteriorly oriented limb force
would produce a nose-up pitching moment and can perturb balance. Unfortunately
the current experiment does not contain sufficient data to further investigate this idea.
One potential method to examine this idea in the future would be by manipulating the
angular momentum of the athletes. For example, by applying a forward pulling force
above an athlete’s centre of mass, it may allow the athlete to create a more forward-
oriented force (the green arrow in Fig. 22.4b) without changing the body lean. This
method would essentially decouple the limb force and effective leg orientation.
Such kinematic constraint may exist for various reasons. First, leaning further in
may have a negative influence on the joint operation ranges of the opposite limb. As
the body leans in, the body centre of mass will be lowered which may in turn
impose a more flexed posture for the opposite limb upon ground contact, resem-
bling ‘Groucho running’ (McMahon et al. 1987). Skeletal muscle’s capability to
generate force changes with its contraction length (Rassier et al. 1998). Such flexed
body posture may position the joint extensor muscles at a length not ideal for force
generation. Meanwhile, the more flexed posture will likely reduce the effective
mechanical advantage of the joint extensors (Biewener 1989). Secondly, in order
for the body to roll into the lean-in position after the initial contact, a larger braking
force may be needed, which can be counter-productive for propulsion. During
acceleration, the effective leg undergoes a rotation about the contact point first,
followed by an explosive leg extension after the body leans in (Jacobs and van
Ingen Schenau 1992). For the body to roll into the push-off position, a head-down
pitch moment is needed. In order to generate such moment, the line of action of the
ground reaction force needs to be placed posterior to the centre of mass, especially
during early stance. Since the body is in a rather upright orientation during this
phase, increasing the nose-down pitching moment may need to be accomplished by
applying a larger backward braking force, which can lead to a larger braking
impulse. Such an increase in braking impulse can compromise the performance
gain from the increased propulsive impulse by leaning more into the ground.
Future experiments are needed to further investigate these speculations. One
potential experimental implementation would be performing acceleration on an
adjustable incline instrumented with force platforms. By varying the incline angle,
the researchers would be able to directly control the perpendicular distance between
the centre of mass and ground, which influences the limb kinematics during
landing, and then observe changes in the capacity to generate limb force. By having
the athletes sprint on various incline angles, the need for forward body roll can be
reduced in a systematic manner. Thus the need to create nose-down pitch moment
can be controlled. It would allow the experimenters to explore the strategy changes
in limb force generation at different phases of stance.
22.5 Conclusion
The current experiment aimed to examine whether limb force generation is at its
peak capacity during maximum-effort acceleration, thus limiting ultimate perfor-
mance. The results showed the athletes retained extra limb force generation capac-
ity, suggesting factors other than limb force constrained performance. Various
speculations were made about why the athletes would not fully utilize the force
generation capacity. Future studies are needed to further examine these ideas.
References
Alexander R (2002) Stability and manoeuvrability of terrestrial vertebrates. Integr Comp Biol
42(1):158–164
Biewener A (1989) Scaling body support in mammals: limb posture and muscle mechanics.
Science 245:45–48
de Lasa M, Mordatch I, Hertzmann A (2010) Feature-based locomotion controllers. ACM Trans
Graph 29(4):131
Herr H, Popovic M (2008) Angular momentum in human walking. J Exp Biol 211(4):467–481
Jacobs R, van Ingen SG (1992) Intermuscular coordination in a sprint push-off. J Biomech
25(9):953–965
Kugler F, Janshen L (2010) Body position determines propulsive forces in accelerated running.
J Biomech 43(2):343–348
Luo G, Stefanyshyn D (2011) Identification of critical traction values for maximum athletic
performance. Footwear Sci 3:127–138
McMahon T, Valiant G, Frederick E (1987) Groucho running. J Appl Physiol 62(6):2326–2337
Mero A (1988) Force-time characteristics and running velocity of male sprinters during the
acceleration phase of sprinting. Res Q Exerc Sport 59(2):94–98
Mohr M, Krustrup P, Bangsbo J (2003) Match performance of high-standard soccer players with
special reference to development of fatigue. J Sports Sci 21(7):519–528
Morin J, Edouard P, Samozino P (2011) Technical ability of force application as a determinant
factor of sprint performance. Med Sci Sports Exerc 43(9):1680–1688
Rassier D, MacIntosh B, Herzog W (1998) Length dependence of active force production in
skeletal muscle. J Appl Physiol 86(5):1445–1457
S€
oderkvist I, Wedin P (1993) Determining the movements of the skeleton using well-configured
markers. J Biomech 26(12):1473–1477
Chapter 23
Optimal Technique, Variability and Control
in Gymnastics
Michael Hiley and M.R. (Fred) Yeadon
Abstract Optimisation is often used in an attempt to explain technique adopted in

skilled sport performance. This might take the form of minimising joint torques in
an expectation that the optimum simulated technique will resemble the actual
performance. If a suitable optimisation criterion can be identified then this may
give some insight into the adopted technique. In all human movement there is
inherent variation so that no two performances are exactly the same. As a conse-
quence skilled technique needs to be successful in a noisy environment and so
optimised technique also needs to be robust to the inherent variation in coordina-
tion. In movements in which there is sufficient time for feedback control to operate
it is to be expected that there will be greater variation in technique in those phases
that adjustments are made. It is also to be expected that there will be little variation
in technique for those phases where accurate coordination is crucial to the success
of the movement. The aspect that often governs elite technique is that of achieving
consistent success rather than some biomechanical measure of movement.
Keywords Simulation • Optimisation • Motor control • Variability
23.1 Introduction
Biomechanics is often used to try to explain the techniques adopted by skilled

sports competitors. While a biomechanical analysis can often provide important
insights, other aspects of technique such as variability and control are important
determinants and should also be considered.
M. Hiley (*) • M.R.Yeadon

School of Sport, Exercise and Health Sciences, Loughborough University, Loughborough, UK
e-mail: m.j.hiley@lboro.ac.uk

294 M. Hiley and M.R. Yeadon
23.2 Optimisation
Computer simulation modelling and optimisation are often used in an attempt to

explain technique adopted in skilled sport performance. Optimisation criteria such
as minimisation of joint torque are often used to attempt to account for techniques
adopted by elite gymnasts. This may be appropriate for many everyday activities
such as walking (Ren et al. 2007; Anderson and Pandy 2001), but in gymnastics and
other sports effort is often maximised in order to achieve the desired performance
outcome. It is therefore important to incorporate realistic strength characteristics in
the simulation model although strength should not be considered as the only limit to
performance.
For example, in the undersomersault on parallel bars the gymnast starts in
handstand, circles beneath the bars, has a short flight phase and then re-grasps the
bars close to the final handstand position (Fig. 23.1a). Torque minimisation
(Fig. 23.1b) results in a technique recommended for beginners (Davis 2005;
Hiley and Yeadon 2012), while the technique of elite gymnasts can be characterised
as having a vertical mass centre velocity at hand release (Fig. 23.1c). The vertical
path has two advantages: firstly the direction of the mass centre velocity changes
less near to release, when compared to the technique obtained from torque
minimisation, and this should lead to a more consistent performance. Secondly,
the undersomersault to handstand forms the basis of more complex skills: typically
the undersomersault with either a half or full twist before reaching the handstand
position. In these skills there is not a flight phase, instead the gymnast makes the
necessary hand changes whilst the force on the bars is low. Having a vertical mass
centre velocity while the body is twisting reduces the task complexity of the hand
changes (i.e. less correction is needed for non-vertical alignment). Although the
Fig. 23.1 Graphics

sequences of (a) recorded
performance, and optimum
simulations, (b) minimising
joint torques and (c)
producing vertical mass
centre path (Hiley and
Yeadon 2012)
23 Optimal Technique, Variability and Control in Gymnastics 295
Fig. 23.2 The two techniques used by elite gymnasts during accelerated backward giant circles
prior to a double layout somersault dismount: (a) traditional technique and (b) the scooped
technique (Hiley and Yeadon 2003a, b)
technique is more demanding in terms of strength, it has advantages in terms of

consistency and skill development.
In technical sports such as gymnastics the accuracy of timing and the accuracy
with which the human motor system can perform movements will have a bearing on
the movement strategies adopted. For example, prior to a dismount on high bar an
accelerated giant circle is used to generate the necessary linear and angular
momentum for the flight phase (Bruggemann et al. 1994). The traditional giant
circle in which the body remains extended through the highest point has been
superseded by a scooped giant circle in which the body flexes through the highest
point and hyperextends at the lowest point of the circle (Fig. 23.2). An attempt to
explain this change in terms of ability to produce more angular momentum failed
(Hiley and Yeadon 2003a). The change was accounted for by considering the
margin for error in timing the release (Hiley and Yeadon 2003b). The scooped
technique exhibited larger release windows for successful dismounts, where the
release window was defined as the period of time during the giant circle where the
gymnast had sufficient linear and angular momentum to successfully perform the
dismount. The larger the release window the more time during the giant circle the
gymnast has the appropriate release conditions. As a result using the scooped
technique the gymnast has the impression that timing the release is easier compared
with the traditional technique. In reality, when using the scooped technique the
timing becomes less critical as there is an extended period of time where a
successful dismount can be performed.
In order to then maximise the angular momentum in the giant circles preceding a
dismount, to either improve performance or investigate performing new dismounts,
a constraint of a sufficiently large release window is required (Hiley and Yeadon
2005). As the amount of angular momentum at release increases the size of the
release window decreases. For example, using a simulation model it was found that
a gymnast is capable of producing sufficient angular momentum to perform a triple
Fig. 23.3 Limiting dismounts using optimised giant circles with full strength: (a) triple piked
somersault, (b) triple somersault with two twists (Hiley and Yeadon 2005)
straight somersault dismount but the release window was found to be rather small
(Hiley and Yeadon 2005). When the angular momentum was limited by having a
suitably large release window, a triple piked somersault dismount or a double
twisting triple somersault was possible (Fig. 23.3).
23.3 Variability
In all sports activities there is inherent variation so that no two performances are
exactly the same. As a consequence skilled technique needs to be successful in a
noisy environment and so optimised technique needs to be robust to the inherent
variation in coordination. It is necessary to determine the level of variability in
order to conduct such optimisation studies. This was investigated for regular and
accelerated giant circles on high bar (Fig. 23.4). Four elite male gymnasts each
performed ten regular giant circles in succession, where the aim was to swing from
handstand to handstand with good form and even pace (Fig. 23.4a) and ten double
straight somersault dismounts, from which the preceding accelerated giant circles
were analysed (Fig. 23.4b). For giant circles the actions performed through the
lowest part of the circle have been shown to be mechanically the most important
(Arampatzis and Bruggemann 1999; Irwin and Kerwin 2007; Yeadon and Hiley
2000) since they provide the necessary input of energy into the system for the
gymnast to swing back above the bar for regular circles and they help generate the
required linear and angular momentum (and hence release window) for the subse-
quent dismount for the accelerated giant circles.
Fig. 23.4 Two types of backward giant circles performed on the high bar: (a) the regular giant
circle and (b) accelerated giant circles
Fig. 23.5 Time histories of the hip and shoulder angles for ten (a) regular and (b) accelerated
giant circles, with graphics to show the gymnast orientation (Hiley et al. 2013)
The trials for each gymnast and each type of giant circle were aligned when the
mass centre was level with the bar (time zero in Fig. 23.5). The amount of
movement variability was quantified in terms of the standard deviations of the
time and angle at turning points between flexion and extension at the hip and
shoulder (Fig. 23.5). The gymnasts’ movement patterns were very consistent
through the lower part of the circle, with average standard deviations (from hip
and shoulder) of 15 and 11 ms for the regular and accelerated circles, respectively.
Through the upper part of the circles the average standard deviation was larger
(93 ms) for the regular giant circles, whereas the average standard deviation for the
accelerated giant circles was 12 ms. It was found that the more elite gymnasts were
more consistent in the mechanically important aspects of technique.
It is speculated that the gymnasts try to perform the mechanically important
aspects of technique as accurately as possible, i.e. try to minimise the amount of
variability in these actions, since they have a direct effect on the successful outcome
of the skill. In order to achieve this, the gymnast uses feedback control through the
upper part of the circle where adjustments can be made (e.g. to the whole body
angular velocity) to ensure that the mechanically important aspects can be timed
correctly. Although the gymnasts appear to be more variable through the upper part
of the circle, they are performing the necessary corrections to prepare for the
mechanically important actions.
The level of variability in gymnasts’ movements has implications for the results
of optimisation studies and the techniques adopted. Optimal technique, by defini-
tion, is sensitive to perturbations, therefore introducing small timing errors can lead
to sub-optimal performance. In the piked triple somersault dismount shown in
Fig. 23.3a it was found that when the timing of the actions at the hip and shoulder
joints of the optimum simulation were perturbed by 30 ms, the resulting simulation
could no longer meet the criteria for a successful dismount (i.e. sufficient aerial
rotation and release window, Fig. 23.6a). Since it is to be expected that a gymnast’s
technique can cope with small errors in timing for consistent performance, a
requirement of robustness to timing perturbations should be included within the
optimisation process. When the technique in the backward giant circle was
optimised to be robust to 30 ms perturbations it was found that sufficient linear
and angular momentum for a triple piked dismount could be achieved with a
realistic release window (Fig. 23.6b).
Gymnastics is a sport that demands consistency. For a gymnast to include a skill
in a routine a success rate of over 90 % would be expected. Reasons for inconsis-
tency include: poor technique, lack of strength, lack of flexibility and poor
Fig. 23.6 The size of release window when the (a) maximised and (b) robust to 30 ms simulations
are perturbed by 30 ms using five different perturbation combinations – [1] the unperturbed
simulation, [2] shoulder and hip actions early, [3] shoulder and hip actions late, [4] shoulder
early with hip late, and [5] shoulder late with hip early (Hiley and Yeadon 2008)
Fig. 23.7 The upstart

(Adapted from the FIG
Code of Points 2009)
coordination precision which refers to how accurately the gymnast can produce the
desired movements. We know that the motor system is noisy and that there will be
variability in the movements when the gymnast tries to perform the same skill a
repeated number of times.
The effect of noise from the motor system on technique and consistency of
performance can be demonstrated with the upstart (sometimes called a kip) on a
bar. The upstart is a fundamental skill in gymnastics which is used to transfer a
gymnast from a swing beneath the bar to a position above the bar (Fig. 23.7). The
gymnast typically swings forwards to an extended body position, closes the hip and
shoulder angle to bring the feet to the bar, then rapidly extends at the hip whilst
continuing to close the shoulder angle, which allows the gymnast to circle round to
the final support position.
Previous research has tried to establish the underlying strategy for the upstart by
optimising the technique using various criteria (Yamasaki et al. 2010). The idea was
that the solution which most closely resembles the gymnast technique will charac-
terise the strategy adopted. Yamasaki et al. (2010) optimised using criteria based on
minimising angle jerk, torque and torque change. They could only find techniques
close to that of the gymnast when a “via point” was imposed, a constraint that forced
the hip and shoulder angle to match the recorded performances at an arbitrary point.
It should therefore be concluded that, in this study, some aspect of human movement
has not been considered. Wolpert (2007) proposed that humans use prior knowledge
of movements and situations in order to choose a strategy that maximises the
likelihood of success at the task. This includes the knowledge that due to noise
from the motor system our movements are never exactly how we planned them.
Therefore we need a technique that can cope with this variability.
Technique in the upstart was subsequently investigated by Hiley and Yeadon
(2013). A recording of a gymnast trial was made, then a simulation model of the
gymnast and bar was used to optimise the technique under various criteria:
minimising joint torque, minimising torque change, maximising success, where
maximising success searches for a solution that produces a successful upstart despite
the presence of noise from the motor system. For all three of the optimisations,
16 parameters defining the joint angle time histories at the hip and shoulder were
manipulated to optimize the score. For the maximising success optimisation, for each
set of parameters produced by the optimisation algorithm 1,000 randomly perturbed
simulations were performed. The perturbations were added to the 16 joint angle time
history parameters using a random number generator with a normal distribution.
Fig. 23.8 Graphics

sequences of: (a) the
matching simulation of the
recorded performance and
the optimal solutions for (b)
minimising torque, (c)
minimising torque change
and (d) maximising success
(Hiley and Yeadon 2013)
Kinematic variability may be due to a combination of noise from the motor system,
errors (in planning and execution), feedback control and covariation. In this case the
perturbations were used to represent the noise from the motor system and were based
on the variability measured in giant circles. Success was defined as being able to
achieve the final orientation above the bar without exceeding the joint torque limits.
The number of successful simulations out of 1,000 was recorded as the score. The
maximising success optimisation therefore looks for a technique that can cope with
the noise added to the planned joint angle time histories.
The optimal solutions for minimising joint torque and torque change deviated
from the recorded performance (Fig. 23.8). Root mean square differences between
the recorded and optimal joint angle time histories confirmed that the maximising
success criterion most closely resembled the gymnast’s technique – without the
need for a via point (Fig. 23.8). It may be concluded that the gymnast’s technique is
characterised by maximising success at the upstart despite the presence of noise and
that gymnasts develop techniques that are able to cope with the level of noise within
their movements. Maximising success really only finds the solution space for the
movement given the constraints placed on the system. If the task is complex, like
the upstart, with a large number of constraints including coping with the noise from
the motor system, it might be that the solution space is small and effectively defines
the technique. Unless noise or variability is included in the optimisation process,
simulation solutions are rarely naturally robust to perturbations. Therefore it is
important to include this aspect of human movement when optimising technique.
23.4 Control
Many sports movements are only possible because the athlete makes continual
adjustments. For example in a hand balance a gymnast must continually correct for
movement of the mass centre (Yeadon and Trewartha 2003). In movements with a
flight phase the athlete will make adjustments towards the end of flight in order to
achieve an appropriate landing. In a double straight somersault failure to make such
corrections will lead to an unwanted half twist due to mechanical instability
together with inherent slight asymmetries in initial conditions or body configuration
(Fig. 23.9).
If in a plain jump, with arms above the head, a gymnast drops one arm to the
side, due to the conservation of angular momentum, the whole body will tilt
(a rotation about the frontal axis). If a gymnast produces a tilt whilst in a straight
somersault the result is a twisting somersault (Yeadon 1993). Therefore asymmet-
rical arm movements can be used to create a twisting somersault. Similarly asym-
metrical arm movements can be used to remove unwanted twist from a non-twisting
somersault (Yeadon and Mikulcik 1996). To prevent the build-up of twist, small
asymmetrical arm movements may be used based on twist angular velocity and
acceleration information provided by the inner ear balance mechanisms. This is
equivalent to a proportional plus derivative control strategy. However, as this is
human movement there must be a delay in the system. Yeadon and Mikulcik (1996)
used a computer simulation model to show that so long as the feedback time delay
in the control system is less than 0.25 somersaults such control can prevent the
build-up of twist (Fig. 23.10). Given the duration of a double straight somersault in
gymnastics it is most likely that these small corrections are made using long loop
reflexes or triggered responses. In other words, the gymnast is unlikely to be aware
of the corrections being made. This technique can also be used to control twist
towards the end of a twisting somersault (Yeadon 2002). Indeed it is likely that
gymnasts first learn this method of control in single somersaults with twist and so
control is in place when the first double straight somersault (without twist) is
attempted.
In order to make adjustments during the aerial phase of a gymnastics movement
it is of advantage to obtain good visual information on the progress of the twist and
somersault throughout the aerial phase. This can be achieved by adjusting head
position so that the landing area can be viewed throughout the skill. From an initial
Fig. 23.9 A rigid configuration with only 1 of asymmetry in arm abduction angles produces
almost a half twist after two somersaults in the straight position
Fig. 23.10 Proportional plus derivative control in straight double somersaults with feedback
delays of (a) 0.02, (b) 0.12, and (c) 0.24 somersaults (Yeadon and Mikulcik 2000)
Fig. 23.11 Use of a virtual

environment to learn the
head movements required
for viewing (Yeadon and
Knight 2012)
video analysis of high bar dismounts at the recent London 2012 Olympic Games it
appears that approximately 50 % of male gymnasts use head movements in order to
view the landing area throughout the majority of the dismount. Analysis is being
conducted to determine whether viewing is related to more consistent and success-
ful landings (i.e. landings that receive fewer deductions from the judges based on
steps and excessively deep hip and knee flexion). Such head movement can be
learned in a virtual environment (Fig. 23.11). The starting point is a simulation of a
twisting somersault. In this example a double straight somersault with a full twist in
the second somersault. A video of the movement can then be generated (rendered)
from any camera view. For the virtual environment the camera is placed within the
head of the simulation model to give a first person view of the skill. Using a three
Fig. 23.12 A full twisting

double somersault dismount
from high bar whilst
viewing the landing area
throughout
dimensional motion sensor attached to the virtual reality head set, the user
can control the orientation of the head within the simulation, which is then rendered
in real-time.
After a few minutes of using the virtual viewing system to learn the appropriate
head movement for viewing during a full twisting double somersault simulation, the
gymnast learned to apply the same head movement strategy to high bar dismounts
and was successful in learning to view during a full twisting double somersault
dismount on the same day (Fig. 23.12). With a basic strategy of tilting the chin
down when the body is upright and tilting the head back when the body is upside
down the gymnast can view the landing area throughout the dismount.
23.5 Conclusion
Sports movements are learned and refined in an environment of constraints com-

prising strength, flexibility, anatomical limits, coordination precision as well as the
mechanical constraints of the particular movement. In movements in which there is
sufficient time for feedback control to operate it is to be expected that there will be
greater variation in technique in those phases in which adjustments are made. It is
also to be expected that there will be little variation in technique in those phases
where accurate coordination is crucial to the success of the movement. The aspect
that often governs elite technique is that of achieving consistent success rather than
some biomechanical measure of movement.
References
Anderson FC, Pandy MG (2001) Dynamic optimisation of human walking. ASME J Biomech Eng
123:381–390
Arampatzis D, Brüggemann G-P (1999) Mechanical and energetic processes during the giant
swing exercise before dismount and flight elements on the high bar and uneven parallel bars.
J Biomech 32:811–820
Brüggemann G-P, Cheetham PJ, Alp Y, Arampatzis D (1994) Approach to a biomechanical profile
of dismounts and release-regrasp skills of the high bar. J Appl Biomech 10:291–312
Davis J (2005) Undersomersaults on parallel bars. Gym Craft 14:6–7
Fédération Internationale de Gymnastique (2009) Code of points. F.I.G, Moutier
Hiley MJ, Yeadon MR (2003a) Optimum technique for generating angular momentum in accel-
erated backward giant circles prior to a dismount. J Appl Biomech 19:119–130
Hiley MJ, Yeadon MR (2003b) The margin for error when releasing the high bar for dismounts.
J Biomech 36:313–319
Hiley MJ, Yeadon MR (2005) Maximal dismounts from high bar. J Biomech 38:2221–2227
Hiley MJ, Yeadon MR (2008) Optimisation of high bar circling technique for consistent perfor-
mance of a triple piked somersault dismount. J Biomech 41:1730–1735
Hiley MJ, Yeadon MR (2012) The effect of cost function on optimum technique of the
undersomersault on parallel bars. J Appl Biomech 28:10–19
Hiley MJ, Yeadon MR (2013) Investigating optimal technique in a noisy environment: application
to the upstart on uneven bars. Hum Mov Sci 32:181–191
Hiley MJ, Zuevsky VV, Yeadon MR (2013) Is skilled technique characterised by high or low
variability? – An analysis of high bar giant circles. Hum Mov Sci 31:171–180
Irwin G, Kerwin DG (2007) Musculoskeletal work of high bar progressions. Sports Biomech
6:360–373
Ren L, Jones RK, Howard D (2007) Predictive modelling of human walking over a complete gait
cycle. J Biomech 40:1567–1574
Wolpert DM (2007) Probabilistic models in human sensorimotor control. Hum Mov Sci
26:511–524
Yamasaki T, Gotoh K, Xin X (2010) Optimality of a kip performance on the high bar: an example
of skilled goal-directed whole body movement. Hum Mov Sci 29:464–482
Yeadon MR (1993) The biomechanics of twisting somersaults: Part III: Aerial twist. J Sport Sci
11:209–218
Yeadon MR (2002) The control of twisting somersaults using asymmetrical arm movements.
In: IV World Congress of Biomechanics Proceedings, Calgary, pp 40–41
Yeadon MR, Hiley MJ (2000) Mechanics of the backward giant circle on the high bar. Hum Mov
Sci 19:153–173
Yeadon MR, Knight JP (2012) A virtual environment for learning to view during aerial move-
ments. Compt Methods Biomech 15:919–924
Yeadon MR, Mikulcik EC (1996) The control of non-twisting somersaults using configurational
changes. J Biomech 29:1341–1348
Yeadon MR, Mikulcik EC (2000) Stability and control of aerial movements. In: Nigg BM,
MacIntosh BR (eds) Biomechanics and biology of movement. Human Kinetics, Champaign,
pp 211–221
Yeadon MR, Trewartha G (2003) Control strategy for a hand balance. Motor Control 7:411–430
Chapter 24
Activity of the Trunk and Leg Musculature
During the Flutter Kick
Koji Kaneoka, Atsushi Imai, and Morimitsu Kohdate
Abstract There are no in-depth biomechanical studies about swimming because

detailed analyses of motion and muscle activity in water are difficult. In the present
study, we measured the activity of trunk and leg muscles during the flutter kick,
using a waterproof, wireless electromyographic system. Flutter kick movements
were recorded by a waterproof high-speed camera, and were divided into four
phases to analyze muscle activity in each phase: (1) hip extension-knee extension,
(2) hip extension-knee flexion, (3) hip flexion-knee flexion, and (4) hip flexion-knee
extension. The femoral biceps and gastrocnemius became active from Phase 1 to
2, the rectus femoris and vastus medialis became active from Phase 3 to 4, and the
internal oblique muscle activity was high from Phase 2 to 3. In this study, the
muscle activity patterns during the flutter kick were described. It was suggested that
for this kick the internal oblique muscle has an important role during the switch
over from hip extension to flexion.
Keywords Electromyography • Swimming • Trunk muscle
24.1 Introduction
Exercise patterns in water typically differ from those on land because of the
viscosity and buoyancy of water. However, detailed analyses of motion and muscle
activity during swimming are difficult. Therefore, no in-depth biomechanical
studies have been performed on swimmers, although a few studies have monitored
arm muscle activity during the front crawl stroke (Conceicao et al. 2010;

K. Kaneoka (*) • A. Imai • M. Kohdate
e-mail: kaneoka@waseda.jp

306 K. Kaneoka et al.
Stirn et al. 2011, 2013). To our knowledge, no study has divided the flutter kick
motion into phases and analyzed muscle activity. Analyzing muscle activity during
the flutter kick may provide useful data to improve swimming performance.
Moreover, to reveal the function of the muscles in water may aid rehabilitation
and fitness conditioning because the flutter kick is a fundamental movement in
water and aquatic exercises are useful in certain situations because they are free
from the influence of gravity. Accordingly, in this chapter we analyze muscle
activity during each phase of the flutter kick.
24.2 Methods
The subjects were 15 healthy men (mean SD age: 20 2 years, height:

172.7 4.2 cm, weight: 67.9 5.6 kg) with experience in competitive swimming.
Informed consent was obtained from all subjects.
Electromyographic analysis was performed using a waterproof wireless electro-
myograph (Bio Log DL-5000; S&ME). The muscles analyzed were the right rectus
abdominis, internal oblique, gluteus medius, erector spinae, rectus femoris, vastus
medialis, biceps femoris, and gastrocnemius (Fig. 24.1). Surface electrodes were
attached in alignment with the direction of the muscle fibers, and a waterproofed
data logger was attached to the skin by double-sided tape and protected from the
water with a medical-quality waterproofing sheet (Fig. 24.2). The maximum vol-
untary contraction (MVC) of each muscle was measured, and the %MVC of muscle
activity during each phase of kicking was calculated for each muscle.
The aquatic exercise was recorded from the right hand side with a single
waterproof high-speed camera (200 Hz HAS-220; DITECT). Body surface markers
were placed on the pelvic line joining the superior anterior iliac spine and superior
posterior iliac spine as well as the right greater trochanter, lateral femoral
epicondyle, and lateral malleolus (Fig. 24.3). Subjects did one performance of the
flutter kick with maximum effort in a 25-m indoor swimming pool. For analysis, the
hip angle (i.e., the angle formed by the line joining the greater trochanter to the
femoral lateral epicondyle and the pelvic line) and knee angle (i.e., the angle
formed by the greater trochanter, femoral lateral epicondyle, and lateral malleolus)
during a single kick were calculated from the markers on the subjects’ bodies using
image analysis software (Dipp-Motion Pro; DITECT). Kick movements were
divided into the following 4 phases on the basis of the behavior of the hip and
knee joints (Fig. 24.4).
1. Hip extension–knee extension
2. Hip extension–knee flexion
3. Hip flexion–knee flexion
4. Hip flexion–knee extension
24 Activity of the Trunk and Leg Musculature During the Flutter Kick 307
Fig. 24.1 Anterior (a) and posterior (b) views of electrode placement. BF biceps femoris, ES
erector spinea, GC gastrocnemius, GM gluteus medius, IO internal oblique, RA rectus abdominis,
RF rectus femoris, and VM vastus medialis
To examine changes in muscle activity during the flutter kick movements, each
phase was divided into ten sections, and %MVC was used to produce a linear
envelope (i.e., mean rectified value).
24.3 Results and Discussion
The raw electromyographic data of each muscle during the is presented in Fig. 24.5.
Figures 24.6 and 24.7 show the activity patterns of the leg and trunk muscles
respectively, during flutter kicking performed with maximum effort.
In Phase 1, which consisted of the hip extension–knee extension motion, the
biceps femoris became active, followed by the gastrocnemius. As the contraction of
Fig. 24.2 Waterproofing

process for the electrode
and data logger
the biceps femoris caused hip extension and knee flexion; the biceps femoris is
likely to be undergoing eccentric contraction during this phase. In Phase 2, which
consisted of the hip extension–knee flexion motion, the biceps femoris and gas-
trocnemius both exhibited muscle activity; this peaked in both muscles at the
transition from Phase 1 to 2, which is the point at which the knee switches from
extension to flexion.
In Phase 3, which consisted of the hip flexion–knee flexion motion, activity in
the rectus femoris was initiated, followed by an increase in the activity of the vastus
medialis. As the contraction of the rectus femoris caused hip flexion and knee
extension, it is likely to be engaged in eccentric contraction during this phase. Then,
the activities of the rectus femoris and vastus medialis peaked during Phase
4, which consisted of a hip flexion–knee extension motion.
Internal oblique muscle activity was evident from Phase 2 to 3. The internal
oblique muscle is the one of deep trunk muscles, and it contributes to the stabili-
zation of the lumbar vertebrae and pelvis. This muscle exhibited a high activity
level when the hip switched from extension to flexion, suggesting that this activity
may stabilize the yaw moment toward the trunk which is associated with the return
motion of the hip. Transition to next kick movements cannot be performed
Fig. 24.3 Right side view

of the body surface marker
placement
smoothly when this return motion of the hip is delayed. In this case forward
momentum cannot be effectively generated and excessive stress may be directed
toward the low back (e.g. lumbar facet) by hyperextension of the trunk. This
suggests that it is important to control hip movements and trunk stability by activity
in the internal oblique muscle during the flutter kick.
In conclusion, as a result of detailed recording and analysis of muscle activity in
each phase of the flutter kick, the activity patterns of the trunk and leg musculature
Fig. 24.4 Four phases as derived from hip and knee motion
Fig. 24.5 The raw electromyographic data obtained during the flutter kick. Phase 1; hip
extension-knee extension, Phase 2; hip extension-knee flexion, Phase 3; hip flexion-knee flexion,
Phase 4; hip flexion-knee extension
Fig. 24.6 The mean electromyographic linear envelope of the lower extremity muscles for all
subjects
Fig. 24.7 The mean electromyographic linear envelope analysis of the trunk muscles for all
subjects
were elucidated. Information gained in the study leads to the probability that when
performing the flutter kick, activity of the internal oblique is important for the
switch over from hip extension to hip flexion.
References
Conceição A, Silva A, Palma S, Silva H, Gamboa H, Louro H (2010) Electromyography in front

crawl technique-case study. Open Sports Sci J 3:67–69
Stirn I, Jarm T, Kapus V, Strojnik V (2011) Evaluation of muscle fatigue during 100-m front
crawl. Eur J Appl Physiol 111:101–113
Stirn I, Jarm T, Kapus V, Strojnik V (2013) Evaluation of mean power spectral frequency of EMG
signal during 100 metre crawl. Eur J Sport Sci 13:164–173
Chapter 25
Open Water Swimming Performance
Reira Hara and Isao Muraoka
Abstract Open water swimming is defined as any swimming competition that

takes place in rivers, lakes, oceans, or water channels. Since the inclusion of open
water swimming as an Olympic event in 2008, it has steadily been gaining popu-
larity worldwide. Many studies have focused on the risks involved in open water
swimming events. The purpose of this chapter is to review and shed light on the
risks and performance factors of open water swimming. It is hoped that this
information will improve both safety and performance worldwide. Most studies
have reported that open water swimming involves risks of hypothermia and
afterdrop. Since hypothermia can lead to death, adequate medical preparations
must be made in order to attend to emergency cases. Another possible risk involved
with the sport is heat stroke. Fédération Internationale de Natation (FINA) sets rules
for water temperature on race days. However, some studies have found that racing
under the FINA rules can still leave some swimmers vulnerable to heat stroke.
Special attention must be given to the race conditions and should include the
possibility of providing food to contestants during the race. The United States
and Brazil performed experiments with their top athletes to investigate physiolog-
ical stresses in relation to race conditions. These studies suggested that pre core
temperature and capillary glycemia were related to their race times. Because these
studies were reported in 2004 (VanHeest et al. 2004), further investigation must be
conducted in open water swimmers to determine the factors that can maximize
performance in this event.
Keywords Open water swimming • Risk • Performance
R. Hara (*)
e-mail: laylahara@yahoo.co.jp
I. Muraoka

314 R. Hara and I. Muraoka
25.1 Introduction
Open water swimming is defined as any swimming competition that takes place in
rivers, lakes, oceans, or water channels (Fédération Internationale de Natation
2013b). Since the inclusion of open water swimming as an Olympic event in
2008, it has steadily been gaining popularity worldwide; events have taken place
with over 25,000 participants (Tipton 2014).
The first open water swimming event was held more than 100 years ago. In this
sport, “ channel swim” is considered the most common ultra-endurance swimming
event. The “English channel swim” between England and France is one of the most
famous channel swims, and athletes are required to cover a distance of 34 km. The
first person officially swam the Channel in 1875 with a time of 21 h 45 min
(Eichenberger et al. 2012). Many swimmers have since challenged this channel
swim and improved the time. In 2013, Rebecca Lewis swam the channel in 9 h
2 min. Rules of the English Channel swim were set such that one boat with the
coach was required to be present with each swimmer. The swimmers can get food
and drink from the coach whenever required. Although the actual distance across
the channel is 34 km, many swimmers end up swimming distances of approxi-
mately 40–50 km because of the strong currents. The water temperature is approx-
imately 16 C, even in the summer. Swimmers have to swim in these severe
conditions, and are not allowed to wear protective or performance enhancing
swimsuits. Although many swimmers have attempted the channel swim in recent
years, the success rate remains at approximately at 60 %. Therefore, the channel
swim is difficult even to just to finish (The channel swim association 2013).
There are other open water swimming events organized by Fédération
Internationale de Natation (FINA). These events do not have such large currents
and are less difficult to finish, but the participants have to swim at higher swimming
speeds. Each year from 1997 to 2006, FINA organized the Open Water World Cup,
which included a whole range of long distance events. Since 2007, FINA decided to
have the World Cup focus only on the 10 km distance. This was a direct conse-
quence of the introduction of the 10 km open water swimming to the Olympic
programme in 2005 (Fédération Internationale de). Many elite athletes participated
in the World Cup as training event for the Olympics. This event requires that one
coach watch each swimmer as a safety precaution. If athletes do not have a coach
with them, they cannot compete in the race (Fédération Internationale de Natation
2012). Similar to the channel swim, the times for the 10 km events have been
improving over the years. At the World Championships in 2010, the 10 km men’s
champion finished the race in 2:00:59.3, while the 10 km women’s champion
finished in 2:05:45.2. At the 2013 World Championships, the men’s champion
finished in 1:49:11.8, while the women’s champion finished in 1:58:19.2. Had the
2010 champions swam at the 2013 World Championships and finished with the
2010 times, the male would have finished in 58th place and the female 44th place
(Fédération Internationale de Natation 2011, 2013a). Presently, coaches can give
food and sports drink to the swimmers at the feeding platforms. FINA has set rules
25 Open Water Swimming Performance 315
requiring that a floating or stationary feeding station be available at least every

2.5 km for sponsored competitions greater than 5 km (Fédération Internationale de
Natation 2012). In addition, the swimmers can eat and drink not only at the feeding
areas but also anywhere along the course. If they want to eat or drink outside the
feeding area, they simply swim with the food placed underneath their swimsuits.
Channel swimming has a long history, and consequently many studies have
focuesed on the channel swim (Dwyer 1983; Eichenberger et al. 2012; Finlay
et al. 1995, 1996; Karamouzis et al. 2002; Knechtle et al. 2010; Thomas
et al. 2000; Wagner et al. 2012). Most of the studies have focused on the risk of
open water swimming because of the number of swimmers who are not able to
finish the race owing to medical problems. Indeed, we lost a famouse athlete in the
FINA 10 km World Cup. This is a clear indication that the participants in such races
need adequate information on the risks of open water swimming. A few studies
have also focused on the performance aspects of the 10 km World Cup, but these do
not contain adequate information in order for coaches and athletes to properly train
for the Olympic Games (Castro et al. 2009; Macaluso et al. 2013; VanHeest
et al. 2004). Coaches and athletes are therefore searching for the best training
methods for open water swimming, but the proper scientific information does not
exist. Before proceeding to review the different aspects of open water swimming
for the Olympics, we will cover what is known about the topic. The purpose of this
chapter is to review and shed light on the risks and performance factors of open
water swimming in order to improve upon the performance and safety factors of
open water events worldwide.
25.2 Risk of Open Water Swimming
Hypothermia and dehydration are the most prevalent medical risks for open water
swimming (Castro et al. 2009; Gerrard 1999). To reduce the frequency of hypo-
thermia, FINA has set rules on the minimum temperature of water allowed during
the race. The water temperature should be a minimum of 16 C, and it should be
checked the day of the race, 2 h before the start, and during the race itself
(Fédération Internationale de Natation 2012). Although, many studies have focused
on hypothermia, we will focus on heat stroke. During the 2010 FINA 10 km open
water World Cup in Fujaurah (Dubai), an elite United States athlete died. Although
the results of the autopsy have not been revealed, concerns have been raised about
the high water temperatures. A FINA official claimed that the water temperature
exceeded 31 C near the finishing time of that race (Macaluso et al. 2013). After this
incident, some researchers focused on heat stroke, and FINA set a new rule: the
water temperature allowed would be a maximum of 31 C (Fédération
Internationale de Natation 2013b). Additionally to reduce the frequency of dehy-
dration, FINA has set another rule for the 10 km World Cup that feeding stations
must be at least 2.5 km apart (Fédération Internationale de Natation 2012).
25.2.1 Hypothermia
Many studies have reported on hypothermia during open water swimming events,
mainly focusing on the channel swims (Dwyer 1983; Eichenberger et al. 2012;
Finlay et al. 1995, 1996; Karamouzis et al. 2002; Knechtle et al. 2010; Thomas
et al. 2000; Wagner et al. 2012). The study by Castro et al. (2009) is the only study
involving experiments performed at the FINA 10 km World Cup (Castro
et al. 2009). Brazilian athletes, who are some of the top open water swimmers in
recent years, participated in this experiment. The water temperature was 21 C,
which is considered an ordinary open water swimming event temperature. Most
athletes finished the race with mild (34–35 C; n ¼ 3) to moderate (30–34 C; n ¼ 7)
hypothermia. These results indicate that even elite athletes have a risk of hypother-
mia that is dependent upon water temperature. This study also mentioned that the
difference between pre and post competition core-temperatures was greater in
female athletes than in male athletes.
Another study focused on afterdrop (Thomas et al. 2000). Afterdrop is a phe-
nomenon described as continued core cooling following removal from cold stress.
The distance of the experimental race was approximately 3 km, and the water
tempreture was 11.7 C. Core temperatures were recorded before and after (every
3 min for 45 min after the race) the open water swimming event. Afterdrop was
noted in 10 of the 11 subjects, even though they were dried off and wore hospital
gowns in a heated room. One participant had to stop the experiment because the
core temperature did not stop decreasing and went below 35 C.
Given the results of several previous studies, it can be concluded that open water
swimming poses a risk of hypothermia and afterdrop. Howebver, under the FINA
rules, the full wet suits which surfers and triathletes wear, are still not allowed. The
only allowed solutions to prevent hypothermia are to use grease or increase body fat
mass, especially for women. Adequate preparation for medical care must be made
in order to attend to emergency cases.
25.2.2 Heat Stroke
We know that open water swimming has the risk of hypothermia. However, one
must not forget the famous athlete who died in warm water temperatures during
open water swimming. Heat stroke occurs when the core body temperature rises
above 40 C, and it is caused by the failure of the brain’s temperature regulatory
center in the hypothalamus. And heat stroke is also leading cause of exercise-
related mortality. Heat stroke occur when the athlete’s rate of heat production
exceeds the rate at which the excess heat produced during exercise can be dissipated
into the environment (Noakes 1995). During swimming when the water tempreture
exceeds the tempreture of the skin, swimmers will experience heat gain and became
heat stroke. This hyperthermic condition can result in death, with 20 % of reported
cases being the result of heat failure or cerebral edema (Brooks et al. 2000).
Research results are in agreement that there is a risk of hypothermia in open
water swimming. However, almost all of the studies were conducted with the water
temperature below 23 C. In open water swimming, especially in the 10 km World
Cup and Olympics, such cold conditions may not always be present. One previous
study (Macaluso et al. 2011) reported the effect of three different water tempera-
tures on core temperature during a 5 km swim in a swimming pool. These results
may currently be the best information for coaches. In this study, the subjects swam
5 km with water temperatures of 23 C, 27 C and 32 C. During the 23 C swim,
pre and post core temperature did not change. On the other hand, during the 27 and
32 C swims, the post swimming temperature was significantly higher than the pre
temperature. Another study reported that swimming in 27–33 C water for only
20 min can be enough to cause heat stroke (Costill et al. 1967). These results, taken
together, suggest that open water swimming in warmer temperatures poses a risk of
heat stroke (Macaluso et al. 2013). Previous reviews on heat stroke suggest the limit
of 33 C as a possible maximum water temperature for open water swimming
events. After the publication of this review, in 2013, FINA consequently set a new
rule that the maximum temperature allowed to be 31 C (Fédération Internationale
de Natation 2013b). However, some studies reported that even at 27 C, some
swimmers experienced heat stroke. Special attention should be paid to the water
temperature, and possible modifications should be made on the rule.
25.2.3 Dehydration (Hypernatremia and Hyponatremia)
Some case of dehydration are diagnosed as hyponatremia, and others are diagnosed
as hypernatremia. Hyponatremia occurs when the sodium level in the blood is low
or the person is in a state of fluid overload. Hypernatremia is almost always an
indication of excessive fluid depletion. It usually occurs with inadequate fluid
intake and increased water loss.
Hyponatremia occurs due to three main factors: (a) overdrinking due to biolog-
ical or psychological factors; (b) inappropriate antidiuretic hormone (ADH) secre-
tion, in particular the failure to suppress ADH-secretion in the face of an increased
total water in the body; and (c) a failure to mobilize Na from the osmotically
inactive sodium stores or inappropriate osmotic inactivation of circulating Na
(Noakes 2011). The main reason for developing hyponatremia involves a behav-
ioral response, such as overdrinking during an endurance performance (Noakes
2011; Noakes et al. 2005).
Sandra (2012) investigated the prevalence of hyponatremia in 25 male and
11 female ultra-endurance swimmers participating in open water events, covering
a total distance of 26.4 km (Wagner et al. 2012). The swimmers were allowed to eat
anything in this event. This study reported than no athlete showed signs of
hypernatremia, while 6 people suffered from hyponatremia. It was also found that
women have a higher risk for hyponatremia as compared with men. These results
might be because women have smaller bodies than men. If women ingest the same
amount of fluid as men, the women would tend to experience more hyponatremia
than men.
On the other hand, Macaluso (2011) reported that open water swimming poses a
risk of hypernatremia (Macaluso et al. 2011). They investigated the prevalence of
hypernatremia in 9 open water male athletes participating in their experiment of
swimming for 5 km in an indoor pool without drinking. This study concluded that
these athletes experienced hypernatremia because they did not have an adequate
fluid intake.
These studies suggest that adequate feeding during the various open water
swimming events is needed.
Completing a channel swim is a difficult goal to attain. On the other hand, the
focus of World Cup swimmers and Olympians is how to swim faster during the
race. Many athletes spill some of their drinking supply by too hurriedly drinking
during the event. Indeed, coaches now instruct swimmers to take more time and
drink enough to get an adequate replenishment of fluid and energy. It is the author’s
opinion that in the World Cup Competition, swimmers do not overdrink. Therefore,
they tend to experience hypernatremia and not hyponatremia. There is not enough
information on dehydration during the 10 km open water swimming event. There-
fore, more investigations are needed in order to determine the optimal food and
fluid intake levels for open water swimmers.
25.3 Performance of the Open Water Swimmer
Several studies have focused on the risks of open water swimming events. On the
other hand, competitive swimming and other swimming events have been the
subject of a number of studies. Olympians want to obtain more scientific informa-
tion for training purposes. Some studies have focused on the performance of the
10 km open water swimmers and not the channel swimmers, and thus will be of
more help.
25.3.1 Anthropometric Data, Anaerobic Threshold Data,

Blood Chemistry, and Hematology Data
Vanheest et al. (2004) reported on the characteristics of elite open water swimmers
(VanHeest et al. 2004). Four male and four female athletes participated in this
study. All subjects were considered to be among the top 6 athletes in the United
States. Their anthropometric profile, lactate threshold data, blood chemistry, and
hematology data were collected and compared with those of pool swimmers who
participated in the World Championships. The open water swimmers were shorter
and lighter than the pool swimmers. The percent muscle mass of open water
swimmers compared to pool swimmers was 6.2 % and 11.9 % lower for males
and females, respectively. Distance swimming dose not require maximul power
outputs typical of those seen in many pool events. So the athlete’s size may be
benefit to their swimming performance and allow them to endure long distance
events more readily (VanHeest et al. 2004). However, the lower body fat percent-
age might place the athlete at risk for thermal stress-related events. Countermea-
sures for hypothermia must be worked out. A lot of channel swimmers used
vaseline to cover their whole body for preventing hypothermia. So open water
swimmers and coaches need to check the water tempreture and to take measures to
meet various situation.
The anaerobic threshold data were collected and were compared to those of other
long distance sports athletes. Using the anaerobic threshold data, marathon runners
had values of Lactate Threshold (LT) at 76–83 % of the velocity peak for males and
83 % for females. That for male open water swimmers is reported to be
88.75 3.1 %, which is similar to other long distance sports athletes. On other
hand, women have a higher pace associated with LT relative to peak pace
(93.75 1.5 %). The data supported the enhanced ability of female swimmers to
maintain a high percentage of peak velocity during swimming. These results
suggested that open water swimmers need a high anaerobic capacity. Although
the characteristics of open water swimmers are known, anaerobic threshold did not
compare with competitive swimmers. In recent years, many open water swimmers
have switched to competitive swimming. The different abilities of open water
swimmers as compared to competitive swimmers must be investigated tob etter
understand the proper training methods for each.
Blood data on the open water swimmers included levels of glucose, urea,
creatinine, cholesterol, triglyceride, ferritin, iron, CPK, AST, LDH, and Cortisol.
Blood parameters were affected by factors such as the intensity of training, diet,
stress, and the amount of training fatigue. In comparison to competitive swimmers,
muscle damage and many other parameters were in the normal range for swimmers
under anaerobic training. Iron status markers were lower than the normal range for
female athletes. Many United States female competitive swimmers are also found
to also have a low iron status. Poor iron status comes from fatigue and can be
avoided by maintaining a good diet. Female athletes have to train with proper
attention to their diet.
Although many characteristics of open water swimmers can be garnered through
this study, the pace of progress in the sport is rapid, and more studies are needed to
develop optimal training methods. Additionally, data from such studies should be
compared to that obtained for other sports. New training methods and the physio-
logical responses of athletes of many sports.
25.3.2 Core Temperature
Castro et al. (2009) reported data for body temperature, capillary glycemia and race
times in the 10 km World Cup (Castro et al. 2009). A total of 7 men and 5 women
participated in this study. On the day before the competition, their core tempera-
tures were measured immediately before warming up and within 1 min after the
athletes got out of the water. After the race, they immediately measured the same
parameters. It was found that all competitors had significantly decreased core
temparetures. The temperature changes were 2.7 1.8 C for men, and
4.2 0.9 C for women. The athletes’ percentages of body fat were very similar
to American open-water swimmers, but no relationship between tbody composition
and performance was established. At least for the male athletes, race times were
inversely correlated to pre competition core temperatures. Most coaches advise
open water swimmers to have an increased percentage of body fat as compared with
competitive swimmers to prevent decreases in their core temperatures.
25.3.3 Capillary Glycemia
Capillary glycemia was measured immediately before warming up and within

3 min after the athletes got out of the water (Castro et al. 2009). Capillary glycemia
was affected by food intake. In this study, one athlete ate a large breakfast and
received no supplements during the race. Other athletes ate an ordinary breakfast
and received various loads of maltodexrin during the race. For the athlete who did
not get the supplements during the race, capillary glycemia was 161 mg/dL before
the competition and fell to 75 mg/dL at the end of the race. Another member had
increased glycemia. In the women, the change of glycemia and race time were
inversely related to time to complete the race. These results indicate that larger
increases in glycemia were related to better performance, with the athletes who had
a fall in their glycemia levels finishing last among the group. The athletes did not
recall amount of food intake they received during the race. Bacause of this, a
conclusion cannot be made as to whether high food intake during the race is good
for performance or not. Common sense suggests that good athletes with high
capillary glycemia could readily sprint. Athletes who finished last probably could
not sprint at the final stretch of the race. A film review of the races, would could add
interesting information to the study. The one athlete who did not eat during the race
suggested that this is not a good strategy to take. In open water swimmming events,
if one were to just have a large breakfast, capillary glycemia will fall and there may
not be enough energy for open water swimming. Recently, all athletes eat breakfast
and also eat during the race. However, the best type and amount of food intake at
the feeding stations remain unknown.
25.4 Conclusion
Previous studies have focused on the risks involved in open water swimming
events. Almost all studies reported that open water swimming poses a risk of
hypothermia and afterdrop. Hypothermia occurs in races in cold water, and can
lead to death. Therefore, medical care needs to be available to quickly respond to
emergencies. In warm water there is the risk of heat stroke. FINA has set the rules
on water temperatures for the race day. Some studies however, report that some
swimmers may still experience heat stroke even if they compete within the FINA
temperature guidelines. Further investigations are needed on the optimal type and
amount of food intake during the race. The United States and Brazil performed
experiments on their top athletes to investigate the physiological characteristics of
open water swimmers. Pre core temperature and capillary glycemia were related to
their race times. Since the data for these studies were collected around 2004
(VanHeest et al. 2004), up to date investigations of open water swimmers are
needed to verify the factors needed for an optimal, safe performance in this sport.
References
Brooks GA, Fahey TD, White TP, Baldwin KM (2000) Textbook of exercise physiology, human
bioenergetics and its applications, 3rd edn. Mayfield Publishing Company, Mountain View
Castro RT, Mendes FS, Nobrega AC (2009) Risk of hypothermia in a new Olympic event: the
10km marathon swim. Clinics 64(4):351–356
Costill DL, Cahill PJ, Eddy D (1967) Metabolic responses to submaximal exercise in three water
temperatures. J Appl Physiol 22(4):628–632
Fédération Internationale de Natation (2011) FINA world championships results Shanghai’11.
http://www.fina.org/H2O/index.php?option¼com_content&view¼category&id¼271&
Itemid¼786
Fédération Internationale de Natation (2012) Rules and regulations. FINA 10km marathon swim-
ming world cup. 2011–2013. http://www.fina.org/H2O/docs/events/10km2013/rules.pdf
Fédération Internationale de Natation (2013a) 15th FINA world championships start & results.
http://www.fina.org/H2O/index.php?option¼com_wrapper&view¼wrapper&Itemid¼1507
Fédération Internationale de Natation (2013b) Federation Internationale de Natation. FINA hand-
book – Constitution and rules 2013–2017. FINA Office, Lausanne, pp 204–212
Dwyer J (1983) Marathon swimmers: physiologic characteristics. J Sports Med Phys Fitness
23(3):263–272
Eichenberger E, Knechtle B, Knechtle P, Aleander RC, Rosemann T, Lepers R (2012) Best
performances by men and women open water swimmers during the ‘English Channel Swim’
from 1900 to 2010. J Sports Sci 30(12):1295–1301
Fédération Internationale de Natation. FINA 10km marathon swimming world cup. http://www.
fina.org/H2O/index.php?option¼com_content&view¼category&layout¼blog&id¼51&
Itemid¼297
Finlay JB, Hartman AF, Weir RC (1995) Post-swim orthostatic intolerance in a marathon
swimmer. Med Sci Sports Exerc 27(9):1231–1237
Finlay JB, Lysynski B, Liggins AB, Weir RC (1996) Monitoring rectal core temperature in
marathon swimming. Med Biol Eng Comput 34(1):97–100
Gerrard DF (1999) Open water swimming. Particular medical problems. Clin Sports Med 18(2):
337–347
Karamouzis L, Karamouzis M, Vrabas LS, Christoulas K, Kyriazis N, Giannouis E, Mandroukas K
(2002) The effects of marathon swimming on serum leptin and plasma neuro peptide Y levels.
Clin Chem Lab Med 40(2):132–136
Knechtle B, Baumann B, Knechtle P (2010) Speed during training and anthropometric measures in
relation to race performance by male and female open water ultra endurance swimmers.
Percept Mot Skills 111(2):463–474
Macaluso F, Di Felice V, Boscanio G, Bonsignore G, Stampone T, Farina F, Morici G (2011)
Effects of three different water temperatures on dehydration in competitive swimmers. Sci
Sports 26(5):265–271
Macaluso F, Barone R, Isaacs AW, Farina F, Morici G, Felice VD (2013) Heat stroke risk for open
water swimmers during long distance events. Wilderness Environ Med 24(4):362–365
Noakes TD (1995) Dehydration during exercise: what are the real danger? Clin J Sport Med
5:123–128
Noakes TD (2011) Changes in body mass alone explain almost all of the variance in the serum
sodium concentrations during prolonged exercise. Has commercial influence impeded scien-
tific endeavor? Br J Sports Med 45(6):475–477
Noakes TD, Sharwood K, Speedy D, Hew T, Reid S, Dugas J, Almond C, Wharam P, Weschler L
(2005) Three independent biological mechanisms cause exercise-associated hyponatremia:
evidence from 2,135 weighed competitive athletic performances. Proc Natl Acad Sci USA
102(51):18550–18555
Sandra W, Beat K, Patrizia K, Christoph AR, Thomas R (2012) Higher prevalence of exercise-
associated hyponatremia in female than in male open-water ultra-endurance swimmers: the
‘Marathon-Swim’ in Lake Zurich. Eur J Appl Physiol 112(3):1095–1106
The Channel Swim Association (2013) 2013 Trophy winners. http://www.
channelswimmingassociation.com/swim-results
Thomas J, Nuckton MD, David M, Claman MD, Goldreich D, Frederick C, Wendt MD, Nuckton
JD (2000) Hypothermia and afterdrop following open water swimming: the Alcatarz/San
Francisco swim study. Am J Emerg Med 18(6):703–707
Tipton MJ (2014) Sudden cardiac death during open water swimming. Br J Sports Med 48(15):
1134–1135
VanHeest JL, Mahoney CE, Herr L (2004) Characteristics of elite open water swimmers. J
Strength Cond Res 18(2):302–305
Wagner S, Knechtle B, Knechtle P, Alexander RC, Rosemann T (2012) Higher prevalence of
exercise-associated hypothermia in female than in male open-water ultra-endurance swim-
mers: the ‘marathon – swim’ in Lake Zurich. Eur J Appl Physiol 112(3):1095–1106
Chapter 26
The Spin on a Baseball for Eight Different
Pitches Thrown by an Elite Professional
Pitcher
Tomoyuki Nagami, Takatoshi Higuchi, and Kazuyuki Kanosue
Abstract We analyzed the spin on a baseball pitched by a professional pitcher, who

utilized eight distinct, different types of pitches in official league games. In the
experiments he pitched each of the eight pitches twice from an official pitching
mound to a catcher. The video image of the pitched baseball from just before until
500 milliseconds after ball release was taken with a high-speed video camera at a rate
of 1,000 frames per second. A special apparatus was used to obtain the direction of
spin axis and the spin rate. Both values varied considerably depending on the type of
pitch. In the four kinds of pitches which were thrown with a grip similar to that of the
four-seam fastball, ball spin was altered by changing the direction of spin axis and/or
orientation of the seams. The other three kinds of pitches thrown with a grip different
from that of the four-seam fastball. These pitches had a slower speed and/or slower
spin rate than the four-seam fastball. Moreover, the subject was able to throw these
different pitches without changing the throwing motion as viewed by a batter.
Mastery of these skills is what made him an elite pitcher with a long career.
Keywords Fastball • Curve ball • Forkball • Spin axis • Spin rate
26.1 Introduction
Baseball pitchers throw a ball toward a catcher to get batters out. To accomplish
this, they throw various kinds of pitches, such as fastballs, curveballs, sliders,
forkballs, and others. These pitches travel at different speeds and trajectories.

T. Nagami (*) • T. Higuchi • K. Kanosue
e-mail: t-nagami@aoni.waseda.jp

324 T. Nagami et al.
A baseball in flight is influenced by three forces: (1) gravity that pulls down the ball,
(2) air resistance or drag operating in the opposite direction of the ball’s transla-
tional movement and, (3) if it is spinning, a lift force perpendicular to the direction
of the translational movement (Bahill and Baldwin 2007). Many studies report that
the amount of lift force on the spinning baseball depends on the spin parameter (the
ratio of the ball spin rate and the ball speed) and the direction of the spin axis
(Alaways and Hubbard 2001; Briggs 1959; Nathan 2007; Watts and Ferrer 1987),
and that movement direction also depends on the direction of the spin axis (Bahill
and Baldwin 2007; Jinji and Sakurai 2006). It is presumable that not only ball
speed, but also these other factors differ among the various types of pitches.
Jinji and Sakurai (2006) made a detailed analysis of the spin rate and spin axis of
fastballs and curveballs thrown by actual pitchers. We, also, analyzed the spin on
fastballs thrown by elite professional and collegiate pitchers (Nagami et al. 2001).
However, there have been few studies that have analyzed the ball spins of pitches
other than fastballs and curveballs, especially those of elite pitchers who throw a
variety of pitches. Moreover, how they throw the various types of pitches is unknown.
In the present chapter, we report on an investigation of pitches thrown by a
professional pitcher. His fastball was not especially fast (140 km/h or less), but due
to his excellent control over eight different pitches he was able to remain active as a
professional player for a long time (his nickname was “Mr. Control”). Thus, the
purpose of this study was to analyze the spin rate and direction of spin axis for those
eight different pitches thrown by this elite pitcher. We show that each type of pitch
has a unique combination of velocity, spin rate, and direction of spin axis, and the
throwing kinematics of those eight different pitches is discussed based on the
outcome of the experiment.
26.2 Methods
26.2.1 Subject
The subject was a right-handed professional pitcher (183 cm, 86 kg, 42 years). Over
the span of his career he played 19 seasons of professional baseball, 18 seasons in
Japan and 1 season in the Major Leagues in the United States. His career record was
117 wins and 144 losses with an ERA of 3.74.
26.2.2 Experimental Design and Procedure
The experiments were conducted at Waseda University, Tokorozawa campus

utilizing a specially designed site which included a pitcher’s mound and home
plate constructed according to official dimensions. The site also contained the
26 The Spin on a Baseball for Eight Different Pitches Thrown by an Elite. . . 325
necessary background for video analysis. The experimental protocol was approved
by the Human Research Ethics Committee of the Faculty of Sport Sciences,
Waseda University. Informed, written consent was obtained from the subject.
The subject threw an official baseball (Mizuno) to a catcher squatting in the
receiving position behind home plate, which was 18.44 m away from the pitcher’s
rubber. He threw eight different, self-enumerated pitches (four-seam fastball,
two-seam fastball, cut fastball, slider, curveball, change-up, forkball, and “shake”),
two strikes for each one. The “shake” is an original pitch created by the subject. Ball
speed was measured with a radar gun (JUGS, USA) positioned behind the catcher.
The letters E, M, and I were marked on each baseball and made it possible to
utilize a feature point detection analysis to quantify ball spin. A high-speed video
camera (frame frequency ¼ 1,000 frame/s (fps), shutter speed ¼ 1/10,000 s, resolu-
tion of 640 * 480 pixels, Fastec Imaging) was set 3 m behind the subject and at the
height of ball release. Ball images were taken during the period from just before
until 500 ms (ms) after release (See Extra).
26.2.3 Analysis of Spin Axis and Spin Rate
Direction of the spin axis was obtained with the same method as reported in a
previous study (Nagami et al. 2001). In this method, a custom-made apparatus
made of aluminum frames (Tri-axis Feature, Furusawa Lab, Japan, Fig. 26.1) was
used. The same ball that had been used in a particular pitch was mounted in the
center of the apparatus. The ball could be rotated on axis A which could then be
tilted around axis B and axis C so that ball spin with any orientation in three
dimensional space was able to be reproduced (Fig. 26.1). The rotation angle around
each axis was measured on a protractor attached in the proper orientation to
evaluate the angle of that axis. The ball image taken in the experiment was
displayed on a video monitor in which the direction of the ball’s translational
movement corresponded to the direction of depth. A video camera was set above
the apparatus, and the image of the camera was displayed on another monitor. The
experimenter compared the marks on the balls in the two video images side-by-side,
and adjusted the relative setting of the ball and direction of the two axes (B and C)
of the apparatus so that the two images coincided. The direction of the spin axis was
obtained when the whole set of experimentally obtained ball images in one rotation
could be matched with the ball images on the apparatus only by rotating the ball
around axis A. The direction of the spin axis was defined by the right-hand rule
(Bahill and Baldwin 2007). We also defined a ball coordinate system. The origin of
the system was located at the center of the ball, the X-axis was directed from the
pitcher’s plate to third base, the Y-axis was directed to home plate, and the Z-axis
indicated a vertical (upward) direction. The orientation of the ball spin axis was
expressed utilizing two parameters, elevation Φ (the angle between the spin axis
and X-Y plane) and azimuth θ (the angle between the X axis and the projection of
the spin axis on the X-Y plane) (Fig. 26.2).
Fig. 26.1 Special apparatus to analyze the ball spin
Fig. 26.2 Definition of the

angles of elevation Φ
and azimuth θ
Ball spin rate just after ball release was calculated from the number of video
frames necessary for one rotation of the ball (half rotation for the “shake”): 1,000/
number of frames (rotations per second, rps). The direction of the spin axis and the
spin rate for each type of pitch were reported as the mean of two pitches.
26.3 Results
Table 26.1 indicates the two components (angles) of the axes together with spin rate
and ball speed of eight different pitches. The spin of the “shake” was so slow that
the direction of spin axis could not be determined. Direction of spin axis and spin
rate both changed considerably depending upon the type of pitch. Figure 26.3
shows changes in the images of ball markers over time. The spin axes are depicted
in the rightmost portions of Fig. 26.3
The four-seam and two-seam fastballs were similar in direction of the spin axis,
spin rate, and ball speed. The only difference was seam orientation, apparently
4 seams for the four-seam fastball and 2 seams for the two-seam fastball. The slider
was unique in that the direction of axis was almost identical with the direction of
trajectory, which was 80 in the X-Z plane. The angle of θ was bigger than 90 ,
indicating a “top spin” only for the curveball. The slider and curveball had a
relatively higher spin rate relative to other pitches. The cut fastball showed the
spin rate and the direction of spin axis just between those of the four-seam and the
slider. Although the direction of spin axis of the change-up was similar to that of the
cut fastball, the spin rate and the ball speed were slower than the cut fastball. The
direction of spin axis of the forkball was similar to that of the slider, and the spin
rate and the ball speed were slower than it. The spin rate and the ball speed of the
“shake” were the slowest among the eight pitches.
26.4 Discussion
In the present study, by analyzing high-speed video images of a spinning ball, we

were able to obtain the spin rate and the direction of spin axis of a baseball thrown
by a professional pitcher. This was accomplished for eight different, distinct types
of pitches. The eight pitch types differed in direction of spin axis as well as in spin
rate. Since the subject had excellent control of all his pitches, he had no problem
throwing the ball twice within the strike zone, without mistake, for all the eight
pitch types. In the present study, each type of pitch was thrown twice. The
difference in θ and Φ between the two pitches were less than 10 , mostly less
than 3 except for the θs of the cut fastball and curveball. Likewise, the differences
in spin rate and ball speed were less than 2 rps and 2 m/s, respectively. These results
indicate a high level of reproducibility for all the pitches. This indicated very good
328
Table 26.1 Direction of spin axis, spin rate, and ball speed for eight different pitches
Direction of spin axis
θ Φ Spin rate Ball speed
Pitches degree degree rps m/s (km/h)
Four-seam fastball 29 (30, 28) 38 (39, 37) 33.9 (33.3, 34.5) 37.4 (37.2, 37.5) (134, 135)
Two-seam fastball 34 (35, 33) 40 (38, 41) 31.3 (31.3, 31.3) 36.4 (36.4, 36.4) (131, 131)
Cut fastball 52 (56, 47) 22 (22, 22) 35.7 (35.7, 35.7) 35.7 (35.6, 35.8) (128, 129)
Slider 80 (79, 80) 4 (5, 3) 38.5 (38.5, 38.5) 34.7 (34.7, 34.7) (125, 125)
Curveball 142 (139, 145) 41 (40, 42) 44.5 (45.5, 43.5) 29.2 (29.7, 28.6) (107, 103)
Change-up 58 (58, 58) 18 (18, 19) 28.2 (28.6, 27.8) 29.6 (28.6, 30.6) (103, 110)
Forkball 84 (85, 82) 12 (11, 13) 8.3 (8.7, 7.9) 32.8 (32.8, 32.8) (118, 118)
“Shake” – – 2.6 (3.6, 1.7) 21.5 (21.4, 21.7) (77, 78)
rps rotation/s
T. Nagami et al.
Fig. 26.3 The spin of eight different pitches. Each row contains images just after ball release at
2 ms intervals. The rightmost figures indicate the calculated spin axes
ball control for the subject. His ability to utilize all eight pitches with excellent
control allowed him to play in the professional leagues well into his 40s.
26.4.1 Characteristics of Each Pitch
Fastballs, both four-seam and two-seam, are characterized by backspin. This pro-
vides an upward lift force that makes the amount of drop smaller as compared to
that of a freely falling ball (Adair 2002). At a given spin rate, the upward lift force is
maximal when the spin axis lies horizontal and perpendicular to the direction of ball
movement. In the present study, the axis of fastballs was tilted at 38 from
horizontal (Φ ¼ 38 ). This may not be specific to the present subject, and all
11 professional and 11 collegiate pitchers whom we analyzed threw fastballs that
tilted more than 20 from the horizontal (Nagami et al. 2001). The elevation Φ
probably depends on pitching style (overarm, three-quarter arm, sidearm, under-
hand arm, and so on), or on throwing arm slot angle in the X-Z plane. Of course,
other factors such as the wrist angle at ball release or a difference in hand shape,
e.g. lengths of the index and middle fingers, would also influence the degree of tilt.
Direction of the spin axis of the four-seam and two-seam fastballs was similar.
However, orientation of the seams was different; with one rotation, the seams cross
a given point four times for the four-seam fastball, while they cross only twice for
the two-seam fastball. This would produce a difference in their trajectories
(Alaways and Hubbard 2001). The present subject said that he threw the
two-seam fastball when he had a feeling that a batter would “swing”. Thus, the
ball would have a speed and spin similar to those of the four-seam fastball, but the
slight difference in trajectory, while not being very different from a four-seam
fastball, would be different enough to lead to a miss-hit if the batter did decide to
swing.
Interestingly, the “slider”, as termed by the subject, had a spin axis in line with
the direction of the throw, similar to the spin of a thrown American football. This
produces a bullet-like spin, which provides no lift and the ball just drops with
gravity. To the batter, this ball looks as if it has a sudden drop. A pitch with this type
of spin has been termed a “gyroball” and it was hotly debated whether or not
Daisuke Matsuzaka, the Boston Redsox pitcher, threw it when he debuted in
American Major League Baseball in 2007. At present, most such debates are largely
based on guesswork, since a direct, accurate analysis of ball spin has not been done.
According to the present subject, he had already been throwing this type of pitch
long before he heard of the name “gyroball”. Very likely, many pitchers have been
throwing this same pitch for a long time, but with different names.
The cut fastball was intermediate in the direction of spin axis, ball spin rate and
ball speed between the four-seam fastball and the slider. This spin would produce
an intermediate flight trajectory between them. The subject would try to get the
hitters out by achieving a flight trajectory with a slight difference from other pitches
so that hitter would not notice the difference during early ball flight.
The spin rate of the curveball (41.8 rps) of the present subject was much greater
than the value obtained for college baseball pitchers in a previous study (31.0 rps,
n ¼ 9) (Jinji and Sakurai 2006). Thus, the curveball of the present subject would
have had a better “break” than the others. Furthermore, the curveball was charac-
terized by its topspin. This provides a larger drop as compared to a freely falling
ball as well as to the other types of pitches. From observation of the video (See
Extra), the curveball was launched in a more upward direction than the other
pitches. This adjustment would be necessary to get the curveball to arrive in the
strike zone.
The forkball also dropped with gravity, because it had only a slight lifting force
due to its slow spin rate. In the present study, the spin rate of the forkball was only
8 rps, one third of the spin rate of the four-seam fastball. On the other hand, ball
speed did not differ much, 32 m/s for the forkball and 36 m/s for the fastball, which
would make the batter’s task of discriminating between the two pitch types difficult.
The change-up pitch is generally used to confuse a batter’s timing. In this
experiment, the change-up was the slowest, next to the “shake” and the curveball,
and had a similar direction of spin axis as the cut fastball. So the flight trajectory of
the change-up would be different from the types of pitches launched in an upward
direction, like the curveball, but similar to the four-seam fastball’s flight trajectory.
The “shake” ball had an even slower rate of spin, 3.6 and 1.7 rps, less than half
that of the forkball. This is a unique pitch that was developed by the present subject.
He throws the ball with the same grip as the forkball, but with a slower pitching
motion, just like when playing catch. Ball speed was very slow (21 m/s, only two
thirds of that of the four-seam fastball). Thus, a batter knows that the “shake” is
being thrown even before ball is actually released. Nevertheless, this pitch can be
useful, because the ball has an unpredictable movement. The “shake” is very
similar to the “knuckleball” in that it flutters due to side forces on the ball, because
its slow spin rate produces changes in the flow separation position by a turbulent
boundary layer triggered by the position of ball seams acting as a surface roughness
(Mizota et al. 1995a, b). This brings about drastic changes in the direction of air
velocity behind the ball and thus produces the side forces. A knuckleball is gripped
against the fingertips, fingernails or knuckles, and apparently different from the
“shake”. The slow or zero spin rate produces a similar effect of fluttering, regardless
how it is thrown. Therefore, the “shake” is aerodynamically identical to a knuck-
leball. According to numerical simulation, a ball has to have a spin rate less than
2 rps to produce a fluttering movement great enough to dazzle batters (Mizota and
Kawamura 2007). The spin rate obtained from two “shake”s in the present study,
1.7 and 3.6 rps, is just around the limit of the effective spin rate. Actually, the
present subject said that the “shake” sometimes did not flutter, probably because the
spin rate was not low enough, and thus, was very risky to throw in a close game.
As was shown for the “knuckle” and “shake”, as well as the “slider” and “gyro-
ball”, pitches thrown by different pitchers, and given different names, might have
the same physical characteristics or, conversely, two pitches with same name
thrown by different pitchers might have dissimilar physical characteristics if the
pitched balls had different spins.
26.4.2 Pitching Grips and the Ball Spin
We now discuss how various kinds of pitches are thrown based on an analysis of the
videos and ball spin. We compared the eight postures of the hand and fingers
viewed from the second base side (the top of Fig. 26.4), and of the throwing motion
as viewed by a batter (the bottom of Fig. 26.4) just before ball release. Compared to
the four-seam fastball, the most common pitch, the grip of the change-up, forkball,
and “shake” were clearly different. It is natural to assume that the pitchers change
the grip to rotate the ball with a different spin. In addition, these three types of
Fig. 26.4 Comparison of grips and the whole body posture for of each pitch
pitches were thrown at a slower ball speed and/or lower spin rate as compared to the
four-seam fastball. Changing the pitching grip is probably effective for decreasing
the ball spin rate and/or the ball speed, and for making the flight trajectory close to
that of a freely falling ball.
On the other hand, the grips of the other four pitch types (two-seam fastball, cut
fastball, slider, and curveball) were almost identical to the grip of the four-seam
fastball. For these pitches, both the index finger and middle finger were put on the
top part of the ball and both the thumb and annular finger were put on the bottom
part of the ball. Although the spin rate of these pitches was similar to that of the
four-seam fastball, the ball spin axes were definitely different. Based on these
features it seems that the subject adjusts the direction of ball movement without
changing the grip by altering seam orientation and/or the direction of spin axis.
Moreover, excluding the “shake”, the throwing motion just before release were
nearly-identical for each of the pitches (the bottom of Fig. 26.4). It would thus be
impossible for hitters to detect the pitch type before and during release. Thus, elite
pitchers seem to possess the ability to throw pitches with various types of spin axis
and spin rate without noticeable changes in their posture.
26.5 Conclusion
We investigated the spin rate, velocity, and orientation of spin axis for eight
different pitches thrown by an elite pitchers. From this experiment, we found that
(1) each type of pitch had a unique combination of velocity, spin rate, and direction
of spin axis, (2) changing the pitching grip was effective for decreasing the ball
speed and/or ball spin rate, (3) even without changing the pitching grip, flight
trajectory could be adjusted by altering the direction of the spin axis, and (4) this
elite pitcher could throw many different pitches without changing the throwing
motion as viewed by a batter. These skills constitute what made him an elite pitcher
with a long career.
References
Adair RK (2002) Spin and the magnus coefficient. In: The physics of the baseball, 3rd edn. Harper
Collins Publishers Inc, New York, Chapter 2
Alaways LW, Hubbard M (2001) Experimental determination of baseball spin and lift. J Sports Sci
19:349–358
Bahill AT, Baldwin DG (2007) Describing baseball pitch movement with right-hand rules.
Comput Biol Med 37:1001–1008
Briggs LJ (1959) Effect of spin and speed on the lateral deflection (curve) of a baseball; and the
Magnus effect for smooth spheres. Am J Phys 27:589–596
Jinji T, Sakurai S (2006) Direction of spin axis and spin rate of the pitched baseball. Sports
Biomech 5(2):197–214
Mizota T, Kawamura Y (2007) 3D-trajectory analysis of side-spin knuckle ball and quasi-steady
side-force in flight. Trans Jpn Soc Mech Eng 73:1981–1986
Mizota T, Kuba H, Okajima A (1995a) Erratic behavior of knuckle ball (1) Quasi-steady flutter
analysis and experiment. J Wind Eng 62:3–13
Mizota T, Kuba H, Okajima A (1995b) Erratic behavior of knuckle ball (2) Wake field and
aerodynamic forces. J Wind Eng 62:15–21
Nagami T, Morohoshi J, Higuchi T, Nakata H, Naito S, Kanosue K (2001) The spin on fastballs
thrown by elite baseball pitchers. Med Sci Sports Exerc 43(12):2321–2327
Nathan AM (2007) The effect of spin on the flight of a baseball. Am J Phys 76(2):119–124
Watts RG, Ferrer R (1987) The lateral force on a spinning sphere: aerodynamics of a curveball.
Am J Phys 55:40–44
Chapter 27
Baseball Hitting Accuracy and Contributing
Factors
Takatoshi Higuchi, Tomoyuki Nagami, and Kazuyuki Kanosue
Abstract The purpose of this chapter is to discern the relationship between spatial
accuracy, timing accuracy, and bat control and hitting accuracy for elite collegiate
baseball batters. Nine college baseball batters performed three tasks. The first task
was hitting a fastball thrown by a pitching machine (HPT). The second task was
observing a pitching machine’s fastball and indicating the location (OPT). The third
task was hitting a ball on a baseball tee (TBT). The subjects’ performance in hitting
accuracy was defined by their success rate in the HPT. The distribution of the point
of ball-bat impact in the TBT represented the subjects’ ability in the bat control. The
fluctuations in the location in pitcher-to-catcher direction between the HPT and the
TBT represented the subjects’ temporal accuracy. The subjects’ spatial accuracy
was defined by their performance in the OPT. Although they were able to control
their bat swings to hit a ball within the effective impact area most of the time in the
Tee Ball Task, timing and spatial components of their performance indicated larger
errors and lower precision. Our results suggest that the perceptual skills involved in
baseball hitting are the main reason why batters fail to hit a ball accurately.
Keywords Batting • Interceptive action • Hand-eye coordination • Ball-bat contact
27.1 Introduction
Successful hitting is an outcome involving the correct prediction of pitched ball

trajectory and accurate execution of the bat swing so that the bat is at the right
location at the right time. To put it the other way around, unsuccessful hitting can be

T. Higuchi (*) • T. Nagami • K. Kanosue
e-mail: t-higuchi@fuji.waseda.jp

336 T. Higuchi et al.
the result of errors in prediction of ball trajectory, timing of ball-bat contact, or

swinging the bat to its intended location (Higuchi et al. 2013a, b). It seems likely that
these errors all occur because these components of accurate hitting all require fine-
tuned perceptual and motor skills. Batters have to predict the trajectory of pitches
with various speeds, spins, and launch angles. To hit a fair ball, the allowable length
of timing error for ball-bat contact is less than 10 ms (Adair 2002). Moreover, the
size of the bat’s barrel surface area that allows batters to hit a ball far and fast is
small; approximately 100 mm wide and 25 mm high (Nathan 2000).
One index generally used to describe a batter’s overall accuracy of ball-bat
contact is the “batting average”. Having a batting average above 0.3 qualifies one as
an excellent batter in Major League Baseball (Adair 2002). No batter has ever
finished an entire season with a batting average above 0.4 (with enough at-bats to be
qualified for a batting championship) since Ted Williams of the Boston Red Sox
achieved 0.406 in 1941. Although the batting average describes how often a batter
reaches at least first base, it does not exactly indicate the batter’s ability to hit
accurately. For example, a ground ball or pop fly often results in a base hit because
the batted ball goes to a location where field players have a hard time covering.
Similarly, batters often make an out even when they hit a line drive if the ball is
caught by a field player. In order to quantitatively evaluate a batter’s ability to hit
accurately, a detailed physical analysis of their ability to make proper contact
between the bat and ball is required. Specifically, it is necessary to evaluate the
batters’ abilities to (1) predict when and where a thrown ball will arrive, (2) swing a
bat to the intended location and time, and most importantly, (3) minimize the
distance between the sweet spot and the contact point at the moment of contact.
The purpose of our investigation was to (1) quantify batter’s performance in spatial
perception, timing accuracy, and bat control, and then (2) analyze the relationship
between the accuracy of ball-bat contact and these three parameters.
27.2 Methods
27.2.1 Subjects
Nine college baseball field players participated in the experiment. Mean age,
height, and body mass (mean SD) were 20.2 1.1 years, 1.76 0.08 m, and
72.3 10.0 kg, respectively. The mean length of their baseball experience was
12 1.9 years (range, 9–16). Their team is a member of the Tokyo Big Six Baseball
League and won the league championship five times and the national championship
twice in the last ten seasons. Seven subjects were right-handed batters, and two
were left-handed batters. The experiment consisted of three tasks which were
completed by each subject within 1 day. The study was approved by the Human
Ethical Committee of Waseda University. Written consent forms were obtained
from all participants. The subjects were paid for their participation on an hourly
27 Baseball Hitting Accuracy and Contributing Factors 337
basis and were naive to the purpose of the experiment until a post-experiment
debriefing was given.
27.2.2 Procedure for the Hitting a Pitch Task
The Hitting a Pitch Task involved hitting 30 fastballs thrown by a two-wheel

pitching machine (Inverter IS, Toa Sports Machine Inc.) in the same way as they
would hit fastballs in a game. The ball release point of the pitching machine was
placed 17 m from home plate and 1.6 m above the ground, which is about the same
as an ordinary pitcher’s release point. Launched ball velocity and ball backspin rate
were maintained at a constant setting of 38.9 m/s (140 km/h) and 35 rotations per
second, respectively. Ball velocity was also measured by a radar gun (The Jugs
Company Japan Ltd., Japan). For consistent ball projection, an experimenter fed
balls to the machine with the same seam orientation. The location of each launched
ball was manipulated randomly and scattered over the strike zone by adjusting the
antero-posterior tilt angle and direction of the machine.
An aluminum bat (SC900 Victory Stage Dream King; length ¼ 0.85 m;
weight ¼ 0.9 kg; Mizuno, Japan), and official Japan amateur league baseballs
(2OH100, Mizuno, Japan) were used. The subjects were instructed to place their
trailing (back) foot in their normally preferred position and to utilize the same foot
position during all the trials. A 10-min-break was given after 15 trials to minimize
the influence of fatigue.
27.2.3 Data Collection and Analysis for the Hitting

a Pitch Task
The movements of bat and ball were recorded using two synchronized high-speed
video cameras (Frame rate ¼ 1,000 Hz, exposure time ¼ 0.5 ms, resolution ¼
680 480 pixels, Trouble Shooter, Fastec Imaging Corporation, USA). Camera
1 was placed 6 m away from home plate at a right angle to the line between the
center of the pitching rubber and the center of home plate, and camera 2 was placed
6 m behind home plate to provide a rear view of the hitting movement (Fig. 27.1).
To establish the actual spin rate of the ball just after ejection from the pitching
machine, the spin rate was determined from images taken by camera 3 (Frame
rate ¼ 1,000 Hz, exposure time ¼ 0.5 ms, resolution ¼ 680 480 pixels, Trouble
Shooter, Fastec Imaging Corporation, USA) for 150 ms after the ball’s ejection.
Camera 3 was located 2 m behind the pitching machine. Linear velocity of a
launched ball was obtained from digitized data taken from 5 ms to 1 ms before
the moment of ball-bat contact. Reflective tape was attached to the barrel end of the
bat and 450 mm down from the barrel end of the bat to aid data analysis.
Fig. 27.1 Experimental

setup and the space-
coordinate system (X global,
Y global, and Z global) with its
origin at the rear point of
home plate
Data involving the location of the bat top, the bat grip, and the launched ball
were obtained from images, digitized, and analyzed using a motion analysis system,
Frame Dias IV (DKH, Japan). One frame (1 ms) before the frame which captured
contact of the ball and bat was selected as the frame to be utilized for analysis of
ball-bat location. Three-dimensional coordinates were obtained using the Direct
Linear Transformation method (DLT) with an approximately 2 m 2 m 2 m size
radial calibration structure with 68 reference points and five reference markers set
vertical and horizontal relative to the ground. The right-hand orthogonal reference
frame was defined by Xglobal, Yglobal, and Zglobal-axes with the origin at the rear
point of home plate (Fig. 27.1). The Yglobal-axis was directed from home plate to the
pitcher’s plate; and the Zglobal-axis indicated a vertically upward direction. The
Xglobal-axis was defined as the cross product of the Yglobal and Zglobal-axes. For
analysis of left-handed batters, a left-hand coordinate system with the same Yglobal
and Zglobal-axes as the right-hand coordinate system was utilized. For calibration,
poles with three reference markers (0, 0.75, and 1.5 m from the bottom) were
vertically set at nine different locations within a 1.5 m 1.5 m square on the
ground. A recording of the calibration points with cameras 1 and 2 was conducted
both before and after the batting tasks. To test the accuracy and reliability of this
measurement method, one investigator digitized two reference markers on a swung
bat for five frames on two separate occasions. The difference between the actual
value (0.450 m) and calculated value (mean standard deviation ¼ 0.453
0.0004) was less than 2 %. For the test-retest reliability of the distance,
r was ¼ 0.953.
To clarify the spatial relationship between the bat’s sweet spot and the ball at the
point of impact, the impact Z displacement and the impact X displacement were
calculated (Fig. 27.2). First, the bat vector was defined as lying on the long axis of
the bat and as being oriented from the bat grip to the top. Then, the impact X
displacement can be computed as

impact X displacement ¼ bx rImp rSS ; ð27:1Þ
where the bx ¼ unit vector that is parallel to the bat vector directed away from the
sweet spot ¼ position of the ball center at impact, and r SS ¼ position of the sweet
spot at impact. The impact X displacement provides a measure of hitting accuracy
in the direction parallel to the bat. The bat coordinate system was adopted to show
Fig. 27.2 Top view (left) and back view (right) at the moment of ball bat impact by a batter
(middle). Impact Z displacement is the distance between the ball center (rImp) and the sweet spot of
the bat (white circle) (rSS) in the direction of bZ
the ball location on bat barrel at the moment of contact. The impact Z displacement
can be computed as

impact Z displacement ¼ bz rImp rSS ; ð27:2Þ
where the bz ¼ unit vector that is perpendicular to the bat vector directed upward in
the vertical plane, r Imp ¼ position of the ball center at impact, and r SS ¼ position of
the sweet spot at impact. The impact Z displacement provides a measure of hitting
accuracy in the direction perpendicular to the bat.
27.2.4 Procedure for the Observing a Pitch Task
The second task, Observing a Pitch Task, was designed to indicate the location of a
pitch right after it passed home plate (See Extras.). Location and settings of the
pitching machine and cameras were identical to those in the Hitting a Pitch Task.
Twenty fastballs which arrived at different locations were thrown for each subject.
Subjects were instructed to observe pitches thrown by the pitching machine and
then indicate the location of the ball with a bat based on their judgment right after
the ball passed home plate. This procedure was designed in order to eliminate the
timing accuracy and bat control components of batting accuracy. It was also
necessary to eliminate visual information about the ball’s trajectory after it passed
the plate. This information would not normally be available to a batter, but could be
potentially be utilized by the subjects to locate the ball’s position. For this reason,
the subjects wore a visual occlusion liquid-crystal apparatus (PLATO, Translucent
Technologies Inc., USA) (Fig. 27.3). The timing of visual occlusion was regulated
by software (ToTaLcontrol 2.0, Translucent Technologies Inc., USA) and subjects’
Fig. 27.3 Image of the two-wheel pitching machine with the photo-sensing diode and visual
occlusion liquid-crystal apparatus
vision was occluded 350 ms after a ball passed a photo-sensing diode (PLDM-10,
Sankei Machinery, Japan, Fig. 27.2). This occurred about 100 ms before the ball
passed home plate. The visual occlusion lasted for 150 ms. The same bat and ball as
the Hitting a Pitch Task were used in this observation task.
27.2.5 Data Collection and Analysis for the Observing

a Pitch Task
The same cameras and coordinate system as used in the Hitting a Pitch Task were
utilized in this task. Pitch trajectory during the period between 150 ms before and
150 ms after the ball passed home plate was recorded. Next, video cameras
recorded the subjects for 30 ms while they were indicating their perceived location
of the passed ball. Calculation of estimated ball-bat contact location was conducted
as follows:
1. The locations of markers on the bat while the subject was indicating his
perceived ball location were digitized for one frame.
2. For digitization of ball center, an experimenter digitized ten continuous frames
of ball images which included the period of time when a launched ball was
passing the vertical plane that contains the indicated bat.
3. Displacement from the bat’s sweet spot to the ball center in the direction of
Yglobal axis was calculated for each of the ten frames. If the optimal choice of the
ten frames was made, Yglobal displacement became smaller and closer to zero
during the first five frames and attained greater negative values during the last
five frames.
4. The frame before the Yglobal displacement decreased to less than 70 mm was
considered to be the ball location just before ball-bat contact and was used to
calculate the estimated ball-bat contact location.
27.2.6 Procedure for the Tee Ball Task
The third task, Tee Ball Task, was to hit a ball off a baseball tee (2ZA770,
MIZUNO, Japan) from five different locations (Fig. 27.2), ten times at each
location. Ball location was changed every trial in a randomized order. The same
bat and ball as those in the Hitting a Pitch Task were used. Since the ball (target)
was at rest on the tee, this task should assess the batter’s skill in bat control, while
largely excluding skill in timing accuracy and visual perception. There were three
kinds of ball location which were on the Yglobal axis with different heights, (a) the
base of the patella¼“low”, (b) mid-height between the acromion and iliac
crest¼“high”, (c) the midpoint between the “low” and “high”¼“middle”. At the
height of c, there were two other ball locations; (d ) in which the ball was located
0.2 m towards the hitter in the Xglobal direction¼“inside”, and (e) in which the ball
was located 0.2 m away from the hitter in the Xglobal direction¼“outside”
(Fig. 27.4). Ball locations in the Yglobal direction were determined based on each
subject’s characteriatics. The subjects were instructed to hit utilizing the same foot
Fig. 27.4 Five ball locations for the Tee Ball Task
position as they did for the fastball hitting task. Three-minute-breaks were given
after 10 trials to minimize the influence of fatigue.
27.2.7 Data Collection and Analysis for the Tee Ball Task
The same cameras and coordinate system as utilized in the Hitting a Pitch Task
were employed in this task. The movement of bat and ball 150 ms before and after
ball-bat contact was recorded from the same cameras and with the same settings as
for the other tasks. The frame (1 ms) before the frame capturing contact of ball and
bat was selected as the frame with which to analyze ball-bat location. The coordi-
nate systems were the same as those utilized in the Hitting a Pitch Task.
27.2.8 Statistical Analyses
The subject’s ability to hit a pitch accurately was determined by the mean of
distance (absolute value from the sweet spot) between the ball center and the
sweet spot and the standard deviation of displacement (relative value from the
sweet spot) of the ball center from the sweet spot in the Hitting a Pitch Task. Shorter
ball-bat distance and smaller standard deviation of the displacement at the moment
of impact indicated greater hitting accuracy.
The mean of distance between the ball center and the sweet spot as well as the
standard deviation of displacement of ball center from the sweet spot in the
Observing a Pitch Task indicated the subject’s performance for the spatial percep-
tion of ball location. The mean of distance between the ball center and the sweet
Fig. 27.5 An example of an approximate line of a subject’s preferred ball-bat contact location and
equation
spot coupled with the standard deviation of displacement of ball center from the
sweet spot in the Tee Ball Task indicated the subject’s performance in bat control –
the batter’s ability to bring the bat to its intended location. The subject’s perfor-
mance in timing accuracy was analyzed based on the mean distance between the
ball-bat contact location and the “preferred impact location”. This was established
from their preferred Yglobal ball location in the Tee Ball Task. First, an approximate
equation of preferred ball locations on the XYglobal plane in each subject’s Tee Ball
Task was calculated (Fig. 27.5). Then, the difference between the Yglobal value of
ball-bat contact location in each trial of the Hitting a Pitch Task trial and the Yglobal
value which was calculated by assigning Xglobal value in the trial to the appropriate
equation was determined.
Statistical analyses were conducted with IBM SPSS statistical software, version
20 (Japan IBM, Japan). Correlations between hitting accuracy and various factors
were calculated using the Pearson product-moment correlation coefficient. Statis-
tical significance was set at p < 0.05. A two-way ANOVA with repeated measures
(factor-task [Hitting a Pitch Task vs. Observing a Pitch Task vs. Tee Ball Task] and
direction of the impact displacement [impact X displacement vs. impact Z dis-
placement] was used to assess difference in the standard deviation of the impact
deviations in each task. Significant results were further analyzed with a post hoc
Bonferroni t test for multiple comparisons. The alpha level for significance was set
at p < 0.01.
27.3 Results
27.3.1 Actual Hitting Accuracy
The mean and standard deviation (SD) of ball location at the moment of ball-bat
contact in the Hitting a Pitch Task for each subject are shown in Fig. 27.6a and
Table 27.1. In the Hitting a Pitch Task, mean (SD) of each subject’s ball-bat
distance at the moment of impact was 40.4 8.6 mm in the Xbat direction and
24.0 3.6 mm in the Zbat direction. The mean (SD) of the standard deviation of
each subject’s ball-bat displacement at the moment of impact was 41.6 7.8 mm in
the Xbat direction and 22.0 4.3 mm in the Zbat direction. These values were used
as indicators of the subjects’ actual performance in hitting accuracy.
27.3.2 Spatial Perception of Ball Location
The mean (SD) values for ball location at the moment of ball-bat contact in the
Observing a Pitch Task for each subject are shown in Fig. 27.6b and Table 27.1. In
the Observing a Pitch Task, mean and standard deviation of each subject’s ball-bat
Fig. 27.6 Locations of average ball location at the moment of ball bat contact for (a) Hitting a
Pitch Task, (b) Observing a Pitch Task, (c) Tee Ball Task
Table 27.1 Mean (SD) of ball-bat distance for three tasks

Task Impact X displacement (mm) Impact Z displacement (mm)
Hitting a Pitch Task 40.4 8.6 24.0 3.6
Observing a Pitch Task 42.5 17.7 53.5 46.2
Tee Ball Task 27.4 4.9 16.6 1.6
distance at the moment of virtual ball-bat contact was 42.5 17.7 mm in the Xbat
direction and 53.5 46.2 mm in the Zbat direction. The mean (SD) of the standard
deviation of each subject’s ball-bat displacement at the moment of impact was
36.8 6.0 mm in the Xbat direction and 39.6 12.1 mm in the Zbat direction. These
values were used assess the subjects’ performance in spatial perception.
27.3.3 Bat Control to the Intended Location
The mean (SD) values for ball location at the moment of ball-bat contact in the
Tee Ball Task for each subject are shown in Fig. 27.6c and Table 27.1. In the Tee
Ball Task, the mean (SD) of each subject’s ball-bat distance at the moment of
ball-bat contact was 27.4 4.9 mm in the Xbat direction and 16.6 1.6 mm in the
Zbat direction. The mean (SD) of the standard deviation of each subject’s ball-bat
displacement at the moment of impact was 26.1 4.9 mm in the Xbat direction and
17.0 3.6 mm in the Zbat direction. These values were used as indicators of each
subject’s skill in controlling the bat.
27.3.4 Timing Accuracy for Intended Impact Location
Ball locations at the moment of impact in the Hitting a Pitch Task with the
approximate lines of ideal impact location to indicate the timing error are shown
in Fig. 27.7. Mean and standard deviation of the distance between preferred ball
location and actual ball location in the Yglobal direction, the indicator of the timing
error, were 153.3 33.2 mm. The mean and standard deviation of the standard
deviation of displacement of the actual ball location from the ideal ball location in
the Yglobal direction were 164.6 25.1 mm.
27.3.5 Correlation with Actual Hitting Accuracy
There was a moderate positive correlation between each subject’s standard devia-
tion of ball-bat contact location in the Xbat direction in the Hitting a Pitch Task and
the standard deviation of virtual ball-bat contact location in the Xbat direction in the
Observing a Pitch Task (r ¼ 0.64, p ¼ 0.06) (Fig. 27.8a). The standard deviation of
Fig. 27.7 Ball location at the moment of impact located on the XYglobal plane for each subject’s
Hitting a Pitch Task. Dashed lines indicate the approximate line obtained from ball locations at the
moment of impact for their Tee Ball Task. Colored circles indicate that the impact Z displacement
was between 0 and 25 mm, and impact X displacement was between 50 and 50 mm
Fig. 27.8 Relationship between the actual hitting accuracy and spatial perception of ball location.
(a) Scatter plot of each subject’s average distance of ball-bat impact location and the standard
deviation of virtual ball-bat impact location in the Xbat direction. (b) Scatter plot of each subject’s
standard deviation of ball-bat impact location in the Xbat direction and the standard deviation of
virtual ball-bat impact location in the Xbat direction
virtual ball-bat contact location in the Xbat direction in the Observing a Pitch Task
also had a significantly strong positive correlation with each subject’s average
distance of ball-bat contact location in the Hitting a Pitch Task (r ¼ 0.73,
p < 0.05) (Fig. 27.8b). There was no significant correlation between the actual
hitting accuracy (¼ performance in the Hitting a Pitch Task) and timing accuracy
(¼ fluctuation of ball-bat contact location in the Yglobal direction) or bat control (¼
performance in the Tee Ball Task).
27.3.6 Precision of Ball-Bat Contact in Three Tasks
Standard deviation of the impact X displacement and impact Z displacement are

shown in Fig. 27.9. A two-way ANOVA indicated a significant interaction of the
factors (F2,16 ¼ 22.65, p < 0.0001). A post hoc paired t-test showed that the standard
deviation of the impact Z displacement in the Hitting a Pitch Task and the Tee Ball
Task were significantly smaller than the respective standard deviation in the impact
X displacement (22.0 4.3 mm vs. 41.6 7.8 mm for the Hitting a Pitch Task
Fig. 27.9 Standard deviation of each subject’s impact displacement (mm) for three tasks: Hitting
a Pitch Task (HPT), Observing a Pitch Task (OPT), and Tee Ball Task (TBT)
(p < 0.001); 17.0 3.6 mm vs. 26.1 4.9 mm for the Tee Ball Task (p < 0.01)).
Another post hoc t-test showed that the impact Z displacements in the Hitting a
Pitch Task and Tee Ball Task were significantly smaller than the impact Z displace-
ments in the Observing a Pitch Task (p < 0.05). Also, the impact X displacement in
the Tee Ball Task was significantly smaller than the impact X displacement in the
Hitting a Pitch Task and Observing a Pitch Task (p < 0.05).
27.4 Discussion
To assess hitting skill of experienced baseball batters, their performance in the three
tasks was measured. The Hitting a Pitch Task was utilized to quantify their accuracy
in an actual hitting situation. Timing accuracy was assessed by measuring the
location of ball-bat contact in the bat-coordinate system and global-coordinate
system, respectively. These parameters were utilized on the assumption that the
batters’ intention was to hit a ball at the sweet spot and ideal Yglobal location in order
to hit the ball far and fast. The Observing a Pitch Task was utilized to test their ability
to predict the arrival location of a flying ball by pointing to the predicted location
instead of swinging the bat at the intended location, which could be affected by their
bat control skill. The Tee Ball Task was utilized to measure their performance to
accurately hit a stationary ball. In this task, once the batters see the ball on a tee, they
know exactly where to hit. Therefore, error in the location of ball-bat contact can
indicate an error in their bat control. Moreover, in the Tee Ball Task, batters were
asked to adjust the ball location in the Yglobal direction because each batter’s ideal
impact location in the pitcher-to-catcher direction differs depending on the vertical
(Zglobal) and horizontal (Xglobal) location. The preferred ball location was utilized to
measure errors in ball-bat contact location in the global-coordinate system.
27.4.1 Performace in Actual Hitting Accuracy
As shown in Fig. 27.6a, eight out of nine subjects’ average impact location was
within the “effective impact area” – a rectangle of 25 mm height and 50 mm
width centered around the sweet spot as defined by Adair (2002). Although the
mean standard deviation of the impact X displacement and the impact Z displace-
ment were smaller than the size of the effective impact area (41.6 mm and 22.0 mm,
respectively), there were many mishits outside the effective impact area because the
average impact location was skewed toward the grip end of the bat and upwards
from the sweet spot. The reason for this tendency is unclear, but some possibilities
will be suggested.
In this experiment, a four-seam fastball was launched from the machine. When a
launched ball is spinning, in addition to the gravitational force and air drag force,
the direction and rate of spin will have an influence on flight trajectory (Bahill and
Baldwin 2007). As a ball spins closer to pure backspin, and at faster rates, the lift
force acting against the drop of the ball becomes greater. Although the ball
backspin rate adopted in this experiment replicated a normal ball backspin rate
thrown by actual pitchers (Nagami et al. 2011), the spin angle might have been
closer to a pure backspin than that thrown by actual pitchers. This would lead the
subjects to swing under the ball.
Another possible explanation is that the fastball has the smallest amount of drop
of the normally thrown pitches. If batters are trained to swing based on the average
trajectory of all the various kinds of pitches, they would tend to swing under the
fastball trajectory more than with other types of pitches. It is also possible that if the
batters intended to hit line drives, they would need to swing 10–20 mm under the
ball rather than at the center of the ball (Bahill and Baldwin 2008).
27.4.2 Performace in Bat Control
Performance in the subjects’ bat control was evaluated by their performance in the
Tee Ball Task. Since the ball was not moving in this task, a batter with good bat
control could hit the ball at the sweet spot repeatedly. The means of impact X
displacement and impact Z displacement, which are indicators of the subject’s
performance in bat control, were 27.4 4.9 mm and 16.6 1.6 mm, respectively.
Based on this result, when the ball was not moving, batters were able to hit the ball
with less deviation from the sweet spot. Standard deviation of the impact X
displacements, which indicates the precision of ball-bat contact in the Xbat direc-
tion, in the Tee Ball Task was significantly smaller than that in the Hitting a Pitch
Task. On the other hand, there was no significant difference in the standard
deviation of the impact Z displacements in the two tasks. Therefore, movement
of the ball impairs the precision of ball-bat contact in the Xbat direction only.
However, the lack of significant correlation between performance in bat control
and actual hitting accuracy suggests there might be a difference in the hitting action
for hitting a flying ball or tee ball. When a batter tries to hit a flying ball, their
perceived affordance to hit the ball affects the execution of hitting action. Gibson
(1986) defined that affordances need not to be conceived by the actor via some sort
of cognitive processing, but rather can be directly perceived. Fajen (2007) termed
“affordance-based control” as a strategy for an actor to make adjustments so that the
intended action is always possible within the limits of their action capability.
Batters swing a bat when a flying ball affords hitting. Since a pitched baseball
arrives at home plate in a split second, quick decision making which utilizes
affordance would be very important. The Tee Ball Task did not task any temporal
constraints to the batters, and they might be using a different decision making
process when they hit the ball off a tee. This would minimize the relationship
between performance in the Hitting a Pitch Task and that in the Tee Ball Task.
27.4.3 Performace in Spatial Perception
Performance in the subjects’ spatial perception was evaluated based on how

accurately they could point to the location of a launched ball in the Observing a
Pitch Task. As shown in Fig. 27.6b, average locations of each subject’s virtual ball-
bat contact location were dispersed widely, especially in the Zbat direction. If they
swung the bat to those locations, they would miss the ball in many trials. However,
there was no trial where any subject swung and missed a ball in the Hitting a Pitch
Task. The dispersion might be caused by the few seconds of time delay from when
the ball passed the batter to when he pointed to the perceived ball location. It is also
possible that there might be a gap between an individuals’ perception about his
predicted location of the moving ball and the location of his swung bat when he was
hitting a ball. There was a significant correlation between the standard deviation of
impact X displacement in the Observing a Pitch Task and the impact X displace-
ment in the Hitting a Pitch Task. This correlation could be a reason for the
significantly greater standard deviation of the subjects’ impact X displacement in
the Hitting a Pitch Task than that of the Tee Ball Task. Subjects with a superior
performance of spatial perception in Xbat location had a greater precision in the
actual hitting accuracy in the Xbat direction.
27.4.4 Performace in Timing Accuracy
Timing error was assessed by measuring the distance between the preferred ball-bat
location and the actual ball-bat location in the Yglobal direction. Based on the
horizontal ball velocity during the 5 ms before ball-bat contact (35.8 0.7 m/s)
and the mean distance (15.3 mm), the average timing error from an ideal ball-bat
contact location was 4.3 ms. This timing error, which was calculated based on ball
speed and fluctuation of contact location in space, was close to the timing error of
elite athletes in other sports such as the 2–5 ms timing error in table tennis return
(Bootsma and Wieringen 1990) or the 2–3 ms timing error in cricket batting (Regan
1997). Interestingly, there was no correlation between timing accuracy and actual
hitting accuracy. As shown in Fig. 27.7, some successful ball-bat contact was made
far away from the batter’s preferred ball location. At the other extreme, there was
also unsuccessful ball-bat contact which occurred near the preferred ball location.
Therefore, whether a batter makes ball-bat contact at the batter’s preferred ball
location is not a good determinant of his overall hitting accuracy.
27.4.5 Batter’s Ability in Accurate Hitting
In our investigation, the performance of collegiate baseball batters with at least

9 years of experience was investigated. When they hit fastballs launched from a
pitching machine, most of them were able to keep their average location of ball-bat
contact within the effective impact area. Precision in the impact Z displacement
when they hit launched balls was not different from when they hit balls off a
baseball tee. On the other hand, precision in the impact X displacement when
they hit launched balls was not different from when they observed, and then pointed
to the locations of launched balls. Based on the subjects’ precision in the Tee Ball
Task (impact X displacement ¼ 26.1 mm, impact Z displacement ¼ 17.0 mm), they
have sufficient skill in bat control to repeatedly hit a ball within the effective impact
area. In fact, they were able to hit a ball within the bat’s effective impact area
(approximately 100 mm wide and 25 mm high) in about 48 % of all trials in the Tee
Ball Task. However, there were larger errors in spatial perception and timing
accuracy. Therefore, it was suggested that deficiencies in perceptual skill in base-
ball hitting might be the main reason why batters fail to hit a ball accurately. Since
errors in spatial perception and timing accuracy might be interrelated, the following
experiments will assume that errors in ball-bat contact with moving balls are errors
in spatio-temporal perceptual skill.
References
Adair RK (2002) The physics of baseball, 3rd edn. HarperCollins Publishers, New York,
pp 121–130
Bahill AT, Baldwin DG (2007) Describing baseball pitch movement with right-hand rules.
Comput Biol Med 37:1001–1008
Bahill AT, Baldwin DG (2008) Mechanics of baseball pitching and batting. In: Ghista D
(ed) Applied biomedical engineering mechanics. CRC Press/Taylor & Francis Asia Pacific,
Boca Raton, pp 445–488, Chapter 16
Bootsma RJ, van Wieringen PCW (1990) Timing an attacking forehand drive in table tennis.
J Exp Psychol Hum Percept Perform 16(1):21–29
Fajen BR (2007) Affordance-based control of visually guided action. Ecol Psychol 19(4):383–410
Gibson JJ (1986) The ecological approach to visual perception. Houghton-Mifflin, Boston
Higuchi T, Morohoshi J, Nagami T, Nakata H, Kanosue K (2013a) The effect of fastball baskspin
rate on baseball hitting accuracy. J Appl Biomech 29:279–284
Higuchi T, Nagami T, Morohoshi J, Nakata H, Kanosue K (2013b) Disturbance in hitting accuracy
by professional and collegiate baseball players due to intentional change of target position.
Percept Mot Skills 116(2):627–639
Nagami T, Morohoshi J, Higuchi T, Nakata H, Naito S, Kanosue K (2011) The spin on fastballs
thrown by elite baseball pitchers. Med Sci Sports Exerc 43(12):2321–2327
Nathan AM (2000) Dynamics of the baseball-bat collision. Am J Phys 68(11):979–990
Regan D (1997) Visual factors in hitting and catching. J Sports Sci 15:533–558
Chapter 28
Automatic Tracking of Player Locations
from Video Image of Football Game
Masaaki Honda and Noriyuki Oosaka
Abstract In this paper, we describe a method for automatically tracking the

location of a player on the pitch from the video image of a football game. In this
method, the player location on the video image is determined by employing binary
image transformation based on the color features of the player’s uniform and a
centroid computation of the binary image. The image area for the position tracking
and the color threshold value in the binarization are automatically adapted in the
tracking process to accommodate the motion of the player as well as the lighting
condition, which is dependent on the pitch location. The field coordinates of the
player location are determined from the video coordinates by using the two
dimensional DLT (Direct Linear Transformation) method. The automatic tracking
method was evaluated for video data which were collected for a football game. The
results indicated that the correctly detected rate of the tracking method was
99.93 %.
Keywords Image processing • Automatic tracking • Football game
28.1 Introduction
In the game analysis of ball games such as international football, locations of the
players and ball are important primitive features for examining qualitatively the
player’s physical performance such as running speed, running distance, and the
effectiveness of the formation utilized by the players in defense and offense (Ueda
and Honda 2014). Presently, these location data are mostly collected using manual
digitization procedure on video data of a game. However, this procedure is time-
consuming. Hard work and much effort are needed to obtain the location data of all
players and their relationship to the ball for the entire game, On the other hand,
potable sensor devise like GPS sensors or acceleration sensors have been used to
measure the location of players. These devises are feasible to use for automatic data
collection, but there remain some problems in their measurement accuracy as well
M. Honda (*) • N. Oosaka

e-mail: hon@waseda.jp

354 M. Honda and N. Oosaka
as the annoyance associated with mounting the devise on player. This latter
problem limits the use of sensor devises to training games. Thus, for official
games, a system for tracking player locations from video image is needed. So far
several studies on the automatic tracking based on the video image processing have
been done (Kanzaki and Ariki 2002; Siozaki et al. 2004; Tokikura and Nishihara
2005; Lim and Haseyama 2008; Oosaka and Honda 2009).
In this chapter, a method for automatically tracking the location of a player from
the video image of football game is described. This method employs image
binarization of the color image based on the color features of the player’s uniform.
The location of the player in the video coordinates is determined by calculating the
centroid on the binary image. The image area for position tracking and the color
threshold value in the binarization are automatically adapted in the tracking process
to accommodate the motion of the player as well as the lighting condition, which is
dependent on pitch location. The location of the player on the field coordinates is
obtained from the video coordinates by using the two dimensional DLT method
after transforming the video coordinates detected as the body location into those of
the foot location. The method was evaluated by using video data taken from an
official football game and the tracking accuracy and the tracking error sources were
examined. Finally, a semi-automatic location tracking system to cope with the
tracking error will be discussed.
28.2 Automatic Tracking of Player Location
The automatic tracking procedure consists of two parts. The first part involves
image processing for tracking of the location of the player on the video image. The
second part involves a transformation from the video coordinates of the player’s
location into the field coordinates.
28.2.1 Video Image Processing
In a football game, the player on the video image is characterized by the distinct
color features of the uniform which contrast with the field pitch which has green
color. The uniform color is often characterized by different colors on the shirt and
trunks. For a given tracking image area, the color image in this area is transformed
into the binary image as

if I i jk T sk Δ and I i jþhk T tk Δ then Bij ¼ 1,
else Bij ¼ 0;

where I i jk i ¼ cx δx , , cx þ δx , j ¼ c y δ y , , c y þ δ y is RGB ðk ¼ 1, 2, 3Þ
values of the pixel (i, j) of the image, cx, cy and δx, δy are the center coordinates of
28 Automatic Tracking of Player Locations from Video Image of Football Game 355
Fig. 28.1 Example of image binarization procedure. Left is the original color image and right is
the binary image
the rectangular tracking area and its sizes, and Bij is the pixel value of the binary
image. Tsk and Ttk are RGB values of the shirt and trunks of the uniform and Δ is the
threshold value in the image binarization. h is distance in pixel from the shirt of the
player to the trunks. The RGB values of the uniform are specified manually to the
player of each team at the beginning of the tracking process. This image
binarization is an extension of conventional binarization of the gray scale image.
Figure 28.1 shows examples of the original color image and the binary image.
The video coordinates of the player location is determined by calculating the
centroid location to the binary image as
X .X
xg ¼ iBi j Bi j
i, j ij
X .X
yg ¼ jBi j Bi j
ij ij
Figure 28.2 shows an example of the detected player location. It should be noted
here that the detected position is located around the breast of the player but not the
foot position. The conversion from the breast position to the foot position is
described in the next session.
The location of the tracking area for each player is specified manually at the
beginning of the tracking, and then is updated according to the detected player
location as

cx ðtÞ ¼ cx ðt 1Þ þ k xg ðtÞ cx ðt 1Þ

c y ðtÞ ¼ c y ðt 1Þ þ k yg ðtÞ c y ðt 1Þ
where k is a updating coefficient and set at 0.8 experimentally. Figure 28.3 shows
the sequence of tracking areas in the tracking process.
Fig. 28.2 Player location

obtained by centroid
computation of the binary
image
Fig. 28.3 Adaptation of the

location of the tracking area
28.2.2 Adaptation of Parameters
The size of the tracking area has an important influence on the tracking perfor-
mance. If the size is enough large, the player mostly exists in the area during the
tracking process. However, in situations where multiple players are inside the area,
the chance for tracking errors increases, namely alternation of the trucked player
with another player. Contrarily, if the size is too small, the probability for the player
to be outside the area when the player moves at a high speed, is increased. This
leads to follow-up errors in the tracking process. In order to reduce these errors, it is
Fig. 28.4 Adaptation of the

tracking area size
necessary to change the size of the tracking area during the tracking process. The
size is changed according to the number of white pixels in the binary image. If the
number is small, the player location is possibly outside the tracking area. Con-
trarily, if the number is sufficiently large, the tracking area should be reduced to
prevent other players from being inside the area. The tracking area sizes δx, δy are
initially set at their minimum values δx/y min. If the number of white pixels of the
binary image at a video frame is less than the threshold value (typically 1), the size
used for the next frame is expanded by multiplying by a constant value larger than
1 within a range not exceeding the maximum value δx/y max. If the number exceeds
the threshold value at a video frame, the size at the next frame is reduced by
multiplying by a constant value less than 1 within a range not exceeding the
minimum value δx/y min. The adaptation procedure of the tracking area size is
shown in Fig. 28.4.
The tracking procedure is performed for every player. A tracking error often
occurs when two players of the same team are close each other and enter inside the
tracking area. The error involves an alternation of the location of the target player
with that of the other player. In order to reduce such alternation errors, when the
tracking area for a player overlaps those of other players for which the tracking is
completed, the overlapped area is eliminated from the tracking area for the current
player as shown in Fig. 28.5.
RGB values of the uniform color are dependent on the location of the field pitch
because the lighting conditions are slightly different as shown in Fig. 28.6. This
figure shows the variances of the RGB values of the uniform color during the
tracking process. Therefore, the color threshold Δ used in the image binarization is
adapted within the range of Δmin to Δmax according to the number of white pixels in
the binary image in a similar way to the adaptation procedure for the tracking area
size.
Fig. 28.5 Elimination of

the overlapped tracking area
Fig. 28.6 Temporal changes of the RGB values of the uniform color
28.2.3 Transformation from Video Location to Field

Location
The field coordinates of player location are obtained from the video coordinates by
using the two-dimensional DLT method. The detected video coordinates are
located around the breast of the player. To determine the correct field coordinates
by using the DLT method, it must know the video coordinates of the foot position.
Thus, the foot position is estimated
from the breast position by modifying the video
coordinates of Y-axis as yg d b f xg ; yg . Here, dbf(xg, yg) is the distance from the
breast position to the foot position in the video coordinates. This distance is
obtained by converting the average distance in the field coordinates into that in
the video coordinates by using an inverse DLT method which involves mapping the
distance in field coordinates into those video coordinates as a function of the video
coordinates (xg, yg).
The field coordinates of player location are obtained from the video coordinates
of the player’s foot location by using the 2-D DLT method (Abdel-Aziz and
Karara 1971). The relationship between the video coordinates (xg, yg) and the
field coordinates (xf, yf) is shown as
L 1 x f þ L 2 y f þ L3
xg ¼ ,
L7 x f þ L8 y f þ 1
L4 x f þ L5y f þ L6
yg ¼ ;
L7 x f þ L8 y f þ 1
where Li ði ¼ 1, 2, , 8Þ are DLT parameters. Then, the field coordinates are

obtained by solving the linear equations as

L xg L7 x f þ L2 xg L8 y f ¼ xg L3
1
L4 yg L7 x f þ L5 yg L8 y f ¼ yg L6 :
The DLT parameters are constant values for a fixed camera location. These values are
obtained in the calibration procedure from both the video and field coordinates of the
four control points which are selected as the four corner points of the football pitch.
28.3 Experiments
28.3.1 Video Recording
Video data was taken from the official football match between Waseda and Keio
Universities. Two high definition (HD) video cameras were used for the video
recording. The resolution of the HD video camera in the present study was 1,440
Fig. 28.7 Video images of left and right half pitches taken by two video cameras
1,080 pixels; the speed of the image capture was 30 frames per second (fps). Video
recording was performed with a fixed camera location with no use of panning or
zoom. Given the size of a football pitch, it would be difficult to cover the whole
pitch with just one camera; even using an HD video camera located at a distance,
the resolution would be unacceptably low, and it would be hard to determine player
location with a high degree of precision. For this reason, two fixed-location cameras
were used, each recording the action in half of the pitch. The cameras used were
Full HD video cameras manufactured by Sony. The video cameras were located on
the top level of the main stand, in such a way as to be able to cover the area from the
corners in their respective halves of the pitch up to the halfway line as shown in
Fig. 28.7.
The recorded images were transferred to a PC using the Edius Neo 2 video
editing software produced by Grass Valley (formerly Thomson GrassValley Cano-
pus). As the two cameras were not synchronized when recording the video footage,
adjustment was implemented on a frame-by-frame basis to allow for discrepancies
in the movement of the ball and players between the footage from the two cameras,
after which the video images were synchronized. The video images of the total
number of 3,600 frames were selected from the video data of the game match and
were used for an evaluation of the method.
28.3.2 Combination of Multiple Images
In the video recording, two video cameras were used to capture a whole football
pitch. When the player location moved from one half-pitch to the other half-pitch,
the tracking procedure needed to be automatically switched from one video image
to the other video image. The use of two video images complicated the tracking
procedure. Therefore, the two video images were combined into one image which
captured the whole football pitch. The combining procedure was performed by
mapping the pixel of one image into the pixel outside another image area. Denoting
Fig. 28.8 Combined image of the left and the right half pitch images
a mapping function of the video coordinates (xv, yv) of the pixel of one image into
the field coordinates (xf, yf) by φ1(xv, yv), and an inverse mapping function of the
field coordinates of the pixel
of another image (reference image) into the video
coordinates by φ1
2 x f ; y f , the video coordinates of the pixel outside the reference
image
area was obtained by the products of by these mapping functions,
φ1
2 x f ; y f φ1(xv, yv). Here, each mapping function was determined by the DLT
method. Then, the video coordinates were truncated to the nearest integer values
and the pixel value outside the reference image area was replaced by the pixel of the
corresponding another image. An example of two separate video images and the
combined image are shown in Fig. 28.8.
28.3.3 Evaluation of Tracking Performance
Given an initial player location specified manually, the tracking procedure deter-
mines sequentially the player location in a frame by frame manner. Thus, if the
procedure is losing the location of the player, it is not possible to find the correct
location thereafter. Therefore, when a tracking error occurred, the automatic track-
ing process was restarted after specifying the correct location manually. The
tracking error rate was obtained by counting the number of error occurrences for
the total number of 79,200 samples; 3,600 video frame and 22 players on the two
teams (with exception of the goal keepers).
Figure 28.9 shows an example of the location trajectories of players on the
combined video image. Table 28.1 shows the correct tracking score for the
non-adaptive tracking method. In this method, both the size of the tracking area
and the color threshold value in the image binarization are fixed during the tracking
process. The best score (99.927 %) is obtained with color threshold value of 55 and
at tracking size of 15 pixels. The best score is obtained around the color threshold
Fig. 28.9 Location trajectories of players displayed on the combined image
value of 55 for every tracking area size. Table 28.2 shows the correct tracking score
for the adaptive tracking method. In this method, the size of the tracking area and
the color threshold value are adapted in the manner previously described. The
minimum value of the color threshold and the size of the tracking area were set at
55 and 15, respectively. These maximum values are listed in the table. The best
score of 99.962 % is obtained for the maximum color threshold value of 80 and at
maximum tracking area size of 15 pixels. This rate is slightly better than that of the
non-adaptive method. The number of the tracking errors in the total samples of
76,200 was 58 samples for the non-adaptive method and 30 samples for adaptive
method. The adaptive method efficiently reduces the number f errors by about half.
The tracking score may appear reasonably high, but this rate means that a tracking
error occurs once every 2 s on average for one of the 22 players. This error rate is
not tolerable for a complete automatic tracking system. Thus, in this system the
experimenter needs to monitor the tracking process, detect errors and correct them
manually. In order to reduce the work load of the experimenter, automatic error
tracking is desirable. If the tracking error involves loosing track of the player, it is
possible to detect the error by examining the number of the white pixels in the
image binarization. On the other hand, if error involves tracking the wrong player
when two or more players are in close proximity, it is still difficult to detect the error
automatically.
Next, location precision for the automatic tracking was evaluated by comparing
the detected location with a manually digitalized location in terms of the video
coordinates. For manual digital processing, the mid-point between the players’ left
and right foot was determined by visual inspection. The difference between loca-
tions of the automatic and manual digitization was 1.2 2.8 pixels in the X-axis
and 2.1 2.4 pixels (average SD) in the Y-axis. Precision of the detected location
in the field coordinates is dependent on the location in the football pitch. When the
28
Table 28.1 Correct tracking score in percentage for non-adaptive tracking method
Color threshold in binarization
35 40 45 50 55 60 65 70 75 80
Size of tracking area 7 99.595 99.713 99.777 99.823 99.851 99.873 99.875 99.866 99.813 99.746
10 99.716 99.797 99.852 99.886 99.902 99.917 99.899 99.860 99.779 99.731
13 99.777 99.832 99.884 99.898 99.918 99.903 99.871 99.817 99.716 99.717
15 99.793 99.841 99.881 99.913 99.927 99.894 99.840 99.794 99.658 99.678
17 99.798 99.845 99.884 99.923 99.908 99.864 99.775 99.744 99.573 99.623
20 99.803 99.845 99.890 99.904 99.891 99.804 99.706 99.643 99.451 99.532
Automatic Tracking of Player Locations from Video Image of Football Game
363
364
Table 28.2 Correct tracking score in percentage for adaptive tracking method
Maximum color threshold in binarization
55 60 65 70 75 80 85 90
Maximum size of tracking area 15 99.942 99.947 99.952 99.956 99.960 99.962 99.961 99.961
20 99.948 99.953 99.958 99.958 99.958 99.957 99.958 99.958
25 99.946 99.952 99.955 99.958 99.961 99.961 99.961 99.961
30 99.948 99.953 99.956 99.960 99.960 99.961 99.961 99.961
35 99.944 99.952 99.956 99.960 99.960 99.960 99.961 99.961
40 99.943 99.947 99.953 99.958 99.958 99.960 99.961 99.961
45 99.942 99.946 99.947 99.955 99.953 99.960 99.961 99.961
50 99.941 99.944 99.947 99.952 99.957 99.960 99.961 99.961
M. Honda and N. Oosaka
video coordinates were altered by 1 pixel in the X-axis and Y-axis at the center
location of each area when the whole football pitch is divided into 3 by 3 grid areas,
the change in the field coordinates was 7.5 cm, 4.3 cm and 11 cm respectively along
the X-axis and 9.4 cm, 26.2 cm, and 50.0 cm along the Y-axis.
Thus, the average error of the detected location in the field coordinates is
estimated to be in the range from 5.2 to 13.2 cm in the X-axis and from 19.7 to
105.0 cm in the Y-axis. Here, it should be noted that both errors in the video and
field coordinates are dependent on the view of the video camera.
28.4 Conclusions
In this chapter, we described an automatic processing method for detecting the

location of a football player from the video image. Experimental results showed
that the tracking score was 99.962 % when using the optimal system settings.
Tracking errors mostly occurred when two players of the same team were in
close proximity. Although the current system still needs a manual correction for
the occasional tracking error, it is an efficient help in the time consuming task of
collecting player location data from a football match. In order to build an error-free
automatic tracking system, some improvements will be required. One such
improvement would involve the ability to process data from multiple cameras
installed in various locations.
References
Abdel-Aziz YI, Karara HM (1971) Direct linear transformation from comparator coordinates into
object space coordinates in close-range photogrammetry. ASP symposium on close range
photogrammetry, pp 1–19
Kanzaki N, Ariki Y (2002) Player tracking in soccer game based on normalized correlation method
using divided template, pp. FIT Forum 2002, pp 189–190
Lim W, Haseyama M (2008) A note on player tracking by using level-set method based on color
component in soccer video. IEICE Tech Rep 107(489):67–70
Oosaka S, Honda M (2009) Automatic detection of player positions from soccer video image
based on the color information, 08-2P-P29. Meeting of Japan Society of Physical Education
Shiozaki T, Ohira S, Honda M, Shirai K (2004) Soccer video indexing based on automatic signal
processing. FIT 2004, I-046
Tokikura M, Nishihara A (2005) Development of realtime player tracing system with adaptive
background renewal. IEICE Tech Rep 105(146):1–6
Ueda F, Honda M (2014) The causal relationship between dominant region and offense-defense
performance –focusing on the time of ball acquisition. Football Sci 11:1–17
Chapter 29
Conceptualization of Coaching Process
and Coaching Practice
Hiroyuki Horino
Abstract As the coaching process includes many elements, in the sports psycho-
logical research arena, a wide variety of coaching themes have been explored.
Expert coaches utilize their acquired knowledge and practice appropriate decision-
making and behavior depending on the situation. Accordingly, investigating the
coaching style of experienced and successful coaches can provide a model to
facilitate the development of less experienced coaches. In this chapter I review
the “conceptualization studies of the coaching process”, largely dealing with
utilizing experienced coaches to create a coaching model (Côté et al. J Sport
Exerc Psychol 17:1–17, 1995a). Subsequently, I will discuss future challenges in
this area and related areas of research. It is essential to integrate various conceptual
models with research developments and to utilize this body of knowledge to create
a practitioner-oriented model that will be applicable to both coaching practice and
athlete development.
Keywords Coaching process • Cognitive conceptualization • Coaching model •

Expert coach
29.1 Introduction
In recent years, competitive sports have become far more specialized. Nowadays, in
order to maintain competitiveness at an international level, short term developmen-
tal plans are not adequate. A population of grass roots athletes needs to be
continually nurtured and helped to develop.
Coaches can play critical roles over a broad range of themes in the process of the
popularization of sports and the development of athletes. Therefore, coaching
education programs were designed to facilitate the ability of coaches have been
developed, and coaching education programs are held in many countries. Unfortu-
nately, recent research indicates that these coaching education programs have little
impact on actual coach development (Cushion et al. 2003; Nelson et al. 2006;
H. Horino (*)
e-mail: horino@waseda.jp

368 H. Horino
Dieffenbach et al. 2011; Gearity 2012). In particular, the structure and content of
the coaching education courses often lack content on important psychological and
pedagogical approaches for talent development applicable to their actual practice
(Nash and Sproule 2012). Nash and Sproule point out that: “coach education
courses are able to deliver the sport specific content but generally are not able to
fulfill the coach’s requirements when it comes to other aspects of coaching, for
example, sport psychology or pedagogy” (p. 30).
Coaches wanting to develop their knowledge base and coaching capabilities
would really like to grasp and understand “the tacit knowledge” of expert coaches.
However, the coaching philosophies and knowledge of expert coaches have not been
fully integrated and categorized. Much of their knowledge is implicit and has been
acquired from their various experiences. While this knowledge is valuable and to
some extent irreplaceable, it is lost when the coaches retire or change occupations.
29.2 Sport Psychological Research on Coaching
29.2.1 Focus for Coaching Research
As the coaching process includes many elements, sport psychological research on

coaching has investigated a wide variety of themes. These include coaching behav-
ior, knowledge, interaction between coaches and athletes, and expertise.
Gilbert and Trudel (2004), in a review, noted that sport psychological research
on coaching increased after 1970, and that the related studies increased consider-
ably in 1990s. Research on “coaching behavior”, which demonstrates how coaches
actually put their knowledge and experience into practice, have been a main theme
and have shown a consistent increase since the 1970s. After the late 1970s, the
percentage of coaching science articles focusing on “characteristics and career
development” increased, but there has been a recent decline. The studies on
“thought”, which include the cognitive processes of expert coaches, have increased.
Topics in this area include decision-making, knowledge, and expertise. According
to Nash and Martindale (2012), from 1993 to 2009, the most researched aspects in
the coaching arena involved the developmental process, coaching behavior, skills,
and decision-making. In addition, studies on expert coaches increased after 2000.
The developmental process wherein coaches become expert involves all aspects of
learning. In addition to formal coaching education courses, information garnered
from casual and informal sources is quite important.
29.2.2 Research Methods
Earlier studies were focused on quantitative analysis, and typically utilized some
form of the questionnaire approach. In recent years, the most popular method has
29 Conceptualization of Coaching Process and Coaching Practice 369
shifted to qualitative methods, which often involve an interview. Such interviews

can be in depth, open-ended, and semi-structured. When using the qualitative
method, data analysis is often performed in the following order: After interviews,
the narrative is transcribed verbatim, and the transcribed data are inductively
analyzed according to the specific procedures and techniques. These can be based
on an applied theory, such as the grounded theory (Strauss and Corbin 1990).
Another procedure involves the inductive analysis process, in which conceptuali-
zation is proceeds as follows: First, meaningful units are extracted from the
interview transcripts (coding meaningful text segments, or creating tags). Second,
units of similar meaning are regrouped into those with the same properties, cate-
gories, and components in stages. Based on such an analysis, researchers can
evaluate the interactions between the components, and construct a conceptual
model or theory inductively (Cote et al. 1993).
The mixed method approach has become the next most popular. This method
involves the observation of coaches during training sessions, as well as the use of
questionnaires and some form of interview. Mixed methods are thought to be
beneficial in mixing the valuable aspects of both quantitative and qualitative
designs and allow an enhanced triangulation, which provides a more robust devel-
opment of theory. The mixed approach also has the potential to allow a more
comprehensive understanding of the research situation (Nash and Martindale
2012). Because of these advantages many researchers in sport psychology have
utilized a combination of quantitative and qualitative methods to investigate vari-
ous domains of coaching with.
Along with a transformation of research topics and methodology, coaching has
attained a more distinct status and, “Coaching is no longer a subset of physical
education or sport psychology but is rather an established vocation for research”
(Abraham and Collins 2011, p. 366).
In addition, “the conceptualization of the coaching process” using a qualitative
analysis has become the main theme of sport psychological researches on coaching.
One purpose of this chapter is to review the “conceptualization studies of the
coaching process”. This will be based mainly on a model (Côté et al. 1995a) of the
expert coach. I will also discuss future challenges in research related to coaching.
29.2.3 Leadership Behavior and Cognitive Processes

of Coaches
Over the last three decades, researchers investigating coaching have made an effort
to clearly define, and increase the quality of, coaching. Early research in this
domain aimed to accurately describe the behavior of expert coaches in order to
transmit a desirable coaching style to novice coaches. The purpose of later studies
shifted to an investigation of “the cognitive process of coaching”. In the coaching
process a coach makes good decisions and utilizes effective actions in three areas:
training, competition, and planning.
370 H. Horino
Chelladurai (1990) described the leadership qualities of coaches with the con-
ceptual “Multidimensional model”.
In his model, coaching behavior is influenced by three factors: situational,
leader, and member characteristics. In addition, the coaching behavior is classified
into three types: required, actual, and preferred. Finally, actual behavior was
prescribed by two other behaviors, and three behaviors were united in their effect
on the performance and satisfaction of the members.
29.3 Coaching Process and Application of Models
After Chelladurai (1990), many researchers have investigated the knowledge and
behavior of expert coaches to conceptualize their coaching process. In their studies,
researchers utilized the approach of breaking down the complicated cognitive
processes of practical coaching into more simple constituent components, and
then integrated these constituents in order to build a conceptual model.
Côté et al. (1995a) support the validity of perceptual modeling as a way to explain
the coaching process as follows: “From a cognitive perspective, the modeling system,
elaborated in an attempt to explain how expert coaches utilize knowledge to develop
elite gymnasts, was consistent with theoretical definitions of mental models (Glaser
1987; Holyoak 1984; Johnson-Laird 1983). According to Côté et al. (1995a), gener-
ally, these authors suggested that “mental models were specific knowledge structures
that were constructed mentally to represent various situations” (p. 13). Moreover
Kitamura et al. (2005) noted, “In the coaching scene, . . . A coach adopts a coaching
behavior based on a prospect how a player will recognize the behavior and how
coaching behavior will influence the athlete’s performance. The frame deciding the
behavior can be explained by mental model (Johnson-Laird 1983)” (p. 18).
As the above authors have noted, it is very effective to construct mental models
of expert coaches to comprehend the processes underlying their methods of
coaching. These processes allow expert coaches to utilize the knowledge that
they have acquired through various experiences, and aids them in the practice of
appropriate behavior and good decision-making as required by the situation.
Accordingly, constructing a model that depicts their coaching style can be a very
effective way of facilitating the development of coaches desirous of improving.
29.3.1 Research for “Conceptualization of Expert Coaches”
Côté et al. (1995a, b) in a pioneer study, explored cognitive processes of expert

coaches utilizing the qualitative research method. In the wake of their studies, many
researchers came to investigate the coaching process of various individual and team
sports. Such studies are shown in Table 29.1.
Table 29.1 Researches about conceptualization of coaching process

Author Theme sport Components
Côté et al. Coaching model Gymnastics Competition, training,
(1995a) (knowledge) organization,
Coach’s personal characteristics,
Gymnast’s personal characteristics
and level of development, contex-
tual factors
Côté et al. Knowledge Gymnastics Minimally involved in
(1995b) competition,
Coach involvement in training,
intervention style, technical skills,
mental skills,
Simulation: Teaching progres-
sions, being supportive, helping
athletes to deal with stress
Kitamura et Coaching model Football Training, motivating, supporting.
al. (2005)
Koga and Coaching model Soccer Development of life skill,
Horino Development of performances,
(2012) Coaches’ behavior and approach
Katoh and Coaching model Soccer Sincere attitudes for anything
Horino Supporting humanistic education
(2011)
Nash et al. Coaching practice Multi sports: Long-term approach,
(2011) individual & team Authentic coaching environment
sports Creating a learning environment
Quality and quantity of training
sessions
Bennie and Coaching philosophy Rugby, cricket Player development on and off the
O’Connor field,
(2010) Role of the coach, develop the
player and the person,
Educate the players, not purely
focused on results
Bennie and Coaching model Rugby, cricket The coach, coaching skill, the
O’Connor environment
(2011)
Abraham The coaching process: Multi sports: Roles, goals, typical actions,
et al. (2006) the coaching individual & team Required knowledge,
schematic sports Support for the schematic,
Factors influencing development.
Côté and Coaching behavior Rowing Plan proactively, create a positive
Sedgwick training environment
(2003) Facilitate goal setting, Build
athletes’ confidence
Teach skills effectively,
Recognize individual differences
Establish a positive rapport with
each athlete.
(continued)
372 H. Horino

Author Theme sport Components
Bloom et al. Routines Basketball, field Game-day routines for coach and
(1997) hockey, ice team,
hockey, Coaches’ emotions and behaviors,
volleyball Team meeting, game evaluation
Bloom and Personal Basketball, field Desire to learn,
Salmela characteristics hockey, ice Ways of acquiring knowledge,
(2000) hockey, Personal approach to coaching
volleyball
d’Arripe- Effectiveness of inter- Judo Stimulating interpersonal rivalry,
Longueville actions between Provoking athletes verbally,
(1998) coaches and athletes displaying indifference
Entering into direct conflict,
showing preferences
Developing specific team cohesion
Irwin et al. Origin of coaching Gymnastics Mentor coaches, trial and error/
(2004) knowledge experimentation,
Past experiences, coaching
courses, squad sessions, observa-
tion, coaching manuals, foreign
coaches,
Côté and Definition and Artistic Coaches’ knowledge,
Gilbert required for expertise gymnastic Athletes’ outcomes,
(2009) Coaching contexts
29.3.2 Early Researches
In the pioneer study of Côté et al. (1995a), they conducted semi-structured inter-
views, in-depth and open-ended, with 17 expert gymnastics coaches who were
involved producing Olympic level athletes. The goal of the investigators was to
conceptualize the coaches; knowledge. They analyzed their qualitative data utiliz-
ing the grounded theory methodology (Glaser and Strauss 1967; Strauss and Corbin
1990). As shown in Fig. 29.1, six components emerged from their analysis. As a
result, they proposed a model of the cognitive processes of expert coaches, entitled
the “Coaching model”. Côté et al. proposed that the coaching process (knowledge)
was categorized into three central components: competition, training, and organi-
zation. Moreover, they denoted three variables. These were the coach’s personal
characteristics, the athletes’ personal characteristics and level of development.
They also proposed some contextual factors, which were defined as peripheral
components. Their stated goal was to aid in the development of athletes by creating
a model describing how expert coaches function Fig. 29.2.
Using a similar qualitative analysis, Côté et al. (1995b) also investigated the
knowledge base of expert coaches. They revealed that expert coaches were mini-
mally involved with the athletes in competition and but in training they were
Goal: Developing Athletes
Athletes’ Personal
Coach’s Personal
Characteristics and Level
Characteristics
of Development
Coach’s Mental Model of

Athletes’ Potential
Competition Training
Training
Contextual Factors
Fig. 29.1 The coaching model (Côté et al. 1995a)
Development of Development of
the sport the humanistic
performance performance
Supporting (Organization)
Fig. 29.2 The key components in the coaching process of expert coaches
involved with teaching progressions, being supportive, and helping athletes to deal
with stress. Succeeding researchers generally utilized the procedure of Cote
et al. (1995a, b) to investigated, identify, and conceptualize the coaching processes
of expert coaches. The conceptualization of the cognitive process was expanded to
team sports by Bennie and O’Connor (2011).
Kitamura et al. (2005) demonstrated that coaching model (coaching mental
model) of expert high school youth football coaches was comprised of three
374 H. Horino
categories: training, motivating and supporting. Koga and Horino (2012) performed
a study to compare professional youth football clubs to high school football clubs.
They found no difference in the coaching model that described coaches at the two
levels. The coaching model of expert youth coaches consisted of three categories:
development of life skill, development of performance, and the coaches’ behavior
and approach. Katou and Horino (2014) also investigated the coaching model of
local youth coaches at the final 16 in a prefectural tournament. The coaching model
for these coaches consisted of three components: diligent attitudes for football,
supporting, and humanistic education. Although the components’ names were
different among the three studies, the youth coaches all regarded training (devel-
opment of the sport performance), education (development of a humanistic perfor-
mance), and support (organization of environment) as key components of their
coaching process. “The humanistic performance” involves a holistic embraces of
abilities that allow the players to make emotional, psychological, and social adjust-
ments which allow them to adequately deal with various situations and in the
achievement of their goals. From the above conceptualizations, it is clear that the
extracted components of each coaching model are very similar to each other.
Despite differences of the club attributes and the competition level, the coaching
models of youth coaches have similar characteristics in that they regard the
development of a humanistic performance as an essential part of youth develop-
ment along with the sport performance. The expert coaches instructed and
supported their athletes in the development of both athletic and humanistic
performance.
Nash et al. (2011) investigated the coaching practice of expert coaches in several
sports. Their observations indicated that the experts’ coaching practice consisted of
four themes: a long-term approach, an authentic coaching environment, the creation
of a learning environment, and the quality and quantity of training sessions.
Furthermore Bennie and O’Connor (2010, 2011) explored professional coach and
player perceptions of effective coaching. They created an “Effective Coaching
Model (ECM)” for professional sports that contained three major concepts: the
coach, coaching skills, and the environment. Bennie and O’Connor (2010)
described the coach as follows: “coaches in these professional settings develop
programs to assist players in acquiring on- and off-field skills. In addition to this,
there is a tendency to focus on learning and improvement as opposed to a win-at-all-
costs attitude. These philosophies highlight elements of a Humanistic approach to
coaching which focuses on the total development of the person” (p. 310). Thus, the
professional coaches as well considered humanistic development to be one of the
most important elements of their coaching.
The research summarized in this section clearly indicates that, irrespective of
competition level, club affiliation, or the athletes’ age, expert coaches place a strong
emphasis on “humanistic education” in their approach to coaching. These coaches
are eager to develop not only good athletes but also good people.
29.3.3 Schematic Model: Structure of Expert Coaches
Abraham et al. (2006) investigated coaching process of expert coaches and con-
cluded that it could be described by six general categories: roles, goals, typical
actions, required knowledge, support for the schematic, and factors influencing
development. They also offered “the Coaching Schematic” as a model which
organized the coaches’ knowledge structure. They held that both implicit and
explicit aspects of their schematic arrangement were practiced by most expert
coaches. “But it is important to note that this knowledge is used to understand
their athletes better. . . the schematic does indeed accurately match the thought
processes and decision making of expert coaches” (p. 562). They declared, “Our
argument would be that, given the broad range of concepts and conceptions and
knowledge within the schematic, it should represent a good starting point, through
context-specific targeting of the most relevant factors, for the design of effective
coach development curricula and practices for volunteer through to expert coaches”
(p. 563).
In the terms of conceptualizing the cognitive process of the coaching process as a
frame, their scheme would seem to be very effective. As there are still few studies
that evaluate the Schematic Model, further studies are necessary.
29.3.4 Cognitive Conceptualization of Other Specific

Subjects
Côté and Sedgwick (2003) interviewed elite rowing athletes and their coaches.
They conceptualized coaching behavior of the expert coaches as characterized by
seven components. The components are organized into a three circle hierarchy
which corresponds to the different types of interaction between the coach and
athlete: environmental maintenance, technical knowledge, and an interpersonal
skill. They declared that the various components were interdependent and interac-
tive within the same level and across different levels of behavior.
Bloom and Salmela (2000) investigated the personal characteristics of expert
coaches in team sports. In this study, the personal characteristics of expert coaches
were condensed into three categories: desire to learn, ways of acquiring knowledge,
and personal approach to coaching. These investigators felt that the coaches’
personal characteristics might be more important than a simple understanding of
their technical and tactical skills.
d’Arripe-Longueville (1998) invested the effectiveness of interactions between
coaches and athletes of judo. They revealed that the strategies of expert judo
coaches could be categorized by six strategies: stimulating interpersonal rivalry,
provoking athletes verbally, displaying indifference, entering into direct conflict,
developing specific team cohesion, and showing preferences. They demonstrated
that the coach-athlete relationships of expert Judo coaches were different from the
376 H. Horino
expert coaches in other sports in that Judo coaches used negative feedback (for
example, encouraged rivalries) and tended to not provide social support for the
athletes. This study suggests that the coaching process may be different for each
sport. Further investigation in this area will determine the extent that such
differenced occur.
29.3.5 Expertise of Coaches
The coaches develop their coaching ability through formal (coach education pro-
grams), nonformal (small group seminars) and informal (daily experiences and
exposure to the environment) (Nelson et al. 2006). Expert coaches can effectively
handle a broad range of situations that frequently occur in the practice of coaching.
In addition to scientific knowledge, practical knowledge acquired through experi-
ence is required in many coaching situations. Piggott (2012) noted that “coaching
knowledge and practices, in both elite and non-elite coaches, are derived over-
whelmingly from informal and non-formal sources” (p. 538). Therefore it is
important to identify the factors required for developing the expertise of expert
coaches and to use this information to aid in the development of novice coaches.
Irwin et al. (2004) conceptualized the process of knowledge acquisition by
expert coaches. They demonstrated that most expert coaches “identified that
knowledge acquisition was facilitated mostly through interactive coaching clinics
and mentorships that promoted critical inquiry and active experimentation”
(p. 425). They felt that experts developed their own knowledge from practical
experience, including trial and error, gained from earlier coaching involvement.
Côté and Gilbert (2009), in an effort to define the coaching effectiveness and
expertise, and proposed that the definition was comprised of three components:
coaches’ knowledge, athletes’ outcome, and coaching context. They concluded that
“a definition of coaching effectiveness and expertise should integrate these three
components by considering the interaction of a coaches’ knowledge and athletes’
outcomes in specific coaching contexts.” (p. 309). Furthermore they demonstrated
that coaching effectiveness and expertise should require professional, interpersonal,
and intrapersonal knowledge and should develop the athletes’ 4 C’s: competence,
confidence, connection, and character. As there exist four sport contexts depending
on the players’ age, they insisted that “Coaching effectiveness should be defined
according to how coaches meet their athletes’ needs and help them fulfill their
goals, as defined by the specific coaching context” (p. 315). As shown in Table 29.1,
four themes were proposed by Nash et al. (2011) as the coaching practice of expert
coaches. These are similar to those of Côté and Gilbert (2009).
Nash and Sproule (2011) compared an expert swimming coach to a novice coach
relative to the construction of their knowledge and the practical application of their
learning experiences. The expert coach acquired experiences in a more efficient
way than did the novices. They suggested that it was very important for novice
coaches to ask questions of both their colleagues and themselves in order to develop
their coaching ability.
As seen above, recent research has focused on a wide range of subjects regarding
the coaching process and how the skills of expert coaches can be conceptualized
and placed into cognitive categories. Such conceptualization can make it easier for
novices to understand the coaching process of expert coaches. However, more
investigation is needed to better understand such a complicated skill more
completely.
29.4 Discussion
In the last few decades, various conceptual models of the complicated coaching
process have been proposed. As the result of overviewing recent studies, it is
considered that expert coaches place significant importance on “development of
the sport performance”, “development of the humanistic performance”, and
“supporting (organization)”.
In the coaching practice, however, competition level, age, and many environ-
mental factors influence the coaching process. In addition, conceptual models of
coaching have been constructed from various theoretical grounds including lead-
ership, expertise, coach-athlete relationships, motivation, and education (Côté and
Gilbert 2009). Because of such a highly-diversity in the coaching process, the focus
in recent studies has been very wide-ranging. As a result, although many models
and conceptual organizations have been proposed, each model tends to be effective
only in a limited situation and sport context. Bennie and O’Connor (2011) noted,
“Many of the existing models have not attempted to symbolize the entirety of
coaches’ actions but rather provide representations of key parts of the coaching
process” (p. 98). Thus, it has not yet been possible to establish a general concep-
tualization of effective coaching in various coaching situations (Cushion 2007;
Vella et al. 2010; Abraham and Collins 2011). Côté and Gilbert (2009) noted that
“in order to facilitate integration of findings across diverse lines of research,
discussion of results within individual studies should be re-framed within an
integrative theoretical framework of coaching effectiveness”(p. 318). Thus the
need for theoretical integration across various conceptual models has been pointed
out. A fully integrated model has to be able to apply to most or ideally all sports
contexts. However, this challenge has yet to be met.
Nash and Martindale (2012) observed that “We need to establish some effective
criteria and base characteristics of expertise from which participant selection, study
focus and eventual interventions may evolve” (p. 992). As they state, the criteria for
identifying expert coaches are lacking. In addition, so as to facilitate validation of
data, researchers will need to use the triangulation method. For example, when
interview data are combined with measures from systematic observation or ques-
tionnaires, the mixed methods can improve the validity of the analysis. In order to
378 H. Horino
resolve these methodological and procedural challenges, further studies which

utilize triangulation to investigate a particular situation are needed.
Finally it is most important for researchers to focus on both athletes and
practitioners. Instead of a model suggested by strictly scientific approaches,
coaches require a more practitioner-oriented model. Such a practical model can
enhance the applicability of the complicated interaction between coaches and
athletes, and can facilitate the development of both the coaches and the athletes.
In sum: First, it is important to integrate various conceptualized models by
means of an accumulation of further research. Second, it is essential to propose a
practitioner-oriented model that is applicable to both coaching practice and athlete
development.
References
Abraham A, Collins D (2011) Taking the next step: ways forward for coaching science. Quest
63:366–384
Abraham A, Collins D, Martindale R (2006) The coaching schematic: validation through expert
coach consensus. J Sports Sci 24:549–564
Bennie A, O’Connor D (2010) Coaching philosophies: perceptions from professional cricket,
rugby league and rugby union players and coaches in Australia. Int J Sport Sci Coach
5:309–320
Bennie A, O’Connor D (2011) An effective coaching model: the perceptions and strategies of
professional team sport coaches and players in Australia. Int J Sport Heal Sci 9:98–103
Bloom G, Salmela J (2000) Personal characteristics of expert team sport coaches. J Sport Pedagog
6:56–76
Bloom G, Durand-Bush N, Salmela J (1997) Pre- and postcompetition routines of expert coaches
of team sports. Sport Psychol 11:127–141
Chelladurai P (1990) Leadership in sports: a review. Int J Sport Psychol 21:328–354
Côté J, Gilbert W (2009) An integrative definition of coaching effectiveness and expertise. Int J
Sport Sci Coach 4:307–323
Cote J, Sedgwick W (2003) Effective behaviors of expert rowing coaches: a qualitative investi-
gation of Canadian athletes and coaches. Int Sport J 7:62–78
Cote J, Salmela J, Baria A, Russel S (1993) Organizing and interpreting unstructured qualitative
data. Sport Psychol 7:127–137
Côté J, Salmela J, Trudel P, Baria A (1995a) The coaching model: a grounded assessment of expert
gymnastic coaches’ knowledge. J Sport Exerc Psychol 17:1–17
Cote J, Salmela J, Russell S (1995b) The knowledge of high-performance gymnastic coaches:
competition and training considerations. Sport Psychol 9:76–95
Cushion C (2007) Modelling the complexity of the coaching process. Int J Sport Sci Coach
2(4):395–401
Cushion C, Armour K, Jones R (2003) Coach education and continuing professional development:
experience and learning to coach. Quest 55:215–230
d’Arripe-Longueville F (1998) The perceived effectiveness of interactions between expert French
judo coaches and elite female athletes. Sport Psychol 12:317–332
Dieffenbach K, Murray M, Zakrajsek R (2011) The coach education internship experience: an
exploratory study. Int J Coach Sci 5:3–25
Gearity B (2012) Poor teaching by the coach: a phenomenological description from athletes’
experience of poor coaching. Phys Educ Sport Pedagog 17:79–96
Gilbert W, Trudel P (2004) Analysis of coaching science research published from 1970–2001. Res
Q Exerc Sport 75:388–399
Glaser R (1987) Learning theory and theories of knowledge. In: deCorte E, Lodewijks H,
Parmentier R, Span P (eds) Learning and instruction. Pergarnon, Toronto, pp 397–414
Glaser G, Strauss A (1967) The discovery of grounded theory: strategies for qualitative research.
In: The discovery of grounded theory: strategies for. Aldine, New York
Holyoak K (1984) Mental models in problem solving. In: Anderson JR, Kosslyn SM (eds)
Tutorials in learning and memory: essays in honor of Gordon Bower. W.H. Freeman, San
Francisco, pp 193–218
Irwin G, Hanton S, Kerwin D (2004) Reflective practice and the origins of elite coaching
knowledge. Reflective Pract 5:425–442
Johnson-Laird P (1983) Mental models: towards a cognitive science of language, inference, and
consciousness, 6th edn. Harvard University Press, Cambridge
Katou A, Horino H (2014) The difference of competition level and coaching mental model -the
case study of high school soccer coach in X prefecture-. Sport Sci 11:146–158
Kitamura K, Saito S, Nagayama T (2005) How do expert coaches enhance athletes’ and team
performance?: construction of a coaching mental model of expert high school soccer coaches
in Japan. Jpn J Sport Psychol 32:17–28
Koga Y, Horino H (2012) The study of J-league club youth coache’s philosophy -through
comparison with the case of high school soccer club coache’s philosophy-. Sport Sci
10:173–182
Nash C, Martindale R (2012) Parameterising expertise in coaching: past, present and future. J
Sports Sci 30:985–994
Nash C, Sproule J (2011) Insights into experiences: reflections of an expert and novice coach. Int J
Sport Sci Coach 6:149–161
Nash C, Sproule J (2012) Coaches perceptions of their coach education experiences. Int J Sport
Psychol 43:33–52
Nash C, Sproule J, Horton P (2011) Excellence in coaching: the art and skill of elite practitioners.
Res Q Exerc Sport 82:229–238
Nelson L, Cushion C, Potrac P (2006) Formal, nonformal and informal coach learning: a holistic
conceptualization. J Sport Sci Coach 1:247–259
Piggott D (2012) Coaches’ experiences of formal coach education: a critical sociological inves-
tigation. Sport Educ Soc 17(4):535–554
Strauss A, Corbin J (1990) In: basics of qualitative research: grounded theory procedures and
techniques. Sage, Newbury/Newbury Park
Vella S, Oades L, Crowe T (2010) Review: the application of coach leadership models to coaching
practice: current state and future directions. Int J Sport Sci Coach 5:425–434
Index
A Anterior cingulate cortex (ACC), 7,

Abductor halluces (AH), 86–88 28–30
Acceleration, 274, 277, 281–291, 301, 353 Anthropometric, 141, 146, 318–319
Accurate coordination, 303 Artificial gravity, 239
Accurate hitting, 336, 351 Asian, vi, 107, 110–111, 116–120
Achilles tendon, 84, 192 Athletic performance, 25–33, 106,
α-Actinin-3 (ACTN3) polymorphisms, 112, 117
106, 112–119 ATP, 106, 107, 110, 111, 151, 152
Actin-myosin interaction, 208 Attached cross bridges, 201, 208
Action monitoring, 30 Automatically, 50, 56, 354, 360, 362
Activation coupling, 42–44, 46 Automaticity, 27, 50, 57
Actual hitting accuracy, 343, 345–350 Automatic tracking, 353–365
Adaptive method, 362 Azimuth, 325, 326
Aerobic exercise, 140
Aerobic fitness, 91–94, 96–99
Afferent feedback, 28 B
Afferent nerves, 84 Backspin, 329, 337, 349
African, 107, 109, 111–112 Badminton players, 73, 75
Afterdrop, 316, 321 Balance control, 26, 29–31, 33
Aging society, 229 Balance performance, 33
Agonist–antagonist, 69 Balance-related brain dynamics, 29
Alpha, 5–7, 15, 26–28, 32, 33, 154, 189, 231, Ball spin rate, 324, 327, 330, 332, 349
233, 235, 237, 238, 343 Baseball, ix, 4, 9, 49, 50, 323–332,
desynchronisation, 27, 33 335–351
rhythms, 6 Bat, 335–343, 345, 348–350
synchronisation, 27, 28 Bat control, 336, 339, 341, 343–345,
Altered neuromuscular control, 82 347–349, 351
American football, 126, 140–142, 144, 147, Batters’ abilities, 336
178, 330 Batters’ intention, 348
Anatomical cross-sectional area (ACSA), Batting average, 336
176–182 Biarticular, 176, 181
Angiotensin I-converting enzyme (ACE), Binarization, 354, 355, 357, 361–364
106, 112–120 Biochemical assessment, 147
Angular momentum, 289, 290, 295–296, Blood, 70, 116, 133, 147, 152, 153, 167, 214,
298, 301 219, 317–319

K. Kanosue et al. (eds.), Sports Performance, DOI 10.1007/978-4-431-55315-1
382 Index
Blood oxygenation level-dependent (BOLD) Conceptualize the coaching processes,

signals, 7, 72 367–378
Body composition, 128, 129, 133, 140–141, Conceptual models, 370, 377
143, 144, 146, 320 Conditioning contraction, 199–209
Body fat, 129, 134, 135, 140, 141, 143–147, Consistency, 8, 131, 294–295, 298, 299
316, 319, 320 Contingent negative variation (CNV),
Body lean, 282, 285, 290 63, 64
Body size, 127–130, 140, 143 Continual adjustments, 301
Bone stability, 227, 241 Contraction type, 159, 200, 208
Bradykinesia, 68 Core-temperatures, 316, 317, 320, 321
Brain, x, 3–10, 14, 15, 25–33, 45, 50, 52, 64, Cortical activation, 6, 26, 27, 33, 68, 73
70, 72, 75, 88, 131, 230, 231, 233–235, Cortical mapping, 31
238, 240, 241, 316 Corticospinal, 8, 15, 17–19, 53, 70, 71
activity, 3–10, 13–19, 26, 29–31, 33, 68, 72, Cost of transport, 249, 250, 252, 253, 256
86–87, 238 Countermeasure, 229, 230, 319
cortical function, 230, 233, 241 Countermovement, 187–195
Cross-country skiing, 247–279
Cross-sectional study, 176–181
C Curveball, 323–325, 327, 328, 330–332
Ca2+ concentration, 200–202 Cut fastball, 325, 327, 328, 330, 331
Calf-raise exercise, 214, 215, 217–220 Cyclists, 109, 177–179, 181, 182
Calmodulin, 201
Camera, 251, 264, 283, 302, 306, 325,
337–340, 342, 359–360, 365 D
Carbohydrate, 82, 145, 146, 151–155 Dehydration, 79, 81, 82, 315, 317–318
Cardiovascular disease, 141–143, 147, Diet, 145–147, 151–155, 319
155, 241 Dietary assessment, 146, 147
Cardiovascular system, 224, 227–228 Dietary management, 125–135
Central nervous system (CNS), 25, 50, 54–56, Dietitians, 145–147
68, 82, 84, 85, 88, 230, 231, 239 Direction of spin axis, 324, 325, 327, 328,
Central origin theory, 82–84 330–332
Central pattern generators (CPGs), 51–53 Direct linear transformation (DLT) method,
Change-up, 325, 327, 328, 331 264, 338, 354, 359, 361
Channel swimming, 314–316, 318, 319 Dismount, 295, 296, 298, 302, 303
Childhood fitness, 91–99 Displacement, 39, 41, 50, 266, 267, 338, 339,
Children, v, vi, x, 92–94, 96–99 341–343, 345–351
Classic technique, 248 Distance, ix, 5, 27, 33, 62, 80, 82, 109, 116,
Closed-loop control system, 60 117, 119, 128, 158, 169, 176–179, 192,
Co-contraction, 69, 75 205, 216, 278, 287, 291, 314, 316, 317,
Cognitive control, 91–93, 95, 97–99 319, 336, 338, 339, 342, 343, 345–347,
Cognitive flexibility, 92–94, 97–99 350, 353, 355, 359, 360
Cognitive function, x, 9, 91–94, 99 Dorsolateral prefrontal cortex (DLPFC),
Cognitive performance, 96, 224, 229–241, 370 14, 72
Cognitive process of coaching, 368–370, Dose-response relationship, 239, 241
372, 375 Drummers, 74
Color threshold, 357, 361–364 Duty factors, 289
Competitive, 117, 128, 142, 159, 176, Dystonia, 67–70
179–182, 259, 260, 263, 367
Competitive swimming, 306, 318–320
Compression garments, 214, 218, 219 E
Compression stockings, 213–220 Eccentric, 159, 162, 165, 167, 189, 203, 207,
Concentric, 159, 162, 165, 167, 189, 200, 203, 208, 308
207–209 Effective impact area, 348, 351
Index 383
Effective leg angle, 285, 287 Female athletes, 125, 128, 129, 131–134, 142,
Electrical stimulation, 80, 81, 84–86, 200–201, 273, 278, 279, 316, 318, 319
203, 205 Fibromyalgia, 162
Electroencephalography (EEG) Field coordinates, 354, 359, 361, 362, 365
biofeedback, 31 Fingers, 16, 18, 28, 37–47, 330–332
neurofeedback, 31–33 First person perspective imagery, 13
Electroencephalography/ Flanker task, 92, 93
electroencephalogram (EEG), 3–7, 9, Flight phase, 262–268, 273–279, 294,
14–16, 25–27, 29–33, 63, 68, 69, 72–73, 295, 301
231, 238 Flutter kick, 305–312
Electrolyte disturbances, 81, 82 Food choices, 145, 146
Electromyography/electromyograph/ Football game, 353–365
electromyographic (EMG), 9, 51–54, Football match, 359, 365
68, 69, 73, 74, 83–88, 190–193, Force-generating capacity, 199–200, 218,
216–219, 306, 307, 310 287, 289, 291
Elevation, 116, 130, 152, 325, 326, 330 Force orientation, 288, 290
Endurance, 54, 109–111, 115, 116, 126, Force-velocity relationship, 194, 202, 208,
132, 135, 151–155, 162, 248, 249, 252, 258
314, 317 Forkball, 323, 325, 327, 328, 330, 331
Endurance performance, 106, 108, 110, 116, Four-legged gait, 250
119, 151–155, 317 Four-seam, 325, 327–332, 348
Energy balance, 126–127, 134, 135, Free skating technique, 248
143, 145 Free technique, 248
Energy requirements, 127, 135, 249 Frontal midline theta, 28
Error detection, 30, 33 Full HD video cameras, 360
Error-free automatic tracking, 365 Functional alterations, 54, 55
Error-related negativity (ERN), 30 Functional magnetic resonance imaging
Estimation equation, 133, 134, 146 (fMRI), 3, 4, 7, 8, 14, 16, 17, 28, 29, 68,
Ethnicity, 117, 119, 145 72, 238
European, 107, 109, 111, 116, 117 Functional networks in the spinal cord, 52
European Space Agency (ESA), 225–229, Functional reorganization, 50, 54, 56
231–236, 239
Event-related desynchronisation/
synchronisation, 15, 27 G
Event-related desynchronization (ERD), GABA, 84, 85
5, 15, 27 GABAergic connections, 71
Event-related potentials (ERPs), 15, 26, 63 Gait patterns, 248–250, 252, 256–257, 259
Exercise Gamma power, 28–30
preference hypothesis, 239 Gastrocnemius, 190, 191, 214, 215, 217–219,
preferences, 241 306, 307
Expert performance, 25 Generalized motor programs (GMP), 61–64
Extreme environments, 224 Genetic, x, 105–121, 143–145
Genetic factors, 105–108, 119, 121, 135
Giant circle, 295–298, 300
F Glycemia, 320, 321
Fascicle length, 176, 190, 191, 193, 194, Glycogen, 151–155
204–209 Golf, 7–9, 32, 33
Fastballs, 323–325, 327–332, 337, 339, 342, Golf putting, 4–7, 9, 10, 27, 28, 32, 33
348, 349, 351 Golgi tendon organs, 82
Fat, 129, 133, 135, 140–147, 151–155, 316, Go/No-go task, 72
319, 320 Graduated pressure profile, 213–220
Fat-free mass (FFM), 126–135, 140, Grasping, 38, 45–47, 368
144–146 Grip, 70, 215, 263, 264, 269, 331–332,
Feedback control, 298, 300, 303 338, 348
Feedback time delay, 301 Grip-and-lift tasks, 70
384 Index
Ground reaction force, 50, 193, 268, Ipsilateral, 17, 29, 63, 72, 73, 83
281–290 Isokinetic, 158, 162, 165–167, 204
Gyroball, 330 Isolated environment, 230
Isometric, 72, 73, 108, 162, 166, 200,
202–208, 219
H
Hand, 5, 7, 15, 17, 18, 27, 28, 39, 45–47, 61, 63,
70, 71, 73–75, 107–111, 116, 117, 119, J
128, 131, 141, 146, 159, 179, 181, 182, Japanese, v, 95, 110–112, 116–117, 119,
200, 208, 219, 277, 294, 301, 306, 125–135, 140–145, 176
317–319, 325, 330–332, 338, 349, 351, Japanese diet, 145
353, 362 Joint angle-torque (force) relationship, 169
Haplogroup, 107, 109–112 Joint stiffness, 193
Haplogroup F, 107, 110, 111 Judo, 128, 141, 142, 372, 375–376
Haplogroup G1, 110, 112 Jump, ix, 110, 176, 187–195, 268, 274, 301
Heat stroke, 315–317, 321
Heavy athletes, 139–147
Heritability, 107–108 K
High bar, 295–297, 302, 303 Kinematics, 250, 251, 255, 266, 268, 283, 285,
High fat diet, 151–155 287, 290, 291, 300, 324
Highly-coordinated, 50, 51 Kinesthetic sense, 220
High-speed video camera, 251, 283, 306, Knee extension, 108, 117, 158, 159, 162, 169,
325, 337 176, 181, 306–308, 310
Hitting, 335–351 Knee extensor strength, 108
H-reflex, 18, 54, 69, 84
Hypertrophy, 157–170, 179–182
Hypothermia, 315–317, 319, 321 L
Lactate threshold (LT), 318, 319
Lateralized readiness potential (LRP), 63, 64
I Leg extension, 180, 181, 290
Ia afferent, 80–81, 84 Leg extension power, 108
Image processing, 354–356 Life expectancy, v, 143
Immobilisation, 227, 241 Lift force, 324, 329
Impulses, 252–255, 257, 282, 284, Limb force generation, 281–291
287–290 Location precision, 362
Inappropriate contraction, 75 Location trajectories, 361, 362
Inappropriate muscle contraction, 67 Locomotion related neural mechanism,
Independence in the neural mechanisms, 55 55–56
Indirect calorimetry, 127, 133 Locomotor-like EMG activities, 52
Inhibitory control, 27, 92 Locomotory function, 53
Insertion/deletion (I/D), 116, 117, 119, 120 Locomotory modes, 55
Insulin resistance, 142, 144, 145, 147 Locomotory movements, 50, 52, 54
Integrated goal, 38 Longevity, 147
Intended impact location, 345 Longitudinal study, 8, 9, 143, 181,
Interlimb coordination, 38, 45 182, 228
Inter-muscle, 182 LORETA, 7, 28, 30, 238
International Space Station (ISS), 224, 225
Inter-person, 37–47
Intracortical inhibition, 15, 69–71 M
Intra-muscle, 182 Magnesium, 81
Intramuscular pH, 219 Magnetic resonance (MR) imaging, 131, 176,
Intra-person, 37–47 178, 179, 218, 219
Ion balance, 81 Marksman, 26, 27
Index 385
Mars500, 231–234 Muscle–tendon unit (MTU), 188, 191,

Maximal-intensity contraction, 201, 193, 195
208, 209 Musculoskeletal system, 224, 227
Maximal voluntary contraction, 86, 201 Music, 31, 67, 73
Maximising success, 299, 300 M-wave, 86–88
Meaningful behavior, 45 Myosin light chain kinase, 201
Medial gastrocnemius (MG), 50, 51, 53, 190, Myosin regulatory light chain phosphorylation
205–207, 217, 219 (MLCP), 201, 202, 208
Medical, v, 219, 232, 315, 316, 321
Menstrual cycles, 132–133
Mental health, 229, 230, 241 N
Metabolic energy, 132, 251, 259 Negative potentials (N1), 16, 29
Metabolic rates, 127, 130, 131 Nerve block, 83, 85
Metabolic syndrome, 140–142 Neural
Mitochondria, 106, 107, 110, 152 activity, 3, 4, 10, 14, 52, 83–85
Mitochondrial biogenesis, 153–155 efficiency, 27
Mitochondrial DNA, 106–107 mechanisms, 3, 49–57, 64, 70–73, 79–88
Momentum, 282, 289, 290, 295, 296, Neurocognitive amelioration, 239
298, 301 Neurofeedback training (NFT), 26, 31–33
Monoarticular, 176, 181 Neuromuscular activity, 192–194
Mood, 224, 229–231, 233, 234, 239–241 Neuromuscular compartments, 168
Mortality, 140, 142, 143 Neurophysiological, 3, 4, 8, 10, 238
Motion analysis, 281, 283, 338
Motor adaptation, 55
Motor control, 3, 6, 7, 10, 27, 30, 59–61, O
63, 72 Obesity, 94, 110, 140, 142, 143, 146, 155
Motor evoked potentials (MEPs), 14, 15, 17, Oddball task, 99
27, 70 Olympic, 110, 116, 141, 177, 180, 181, 249,
Motoric (anatomical) information, 38 314, 315, 317, 372
Motor neuron, 83, 84, 88, 189, 201 Olympic Games, 248, 302, 315
Motor Schema Theory, 59–64 Open-loop control system, 59, 60
Motor unit, 83, 85, 219 Open water swimming, 313–321
Motor unit potentials, 83, 85 Optimisation, 227, 231, 294–296, 298–300
Movement-related cortical potentials Optimum angle, 169
(MRCPs), 6, 69, 72 Organ tissue mass, 129, 131
20-m shuttle run test, 94, 95 Osteoarthritis, 162
Multi-limb skill, 73 Overfeeding, 139, 140, 143–147
Multiple images, 360–361 Oxidative phosphorylation system (OXPHOS),
Multistressor, 229 106, 110
Muscle Oxygen uptake capacity, 249
activation levels, 180, 209
contractile properties, 219
cramp, 79–88 P
fatigue, 82, 199–200, 214, 218–220 Parallel bars, 294
fibers, 82, 86, 112, 116, 168, 188–194, Paretic hand, 70
199–209, 214, 219, 306 Parietal cortex, 7, 14, 16–17, 29, 30, 45
fiber types, 168 Parkinson’s disease, 68–69
motor point, 80, 85 Perceived physical state, 231, 233, 234, 237
spindles, 73, 82, 189 Performance enhancement, 32, 187–195, 314
thickness, 157, 158, 169, 170, 176, 178, Peripheral muscle fatigue, 214, 218
179, 182 Peripheral origin theory, 82, 85–88
volumes, 169, 177, 178, 180, 181 Personalised EEG profile, 32
Muscle–tendon interaction, 188–191, 207 Perturbations, 29, 30, 298–300
386 Index
Phase-dependent manner, 53 Relaxation, 67–75, 168, 238

Photo-sensing diode, 340 Release windows, 295, 296, 298
Physical active lifestyle, 241 Remote effect, 73
Physical activity, 50, 54, 91, 92, 95, 98, 99, Repetition maximum, 162, 165
125, 135, 239, 241 Resistance training, 139, 143, 157–170, 182,
Physiological, x, 4, 30, 75, 105, 131, 220, 224, 193, 239
227, 229, 230, 234, 239, 240, 321 Resistive exercise training, 239
Pianists, 74 Respiration coupling, 250–252, 254,
Pitching grip, 331–332 255, 259
Plantar flexion, 190, 203, 204, 213–220 Resting energy expenditure (REE), 125–135
Plastic reorganization, 54 Rifle shooters, 28, 32
Player locations, 353–365 Risk, 25, 80, 140–147, 215, 315–319, 321
Plyometric exercise, 193–195 Roller skiing, 251, 269, 270, 274–277
Pole reaction forces, 262, 270, 271, 275–277 Rugby, 140, 141, 371
Positive feedback, 84 Running, ix, 49, 51, 55, 56, 109, 128, 176, 182,
Postactivation potentiation (PAP), 199–209 187, 188, 214, 218, 240, 248, 249,
Power, 4–6, 26–28, 30–33, 106, 108, 110–112, 256–258, 270, 278, 290, 353
115–117, 119, 132, 139, 181, 182, R577X, 112, 116–119
190, 194, 200, 219, 255, 263, 269,
285, 319
Power-velocity relationships, 252, 254 S
Practitioner-oriented model, 378 Schema, 59–64, 71, 371, 375
Precision of ball-bat contact, 347–349 School kids, 235
Precue paradigm, 64 Self-determined intensity, 240
Precuneus, 235, 238 Sensorimotor rhythm, 32
Preferred impact location, 343 Shooters, 26–28, 32, 33, 337
Prefrontal cortex, 14, 28, 30, 72, 92 Short arm human centrifuge (SAHC), 239, 240
Prefrontal cortex activity, 240 Shortening of the muscle, 85, 200, 202
Premotor cortex (PM), 14, 16 Short interval intracortical inhibition (SICI), 71
Pre-supplementary motor area (pre-SMA), 72 Simulation, 80, 191, 294, 295, 298–303,
Primary motor cortex (M1), 14, 15, 17, 30, 68 331, 371
Profile of mood state, 231 Simultaneous relaxation, 73, 75
Proprioceptive, 18, 53, 60, 73 Simultaneous relaxation and contraction,
Propulsion, 248, 250–252, 254, 257–260, 262, 73, 75
288, 290 Single poling technique, 248
Propulsive force, 50, 282, 285–289 1-Skate technique, 248–260
Psycho-physiological stressors, 229 2-Skate technique, 248–249
Skating technique, 248, 268–273
Ski reaction forces, 268–272, 275–277
Q Slider, 323, 325, 327, 328, 330–332
Quadriceps femoris, 158–159, 175–183 Sociological dimensions of health, 242
Quantitative, ix, 175–183, 336, 368, 369 Sodium, 81, 317
Quiet eye (QE), 6, 9 Soleus, 84, 86, 87, 168, 214, 216,
217, 219
Spatial information, 26, 38, 43, 44
R Spatial perception, 342–346, 350, 351
Readiness potentials (RP), 63, 72 Specialized, 50, 52, 56, 367
Rectus femoris (RF), 51, 162, 176–181, Specificity, 54–57, 63, 113, 114, 167
306–308 Specificity of functional alterations, 55
Reference frame, 45 Spinal cord, 18, 50, 52, 53, 69, 82–86
Rehabilitation, 19, 30, 227, 228, 241, 306 Spinal reflex, 18, 83, 84, 86
Relative direction, 38, 44, 46 Spin axis, 324–332
Relative phase, 38, 41–47 Spin parameter, 324
Index 387
Spin rate, 324–332, 337 Throwing motion, 331, 332

Sport psychological research on coaching, Tibial nerve, 80, 86, 204
368–370 Timing accuracy, 336, 339, 341, 343, 345, 347,
Sports 348, 350, 351
nutrition, 127 Tonic vibration reflex, 84
performance, ix, x, 3–10, 26–29, 33, 135, Top-down sustained attention, 28
182, 183, 294, 374, 377 Top spin, 327
Sport-specific, 176, 182, 368 Track and field, 116, 118, 119, 142
Sprint, 109–112, 115–117, 119, 170, 179, 182, Tracking score, 361–365
248, 263, 269, 274, 278, 281–283, 287, Tracking size, 361
289, 291, 320 Trail Making Test (TMT), 94–97
Sprinter, 54, 109, 116, 118, 169, 176–180, Training procedures, 54
192, 258 Transcranial direct current stimulation
Stepping-like movements, 51, 53 (tDCS), 17
Stereotypical movement patterns, 50, 56 Transcranial magnetic stimulation (TMS),
Stimulation frequency, 201, 208 3, 14–15, 17, 27, 53, 68–71
Stimulus – response compatibility, 93 Transfered, 46, 56, 106, 259, 299, 360
Strain gage, 263, 264, 269 Transverse relaxation time, 168
Strength, 13, 106, 108, 132, 139, 143, 147, 157, Triiodothyronine, 131
194, 195, 227, 228, 262, 278, 282, 287, Triple piked somersault, 296, 298
294–296, 298, 303 T2-weighted magnetic resonance imaging,
Stressors, 229, 230, 233 218, 219
Stretching, 79, 188, 189, 191 Twist, 294, 296, 301, 302
Stretching to the muscle, 85 Twitch contractions, 200, 202–209
Stretch-shortening cycle (SSC), 188, 189, 192, Two-seam, 325, 327–330, 332
193, 195, 273 Type of sport, 128, 132
Stroke, 5, 68, 70, 305, 315–317, 321
Strongly-bounding state, 201
Stroop task, 93 U
Structure of expert coaches, 375 Ultrasonography, 176, 178, 179, 188, 192,
Subhaplogroups, 109, 112–114 193, 205
Substantia nigra, 68 Ultrasound, 146, 205
Sumo, 131, 141, 143, 147, 176–178, 180 Undersomersault, 294
Supplementary motor area (SMA), 14, 16, Uniform color, 354, 357, 358
72, 230 Unstable postures, 29
Swim, 55, 314, 315, 317, 318 Upstart, 299, 300
Swimming, 52, 55, 119, 305, 306,
313–321, 376
Synergistic muscle, 180 V
Variability, 32, 38, 83, 117, 127, 293–303
Variant, 111–112, 119
T Vasti, 159, 162, 176, 181
Tacit knowledge of expert coaches, 368 Vastus intermedius (VI), 159, 176–180, 182
Tactile, 18 Vastus lateralis (VL), 117, 159, 168,
Task difficulty, 30, 33, 91–99 176–181, 192
Task-switching tasks, 98 Vastus medialis, 159, 176, 179, 180, 182,
Team sports, 4, 9, 242, 371, 373, 375 306–308
Tendon elasticity, 192, 193 Velocity, 28, 56, 117, 182, 193, 199–209, 218,
Tendons, 18, 80, 82, 84, 188–195, 203 219, 263–268, 271, 273–278, 282, 289,
Theoretical integration across, 377 294, 298, 301, 319, 324, 331, 332,
Theta power, 5, 6, 28, 30–32 337, 350
Third person perspective imagery, 13 Vertical force, 282, 286–289
Three-dimensional (3-D) direct linear Vertical impulses, 288, 289
transformation method, 264, 338 Video coordinates, 354, 355, 359,
Threshold frequency, 80–82, 84, 85 361, 365
388 Index
Video image, 264, 325, 327, 353–365 Voxel-based morphometry (VBM), 8

Viewing, 239, 251, 302, 303 V2 skating, 261–279
Virtual environment, 6, 302
Virtual realities, 241, 303
Visceral fat, 140, 143, 144, 146 W
Visual, 5, 16, 18, 19, 29, 30, 38, 46, Walking, 49–56, 161, 176, 187, 188, 227, 249,
339–341, 362 257, 269, 290, 294
Visual information, 7, 9, 18–19, 46, Weightlifting, 142, 177, 178, 180, 192
301, 339 World Health Organisation (WHO), 242
Visual occlusion, 339, 340
Voluntary contraction, 80, 83, 84,
86–88, 201 Z
VO2max, 108, 109, 269 Zero-gravity, 224, 227

2015 Book SportsPerformance

Uploaded by

Copyright:

Available Formats

2015 Book SportsPerformance

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2015 Book SportsPerformance

Uploaded by

Copyright:

Available Formats

Kazuyuki Kanosue Editor in Chief

Tomoyuki Nagami • Jun Tsuchiya

ISBN 978-4-431-55314-4 ISBN 978-4-431-55315-1 (eBook)

Library of Congress Control Number: 2015942559

Springer Tokyo Heidelberg New York Dordrecht London

Printed on acid-free paper

Springer Japan KK is part of Springer Science+Business Media (www.springer.com)

Program Leader Kazuyuki Kanosue

This book focuses on sports performance. According to the Longman Dictionary of

Saitama, Japan Tomoyuki Nagami

Part I The Sporting Brain

Part II The Physiological Basis of Sports Performance

Part III Performance and Coaching in Various Sports

22 Limb Force Generation as a Limiting Factor for Maximum-Effort

Keywords Athlete • Golf • Aiming • Alpha • Theta

© Springer Japan 2015 3

1.2 Brain Activity During Sports Performance: Golf

EEG techniques can be used to detect psychological and physiological responses

1.3 Golfers’ Brain

Several studies have focused on the characteristics of athletes’ brains, by comparing

Jäncke and colleagues (2009) investigated the relationship between a particular

1.4 Future Studies

In this section, I describe approaches to improve sports sciences by using neuro-

Keywords Motor imagery • Mental practice • Brain imaging • Functional

© Springer Japan 2015 13

2.2 Measurement of Brain Activity

2.2.1 fMRI and PET Studies

2.2.2 TMS Studies

TMS has proven to be a valuable technique. TMS is performed by passing a high-

2.2.3 MEG and EEG Studies

measured a negative potential recorded at FCz in response to a visual cue with

2.3 Role of Each Region

2.3.1 Supplementary Motor Area and Premotor Cortex

2.3.2 Parietal Cortex

accurately. According to Desmurget and colleagues (2009), stimulation of the

2.3.3 Primary Motor Cortex

While some researchers found M1 activation during motor imagery (Porro

2.3.4 Spinal Cord

2.4 Influence of Somatosensory and Visual Information

Visual information also influences corticospinal excitability during motor imag-

Fadiga L, Buccino G, Craighero L, Fogassi L, Gallese V, Pavesi G (1999) Corticospinal excit-

Andreas Mierau, Thorben Hülsdünker, and Heiko K. Strüder

Abstract Large populations of synchronously active cortical neurons produce

Keywords EEG • Cortex • Sport • Expert • Golf • Shooting

A. Mierau (*) • T. Hülsdünker • H.K. Strüder

© Springer Japan 2015 25

3.1 Cortical Oscillations and Sport Performance

sensorimotor cortex. Therefore, alpha synchronisation over the contralateral sen-

3.2 Brain Activity During Balance Control

3.3 EEG Neurofeedback Training to Enhance Athletic

The basic principle of cortical oscillations-based neurofeedback training (NFT) is

3.4 Concluding Remarks and Suggestions for Future

Tetsuro Muraoka, Yuki Watanabe, and Kazuyuki Kanosue

Abstract When performing intra- or inter-person coordination of cyclical move-

Keywords Coordination • Interlimb • Digits

© Springer Japan 2015 37

Rhythmic interlimb coordination is constrained by perceptual-cognitive, musculo-

study in which we investigate whether intra- and inter-person coordination between

In experiment 1, subjects performed cyclical extension-flexion movements of the

“opposite-direction” in tasks 2 and 3. Thus, tasks 1, 2, 3, and 4 were referred to as