This article was downloaded by: [Tulane University]

On: 02 September 2014, At: 02:13

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Journal of Sports Sciences

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

Determinants of countermovement jump performance:

a kinetic and kinematic analysis
a a a
Stuart McErlain-Naylor , Mark King & Matthew Thomas Gerard Pain
a
School of Sport, Exercise and Health Sciences, Loughborough University, Loughborough, UK
Published online: 29 May 2014.

To cite this article: Stuart McErlain-Naylor, Mark King & Matthew Thomas Gerard Pain (2014) Determinants of
countermovement jump performance: a kinetic and kinematic analysis, Journal of Sports Sciences, 32:19, 1805-1812, DOI:
10.1080/02640414.2014.924055

To link to this article: http://dx.doi.org/10.1080/02640414.2014.924055

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 Journal of Sports Sciences, 2014
Vol. 32, No. 19, 1805–1812, http://dx.doi.org/10.1080/02640414.2014.924055

Determinants of countermovement jump performance: a kinetic and

kinematic analysis

STUART MCERLAIN-NAYLOR, MARK KING & MATTHEW THOMAS GERARD PAIN

School of Sport, Exercise and Health Sciences, Loughborough University, Loughborough, UK

(Accepted 10 May 2014)

Abstract
This study aimed to investigate the contributions of kinetic and kinematic parameters to inter-individual variation in
Downloaded by [Tulane University] at 02:13 02 September 2014

countermovement jump (CMJ) performance. Two-dimensional kinematic data and ground reaction forces during a CMJ
were recorded for 18 males of varying jumping experience. Ten kinetic and eight kinematic parameters were determined for
each performance, describing peak lower-limb joint torques and powers, concentric knee extension rate of torque develop-
ment and CMJ technique. Participants also completed a series of isometric knee extensions to measure the rate of torque
development and peak torque. CMJ height ranged from 0.38 to 0.73 m (mean 0.55 ± 0.09 m). CMJ peak knee power, peak
ankle power and take-off shoulder angle explained 74% of this observed variation. CMJ kinematic (58%) and CMJ kinetic
(57%) parameters explained a much larger proportion of the jump height variation than the isometric parameters (18%),
suggesting that coachable technique factors and the joint kinetics during the jump are important determinants of CMJ
performance. Technique, specifically greater ankle plantar-flexion and shoulder flexion at take-off (together explaining 58%
of the CMJ height variation), likely influences the extent to which maximal muscle capabilities can be utilised during the
jump.

Keywords: countermovement jump, kinetics, kinematics, technique, rate of torque development

Introduction earlier, neurally mediated rate of torque development

(Tillin, Pain, & Folland, 2013).
The countermovement jump (CMJ) is a key perfor-
There are discrepancies among the results of the
mance requirement in many sports. Research has
few studies investigating ankle, knee and hip joint
shown positive relationships between lower-limb
contributions during the CMJ. Hubley and Wells
strength and power measures and CMJ performance
(1983) found the greatest contributor to be the knee
(Ashley & Weiss, 1994; Nuzzo, McBride, Cormie, &
joint (49% of the total positive work), whilst Fukashiro
McCaulley, 2008; Sheppard et al., 2008; Wisløff,
and Komi (1987) found it to be the hip joint (51%).
Castagna, Helgerud, Jones, & Hoff, 2004).
More recently, Vanezis and Lees (2005) obtained
However, further research is needed to explain quan-
values (30% at the knee and 42% at the hip) that
titatively the relative contributions of kinetic and kine-
were in closer agreement with Fukashiro and Komi
matic variables to this movement. Of the kinetic
(1987) than with Hubley and Wells (1983). A novel
determinants, greater rate of force or torque develop-
finding by Vanezis and Lees (2005) was a negative
ment has frequently been associated with increased
relationship between hip work and knee work, indicat-
CMJ performance (De Ruiter, Van Leeuwen,
ing a technique difference between participants. The
Heijblom, Bobbert, & De Haan, 2006; Marcora &
authors suggested that this difference could account
Miller, 2000; McLellan, Lovell, & Gass, 2011;
for previous discrepancies in the literature, implying
Thompson et al., 2013). Both rate of force develop-
that technique determines the relative contribution of
ment and rate of torque development measure the
different lower-limb joints. The inclusion of an arm
capabilities of skeletal muscle to rapidly generate
swing increased jump height by approximately 10 cm,
muscle forces and for the purposes of the present
supporting previous arm swing-induced performance
study will be referred to as rate of torque develop-
improvements (Feltner, Fraschetti, & Crisp, 1999;
ment. Furthermore, it seems that the maximum
Shetty & Etnyre, 1989). Likely contributors to this
force-dependent, post-50 ms, rate of torque develop-
effect include the increase in work done at the hip
ment is more strongly related to CMJ height than the
joint (Hara, Shibayama, Takeshita, & Fukashiro,

Correspondence: Mark King, School of Sport, Exercise and Health Sciences, Loughborough University, Loughborough, UK. E-mail: [email protected]

© 2014 Taylor & Francis

 1806 S. McErlain-Naylor et al.

2006; Lees, Vanrenterghem, & De Clercq, 2004) and difference between good and poor jumpers are iden-
maximised pre-take-off mass centre displacement tified. This necessitates the recruitment of a hetero-
(Cheng, Wang, Chen, Wu, & Chiu, 2008; Harman, geneous ability range to the sample population so
Rosenstein, Frykman, & Rosenstein, 1990; Payne, that the effects of variability in each of the kinetic
Slater, & Telford, 1968) in jumps with an arm and kinematic variables can be observed. The pur-
swing. Simulation studies of squat jumping show the pose of the present study is therefore to quantify the
augmented hip work to be due to a slowing of hip relative contributions of these factors in order to
extension enabling the musculature to work on a identify the most important determinants of CMJ
more favourable region of the force–velocity curve height.
(Blache & Monteil, 2013; Cheng et al., 2008;
Domire & Challis, 2010). Approximately one-third
Methods
of the arm swing-related performance improvement
results from the work and energy induced at the Eighteen physically active males (21.2 ± 2.2 years,
shoulder joint (Domire & Challis, 2010). 1.80 ± 0.08 m, 78.1 ± 9.2 kg, mean ± SD) partici-
Countermovement depth has also been linked to pated in this investigation. Participants with large
CMJ performance. Moran and Wallace (2007) found variation in jumping experience were selected so as
that increasing the knee joint range of motion from not to distort the importance of individual variables.
Downloaded by [Tulane University] at 02:13 02 September 2014

70° to 90° resulted in a 17% improvement in CMJ The testing procedures were explained to each par-
height. Similarly, high ankle dorsi-flexion range of ticipant, and informed consent was obtained in
motion has been shown to contribute to CMJ perfor- accordance with the Loughborough University
mance in men but not women (Georgios, Fotis, Ethical Advisory Committee.
Thomas, Vassilios, & Iraklis, 2006). Simulation stu- Participants attended two laboratory testing ses-
dies have shown an increase in squat depth to improve sions: (1) isometric knee extension measurement and
squat jump performance due to an increase in time to (2) anthropometric and CMJ measurement. They
develop joint torques (Bobbert, Casius, Sijpkens, & were required to refrain from strenuous physical activ-
Jaspers, 2008; Domire & Challis, 2007). Proposed ity for 36 h prior to each session. The knee extensor
mechanisms for the benefit of the countermovement contractile properties of the dominant leg were tested
phase in a CMJ include the development of active using a dynamometer (Con-Trex; CMV Aargau,
state prior to concentric action (Bobbert & Casius, Switzerland; hip angle 100°; frequency 512 Hz).
2005; Bobbert, Gerritsen, Litjens, & Van Soest, Following dynamic stretching and submaximal
1996), tendon elastic recoil (Alexander, 1995), and warm-up trials of incremental intensity, isometric uni-
the enhancement of subsequent force following mus- lateral knee extension torque was measured at five
cle stretch (Edman, Elzinga, & Noble, 1978). angles (15°, 30°, 45°, 60° and 75°; 0° indicated a
Few researchers have compared kinetic and kine- fully extended leg) in a randomised order. Two trials
matic CMJ determinants. Vanezis and Lees (2005) were recorded at each angle, separated by 2 min rest: a
concluded that kinematic technique factors were less 5 s maximal voluntary contraction and then a measure
important than muscle capabilities, although their of rate of torque development, with the participant
technique analysis was limited to the timing of the instructed to increase their knee extension torque as
lowest vertical mass centre position and the use (or fast as possible (Sahaly, Vandewalle, Driss, & Monod,
not) of an arm swing. An increase in strength does not 2001). The participants rested for 3 min between each
always result in a subsequent performance improve- knee angle. The peak isometric knee extension torque
ment (Clutch, Wilton, McGown, & Bryce, 1983), was identified as the highest of the angle-specific peak
perhaps due to the need to adapt coordination follow- torques. The rate of torque development trial at the
ing strength gains (Bobbert & Van Soest, 1994). angle corresponding to peak isometric torque was
Supporting the importance of appropriate technique used to obtain the rate of change of joint torque
utilisation, Luhtanen and Komi (1978) reported that in 50 ms intervals (RTD0–50, RTD50–100 and
well-trained participants were able to utilise only 76% RTD100–150) from 0 to 150 ms after the onset of
of the available mechanical energy during a CMJ but contraction (identified manually; Tillin, Jimenez-
that optimal coordination could increase this to 84%. Reyes, Pain, & Folland, 2010). This enabled the
It is evident that in order to gain a broad understand- investigation of the earlier agonist neural drive domi-
ing of the determinants of CMJ performance, it is nated and later maximal voluntary torque dominated
necessary to study both kinetic and kinematic rate of torque development (Andersen & Aagaard,
variables. 2006; Tillin, Pain, & Folland, 2012). All isometric
If the findings of this study are to be practically parameters were normalised to body mass.
applicable when considering progression from poor For the CMJ measurement, thirty-eight 14 mm
to good countermovement jumping ability, then it is retro-reflective markers were attached to each partici-
important that variables contributing to the pant, positioned over bony landmarks. The metatarso-
 Determinants of countermovement jump performance 1807

phalangeal, ankle, knee, shoulder, elbow and wrist from 0 to 200 ms of knee extension in 50 ms intervals
joint centres were calculated from a pair of markers (CMJ RTD0–50, CMJ RTD50–100, CMJ RTD100–150
placed across the joint so that their mid-point coin- and CMJ RTD150–200). Eight kinematic parameters
cided with the joint centre, similarly to Ranson, King, were also defined: minimum absolute joint angles and
Burnett, Worthington, and Shine (2009). The centre those at take-off for the ankle, knee, hip and shoulder.
of the neck was defined as the midpoint between two Shoulder extension beyond the line of the greater
markers positioned over the sternoclavicular notch trochanter to glenohumeral joint was defined as nega-
and the C7 vertebra. The centre of the head was tive, with flexion forwards from this line positive.
defined as the average position of four markers, and All statistical analyses were performed within
the hip joint centres were calculated from four markers SPSS v.20 (IBM Corporation, Armonk, NY,
placed over the left and right anterior and posterior USA). To address the aim of the study and identify
superior iliac spine (Davis, Õunpuu, Tyburski, & which of the isometric, CMJ kinematic and CMJ
Gage, 1991). Participants were given the chance to kinetic (independent) variables best explained the
perform a self-selected warm-up and to practice before variation in CMJ height (dependent variable), for-
performing three maximal CMJs using a natural tech- wards stepwise linear regressions were used.
nique of their selection, including arm swing. They Predictor variables included in these three regression
were permitted to rest between trials for as long as models were put forward as “candidate” variables to
Downloaded by [Tulane University] at 02:13 02 September 2014

they felt necessary, with a minimum rest period of 15 s an overall regression model. Scatterplot and Pearson
imposed (Read & Cisar, 2001). Trials were recorded product moment correlation analyses revealed a sig-
using a 17 camera (M2 MCam) Vicon Motion nificant (r = 0.68, P < 0.01) quadratic relationship
Analysis System (OMG Plc, Oxford, UK) operating between CMJ RTD0–50 and CMJ height, and thus
at 480 Hz. Ground reaction forces were measured an exponentiation transformation was performed on
using an AMTI force platform (600 × 400 mm, CMJ RTD0–50, raising each value to the power of
960 Hz). two prior to its inclusion in the linear regression
The CMJs were manually labelled and processed, analyses. The requirement for the inclusion of a
and all data were synchronised in Vicon’s software. parameter in the regression equations was
Two-dimensional position data were used, with the P < 0.05. Similarly, regression models were rejected
assumption of negligible movement in the medio- if coefficient 95% confidence intervals included zero
lateral plane. All joint centre trajectories were filtered or if correlations, tolerance statistics or variance
using a recursive fourth-order low-pass Butterworth inflation factors showed any evidence of multicolli-
filter with a cut-off frequency of 6 Hz determined nearity (Bowerman & O’Connell, 1990; Draper &
based on a residual analysis and qualitative evalua- Smith, 1998; Field, 2013; Menard, 1995; Myers,
tion of the data (Winter, 1990). Unilateral joint cen- 1990). To confirm the normality of the standardised
tre positions were assumed to represent the bilateral residuals in the regression models, Shapiro–Wilk
location, and the errors in jump height and peak tests for normality were performed. The P-values
joint torques caused by this assumption were calcu- ranged from 0.22 to 0.88, indicating no evidence
lated for one participant. These were found to be less against the assumption of normality of the residuals.
than 1%, with the error in mass centre displacement The percentage of variance in the dependent variable
and joint torques remaining small throughout the (CMJ height) explained by the independent variable
movement. (s) in a regression was determined by Wherry’s
Participant-specific segmental inertia parameters (1931) adjusted R2-value. This represents an
were computed from anthropometric measurements attempt to estimate the proportion of variance that
using Yeadon’s (1990) geometric inertia model of the would be explained by the model had it been derived
human body. The average centre of mass height dur- from the population (young physically active males)
ing the approximately 2 s period of stationary stand- from which the sample was taken. To overcome the
ing prior to the jump was defined as zero potential limitation of stepwise regressions relying on
displacement, and thus the CMJ height was deter- a single best model, the explained variance for all
mined as the maximum vertical mass centre displace- possible regressions with the same number of pre-
ment, with the highest jump for each participant used dictor variables as the stepwise solution was deter-
for further analysis. Inverse dynamics was used to mined for comparison. Pearson product moment
obtain body mass normalised peak ankle, knee and correlation was used to establish relationships, with
hip net joint torques and powers, with extension tor- a P-value <0.05 indicating statistical significance.
ques presented as positive. In order to provide meth-
odological consistency with the isometric rate of
Results
torque development and facilitate the investigation
of different time periods during knee extension, The 18 males participating in this study achieved CMJ
CMJ rate of torque development was computed heights of 0.38–0.73 m (mean 0.55 ± 0.09 m). There
 1808 S. McErlain-Naylor et al.

was substantial variation in the isometric parameters

(Table I), with a mean peak isometric knee extension
torque of 3.62 ± 0.68 N m kg−1. Of the CMJ kine-
matic parameters (Table I), the shoulder showed the
largest variation, indicating a technique difference at
this joint. Mean peak powers at the ankle, knee and
hip were 18.00 ± 4.20, 22.02 ± 4.94 and
9.83 ± 3.54 W kg−1, respectively (Table I). The
mean concentric CMJ rate of torque development
was negative for all four 50 ms intervals.
The best individual predictor of CMJ height was
peak power at the knee joint, explaining 44% of the
observed variation (P < 0.01). This increased to
74%, with the addition of one CMJ kinematic para-
meter (take-off shoulder angle) and a further CMJ
kinetic parameter (peak ankle power; Figure 1). Figure 1. Predicted countermovement jump (CMJ) height against
Higher jumps were associated with greater peak actual CMJ height for the overall three-parameter stepwise solu-
Downloaded by [Tulane University] at 02:13 02 September 2014

powers at the knee and ankle and greater shoulder tion (Table II). Seventy four per cent of the variation in CMJ
height is explained by CMJ peak knee power, take-off shoulder
flexion at take-off. The CMJ kinematic regression
angle and CMJ peak ankle power. With a higher percentage of the
showed the shoulder angle at take-off to be the great- variation in CMJ height explained, the closer the data points lie to
est kinematic predictor of jump height (R2 = 0.26, the dashed line y = x (predicted height = actual height).
P < 0.05). Greater shoulder flexion and ankle plan-
tar-flexion at take-off predicted greater jump heights
(together explaining 58% of the variation). Two Discussion
CMJ kinetic parameters (peak knee power and peak The present study has identified the parameters that
ankle power) explained 57% of the variation (Table best explain CMJ height. In particular, 74% of the
II). Increases in these parameters were associated performance variation can be explained using just
with greater CMJ heights. Further analysis showed three parameters: CMJ peak knee power, take-off
that an alternative CMJ kinetic regression model shoulder angle and CMJ peak ankle power. Two
including peak knee torque and peak ankle power CMJ kinematic parameters (take-off shoulder angle
also explained 57% of the variation in jump height. and take-off ankle angle) explained 58% of the jump
The peak isometric knee extension torque alone height variation, whilst two CMJ kinetic parameters
accounted for 18% of the jump height variation (peak knee power or peak knee torque and peak
(P < 0.05; Table II), with insufficient evidence to ankle power) and one isometric parameter (peak
support the addition of any further isometric para- isometric knee extension torque) explained 57%
meters to the regression equation (i.e. P > 0.05). and 18%, respectively.
The correlation between peak isometric knee exten- The inclusion of peak power at the knee and ankle
sion torque and CMJ peak knee power was non- joints in the overall regression model supports pre-
significant (r = 0.267; P = 0.142), with only 7% of vious claims that CMJ performance is positively
the variation in peak knee power explained by peak associated with lower-limb power (Ashley & Weiss,
isometric torque. 1994; Nuzzo et al., 2008; Sheppard et al., 2008;

Table I. Summary of parameters.

Isometric parameters Mean ± SD CMJ kinetic parameter Mean ± SD CMJ kinematic parameter Mean ± SD

Peak torque (N∙m∙kg−1) 3.62 ± 0.68 PT ankle (N∙m∙kg−1) 2.79 ± 0.40 Minimum ankle angle (°) 84 ± 9
RTD0–50 (N∙m∙kg−1∙s−1) 11.67 ± 8.12 PT knee (N∙m∙kg−1) 3.31 ± 0.62 Minimum knee angle (°) 81 ± 16
RTD50–100 (N∙m∙kg−1∙s−1) 18.96 ± 7.92 PT hip (N∙m∙kg−1) 2.20 ± 0.44 Minimum hip angle (°) 75 ± 15
RTD100–150 (N∙m∙kg−1∙s−1) 10.84 ± 5.07 PP ankle (W∙kg−1) 18.00 ± 4.20 Minimum shoulder angle (°) −67 ± 26
PP knee (W∙kg−1) 22.02 ± 4.94 TO ankle angle (°) 137 ± 12
PP hip (W∙kg−1) 9.83 ± 3.54 TO knee angle (°) 174 ± 14
CMJ RTD0–50 (N∙m∙kg−1∙s−1) −0.93 ± 6.07 TO hip angle (°) 172 ± 5
CMJ RTD50–100 (N∙m∙kg−1∙s−1) −4.75 ± 8.12 TO shoulder angle (°) 103 ± 37
CMJ RTD100–150 (N∙m∙kg−1∙s−1) −1.18 ± 8.15
CMJ RTD150–200 (N∙m∙kg−1∙s−1) −0.66 ± 7.14

Note: CMJ: countermovement jump; RTD0–50, RTD50–100, RTD100–150, RTD150–200: rate of torque development from 0 to 50, 50 to 100,
100 to 150, 150 to 200 ms (of concentric knee extension for the CMJ RTD); PT: peak torque; PP: peak power; TO: take-off.
 Determinants of countermovement jump performance 1809

Table II. Regression equations predicting countermovement jump height from computed variables using stepwise linear regression.

95% confidence intervals

Parameter(s) Coefficient Lower bound Upper bound P Percentage explained

Isometric regression
Peak isometric knee extension torque 0.064 0.002 0.127 0.045 18
CMJ kinematic regression
TO shoulder angle 0.0016 0.0007 0.0024 0.001 58
TO ankle angle 0.004 0.002 0.007 0.003
CMJ kinetic regression 1
CMJ peak knee power 0.011 0.005 0.017 0.002 57
CMJ peak ankle power 0.008 0.001 0.016 0.032
CMJ kinetic regression 2
CMJ peak knee torque 0.087 0.036 0.138 0.002 57
CMJ peak ankle power 0.009 0.002 0.017 0.018
Overall regression
CMJ peak knee power 0.010 0.005 0.015 0.001 74
TO shoulder angle 0.001 0.0004 0.002 0.005
CMJ peak ankle power 0.008 0.002 0.014 0.010
Downloaded by [Tulane University] at 02:13 02 September 2014

Note: CMJ: countermovement jump; TO: take-off; CMJ kinetic regression 1: stepwise solution; CMJ kinetic regression 2: alternative
solution. P < 0.05 indicates a significant relationship.

Vanezis & Lees, 2005). The work-energy–power during the jump are likely important determinants of
relationship makes it inevitable that greater joint CMJ height in the current sample of participants,
powers result in more positive work done and so where jumping ability varied greatly (from 0.3 to
greater total body kinetic energy and mass centre 0.73 m). Experienced jumpers would be expected
vertical velocities at take-off. As an indicator of max- to use similar, close to optimal, techniques and
imal capabilities of the knee extensor musculature, a thus muscle capabilities may distinguish between
greater peak isometric knee extension torque enables their performances, as reported by Vanezis and
greater joint torques and powers to be produced Lees (2005).
during the CMJ. However, whilst the inclusion of The lack of significant finding relating to the initial,
peak isometric torque in the isometric regression neurally mediated isometric RTD0–50 is in agreement
furthers the existing evidence for a relationship with Tillin et al. (2013) but not De Ruiter et al.
between strength and CMJ height (Ashley & Weiss, (2006). It seems likely that the countermovement
1994; Sheppard et al., 2008; Wisløff et al., 2004), phase of the jump diminishes the importance of fast
CMJ peak knee power explained a much greater neural activation by enabling the development of an
proportion of the performance variation (44% versus active state prior to the onset of concentric muscle
18%). Similarly, Young, Wilson, and Byrne (1999) contraction and thus increasing the time available to
showed that CMJ height is more closely related to activate the musculature and to generate extension
measures of speed-strength qualities than maximum joint torques (Bobbert & Casius, 2005; Bobbert
strength. Indeed the low R2 (0.07) and lack of sig- et al., 1996). In the very early stages of knee exten-
nificant correlation between body mass normalised sion, whilst the total length of the knee extensor mus-
peak isometric torque and CMJ peak knee power culo-tendon units decreases, the contractile elements
parameters suggest it is not maximal muscle strength may still be being stretched as the tendon begins its
that causes the strong relationships between CMJ elastic recoil and so knee extension begins with large
kinetic variables and jump height. Given that the eccentric muscle forces (Alexander, 1995). All of
isometric parameters explained only 18% of the var- these factors reduce the importance of fast initial
iation and 58% can be explained by CMJ kinematic rate of torque development during CMJs.
parameters, it seems likely that technique (kinematic The association between later (post-50 ms) rate of
parameters) determines the extent to which the max- torque development and CMJ height is dependent on
imal muscle capabilities (isometric parameters) can absolute maximal force (Tillin et al., 2013), and so
be utilised during the jump (to produce the CMJ with peak isometric knee extension torque already
kinetic parameters). Indeed, Bobbert and Van included in the stepwise regressions, the inclusion of
Soest (1994) showed that an increase in muscle the later isometric rates of torque development did not
strength only improves CMJ performance if techni- significantly improve the prediction of jump height.
que is adapted to the specific muscle capabilities. Previous studies have used correlation coefficients
Thus, both the technique used and the joint kinetics rather than stepwise regressions to assess the rate of
 1810 S. McErlain-Naylor et al.

torque development–jump height relationship and so plantar-flexion at take-off increase the “stretch
were not affected by this issue (De Ruiter et al., 2006; height” and thus pre-take-off displacement, and
Marcora & Miller, 2000; McLellan et al., 2011; both were included in the CMJ kinematic regression.
Thompson et al., 2013; Tillin et al., 2013). These Because CMJ height was calculated relative to stand-
assertions are further supported by a significant corre- ing position, pre-take-off displacement was included,
lation between peak isometric torque and RTD100–150 and thus jump height may be affected by anthropo-
(r = 0.546; P = 0.01) in the present study. metric variables such as foot length. However, the
The significant quadratic relationship between degree to which any anthropometric advantage is
CMJ RTD0–50 and CMJ height (r = 0.68, P < 0.01) reflected in the stretch height is dependent on tech-
was explained by a significant (r = −0.48, P < 0.05) nique such as shoulder flexion and ankle plantar-flex-
negative correlation between CMJ RTD0–50 (variable ion. An analysis of individual participant data suggests
x) and the knee extension torque at concentric onset that the degree of ankle plantar-flexion and shoulder
(variable y). This relationship takes the form flexion during the propulsion phase distinguishes the
x + ay ≈ constant (i.e. there is a trade-off between highest two jumpers from the rest of the participants
the two variables), with (x2 + y2) positively correlating and explains the underestimation of their jump
to CMJ height (r = 0.69, adjusted R2 = 0.45, heights by the CMJ kinetic and isometric parameter
P < 0.01). This (x2 + y2) relationship means that the regression models. Exclusion of these participants
would increase the adjusted R2 for these two regres-
Downloaded by [Tulane University] at 02:13 02 September 2014

higher jumpers either produced high eccentric knee

extension torques (greater y, resulting in an apparent sions to 0.66 and 0.40, respectively, illustrating the
benefit of negative rates of torque development as importance of recruiting a heterogeneous sample so
torque subsequently decreased during knee exten- as not to overestimate the importance of individual
sion) or were able to maintain their knee extension factors in the progression from poor to good counter-
torque during early concentric contraction (greater x), movement jumping.
with those participants in the mid-range for both One limitation of the present study is the introduc-
variables (neither high x nor high y) achieving the tion of errors by any movement outside of the sagittal
lowest jump heights. plane, although this is expected to have been negligible.
Despite the discussed benefits of the counter- Furthermore, isometric knee extensions were mea-
movement phase, the minimum knee and ankle sured at five discrete joint angles, and so the true peak
angles were not included in the stepwise solutions. isometric torque is likely at an intermediate angle. In a
This is in contrast with previous findings that review by Jakobi and Chilibeck (2001), five out of eight
increased knee and ankle joint ranges of motion studies showed no bilateral deficit in isometric knee
result in greater jump heights (Georgios et al., extension. The effect is present, however, in explosive
2006; Moran & Wallace, 2007). Further analysis of voluntary contractions such as the isometric rate of
individual participant data in the present study torque development trials in this study (Buckthorpe,
showed that the highest jumper was the participant Pain, & Folland, 2013). The potential implications of
with the greatest knee flexion. In theory, there is no this deficit in the present study are minor, with the
limit to the relationship between increased squat application of unilateral measures to the investigation
depth and increased squat jump height (Domire & of a bilateral performance task unlikely to distort the
Challis, 2007); however, jumps from a deep squat observed relationships. The 74% of CMJ height varia-
are rarely optimally coordinated due to a lack of tion explained by the overall three-parameter regres-
practice with this technique. This same issue is likely sion suggests that the important aspects of performance
present in inexperienced countermovement jumpers have been identified. In particular, those wishing to
and may explain why the link between minimum improve their CMJ height should seek to maximise
knee angle and CMJ height was only observed in power at the knee and ankle joints and utilise greater
the best-performing participant. ankle plantar-flexion and shoulder flexion. These
Previous studies have simply compared jumps with results are likely to provide a valuable framework
and without an arm swing (Feltner et al., 1999; upon which to base coaching and conditioning as ath-
Harman et al., 1990; Payne et al., 1968; Shetty & letes progress from poor to good countermovement
Etnyre, 1989; Vanezis & Lees, 2005), whereas the jumping. Future studies should continue to explore
present study investigated shoulder angles at key the interaction between kinetic and kinematic factors,
points in the arm swing movement. Greater shoulder including joint ranges of motion and the timings of
flexion at take-off was a strong predictor of CMJ muscle activations, possibly using methods beyond
height, likely indicating greater use of the arm swing, the scale of the current study such as simulation mod-
and thus a slowing of hip extension leading to greater elling or electromyography. It is also important to
work done at the hip as well as the shoulder (Blache & address whether these results are independent of
Monteil, 2013; Cheng et al., 2008; Domire & Challis, anthropometric differences and whether the same
2010). Both greater shoulder flexion and ankle results are observed in a female population.
 Determinants of countermovement jump performance 1811

In conclusion, the purpose of the study was to De Ruiter, C. J., Van Leeuwen, D., Heijblom, A., Bobbert, M. F.,
quantify the relative contributions of kinetic and & De Haan, A. (2006). Fast unilateral isometric knee extension
torque development and bilateral jump height. Medicine and
kinematic parameters in order to identify the most Science in Sports & Exercise, 38, 1843–1852.
important determinants of CMJ performance. The Domire, Z. J., & Challis, J. H. (2007). The influence of squat
findings suggest that both kinetic and kinematic fac- depth on maximal vertical jump performance. Journal of Sports
tors during the jump are important determinants of Sciences, 25, 193–200.
CMJ performance, with technique influencing the Domire, Z. J., & Challis, J. H. (2010). An induced energy analysis to
determine the mechanism for performance enhancement as a
extent to which maximal muscle capabilities can be result of arm swing during jumping. Sports Biomechanics, 9, 38–46.
utilised during the jump. The study has revealed the Draper, N. R., & Smith, H. (1998). Applied regression analysis (3rd
importance of lower-limb joint powers and pre- ed.). New York, NY: Wiley.
viously underestimated, coachable technique factors Edman, K. A. P., Elzinga, G., & Noble, M. I. M. (1978).
including greater ankle plantar-flexion during the Enhancement of mechanical performance by stretch during
tetanic contractions of vertebrate skeletal muscle fibres.
jump and shoulder flexion during the arm swing. Journal of Physiology, 281, 139–155.
Both the kinetic and kinematic variables during the Feltner, M. E., Fraschetti, D. J., & Crisp, R. J. (1999). Upper
jump explained a large proportion of the perfor- extremity augmentation of lower extremity kinetics during
mance variation, and further research is needed to countermovement vertical jumps. Journal of Sports Sciences,
fully understand the interactions between these two 17, 449–466.
Field, A. P. (2013). Discovering statistics using IBM SPSS statistics
Downloaded by [Tulane University] at 02:13 02 September 2014

sets of factors. (4th ed.). London: Sage.

Fukashiro, S., & Komi, P. V. (1987). Joint moment and mechan-
ical power flow of the lower limb during vertical jump.
References International Journal of Sports Medicine, 8, S15–S21.
Georgios, P., Fotis, K., Thomas, N., Vassilios, P., & Iraklis, K.
Alexander, R. M. (1995). Leg design and jumping technique for (2006). Influence of the ankle joint dorsiflexion on the execu-
humans, other vertebrates and insects. Philosophical Transactions tion of vertical jumps. In ISBS-Conference Proceedings Archive
of the Royal Society B: Biological Sciences, 347, 235–248. (Vol. 1, pp. 448–451), Salzburg.
Andersen, L. L., & Aagaard, P. (2006). Influence of maximal Hara, M., Shibayama, A., Takeshita, D., & Fukashiro, S. (2006).
muscle strength and intrinsic muscle contractile properties on The effect of arm swing on lower extremities in vertical jump-
contractile rate of force development. European Journal of ing. Journal of Biomechanics, 39, 2503–2511.
Applied Physiology, 96, 46–52. Harman, E. A., Rosenstein, M. T., Frykman, P. N., &
Ashley, C. D., & Weiss, L. W. (1994). Vertical jump performance Rosenstein, R. M. (1990). The effects of arms and counter-
and selected physiological characteristics of women. Journal of movement on vertical jumping. Medicine and Science in Sports
Strength & Conditioning Research, 8, 5–11. and Exercise, 22, 825–833.
Blache, Y., & Monteil, K. (2013). Effect of arm swing on effective Hubley, C. L., & Wells, R. P. (1983). A work-energy approach to
energy during vertical jumping: Experimental and simulation determine individual joint contributions to vertical jump per-
study. Scandinavian Journal of Medicine & Science in Sports, 23, formance. European Journal of Applied Physiology and
e121–e129. Occupational Physiology, 50, 247–254.
Bobbert, M. F., & Casius, L. J. R. (2005). Is the effect of a Jakobi, J. M., & Chilibeck, P. D. (2001). Bilateral and unilateral
countermovement on jump height due to active state develop- contractions: Possible differences in maximal voluntary force.
ment? Medicine and Science in Sports and Exercise, 37, 440–446. Canadian Journal of Applied Physiology, 26, 12–33.
Bobbert, M. F., Casius, L. J. R., Sijpkens, I. W. T., & Jaspers, R. T. Lees, A., Vanrenterghem, J., & De Clercq, D. (2004).
(2008). Humans adjust control to initial squat depth in vertical Understanding how an arm swing enhances performance in
squat jumping. Journal of Applied Physiology, 105, 1428–1440. the vertical jump. Journal of Biomechanics, 37, 1929–1940.
Bobbert, M. F., Gerritsen, K. G. M., Litjens, M. C. A., & Van Luhtanen, P., & Komi, P. V. (1978). Segmental contribution to
Soest, A. J. (1996). Why is countermovement jump height forces in vertical jump. European Journal of Applied Physiology
greater than squat jump height? Medicine and Science in Sports and Occupational Physiology, 38, 181–188.
and Exercise, 28, 1402–1412. Marcora, S., & Miller, M. K. (2000). The effect of knee angle on
Bobbert, M. F., & Van Soest, A. J. (1994). Effects of muscle the external validity of isometric measures of lower body neu-
strengthening on vertical jump height: A simulation study. romuscular function. Journal of Sports Sciences, 18, 313–319.
Medicine and Science in Sports and Exercise, 26, 1012–1020. McLellan, C. P., Lovell, D. I., & Gass, G. C. (2011). The role of
Bowerman, B. L., & O’Connell, R. T. (1990). Linear statistical rate of force development on vertical jump performance.
models: An applied approach (2nd ed.). Belmont, CA: Duxbury. Journal of Strength and Conditioning Research, 25, 379–385.
Buckthorpe, M. W., Pain, M. T. G., & Folland, J. P. (2013). Menard, S. (1995). Applied logistic regression analysis: Sage
Bilateral deficit in explosive force production is not caused by University series on quantitative applications in the social sciences.
changes in agonist neural drive. Plos ONE, 8, e57549. Thousand Oaks, CA: Sage.
Cheng, K. B., Wang, C., Chen, H., Wu, C., & Chiu, H. (2008). Moran, K. A., & Wallace, E. S. (2007). Eccentric loading and
The mechanisms that enable arm motion to enhance vertical range of knee joint motion effects on performance enhancement
jump performance-a simulation study. Journal of Biomechanics, in vertical jumping. Human Movement Science, 26, 824–840.
41, 1847–1854. Myers, R. (1990). Classical and modern regression with applications
Clutch, D., Wilton, M., McGown, C., & Bryce, G. R. (1983). The (2nd ed.). Boston, MA: Duxbury.
effect of depth jumps and weight training on leg strength and Nuzzo, J. L., McBride, J. M., Cormie, P., & McCaulley, G. O.
vertical jump. Research Quarterly for Exercise and Sport, 54, 5–10. (2008). Relationship between countermovement jump perfor-
Davis, R. B., Õunpuu, S., Tyburski, D., & Gage, J. R. (1991). A mance and multijoint isometric and dynamic tests of
gait analysis data collection and reduction technique. Human strength. Journal of Strength & Conditioning Research, 22,
Movement Science, 10, 575–587. 699–707.
 1812 S. McErlain-Naylor et al.

Payne, A. H., Slater, W. J., & Telford, T. (1968). The use of a Tillin, N. A., Jimenez-Reyes, P., Pain, M. T. G., & Folland, J. P.
force platform in the study of athletic activities. A preliminary (2010). Neuromuscular performance of explosive power ath-
investigation. Ergonomics, 11, 123–143. letes versus untrained individuals. Medicine & Science in Sports &
Ranson, C., King, M., Burnett, A., Worthington, P., & Shine, K. Exercise, 42, 781–790.
(2009). The effect of coaching intervention on elite fast bowl- Tillin, N. A., Pain, M. T. G., & Folland, J. P. (2012). Short-term
ing technique over a two year period. Sports Biomechanics, 8, training for explosive strength causes neural and mechanical
261–274. adaptations. Experimental Physiology, 97, 630–641.
Read, M. M., & Cisar, C. (2001). The influence of varied rest Tillin, N. A., Pain, M. T. G., & Folland, J. (2013). Explosive
interval lengths on depth jump performance. Journal of Strength force production during isometric squats correlates with ath-
and Conditioning Research, 15, 279–283. letic performance in rugby union players. Journal of Sports
Sahaly, R., Vandewalle, H., Driss, T., & Monod, H. (2001). Sciences, 31, 66–76.
Maximal voluntary force and rate of force development in Vanezis, A., & Lees, A. (2005). A biomechanical analysis of good
humans – Importance of instruction. European Journal of and poor performers of the vertical jump. Ergonomics, 48,
Applied Physiology, 85, 345–350. 1594–1603.
Sheppard, J. M., Cronin, J. B., Gabbett, T. J., McGuigan, M. R., Wherry, R. J. (1931). A new formula for predicting the shrinkage
Etxebarria, N., & Newton, R. U. (2008). Relative importance of the coefficient of multiple correlation coefficient. The Annals
of strength, power, and anthropometric measures to jump per- of Mathematical Statistics, 2, 440–457.
formance of elite volleyball players. Journal of Strength & Winter, D. A. (1990). Biomechanics and motor control of human
Conditioning Research, 22, 758–765. movement. New York, NY: Wiley.
Shetty, A. B., & Etnyre, B. R. (1989). Contribution of arm move- Wisløff, U., Castagna, C., Helgerud, J., Jones, R., & Hoff, J.
ment to the force components of a maximum vertical jump. (2004). Strong correlation of maximal squat strength with
Downloaded by [Tulane University] at 02:13 02 September 2014

The Journal of Orthopaedic and Sports Physical Therapy, 11, 198– sprint performance and vertical jump height in elite soccer
201. players. British Journal of Sports Medicine, 38, 285–288.
Thompson, B. J., Ryan, E. D., Sobolewski, E. J., Smith, D. B., Yeadon, M. R. (1990). The simulation of aerial movement – II. A
Akehi, K., Conchola, E. C., & Buckminster, T. (2013). mathematical inertia model of the human body. Journal of
Relationships between rapid isometric torque characteristics Biomechanics, 23, 67–74.
and vertical jump performance in division I collegiate Young, W., Wilson, G., & Byrne, C. (1999). Relationship
American football players: Influence of body mass normalisa- between strength qualities and performance in standing and
tion. Journal of Strength and Conditioning Research, 27, run-up vertical jumps. The Journal of Sports Medicine and
2737–2742. Physical Fitness, 39, 285–293.