JOURNAL OF SPORTS SCIENCES, 2016

http://dx.doi.org/10.1080/02640414.2016.1221521

A simple method for assessment of muscle force, velocity, and power producing
capacities from functional movement tasks
Milena Z. Zivkovica, Sasa Djurica, Ivan Cuka,b, Dejan Suzovica and Slobodan Jaricc,d
a
Faculty of Sport and Physical Education, The Research Centre, University of Belgrade, Belgrade, Serbia; bDepartment of Kinesiology, College of
Sports and Health, Belgrade, Serbia; cDepartment of Kinesiology and Applied Physiology, University of Delaware, Newark, DE, USA; dBiomechanics
and Movement Science Graduate Program, University of Delaware, Newark, DE, USA

ABSTRACT ARTICLE HISTORY

A range of force (F) and velocity (V) data obtained from functional movement tasks (e.g., running, Accepted 2 August 2016
jumping, throwing, lifting, cycling) performed under variety of external loads have typically revealed
KEYWORDS
strong and approximately linear F–V relationships. The regression model parameters reveal the max- Regression; mechanics;
imum F (F-intercept), V (V-intercept), and power (P) producing capacities of the tested muscles. The aim parameter; output; load
of the present study was to evaluate the level of agreement between the routinely used “multiple-load
model” and a simple “two-load model” based on direct assessment of the F–V relationship from only 2
external loads applied. Twelve participants were tested on the maximum performance vertical jumps,
cycling, bench press throws, and bench pull performed against a variety of different loads. All 4 tested
tasks revealed both exceptionally strong relationships between the parameters of the 2 models (median
R = 0.98) and a lack of meaningful differences between their magnitudes (fixed bias below 3.4%).
Therefore, addition of another load to the standard tests of various functional tasks typically conducted
under a single set of mechanical conditions could allow for the assessment of the muscle mechanical
properties such as the muscle F, V, and P producing capacities.

Introduction Chevalier, & Monod, 2002; Nikolaidis, 2012), and running

(Jaskolska, Goossens, Veenstra, Jaskolski, & Skinner, 1999;
Routine procedures for testing muscle function have been
Morin, Samozino, Bonnefoy, Edouard, & Belli, 2010; Rabita
usually conducted under a single mechanical condition. As a
et al., 2015). Note that such tasks reveal only the outputs that
result, the muscle capacities, such as those for producing high
reflect the capacity of the active muscles as a whole to produce
F, V and P outputs, cannot be distinguished from single out-
high external F and V, but not the same capacities of individual
comes of such tests (Jaric, 2015). This inevitably leads to a
muscles (McMahon, 1984; Winter et al., 2016). Nevertheless, the
fundamental problem in the contemporary literature both
applied regression modelling inevitably revealed the maximum
regarding the design of various research and testing proce-
F (i.e., the F-intercept; Fmax), maximum V (V-intercept;
dures, as well as an ambiguity in interpreting their results.
Vmax = Fmax/a), and maximum power (P) [Pmax = (Fmax · Vmax/
A solution of the discussed problem could be based on a
4); (Driss et al., 2002; Jaric, 2015; Samozino et al., 2012)]. In the
number of recent studies that have been focused upon the
further text, this approach will be referred to as the multiple-
force–velocity (F–V) relationship of muscular systems perform-
load model. Of particular importance could be that the multiple-
ing various functional movement tasks [i.e., multi-join move-
load model parameters depicting the external F, V, and P out-
ments represented in either everyday behaviour or various
puts of the tested muscles (i.e., Fmax, Vmax, and Pmax, respec-
sport activities such as jumping, cycling, or lifting; (Jaric,
tively) proved to be highly reliable and at least moderately valid
2015)]. Specifically, a manipulation of external loads provided
with respect to the same variables directly measured through
a range of F and V data that allowed for applying various
standard tests (Jaric, 2015). Therefore, the F–V relationship
regression models revealing V-associated decrease in F.
could provide a comprehensive and valuable set of information
Outcomes of the applied regressions typically revealed excep-
regarding different mechanical capacities of the tested muscles,
tionally strong and linear F–V relationship (i.e., F = Fmax−aV)
particularly when compared with the standard testing proce-
from functional tasks such as lifting (Garcia-Ramos, Jaric, Padial,
dures typically performed under a single set of mechanical
& Feriche, 2016; Sanchez-Medina, Gonzalez-Badillo, Perez, &
conditions (Bobbert, 2012; Jaric, 2015; Rabita et al., 2015;
Pallares, 2014; Sreckovic et al., 2015) jumping and leg push
Samozino et al., 2012, 2014).
offs (Cuk et al., 2014; Feeney, Stanhope, Kaminski, Machi, &
Several authors have already suggested that the method of
Jaric, 2016; Giroux, Rabita, Chollet, & Guilhem, 2015; Meylan
obtaining the linear F–V relationship from loaded functional
et al., 2015; Nikolaidis, 2012; Samozino, Rejc, Di Prampero,
movements procedure could be developed into a routine
Belli, & Morin, 2012, 2014), cycling (Driss, Vandewalle, Le

CONTACT Slobodan Jaric jaric@udel.edu Department of Kinesiology and Applied Physiology, University of Delaware, Rust Arena, Rm. 143, 541 South
College Avenue, Newark, DE 19716, USA
© 2016 Informa UK Limited, trading as Taylor & Francis Group
2 M. Z. ZIVKOVIC ET AL.
procedure for testing the muscle mechanical capacities (Cuk
et al., 2014; Garcia-Ramos et.al., 2016; Jaric, 2015; Meylan et al.,
2015; Nikolaidis, 2012; Sreckovic et al., 2015). However, the
same procedure could be too demanding due to a need for a
number of loading conditions and, therefore, being prone to
fatigue, as well as due to the required regression modelling. A
possible solution of the stated problems could be based on
the 2 findings. First, since the F–V relationship obtained from a
series of loaded functional movements proved to be excep-
tionally strong, the number of the different loads applied
could have only a minor effect upon the shape of the
observed F–V relationship. Second, due to the typically linear
shape of the observed F–V relationships, the magnitude of the
applied loads should also play a minor role. Therefore, a
plausible simplification of the standard procedure of obtaining
F–V relationship referred to as the multiple-load model could
be based on drawing a line through just 2 pairs of F and V
data obtained from 2 distinctive loads (i.e., the two-load
model). Such a two-load model could accurately replicate the
F–V relationship obtained from a multiple-load model applied
on a number of trials performed under different loads. As a
result, the two-load model might discern among the F, V, and
P capacities of the tested muscles in the same way as the
parameters of the typically applied multiple-load model does,
while still avoiding ambiguity of interpretation of routine tests
typically performed under a single mechanical condition.
To explore to what extent the two-load model replicates
the actual F–V relationship obtained from the multiple-load
model, we conducted 4 functional movement tests under
various external loads. We specifically hypothesised that
there would be a high level of agreement between the para-
meters depicting the F, V, and P producing capacity of the
tested muscles (i.e., Fmax, Vmax, and Pmax, respectively)
obtained from the two-load models and the corresponding
standard multiple-load models. If the hypothesised concurrent
validity of the evaluated two-load model can be confirmed,
that could motivate its routine use for a comprehensive
assessment of muscle mechanical capacities from various
functional movements.
Methods
Participants
We employed the sample size estimate (Cohen, 1988) for an
alpha level 0.05 and P 0.8 based on previously observed
effects of similar magnitudes of loading (Cuk et al., 2014;
Sreckovic et al., 2015) assessed by one-way ANOVA. Between
3–12 participants appeared to be necessary to detect the
differences among the F and V outputs obtained from differ-
ent tests and loading conditions. Therefore, we recruited 12
healthy male participants (age 22.1 ± 3.4 years; body height
184.1 ± 7.1 cm; body mass 80.8 ± 8.2 kg, body mass index
24.5 ± 1.5 cm/kg2, and per cent body fat 11.2 ± 2.8 %; data
shown as mean ± SD). IPAQ questionnaire (Taylor-Piliae et al.,
2006) revealed low, moderate, and high level of physical
activity in 3, 5, and 4 participants, respectively. None of them
reported either medical problems or recent injuries. They were
informed regarding the potential risks associated with the
applied testing protocol and also instructed to avoid any
unusual strenuous activities over the course of the study.
Both the experimental protocol and the informed consent
signed by the participants were in accordance with the
Declaration of Helsinki and approved by Institutional Review
Board.
Functional movement tests
Participants were tested on maximum counter-movement
jumps without arm swing (JUMP) on a F plate (AMTI;
Watertown, USA) wearing a weighted vest and belt (MiR Vest
Inc; San Hose, USA). Although loaded bars positioned upon
subjects’ shoulders have been more frequently employed, we
selected the applied loading method because it brings the
load closer to the body centre of mass, reduces loading of the
spine, and less affect the jumping pattern. No specific instruc-
tions were given regarding the depth of the counter move-
ment. Two trials per load were performed and the rest period
between consecutive jumps was 1 min and 2–3 min between
different loading magnitudes (Feeney et al., 2016; Leontijevic
et al., 2012).
A 6-s maximal cycling sprint test (CYCLING) performed on a
Monark 834E leg cycle ergometer (Monark, Varberg, Sweden)
consisted entirely of testing the maximum P output (Jaskolska
et al., 1999). Participants were instructed to perform an “all
out” effort from the very beginning of the test and to remain
seated during the entire sprint. The seat height was individu-
ally adjusted to individual satisfaction, and toe clips with
straps were used to prevent the feet slipping off the pedals
(Driss et al., 2002). One trial per each resistance was performed
and the rest period between consecutive sprints was 4 min.
Maximum bench press throws test (BPRESS) was performed
on a Smith machine [see (Sreckovic et al., 2015) for details]
with the instruction “to throw the bar as high as possible.”
Two stoppers were standing aside and catching the bar during
its descending trajectory. For safety reasons, mechanical stops
not only kept the bar about 1 cm above the chest prior to the
test, but also prevented (if needed) the collision of the bar
with the subject’s chest. To avoid fatigue, the rest periods
between 2 consecutive trials were 45 s, while the rest between
different loading conditions was about 3 min (Leontijevic,
Pazin, Kukolj, Ugarkovic, & Jaric, 2013; Sreckovic et al., 2015).
During the maximum bench press pull test (BPULL) the
participants were instructed to lay face down the high bench
and placing their chin on the padded edge of the bench. Their
lower legs and the upper body were strapped to the bench
with padded rope. They pulled up the bar with maximum
effort until the bar struck the cushioned underside of the
bench. The rest periods were the same as in BPRESS.
External loading
The applied external loads in JUMP were 0, 8, 16, 24, and 32 kg
of weights that allowed all participants to perform the jumps.
In total, they performed 10 jumps (5 loads × 2 trials). Based on
a pilot testing, all subjects were able to jump (i.e., to reveal a
flight phase) with the 32 kg of added weights. Regarding
CYCLING, participants performed 5 sprints against the external

loads of 2, 4, 6, 8, and 10 kg (i.e., 1 trial per load). The
individual maximum load lifted in BPRESS and BPULL
appeared to be between 62 and 90 kg. Therefore, the applied
external loads were 20, 27.5, 35, 42.5, 50, and 57.5 kg. The
minimum load of 20 kg consisted of the bar and arm seg-
ments (Sreckovic et al., 2015), while the weight plates were
added for higher loads. In total, participants completed 12
bench press throws and 12 bench pulls (i.e., 6 loads × 2 trials
each).
Experimental protocol
Each participant completed 4 sessions separated by rest per-
iods of 5–7 days. The first session consisted of anthropometric
measurements, followed by familiarisation with 2 functional
tests. Specifically, body height and body mass were assessed
by a standard anthropometer and digital scale, respectively.
For the assessment of per cent body fat, a bioelectric impe-
dance method (In Body 720; USA) was used. Familiarisation
with another 2 tests was conducted during the second ses-
sion. The third and fourth testing sessions served for data
collecting. Note that the order of the tests was partially ran-
domised for each participant. Namely, to avoid fatigue, each
familiarisation and testing session involved 1 test of upper
limbs and 1 test of lower limbs. The sequence of loads was
randomised for each particular test. For all tests except
CYCLING, the first trial was used as practice trial, while the
second trial was used for further analysis. Standard 10-min
warm up procedures of the lower and upper body muscles
were conducted prior to each test. A strong verbal encourage-
ment was systematically provided.
Data analysis
The first step of data analysis was the calculation of the F and
V values averaged either across the trials concentric phase (in
JUMP, BPRESS, and BPULL), or from the entire test interval
(CYCLING). A custom-designed LabVIEW (National
Instruments, 2013; USA) programme was used to acquire and
process the vertical component of the ground reaction F
exerted during JUMP. The F signal was sampled at 1000 Hz
and low-pass filtered with a second-order recursive 10 Hz low-
pass Butterworth filter. Integration of the acceleration signal
calculated from F was conducted to calculate V (Cuk et al.,
2014). Regarding CYCLING, the device software was used to
acquire P and the frequency data. To assess the corresponding
linear measures, V was calculated from the frequency and the
crank length, while F was calculated as P divided by V.
Another custom-made Lab VIEW programme was used to
assess F and V from BPRESS and BPULL. 3D kinematic record-
ing (Qualisys AB, Gothenburg, Sweden) sampled at a rate of
240 Hz with recursive Butterworth 10 Hz low-pass filter was
used to assess the vertical position of the bar. V and accelera-
tion of the bar were calculated from the first and second
derivative of the position, respectively, while F was calculated
as a sum of the weight (i.e., mass times gravity acceleration)
and inertia (mass times acceleration) of the total mass lifted.
The second step of the data analysis was the assessment of
2 models and their parameters. The individual multiple-load
JOURNAL OF SPORTS SCIENCES 3
model was obtained from a linear regression applied on the
data observed from all loading conditions, while the two-load
model was drawn through the data obtained from only the
lightest and heaviest load. Each model separately revealed the
parameters Fmax (F-intercept), Vmax (V-intercept), and Pmax
(Pmax = Fmax · Vmax/4).
Statistical analyses
The analyses were performed on the parameters Fmax, Vmax,
and Pmax and correlation coefficients obtained separately from
the multiple-load and two-load model. Descriptive statistics
were calculated as mean and standard deviation, while the
correlation coefficients were presented through their median
values and ranges. Initial testing revealed that none of the
dependent variables significantly deviated from their normal
distribution (Kolmogorov–Smirnov test). Student’s t-test was
employed to test the differences between the same para-
meters observed from 2 models applied on the particular
sets of data, while Pearson’s correlations were used to test
their relationships. To assess systematic bias and the limits of
agreement of the 2 models, we also used Bland–Altman plots
(Bland & Altman, 1986). The level of statistical significance was
set to P < 0.05. All statistical tests were performed using SPSS
21 (IBM, Armonk, NY).
Results
Figure 1(a) shows the averaged across the participants F and V
data obtained from individual loads applied on 4 functional
performance tests. The multiple-load model revealed excep-
tionally strong and approximately linear relationships. When
applied on individual sets of data, the same model revealed
the correlation coefficients 0.951 (0.877–0.992) in JUMP, 0.995
(0.978–0.999) in CYCLING, 0.984 (0.963–0.991) in BPRESS, and
0.990 (0.940–0.997) in BPULL [all data shown as median
(range)]. Figure 1(b) illustrates the similarity of the multiple-
load and two-load model outcomes observed from a repre-
sentative set of individual data. Although the two-load model
(solid line) was observed from only the first and the last point
used to obtain the linear regression (i.e., the multiple-load
model; all points and dashed line), the 2 lines are almost
overlapping. As a result, the values of the parameters of the
2 models should have been very similar.
The following figures show a high level of agreement
between the 2 models through the similarity of their para-
meters observed across the participants. Specifically, Figure 2
presents the differences between the magnitudes, the same
parameters observed from the multiple-load and two-load
model applied separately on different functional performance
tasks. Although 4 out of 12 comparisons revealed significant
differences (P < 0.05; paired t-test), note that the magnitudes
of the differences were exceptionally small. Figure 3 shows the
relationships between the same parameters. All correlation
coefficients proved to be strong (all r > 0.952; P < 0.01).
Although the correlation coefficients obtained from BPRESS
appear to be somewhat lower than in the remaining 3 tests,
none of the 12 correlation coefficients was either above or
below the 95% confidence intervals of others.
4 M. Z. ZIVKOVIC ET AL.

Figure 1. (a) The linear regressions obtained from the averaged across the
participant data recorded from 4 functional movement tests. The individual
points represent F and V values recorded from particular loads, while the
correlation coefficients are also indicated. (b) Comparison of the 2 models
applied on a representative set of individual data obtained from BPRESS.
While the standard “multiple-load model” is based on a linear regression
applied on all experimentally recorded F and V values (both the open and filled
squares; dashed line), the “two-load model” represents the line drawn through
only the first and last point (circled; solid line) of the tested individual.

The Bland–Altman plots also revealed a high level of agree-
ment between the 2 models. Specifically, for all parameters the
fixed bias was below 3.4%, while the proportional bias was
R < 0.60 (all P > 0.05). Finally, the averaged 95% limit of
agreements was 5.6%, ranging from 2.9% (Vmax obtained from
CYCLING) to 10.7% (Vmax obtained from JUMP) (Figure 4).
Discussion
Within the present study, we compared a model observed
from just 2 distinctive loads with the outcomes of the stan-
dard multiple-load model used to obtain the F–V relationship
from various tasks performed under a number of loading
conditions. The results obtained from 4 different functional
movement tasks strongly supported our hypothesis. Namely,
both the absolute values, the relationships and the bias
observed between the models’ parameters revealed a high
Figure 2. The averaged across the participant values of the parameters Fmax
(top panel), Vmax (middle panel), and Pmax (bottom panel) obtained from the
multiple-load (open bars) and two-load model (filled bars) for different tests
(means with SD error bars). Significant differences (*P < 0.05) between the 2
models are also presented (corresponding effect sizes in parentheses).
level of agreement between the multiple-load and two-load
model.
Previous studies revealed that the F–V relationships
observed from the multiple-load model applied on loaded
functional movements proved not only to be linear and
strong, but also that their parameters Fmax, Vmax, and Pmax
could be reliable and valid indices of muscle F, V, and P
producing capacities, respectively (Jaric, 2015). The obtained
results strongly suggest that the two-load model that requires
neither the regression modelling nor more than 2 different
loads applied provides virtually identical outcomes regarding
the magnitudes of the F–V parameters. This could be

Figure 3. Correlation coefficients between the same parameters obtained from
the multiple-load and two-load model (corresponding 95% confidence intervals
also shown).
considered as an outcome that supports the concurrent valid-
ity of the two-load model with respect to the multiple-load
model already extensively used in literature. It is plausible to
assume that the level of agreement between the outcomes of
the two-load and multiple-load model originate from both the
strength and linearity of the F–V relationships of muscles
Figure 4. Bland–Altman plots illustrating the agreement of the 3 parameters (Fmax, Vmax, and Pmax) obtained from 4 different tests by means of the multiple-load
and two-load model. Dashed lines show the fixed bias, while the dotted lines show 95% limits of agreements.
JOURNAL OF SPORTS SCIENCES 5
performing various multi-joint functional tasks [(Jaric, 2015);
see also Introduction].
Various functional movements, such as maximum perfor-
mance jumping, cycling, running, throwing, and lifting, have
been routinely employed to assess the mechanical capacities
of involved muscles, detect differences among various popu-
lations, and explore the effects of various rehabilitation and
training interventions. As already pointed out in Introduction,
they have been typically tested under a single set of mechan-
ical conditions including a single external load. Just a quick
glance at Figure 1(b) reveals the main problem of such an
approach: a single experimental point does not allow for the
assessment of the F–V relationship and, therefore, the muscle
F, V, and P producing capacities cannot be assessed. Moreover,
the position of the point relative to the maximum F (i.e., zero
V), V (zero F), and P (the middle section of the line) of the
tested muscles cannot be assessed. One could only speculate
to what extent such an approach has created ambiguity in
literature regarding whether the outcomes of standard func-
tional movement tests reveal F, V, or P producing capacities of
the tested muscles.
The present data, however, suggest that adding of just
another external load could allow for using the two-load
model that distinguishes among the discussed muscle capa-
cities. Conversely, the obtained results also suggest that add-
ing more than just 1 external load may not be needed. That
eliminates a need for a linear regression modelling and also
makes the procedure both quicker and less prone to fatigue.
6 M. Z. ZIVKOVIC ET AL.
Finally, a question could be the method of selection of the
particular 2 loads. A plausible solution could be the selection
of 2 more, rather than less distinctive loads (as shown in
Figure 1(b)). Namely, if each individual load provides F and V
magnitudes with a similar error score, the error of the
obtained F–V relationship (and, therefore, of Fmax, Vmax, and
Pmax) would be smaller if calculated from more distant experi-
mental points.
Regarding the possible limitations of the discussed two-
load model, note that it could be applied only on multi-joint
functional movement tasks, since the F–V relationship
observed from in vitro muscles and single-joint movements
is generally considered to be curvilinear (Kaneko, Fuchimoto,
Toji, & Suei, 1983; McMahon, 1984). Nevertheless, it should be
noted that the functional movement tasks are not only exten-
sively used in routine testing, but also more ecologically valid
and typically more familiar for participants (Jaric, 2015). In
addition, due to their presumed linearity, any 2 parameters
(i.e., Fmax, Vmax, and Pmax) obtained from either of the models
allow for calculation of the third parameter. Therefore, despite
both the apparent and distinctive physiological meaning of
the obtained F, V, and P producing capacity of the tested
muscles, they are not fully independent. An important limita-
tion could be that the outcomes of the multiple-load model
have been often based on relatively narrow range of F and V
data and, therefore, Fmax and Vmax were the outcomes of
distant extrapolation. Furthermore, it remains underexplored
how different types of loads affect both the F–V relationship in
general and the concurrent validity of the two-load model
(Leontijevic et al., 2012, 2013). Somewhat lower concurrent
validity obtained from BPRESS could also need to be explored.
Furthermore, studies of the load associated changes in of work
and P contributions of individual joins could shed a light on
the mechanisms underlying the obtained F–V patterns
(Feeney et al., 2016). Finally, despite a high level of agreement
of the two-load model with the multiple-load model that
already revealed highly reliable parameters [(Jaric, 2015); see
also Introduction], the reliability of the two-load model appar-
ently needs to be evaluated.
Conclusions
A rather consistent set of findings observed from 4 distinc-
tive functional movement tasks support the concurrent
validity of the evaluated two-load model with respect to
the standard multiple-load models applied to assess the
muscle F–V relationships. Therefore, addition of just another
high external load to the standard tests of muscle capacities
based on functional movements typically conducted under a
single set of external conditions (e.g., maximum jumping,
running, cycling, rowing, lifting, throwing) could discern
among the F, V, and P producing capacities of the tested
muscles. As a result, the use of the two-load model could
both improve routine testing methods and resolve a number
of debated issues in literature regarding the interpretation
of the outcomes of various functional movement tests. With
respect to the multiple-load model routinely applied in the
literature, the two-load model provides a simpler and
quicker testing procedure that is also less prone to fatigue.
Nevertheless, further research is needed to standardise the
testing procedures regarding the magnitudes of external
loads applied, their type (Leontijevic et al., 2013), and addi-
tionally explore the reliability, validity, and sensitivity of the
observed parameters.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
This work was supported in part by the National Institute of Health, USA:
[Grant Number R21AR06065]; and the Serbian Research Council: [Grant
Number 175037].
References
Bland, J. M., & Altman, D. G. (1986). Statistical methods for assessing
agreement between two methods of clinical measurement. The
Lancet, 327, 307–310. doi:10.1016/S0140-6736(86)90837-8
Bobbert, M. F. (2012). Why is the force-velocity relationship in leg press
tasks quasi-linear rather than hyperbolic? Journal of Applied Physiology,
112, 1975–1983. doi:10.1152/japplphysiol.00787.2011
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd
ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.
Cuk, I., Markovic, M., Nedeljkovic, A., Ugarkovic, D., Kukolj, M., & Jaric, S.
(2014). Force-velocity relationship of leg extensors obtained from
loaded and unloaded vertical jumps. European Journal of Applied
Physiology, 114, 1703–1714. doi:10.1007/s00421-014-2901-2
Driss, T., Vandewalle, H., Le Chevalier, J. M., & Monod, H. (2002). Force-
velocity relationship on a cycle ergometer and knee-extensor strength
indices. Canadian Journal of Applied Physiology, 27, 250–262.
doi:10.1139/h02-015
Feeney, D., Stanhope, S. J., Kaminski, T. W., Machi, A., & Jaric, S. (2016).
Loaded vertical jumping: Force-velocity relationship, work, and power.
Journal of Applied Biomechanics, 32, 120–127.
Garcia-Ramos, A., Jaric, S., Padial, P., & Feriche, B. (2016). Force-velocity
relationship of upper-body muscles: Traditional vs. ballistic bench
press. Journal of Applied Biomechanics, 32, 178–185.
Giroux, C., Rabita, G., Chollet, D., & Guilhem, G. (2015). What is the best
method for assessing lower limb force-velocity relationship?
International Journal of Sports Medicine, 36, 143–149.
Jaric, S. (2015). Force-velocity relationship of muscles performing multi-
joint maximum performance tasks. International Journal of Sports
Medicine, 36, 699–704. doi:10.1055/s-00000028
Jaskolska, A., Goossens, P., Veenstra, B., Jaskolski, A., & Skinner, J. S. (1999).
Comparison of treadmill and cycle ergometer measurements of force-
velocity relationships and power output. International Journal of Sports
Medicine, 20, 192–197. doi:10.1055/s-1999-970288
Kaneko, M., Fuchimoto, T., Toji, H., & Suei, K. (1983). Training effect of
different loads on the force-velocity relationship and mechanical power
output in human muscle. Scandinavian Journal of Medicine & Science in
Sports, 5, 50–55.
Leontijevic, B., Pazin, N., Bozic, P. R., Kukolj, M., Ugarkovic, D., & Jaric, S.
(2012). Effects of loading on maximum vertical jumps: Selective effects
of weight and inertia. Journal of Electromyography and Kinesiology, 22,
286–293. doi:10.1016/j.jelekin.2011.12.002
Leontijevic, B., Pazin, N., Kukolj, M., Ugarkovic, D., & Jaric, S. (2013).
Selective effects of weight and inertia on maximum lifting.
International Journal of Sports Medicine, 34, 232–238.
McMahon, T. A. (1984). Muscles, reflexes, and locomotion. Princeton:
Princeton University Press.
Meylan, C. M., Cronin, J. B., Oliver, J. L., Hughes, M. M., Jidovtseff, B., &
Pinder, S. (2015). The reliability of isoinertial force-velocity-power profil-
ing and maximal strength assessment in youth. Sports Biomechanics, 14,
68–80. doi:10.1080/14763141.2014.982696

Morin, J. B., Samozino, P., Bonnefoy, R., Edouard, P., & Belli, A. (2010). Direct
measurement of power during one single sprint on treadmill. Journal of
Biomechanics, 43, 1970–1975. doi:10.1016/j.jbiomech.2010.03.012
Nikolaidis, P. T. (2012). Age- and sex-related differences in force-velocity char-
acteristics of upper and lower limbs of competitive adolescent swimmers.
Journal of Human Kinetics, 32, 87–95. doi:10.2478/v10078-012-0026-4
Rabita, G., Dorel, S., Slawinski, J., Saez-de-Villarreal, E., Couturier, A., Samozino,
P., & Morin, J. B. (2015). Sprint mechanics in world-class athletes: A new
insight into the limits of human locomotion. Scandinavian Journal of
Medicine & Science in Sports, 25, 583–594. doi:10.1111/sms.12389
Samozino, P., Rejc, E., Di Prampero, P. E., Belli, A., & Morin, J. B. (2012).
Optimal force-velocity profile in ballistic movements–altius: Citius or
fortius? Medicine & Science in Sports & Exercise, 44, 313–322.
doi:10.1249/MSS.0b013e31822d757a
Samozino, P., Rejc, E., di Prampero, P. E., Belli, A., & Morin, J. B. (2014).
Force-velocity properties’ contribution to bilateral deficit during ballis-
tic push-off. Medicine & Science in Sports & Exercise, 46, 107–114.
doi:10.1249/MSS.0b013e3182a124fb
JOURNAL OF SPORTS SCIENCES 7
Sanchez-Medina, L., Gonzalez-Badillo, J. J., Perez, C. E., & Pallares, J. G.
(2014). Velocity- and power-load relationships of the bench pull vs.
bench press exercises. International Journal of Sports Medicine, 35,
209–216.
Sreckovic, S., Cuk, I., Djuric, S., Nedeljkovic, A., Mirkov, D., & Jaric, S. (2015).
Evaluation of force-velocity and power-velocity relationship of arm
muscles. European Journal of Applied Physiology, 115, 1779–1787.
doi:10.1007/s00421-015-3165-1
Taylor-Piliae, R. E., Norton, L. C., Haskell, W. L., Mahbouda, M. H., Fair, J.
M., Iribarren, C., . . . Fortmann, S. P. (2006). Validation of a new brief
physical activity survey among men and women aged 60-69 years.
American Journal of Epidemiology, 164, 598–606. doi:10.1093/aje/
kwj248
Winter, E. M., Abt, G., Brookes, F. B., Challis, J. H., Fowler, N. E., Knudson,
D. V., . . . Yeadon, M. R. (2016). Misuse of “Power” and other mechanical
terms in sport and exercise science research. Journal of Strength and
Conditioning Research, 30, 292–300. doi:10.1519/
JSC.0000000000001101