Journal of Applied Biomechanics, 2018, 34, 104-110

https://doi.org/10.1123/jab.2016-0356
© 2018 Human Kinetics, Inc. ORIGINAL RESEARCH

Association of Sprint Performance With Ground Reaction Forces

During Acceleration and Maximal Speed Phases in a Single Sprint
Ryu Nagahara, Mirai Mizutani, Akifumi Matsuo, Hiroaki Kanehisa, and Tetsuo Fukunaga
National Institute of Fitness and Sports in Kanoya

We aimed to clarify the mechanical determinants of sprinting performance during acceleration and maximal speed phases of a
single sprint, using ground reaction forces (GRFs). While 18 male athletes performed a 60-m sprint, GRF was measured at every
step over a 50-m distance from the start. Variables during the entire acceleration phase were approximated with a fourth-order
polynomial. Subsequently, accelerations at 55%, 65%, 75%, 85%, and 95% of maximal speed, and running speed during the
maximal speed phase were determined as sprinting performance variables. Ground reaction impulses and mean GRFs during the
acceleration and maximal speed phases were selected as independent variables. Stepwise multiple regression analysis selected
propulsive and braking impulses as contributors to acceleration at 55%–95% (β > 0.72) and 75%–95% (β > 0.18), respectively, of
maximal speed. Moreover, mean vertical force was a contributor to maximal running speed (β = 0.48). The current results
demonstrate that exerting a large propulsive force during the entire acceleration phase, suppressing braking force when
approaching maximal speed, and producing a large vertical force during the maximal speed phase are essential for achieving
greater acceleration and maintaining higher maximal speed, respectively.

Keywords: running speed, GRF, impulse, propulsion, braking

Ground reaction force (GRF) during human running has been
examined for many decades,1–9 because, apart from gravitational
force, only GRF can accelerate or decelerate the human body
(ignoring air resistance). In the vertical and anterior–posterior
directions, the rates of force production, as well as peak and
mean forces, increase with increasing steady running speed.9–17
Effective vertical impulse, being an integrated value of vertical
force representing changes in the center of gravity vertical velocity
within a specific duration, increases up to moderate speed with
increasing steady running speeds, and decreases thereafter until
maximal running speed is reached.9,18 In contrast, both braking and
propulsive impulses increase with increasing steady running speed,
although published data exist only for speeds up to 7 m/s.12,15
These impulses reflect a large decrease and then increase in the
forward velocity of the center of gravity during the support phase at
higher speeds. As running speed during over-ground accelerated
sprinting increases, propulsive force and impulse decrease, while
braking force and impulse increase, resulting in a decrease in net
anterior–posterior impulse.6–8 In the vertical direction, the peak and
mean forces increase until approximately the 14th step and level off
thereafter, whereas effective impulse increases up to the 6th step,
and then plateaus during accelerated sprinting.7
Previous studies have also examined how GRF measures can
be associated with better maximal-effort sprinting (hereafter,
sprinting) performance,6,8,9 but findings differ among reports.
Weyand et al.9 have verified that a larger vertical force during a
short support duration was the critical factor for higher maximal
steady running speed on a treadmill with no preceding acceleration.
In contrast, Morin et al6 and Rabita et al8 found no significant
relationship between averaged vertical force or impulse over a
Nagahara, Mizutani, Matsuo, Kanehisa, and Fukunaga are with the National
Institute of Fitness and Sports in Kanoya, Kagoshima, Japan. Address author
correspondence to Ryu Nagahara at nagahara@nifs-k.ac.jp.
104
40-m distance and either maximal speed or mean running speed
over 40 m. They demonstrated that a larger propulsive force or
impulse during the acceleration phase after clearing the blocks was
essential for better accelerated sprinting performance. These con-
trasting findings may be because prior studies6,8 have examined the
relationship between the entire sprinting performance over a 40-m
acceleration, and averaged GRF or impulse during the correspond-
ing distance, while Weyand et al9 verified the corresponding
relationship at maximal speed.
No studies have examined associations of GRF measures
during the acceleration and maximal speed phases with running
acceleration and maximal running speed, respectively, over the
distance of the entire acceleration phase in a single sprint within the
same subjects. Therefore, whether the aforementioned findings can
be applied to a single sprint remains in question. Certainly, Morin
et al6 and Rabita et al8 reported that a larger propulsive force or
impulse, rather than a smaller braking force or impulse, was
essential to achieve better sprinting performance over a 40-m
distance or the 0 to 20-m section of a 40-m sprint. However,
they have used a macroscopic approach examining the relationship
between the entire sprinting performance over the 40-m distance or
in each 20-m section of the 40-m sprint and the averaged GRF or
impulse during the corresponding distance. In the previous studies,
therefore, there is a possibility that the influence of smaller braking
force at a specific spot during the acceleration phase might have
been masked by considering the entire distance. The importance of
increase in step length or frequency with respect to better accelera-
tion alters at approximately the 4th and 15th steps during the entire
acceleration phase by employing a step-to-step approach.19 There-
fore, when considering a more detailed interval during accelerated
sprinting, it is assumed that the association of a smaller braking
force or impulse, which contributes to increase running speed, with
larger acceleration at the corresponding spot may be elucidated.
This study aimed to clarify the mechanical determinants of
sprinting performance from the aspects of acceleration and

maximal speed phases in terms of GRFs during a single sprint. We
hypothesized that greater propulsive force during the entire accel-
eration phase, smaller braking force during the late acceleration
phase, and greater vertical force during the maximal speed phase
are essential for improving sprinting performance. Comprehen-
sively clarifying the mechanical determinants of sprinting perfor-
mance in terms of GRFs is of great importance for understanding
human locomotor functions, and knowledge about variables that
are essential for better sprinting performance will be beneficial for
the development of training programs to improve performance.
Methods
Eighteen male athletes (mean ± SD: age, 20.6 ± 1.2 y; stature,
1.73 ± 0.03 m; body mass, 66.5 ± 4.1 kg; personal best 100-m
race time, 11.28 ± 0.36 s [range, 10.54–12.30 s]) belonging to
university athletic clubs were recruited as participants in this study.
The aim, risks of involvement, and experimental conditions of the
study were explained before the experiment, and written informed
consent was obtained. The experimental procedures were conducted
with approval from the research ethics committee of the institute.
After warming up, the participants, wearing spiked shoes,
sprinted 60-m with maximal effort from starting blocks. The
Downloaded by IOWA STATE UNIV on 01/18/19
number of sprints ranged from 2 to 5. The rest period between
each trial was more than 10 min. Fifty-four force platforms
(1000 Hz) connected to a single computer (TF-90100, TF-3055,
TF-32120, Tec Gihan, Uji, Japan) were used to collect GRFs
during sprinting through 52 m from approximately 1.5 m behind
the starting line to the 50.5-m mark. Each starting block was bolted
to a separate force platform.
In this study, the variables representing sprinting performance
were running accelerations at specific percentages of maximal
running speed during the acceleration phase and running speed
during the maximal speed phase. There are several measures, such
as running speed, acceleration and horizontal external power, for
representing sprinting performance.20 For athletes, the real nature
of sprinting is to reach the finish in the least possible time. The
Figure 1 — Serial ground reaction force signals during accelerated sprinting over the 50-m distance for a typical participant. The numbers in each panel
indicate step numbers from the first step after block clearance.
JAB Vol. 34, No. 2, 2018
Determinants of Sprint Performance 105
integral of acceleration is a velocity and the mean velocity of a
specific distance has the same meaning as the time for covering that
distance. Moreover, the maximal speed is strongly correlated with
100-m race time,21 and the maximal speed mechanically results
from the preceding acceleration during the acceleration phase
(apart from an air resistance). This study focused on the accelera-
tion phase until the maximal speed, and the distances of the
acceleration phase differed among athletes. Taking these aspects
into account, we adopted the running acceleration as a measure of
sprint performance. GRF variables for the acceleration phase and
the maximal speed phase were defined as independent variables in
the regression analyses to examine their relationships with sprint-
ing performance.
Running speed at each step over the 50-m distance was
obtained from the GRF data (typical GRF signals being shown
in Figure 1). The foot strike at each step after clearing the blocks
was determined as the vertical GRF exceeded 20 N. The duration
for every step was defined as the time from the foot strike of one leg
to the next foot strike of the other leg. Step frequency was
calculated as the inverse of step duration. The position of the
ground contact foot after clearing the blocks was defined as the
center of pressure of GRF on the ground (mean of the center of
pressure position for 0.01 s during the middle of the support phase)
at each step. Step length was calculated as the distance between the
positions of the center of pressure for 2 consecutive steps in the
running direction. Running speed was calculated as the product of
step length and frequency. The fastest trial for each participant was
used for further analysis according to mean raw running speed over
the 50-m distance (from the start to the foot strike just before
reaching the 50.5-m mark).
GRF variables were calculated for each step from the first step
after the block clearance to the following steps. Using time
integration of vertical, propulsive and braking forces, vertical,
propulsive and braking impulses during the support phase were
computed. Net anterior–posterior impulse was calculated as the
sum of propulsive (positive value) and braking (negative value)
impulses. Respective GRF components were averaged for each
 106 Nagahara et al
support phase to obtain mean propulsive, braking, net anterior–
posterior, and vertical forces.
To examine the association among the variables at the exact
same percentage of maximal running speed during the acceleration
phase, step-to-step changes in variables during the acceleration
phase (up to maximal speed) for each participant were approxi-
mated against the time axis using a fourth-order polynomial
Downloaded by IOWA STATE UNIV on 01/18/19
Figure 2 — Examples of the approximation and first derivative (for
running speed only) of (A) running speed, (B) propulsive impulse,
(C) braking impulse, (D) net anterior−posterior impulse, (E) vertical
impulse, (F) mean propulsive force, (G) mean braking force, (H) net
anterior−posterior mean force, and (I) mean vertical force. The values were
from the same typical participant as shown in Figure 1. Gray, black and
dotted lines show the raw, approximated and first derivative data for each
variable, respectively. Vertical arrows indicate the variables at the 22nd
step and horizontal 2-headed arrows show the range of added data for
variables.
JAB Vol. 34, No. 2, 2018
(Figure 2).19,22 The order of approximation for running speed
was the same as that used in previous studies.19,22 For the GRF
variables, we also adopted a fourth-order polynomial approxima-
tion in the same manner as for running speed to obtain the variables
at the exact same percentage of maximal speed and to cancel
bilateral differences and variability of cyclic movement; the
approximation exhibited a good fit with the values, as shown in
Figure 2. In this study, we excluded the block clearance phase,
because GRF during this phase represents specific muscular and
technical abilities.8 Because there were bilateral differences and
cyclic movement variability in values, as can be seen in the
respective variables (Figure 2), an endpoint value of polynomial
results in an artifact, while the changes in variables can be assumed
to stabilize near the end of sprint acceleration. Therefore, to
eliminate the influence of the endpoint, we added the mean values
of the last 2 steps (data at maximal speed and just before the
maximal speed) for 3 steps after the maximal speed step for
the polynomial endpoint. The duration of the last step was used
as the time interval for the added data. Acceleration was calculated
as the first derivative of the polynomial of the approximated
running speed. Using the polynomial equation, acceleration and
GRF variables at each exact percentage of maximal speed during
the acceleration phase were computed. Initially, during this pro-
cess, a time point at which running speed was close to a specific
percentage (eg, 55%) of maximal speed was detected with 0.001-s
intervals using the approximated running speed. Using this time
point as input, variables at the specific percentages of maximal
speed were computed using the fourth-order (GRF variables) or
third-order (acceleration) polynomial equation. Running speed at
the first step exceeded half of maximal speed for a few participants,
but was less than 55% of maximal speed for all participants.
Therefore, the acceleration and GRF variables during the acceler-
ation phase were calculated at the following specific percentages of
maximal speed: 55, 65, 75, 85, and 95%.
Maximal running speed was obtained from fourth-order
approximated step-to-step running speed, because the raw values
included the bilateral difference and cyclic movement variability.
In addition, the GRF variables at the step where the maximal
running speed appeared were also obtained from fourth-order
approximated step-to-step GRF variables.
Descriptive data are presented as means and SDs. A stepwise
multiple regression analysis was performed, with acceleration at
each percentage of maximal speed and maximal speed as depen-
dent variables, and GRF variables at each percentage of maximal
speed and at maximal speed as independent variables. To avoid the
effect of multicollinearity of variables, the stepwise multiple
regression analysis was conducted separately for each of the
impulse and mean force variables. Moreover, with the same
reasoning, stepwise multiple regression analysis was performed
with propulsive, braking and vertical impulses or mean forces as
independent variables at first. Subsequently, the analysis was
performed with net anterior–posterior and vertical impulses or
mean forces as independent variables. The significance level
was set at P < .05. All statistical values were calculated using
JMP 12 statistical software (SAS Institute Japan Ltd, Tokyo,
Japan).
Results
The mean running speed over the 50-m distance and maximal
running speed were 8.16 ± 0.25 m/s and 9.59 ± 0.32 m/s, respec-
tively. The acceleration at the specific percentages of maximal
 Determinants of Sprint Performance 107

speed showed an almost linear relationship with percentage of
running speed (Figure 3C). As running speed increased, propulsive
and net anterior–posterior impulses and mean forces decreased
(Figure 3D, 3F, 3H, and 3J), while braking impulse and mean force
increased (Figure 3E and 3I). The vertical impulse was unchanged
throughout the acceleration phase (Figure 3G), while the mean
vertical force increased with an increase in running speed
(Figure 3K).
The greater propulsive, smaller braking and smaller vertical
impulses were identified as contributors to greater acceleration at
55%–95%, 75%–95% and 55%–95%, respectively, of maximal
Downloaded by IOWA STATE UNIV on 01/18/19
speed, while no impulse variables were selected at maximal speed
phase (Table 1). Moreover, greater propulsive, smaller braking and
smaller vertical mean forces emerged as contributors for greater
acceleration at 55%–75%, 75% and 75%, respectively, of maximal
speed. Greater vertical mean force was identified as a contributor
to maximal speed. The greater net anterior−posterior and smaller
vertical impulses were identified as contributors to greater
acceleration at 65%–95% of maximal speed, while the greater
net anterior−posterior and smaller vertical mean forces contributed
to greater acceleration at 65%–95% and 75%–95%, respectively, of
maximal speed (Table 2).

Figure 3 — Mean and individual values of changes in (A) distance, (B) speed, (C) acceleration, (D) propulsive impulse, (E) braking impulse, (F) net
anterior–posterior impulse, (G) vertical impulse, (H) mean propulsive force, (I) mean braking force, (J) net anterior−posterior mean force, and (K) mean
vertical force with respect to percentage of maximal speed. Black and gray lines show the mean and individual data for each variable.

JAB Vol. 34, No. 2, 2018

 108 Nagahara et al

Table 1 Results of Stepwise Multiple Regression Analysis Using Propulsive, Braking, and Vertical Impulses or Mean
Forces as Independent Variables
Impulse Mean force
Dependent
variables Adjusted R2 Propulsive Braking Vertical Adjusted R2 Propulsive Braking Vertical
Acceleration at 55% 0.40 (0.06) 0.72 (0.03) −0.22 (0.32) −0.65 (0.05) 0.36 (0.01) 0.60 (0.01)
of maximal speed
Acceleration at 65% 0.65 (<0.001) 0.98 (<0.001) −0.78 (0.001) 0.57 (<0.001) 0.75 (<0.001)
of maximal speed
Acceleration at 75% 0.88 (<0.001) 1.08 (<0.001) 0.18 (0.08) −0.71 (<0.001) 0.80 (<0.001) 1.43 (<0.001) 0.33 (0.04) −0.80 (<0.001)
of maximal speed
Acceleration at 85% 0.87 (<0.001) 1.07 (<0.001) 0.50 (<0.001) −0.53 (<0.001) No variables selected
of maximal speed
Acceleration at 95% 0.81 (<0.001) 1.26 (<0.001) 1.02 (<0.001) −0.42 (0.01) No variables selected
of maximal speed
Maximal speed No variables selected 0.23 (0.04) 0.48 (0.04)
Note. Values are shown as adjusted R2 (P-value) on the far left in the respective columns of impulse and mean force and β (P-value) of the respective selected variables on the
other column. The row with a notation of “No variables selected” indicates that no significant regression equation was made with the tested independent variables for
predicting sprinting performance at specific percentages of maximal speed. The blank cell shows that the variable was not selected as a predictor for regression equation.
Downloaded by IOWA STATE UNIV on 01/18/19

Table 2 Results of Stepwise Multiple Regression Analysis Using net Anterior–Posterior and Vertical Impulses or
Mean Forces as Independent Variables
Impulse Mean force
Dependent
variables Adjusted R2 Net anterior–posterior Vertical Adjusted R2 Net anterior–posterior Vertical
Acceleration at 55% No variables selected 0.16 (0.01) 0.40 (0.01)
of maximal speed
Acceleration at 65% 0.56 (0.002) 0.92 (<0.001) −0.76 (0.003) 0.53 (<0.001) 0.73 (<0.001)
of maximal speed
Acceleration at 75% 0.83 (<0.001) 1.05 (<0.001) −0.68 (<0.001) 0.90 (<0.001) 1.02 (<0.001) −0.26 (0.01)
of maximal speed
Acceleration at 85% 0.87 (<0.001) 1.07 (<0.001) −0.52 (<0.001) 0.85 (<0.001) 0.97 (<0.001) −0.35 (0.01)
of maximal speed
Acceleration at 95% 0.80 (<0.001) 1.08 (<0.001) −0.44 (0.007) 0.86 (<0.001) 0.94 (<0.001) −0.39 (0.001)
of maximal speed
Maximal speed No variables selected 0.23 (0.04) 0.48 (0.04)
Note. Values are shown as adjusted R2 (P-value) on the far left in the respective columns of impulse and mean force and β (P-value) of the respective selected variables on the
other column. The row with a notation of “No variables selected” indicates that no significant regression equation was made with the tested independent variables for
predicting sprinting performance at specific percentages of maximal speed. The blank cell shows that the variable was not selected as a predictor for regression equation.

Discussion ways to achieve a larger increase in running speed, the importance

This study aimed to clarify the mechanical determinants of sprint-
ing performance from aspects of acceleration and maximal speed
phases in terms of GRFs during a single sprint. As hypothesized,
the current results generally demonstrate that the larger propulsive
impulse and mean force during the entire acceleration phase,
smaller braking and vertical impulses and mean forces when
approaching maximal speed, and larger mean vertical force during
the maximal speed phase are important in a single sprint for better
sprinting performance.
Stepwise multiple regression analysis revealed that the greater
propulsive and smaller braking impulses were the contributors for
greater acceleration at 55%–95% (2.9 ± 0.2 m to 23.1 ± 2.2 m
mark) and 75%–95% (7.5 ± 0.6 m to 23.1 ± 2.2 m mark), respec-
tively, of maximal speed. Although both a larger propulsive
impulse and a smaller braking impulse are mechanically reasonable
JAB Vol. 34, No. 2, 2018
of a smaller braking impulse appeared at 75% (7.5 ± 0.6 m mark) or
more of maximal speed. This indicates the importance of a small
braking impulse for better performance at relatively high speeds.
The propulsive, braking and vertical mean forces were selected as
predictors of acceleration at 75% of maximal speed (7.5 ± 0.6 m
mark) as with the case of impulses. However, no mean force
variable was selected as a contributor for accelerations at 85 and
95% of maximal speed (12.4 ± 1.0 m and 23.1 ± 2.2 m marks),
indicating that the accelerations at 85 and 95% of maximal speed
cannot be predicted with combinations of the propulsive, braking
and vertical mean forces. Taking into account that the net anterior–
posterior mean force was selected at all examined percentages of
maximal speed, the combination of magnitudes of propulsive and
braking mean forces (greater propulsive mean force with greater
braking mean force or smaller propulsive mean force with smaller
braking mean force) for greater acceleration through greater net

anterior–posterior mean force may differ among athletes at higher
speeds when approaching maximal speed. The current results also
showed that the smaller vertical impulse predicted better accelera-
tion performance during the entire acceleration phase. Moreover,
the advantage of the smaller mean vertical force was found at
higher speeds during the acceleration phase. These findings indi-
cate that, in addition to the necessity of propulsive force for better
acceleration, the forward-leaning GRF is important for achieving
better acceleration performance during the acceleration phase of
sprinting.
The current findings regarding the importance of greater
propulsive impulse and mean force with smaller vertical impulse
and mean force at specific percentages of maximal speed during the
entire acceleration phase of sprinting are in line with previous
studies.6,8 Morin et al6 found that averaged propulsive impulse over
the entire running distance (40 m) was associated with sprinting
performance (mean speed over 40 m). Moreover, Rabita et al8
revealed that the mean propulsive force and ratio of net anterior
−posterior mean force to total applied force over a 40-m distance
was highly correlated with sprinting performance. In their results,
however, there was no significant correlation between averaged
braking impulse and sprinting performance.6 While the previous
study investigated the relationship between the entire sprinting
Downloaded by IOWA STATE UNIV on 01/18/19
performance and GRFs using a macroscopic approach (averaged
over the entire acceleration), the present study examined the
predictors of acceleration in terms of GRF variables at each
percentage of maximal speed. This is a possible reason for the
discrepancy between the current and previous studies, because
the influence of smaller braking force at a specific spot during the
acceleration phase might be masked by considering the entire
distance.6 Further, the difference in performance levels between
the present and previous studies (best record of 100-m race: 10.54–
12.30 s versus 9.95–10.60 s) could be another reason for the
disparity.
For the vertical force during the maximal speed phase, a large
force rather than a large impulse is theoretically necessary for better
sprinting performance, because the support time at high maximal
speed is very short and sprinters have to produce the requisite
quantity of vertical impulse to maintain sprinting during that
limited time. This concept was accepted in a previous study.9 In
contrast, Morin et al6 and Rabita et al8 could not verify the
importance of vertical force. The reason for this inconsistency is
that Morin et al6 and Rabita et al8 examined the relationship of the
averaged vertical force or impulse over a 40-m distance to sprinting
performance (maximal speed and mean running speed over 40 m),
while Weyand et al9 adopted the variables at steady maximal speed
only (ie, with no preceding acceleration). The mean vertical force
as the predictor of running speed during the maximal speed phase
(45.5 ± 3.2 m mark) in this study supports the findings of Weyand
et al.9 Moreover, the current results indicate that smaller vertical
impulse and mean forces are more conductive to greater accelera-
tion at all examined range of running speed. Consequently, it can be
argued that greater vertical force during the maximal speed phase,
rather than the acceleration phase, is essential for achieving better
sprinting performance.
A limitation of the present study is that the average distance
from the starting line to 95% of maximal speed was 23.1 m, which
is relatively short compared with the distance of entire acceleration
phase (>40 m). To compare the serial changes in variables between
individuals, the x-axis needs to be standardized (eg, with time,
distance, speed, or percentage of maximal speed). There is always a
biasing problem. The range of performance levels in our study
JAB Vol. 34, No. 2, 2018
Determinants of Sprint Performance 109
(personal best 100-m race time: 10.54–12.30 s) was not an
international standard. Thus, there is a possibility that the results
may differ for world-class sprinters.
In conclusion, the current results indicate that, in a single
maximal sprint, exertion of a large propulsive force during the
entire acceleration after clearing the blocks, suppression of braking
and vertical forces when approaching maximal speed, and produc-
tion of a large vertical force at maximal speed are essential for
improving performance. These findings may deepen understanding
of sprinting performance and be beneficial for athletes and coaches
as they attempt to improve sprint acceleration performance.
Acknowledgments
We thank Dr Neil Bezodis (Swansea University, UK) for his comments on
this study. There was no financial support to conduct this study. This
research was conducted at the National Institute of Fitness and Sports in
Kanoya.
References
1. Cavagna GA, Komarek L, Mazzoleni S. The mechanics of sprint
running. J Physiol. 1971;217:709–721. PubMed doi:10.1113/
jphysiol.1971.sp009595
2. Fukunaga T, Matsuo A, Ichikawa M. Mechanical energy output and
joint movements in sprint running. Ergonomics. 1981;24:765–772.
PubMed doi:10.1080/00140138108924898
3. Hunter JP, Marshall RN, McNair PJ. Interaction of step length and
step rate during sprint running. Med Sci Sports Exerc. 2004;36:261–
271. PubMed doi:10.1249/01.MSS.0000113664.15777.53
4. Hunter JP, Marshall RN, McNair PJ. Relationships between ground
reaction force impulse and kinematics of sprint-running acceleration.
J Appl Biomech. 2005;21:31–43. PubMed doi:10.1123/jab.21.1.31
5. Mero A, Komi PV. Force-, EMG-, and elasticity-velocity relation-
ships at submaximal, maximal and supramaximal running speeds in
sprinters. Eur J Appl Physiol Occup Physiol. 1986;55:553–561.
PubMed doi:10.1007/BF00421652
6. Morin JB, Slawinski J, Dorel S, et al. Acceleration capability in elite
sprinters and ground impulse: push more, brake less? J Biomech.
2015;48:3149–3154. PubMed doi:10.1016/j.jbiomech.2015.07.
009
7. Plamondon A, Roy B. Kinematics and kinetics of sprint acceleration.
Can J Appl Sport Sci. 1984;9:42–52. PubMed
8. Rabita G, Dorel S, Slawinski J, et al. Sprint mechanics in world-class
athletes: a new insight into the limits of human locomotion. Scand J
Med Sci Sports. 2015;25:583–594. PubMed doi:10.1111/sms.12389
9. Weyand PG, Sternlight DB, Bellizzi MJ, Wright S. Faster top running
speeds are achieved with greater ground forces not more rapid leg
movements. J Appl Physiol. 2000;89:1991–1999. PubMed
10. Belli A, Kyrolainen H, Komi PV. Moment and power of lower limb
joints in running. Int J Sports Med. 2002;23:136–141. PubMed doi:10.
1055/s-2002-20136
11. Clark KP, Weyand PG. Are running speeds maximized with simple-
spring stance mechanics? J Appl Physiol. 2014;117:604–615.
PubMed doi:10.1152/japplphysiol.00174.2014
12. Hamill J, Bates BT, Knutzen KM, Sawhill JA. Variations in ground
reaction force parameters at different running speeds. Hum Mov Sci.
1983;2:47–56. doi:10.1016/0167-9457(83)90005-2
13. Kyrolainen H, Komi PV, Belli A. Changes in muscle activity patterns
and kinetics with increasing running speed. J Strength Cond Res.
1999;13:400–406.
 110 Nagahara et al
14. Luhtanen P, Komi PV. Force-, power-, and elasticity-velocity
relationships in walking, running, and jumping. Eur J Appl
Physiol Occup Physiol. 1980;44:279–289. PubMed doi:10.1007/
BF00421627
15. Munro CF, Miller DI, Fuglevand AJ. Ground reaction forces in
running: a reexamination. J Biomech. 1987;20:147–155. PubMed
doi:10.1016/0021-9290(87)90306-X
16. Nigg BM, Bahlsen HA, Luethi SM, Stokes S. The influence of
running velocity and midsole hardness on external impact forces
in heel-toe running. J Biomech. 1987;20:951–959. PubMed doi:10.
1016/0021-9290(87)90324-1
17. Nilsson J, Thorstensson A. Ground reaction forces at different speeds
of human walking and running. Acta Physiol Scand. 1989;136:
217–227. PubMed doi:10.1111/j.1748-1716.1989.tb08655.x
18. Dorn TW, Schache AG, Pandy MG. Muscular strategy shift in human
running: dependence of running speed on hip and ankle muscle
Downloaded by IOWA STATE UNIV on 01/18/19
JAB Vol. 34, No. 2, 2018
performance. J Exp Biol. 2012;215:1944–1956. PubMed doi:10.
1242/jeb.064527
19. Nagahara R, Naito H, Morin JB, Zushi K. Association of acceleration
with spatiotemporal variables in maximal sprinting. Int J Sports Med.
2014;35:755–761. PubMed doi:10.1055/s-0033-1363252
20. Bezodis NE, Salo AI, Trewartha G. Choice of sprint start performance
measure affects the performance-based ranking within a group of
sprinters: which is the most appropriate measure? Sports Biomech.
2010;9:258–269. PubMed doi:10.1080/14763141.2010.538713
21. Slawinski J, Termoz N, Rabita G, et al. How 100-m event analyses
improve our understanding of world-class men’s and women’s sprint
performance. Scand J Med Sci Sports. 2017;27:45–54. PubMed
doi:10.1111/sms.12627
22. Bezodis NE, Salo AI, Trewartha G. Measurement error in estimates of
sprint velocity from a laser displacement measurement device. Int J
Sports Med. 2012;33:439–444. PubMed doi:10.1055/s-0031-1301313