Journal of Electromyography and Kinesiology 25 (2015) 629–636

Contents lists available at ScienceDirect

Journal of Electromyography and Kinesiology

journal homepage: www.elsevier.com/locate/jelekin

Neuromuscular strategies contributing to faster multidirectional agility

performance
Tania Spiteri a,b,⇑, Robert U. Newton a, Sophia Nimphius a
a
School of Health Science, The University of Notre Dame Australia, Fremantle, Australia
b
Centre for Exercise and Sport Science Research, Edith Cowan University, Perth, Australia

a r t i c l e i n f o a b s t r a c t

Article history: The aim of this study was to first determine differences in neuromuscular strategy between a faster and
Received 10 December 2014 slower agility performance, and second compare differences in muscle activation strategy employed
Received in revised form 14 April 2015 when performing two closely executed agility movements. Participants recruited from an elite female
Accepted 14 April 2015
basketball team completed an ultrasound to determine quadriceps muscle-cross sectional area; reactive
isometric mid-thigh pull to determine the rate of muscle activation, rate of force development, pre-motor
time and motor time; and multidirectional agility tests completing two directional changes in response to
Keywords:
EMG
a visual stimulus. Peak and average relative muscle activation of the rectus femoris, vastus medialis, vas-
Co-contraction tus lateralis, biceps femoris, semitendinosus and gastrocnemius were measured 100 ms prior to heel
Pre-activation strike (pre-heel strike) and across stance phase for both directional changes. Faster agility performance
RFD was characterized by greater pre-heel strike muscle activity and greater anterior muscle activation dur-
Decision-making ing stance phase resulting in greater hip and knee extension increasing propulsive impulse. Differences
Processing speed between directional changes appear to result from processing speed, where a greater delay in refractory
times during the second directional change resulted in greater anterior muscle activation, decelerating
the body while movement direction was determined.
Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction Swanik et al., 2007). As a result, slower agility performances are

often characterized by a subsequent delay in processing speed,
Performance during agility maneuvers is determined by an ath- resulting in a longer decision-making time, affecting both
lete’s ability to identify relevant cues within their environment, neuromuscular control and compromising performance outcomes
make the correct decision and rapidly accelerate out of the direc- (Dault et al., 2001; Woo et al., 2006).
tional change (Spiteri et al., 2013; Spiteri and Nimphius, 2013). Producing a faster response to a stimulus when changing direc-
During such a maneuver, efficient neuromuscular control is essen- tion, can result in muscle pre-activation, which has been shown to
tial to execute a coordinated motor response through the integra- protect against injury and increase subsequent movement execu-
tion between cortically-programmed and reflex-mediated muscle tion (Bencke and Zebis, 2011; McBride et al., 2008). Increasing
actions. It is therefore contingent to maintain dynamic restraint preparatory muscle activation through a faster reaction time can
during these high velocity dynamic movements (Swanik et al., increase rate of force development (RFD) and muscular stiffness
2007), reducing the amount of external distractions to produce a during the early phase of movement, enabling more force and
faster performance. Compared to closed environment motor skills impulse to be applied throughout the movement (Aagaard, 2003;
such as sprinting and change of direction; agility requires greater Aagaard et al., 2002; Sleivert and Taingahue, 2004). Greater force
attentional focus to filter irrelevant information and simultane- production during both the braking and propulsive phase of agility
ously execute a complex motor program (Landers et al., 1985; movements has been observed when athletes produce a faster
decision-making time, which subsequently increases exit velocity
out of the directional change due to the higher net impulse
⇑ Corresponding author at: School of Health Science, The University of Notre (Spiteri et al., 2014a). Currently, a majority of research investigat-
Dame Australia, 33 Fillimore Street, Fremantle, WA 6959, Australia. Tel.: +61 8 9433
ing muscle activation strategies during agility movements have sit-
0974.
E-mail addresses: [email protected] (T. Spiteri), [email protected]
uated from an injury prevention perspective, reporting decreased
(R.U. Newton), [email protected] (S. Nimphius). neuromuscular control as a result of insufficient time for the

http://dx.doi.org/10.1016/j.jelekin.2015.04.009
1050-6411/Ó 2015 Elsevier Ltd. All rights reserved.
630 T. Spiteri et al. / Journal of Electromyography and Kinesiology 25 (2015) 629–636

central nervous system to implement appropriate postural adjust- slower group, similar to previous research (Spiteri et al., 2013). A pri-
ment strategies (Besier et al., 2001a), and decreased lower body ori power analysis was performed using results that allowed detec-
strength reducing the amount of structural support during the tion of a significant difference in quadriceps muscle activity in men
movement (Spiteri et al., 2014a). Therefore, it appears both the and women (Padua et al., 2005). Using G⁄Power (Faul et al., 2007) for
nervous and muscular systems can influence performance out- independent means analysis (a = 0.05; b = 0.80; d = 1.54) from pre-
comes as a result of insufficient control (Aagaard, 2003; Aagaard viously mentioned results it was determined at least 12 total partic-
et al., 2002; Hakkinen et al., 1998; Narici et al., 1996), however ipants (6 in each group) were required to achieve an actual power of
the relative involvement and functional significance of these neu- 0.83. Participant characteristics of the faster (n = 6) and slower
romuscular factors to achieve a faster agility performance has yet (n = 6) groups are displayed in Table 1.
to be examined.
Cerebral performance is indirectly measured during reaction
2.2. Procedures
time tasks by quantifying the time between stimulus presentation
and the onset of movement execution providing a measure of pro-
2.2.1. Measurement of muscle cross-sectional area (CSA)
cessing speed (Swanik et al., 2007). Reaction time can be further
Images were captured using a B-mode axial-plane ultrasound
dichotomized into pre-motor and motor time. Pre-motor time is
machine (Aloka SSD-a 10, 6.1.0, Aloka Co., Ltd., Tokyo, Japan), with
the time between stimulus identification and the onset of elec-
a 10 MHz linear-array probe (60-mm width) in extended field of
tromyographic activity, and is reflective of cognitive processes
view mode (gain: 79, constraint: 13, sampling frequency: 24 Hz).
(Van Donkelaar and Franks, 1991). Motor time is the time between
Participants were instructed to relax their leg and rest in a supine
initial muscle activity and movement execution, therefore encom-
position for 20 min to allow fluid shifts in the lower body to stabi-
passing electromechanical delay and is often considered to be the
lize (Noorkoiv et al., 2010). A continuous single image of the
neuromuscular component of processing speed (Van Donkelaar
quadriceps was taken from 50% on the line between the greater
and Franks, 1991). Many factors can affect processing speed and
trochanter and the lateral epicondyle of the femur (Noorkoiv
movement execution including speed of movement execution,
et al., 2010). During the imaging, minimal pressure was applied
number of response alternatives and location of the stimulus
to the skin to avoid compressing the muscle and transmission gel
within the visual field (Brisswalter et al., 2002). During competi-
was used to improve acoustic coupling and image quality (Narici
tion, athletes are required to perform multiple directional changes
et al., 1996; Noorkoiv et al., 2010). Cross-sectional area of the
in close proximity, which requires a rapid response to successive
quadriceps (rectus femoris, vastus lateralis and vastus medialis)
stimuli to pursue opponents. When responding to two closely
of the participant’s dominant leg was measured using ImageJ digi-
spaced stimuli a delay in processing speed is often observed; sub-
tising software (1.41, Wayne Rasband, National Institutes of
sequently termed the psychological refractory period (Knott,
Health, USA), with the average muscle CSA of the three scans used
1970). While a delayed response to a stimulus has been shown
for analysis. Limb dominance or ‘‘preferred limb’’ was defined as
to compromise performance outcomes, a greater delay in process-
the limb that participants used as their preferred takeoff foot when
ing speed to a second stimulus may result in a greater decrease in
performing a lay-up.
neuromuscular control further affecting performance.
Therefore, the primary purpose of this study was to first quan-
tify differences in decision-making time, pre-motor time, motor 2.2.2. Determination of pre-motor and motor time
time, muscle cross-section area (CSA), RFD and muscle activation Participants were instructed to assume an isometric mid-thigh
strategies between faster and slower agility performances. The sec- pull position, with both legs positioned shoulder width apart on
ond purpose was to compare differences in muscle activation the center of a force plate sampling at 2000 Hz (AMTI,
strategies between two closely executed agility movements. BP12001200, Watertown, USA). Hip and knee angles of both legs
were positioned at 140° (Haff and Dumke, 2012). Participants were
instructed to pull on a bar attached to tension removed adjustable
2. Methods straps, driving their feet into the ground ‘‘as hard and as fast as
possible’’ for 5 s, in response to a visual stimulus (light stimulus).
2.1. Participants The light stimulus was positioned at chest height, 3 m in front of
the participant. Participants completed a total of three trials, with
Twelve (N = 12) female basketball athletes (age: 24.25 ± 2.55 yrs; a 30-s rest interval between trials. Reliability of this protocol to
height: 177.69 ± 7.25 cm; body mass: 75.56 ± 14.55 kg) playing for a obtain a measure of peak force was performed in a separate study
professional basketball team within the Women’s National prior to testing (ICC = 0.88, CV = 3.0%).
Basketball League (WNBL) in Australia were recruited for this study. Electromyography (EMG) activity of the rectus femoris, vastus
All participants were recruited from the same WNBL team consisting medialis, vastus lateralis, biceps femoris, semitendinosus and med-
of three guards, six forwards and three centers. To be included for ial gastrocnemius of the dominant leg was recorded using wireless
participation within the study, participants were required to have pre-amplified active surface electrodes sampling at 1000 Hz (Wave
played basketball for a minimum of five years and partake in a min- Cometa, Milano, Italy). Prior to electrode placement, area was
imum of one competitive game and two structured skills training shaved and skin cleaned with alcohol swabs. Two monopolar Ag–
sessions each week. All participants were required to be injury free AgCl surface electrodes were then placed on the muscle belly in
(of the lower limbs) at the time of testing, and report no previous his- the direction of the muscle fibers, in accordance to SENIAM guide-
tory of major lower limb injuries (e.g. anterior-cruciate ligament lines (Hermens et al., 2000). All force, EMG signals and timing data
injuries). Ethics approval was obtained from the University Human was collected and synchronized using Powerlab A/D system
Research Ethics Committee prior to testing and procedures were (ADInstruments, Colorado Springs, CO, USA) and LabChart 5.0 soft-
explained and informed consent was obtained prior to testing. ware (ADInstruments, Colorado Springs, CO, USA). The variables
Participants were separated into faster and slower groups based examined include, pre-motor time (ms), motor time (ms), rate of
on their total running time achieved during the agility test. EMG rise (RER) across 30, 50 and 75 ms time intervals for each
Participants above the 50th percentile were assigned to the faster muscle (Aagaard, 2003), relative RFD across 30, 50, 90 and
group and those below the 50th percentile were assigned to the 100 ms time intervals (Aagaard, 2003), and relative peak force.
 T. Spiteri et al. / Journal of Electromyography and Kinesiology 25 (2015) 629–636 631

Onset of muscle activity was determined visually by the researcher 5. Change direction by 45° to the left and pass the ball right.
based upon the raw EMG signals for each muscle (Rand and 6. Change direction by 45° to the right and pass the ball left.
Ohtsuki, 2000). Pre-motor time was determined by averaging the 7. Fake right, change direction by 45° to the left, and pass the ball
time between stimulus presentation and onset of muscle activity left.
for each muscle (Johnson et al., 2012). Motor time was determined 8. Fake left, change direction by 45° to the right, and pass the ball
by averaging the time between onset of muscle activity for each right.
muscle and onset of vertical force production (Johnson et al.,
2012; Zhou et al., 1996). These variables were averaged across These pre-recorded videos were triggered by an automated pro-
three trials and calculated using a custom written MatLab program gram within the Kinematic Measurement System software (Version
(R2010a, The Mathworks Inc, Chatswood, NSW, AU). 13.0, Innervations, Perth, Australia) which also recorded switch
timings of the light gates with 1 ms resolution. Four in-ground force
2.2.3. Multidirectional agility test plates positioned in a square layout (Fig. 1) were used to capture
Subjects began behind a set of infra-red timing gates (Fitness subjects’ foot plant for the first directional change. Two high-speed
Technology, Adelaide, Australia) on a marked line 19 m opposite video cameras (Sony HDD Camcorder HDRXR550V, Sony Australia)
a projection screen, and were instructed to run in a straight line sampling at 120 Hz were positioned to the left and right of the agi-
toward the projected image (Fig. 1). Once the subject reached the lity course adjacent to both directional changes to capture the
timing gates positioned 7.5 m from the starting position the first visual stimulus and the athlete changing direction to determine
visual stimulus (video clip) was programmed to automatically decision making time. Decision time (ms) was also measured dur-
start. Subjects then responded to the video by changing direction ing each agility trial using Silicon Coach (version, company, city,
on one of four in-ground force plates (Kistler, Type 9290AD, country), and was defined as the time between occlusion of the
USA), 45 ± 5° to the left or right moving in the same direction as video clip to the participant’s first definitive foot strike initiating
the stimulus (i.e. from a defenders perspective). Once changing change of direction movement (Farrow et al., 2005; Serpell et al.,
direction, subjects then ran a further 3.5 m, triggering timing gates 2009). Subjects completed a total of eight trials where the average
positioned 1.5 m after changing direction, to trigger the second approach velocity (m/s), total running time (s), decision time (s) for
visual stimulus (video clip). Subjects then performed a second the first and second directional change and running time to com-
directional change, 45 ± 5° either to the left or right in the same plete the first and second directional change (s) was determined
direction as the stimulus through timing gates to complete the test. for the dominant leg. Reliability of this protocol to determine total
For both directional changes athletes responded to one of these running time (s) was performed in a separate study prior to testing,
eight projected movement patterns (Spiteri et al., 2014b): resulting in high test re-test reliability (ICC = 0.81, CV = 3.3%).

1. Change direction by 45° to the left. 2.3. Data analysis

2. Change direction by 45° to the right.
3. Change direction by 45° to the left and pass the ball left. During the agility trials all force and EMG was synchronized
4. Change direction by 45° to the right and pass the ball right. using Vicon Nexus software (Ver. 1.6.1; Oxford Metrics Ltd.,

Fig. 1. Multidirectional agility test.

632 T. Spiteri et al. / Journal of Electromyography and Kinesiology 25 (2015) 629–636

Oxford, United Kingdom). As the first directional change occurred Table 1

over in-ground force plates, heel strike was defined as the point Participant characteristics and performance times (mean ± SD) for faster and slower
groups during the agility test.
at which vertical ground reaction force exceeded 10 N and toe
off defined as the point at which the vertical GRF is below 10 N Faster (n = 6) Slower (n = 6) p d
(McLean et al., 2004). To determine stance phase for the second Participant characteristics
directional change three retro-reflective markers where placed Age (yrs) 24.33 ± 2.33 23.75 ± 2.69 0.73 0.23
on the lateral malleolus, second metatarsal head and calcaneus Height (cm) 173.8 ± 6.39 179.05 ± 9.69 0.32 0.64
Weight (kg)^ 68.5 ± 5.76 79.7 ± 5.12 0.03 2.05
to determine heel strike and toe off events and using Vicon Rectus femoris CSA (cm2) 20.54 ± 3.42 18.64 ± 3.12 0.13 0.58
Nexus software. All EMG recordings were full-wave rectified, Vastus lateralis CSA (cm2) 22.32 ± 4.02 20.83 ± 3.87 0.19 0.38
and filtered using a band-pass 10–250 Hz, filter (Sigward and Vastus medialis CSA (cm2) 18.50 ± 2.98 16.23 ± 2.04 0.51 0.89
Powers, 2006). Root mean square of the signal was calculated Rel. peak force/RF CSA 0.11 ± 0.03 0.08 ± 0.01 0.03 1.34
(N kg 1) cm 2^
using a 10 ms window across two phases (1) pre-heel strike
Rel. peak force/VM CSA 0.08 ± 0.02 0.06 ± 0.01 0.05 1.26
(defined as 100 ms prior to heel strike) (Sigward and Powers, (N kg 1) cm 2^
2006) and (2) stance phase (heel strike to toe off). The EMG sig- Rel. peak force/VL CSA 0.06 ± 0.01 0.05 ± 0.01 0.10 0.99
nals obtained during the first and second directional change dur- (N kg 1) cm 2
ing the agility assessment were normalized to MVC EMG signals Rel. RFD 30 ms 103.24 ± 69.88 52.85 ± 29.25 0.19 0.94
(N kg 1) ms 1
obtained during the isometric mid-thigh pull. The following
Rel. RFD 50 ms 91.85 ± 40.51 48.34 ± 31.19 0.08 1.20
parameters were calculated for each trial using a custom (N kg 1) ms 1
MATLAB program and expressed as a percentage of MVC (deter- Rel. RFD 90 ms 77.09 ± 21.39 41.98 ± 25.07 0.05 1.50
mined during isometric mid thigh pull???): (1) peak amplitude (N kg 1) ms 1^
Rel. RFD 100 ms 73.47 ± 19.42 41.19 ± 23.47 0.02 1.50
across pre-heel strike and stance phase for each muscle; (2) aver-
(N kg 1) ms 1^
age amplitude during pre-heel strike and stance phase for each Rel. peak force (N kg 1) 1.28 ± 0.39 1.38 ± 0.39 0.80 0.25
muscle; and (3) co-activation index of agonist to antagonist mus-
Performance times
cle activity for the hamstrings to quadriceps muscles (Serpell Approach speed (s) 1.85 ± 0.20 1.89 ± 0.23 0.62 0.18
et al., 2014). First COD time (s)^ 1.72 ± 0.50 2.36 ± 0.12 0.04 0.43
Second COD time (s) 0.83 ± 0.22 1.02 ± 0.11 0.23 1.09
Total running time (s)^ 4.47 ± 0.62 5.34 ± 0.25 0.04 1.84
First COD decision time 0.15 ± 0.03 0.20 ± 0.12 0.03 0.57
2.4. Statistical analysis (ms)^
Second COD decision time 0.17 ± 0.11 0.21 ± 0.09 0.07 1.03
(ms)
All results are presented as means ± SD. Independent t tests
were conducted to examine differences in participant characteris- CSA = cross sectional area; Rel = relative; RF = rectus femoris; VL = vastus lateralis;
tics (height, body mass, age, and muscle CSA), performance times VM = vastus medialis; RFD = rate of force development; N = newton; kg = kilogram;
COD = change of direction.
(approach velocity, decision time, change of direction time, and ^
Indicates significant difference (p 6 0.05) in means between faster and slower
total running time), pre-motor, motor time, and RER. A 2  2 athletes.
multivariate analysis (MANOVA) was conducted to examine dif-
ferences in maximal and average EMG amplitude between faster
Table 2
and slower athletes and first and second directional changes
Rate of EMG rise (ROR; mV) across 30, 50 and 75 ms windows during the maximal
across pre-heel strike and stance phase. A follow up 1-way anal- voluntary contraction (mean ± SD).
ysis of variance (ANOVA) was conducted on each dependent vari-
Faster (n = 6) Slower (n = 6) p d
able to determine precisely where significant differences
occurred. A significance level of p 6 0.05 was employed through- Rectus femoris
ROR 30 ms 210 ± 9.01 160 ± 9.20 0.58 0.55
out all statistical analysis. Effect sizes (d) were calculated for
ROR 50 ms 260 ± 12.80 200 ± 16.32 0.60 0.72
group comparisons by dividing the difference between groups ROR 75 ms 290 ± 11.03 230 ± 10.40 0.64 0.57
by the pooled standard deviation (Cohen, 1988). The magnitude
Vastus lateralis
of effect size calculations were interpreted following Hopkins ROR 30 ms 380 ± 13.21 290 ± 13.58 0.07 0.93
(Hopkins, 2002) guidelines, with trivial = 60; small = 0–0.2; ROR 50 ms 440 ± 11.12 340 ± 14.99 0.07 1.21
moderate = 0.2–0.6; large = 0.6–1.2; very large = 1.2–2.0; nearly ROR 75 ms 490 ± 9.37 380 ± 15.54 0.06 1.51
perfect = 2.0–4.0; perfect = P4.0. All statistical computations were Vastus medialis
performed using a statistical analysis program (SPSS, Version ROR 30 ms^ 330 ± 12.23 140 ± 15.63 0.02 2.21
17.0; Chicago, Illinois). ROR 50 ms^ 390 ± 14.21 150 ± 9.45 0.01 2.40
ROR 75 ms^ 450 ± 21.03 160 ± 13.45 0.03 1.93
Biceps femoris
ROR 30 ms 410 ± 18.39 360 ± 18.42 0.21 0.28
ROR 50 ms 440 ± 19.62 400 ± 14.71 0.38 0.29
3. Results
ROR 75 ms 460 ± 17.92 420 ± 15.12 0.26 0.32
Semitendinosus
Participant characteristics (mean ± SD) for faster and slower
ROR 30 ms 460 ± 16.33 430 ± 13.51 0.26 0.21
groups are shown in Table 1. While there were no significant dif- ROR 50 ms 550 ± 18.54 510 ± 14.38 0.25 0.25
ferences in height, age and muscle CSA between faster or slower ROR 75 ms 650 ± 15.67 590 ± 14.88 0.20 0.62
groups, faster athletes had a significantly lower body mass Gastrocnemius
compared to slower athletes. Faster athletes also demonstrated ROR 30 ms^ 480 ± 11.26 210 ± 11.42 0.03 2.91
significantly greater relative peak force per unit of CSA for the ROR 50 ms^ 530 ± 15.44 260 ± 13.57 0.01 2.50
rectus femoris and vastus medialis (Table 1), and significantly ROR 75 ms^ 550 ± 13.73 290 ± 14.62 0.01 2.66

greater relative isometric RFD at 90 and 100 ms time points ^

Indicates significant difference (p 6 0.05) in means between faster and slower
(Table 1). athletes.
 T. Spiteri et al. / Journal of Electromyography and Kinesiology 25 (2015) 629–636 633

Comparison of performance times (Table 1) indicated that faster greater rectus femoris and vastus medialis activation during the
athletes demonstrated a significantly faster first COD time, deci- first directional change, and significantly greater biceps femoris
sion time for the first directional change, and total running time and medial gastrocnemius activity during both the first and second
compared to slower athletes. There was no significant difference directional change compared to slower athletes. Faster athletes
between faster and slower athletes for approach velocity, second also demonstrated significantly greater peak vastus medialis and
COD time or decision time for the second directional change biceps femoris activation during first and second directional
(Table 1). Faster athletes also demonstrated a significantly faster changes, greater rectus femoris activation during the first direc-
pre-motor time (faster: 0.18 ± 0.02 ms; slower: 0.25 ± 0.03 ms, tional change, and greater medial gastrocnemius activation during
p = 0.02, d = 2.74), motor time (faster: 0.02 ± 0.006 ms; slower: the second directional change (Table 4). Finally the co-activation
0.03 ± 0.001 ms, p = 0.03, d = 2.32), and RER for the vastus medialis, strategy of the hamstrings and quadriceps across stance phase
vastus lateralis and medial gastrocnemius across all time windows revealed faster athletes demonstrate greater hamstring activation
compared to slower athletes (Table 2). during the first directional change (1.17) compared to slower ath-
The comparison of pre-heel strike average muscle activity letes (1.11), which further increases during the second directional
(Table 3), indicates faster athletes demonstrate significantly change (faster: 1.54, slower: 1.06).
greater rectus femoris and vastus medialis activation during both
first and second directional changes, while biceps femoris and 4. Discussion
medial gastrocnemius was significantly greater during the second
directional change compared to slower athletes. When examining The purpose of this study was to first quantify differences in
peak muscle activity across pre-heel strike (Table 4), faster athletes neuromuscular characteristics between faster and slower agility
demonstrated significantly greater rectus femoris, vastus medialis, performances. Findings from this study are that faster athletes
and biceps femoris activation during first and second directional demonstrated a faster total agility time, first COD time and deci-
changes, and greater semitendinosus activation during the second sion time for the first directional change compared to slower ath-
directional change. letes. These differences noted in performance between groups
The comparison of stance phase average muscle activity may be explained by body mass; rate and time to activate muscle
(Table 3), indicates faster athletes demonstrate significantly as measured by RER, pre-motor time and motor time; amplitude of

Table 3
Average amplitude muscle activation as a percentage of MVC (mean ± SD) during pre-activation and stance phase.

First directional change Second directional change

Faster (n = 6) Slower (n = 6) p ES Faster (n = 6) Slower (n = 6) p ES
Pre-activation (100 ms prior to heel strike)
Rectus femoris 41 ± 9^,* 29 ± 10 0.05 1.26 28 ± 10^ 16 ± 4 0.05 1.58
Vastus lateralis 23 ± 10 18 ± 7 0.12 0.58 20 ± 9^ 12 ± 4 0.03 1.15
Vastus medialis 24 ± 8^ 10 ± 4 0.05 2.21 16 ± 6^ 8±3 0.04 1.69
Biceps femoris 30 ± 10 18 ± 6 0.24 1.46 38 ± 11^ 18 ± 7 0.05 2.17
Semitendinosus 22 ± 7 12 ± 4 0.11 1.75 23 ± 11 17 ± 6 0.12 0.68
Gastrocnemius 29 ± 11 19 ± 8 0.14 1.04 25 ± 9^ 17 ± 7 0.05 0.98

Stance phase (heel strike to toe off)

Rectus femoris 88 ± 23^,* 35 ± 15 0.03 2.73 35 ± 10 27 ± 9 0.12 0.85
Vastus lateralis 24 ± 5 20 ± 6 0.06 0.76 20 ± 7 18 ± 6 0.21 0.31
Vastus medialis 34 ± 8^,* 15 ± 7 0.02 2.53 15 ± 5 11 ± 7 0.58 0.66
Biceps femoris 40 ± 12^,* 15 ± 9 0.02 2.43 34 ± 6^ 13 ± 5 0.02 1.99
Semitendinosus 28 ± 7 24 ± 7 0.24 0.57 16 ± 4 18 ± 6 0.11 0.39
Gastrocnemius 70 ± 18^,* 48 ± 12 0.02 1.44 47 ± 11^ 38 ± 8 0.04 0.94
^
Indicates significant difference (p 6 0.05) in average amplitude muscle activation between faster and slower athletes.
*
Indicates significant difference (p 6 0.05) in average amplitude muscle activation between faster athletes first and second directional change.

Table 4
Average peak muscle activation as a percentage of MVC (mean ± SD) during pre-activation and stance phase.

First directional change Second directional change

Faster (n = 6) Slower (n = 6) p ES Faster (n = 6) Slower (n = 6) p ES
Pre-activation (100 ms prior to heel strike)
Rectus femoris 64 ± 21^ 24 ± 11 0.02 2.39 41 ± 13^ 17 ± 6 0.01 2.37
Vastus lateralis 35 ± 16 18 ± 4 0.06 1.46 20 ± 5 16 ± 4 0.58 0.88
Vastus medialis 37 ± 10^,* 17 ± 3 0.03 2.71 34 ± 12^ 12 ± 6 0.02 2.32
Biceps femoris 47 ± 22^,* 15 ± 6 0.02 1.98 71 ± 28^ 23 ± 7 0.01 2.35
Semitendinosus 32 ± 12 12 ± 6 0.14 2.11 41 ± 18^ 22 ± 5 0.02 1.44
Gastrocnemius 35 ± 17 22 ± 12 0.27 0.88 50 ± 26 20 ± 9 0.26 1.54

Stance phase (heel strike to toe off)

Rectus femoris 120 ± 35^,* 47 ± 24 0.01 2.43 43 ± 14 29 ± 11 0.32 1.11
Vastus lateralis 35 ± 17 29 ± 17 0.24 0.35 28 ± 8 23 ± 10 0.21 0.55
Vastus medialis 50 ± 18^ 20 ± 9 0.02 2.11 36 ± 17^ 13 ± 4 0.05 1.86
Biceps femoris 49 ± 12^ 25 ± 10 0.02 2.17 73 ± 25^ 30 ± 13 0.03 2.16
Semitendinosus 33 ± 19 37 ± 17 0.53 0.22 36 ± 20 34 ± 15 0.62 0.11
Gastrocnemius 97 ± 32 68 ± 33 0.15 0.89 82 ± 24^ 36 ± 19 0.05 2.12
^
Indicates significant difference (p 6 0.05) in average peak muscle activation between faster and slower athletes.
*
Indicates significant difference (p 6 0.05) in average peak muscle activation between faster athletes first and second directional change.
634 T. Spiteri et al. / Journal of Electromyography and Kinesiology 25 (2015) 629–636

activation as measured by average and peak muscle activity; and contribution of semitendinosus and gastrocnemius muscle activity.
muscle activation strategy employed as determined by the ham- Previous studies have observed greater pre-activity of the exten-
string to quadriceps ratio. The second purpose of this study was sors; biceps femoris and semitendinosus, and the gastrocnemius
to compare differences in muscle activation strategies during two with increases in running velocity (Kyröläinen et al., 2005). As
closely executed agility movements. Surprisingly, as no differences velocity increases, the ground reaction force required to decelerate
were observed in decision time during the second directional and change direction, can peak at up to three times body mass
change, the differences observed between two closely executed (Spiteri et al., 2013). Therefore, greater pre-activity observed dur-
directional changes appears to be the direct result of the individu- ing the second directional change appears to be a preparatory
als’ capacity to activate and their subsequent muscle activation requirement for greater EMG activity during the initial braking
strategy employed. phase of the movement to tolerate increased ground reaction
When examining isolated reaction time, multiple studies have forces as a result of increased entry velocity.
reported a faster performance when a faster initial reaction time Across the stance phase during first and second directional
is observed (Abernethy and Russell, 1987; Spiteri et al., 2012). changes, faster athletes demonstrated significantly greater vastus
Previous findings are supported by the current study as a result medialis, biceps femoris and gastrocnemius muscle activity, with
of observing a faster pre-motor time and motor time in response rectus femoris amplitude only significantly different during the
to a light stimulus within faster athletes. While the use of a non- first directional change. Greater muscle activity can result from
specific stimuli has been found to eclipse differences between ath- multiple factors, including rate of neural firing, number of motor
letes of varying expertise (Baker et al., 2003), a quicker first COD units activated, muscle CSA, and muscle architecture (Aagaard,
decision time, suggests faster athletes have the isolated ability to 2003; Aagaard et al., 2000). While no significant difference was
produce a rapid response to a stimulus in game situations. observed in quadriceps CSA in isolation between groups, signifi-
However, similar to the agility test in the current study, game envi- cantly greater relative peak force per rectus femoris and vastus
ronments typically require a faster response to successive stimuli. medialis CSA was observed for faster athletes (Table 1). The
Typically, when stimuli are presented in pairs, reaction time to the increased capacity to produce more force per unit of CSA, resulted
second stimulus is delayed; termed the psychological refractory in faster athletes producing greater contractile RFD during the
period (Knott, 1970; Welford, 1952). The non-significant difference early phase of isometric mid-thigh pull (<100 ms), despite no dif-
in the second COD decision time between groups is similar to pre- ference in relative peak isometric force between faster and slower
vious studies examining differences in paired reaction time athletes. Therefore, we can in part assume the difference in muscle
between athletes and non-athletes (Knott, 1970). Delayed process- activity, is a result of neural factors as faster athletes demonstrated
ing speed in response to the second stimulus supports the theory a significantly greater RFD at 90 and 100 ms (Table 1), and greater
by Welford (1952) that suggests while the first stimulus is being RER for the vastus medialis and gastrocnemius across 30, 50 and
processed the second is being ‘‘held in storage’’ as a result of single 75 ms time intervals (Table 2). Typically when changing direction
channel processing. As response to the first stimulus appears to be at 45° angles the stance phase is approximately 250 ms in dura-
given extra emphasis and in the case of faster athletes producing a tion (Spiteri et al., 2013). Therefore the ability to rapidly activate
significantly faster first decision time and subsequent faster first within this time period as observed by faster athletes, increases
COD time, the refractory times of processing speed to the second the ability to develop muscular force and accelerate out of the
stimulus were unexpectedly slow. These findings together suggest directional change. Faster rate of activation and amplitude of vas-
that when responding to two closely spaced stimuli, differences tus medialis and gastrocnemius activation has been associated
observed between groups appear to be attributed to the ability to with greater propulsive ability (Kyröläinen et al., 2005), enabling
activate the musculoskeletal system (faster RER) as measured dur- faster athletes of the current study to produce a faster propulsive
ing the isometric mid thigh pull. Faster rate of activation of the phase of the agility movement and achieve a faster first COD time
neuromuscular system is advantageous to performance outcomes, compared to slower athletes. Greater activation of biarticular mus-
particularly during high velocity movements like agility, enabling cles, specifically the biceps femoris, has been shown to increase the
the body to pre-activate and rapidly respond to changes within storage of elastic energy during the swing phase of cyclical move-
the environment. ments, releasing this stored energy at ground contact, increasing
Muscle pre-activation is a centrally programmed mechanism propulsion in the intended direction (Simonsen et al., 1985).
(Ferris and Farley, 1997), with the amplitude and timing of muscu- Increased activation of the biceps femoris across the movement
lar activation dependent upon the expectancy of ground contact suggests faster athletes utilized this stored elastic energy during
and movement velocity (Ferris and Farley, 1997; Serpell et al., the braking phase of the movement, demonstrating greater hip
2014). As observed in faster athletes of the current study, a quicker extension during the propulsive phase of the movement increasing
decision time resulted in greater pre-heel strike muscle activity exit velocity. Similarly, greater rectus femoris activation during the
manifesting into quicker ground contact times. Greater pre-activa- first directional change suggests faster athletes accelerated out of
tion of the muscles has been shown to increase muscular stiffness the propulsive phase with greater hip and knee extension com-
preparing the lower body for the greater ground reaction forces pared to slower athletes assisting to produce a faster first COD per-
and moments that are applied at heel strike (Serpell et al., 2012). formance. However, as no significant difference was observed in
Faster athletes in the current study demonstrated a significantly rectus femoris muscle activation during the second directional
greater peak and average rectus femoris and vastus medialis mus- change, and similarly no difference observed in second COD time,
cle activity, and biceps femoris peak muscle activity during the first the absence of performance differences between groups suggests
directional change, demonstrating they approached heel strike a different movement strategy may have been employed when
with greater hamstring and quadriceps co-activation. Increased required to respond and change direction to successive stimuli.
muscle stiffness, particularly through greater hamstring (biceps During the first directional change, faster athletes demon-
femoris) activity prior to and at heel strike, reduces shear forces strated a greater hamstring to quadriceps ratio compared to slower
at the knee (Lloyd, 2001) and increases resultant impulse during athletes. This ratio increased during the second directional change
the braking phase of agility movements, subsequently increasing for faster athletes. However, slower athletes demonstrated a
an athletes’ exit velocity (Spiteri et al., 2013,a). During the second decrease in co-contraction. This suggests faster athletes utilize a
directional change the velocity that faster athletes approached the different activation strategy between the two directional changes,
directional change appears to have contributed to greater demonstrating greater anterior chain activation during the first
 T. Spiteri et al. / Journal of Electromyography and Kinesiology 25 (2015) 629–636 635

directional change and posterior chain activation during the sec- times and processing speed when responding to the second stimu-
ond directional change. During the first directional change signifi- lus, resulting in the different muscular strategy employed.
cantly greater rectus femoris and vastus medialis activation was
observed for faster athletes, increasing knee and hip extension dur- Funding statement
ing the propulsive phase of the movement resulting in a signifi-
cantly faster first COD time. As faster athletes entered the second No external funding was received for this work.
directional change at a higher velocity and displayed a longer deci-
sion time, greater hamstring activation would be required to decel-
Conflict of interest
erate the body enabling more time to determine subsequent
movement direction. This finding supports previous research
There are no conflicts of interest concerning this paper.
observing greater force and impulse during the braking phase of
agility movements where processing speed and knowledge of
References
movement direction is delayed (Spiteri et al., 2014a). In contrast,
the slower group demonstrated a longer decision time during both Aagaard P. Training-induced changes in neural function. Exerc Sport Sci Rev
directional changes, a reduced ability to rapidly activate as 2003;31:61–7.
assessed by RER and a reduced hamstring to quadriceps activation Aagaard P, Simonsen E, Andersen J, Magnusson S, Halkjaer-Kristensen J, Dyhre-
Poulsen P. Neural inhibition during maximal eccentric and concentric
ratio. These qualities are commonly observed in athletes that exhi- quadriceps contraction: effects of resistance training. J Appl Physiol
bit greater knee moments (Besier et al., 2001b; Lloyd, 2001) and 2000;89:2249–57.
therefore insufficient neuromuscular control during directional Aagaard P, Simonsen EB, Andersen JL, Magnusson P, Dyhre-Poulsen P. Increased rate
of force development and neural drive of human skeletal muscle following
changes. Previous research has identified that a longer decision resistance training. J Appl Physiol 2002;93:1318–26.
time reduces time to implement appropriate postural adjustment Abernethy B, Russell DG. Expert-novice differences in an applied selective attention
strategies (Besier et al., 2001a) increasing injury risk and decreas- task. J Sport Psychol 1987;9:326–45.
Baker J, Cote J, Abernethy B. Learning from the experts: practice activities of expert
ing performance outcomes. While a reduced co-contraction decision makers in sport. Res Q Exerc Sport 2003;74:342–7.
increases injury risk, it also decreases amount of force and impulse Bencke J, Zebis MK. The influence of gender on neuromuscular pre-activity during
that can be applied throughout the movement (Spiteri et al., 2013, side-cutting. J Electromyogr Kinesiol 2011;21:371–5.
Besier TF, Lloyd DG, Ackland TR, Cochrane JL. Anticipatory effects on knee joint
2014a), resulting in a slower agility performance. loading during running and cutting maneuvers. Med Sci Sports Exerc
While the current study provides an understanding of the neu- 2001a;33:1176–81.
romuscular factors contributing to a faster multidirectional agility Besier TF, Lloyd DG, Cochrane JL, Ackland TR. External loading of the knee joint
during running and cutting maneuvers. Med Sci Sports Exerc
performance, these findings are limited to a small population
2001b;33:1168–75.
(n = 12) of elite female basketball athletes. While this small sample Brisswalter J, Collardeau M, René A. Effects of acute physical exercise characteristics
size may have prevented realization of further statistical signifi- on cognitive performance. Sports Med 2002;32:555–66.
cance in some variables, the athletes within the study were Cohen J. Statistical power analysis for the behavioural sciences. 2nd ed. Hillsdale
(NJ): Lawrence Earlbaum Associates; 1988.
recruited from the same team and therefore exposed to the same Dault MC, Frank JS, Allard F. Influence of a visuo-spatial, verbal and central
training stimulus, physical conditioning, and playing experience executive working memory task on postural control. Gait Posture
reducing the amount of variability within dynamic EMG measure- 2001;14:110–6.
Farrow D, Young W, Bruce L. The development of a test of reactive agility for
ments due to other factors such as level of expertise and training netball: a new methodology. J Sci Med Sport 2005;8:52–60.
history. It should also be noted that while known error exists when Faul F, Erdfelder E, Lang A-G, Buchner A. G⁄ Power 3: a flexible statistical power
examining muscle activity during dynamic multi-joint movements, analysis program for the social, behavioral, and biomedical sciences. Behav Res
Method 2007;39:175–91.
precautions were taken during the data collection and data analy- Ferris DP, Farley CT. Interaction of leg stiffness and surface stiffness during human
sis phase to reduce this error, however, this should be taken into hopping. J Appl Physiol 1997;82:15–22.
consideration when interpreting the findings of the current study. Haff GG, Dumke C. Laboratory Manual for Exercise Physiology: Human Kinetics;
2012.
A notable strength of this study is that all participants were elite Hakkinen K, Newton RU, Gordon SE, McCormick M, Volek JS, Nindl BC, et al. Changes
female athletes which is often difficult to achieve in research and in muscle morphology, electromyographic activity, and force production
allows greater translation to high performance sport. characteristics during progressive strength training in young and older men. J
Gerontol 1998;53A:415–23.
Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of recommendations
for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol
4.1. Conclusion 2000;10:361–74.
Hopkins WG. A new view of statistics. Internet Society for Sport Science. New View
of Statistics Web Site; 2002.
In summary, findings from the present study suggest are that Johnson ST, Kipp K, Hoffman MA. Spinal motor control differences between the
faster rate of muscle activation and greater pre-heel strike muscle sexes. Eur J Appl Physiol 2012;112(11):3859–64.
activity is indicative of an increased preparatory response for Knott J. The effects of the phenomenon of the psychological refractory period upon
college athletes and non-athletes. Masters Thesis & Specialist Projects; 1970.
greater muscle activity upon entry into the stance phase of the agi- Kyröläinen H, Avela J, Komi PV. Changes in muscle activity with increasing running
lity movement. Greater activation of the biceps femoris and rectus speed. J Sport Sci 2005;23:1101–9.
femoris, along with a faster rate of activation of the vastus medialis Landers D, Wang M, Courtet P. Peripheral narrowing among experienced and
inexperienced rifle shooters under low-and high-stress conditions. Res Q Exerc
and gastrocnemius resulted in greater hip and knee extension Sport 1985;56:122–30.
during the propulsive phase of the movement, assisting to produce Lloyd DG. Rationale for training programs to reduce anterior cruciate ligament
a faster agility performance. When examining the movement strat- injuries in Australian football. J Orthod Sports Phys Ther 2001;31:645–54.
McBride JM, McCaulley GO, Cormie P. Influence of preactivity and eccentric muscle
egy in response to two closely spaced stimuli it appears an offensive activity on concentric performance during vertical jumping. J Strength Cond Res
strategy is employed when responding to the first stimulus, as a fas- 2008;22:750–7.
ter decision making time and greater anterior chain activation of McLean SG, Lipfert SW, van den Bogert AJ. Effect of gender and defensive opponent
on the biomechanics of sidestep cutting. Med Sci Sports Exerc
the vastus medialis and rectus femoris was observed leading to a
2004;36:1008–16.
faster first COD time. In contrast when performing the second direc- Narici MV, Hoppeler H, Kayser B, Landoni L, Claassen H, Gavardi C, et al. Human
tional change a longer decision time and greater hamstring activa- quadriceps cross-sectional area, torque and neural activation during 6 months
tion was demonstrated as a result of greater deceleration to strength training. Acta Physiol Scand 1996;157:175–86.
Noorkoiv M, Nosaka K, Blazevich AJ. Assessment of quadriceps muscle cross
determine subsequent movement direction, indicative of a defen- sectional area by ultrasound extended-field-of-view imaging. Eur J Appl Physiol
sive approach. However, this is likely due to the delay in refractory 2010;109:631–9.
636 T. Spiteri et al. / Journal of Electromyography and Kinesiology 25 (2015) 629–636

Padua DA, Carcia CR, Arnold BL, Granata KP. Gender differences in leg stiffness and Tania Spiteri – Employment: Lecturer, Sport Science
stiffness recruitment strategy during two-legged hopping. J Mot Behav and Sport Biomechanist, Wheelchair Basketball
2005;37:111–26. Australia. Qualifications: MSc, ESSAM. Research
Rand MK, Ohtsuki T. EMG analysis of lower limb muscles in humans during quick Interests: Agility; Change of Direction; Athletic pro-
change in running directions. Gait Posture 2000;12:169–83. filing and development; Skill Acquisition; Gender
Serpell BG, Ford M, Young WB. The development of a new test of agility for rugby differences in performance.
league. J Strength Cond Res 2009;24(12):3270–7.
Serpell BG, Ball NB, Scarvell JM, Smith PN. A review of models of vertical, leg, and
knee stiffness in adults for running, jumping or hopping tasks. J Sports Sci
2012;30:1347–63.
Serpell BG, Ball NB, Scarvell JM, Buttfield A, Smith PN. Muscle
pre-activation strategies play a role in modulating Kvert for change of
direction manoeuvres: an observational study. J Electromyogr Kinesiol
2014;24(5):704–10.
Sigward SM, Powers CM. The influence of gender on knee kinematics, kinetics and
muscle activation patterns during side-step cutting. Clin Biomech 2006;21:
41–8.
Simonsen EB, Thomsen L, Klausen K. Activity of mono-and biarticular leg muscles Professor Robert Newton – Employment:
during sprint running. Eur J Appl Physiol Occup Physiol 1985;54:524–32. Foundation Professor, Exercise and Sport Science.
Sleivert G, Taingahue M. The relationship between maximal jump-squat power and Qualifications: PhD, CSCS⁄D, AEP, FESSA, FNSCA.
sprint acceleration in athletes. Euro J Appl Physiol 2004;91:46–52. Research Interests: Muscular strength and power;
Spiteri T, Nimphius S. Relationship between timing variables and plant foot kinetics Athletic profiling and long-term monitoring; Exercise
during change of direction movements. J Aust Strength Cond 2013;21(S1):
as medicine; Prostate Cancer.
56–62.
Spiteri T, Cochrane JL, Nimphius S. The evaluation of a new lower body reaction
time test. J Strength Cond Res 2012;27(1):174–80.
Spiteri T, Cochrane JL, Hart NH, Haff GG, Nimphius S. Effect of strength on plant foot
kinetics and kinematics during a change of direction task. Eur J Sport Sci
2013;13(6):646–52.
Spiteri T, Hart NH, Nimphius S. Offensive and defensive agility: a sex comparison of
lower body kinematics and ground reaction forces. J Appl Biomech 2014a;30:
514–20.
Spiteri T, Nimphius S, Specos C, Hart NH, Sheppard J, Newton R. The contribution of
strength characteristics to change of direction and agility in female basketball
athletes. J Strength Cond Res 2014b;28:2415–23. Dr. Sophia Nimphius – Employment: Senior
Swanik CB, Covassin T, Stearne DJ, Schatz P. The relationship between
Lecturer, Strength and Conditioning. Qualifications:
neurocognitive function and noncontact anterior cruciate ligament injuries.
PhD, CSCS⁄D. Research Interests: Profiling and long-
Am J Sports Med 2007;35:943–8.
term training adaptations of youth and elite athletes;
Van Donkelaar P, Franks I. The effects of changing movement velocity and
complexity on response preparation: evidence from latency, kinematic, and Measurement and development of strength and
EMG measures. Exp Brain Res 1991;83:618–32. power; Gender differences in athletic performance.
Welford AT. The ‘psychological refractory period’ and the timing of high-speed
performance—a review and a theory. Br. J. Psychol. 1952;43:2–19.
Woo E, Burns Y, Johnston L. The effect of task uncertainty on muscle activation
patterns in 8–10 year old children. Physiother Res Int 2006;8:143–54.
Zhou S, McKenna MJ, Lawson DL, Morrison WE, Fairweather I. Effects of fatigue and
sprint training on electromechanical delay on knee extensor muscles. Eur J Appl
Physiol 1996;72:410–6.