See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/258522939

Force-Velocity Profile: Imbalance Determination and Effect on Lower Limb

Ballistic Performance

Article in International Journal of Sports Medicine · November 2013

DOI: 10.1055/s-0033-1354382 · Source: PubMed

CITATIONS READS

147 11,337

6 authors, including:

Pierre Samozino Pascal Edouard

Université Savoie Mont Blanc Université Jean Monnet
223 PUBLICATIONS 5,937 CITATIONS 259 PUBLICATIONS 4,127 CITATIONS

SEE PROFILE SEE PROFILE

Sebastien Sangnier Matt Brughelli

Association sportive saint etienne Auckland University of Technology
6 PUBLICATIONS 276 CITATIONS 103 PUBLICATIONS 4,932 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Maturation and training adaptations View project

Fatigue View project

All content following this page was uploaded by Pierre Samozino on 19 May 2014.

The user has requested enhancement of the downloaded file.

Personal pdf file for
P. Samozino, P. Edouard, S. Sangnier, M. Brughelli,
P. Gimenez, J.-B. Morin

With compliments of Georg Thieme Verlag www.thieme.de

Force-Velocity Profile:
Imbalance Determination and
Effect on Lower Limb Ballistic
Performance

DOI 10.1055/s-0033-1354382
Int J Sports Med

For personal use only.

No commercial use, no depositing in repositories.

Publisher and Copyright

© 2013 by
Georg Thieme Verlag KG
Rüdigerstraße 14
70469 Stuttgart
ISSN 0172-4622

Reprint with the

permission by
the publisher only
 Training & Testing

Force-Velocity Profile: Imbalance Determination and

Effect on Lower Limb Ballistic Performance

Authors P. Samozino1, P. Edouard2, 3, S. Sangnier4, 5, M. Brughelli6, P. Gimenez3, J.-B. Morin3

Affiliations Affiliation addresses are listed at the end of the article

Key words Abstract ment quality (r² = 0.931, P < 0.001, SEE = 0.015 m),
●
▶ jump
▼ significant contributions of Pmax, F-v imbalance
●
▶ muscle mechanical
This study sought to lend experimental support and lower limb extension range (hPO) to explain
properties
to the theoretical influence of force-velocity (F-v) interindividual differences in jumping perform-
●
▶ maximal power output

●
▶ optimal force-velocity profile mechanical profile on jumping performance ance (P < 0.001) with positive regression coef-
●
▶ explosive push-off independently from the effect of maximal power ficients for Pmax and hPO and a negative one for
●
▶ strength training output (Pmax). 48 high-level athletes (soccer play- F-v imbalance. This experimentally supports that
ers, sprinters, rugby players) performed maximal ballistic performance depends, in addition to
squat jumps with additional loads from 0 to 100 % Pmax, on the F–v profile of lower limbs. This adds
of body mass. During each jump, mean force, support to the actual existence of an individual
velocity and power output were obtained using a optimal F-v profile that maximizes jumping per-
simple computation method based on flight time, formance, a F-v imbalance being associated to a
and then used to determine individual linear F-v lower performance. These results have potential
relationships and Pmax values. Actual and optimal strong applications in the field of strength and
F-v profiles were computed for each subject to conditioning.
quantify mechanical F-v imbalance. A multiple
regression analysis showed, with a high-adjust-

Introduction curvilinear (inverse U shape) with an apex cor-

▼ responding to an optimal F-v profile maximizing
Ballistic performances are a key factor in numer- performance. For each individual, an optimal F-v
ous sport activities and can be defined as the profile can then be accurately determined and
accepted after revision
ability to accelerate a mass as much as possible in represents the best balance between its force and
August 02, 2013
the shortest time possible, be it one’s own body velocity capabilities. For a given Pmax, an unfavo-
Bibliography mass (e. g. sprints or jumps) or an external mass rable balance between force and velocity quali-
DOI http://dx.doi.org/ (e. g. throws). Success in such performances has ties can lead to a ~30 % lower performance [30].
10.1055/s-0033-1354382 been closely related to the maximal power out- Thus, considering both Pmax and the individual
Published online: 2013 put (Pmax) limbs can develop [11, 32, 33]. Instan- F-v profile compared to the optimal one might be
Int J Sports Med taneous power output is the product of the of great interest for exploring ballistic perform-
© Georg Thieme
external force developed by velocity. A recently ances and optimizing athletic training.
Verlag KG Stuttgart · New York
validated theoretical integrative approach of The aforementioned theoretical approach took
ISSN 0172-4622
jumping showed that ballistic performance is account of both movement dynamics of the body
Correspondence also influenced by the force-velocity (F-v) center of mass during jumping and external
Dr. Pierre Samozino mechanical profile (henceforth referred to as “F-v mechanical limits of the entire lower limb neu-
Laboratory of Exercise profile”) of the lower limb neuromuscular sys- romuscular system [29, 30]. The latter was con-
Physiology (EA4338) tem (henceforth referred to as “lower limbs”), sidered as a complete force generator system
Université de Savoie independently from Pmax [30]. This F-v profile, characterized by Pmax, F-v profile and range of
UFR CISM – Technolac
normalized to body mass, represents the ratio motion values. This approach was based on equa-
Le Bourget du Lac 73376
between the external force developed and veloc- tions recently validated in comparison to experi-
France
Tel.: + 33/4/79 75 81 77 ity maximal capabilities, and can be determined mental measurements [30]. However, no
Fax: + 33/4/79 75 81 48 by the slope of the F-v relationship [30]. The rela- experimental data has thus far supported the
pierre.samozino@univ-savoie.fr tionship between performance and F-v profile is theoretical influence of F-v profile on ballistic

Samozino P et al. Force-Velocity Profile: Imbalance Determination … Int J Sports Med

 Training & Testing

performance and the existence of an individual optimal F-v pro- ⎛ h ⎞ 1

Fabs = mg ⎜ + 1⎟
file value, mainly due to (i) the curvilinear relationship between ⎝ hPO ⎠
F-v profile and performance preventing direct correlation analy-
sis and (ii) the difficulty in experimentally isolating the effect of gh 2
v=
the F-v profile independently from the large effect of Pmax. 2
Indeed, these 2 variables, though theoretically independent,
often change together similarly in studies testing athletes with with m being the total mass (sum of body mass + additional load,
different training backgrounds [32, 33] or different training pro- in kg), g the gravitational acceleration (9.81 m.s − 2), hPO the verti-
tocols [6, 7, 10, 15, 22]. The aim of this study was to bring experi- cal distance covered by the CM during push-off corresponding to
mental support to the theoretical influence of F-v profile on the extension range of lower limbs (in m) and h the jump height
jumping performance independently from the effect of Pmax. In (in m). Here, hPO was determined as the difference between the
addition to the previously demonstrated positive relation with extended lower limb length with maximal foot plantar flexion
Pmax, it was hypothesized that jumping performance would be (great trochanter-toe distance) and the vertical distance
negatively related to the magnitude of the F-v imbalance. between the great trochanter and ground in the starting posi-
tion. h was obtained from flight time applying fundamental laws
of dynamics [3], flight time being measured using an infrared
Method timing system (Optojump, Microgate, Bolzano, Italy). For the fol-
▼ lowing analyses, force values were normalized to the subject’s
Population body mass (i. e., the moving mass during unloaded jumping
48 international and national-level athletes (31 soccer players, movements), since jumping performance has been directly
11 sprinters and 6 rugby players) participated in the study (age related to normalized force ( F in N.kg − 1, [29]). From F and
(mean ± SD) 20.9 ± 4.4 years; body mass 75.8 ± 12.0 kg; height v values, individual linear F-v relationships were determined
1.79 ± 0.06 m). All subjects were free of musculoskeletal pain or by least-squares linear regressions [5, 27, 33] to obtain for each
injury during the period of the study and gave their written subject (i) the F-v profile normalized to body mass (SFv, slope of
informed consent to participate in this study after being the F-v curve, in N.s.kg − 1.m − 1) and (ii) Pmax (in W.kg − 1) com-
informed about the procedures, which complied with the Decla- puted here by:
ration of Helsinki. This study met the ethical standards of the
International Journal of Sports Medicine [16]. F0 .v0 3
Pmax =
4
Procedure and data analysis
Each subject performed one session of maximal squat jump tests with F0 and v0 being the theoretical maximal force the lower
to determine their Pmax, F-v profile and F-v imbalance. For each limbs can produce at null velocity (extrapolated intercept with
subject, the vertical distance between the ground and the right force-axis, in N.kg − 1) and the theoretical maximal velocity at
leg great trochanter was measured in a 90 °-knee angle crouch which lower limbs can extend under zero load (extrapolated
position [28]. After a 10-min warm-up, each subject performed intercept with velocity-axis, in m.s − 1), respectively [30, 32]. Note
maximal squat jumps in 5 randomized loading conditions: addi- that Pmax must also be expressed here relative to the body mass
tional loads of 0, 25, 50, 75 and 100 % of their body mass. Sub- in unloaded condition (in W.kg − 1, [30]), and not relative to body
jects stood still holding a barbell across their shoulders for mass exponent 2/3 as suggested by scaling models and experi-
additional-loads conditions or with arms crossed on the torso mental data when the aim is to scale muscle power to the body
for the unloaded condition. They then initiated the squat jump size [23]. From Pmax and hPO values, the theoretical optimal SFv
with a downward movement to reach their individual 90 °-knee (SFvopt normalized to body mass, in N.s.kg − 1.m − 1) maximizing
angle starting crouch position (measured beforehand as the ver- jumping performance were computed for each subject using
tical distance between the ground and the right leg great tro- equations proposed by Samozino et al. (appendix section in
chanter, and checked by the experimenter using a vertical rule [30]). The F-v imbalance (FvIMB, in %) were then individually
before trials [28]). After maintaining this position for at least 1 s, computed as:
they were asked to apply force as rapidly as possible and to jump
S Fv 4
for maximum height. Countermovement was verbally restricted FvIMB = 100. 1 −
S Fv opt
and carefully checked. Subjects were instructed to keep constant
downward pressure on the barbell throughout the jump with
the chest upright, and asked to touch down the ground in the The hypothetic maximal jump each subject could reach if he
same leg position as they took off, i. e., extended leg in foot presented an optimal F-v profile (hmax, in m) was computed from
plantar flexion. If these requirements were not met, the trial was his actual Pmax and hPO values, and using the following equation
repeated. Subjects had to perform 2 correct trials by condition, derived from previously validated equations [29, 30]:
only the best one being used in the analysis. Before this session,
subjects were allowed enough supervised practice to correctly 2
hPO 2 ⎛ S Fv opt 2 2 S opt ⎞ 5
perform squat jumps without countermovement. hmax = ⎜ + (2 − Pmax S Fv opt − g) + Fv ⎟
During each trial, the mean vertical force developed by the lower 2g ⎝ 4 hPO 2 ⎠
limbs during push-off ( Fabs , absolute force in N) and the corre-
sponding mean center of mass (CM) vertical velocity ( v , in The theoretical loss of performance due to the F-v imbalance
m.s − 1) were determined using the equations and the methodol- was computed for each subject as the percent difference between
ogy proposed and validated by Samozino et al. [28]: actual unloaded jump height and hmax (in percentage of hmax).

Samozino P et al. Force-Velocity Profile: Imbalance Determination … Int J Sports Med

 Training & Testing

ple regression analysis was performed with maximal unloaded

30 jump height as the dependent variable, and FvIMB, Pmax (normal-
28 ized to body mass) and hPO as independent variables. This multi-
ple regression model was based on the previously proposed
26 theoretical approach [30]. The effect of F-v profile was tested in
y= –15.0 x +38.2 regression analysis through FvIMB since the relationship between
24
Force (N.kg–1)

F-v profile and performance is curvilinear, while that between

22 performance and FvIMB should be simply decreasing. The crite-
rion for statistical significance was set at α = 0.05.
20
y= –8.99 x +29.6
18
r2 = 0.98
Results
16 FvIMB = 40.0 %
▼
14 Individual F-v relationships were well fitted by linear regressions
(r² = 0.87–1.00; P ≤ 0.05). ●
▶ Fig. 1 presents an F-v curve for a typ-
12
0.2 0.6 1.0 1.4 1.8 2.2 ical subject, compared to the theoretical F-v curve characteriz-
Velocity (m.s–1) ing his optimal F-v profile that would maximize his jumping
performance without changing Pmax. Mean values ( ± SD) of max-
Optimal F-v profile Actual F-v profile
imal unloaded jump height, hPO, Pmax, SFv, SFvopt and FvIMB
were 0.38 ± 0.06 m, 0.36 ± 0.05 m, 29.7 ± 6.13 W.kg − 1, − 11.6 ±
Fig. 1 Typical linear normalized force-velocity relationship (solid line)
7.36 N.s.kg − 1.m − 1, − 15.0 ± 1.30 N.s.kg − 1.m − 1 and 42.6 ± 34.4 %,
and the associated mechanical F-v profile (dotted line). Each circle repre-
respectively. When considering simple correlation analyses,
sents a squat jump at a given additional load. The individual optimal F-v
profile maximizing this subject’s jumping performance with no change in
jumping performance was significantly correlated to Pmax and
Pmax is represented by the dashed-line. hPO, but not to SFv and FvIMB (● ▶ Table 1). No simple correlation

was found between the 3 independent variables (Pmax, FvIMB

and hPO). The multiple regression analysis indicated that, when
Table 1 Correlation matrix for all variables (n = 47, one outlier was removed). considered together, the 3 predictor variables accounted for a
significant amount of jumping performance variability
Pmax SFv FvIMB hPO (P < 0.001, ● ▶ Table 2). Furthermore, the regression model

unloaded h 0.782 − 0.247 0.066 0.303 showed a high quality of adjustment (r² = 0.931, P < 0.001) and
P < 0.001 P = 0.094 P = 0.659 P = 0.042 minimal error (SEE = 0.015 m) (● ▶ Table 2, ● ▶ Fig. 2). It should be

Pmax − 0.206 0.218 − 0.254 noted that, when only Pmax and hPO were considered as inde-
P = 0.164 P = 0.141 P = 0.085 pendent variables, the regression model also showed a good
SFv 0.0423 0.0971 quality of adjustment (r² = 0.875, SEE = 0.020, P < 0.001), but
P = 0.778 P = 0.516
lower than that when FvIMB was considered. The loss of perform-
FvIMB 0.170
ance due to individual F-v imbalance was 6.49 ± 6.25 %. ● ▶ Fig. 3
P = 0.252
presents the actual jump height reached by each subject (h
h: jump height; Pmax: maximal power output; SFv: Force-velocity mechanical profile;
FvIMB: Force-velocity imbalance; hPO: lower limb extension range
expressed relatively to hmax) according to their F-v profile (SFv
expressed relative to their personal SFvopt), as well as the corre-
sponding theoretical changes predicted by the model (equation
Table 2 Multiple regression analysis for the prediction of jump height 5 with hPO and Pmax values arbitrarily set to the average values for
(n = 47, one outlier was removed). the entire group).
Multiple regression model r² SEE (m) P
0.931 0.015 < 0.001
Discussion
Independent variables Coefficient t P
▼
Pmax 0.0122 23.0 < 0.001 The main finding of this study is the significant contribution of
FvIMB − 0.0810 − 5.92 < 0.001 mechanical F-v imbalance to explain jumping performance vari-
hPO 0.617 13.9 < 0.001
ability. This constitutes experimental support for (i) the influ-
Constant − 0.174 − 7.13 < 0.001
ence of the normalized F-v profile (characterized by the slope of
SEE: Standard Error of Estimate; Pmax: maximal power output; FvIMB: Force-velocity
the F-v relationship, SFv) on jumping performance independ-
imbalance; hPO: lower limb extension range
ently from the large effect of Pmax, and (ii) the existence of an
optimal F-v profile maximizing performance for each individual.
Statistical analyses These results lend experimental support to the previously pro-
All data are presented as mean ± SD. Normality and homogeneity posed theoretical approach [30].
of variance were checked before analyses. One highly strength- Simple correlations showed that jumping performance was
trained subject had jump height, Pmax and FvIMB values more highly associated with lower limb maximal power output, as
than 3 SD beyond average values for the group, and was there- previously observed with similar magnitude (e. g. [13, 32, 33], r
fore considered as outlier for statistical analyses. The degree of ranging from 0.65 to 0.84), but not directly to SFv or FvIMB. This
linear relationship between variables was examined using Pear- confirmed that jumping performance mainly depends on Pmax
son’s product moment correlation. To test the independent [30]. The absence of experimental correlation between perform-
implication of Pmax and FvIMB on jumping performance, a multi- ance and SFv was expected knowing the theoretical curvilinear

Samozino P et al. Force-Velocity Profile: Imbalance Determination … Int J Sports Med

 Training & Testing

force qualities (until FvIMB of 248 %, ●▶ Fig. 3). A subject present-

0.60 y=x ing a F-v imbalance (as computed here from SFvopt maximizing
0.55 vertical jumps) means that he does not develop his Pmax against
Jump height predicted by model (m)

r=0.965 his body mass during a vertical jump [30]. This could be related
0.50 to the Maximum Dynamic Output hypothesis proposed by Jaric
and Markovic [19], which stated that the neuromuscular system
0.45
is optimized, and in turn develops Pmax, in ballistic movements
0.40 realized against its own body weight and inertia since it is likely
designed to work optimally against loads most usually sup-
0.35 ported and mobilized [19, 24, 25]. However, when body mass is
mobilized horizontally (horizontal jumps or displacements), the
0.30
mechanical constraint to the movement is lower than the load
0.25 represented by body weight during a vertical jump. Conse-
quently, individuals used to running horizontally (as most of the
0.20 present subjects, notably soccer players or sprinters, ● ▶ Fig. 3)
0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60
Jump height measured (m)
can present optimal load-maximizing power during vertical
jumping lower than their own body mass and thus present an
Fig. 2 Correlation between model-predicted and measured jump F-v imbalance towards velocity qualities for vertical jumps. In
heights. Each point represents a subject: white points for soccer players, contrast, rugby players used to and trained to perform displace-
grey points for sprinters and black points for rugby players. The outlier ments against resistive forces and to mobilize high training loads
(the point with a cross) was not considered in the correlation analysis. The presented F-v imbalances towards force qualities, which is the
solid line represents the identity line. case here for 4 out of 6 rugby players (● ▶ Fig. 3). During the

present protocol, the latter developed their Pmax during vertical

jumps against loads higher than their body mass (i. e., for addi-
120
tional loads from 25 to 100 % of their body mass). This would
explain the large range of F-v imbalance observed here and sup-
100
port the recently proposed and debated influence of training
Jump height (% hmax)

80 history on the optimal load-maximizing power during vertical

jumps [19, 21, 24–26]. Further studies are required to explore
60 the exact origins of these large individual differences in F-v
imbalance.
40 The negative effect of the F-v imbalance on performance was
quantified for each subject through the difference between their
20 actual jump height reached during unloaded condition and the
maximal height they would have reached if they had presented
0
0 50 100 150 200 250 300 350 400 the same Pmax with an optimal F-v profile. The individual loss of
SFV (% SFVopt) performance due to the F-v imbalance ranged from 0 for sub-
Velocity profile Force profile jects presenting optimal profile to ~30 % for the outlier (●▶ Fig. 3),

which is in line with the theoretical simulations performed by

Fig. 3 Actual jump height reached in unloaded condition (h expressed Samozino et al. [30]. Furthermore, since FvIMB is computed as the
relatively to hmax) according to F-v profile (SFv expressed relatively to their difference between a subject’s theoretically optimal and actual
personal SFvopt). Each point represents a subject: white points for soccer F-v profiles, the significant implication of F-v imbalance in
players, grey points for sprinters and black points for rugby players (a jumping performance provides experimental support for the
cross shows the outlier). The solid line represents theoretical changes
actual existence of an optimal F-v profile, specific to each indi-
predicted by the model (equation 5 with hPO and Pmax values arbitrarily set
vidual, that maximizes jumping performance for a same given
to the average values for the entire group).
Pmax. This finding is of interest for better understanding the rela-
tionships between lower limb neuromuscular system mechani-
relationship between these 2 variables [30]. Since Pmax and FvIMB cal properties and functional performance, notably for
are independent in theory, which was not contradicted here optimizing sport performance and training considering individ-
experimentally (absence of correlation between these 2 varia- ual F-v profile in addition to Pmax. The F-v profile has been shown
bles), the effect of FvIMB on jumping performance can be biased to be sensitive to specific training and to the additional loads
by the very large effect of Pmax. However, the multiple regression used by athletes during strength training and conditioning:
analysis, testing the effect of each independent variable inde- moving light loads (e. g., < 30 % of 1 repetition maximum) during
pendently from each other, showed that interindividual variabil- maximal effort to orient F-v profile towards velocity qualities
ity in jumping performance is partly explained by FvIMB. For a and moving heavy loads ( > 75–80 % of 1 repetition maximum) to
given Pmax, the higher FvIMB, the lower the jumping performance, improve strength capabilities [7, 20]. Consequently, training
which validates our hypothesis. Subjects presented here FvIMB loads can be selected according to athlete’s individual F-v imbal-
ranging from 2.07 % (quasi perfect balance between force and ance, which would lead to improved performance through both
velocity qualities) to 94.3 % (very high F-v imbalance), and even an increase in Pmax and an optimization of F-v profile.
248 % for the outlier. These F-v imbalances concerned both F-v Simple correlations and multiple regressions showed that jump-
profiles excessively oriented towards velocity capabilities (until ing performance was positively related to hPO. Thus, for a given
FvIMB of 64.4 %) and F-v profiles excessively oriented towards Pmax and F-v profile, individuals with higher hPO (i. e., higher

Samozino P et al. Force-Velocity Profile: Imbalance Determination … Int J Sports Med

 Training & Testing

lower limb range of motion) presented higher performances opment), some morphological factors (e. g. cross sectional area,
since they can develop a higher mechanical work during the fascicle length, pennation angle, tendon properties, anatomical
push-off phase. Even if this finding is not of great interest for joint configuration), neural mechanisms (e. g. motor unit recruit-
training purposes (since hPO represents the personal optimal ment, firing frequency, motor unit synchronization, inter-mus-
lower limb range of motion), it aids in understanding interindi- cular coordination) and segmental dynamics [4, 7–9]. Our
vidual and interspecies differences in jumping performance, as approach is also based on mechanical outputs averaged over the
previously discussed for humans with the influence of the start- entire lower limb extension movement which appeared to be
ing position [1, 12, 31], and for animals through the effect of hind more representative of the muscular effort analyzed than
limb length or joint range of motion [1, 14, 18]. instantaneous values [2]. Finally, it is worth noting that this
It is worth noting that multiple regression analysis determined study focused only on squat jump performance that does not
that 93.1 % of the variation in jumping performance could be involve all mechanical and neuromuscular mechanisms partici-
explained by variation in lower limb maximal power capabili- pating in the more natural counter-movement jumps. Even if
ties, F-v imbalance and lower limb extension range. In addition one could expect to observe similar findings, further studies are
to the very low SEE (1.5 cm), this represents a very high predic- currently undertaken to explore the effect of F-v imbalance on
tion quality of human performance, from only 3 lower limb such kind of jumping performance.
mechanical properties. However, based on the theoretical To conclude, this study lends experimental evidence that ballis-
approach on which the regression analysis was based [30], these tic performance depends, in addition to Pmax, on the normalized
three variables should explain the entire interindividual varia- F-v profile of lower limbs (slope of the F-v relationship) charac-
bility in performance, which was not the case here (r² = 0.931). terizing the ratio between their maximal normalized force and
This difference can be due to measurement inaccuracies or to maximal velocity capabilities. Even if Pmax remains the main
the imperfect reliability of the human performance, as reported determinant, an F-v imbalance is associated with a lower per-
by Hopkins [17], where a standard error of measurement of ~7 % formance. This result constitutes experimental support to the
for explosive movements was found. The validation of the equa- previous theoretical approach expressing ballistic performance
tions used in the theoretical approach reported errors between as a function of lower limbs mechanical properties [29, 30], and
theoretical and experimental values of about 6 % [30]. Moreover, presents numerous direct practical applications, notably in
a multiple regression analysis supposed linear relationships strength training and conditioning. Finally, this study proposes
between dependent and independent variables, which is not an applied approach to accurately determine (in field or labora-
exactly the case here (equation 6 in [30]), and which can con- tory conditions) both the individual maximum power output
tribute to increasing the model error of estimate. Finally, Pmax and the F-v imbalance, which could be developed in a more gen-
was computed using equation 3 from extrapolation of F-v curves eral method for testing the adaptation of mechanical properties
on force and velocity-axis instead of using the apex of the power- of the lower limb neuromuscular system in a variety of maxi-
velocity relationship [27, 33], the latter requiring more than 5 mum performance tasks.
jumping conditions to properly model power-velocity polyno-
mial regressions. Nevertheless, no difference in Pmax between
these two methods has been reported with a bias lower than 2 % Acknowledgements
[30]. It should be noted that Pmax, jump height and hPO values ▼
obtained here are in line with previous studies [5, 27, 28, 30]. The authors thank James de Lacey (PhD student, Auckland Uni-
This study was based on a biomechanical approach aimed at versity of Technology, New Zealand) for assisting with the data
describing the mechanical outputs that result from the action of collection of the rugby players and all the subjects tested for
the entire lower limb neuromuscular system, and not modeling their imbalance and their powerful implication in the protocol.
the complex musculoskeletal structures at the origin of these
Affiliations
outputs. The main limit of this approach could be the macro- 1
Laboratory of Exercise Physiology (EA4338), University of Savoie, Le Bourget
scopic level from which the multi-segmental neuromuscular du Lac, France
2
system is considered, inducing (i) the description of its mechan- Department of Clinical and Exercise Physiology, Sports Medicine Unity,
University-Hospital of Saint-Etienne, France
ical external capabilities by the empirically-determined F-v rela- 3
Laboratory of Exercise Physiology (EA4338), University of Lyon, Saint
tionships, and (ii) the application of principles of dynamics to a Etienne, France
4
whole body viewed as a system (these points have been dis- Centre d’Etude des Transformations des Activités Physiques et Sportives (EA
3832), University of Rouen, France
cussed in detail in [29, 30]). The bias induced by the simplifica- 5
Association Sportive de Saint-Etienne, France
6
tions and approximations associated with this approach were Sports Performance Research Institute New Zealand, AUT University,
shown to be low ( < 6 %) and trivial, which supported its validity Auckland, New Zealand
[30]. This validity is also strengthened, as shown in ● ▶ Fig. 3, by

the high agreement between measured performances and theo-

References
retical changes predicted by the model. The F-v linear model 1 Alexander RM. Leg design and jumping technique for humans, other
used to characterize the dynamic external capabilities of the vertebrates and insects. Philos Trans R Soc Lond B Biol Sci 1995; 347:
lower limb neuromuscular system has been previously well sup- 235–248
2 Andrews GC. Biomechanical measures of muscular effort. Med Sci
ported and discussed during multi-joint lower limb extensions
Sports Exerc 1983; 15: 199–207
[4, 5, 27, 30, 33]. These overall dynamic external capabilities are 3 Asmussen E, Bonde-Petersen F. Storage of elastic energy in skeletal
the final result of the different mechanisms involved in limbs muscles in man. Acta Physiol Scand 1974; 91: 385–392
extension, and characterize the mechanical limits of the entire 4 Bobbert MF. Why is the force-velocity relationship in leg press tasks
quasi-linear rather than hyperbolic? J Appl Physiol 2012, doi:10.1152/
neuromuscular function in vivo. They are a complex integration
japplphysiol.00787.02011
of numerous individual muscle mechanical properties (e. g.
intrinsic F-v and length-tension relationship, rate of force devel-

Samozino P et al. Force-Velocity Profile: Imbalance Determination … Int J Sports Med

 Training & Testing

5 Bosco C, Belli A, Astrua M, Tihanyi J, Pozzo R, Kellis S, Tsarpela O, Foti 20 Kaneko M, Fuchimoto T, Toji H, Suei K. Training effect of different loads
C, Manno R, Tranquilli C. A dynamometer for evaluation of dynamic on the force-velocity relationship and mechanical power output in
muscle work. Eur J Appl Physiol 1995; 70: 379–386 human muscle. Scand J Sports Sci 1983; 5: 50–55
6 Cormie P, McCaulley GO, McBride JM. Power versus strength-power 21 Markovic G, Jaric S. Positive and negative loading and mechanical
jump squat training: influence on the load-power relationship. Med output in maximum vertical jumping. Med Sci Sports Exerc 2007;
Sci Sports Exerc 2007; 39: 996–1003 39: 1757–1764
7 Cormie P, McGuigan MR, Newton RU. Adaptations in athletic perform- 22 McBride JM, Triplett-McBride T, Davie A, Newton RU. The effect of
ance after ballistic power versus strength training. Med Sci Sports heavy- vs. light-load jump squats on the development of strength,
Exerc 2010; 42: 1582–1598 power, and speed. J Strength Cond Res 2002; 16: 75–82
8 Cormie P, McGuigan MR, Newton RU. Influence of strength on magni- 23 Nedeljkovic A, Mirkov DM, Bozic P, Jaric S. Tests of muscle power out-
tude and mechanisms of adaptation to power training. Med Sci Sports put: the role of body size. Int J Sports Med 2009; 30: 100–106
Exerc 2010; 42: 1566–1581 24 Nuzzo JL, McBride JM, Dayne AM, Israetel MA, Dumke CL, Triplett NT.
9 Cormie P, McGuigan MR, Newton RU. Developing maximal neuromus- Testing of the maximal dynamic output hypothesis in trained and
cular power: part 1 – biological basis of maximal power production. untrained subjects. J Strength Cond Res 2011; 24: 1269–1276
Sports Med 2011; 41: 17–38 25 Pazin N, Berjan B, Nedeljkovic A, Markovic G, Jaric S. Power output in
10 Cronin J, McNair PJ, Marshall RN. Velocity specificity, combination vertical jumps: does optimum loading depend on activity profiles?
training and sport specific tasks. J Sci Med Sport 2001; 4: 168–178 Eur J Appl Physiol 2013; 113: 577–589
11 Cronin J, Sleivert G. Challenges in understanding the influence of maxi- 26 Pazin N, Bozic P, Bobana B, Nedeljkovic A, Jaric S. Optimum loading for
mal power training on improving athletic performance. Sports Med maximizing muscle power output: the effect of training history. Eur
2005; 35: 213–234 J Appl Physiol 2011; 111: 2123–2130
12 Domire ZJ, Challis JH. The influence of squat depth on maximal vertical 27 Rahmani A, Viale F, Dalleau G, Lacour JR. Force/velocity and power/
jump performance. J Sports Sci 2007; 25: 193–200 velocity relationships in squat exercise. Eur J Appl Physiol 2001; 84:
13 Driss T, Vandewalle H, Monod H. Maximal power and force-velocity 227–232
relationships during cycling and cranking exercises in volleyball play- 28 Samozino P, Morin JB, Hintzy F, Belli A. A simple method for measur-
ers. Correlation with the vertical jump test. J Sports Med Phys Fitness ing force, velocity and power output during squat jump. J Biomech
1998; 38: 286–293 2008; 41: 2940–2945
14 Harris MA, Steudel K. The relationship between maximum jumping 29 Samozino P, Morin JB, Hintzy F, Belli A. Jumping ability: A theoretical
performance and hind limb morphology/physiology in domestic cats integrative approach. J Theor Biol 2010; 264: 11–18
(Felis silvestris catus). J Exp Biol 2002; 205: 3877–3889 30 Samozino P, Rejc E, Di Prampero PE, Belli A, Morin JB. Optimal force-
15 Harris NK, Cronin JB, Hopkins WG, Hansen KT. Squat jump training velocity profile in ballistic movements. Altius, citius or fortius? Med
at maximal power loads vs. heavy loads: effect on sprint ability. Sci Sports Exerc 2012; 44: 313–322
J Strength Cond Res 2008; 22: 1742–1749 31 Selbie WS, Caldwell GE. A simulation study of vertical jumping from
16 Harriss DJ, Atkinson G. Update – Ethical standards in sport and exercise different starting postures. J Biomech 1996; 29: 1137–1146
science research. Int J Sports Med 2011; 32: 819–821 32 Vandewalle H, Peres G, Heller J, Panel J, Monod H. Force-velocity rela-
17 Hopkins WG, Schabort EJ, Hawley JA. Reliability of power in physical tionship and maximal power on a cycle ergometer. Correlation with
performance tests. Sports Med 2001; 31: 211–234 the height of a vertical jump. Eur J Appl Physiol Occup Physiol 1987;
18 James RS, Wilson RS. Explosive jumping: extreme morphological and 56: 650–656
physiological specializations of Australian rocket frogs (Litoria nas- 33 Yamauchi J, Ishii N. Relations between force-velocity characteristics
uta). Physiol Biochem Zool 2008; 81: 176–185 of the knee-hip extension movement and vertical jump performance.
19 Jaric S, Markovic G. Leg muscles design: the maximum dynamic out- J Strength Cond Res 2007; 21: 703–709
put hypothesis. Med Sci Sports Exerc 2009; 41: 780–787

Samozino P et al. Force-Velocity Profile: Imbalance Determination … Int J Sports Med

View publication stats