Content uploaded by Benjamin Lee
Author content
All content in this area was uploaded by Benjamin Lee on Sep 21, 2021
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=rjsp20
Download by: [Mount Royal University] Date: 04 October 2016, At: 06:27
Journal of Sports Sciences
ISSN: 0264-0414 (Print) 1466-447X (Online) Journal homepage: http://www.tandfonline.com/loi/rjsp20
The effect of core training on distal limb
performance during ballistic strike manoeuvres
Benjamin Lee & Stuart McGill
To cite this article: Benjamin Lee & Stuart McGill (2016): The effect of core training on
distal limb performance during ballistic strike manoeuvres, Journal of Sports Sciences, DOI:
10.1080/02640414.2016.1236207
To link to this article: http://dx.doi.org/10.1080/02640414.2016.1236207
Published online: 03 Oct 2016.
Submit your article to this journal
View related articles
View Crossmark data
The effect of core training on distal limb performance during ballistic strike
manoeuvres
Benjamin Lee and Stuart McGill
Spine Biomechanics Laboratory, Department of Kinesiology, Faculty of Kinesiology, University of Waterloo, Waterloo, ON, Canada
ABSTRACT
Ballistic limb motion is enabled by proximal “core”stiffness. However, controversy exists regarding the
best method of training this characteristic. This study sought to determine the most effective core
training method to enhance distal limb athleticism. A total of 12 participants (24 ± 3 years, 1.8 ± 0.05 m,
76.8 ± 9.7 kg) consisting of Muay Thai athletes performed a core training protocol (Isometric vs.
Dynamic, with Control) for 6 weeks, using a repeated measures design to assess performance (peak
strike velocity, peak impact force, muscular activation) in various strikes. Isometric training increased
impact force in Jab (554.4 ± 70.1 N), Cross (1895.2 ± 203.1 N), Combo (616.8 ± 54.9 N), and Knee
(1240.0 ± 89.1 N) trials (P< 0.05). Dynamic training increased strike velocity in Jab (1.3 ± 0.2 m · s
−1
),
Cross (5.5 ± 0.9 m · s
−1
), Combo (0.7 ± 0.1, 2.8 ± 0.3 m · s
−1
), and Knee (3.2 ± 0.3 m · s
−1
) trials (P< 0.05).
Isometric training increased Combo impact force 935.1 ± 100.3 N greater than Dynamic and
931.6 ± 108.5 N more than Control (P< 0.05). Dynamic training increased Jab strike velocity
1.3 ± 0.1 m · s
−1
greater than Isometric and 0.8 ± 0.1 m · s
−1
more than Control (P< 0.05). It appears
that both static and dynamic approaches to core training are needed to enhance both velocity and
force in distal limbs.
ARTICLE HISTORY
Accepted 7 September 2016
KEYWORDS
Muay Thai; martial arts; per-
formance; core stability;
proximal stiffness; athleti-
cism; spine
Introduction
Striking martial arts such as Muay Thai involve ballistic limb
motion to deliver high impact force at maximal velocities to an
intended target. For centuries, these athletes have sought out
the most effective training methods to enhance their striking
speed, impact force, and other markers of performance. The
body is an articulated linkage where proximal stiffness and
stability is necessary to enable rapid movements of distal
segments (McGill, 2016). This mechanism requires the core
musculature (muscles proximal to the ball and socket joints
of the hips and shoulders) to prevent spine motion and but-
tress the torso, rather than generate movement akin to the
musculature of the distal limb segments. In this way, the core
muscles increased spinal stiffness and stability to prevent
unwanted torso motion when exerting against external loads
(Bergmark, 1987). Extending this concept to enhancement of
athletic performance, an athlete must generate sufficient stiff-
ness in the high mass torso for several reasons. First, proximal
stiffness in the torso reduces small eccentric movements,
representing a “giving way”or energy loss, to enhance distal
limb motion velocity (McGill, 2014). Second, the spine is a
flexible column that must be stiffened to enable it to bear
more load without buckling (Bergmark, 1987). There is also
evidence of rapid muscle activation/relaxation sequencing in
high-performance athletes who seek more effective strikes
with higher limb speed and higher contact force at the term-
inal end of the linkage (hands or feet) (McGill, 2014; McGill,
Chaimberg, Frost, & Fenwick, 2010). When muscles contract,
they create both –force and stiffness (Brown & McGill, 2009a).
Hence, the speed strength paradox is formed where muscle
force is needed in a pulse-like fashion to initiate limb speed
followed with relaxation to decrease stiffness to enhance clos-
ing velocity between the hand and the target. Thus, it appears
that an interplay between activation and relaxation, working
to create both an inertial high mass torso with rapidly moving
limb segments is a necessary condition of proficient striking
performance. There has been little formal investigation into
enhancement of these mechanisms. This study sought to
determine the most effective core training method (between
Isometric or Dynamic approaches) to enhance distal limb
athletic performance (impact force and strike velocity).
Numerous investigations into the relationship between
core stiffness/stability, training, and athletic performance
have yielded varied results. Nesser, Huxel, Tincher, and
Okada (2008) found weak correlations between Isometric
core stability and strength, speed, and power in football
players, and concluded that implementing core training pro-
tocols would not substantially enhance athletic performance.
However, this investigation only correlated performance and
endurance measures without examining if training core endur-
ance affected performance. Taanila et al. (2009) showed
improvements in military testing parameters, including sit-up
endurance tests, after training with Isometric core exercises.
Both Willardson (2007) and McGill (2014) have discussed the
CONTACT Stuart McGill mcgill@uwaterloo.ca Spine Biomechanics Laboratory, Department of Kinesiology, Faculty of Kinesiology, University of Waterloo,
200 University Ave. W., Waterloo, ON N2L 3G1, Canada
JOURNAL OF SPORTS SCIENCES, 2016
http://dx.doi.org/10.1080/02640414.2016.1236207
© 2016 Informa UK Limited, trading as Taylor & Francis Group
use of core stability exercise to enhance sport performance by
buttressing the torso to facilitate distal limb motion. McGill
et al. (2010) commented on the speed/strength and force/
stiffness paradox after observing activation patterns termed
as the “double peak”in elite mixed martial arts (MMA) athletes
when striking. Muscles when activated create both, force and
stiffness. But stiffness slows rapid motion so that pulse pat-
terns of activation are needed to create high limb velocity.
Specifically, a muscular pulse in the torso and limbs initiate
limb movement. This is followed by a relaxation phase as the
closing velocity of the hand/foot to the target increases. Elite
athletes then produce a second impulse at impact to create a
stiffened resulting in a larger “effective mass”and a higher
strike force. Coaches refer to this as “getting the body behind
the force”or “turning the body to stone”, so there is minimal
energy loss. McGill (2014) reported a similar pulsing phenom-
enon in sprinters and golfers at impact. The idea of enhancing
effective mass to heighten impact force was hypothesised by
Blum (1977), Neto, Magini, and Saba (2007), and Pain and
Challis (2002). Within the world of martial arts, research dating
back to 1985 addressed performance differences between
novice and elite boxers (Filimonov, Koptsev, Husyanov, &
Nazarov, 1985). We were generally interested in training tech-
niques to enhance elite strike performance. While there has
been some work examining the role of strength and condi-
tioning training to improve martial arts performance (Turner,
2009), we could not find data that assessed torso stiffness,
limb speed, or strike effectiveness. Anecdotally, methods of
training muscular relaxation have been used by martial artists
(Little, 2001; Tsatsouline, 2006). However, the lack of investiga-
tion into which training exercises enhance strategic muscle
pulses and core stiffness, and whether they actually enhance
performance motivated this research. We chose a population
of Muay Thai trained athletes given their athletic objectives of
limb speed and strike force enhancement.
Isometric core exercises have been demonstrated to suffi-
ciently create activation of the core musculature (Axler &
McGill, 1997; Callaghan Gunning, & McGill, 1998; Kavcic,
Grenier, & McGill, 2004; McGill & Karpowicz, 2009), while spar-
ing the spine from excessive loads and injurious movement
patterns. Dynamic “core exercises”have similarly been
assessed for muscle activation patterns and in some cases
joint loading (McGill, Cannon, & Anderson, 2014, McGill,
Karpowicz, & Fenwick, 2009; McGill, Karpowicz, Fenwick, &
Brown, 2009, McGill, Marshall, & Andersen, 2013; McGill,
McDermott, & Fenwick, 2009). However, their ability to influ-
ence speed/force “striking”performance has not been
assessed. Collectively, this large body of evidence suggests
that core exercises influence stiffness and corresponding
joint loading but the influence on speed/strength perfor-
mance remains unknown. This study tested an approach that
has been proven to enhance short-term stiffness in the torso
(Lee & McGill, 2015) which to the authors’knowledge, but the
effect on athleticism remains unknown. The specific question
was: What core training style is best suited to enhance strike
force and speed of Muay Thai striking –an Isometric or
Dynamic training approach? It was hypothesised that
Dynamic core training would better improve strike speed
than an Isometric approach or no (Control) training, while
Isometric core training would enhance strike impact force
better than a Dynamic approach or no (Control) training
given existing evidence for enhancing “effective mass”.
Methods
Experimental approach to the problem
A repeated measures test/retest protocol was used to examine
changes in biomechanical performance measures (impact force,
strike velocity, and core and hip electromyography signals) after
a 6-week core training protocol consisting of Isometric bracing or
Dynamic movement exercises in 12 male Muay Thai athletes. All
participants were recruited and trained between March 2013 and
June 2013 during daytime hours. As physiological markers of
health and performance were not within the scope of the study
controls for nutrition and hydration were not used. Participants’
Muay Thai strike performance was measured before and after a
6-week training period (or waiting period for the Control group).
After the initial data collection, participants were divided into an
Isometric training group, Dynamic training group, or Control
group. Isometric and Dynamic training groups performed a train-
ing programme progressing in intensity-based on static-bracing
exercises and movement/speed-based exercises, respectively.
Participants
A total of 12 young healthy (24.2 ± 2.9 years, 1.8 ± 0.05 m,
76.8 ± 9.7 kg) were selected from a population of club Muay
Thai fighters; a martial art native to Thailand involving stand-
ing striking with the fists, elbows, knees, and shins. Exclusion
criteria consisted of any individuals who have experienced low
back pain or injury currently or within the past year.
Participants were trained in Muay Thai boxing for at least
1 year (range: 1.5–6 years of consistent training) with the
majority (10) having competitive amateur records and 2 parti-
cipants being provincial and international amateur champions
in their respective weight classes.
All participant recruitment and data collection procedures
were performed in accordance with University Office of
Research Ethics guidelines. The participants were informed of
the purpose and method of the study to ensure that they
understood completely, and each provided written informed
consent to participate. Participants were also informed that at
any time during the data collection or training protocol they
were free to withdraw from the study. Written informed con-
sent was gained in agreement with University guidelines.
Procedures
Muay Thai strike performance was assessed using strike
impact force strike velocity and electromyography signals. All
3 were measured during 4 trials of Muay Thai strikes,
described below, before, and after a 6-week intervention of
core training or rest. The training intervention consisted of 3
groups, each with 4 participants; 1 group performed Isometric
core exercises, 1 group performed Dynamic core exercises and
the Control group performed no special exercises during this
period.
2B. LEE AND S. MCGILL
Strike trials
The martial art of Muay Thai involves 2 competitors striking at
each other while standing using points of their fists, elbows,
knees, and shins. To gather information regarding the efficacy
of these strikes, 4 trials were selected typical to strikes used in
Muay Thai training and competition. All strikes were per-
formed from the participant’s dominant stance in which a
right-handed participant would stand with their left leg and
arm in front and right leg behind them, in a shoulder width
stance; and vice versa for a left-handed participant. The 4 trials
used were a Jab (lead hand strike with the fist), Cross (rear
hand strike with the fist), Knee (rear leg strike impacting with
the tip of the knee), and Jab–Cross combination (aka “Combo”,
a Jab followed in succession by a Cross), as illustrated in
Figure 1. Three repetitions of each trial were performed in a
randomised order as to eliminate any bias or learning effect,
with 1–2 min of rest between each trial to reduce peripheral
fatigue between trials. Researchers also confirmed before each
trial that the participant was adequately rested and ready to
perform.
Instrumentation
Electromyography (EMG) signals, whole body kinematics, and
strike force were collected during strike trials. Data sources
were connected to Vicon MX Ultranet hardware (Vicon MX,
Vicon Motion Systems, Oxford, UK) and synchronised during
data collection using Vicon Nexus 1.8 software (Vicon MX,
Vicon Motion Systems, Oxford, UK). All signals were collected
at various frequencies and down-sampled to 60 Hz during
post-processing.
Figure 1. Examples of each of the strikes tested pre/post-core training. (a) Jab, (b) Cross, (c) Jab–Cross Combination (“Combo”), (d) Knee.
JOURNAL OF SPORTS SCIENCES 3
Electromyography
EMG signals were collected on bilateral core and hip muscu-
lature using pre-gelled, disposable, monopolar Ag–Cl disc-
shaped surface electrodes (30 mm diameter, Medi-traceTM
100 Series Foam Electrodes, Covidien, MA, USA) placed on
the skin over each muscle of interest (rectus abdominis [RA],
external oblique [EO], internal oblique [IO], latissimus dorsi
[LD], upper erector spinae [UES], lower erector spinae [LES],
gluteus maximum [GMax], gluteus medius [GMed]). Briefly,
normalised signals were obtained as follows. Signals were
amplified (±2.5 V; AMT-8, Bortec, Calgary, Canada; bandwidth
10–1000 Hz, common mode rejection ratio (CMRR) = 115 db at
60 Hz, input impedance = 10 GX) and sampled at 2048 Hz,
low-pass filtered with a 500 Hz, rectified and low-pass filtered
at 2.5 Hz (single pass second order) to mimic the frequency
response of torso muscle after Brereton and McGill (1998) and
normalised to the maximum voltage produced during max-
imum voluntary contraction (MVC) trials to produce a linear
envelope mimicking the muscle force output; a technique
used many times before [5]. Maximum effort to elicit max-
imum neural drive was the goal of the MVC trials. MVC’s were
obtained using 4 postures: (1) a modified sit-up position in
which participants isometrically attempted to produce trunk
flexion, side bend, and twist motions against resistance, (2)
Isometric trunk extension while cantilevered in a prone posi-
tion over the edge of a table (Biering-Sorensen position)
against external resistance, (3) Isometric wide grip pull-up
posture in which the participant attempted to isometrically
pull against a horizontal bar while being resisted with instruc-
tions of maintaining a maximally tight grip and attempting to
“bend the bar”while pulling vertically, and (4) side lying
abduction position where participants lay on 1 side and
attempted to produce hip abduction against resistance. All
EMG signals were normalised to a per cent MVC which
allowed day-to-day comparisons of muscle activation ampli-
tude with maximal values of the enveloped signal reported.
Limb kinematics
Three-dimensional (3D) whole body kinematics were recorded
using an infrared motion analysis system (Vicon MX, Vicon
Motion Systems, Oxford, UK) using an 8 camera set-up
sampled at 60 Hz. Rigid clusters were attached on the thorax,
pelvis, upper arms, forearms, hands, thighs, shanks, and feet;
each cluster with 4 reflective markers except the thighs, con-
taining 5 markers. Jab, Cross, and Jab–Cross limb velocities
were measured from the displacement of the striking hand’s
cluster. The location of the hand cluster was calculated by
taking the average of the 3D location coordinates 4 markers.
Velocity was then calculated using a numerical differentiation
of each coordinate, as per the following equations:
Vx¼X2X1
t2t1
(1)
Vy¼Y2Y1
t2t1
(2)
Vz¼Z2Z1
t2t1
:(3)
Velocities about each axis were then normalised to a scalar
magnitude using the equation:
Vlimb ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
V2
xþV2
yþV2
z
q:(4)
Strike force
A portable “pancake”force transducer designed for impact use
(AMTI, Massachusetts, USA) was mounted to a fixed surface to
record strike force. A steel cylinder was mounted to the trans-
ducer with a padded body protector, typically worn to protect
boxing trainers from strikes to the body, wrapped around the
circumference to act as a target. As all strikes across all parti-
cipants were measured in this manner, the damping effect of
the body protector was not accounted for in impact force
measurements. Impact data were sampled at 2160 Hz and
filtered with a second-order dual pass Butterworth filter with
cut-off frequency of 100 Hz. Cut-off frequency was determined
using a fast Fourier transformation; filtering, processing, and
analysis were performed via MATLAB software (Version r2012a;
The MathWorks Inc., Natick, Massachusetts, USA).
Core training protocols
Participants trained for 6 weeks using either Isometric or
Dynamic core exercises (the Control group did not train). All
participants were asked to refrain from performing any core
exercises outside of those assigned by researchers during the
study. The Isometric training group performed static exercises
designed to challenge the core musculature via bracing cues.
The Dynamic training group performed exercises based on
torso movement. Both training programmes were matched
for volume and intensity and periodised to increase challenge
every 2 weeks, dividing each programme into 3 phases (Tables
1and 2for a description of the progressive programmes).
Statistical analyses
Statistical tests were performed using IBM® SPSS® Statistical
software (Version 19, IBM Corporation, Somers, New York,
USA). 3 × 2 repeated measures analysis of variance (ANOVA)
(3 groups, before and after training) was conducted for com-
paring peak impact force peak EMG amplitudes, and peak
strike velocity; within and between training groups. Where
applicable, post hoc analyses were performed using the
Tukey honest significant difference (HSD) test when a signifi-
cant effect was detected with statistical significance set at
P≤0.05.
Results
Impact force, strike velocity, and torso and gluteal EMG were
measured before and after a 6-week bout in the 3 groups:
Isometric, Dynamic, or no core training (Control group).
4B. LEE AND S. MCGILL
Table 1. Isometric training protocol.
Exercise Sets ×Reps Freq Sets ×Reps Freq Comments
Week 1 Week 2
Plank 5,4,3,2,1 × 10 s 4× per week 5,4,3,2,1 × 10 s 7× per week Focus on quality of core contraction and
postural cues. Descending pyramid
sets (Start at 5 reps at 10 s each, next
set decrease 1 rep, continue to
decease 1 rep per set)
Bird dog 5,4,3,2,110 s 4× per week 5,4,3,2,1 × 10 s per side 7× per week
Side bridge 5,4,3,2,1 × 10 s 4× per week 5,4,3,2,1 × 10 s per side 7× per week
Torsional buttress 5,4,3,2,1 × 10 s per side 7× per week Focus on quality of core contraction and
postural cues. Use a hold time before
shaking begins, maximum 10 s
Week 3 Week 4
Anterior pallof press 5,4,3,2,1 × 10 s 4× per week Same volume, increase load 4× per week Focus on quality of core contraction and
postural cues
Posterior pallof press 5,4,3,2,1 × 10 s 4× per week Same volume, increase load 4× per week
Suitcase hold 5,4,3,2,1 × 10 s per side 4× per week Same volume, increase load 4× per week
Anti-rotation pallof press 5,4,3,2,1 × 10 s per side 4× per week Same volume, increase load 4× per week
Week 5 Week 6
Stir the pot 5×10 s per direction 4× per week 5×10 s per direction 4× per week Begin on knees, progress to toes. If 10 s
are not feasible, train below and
progress through the phase
Inverted row Up to 5×10 4× per week 5×10 4× per week If 10 reps are not feasible, perform as
many reps as possible and maintain
static posture. Focus on keeping torso
straight (avoid hip hiking/sagging)
Kettlebell unilateral rack walk 3×30 m walk per side 4× per week Same volume, increase load 4× per week Focus on core contraction and upright
posture (avoid lateral lean)
Half kneeling woodchop Up to 5×10 per side 4× per week 5×10 per side 4× per week If 10 reps are not possible, perform as
many reps as possible and progress
through the phase.
JOURNAL OF SPORTS SCIENCES 5
Table 2. Dynamic training protocol.
Exercise Sets ×Reps Freq Sets ×Reps Freq Comments
Week 1 Week 2
Curl up Up to 5×10 4× per week Up to 5×10 7× per week Focus on quality of muscular contraction;
visualise muscular activation throughout
motion. A total of 10 repetitions per set
with sets performed until marked
muscular fatigue sets in (up to 5 sets)
Superman Up to 5×10 4× per week Up to 5×10 7× per week
Side curl up Up to 5×10 per side 4× per week Up to 5×10 per side 7× per week
Twisting curl up Up to 5×10 per side 4× per week Up to 5×10 per side 7× per week
Week 3 Week 4
Advanced curl up (limbs extended) Up to 5×5–10 4× per week Up to 5×5–10 4× per week Begin with 5 × 5 and progress repetitions to
10. If 10 reps per side are too easy add/
increase weight
Back extension Up to 5×5–10 4× per week Up to 5×5–10 4× per week
Russian barbell twist Up to 5×5–10 per side 4× per week Up to 5×5–10 per side 4× per week
Week 5 Week 6
Curl up twitch Up to 5×5–10 4× per week Up to 5×5-10 4× per week Begin unweighted and focus on twitch
speed and rate of activation/relaxation.
Begin with 5×5 and progress repetitions
to 10. If 10 reps per side are too easy,
add/increase weight
Superman twitch Up to 5×5–10 4× per week Up to 5×5–10 4× per week
Lateral medball throw Up to 5×5-10 per side 4× per week Up to 5×5-10 per side 4× per week Ball velocity comes from torso movement,
not arms. Begin with 5×5 and progress
repetitions to 10. If 10 reps per side are
too easy, add/increase weight
Rotational medball throw Up to 5×5-10 per side 4× per week Up to 5×5-10 per side 4× per week Ball velocity comes from torso movement,
not arms. Begin with 5×5 and progress
repetitions to 10. If 10 reps per side are
too easy, add/increase weight
6B. LEE AND S. MCGILL
Impact force
Isometric and Dynamic trainings increased impact force in
almost all strike trials. Jab impact force increased by 17.9%
after Isometric training (2539.3 ± 89.1–3093.7 ± 69.4 N)
(F(1,3) = 70.8, P<0.001,β= 1.0), and by 18.3% after Dynamic
training (2614.7 ± 493.1–3199.6 ± 437.9 N) (F(1,3) = 34.1,
P=0.01,β= 0.9). Knee impact force increased by 13.1% after
Isometric training (8242 ± 132.3–9482 ± 152.8 N) (F(1,3) = 16.2,
P=0.03,β= 0.8), compared with a 7.1% increase after Dynamic
training (F(1,3) = 10.3, P=0.05,β= 0.7). Cross impact force
increased by 27% after Isometric training (5008.6 ± 76.3–
5008.6 ± 76.3 N) (F(1,3) = 73.4, P<0.001,β=1.0)but
Dynamic training did not elicit any significant increases or
decreases. The Control group did not experience any significant
changes in impact force.
Comparing between training groups, Isometric training
increased impact force more than Dynamic training for Cross,
Combo, and Knee trials, and was superior to the Control group
for all strike trials. Dynamic training yielded greater effects in
increasing impact force than the Control group for Jab and
Knee trials, but was not more effective than the Isometric
group for any trials. The Control group did not show any effects
in increasing impact force more than Isometric or Dynamic
training. Isometric training increased Cross impact force by
1895.2 ± 155.2 N, compared with a 232.2 ± 30.9 N decrease
after Dynamic training (F(1,6) = 36.5, P<0.001,β=0.8)and
229.5 ± 40.1 N decrease after Control (F(1,6) = 40.1, P<0.001,
β= 0.8). Similarly, Isometric training increased Knee impact
force by 1240.0 ± 143.2 N, while after Dynamic training Knee
impact force decreased by 660.3 ± 101.8 N (F(1,6) = 6.9,
P=0.04,β= 0.7) and 299.7 ± 41.5 N decrease after Control (F
(1,6) = 44.5, P<0.001,β=1.0).
These results are summarised in Tables 3 and 4, and
Figure 2.
Strike velocity
Isometric and Dynamic trainings both increased peak limb
velocity for almost all strike trials. Peak strike velocity
increased within the Isometric training group during Cross
trials by 23.9% (6.7 ± 0.7–8.8 ± 0.9 m · s
−1
) (F(1,3) = 34.3,
P= 0.01, β= 0.8) and the second strike of Combo trials by
5.2% (7.1 ± 0.5–9.0 ± 1.0 m · s
−1
) (F(1,3) = 16.5, P= 0.03,
β= 0.7). Dynamic training increased peak strike velocity for
Jab trials by 22.4% (4.5 ± .4–5.8 ± .6 m · s
−1
) (F(1,3) = 103.2,
P= 0.006, β= 0.9), Cross trials by 45.5% (6.6 ± 0.9–
12.1 ± 1.3 m · s
−1
) (F(1,3) = 140.9, P< 0.001, β= 1.0), Combo
trials by 13.2% (7.6 ± 0.8–10.4 ± 1.0 m · s
−1
) (F(1,3) = 34.7,
P= 0.01, β= 0.8), and Knee trials by 29.1% (7.8 ± 0.6–
11.0 ± 1.0 m · s
−1
) (F(1,3) = 12.6, P= 0.04, β= 0.7). No changes
were measured within the Control group.
Comparing between training groups, Isometric training had
a greater effect on increasing strike velocity than Control for
Cross, Combo, and Knee trials (P< 0.05 for al trials), but did
not increase strike velocity more than Dynamic training during
any trials. Isometric training increased strike velocity during
Cross trials by 2.1 ± 0.3 m · s
−1
compared with a
0.3 ± 0.1 m · s
−1
increase after Control (F(1,6) = 13.9,
P= 0.01, β= 0.8), and increased the first and second strikes
of the Combo trials by 0.3 ± 0.04 and 1.9 ± 0.1 m · s
−1
increase,
respectively, compared with a 0.6 ± 0.1 and 0.5 ± 0.05 m · s
−1
decrease after Control (F(1,6) = 9.0, P= 0.03, β= 0.8; F
(1,6) = 8.5, P= 0.03, β= 0.7). Isometric training also increased
Knee strike velocity by 1.2 ± 0.1 m · s
−1
compared with a
0.4 ± 0.05 m · s
−1
increase after Control (F(1,6) = 9.8,
P= 0.02, β= 0.8). Dynamic training had a greater effect on
increasing strike velocity compared with Isometric training
and Control for all trials. Jab velocity increased by
1.3 ± 0.1 m · s
−1
after Dynamic training, compared with a
0.0 ± 0.04 m · s
−1
change after Isometric training
Table 3. Detailed statistics for impact force within each training group.
Trial
Stats (F-value, P-value, power)
Isometric Dynamic Control
Jab Pre F(1,3) = 70.8, P< 0.001,
power = 1.0
F(1,3) = 34.1, P= 0.01,
power = 0.9
F(1,3) = 5.1, P> 0.05,
power = 0.2Post
Cross Pre F(1,3) = 73.4, P< 0.001,
power = 1.0
F(1,3) = 2.6 , P> 0.05,
power = 0.1
F(1,3) = 4.9, P> 0.05,
power = 0.2Post
Combo Pre F(1,3) = 8.04,
P> 0.05,
power = 0.4
F(1,3) = 10.2,
P= 0.05,
power = 0.7
F(1,3) = 2.0,
P> 0.05,
power = 0.1
F(1,3) = 2.3,
P> 0.05,
power = 0.1
F(1,3) = 5.2,
P> 0.05,
power = 0.2
F(1,3) = 2.5,
P> 0.05,
power = 0.1
Post
Knee Pre F(1,3) = 16.2, P= 0.03,
power = 0.8
F(1,3) = 10.3, P= 0.05,
power = 0.7
F(1,3) = 2.4, P> 0.05,
power = 0.1Post
Table 4. Detailed statistics for impact force between each training group.
Trial
Stats (F-value, P-value, power)
Iso/Dyn Dyn/Con Iso/Con
Jab Pre F(1,6) = 8.2, P= 0.03,
power = 0.7
F(1,6) = 6.1, P= 0.05,
power = 0.7
F(1,6) = 0.3, P> 0.05,
power = 0.07Post
Cross Pre F(1,6) = 36.5, P< 0.001,
power = 0.8
F(1,6) = 0.3, P> 0.05,
power = 0.07
F(1,6) = 40.1, P< 0.001,
power = 0.8Post
Combo Pre F(1,6) = 3.8,
P> 0.05,
power = 0.2
F(1,6) = 44.4,
P< 0.001,
power = 0.8
F(1,6) = 2.4,
P> 0.05,
power = 0.3
F(1,6) = 0.4,
P> 0.05,
power = 0.1
F(1,6) = 1.9,
P< 0.05
power = 0.2
F(1,6) = 41.2,
P< 0.001,
power = 0.9
Post
Knee Pre F(1,6) = 6.9, P= 0.04,
power = 0.7
F(1,6) = 6.0, P= 0.05,
power = 0.7
F(1,6) = 44.5, P< 0.001,
power = 1.0Post
JOURNAL OF SPORTS SCIENCES 7
(F(1,6) = 34.2, P< 0.001, β= 1.0) and 0.5 ± 0.05 m · s
−1
increase
after Control (F(1,6) = 8.7, P= 0.03, β= 0.7). During Cross trials,
Dynamic training increased strike velocity more than Isometric
training (5.5 ± 0.6 m · s
−1
increase compared with a
2.1 ± 0.3 m · s
−1
increase, respectively) (F(1,6) = 6.0, P= 0.05,
β= 0.7), and Control (0.3 ± 0.05 m · s
−1
increase after Control)
(F(1,6) = 18.3, P= 0.008, β= 0.9). Dynamic training increased
Combo velocity by 0.7 ± 0.08 and 2.8 ± 0.2 m · s
−1
for the first
and second strikes, respectively, compared with Isometric
training which increased velocity by 0.3 ± 0.05 and
1.9 ± 0.2 m · s
−1
, respectively (F(1,6) = 6.1, P= 0.05, β= 0.7;
F(1,6) = 6.3, P= 0.05, β= 0.7), and Control where strike velocity
decreased by 0.6 ± 0.1 and 0.5 ± 0.05 m · s
−1
(F(1,6) = 14.3,
P< 0.01, β= 0.8; F(1,6) = 69.6, P< 0.001, β= 1.0). Knee velocity
increased by 3.2 ± 0.3 m · s
−1
after Dynamic training compared
with a 1.2 ± 0.2 m · s
−1
increase after Isometric training, and a
0.4 ± 0.1 m · s
−1
increase after Control (F(1,6) = 13.9, P= 0.009,
β= 0.9). The Control group did not significantly increase in
strike velocity more than Isometric or Dynamic training.
These results are summarised in Tables 5 and 6,and
Figure 3.
Electromyography
Significant increases in peak EMG amplitudes were measured
within both Isometric and Dynamic training groups for all
trials. During Jab trials, Isometric training elicited a 37%
increase in EMG amplitudes across all musculature on average,
Figure 2. Peak impact force during (a) Jab, (b) Cross, (c) Jab–Cross Combination (“Combo”), (d) Knee strikes.
*–statistically significant (P< 0.05) difference within training group. ** –statistically significant (P< 0.05) difference between Isometric and Dynamic training
groups. *** –statistically significant (P< 0.05) difference between Isometric and Control training groups. **** –statistically significant (P< 0.05) difference between
Dynamic and Control training groups.
Table 5. Detailed statistics for strike velocity within each training group.
Trial
Stats (P-value, F-value, power)
Isometric Dynamic Control
Jab Pre F(1,3) = 0, P> 0.05,
power = 0
F(1,3) = 103.2, P= 0.006,
power = 0.9
F(1,3) = 0.4, P> 0.05,
power = 0.07Post
Cross Pre F(1,3) = 34.3, P= 0.01,
power = 0.8
F(1,3) = 140.9, P< 0.001,
power = 1.0
F(1,3) = 0.4, P> 0.05,
power = 0.07Post
Combo Pre F(1,3) = 4.4,
P> 0.05,
power = 0.4
F(1,3) = 16.5,
P= 0.03,
power = 0.7
F(1,3) = 10.6,
P= 0.05,
power = 0.7
F(1,3) = 34.7,
P= 0.01,
power = 0.8
F(1,3) = 0.3,
P> 0.05,
power = 0.06
F(1,3) = 1.0,
P> 0.05,
power = 0.09
Post
Knee Pre F(1,3) = 15.8, P= 0.03,
power = 0.7
F(1,3) = 12.6, P= 0.04 F(1,3) = 1.2, P> 0.05,
power = 0.1Post
8B. LEE AND S. MCGILL
with the largest increases (range: 50–72%) in the left back
(LLAT, LUES, LLES) and left abdominal (LRA, LIO) musculature
(F(1,3) = 10.2–120.5, P= 0.001–0.5, β= 0.7–1.0). In comparison,
Dynamic training elicited a 35% average increase in overall
EMG signal during Jab trials, with the largest increases in the
left and right abdominal musculature (LRA, LEO, LIO, RRA, REO,
RIO; range: 30–69%) (F(1,3) = 11.5–140.1, P= 0.001–0.04,
β= 0.7–1.0). Isometric training increased EMG amplitudes by
34% during Cross trials, with the largest increases (range:
30–60%) occurring in the right back (RLAT, RUES, RLES) and
abdominal (REO, RIO) musculature (F(1,3) = 10.9–64.9,
P= 0.001–0.05, β= 0.7–1.0). Dynamic training also increased
EMG amplitudes on average by 35%, with the greatest
increases (range: 36–58%) in the right back (RLAT, RUES,
RLES) and gluteal (RGMax, RGMed) musculature (F
(1,3) = 15.8–34.0, P= 0.01–0.03, β= 0.7–0.9). Combo trials
experienced a 17% average increase in EMG amplitudes after
Isometric training, and 23% average increase after Dynamic
training. Isometric training yielded the greatest increases
(range: 18–39%) in the right abdominal (RRA, REO, RIO) and
gluteal (RGMed) musculature (F(1,3) = 10.4–34.1, P= 0.01–0.05,
β= 0.7–0.9). Dynamic training had the greatest effect (range:
26–40%) on right abdominal (RRA, REO, RIO) and gluteal
(RGMax, RGMed) musculature (F(1,3) = 16.4–103.9, P= 0.005–
0.03, β= 0.7–0.9). Knee trials experienced a 25% average
increase in EMG amplitudes after Isometric training, compared
with a 20% increase after Dynamic training. Isometric training
had the greatest effect (range: 24–36%) on the left and right
abdominal (LEO, LIO, RRA, LEO, RIO) and right gluteal (RGMax,
RGMed) musculature (F(1,3) = 10.1–72.8, P= 0.001–0.05,
Table 6. Detailed statistics for strike velocity between each training group.
Trial
Stats (P-value, F-value, power)
Iso/Dyn Dyn/Con Iso/Con
Jab Pre F(1,6) = 34.2, P< 0.001,
power = 1.0
F(1,6) = 8.7, P= 0.03,
power = 0.7
F(1,6) = 3.5, P> 0.05,
power = 0.5Post
Cross Pre F(1,6) = 6.0, P= 0.05,
power = 0.7
F(1,6) = 18.3, P= 0.008,
power = 0.9
F(1,6) = 13.9, P= 0.01,
power = 0.8Post
Combo Pre F(1,6) = 6.1,
P= 0.05,
power = 0.7
F(1,6) = 6.3,
P= 0.05,
power = 0.7
F(1,6) = 14.3,
P< 0.01,
power = 0.8
F(1,6) = 69.6,
P< 0.001,
power = 1.0
F(1,6) = 9.0,
P= 0.03,
power = 0.8
F(1,6) = 8.5,
P= 0.03,
power = 0.7
Post
Knee Pre F(1,6) = 6.0, P= 0.05,
power = 0.7
F(1,6) = 13.9, P= 0.009,
power = 0.9
F(1,6) = 9.8, P= 0.02,
power = 0.8Post
Figure 3. Peak strike velocity during (a) Jab, (b) Cross, (c) Jab–Cross Combination (“Combo”), (d) Knee strikes.
*–statistically significant (P< 0.05) difference within training group. ** –statistically significant (P< 0.05) difference between Isometric and Dynamic training
groups. *** –statistically significant (P< 0.05) difference between Isometric and Control training groups. **** –statistically significant (P< 0.05) difference between
Dynamic and Control training groups.
JOURNAL OF SPORTS SCIENCES 9
β= 0.7–1.0), while Dynamic training had the greatest effect
(range: 10–42%) on right abdominal (RRA, REO, RIO) and glu-
teal (RGMax, RGMed) musculature (F(1,3) = 14.5–84.3,
P= 0.001–0.01 β= 0.8–1.0).
Comparing between training groups, Isometric and Dynamic
trainings increased peak EMG amplitudes more than Control for
almost all musculature in all trials but comparisons between
Isometric and Dynamic trainings revealed that some muscles
responded more to Isometric training, while other muscles
responded greater to Dynamic training. During Jab trials,
Isometric training increased overall peak EMG amplitudes simi-
larly to that of Dynamic training (37% vs. 35%) but left back
(LLAT, LLES), left abdominal (LRA, LIO), RLES, RIO, and RGMed
musculature experienced greater increases than with Dynamic
training (F(1,6) = 6.3–40.8, P=0.001–0.05, β=0.7–0.8). Dynamic
training had a greater effect than Isometric training for left
gluteal (LGMax, LGMed), RUES, and RRA musculature (F
(1,6) = 9.3–34.2, P=0.001–0.02, β=0.8–1.0). Isometric training
increased overall peak EMG amplitudes by 103% more than
Control on average, with significant increases in left back
(LLAT, LLES), left abdominal (LRA, LIO), LGMed, right back
(RUES, RLES), right abdominal (RRA, REO, RIO), and RGMed
musculature (F(1,6) = 6.0–13.8, P=0.01–0.05, β=0.7–0.8).
Dynamic training also increased overall peak EMG amplitudes
more than Control, on average 103% more, with left back (LLAT,
LLES), left abdominal (LRA, LIO), left gluteal (LGMax, LGMed),
and right back, abdominal, and gluteal (RUES, RRA, REO, RIO,
RGMed) musculature experiencing the greatest changes (F
(1,6) = 8.1–36.9, P=0.001–0.03, β=0.8–0.9).
During Cross trials, Isometric and Dynamic trainings yielded
similar responses in EMG amplitude (34% vs. 35% overall aver-
age increase). Isometric training increased peak EMG ampli-
tudes greater than Dynamic training for left back and gluteal
(LLES, LGMax), right abdominal (REO, RIO), and right gluteal
(RGMax, RGMed) musculature (F(1,6) = 6.6–49.1, P=0.001–
0.05, β=0.7–1.0); while Dynamic training had a greater effect
in left back (LLAT, LUES), left abdominal (LRA, LEO), and right
back and abdominal (RLAT, RUES, RLES, RRA) musculature (F
(1,6) = 9.0–50.1, P=0.001–0.03, β=0
.8–1.0). Isometric training,
on average, elicited an overall increase 76% greater than
Control for all musculature (F(1,6) = 6.2–35.5, P=0.001–0.05,
β=0.7–0.9), except for the RRA in which the Control group had
a greater response of increasing peak EMG amplitudes (F
(1,6) = 6.1, P=0.05,β= 0.7). Dynamic training elicited on
average a 77% greater increase in peak EMG amplitude than
Control for all musculature except for some gluteal (LGMax,
RGMax) and abdominal (REO) muscles (F(1,6) = 6.4–60.2,
P=0.001–0.05, β=0.7–1.0). The Control group had a greater
effect on increasing peak EMG amplitudes than Dynamic train-
ing for REO muscles (F(1,6) = 7.0, P=0.04,β=0.8).
Combo trial performance revealed Dynamic training had a
25% overall greater effect on increasing peak EMG amplitudes
compared with Isometric training. Dynamic training had a
greater effect on some left and right back, abdominal and
gluteal (LUES, LEO, LGMax, RUES, REO, RIO, RGMax) musculature
(F(1,6) = 8.8–40.4, P=0.03–0.001, β=0.8–1.0); while Isometric
training increased LIO, LGMed, and RLAT EMG amplitudes
greater than Dynamic training (F(1,6) = 6.0–9.0, P=0.03–0.05,
β=0.7–0.8). Isometric training increased overall peak EMG
amplitudes 84% more than Control, with significant increases
in left abdominal (LEO, LIO), left gluteal (LGMax, LGMed), right
back (RLAT, RLES), right abdominal (RRA, RIO), and RGMed
musculature (F(1,6) = 10.0–35.9, P=0.001–0.02, β=0.8–1.0);
while the Control group had a greater effect on than Isometric
training in the RUES musculature (F(1,6) = 6.3, P=0.05,β=0.7).
Dynamic training elicited a similar effect, yielding a 88% greater
increase in peak EMG amplitudes than Control, with significant
differences in response for some left back (LUES), left abdom-
inal (LRA, LEO), left gluteal (LGMax, LGMed), right back (RUES,
RLES), right abdominal (RRA, REO, RIO), and right gluteal
(RGMax, RGMed) musculature (F(1,6) = 6.0–39.1, P=0.001–
0.05, β=0.7–1.0); while the Control group did not show any
greater effects than Dynamic training.
During Knee trials, Isometric training increased overall
peak EMG amplitudes 17% greater than Dynamic training
with significant changes in some left back, abdominal, and
gluteal(LLES,LEO,LGMax),rightback(RLAT,RLES),and
right gluteal (RGMAX, RGMed) musculature (F(1,6) = 6.0–
13.8, P=0.01–0.05, β=0.7–0.9); while Dynamic training
had a greater effect on LGMed, RUES, and RIO musculature
(F(1,6) = 6.0–7.1, P=0.04–0.05, β=0.7–0.8). Isometric
training increased overall EMG amplitudes 99% more than
Control with significant differences in left back (LUES, LLES),
left gluteal (LGMax, LGMed), right back (RLAT, RUES, RLES),
right abdominal (RRA, REO, RIO), and right gluteal (RGMax,
RGMed) muscles (F(1,6) = 6.2–26.3, P=0.005–0.05, β=0.7–
1.0). Dynamic training also increased overall EMG ampli-
tudes 99% more than Control with significant changes in
theleftback(LUES,LLES),LIO,leftgluteal(LGMax,LGMed),
rightback(RUES,RLES),rightabdominal(RRA,REO,RIO),
and right gluteal (RGMax, RGMed) musculature (F(1,6) = 6.0–
39.9, P=0.001–0.05, β=0.7–1.0). The Control group did not
experience any changes in EMG amplitudes that were
greater than Isometric or Dynamic training.
Interestingly, an observed pattern of muscular activation
and relaxation was recorded in some muscles after Dynamic
training during all strike trials. Figure 5 exemplifies the left and
right back musculature before and after Dynamic training
during a Jab trial. The pretraining EMG pattern shows a typical
muscular profile during the strike; while after training, a dis-
tinct pattern of activation and relaxation (“double peak”)is
observed in the UES muscles.
Peak EMG amplitudes before and after Isometric and
Dynamic trainings and Control are summarised in Figure 4
with sample muscular activation patterns shown in Figure 5.
Discussion
The results suggest that the Isometric training protocol was
superior for enhancing impact force than the Dynamic proto-
col or Control, while the Dynamic training protocol was super-
ior in enhancing strike velocity compared with Isometric
training or Control groups. Interestingly, while Isometric and
Dynamic training groups both experienced increases in peak
EMG activity, the Dynamic training group was observed to
experience changes in the measured motor patterns involving
a shift from a typical single peak of activation to peaks of
muscular activation and relaxation.
10 B. LEE AND S. MCGILL
Figure 4. Peak EMG amplitudes during (a) Jab, (b) Cross, (c) Jab–Cross Combination (“Combo”), (d) Knee strikes.
*–statistically significant (P< 0.05) difference within training group. ** –statistically significant (P< 0.05) difference between Isometric and Dynamic training
groups. *** –statistically significant (P< 0.05) difference between Isometric and Control training groups. **** –statistically significant (P< 0.05) difference between
Dynamic and Control training groups.
Figure 5. (a) Back (a, b), Abdominal (c, d), and Gluteal (e, f) EMG signals for the Jab strike before and after Dynamic core training. Note the difference in patternin
the UES (LUES, RUES) highlighted in Figure 5(a,b).
JOURNAL OF SPORTS SCIENCES 11
The use of EMG to investigate motor control strategies in
martial arts striking has grown in recent years (Machado, Osório,
Silva, & Magini, 2010; Neto & Magini, 2008; Quinzi, Camomilla,
Di Mario, Felici, & Sbriccoli, 2015; Sbriccoli et al., 2010;Sorensen,
Zacho, Simonsen, Dyhre-Poulsen, & Klausen, 1996), but investi-
gation of the use of core musculature in martial arts perfor-
mance is limited to 1 study the authors are aware of (McGill
et al., 2010). In processing EMG signals, a low pass filter cut-off
frequency of 2.5 Hz was selected to best translate the EMG
signal to that of the force generated by the torso musculature
(Brereton & McGill, 1998). However, this cut-off was determined
in participants who were not highly trained individuals.
Evidence exists that trained individuals are able to recruit a
greater proportion of motor units as well as a reduced electro-
mechanical delay in muscular activation. It is possible that the
participants in this study possessed this characteristic and their
faster response with less electromechanical delay would have
been better represented with a higher cut-off frequency than
an untrained population. However, work by McGill et al. (2010)
with elite MMA fighters investigated this issue and found that
EMG data processed at 3.5 and 4.5 Hz cut-off frequencies did
not affect the pattern of the EMG signal. Given that McGill’s
population of elite MMA fighters falls further away from the
untrained population than club Muay Thai fighters, the use of a
2.5 Hz cut-off frequency for low pass filtering appears justifiable.
Other limitations in the interpretation of this study include
possible day-to-day differences in EMG amplitude. For example,
skin/adipose thickness is known to influence EMG amplitude.
Participants did not change body mass more than 1 kg over the
trial suggesting this is a remote possibility. Further, the EMG
amplitudes were normalised which removes any influence of
differences in day-to-day absolute amplitudes. While there is no
method to compare Dynamic training and Isometric training to
create an equitable workload, we were very conscious of this
and tried to create an equitable volume in terms of duration
and load challenge. Finally, the numbers of participants were
relatively small given the availability of qualified and suitable
athletes. However, even with the limited numbers significance
in effect was found. This suggests that the effect size is prob-
ably larger than was found even with this smaller sample size.
There appears to be a link between enhanced impact force
and increases in torso stiffness. Evidence supports the notion
that Isometric core training enhances torso stiffness (Lee &
McGill, 2015) and effective mass, thus allowing the striker to
impart greater impact forces while minimising or “stiffening
out”any torso eccentric micro-movements. Stiffer core mus-
cles stabilising the spine appear to prevent “energy leaks”or
“giving way”(McGill, 2014 documents several examples). This
appears to be a viable candidate mechanism for enhancing
impact force. This general notion appears to be corroborated
with the increases in peak torso EMG activity, as increased
muscular activation and enhanced synchronicity of core/hip
muscular activation is linked to enhancements in core stiffness
(or what has been termed as a functional “superstiffness”,
Brown & McGill, 2009b; McGill, 2014). It is postulated that
enhanced strike velocity following Dynamic core training is
due to changes in muscle activation patterns (Figure 5).
Existing evidence links rates of muscular relaxation to creating
high-speed athletic movement; literature supports this claim
for martial arts striking (McGill et al., 2010), golf driving, and
sprinting (McGill, 2014). When comparing the differences
between the Isometric and Dynamic protocols, periods of
muscular relaxation in the Dynamic exercises appear to be
the main difference, particularly during the third block of
Dynamic exercises. While little evidence exists on how mus-
cular relaxation can be trained, except for work by Matveyev
(1981), the idea of training rates of muscular relaxation to
mimic this high force, high velocity motor pattern agrees
with the “Principle of Dynamic Correspondence”explained
by Verkhoshansky and Verkhoshansky (2011).
Typical of many studies, this investigation sheds new light
into the core training methods of Muay Thai practitioners, yet
spawns new questions to be answered in the future. Both
Isometric and Dynamic trainings enhanced separate properties
associated with powerful Muay Thai striking. While Isometric
training was superior in enhancing impact force and Dynamic
training was superior for enhancing strike velocity, investiga-
tion into how the 2 programmes performed together would
provide insight on enhancing both aspects. Further, what is
the role of increased EMG activity to the enhanced athletic
characteristics –does increased EMG signal alone affect
impact force and strike speed, or do synchronisation of mus-
cular activation and altered patterns of activation/relaxation
play a larger role? Further, as stated above, little informed
insight exists in how muscular relaxation can be trained. This
study provided some evidence to suggest that Dynamic train-
ing can alter motor patterns to create activation–relaxation
cycles, but further investigation is needed to understand
mechanisms that will underpin eventual application.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
This work was supported by Natural Sciences and Engineering Research
Council of Canada (NSERC).
References
Axler, C. T., & McGill, S. M. (1997). Low back loads over a variety of
abdominal exercises: Searching for the safest abdominal challenge.
Medicine & Science in Sports & Exercise,29(6), 804–811. doi:10.1097/
00005768-199706000-00011
Bergmark, A. (1987). Stability of the lumbar spine A study in mechanical
engineering. Acta Orthopaedica Scandinavica Supplementum,230(60),
25–38.
Blum, H. (1977). Physics and the art of kicking and punching. American
Journal of Physics,45,61–64. doi:10.1119/1.10911
Brereton, L. C., & McGill, S. M. (1998). Frequency response of spine exten-
sors during rapid isometric contractions: Effects of muscle length and
tension. Journal of Electromyography and Kinesiology,8(4), 227–232.
doi:10.1016/S1050-6411(98)00009-1
Brown, S. H. M., & McGill, S. M. (2009a). The intrinsic stiffness of the in vivo
lumbar spine in response to quick releases: Implications for reflexive
requirements. Journal of Electromyography and Kinesiology,19, 727–736.
doi:10.1016/j.jelekin.2008.04.009
Brown, S. H. M., & McGill, S. M. (2009b). Transmission of muscularly
generated force and stiffness between layers of the rat abdominal
wall. Spine,34(2), E70–E75. doi:10.1097/BRS.0b013e31818bd6b1
12 B. LEE AND S. MCGILL
Callaghan, J. P., Gunning, J. L., & McGill, S. M. (1998). Relationship between
lumbar spine load and muscle activity during extensor exercises.
Physical Therapy,78(1), 8–18.
Filimonov, V. I., Koptsev, K. N., Husyanov, Z. M., & Nazarov, S. S. (1985).
Boxing: Means of increasing strength of the punch. National Strength &
Conditioning Association Journal,7(6), 65–66. doi:10.1519/0744-0049
(1985)007<0065:MOISOT>2.3.CO;2
Kavcic, N., Grenier, S. G., & McGill, S. M. (2004). Quantifying tissue loads
and spine stability while performing commonly prescribed low back
stabilization exercises. Spine,29(20), 2319–2329. doi:10.1097/01.
brs.0000142222.62203.67
Lee, B. C. Y., & McGill, S. M. (2015). Effect of long-term isometric training on
core/torso stiffness. The Journal of Strength & Conditioning Research,29
(6), 1515–1526. doi:10.1519/JSC.0000000000000740
Little, J. R. (2001). Bruce Lee: A warrior’s journey. Chicago, IL: Contemporary
Books.
Machado, S. M., Osório, R. A. L., Silva, N. S., & Magini, M. (2010).
Biomechanical analysis of the muscular power of martial arts athletes.
Medical & Biological Engineering & Computing,48(6), 573–577.
doi:10.1007/s11517-010-0608-z
Matveyev, L. (1981). Fundamentals of sports training. Moscow: Progress.
McGill, S. M. (2014). Ultimate back fitness and performance (5th ed.).
Waterloo: Backfitpro.
McGill, S. M. (2016). Low back disorders (3rd ed.). Champaign, IL: Human
Kinetics.
McGill, S. M., Cannon, J., & Anderson, J. (2014). Muscle activity and spine
load during anterior chain whole body linkage exercises: The body saw,
hanging leg raise and walkout from a pushup. Journal of Sports
Sciences, 33(4), 419-426.
McGill, S. M., Chaimberg, J. D., Frost, D. M., & Fenwick, C. M. J. (2010).
Evidence of a double peak in muscle activation to enhance strike speed
and force: An example with elite mixed martial arts fighters. The Journal
of Strength & Conditioning Research,24(2), 348–357. doi:10.1519/
JSC.0b013e3181cc23d5
McGill, S. M., & Karpowicz, A. (2009). Exercises for spine stabilization:
Motion/Motor patterns, stability progressions and clinical technique.
Archives of Physical Medicine and Rehabilitation,90, 118–126.
doi:10.1016/j.apmr.2008.06.026
McGill, S. M., Karpowicz, A., & Fenwick, C. (2009). Ballistic abdominal exercises:
Muscle activation patternsduring three activities along the stability/mobi-
lity continuum. Journal of Strength and Conditioning Research,23(3), 898–
905. doi:10.1519/JSC.0b013e3181a06401
McGill, S. M., Karpowicz, A., Fenwick, C., & Brown, S. H. M. (2009). Exercises
for the torso performed in a standing posture: Spine and hip motion
and motor patterns and spine load. Journal of Strength and
Conditioning Research,23(2), 455–464. doi:10.1519/JSC.0b013e3
181a0227e
McGill, S. M., Marshall, L., & Andersen, J. (2013). Low back loads while
walking and carrying: Comparing the load carried in one hand or in
both hands. Ergonomics,56(2), 293–302. doi:10.1080/
00140139.2012.752528
McGill, S. M., McDermott, A., & Fenwick, C. M. (2009). Comparison of
different strongman events: Trunk muscle activation and lumbar spine
motion, load, and stiffness. The Journal of Strength & Conditioning
Research,23(4), 1148–1161. doi:10.1519/JSC.0b013e318198f8f7
Nesser, T. W., Huxel, K. C., Tincher, J. L., & Okada, T. (2008). The relationship
between core stability and performance in division i football players.
Journal of Strength and Conditioning Research,22(6), 1750–1754.
doi:10.1519/JSC.0b013e3181874564
Neto, O. P., Magini, M., & Saba, M. M. F. (2007). The role of effective mass
and hand speed in the performance of kung fu athletes compared with
nonpractitioners. Journal of Applied Biomechanics,23(2), 139–148.
doi:10.1123/jab.23.2.139
Neto, O. P., & Magini, M. (2008). Electromiographic and kinematic char-
acteristics of Kung Fu Yau-Man palm strike. Journal of Electromyography
and Kinesiology,18(6), 1047–1052.
Pain, M. T. G., & Challis, J. H. (2002). Soft tissue motion during impacts:
Their potential contributions to energy dissipation. Journal of Applied
Biomechanics,18, 231–242. doi:10.1123/jab.18.3.231
Quinzi, F., Camomilla, V., Di Mario, A., Felici, F., & Sbriccoli, P. (2015).
Repeated kicking actions alter technical execution in elite karate
practitioners. International Journal of Sports Physiology and
Performance, 11(3), 363-369.
Sbriccoli, P., Camomilla, V., Di Mario, A., Quinzi, F., Figura, F., & Felici, F.
(2010). Neuromuscular control adaptations in elite athletes: The case of
top level karateka. European Journal of Applied Physiology,108(6), 1269–
1280. doi:10.1007/s00421-009-1338-5
Sorensen, H., Zacho, M., Simonsen, E. B., Dyhre-Poulsen, P., & Klausen, K.
(1996). Dynamics of the martial arts high front kick. Journal of Sports
Sciences,14(6), 483–495. doi:10.1080/02640419608727735
Taanila, H., Suni, J., Pihlajamäki, H., Mattila, V. M., Ohrankämmen, O.,
Vuorinen, P., & Parkkari, J. (2009). Musculoskeletal disorders in physi-
cally active conscripts: A one-year follow-up study in the Finnish
Defence Forces. BMC Musculoskeletal Disorders,10(1), 89. doi:10.1186/
1471-2474-10-89
Tsatsouline, P. (2006). Enter the kettlebell. Minneapolis, MN: Dragon Door
Publications.
Turner,A.N.(2009). Strength and conditioning for Muay Thai athletes. Strength
& Conditioning Journal,31(6), 78–92. doi:10.1519/SSC.0b013e3181b99603
Verkhoshansky, Y., & Verkhoshansky, N. (2011). Special strength training:
Manual for coaches. Rome: Verkhoshansky SSTM.
Willardson, J. M. (2007). Core stability training: Applications to sports
conditioning programs. The Journal of Strength & Conditioning
Research,21(3), 979–985.
JOURNAL OF SPORTS SCIENCES 13