Secondary Logo

Journal Logo

Original Research

Relationships Between Ground Reaction Impulse and Sprint Acceleration Performance in Team Sport Athletes

Kawamori, Naoki1,2; Nosaka, Kazunori1; Newton, Robert U.1

Author Information
Journal of Strength and Conditioning Research: March 2013 - Volume 27 - Issue 3 - p 568-573
doi: 10.1519/JSC.0b013e318257805a
  • Free



Sprint running is a fundamental activity in many team sports. A faster athlete has an obvious advantage during decisive periods of a match because he or she has a greater chance of getting to a ball or moving into open space before an opponent. Maximal sprinting over 100 m has consistently shown 3 phases of speed generation: positive acceleration, maintenance of maximum speed, and deceleration (or negative acceleration) (14). In team sports, the positive acceleration ability is of particular importance because sprint efforts during team sport competitions are generally of short duration (e.g., 10–20 m, 2–3 seconds) (13). Coaching and conditioning literature commonly use the word "acceleration" to mean a positive horizontal acceleration (increasing running speed) or even short sprint performance, so the colloquial meaning of the word will be used for the remainder of this article.

The acceleration of the athlete's center of mass during sprint running is determined by body mass and 3 external forces acting on the body: (a) ground reaction force (GRF), (b) gravitational force, and (c) air or wind resistance (6). As an athlete has the most influence on the GRF of these 3 external forces, it is likely that GRF has a significant impact on sprint acceleration performance (6,8). For analytical purposes, GRF during sprint running can be resolved into 3 orthogonal components (i.e., vertical, anterior-posterior, medial-lateral), of which the vertical and anterior-posterior components are usually of most interest (6). Anterior-posterior GRF (hereafter termed “horizontal” GRF) for each foot strike can be further subdivided into braking and propulsive phases. Each GRF component can be analyzed in terms of kinetic (e.g., peaks, means, impulses) and temporal (e.g., durations of certain phases) characteristics, in relation to sprint acceleration performance.

According to the impulse-momentum relationship (Newton's Second Law), net horizontal GRF impulse normalized to body mass is the major determining factor of the change in the horizontal velocity of the athlete during ground contacts. However, simply trying to maximize the net horizontal GRF impulse may not be the best approach to improve sprint acceleration performance because an athlete still needs to produce GRF in a vertical direction to reverse the downward motion/velocity of the body upon landing and then to propel the body upward to create flight time long enough to reposition the lower limbs. Moreover, simply attempting to maximize net horizontal GRF impulse may result in longer ground contact time and lower step frequency, which could be detrimental to sprint acceleration performance. Therefore, there should be an optimal combination of the magnitude, direction, and duration of GRF that maximizes sprint acceleration performance; however, a few attempts have been made to find such a combination.

Mero (8) investigated sprint start of track sprinters and reported that running velocity at the end of the first ground contact after leaving starting blocks was significantly correlated (r = 0.62–0.71) with horizontal propulsive GRF (maximal, average, and impulse) and also with vertical GRF (r = 0.41–0.50). Similarly, Hunter et al. (6) reported that sprint velocity at 16 m after a start had significant correlations with net horizontal (r2 = 0.61), propulsive (r2 = 0.57), and vertical (r2 = 0.17) impulses normalized to body mass in a subject population that included both track and field athletes and team sport athletes. The results of these studies provide insights into the optimal GRF pattern for better sprint acceleration performance, but it is questionable whether they can be directly applied to team sport athletes because differences in sprint running techniques likely exist between track sprinters and team sport athletes (e.g., shoe cleats, running surface, running posture, height of foot during recovery). Moreover, team sport athletes usually do not start sprinting from starting blocks or from a crouching position during a game; a standing start is more specific to team sport athletes. In fact, considerable biomechanical differences (e.g., step length, knee and hip angles at push-off) between a crouching start and a standing start have been reported during the initial 10 m of sprinting (11).

Therefore, the purpose of this study was to investigate the relationships between GRF parameters and sprint acceleration performance (0–10 m) from a standing start in team sport athletes (no track sprinters included). This study particularly focused on impulse measures of GRF because they reflect the acceleration (or the rate of change in velocity) of a runner's center of mass during each foot contact, when normalized to body mass (6). The importance of this research for strength and conditioning will be increased knowledge of specific application of impulse during sprinting from a standing start in team sport populations, which is critical to inform resistance training program design and sprint drills to enhance performance.


Experimental Approach to the Problem

To determine the relationships between GRF impulses (normalized to body mass) and sprint acceleration performance from a standing start, 10-m sprint time and GRF data were collected in the same trials. Then, we evaluated the correlations between GRF impulses and sprint times. The test distance of 10 m was chosen because (a) a 10-m sprint test is often used to assess sprint acceleration ability of team sport athletes (1,3) and (b) the initial 10 m of sprint running has been shown to be a specific component representing initial acceleration ability (4).


Thirty physically active men of a team sport background (mean ± SD: age 23 ± 4 years, height 181 ± 6 cm, body mass 79 ± 8 kg [range, 64–101 kg]) were recruited for this study. All the participants were required to have played team sports (soccer, basketball, field hockey, rugby union, and Australian Rules football) for at least 5 years at the time of the data collection to be included in this study so that they were considered to have sprint techniques that are specific to team sport athletes and are different from those of track sprinters. Their competitive levels ranged from recreational athletes to state team members. The sample size has been determined based on the previous studies that examined the associations between GRF and sprint performance (6,8,15). This project was reviewed and approved by the institutional ethics committee. The participants were informed of the study requirements, benefits, and possible risks and then gave their written informed consent before participation.


The sprint test was performed in a biomechanics laboratory with a hard flat surface, and the participants wore their own athletic shoes. The participants were asked to refrain from strenuous exercise for 24 hours before the testing session. After performing a standardized warm-up consisting of light jogging, dynamic stretching exercises, and submaximal sprints of increasing intensity, the participants performed 6 maximal effort sprints over 10 m from a standing start with approximately 2-minute rest between trials. The starting technique of the sprint test was standardized using a parallel start in which the participants started in a standing position with the toes of both feet parallel at 0.3 m behind the first timing gate and moved forward with the first step (no step backward was allowed) (2). The present study used the parallel start to minimize the within-subject and between-subject variation in starting technique by limiting the rocking or swinging motion before the start and thus standardizing the sprint velocity when breaking the first timing gate. The sprint time was measured using a dual-beam electronic timing system (Swift Performance Equipment, Lismore, Australia), which had an accuracy of 0.01 seconds.

Ground reaction force was collected at a sampling frequency of 1,000 Hz during the first ground contact and at 8 m after the start, using 3 force plates recessed in series (Type 9287BA; Kistler Instrument Corp., Winterthur, Switzerland), which were level with the surrounding surface. The GRFs during the first ground contact and at 8 m from the start were collected in separate trials (3 trials for each), by changing the starting line and the positions of the timing gates. The same foot was involved in hitting the force plates for each participant over the 3 trials for each data collection point.

The recorded GRF data (Figure) were filtered using a fourth-order recursive, zero phase-shift Butterworth low-pass filter with a cutoff frequency of 100 Hz. From the filtered GRF-time data, the contact time was determined as the duration between the instants of foot strike and takeoff, which were defined as when the vertical GRF first rose above 10 N and dropped below 10 N, respectively. The vertical (effective) impulse was determined as the area under the vertical GRF-time curve minus body weight impulse over the time of ground contact. The braking and propulsive impulses were obtained by integrating all the negative and positive values of horizontal GRF, respectively, over the time of ground contact. The net horizontal impulse was calculated as propulsive impulse minus the absolute value of braking impulse (6). The resultant impulse was calculated as the area under the resultant GRF-time curve over the time of ground contact. The resultant GRF in this case is the vector addition of vertical, horizontal (anterior-posterior), and medial-lateral GRF. All the impulse measures were normalized to body mass, to reduce the covariate effects of body mass (10), and so that they represent the changes in velocity of center of mass during the ground contact.

typical example of GRF-time curve during the 10-m sprint test for horizontal GRF during the first ground contact after the start (A), vertical GRF during the first ground contact after the start (B), horizontal GRF at 8 m from the start (C), and vertical GRF at 8 m from the start (D). A–C) Ground contact phase during the first ground contact. A, B) Braking phase during the first ground contact. B, C) Propulsion phase during the first ground contact. D–H) Ground contact phase at 8 m. D–G) Braking phase at 8 m. E–H) Propulsion phase at 8 m. GRF = ground reaction force.

Statistical Analyses

The test-retest reliability for all variables was determined using a subgroup of 11 participants with 3–7 days between the 2 separate testing occasions. The intraclass correlation coefficient (ICC) and coefficient of variation (CV) were calculated for each variable. Data from the 3 trials for each condition (i.e., first ground contact and at 8 m from the start) were averaged and used for the calculation of ICC and CV.

For the correlation analyses, data from the 3 trials for each condition (i.e., first ground contact and at 8 m from the start) were averaged and used. Pearson product moment coefficient of correlation (r) was used to examine the relationships among 10-m sprint time and GRF impulses. All statistical analyses were conducted using SPSS (Version 11.5; SPSS, Inc., Chicago, IL, USA). Criterion for statistical significance was set at an alpha level of p ≤ 0.01.


The test-retest reliability data are shown in Table 1. The descriptive data shown in Table 2 indicate that there were no differences in 10-m sprint time between the trials in which GRF data were collected at the first ground contact (the initial 3 trials) and at 8 m after the start (the latter 3 trials).

Table 1
Table 1:
Test-retest reliability based on ICC and CV for 10-m sprint time and 5 different relative GRF impulses at the first ground contact and at 8 m from the start.*
Table 2
Table 2:
Mean ±SD for 10-m sprint time and 5 relative GRF impulses at the first ground contact and at 8 m from the start.*

During the first ground contact, no impulse measures had significant correlations with 10-m sprint time (Table 3). In contrast, at 8 m from the start, relative net horizontal and propulsive impulses had significant correlations with 10-m sprint time, but relative resultant, vertical, and braking impulses did not (Table 3).

Table 3
Table 3:
Pearson correlation coefficients between 10-m sprint time and 5 relative GRF impulses at the first ground contact and at 8 m from start.*

The correlations among relative GRF impulses recorded at 8 m from the start can be observed in Table 4. The correlations among relative GRF impulses recorded during the first ground contact are not shown because none of these variables showed significant correlations with the 10-m sprint time.

Table 4
Table 4:
Intercorrelation matrix between 5 relative GRF impulses at 8 m from start.*


The main findings of this study were that the 10-m sprint time correlated weakly and negatively with the relative net horizontal and propulsive impulses but not with the relative resultant, vertical, and braking impulses measured at 8 m from the start. This indicates that the faster subjects in this study applied ground reaction impulse in a more horizontal direction in achieving better sprint acceleration. On the other hand, none of the impulse measures collected at the first ground contact were correlated with the sprint time.

The present finding that faster participants over 10 m produced larger net horizontal impulse at 8 m from the start was expected because (a) applying larger net horizontal impulse relative to body mass results in larger horizontal acceleration of the center of mass during each ground contact (according to the impulse-momentum relationship) and (b) achieving larger horizontal acceleration during each ground contact is likely to lead to better overall sprint performance over 10 m, if this does not result in excessive increases in contact time and flight time, which would ultimately reduce step frequency. Hunter et al. (6) also reported a high significant correlation (r = ∼0.78) between sprint velocity and relative net horizontal impulse, both measured at 16 m from the start. However, the magnitude of the correlation in the present study was not as high (r = −0.52), with the net horizontal impulse at 8 m explaining only 27% of variance in 10-m sprint time. The higher correlation found by Hunter et al. (6) might be explained by a relatively heterogeneous group of track and field athletes and team sport athletes they employed and the fact that they had their subjects sprint on a synthetic track wearing spiked track shoes.

On the other hand, no significant correlations were found between sprint time and relative net horizontal impulse during the first ground contact. This is somewhat surprising as we expected that net horizontal impulse production would be more important immediately after the start where an athlete needs to overcome the inertia of the body to quickly accelerate from a stationary start. The parallel start we used in this study may in part explain this unexpected result. When sprint is initiated from a parallel start, the horizontal distance between the body center of mass and the point of foot contact is short. Such characteristics of the parallel start would make it difficult to apply force to the ground horizontally during the initial few steps (7) and might have led to the lack of strong correlations between sprint time and net horizontal impulse during the first ground contact. Although we used the parallel start in this study to minimize the variation in starting techniques, such a starting technique may have been somewhat unnatural and affected the result. Thus, replicating the present study using different starting techniques (e.g., split start) would be of future interest.

We also hypothesized that participants with faster 10-m sprint time would produce smaller braking impulse initially and larger propulsive impulse later during ground contacts, a combination of which would maximize net horizontal impulse and ultimately lead to greater horizontal acceleration. However, we did not find strong evidence that smaller braking impulse was associated with better sprint acceleration performance. This is presumably because the magnitude of braking impulse in the initial acceleration phase is so small that an attempt to minimize the braking impulse does not contribute meaningfully to maximizing the net horizontal impulse, which therefore does not lead to better 10-m sprint performance. On the other hand, faster participants produced larger propulsive impulse at 8 m from the start, with shared variance of 44% between propulsive impulse and 10-m sprint time. This finding partially supports the above hypothesis and also agrees with the results of previous research (6,8,12). In contrast, no significant correlations were observed between sprint time and propulsive impulse during the first ground contact, and this may be explained again by the starting technique we employed (i.e., parallel start).

The magnitudes of relative vertical (effective) impulse had no significant relationships with sprint time, both during the first ground contact and at the 8-m mark from a start. Thus, applying greater impulse in a vertical direction during ground contacts, as suggested to be important in achieving faster maximal sprinting speed by Weyand et al. (15), may not be so important in achieving high acceleration during the sprint start. Theoretically, producing larger impulse in a vertical direction during ground contacts would result in greater vertical velocity of the center of mass at takeoff, which subsequently leads to longer flight time (5). Because an athlete can horizontally accelerate his or her center of mass only when in contact with the ground, spending unnecessarily long time in the air and less time on the ground may not be desirable, especially in the acceleration phase of sprinting. Furthermore, producing large impulse in a vertical direction potentially means less impulse can be directed horizontally, which may not be desirable considering the seemingly important role of the horizontal impulse production in sprint acceleration (6). In fact, this is partially supported by the significant negative correlation found between the relative vertical impulse and the relative propulsive impulses.

Whereas separately analyzing “components” of GRF impulse such as vertical and horizontal impulses is a useful way to identify optimal pattern and direction of impulse production, it should be noted that those components are not independent of each other but of a single entity (i.e., resultant impulse) (6). Therefore, we also quantified resultant impulse and analyzed its relation with sprint acceleration performance. Surprisingly, we did not find significant correlations between the relative resultant impulse and the 10-m sprint time. In fact, all the correlations were “positive,” indicating that faster participants tended to produce smaller resultant impulses. Considering that some components of impulse (e.g., net horizontal and propulsive impulses) had “negative” significant correlations with sprint time, it could be argued that the “direction” of impulse application is more important to achieve better sprint acceleration than simply producing large magnitude of (resultant) impulse irrespective of its direction during ground contact (9).

It is important to note that there were some limitations of this study. First, we analyzed the GRF recorded only during the 2 ground contacts and these were measured in different trials, and it was assumed that these would represent GRF production patterns over the 10-m sprint distance. Second, the present findings are likely to be only applicable to sprint acceleration initiated from the parallel start. The starting techniques employed could affect the pattern of GRF production and their importance to the initial sprint acceleration performance. Third, the results are not necessarily applicable to other phases of sprinting and outside the caliber and the type of athletes we tested. Finally, causation or long-term training effects cannot automatically be assumed for all results in this article.

In conclusion, the ability to produce large net horizontal and propulsive impulses, or in other words applying impulse in a more horizontal direction, appears to be important to achieve high acceleration during 10-m sprints from a standing start. This should be considered in training and practice to improve sprint acceleration performance. Future research is required to validate the application hypothesis that altering the magnitude and direction of GRF impulse through training and practice improves sprint acceleration performance.

Practical Applications

This study showed that the magnitude of relative ground reaction impulse (resultant impulse) is not correlated with sprint acceleration performance and that the direction of impulse application is likely to be more important so that applying impulse in a more horizontal direction may lead to faster sprint acceleration. Because correlations do not prove causation or training effects, future studies should investigate, using a longitudinal (pretest-posttest) experimental design, whether the ability to apply impulse more horizontally could be trained or improved, by what means (practice or training exercises typically used to emphasize horizontal force/impulse production), and whether the improved ability to apply impulse horizontally could actually enhance sprint acceleration performance.


The authors thank Jonathon Green for his technical assistance and the participants for their involvement in this research. There was no financial assistance with the project.


1. Baker D, Nance S. The relation between running speed and measures of strength and power in professional rugby league players. J Strength Cond Res 13: 230–235, 1999.
2. Cronin JB, Green JP, Levin GT, Brughelli ME, Frost DM. Effect of starting stance on initial sprint performance. J Strength Cond Res 21: 990–992, 2007.
3. Cronin JB, Hansen KT. Strength and power predictors of sports speed. J Strength Cond Res 19: 349–357, 2005.
4. Delecluse C, Van Coppenolle H, Willems E, Van Leemputte M, Diels R, Goris M. Influence of high-resistance and high-velocity training on sprint performance. Med Sci Sports Exerc 27: 1203–1209, 1995.
5. Hunter JP, Marshall RN, McNair PJ. Interaction of step length and step rate during sprint running. Med Sci Sports Exerc 36: 261–271, 2004.
6. Hunter JP, Marshall RN, McNair PJ. Relationships between ground reaction force impulse and kinematics of sprint-running acceleration. J Appl Biomech 21: 31–43, 2005.
7. Kraan GA, van Veen J, Snijders CJ, Storm J. Starting from standing; why step backwards? J Biomech 34: 211–215, 2001.
8. Mero A. Force-time characteristics and running velocity of male sprinters during the acceleration phase of sprinting. Res Q Exerc Sport 59: 94–98, 1988.
9. Morin JB, Edouard P, Samozino P. Technical ability of force application as a determinant factor of sprint performance. Med Sci Sports Exerc 43: 1680–1688, 2011.
10. Mullineaux DR, Milner CE, Davis IS, Hamill J. Normalization of ground reaction forces. J Appl Biomech 22: 230–233, 2006.
11. Salo A, Bezodis I. Which starting style is faster in sprint running—standing or crouch start? Sports Biomech 3: 43–53, 2004.
12. Sleivert G, Taingahue M. The relationship between maximal jump-squat power and sprint acceleration in athletes. Eur J Appl Physiol 91: 46–52, 2004.
13. Spencer M, Bishop D, Dawson B, Goodman C. Physiological and metabolic responses of repeated-sprint activities: Specific to field-based team sports. Sports Med 35: 1025–1044, 2005.
14. Volkov NI, Lapin VI. Analysis of the velocity curve in sprint running. Med Sci Sports 11: 332–337, 1979.
15. 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 89: 1991–1999, 2000.

biomechanics; kinetics; running; speed; horizontal velocity

© 2013 National Strength and Conditioning Association