Sprint running performance is determined by 3 main factors: the ability to develop maximal forward acceleration, the maximal speed (MS) attained, and the ability to maintain speed against the onset of fatigue (16). Simulation models have shown that for a 100-m sprint, the best pacing strategy is an all-out effort, even if this causes a strong reduction of speed at the end of the race (5). Consequently, the speed profile is used to subdivide the 100-m sprint performance into 4 phases: a start action, an acceleration phase, a phase of MS, and a deceleration phase. During the start action, the sprinter accelerates the body's center of mass in a forward and upward direction. He strives to clear the starting blocks in the shortest possible time and with the highest possible horizontal speed (6). The acceleration phase in sprint running is subdivided into an initial acceleration (IA) phase (0–10 m) and a transition phase (TP) (10–30 m). The IA phase is characterized by the forward leaning position of the sprinter's body, and speed development depends mainly on the powerful extension of all lower leg joints. Meanwhile, the trunk of the runner raises during the TP and evolves into a fully upright position. As a result, the mechanics of forward propulsion change until the end of the TP, with acceleration being determined by the swing back velocity of the support leg at touchdown (11). During the phase of MS, the trunk remains upright and the time required to rotate the legs forward and backward, relative to the hip joint, will limit further acceleration (5,19). From here onward, the sprinter aims to maintain speed and to postpone deceleration. The performance in each of these phases—start phase, IA, TP, and phase of maximum speed—therefore depends on specific technical skills that relate to the particular biomechanical demands of each individual phase. A sprinter can therefore be successful in some phases but less successful in other phases (5). Consequently, physical and technical training of sprint athletes should focus on the requirements of each of these phases of sprint performance. However, our knowledge on how speed is developed in each of these phases is still lacking.
During sprinting, speed is determined by both step length and step rate. An indepth analysis of the changes in both these factors can clarify the mechanisms contributing to speed development and, therefore, performance in each of the phases. Because step length correlates negatively to step rate (7,10), for a given speed, changes in step length affect step rate and vice versa. Step length is known to increase throughout the first 3 phases to reach a plateau during the phase of MS (7,12,14). However, the changes in step rate are less clear: Maximal step rate is reported either from the beginning (12) or between the 10- and 30-m mark (7) of the sprint. Currently, there is no consensus found in the literature as to whether step rate (4,11) or step length is the most critical factor for sprinting performance (2). Some authors suggest that both step length and step rate need to increase simultaneously (2,9,13–15). It is clear that both these factors rely on conflicting requirements: step rate is favorably influenced by low moments of inertia of the legs and, therefore, low masses and short limb length. In contrast, step length is positively influenced by explosive strength and, therefore, muscle mass (3,7,8,11,18).
Based on the evidence published to date, it is unclear how step rate and step length interact to determine sprint performance during the different phases of the sprint and which strategy is most successful in developing and maintaining running speed: should one emphasize step length by maximizing the forward impulse or is it better to emphasize step rate by diminishing contact time (CT) and inertia. In daily practise, coaches are confronted with these conflicting demands and question the relative importance of step rate and step length in the development of sprint performance of their athletes.
This study focuses on the role of step length and step rate in optimizing sprinting performance in high-level athletes throughout the different phases of a sprint. We included the parameters describing sprinting performance (IA, acceleration in the TP, MS, step length, and step rate) and performance determining factors (explosive strength of the lower limbs and leg length) in men and women. In this analysis, we will exploit the differences in sprinting performance between female and male athletes. We hypothesize that men will excel in all phases of a sprint run and that they will present larger step length, even when normalized for leg length, because of superior explosive strength of the leg extensors. Therefore, we hypothesize that flight times (FTs) and CTs are increased in the male compared to female sprinters resulting in longer step lengths.
However, we are unable to formulate a clear hypothesis on the differences in step rate between both groups during the different phases of the run, as the literature suggest that step rate depends on genetic factors and running technique (8). We, however, assume a negative relation between step rate and leg length, because inertia and, therefore, leg length affects step rate. Given this relation, we hypothesize that a constant relation between both parameters is maintained throughout the 3 phases of the sprint. Furthermore, we do not expect different interaction between step length and step rate in men compared with women.
Experimental Approach to the Problem
In this study, we analyze the sprint performance of high-level male (n = 10) and female (n = 10) athletes. All of them performed 2 all-out 60-m sprint trials. Force-time characteristics of the start action were recorded. Running speed and acceleration were measured using a laser to accurately quantify performance during IA, TP, and phase of MS. Step length and step rate were recorded for each step (Optojump). Step length was normalized for leg length. Explosive strength of the lower limb muscles was quantified using vertical jump performance. Statistical analysis was performed to characterize, in each of the phases, the interaction between step length and step rate on the one hand (independent variables), and the relationship with performance on the other (dependent variables). The impact of leg length and explosive power (independent variables) on all performance characteristics was further analyzed within and across both sexes.
The best performing male (n = 20) and female (n = 20) athletes of the 100-m sprint rankings (Flanders) of 2009 were invited to participate in this study. The tests were organized in collaboration with the National Athletic Federation. The data of the top 10 male and female performers during the tests were included for further analysis. Sixty-meter test times ranged between 7.00 and 7.26 seconds in men and 7.73 and 7.98 seconds in women. Their seasonal best times on 60 m ranged between 6.82 and 7.19 seconds for men and between 7.48 and 7.94 seconds for women. All participating athletes followed a double periodization to perform in both winter indoor and summer outdoor seasons. The tests were conducted in December, 6 weeks before the national championships, which were considered as the main target of the winter season. In the period preceding this test, the athletes spent a significant amount of time on sprint- and start-specific training.
The study was approved by the University's Human Ethics Committee in accordance with the declaration of Helsinki. All the participants gave written informed consent.
All data were collected in an indoor track and field hall (host of the European Indoor Championship in 2000) in late December when the athletes were in final preparation of the indoor season. After an individualized warming-up, all the sprinters performed 2 maximal 60-m sprints. During both sprints, start characteristics were obtained using an instrumented start block. Speed and acceleration parameters were recorded by means of an ULS laser device. Because only 40-m Optojump was available, step length and step rate data were obtained over the 2 sprints. The first sprint was used to collect step length and step rate data on the first 40 m (i.e., IA and TP) and the second sprint to collect data on the last 20 m of the sprint (i.e., phase of MS) (Figure 1). Starting commands, that triggered the data capture devices, were identical to those used in the competition. Anthropometric measurements of the athletes were documented based on 3 parameters: Body weight was measured by means of a weight scale, accurate up to 100 g (Seca, Birmingham, United Kingdom); body height, measured from the vertex to the ground and leg length, measured from the top of the trochanter to the top of the medial maleolus, were both measured by means of an anthropometer (GPM Swiss made, New Delhi, India); and finally, explosive strength of the leg extensors was deduced from the maximal jump height (h) during a countermovement jump calculated based on h = 1/8 × g × FT2 with the FT recorded on a contact mat (Schmerzal, Wuppertal, Germany) and g = 9.81 m·s−2.
Load cells (Feteris, Den Haag, The Netherlands), mounted on the back of each starting block recorded the horizontal force-time characteristics of each foot on the block during the entire start action at a sampling frequency of 1,000 Hz. Analog signals were amplified and after AD conversion were processed using in-house made software (FAST ULS). The following parameters were included for further analysis: (a) Start time (T-start, seconds) reflecting the take-off duration and determined from the onset of force generation on one of the blocks until the force signal drops to zero, (b) start speed (V-start) calculated from the sum of the impulses on both blocks divided by body mass, and (c) Start acceleration (A-start = V-start/T-start). All parameters were corrected for individual reaction time (6), that is, the time delay between the starting command and a force increase exceeding a sensitivity threshold set at 7.2 N. The intraclass correlation coefficients for T-start, V-start, and A-start were 0.88, 0.96, and 0.71 respectively, indicative of good reliability of the start data.
Running speed was recorded continuously during the whole running distance by means of the Universal Laser Sensor (ULS, Laser Technology Inc., Centennial, CO, USA). The device was positioned behind the starting line, and the laser was projected onto the lower back of the subject, the most stable segment during running. The ULS data were sampled at 4,000 Hz and measured the distance to the sprinters' body with an accuracy of 0.001 m. Raw data samples were averaged over 13 samples, therefore reducing the frequency to 307.7 Hz. Speed and acceleration were calculated by means of numerical differentiation of the position data using customized software tools. To remove fluctuations in speed present within 1 step, a moving average over 0.30 seconds, was applied. The following parameters were included for further analysis: (a) Speed (meters per second) and time (seconds) at every 5-m point, (b) IA, that is, acceleration over the first 10 m (A10 = V10/T10) and acceleration during TP, that is, between 10 and 30 m (A10–30 = [V30 − V10]/[T30 − T10]) were calculated, and (c) MS (Vmax), the exact time and distance at which the MS was attained were used to reflect the duration and distance covered during the acceleration phase. The intraclass correlation coefficients for A10, A10–30, and Vmax were 0.97, 0.87, and 0.98, respectively, which shows high reliability of the laser data.
Step length and step rate were recorded continuously during sprint running by means of the Optojump system (Microgate, Bolzano, Italy). This system is an optical measurement system consisting of 40 sender and 40 receiver bars, each 1 m in length. The bars were positioned along the opposite sides of the track. Each of the bars contains 32 infrared light emitting diodes, resulting in a system accuracy of 0.031 m at a sampling frequency of 1,000 Hz. Relevant differences reported in the manuscript exceed the system accuracy. During the first sprinting trial, we recorded the step characteristics over the initial 40 m, whereas the step characteristics between 40 and 60 m were recorded during the second trial. Based on the foot position over time, step length, step rate, CT, and FT are calculated. The intraclass correlation coefficients for averaged step length and step rate were all >0.95, which shows a high reliability of the Optojump data. Step length was normalized for leg length according to the formula of Alexander (1):
To characterize changes in step length and step rate during the phase of IA, step characteristics of the first 6 steps were analyzed for each step individually. During the TP and in the phase of MS, step characteristics were averaged between 10 and 30 m and between 40 and 60 m, respectively.
Changes in step length and step rate were analyzed with repeated measures analysis of variance (ANOVA) with the progress of step length/step rate within the different phases of the run and between the sexes as factors. When appropriate, post hoc analysis was performed using Tukey's tests. Paired t-tests were used to compare performance parameters between men and women. Finally, Pearson correlations were calculated between step length and step rate on the one hand and with the performance in the different phases of sprinting (A10, A10–30, Vmax) and with the underlying factors (explosive strength and leg length) on the other hand. Before using parametric tests, the assumption of normality using a Kolmogorov-Smirnov test and the assumptions of linearity using a bivariate regression analysis were verified. The data are expressed as means and SDs. Where relevant, the effect size (ES) was calculated. The intraclass correlation coefficient (model 2, type 1) was calculated to determine test-retest reliability of the parameters recorded with the instrumented starting block, the ULS laser device, and the Optojump system. Statistical analysis was performed using Statistica 9.0 software (Statsoft, Tulsa, OK, USA). A significance level of p ≤ 0.05 is used.
The performance characteristics of the participating athletes are shown in Table 1.
During the start action and throughout the run, acceleration and speed were significantly higher in men compared with women (Table 1). Although no difference in start time was found, start speed and start acceleration were, respectively, 13.8 and 16.2% higher in men (p = 0.01; p = 0.006). Likewise IA, acceleration in the TP and MS were higher in men compared with that in women: 6.5, 34.5, and 12.7%, respectively (p = 0.048; p < 0.001; p < 0.001). The length of the acceleration phase varied between 35.09 and 48.92 m in women compared with 41.64 and 59.64 m in men. Men reached their maximum speed, on average, at 50.75 m, which is 20.1% further compared with women (p = 0.002). Despite these differences in distance of acceleration, the duration of the acceleration did not differ significantly between both sexes (p = 0.121).
A repeated measures ANOVA showed no significant interaction effect within step rate over the different steps and phases and between sexes (F = 2.236; p = 0.069; ES = 0.048). No differences were found between men and women (F = 3.483; p = 0.078; ES = 0.423) (Table 2). Furthermore, step rate does not increase significantly over the different steps and phases (F = 2.236; p = 0.106; ES = 0.185). Step rate during the first step of IA already attained 95% of the step rate at MS (Figure 2A).
A similar analysis was performed for step length. Men generated longer step lengths compared with those generated by women (F = 11.437; p = 0.003; ES = 0.389). Step lengths increased significantly during the sprint (F = 154.4; p < 0.001; ES = 0.932; post hoc Tukey: p < 0.05). In the TP, athletes had already reached 90% of their maximal step length (p < 0.001). Maximum step length is attained during the phase of MS (Figure 2B). Although step length during the first and second steps did not differ significantly between men and women (p = 0.96), step length was 10.3 and 11.5% longer in men compared with those in women during the TP and the phase of MS, respectively (p = 0.003; ES = 0.932) (Table 2). In addition, the interaction effect of normalized step length over the different steps and phases and between sexes showed no significant effect (F = 0.976; p = 0.490; ES = 0.363). Step length normalized for leg length increased from IA over TP until the phase of MS (F = 133.9; p < 0.001; ES = 0.987). However, no higher normalized step lengths were found in men compared with those in women (F = 1.916; p = 0.183; ES = 0.096) (Figure 2C).
Flight times were identical for men and women, but male sprinters showed shorter CTs in the TP (p = 0.046) and in the phase of MS (p = 0.050). From IA to TP, FTs increased in both men (p = 0.007) and women (p = 0.004), whereas CTs decreased in men (p = 0.025) and women (p < 0.001). Only in male athletes, a further reduction in CT was found during the phase of MS (p < 0.001, Figures 3A, B).
Step characteristics were not correlated to performance in the IA, in the TP, nor in the phase of MS (Table 3).
A high negative correlation was found between step length and step rate during the phase of MS in men (r = −0.94) and women (r= −0.77). Only in women, this negative correlation was already present during the TP. In men, a negative correlation was found between step length and step rate in the IA phase (p = 0.054, Table 4).
Explosive strength correlated negatively to start time in men (r = −0.57). No correlations were found between explosive strength and either of the step characteristics. In contrast, a trend toward a positive correlation between explosive strength and step length normalized for leg length was found between the IA phase (r = 0.73; p = 0.086). A similar relation was suggested in the TP (r = 0.61; p = 0.063) and in the phase of maximum speed (r = 0.58; p = 0.079) in male athletes. However, no correlations were found between explosive strength and normalized step length in female athletes. No correlations were found between leg length on the one hand and step characteristics, FT, and CT on the other hand.
Running speed, acceleration, and sprint times over 60 m were analyzed in high-level male and female sprinters. Because only small differences in performance were seen between the test time and the seasons' best over 60 m of that winter indoor season, we can suggest that the interaction between step length and step rate will not differ significantly between these 2 conditions. Men achieved higher accelerations and higher speeds in all phases: start action, IA phase, TP, and phase of MS. However, the dominance of male athletes' performance is most pronounced in the TP, whereas only small differences between the sexes are found in IA.
Step rate did not differ between the different phases of sprint running in both male and female athletes: Sprinters are capable of developing, from the first step on, a step rate ≥95% of the step rate attained at MS. This finding is in accordance with the results of Ito et al. (12) and contrasts with the findings of Gajer who states that maximal step rate is only attained between 10 and 30 m (7) and Coh et al. who showed that step rate increases over the first 20 m (4,7,12). During the acceleration phase, the trunk of the runner changes from forward leaning to an upright position, therefore affecting the biomechanics of forward propulsion (5). Despite this positional change, step rate is not affected (19). However, CTs decrease and FTs increase. These findings are similar to the ones reported by Coh et al. and Bezodis and coworkers (4,17). Notwithstanding the constant step rate, step length increases consistently over the sprint in male and female athletes to achieve maximal step length when MS is attained.
In our data analysis, we exploited the differences in performance between high-level male and female athletes to gain insight into the contribution of step length and step rate on the one hand and performance determining factors on the other. Such a comparison is, to the best of our knowledge, nonexistent in the literature. Comparison between faster and slower sprinters is only found for male athletes. The results are mainly in agreement with those of other authors identifying no significant differences in step rate between the groups (12) although others identified differences in both step rate and step length (4,10). By comparing male and female athletes, this study is able to identify how step characteristics clearly differ between both sexes and how this relates to differences in performance determining parameters.
Comparing step characteristics between female and male athletes, step rate did not differ significantly. Although FTs are almost equal for both sexes, CTs are shorter in men during TP and MS. Although step rate does not change significantly throughout the sprint, the interaction between CT and FT differs between the sexes, with shorter CTs in men compared with those in women during TP and maximal running speed.
Step lengths are significantly longer in male athletes compared with those in female athletes from the third step onward, during the TP, and during the phase of MS. In the latter phases, step length is >10% longer in men compared with that in women. It is obvious that longer leg length is a factor predisposing to longer step length. When step length is normalized for leg length, during IA, the difference between the sexes is reduced to <3%. In contrast, at MS, step length remains notably longer (+7.21%) therefore strongly suggesting that, in this phase, leg length per se is not the only performance determining factor.
Male athletes run faster with longer step lengths, even when step length is expressed relative to leg length, than female athletes do despite similar step rate and even shorter CTs. The male athletes' similar step rate may be explained by the presence of superior explosive lower limb strength. This allows men to attain a similar step rate compared with women despite shorter CTs and increased inertia (because of higher body mass and longer leg length).
Correlation analysis showed that neither step length nor step rate relates to the sprint running performance (Table 3). This was confirmed for both sexes and for each phase of sprinting performance individually. This suggests that performance in high-level athletes is not exclusively determined by step rate or step length solely. Therefore, the interaction between both step characteristics needs further analysis. The inverse relation between both step components is most obvious in the phase of MS because for both sexes a negative correlation between step length and step rate is reported consistent with literature (4,7,10). As step rate is constant during the sprint and given the inverse relation between step length and step rate, we can conclude that during the phase of maximal running speed men and women invest in developing step length while keeping step rate at the same level. However, we notice that within this group of high-level male or female sprinters, maximum running speed is not related to specific differences in step length or step rate.
During IA, a negative relation between step rate and step length is only found in men. In this phase, the sprinter's body leans forward offering him the perfect position to transfer explosive strength of the lower limbs into forward acceleration. Given the higher explosive strength in male athletes, it was expected that they could exploit this to further excel over female athletes. In contrast, the differences in performance between men and women are minimal in the phase of IA: with start speed and start acceleration 13.8 and 16.2% higher and MS being 12.7% higher in men, the IA differs only by 6.5% between the sexes. Furthermore, no significant difference in normalized step length was found between men and women in this phase. Both these findings suggest that men do not take full advantage of their higher explosive strength to develop step length but invest during the first steps in developing optimal step rate even if this limits step lengths.
During the TP, however, men do not present this negative correlation (r = −0.19) between both step characteristics, whereas in women step rate and length are highly negatively correlated. This suggests that men can develop high step rates in the TP without a negative interference with step length. As a result, men have a 34.5% higher acceleration compared with women in that phase. From the TP onward, normalized step length remains 6.39% longer in men compared with women. This may indicate that during this phase, women lack the explosive strength to develop higher step length at an identical step rate. Because ours is the first study to report these parameters in a group of female and male high-level athletes, further research is however needed to confirm and clarify these differences.
This study aimed to better understand the interaction between step rate and step length in high-level sprinting performance in male and female athletes. The data of this study clearly indicate that success at this level of performance is not specifically related to the potential to develop long step length or high step rate. More specifically, we show that the interaction between both step characteristics differ between male and female athletes and between the 3 phases of sprint performance. Both male and female athletes present maximal step rate from the beginning of the sprint. Step length changes differ in male and female athletes during the sprint run. Men develop higher speeds on leaving the starting blocks but are unable to take full advantage of their explosive strength and leg length to develop longer step lengths during the phase of IA. However, during the TP, men are able to develop higher step lengths for an identical step rate, consequently they outscore the female athletes by 34.5% in acceleration during the TP. However, female athletes struggle with a negative interaction between both step characteristics during the TP and fail to increase their step length. In the phase of MS, the inverse relation between step length and step rate is most clear: although athletes can maintain a constant step rate, they fail to further increase step length.
Sprint coaches need to be aware of the multidimensional structure of sprint performance when designing a training program.
Although step rate remains constant over the 3 phases of sprinting (IA, TP, and MS), step length and step rate interact differently from phase to phase within and across sexes. At MS, there is no indication that one of both step characteristics solely discriminates performance. Therefore, the step length of the athlete at top speed can only increase within boundaries dictated by the step rate that the athlete already adopted during acceleration. Consequently, optimal training of maximal running speed requires an individualized approach aiming to optimize the ratio between step length and step rate.
The higher MSs of men compared with that of women result from a longer acceleration distance and a higher acceleration in the TP. The latter is most probably related to the higher explosive strength in men. This confirms strength and power training as crucial components of a sprint training program. However, the limited differences between male and female athletes during IA suggest that explosive strength is not a performance determining factor in this phase. Despite the longer CTs and the forward lean of the upper body during IA, athletes cannot take full advantage of their muscular strength and power to enhance performance during the first steps. An appropriate running technique is therefore essential to guarantee an optimal transfer of explosive strength into sprint acceleration and performance. Optimization of the individual athletes' performance can then be achieved through the integration of these insights with the expertise of the personal coach.
Sofie Debaere is funded by the Flemish Policy Research Centre for Culture, Youth and Sports, supported by the Flemish Government, Belgium. The results of this study do not constitute endorsement of the product by the authors or the National Strength and Conditioning Association. There is no disclosure of funding to report for this study.
1. Alexander RM. Terrestrial locomotion. In: Mechanics and Energetics of Animal Locomotion. Alexander R. M., Goldspink G., eds. London, United Kingdom: Chapman & Hall, 1977. pp. 168–174.
2. Armstrong L, Costill DL, Gehlsen G. Biomechanical comparison of university sprinters and marathon runners. Track Tech 87:2781–2782, 1984.
3. Babic V, Harasin D, Dizdar D. Relations of the variables of power and morphological characteristics to the kinematic indicators of maximal speed running. Kinesiology 39:28–39, 2007.
4. Coh M, Milanovic D, Kampmiller T. Morphological and kinematic characteristics of elite sprinters. Coll Antropol 25:605–610, 2001.
5. Delecluse C. Influence of strength training on sprint running performance. Sports Med 24:147–156, 1997.
6. Delecluse C, Diels R, Goris M. The F.A.S.T. project—A scientific follow-up of sprinting abilities. New Stud Athl 11:141–143, 1996.
7. Gajer B, Thénaut-Mathieu C, Lehénaff D. Evolution of stride and amplitude during course of 100m event in athletics. New Stud Athl 14:43–50, 1999.
8. Harland MJ, Steele JR. Biomechanics of the sprint start. Sports Med 23:11–20, 1997.
9. Hoyt DF, Wickler SJ, Cogger EA. Time of contact and step length: The effect of limb length, running speed, load carrying and incline. J Exp Biol 203:221–227, 2000.
10. Hunter JP, Marshall RN, McNair PJ. Interaction of step length and step rate
during sprint running. Med Sci Sports Exerc 36:261–271, 2004.
11. 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.
12. Ito A, Ishikawa M, Isolehto J, Komi PV. Changes in step width, step length, and step frequency of the world's top sprinters during the 100 metres. New Stud Athl 21:35–39, 2006.
13. Kunz H, Kaufmann DA. Biomechanical analysis of sprinting: Decathletes versus champions. Br J Sports Med 15:177–181, 1981.
14. Luhtanen P, Komi PV. Force-, power-, and elasticity-velocity relationships in walking, running, and jumping. Eur J Appl Physiol Occup Physiol 44:279–289, 1980.
15. Mann R. V, Kotmel J, Herman J, Johnson B, Schultz C. Kinematic trends in elite sprinters. In: Proceedings of the International Symposium of Biomechanics in Sports. Sports biomechanics. Terauds J., Barthels K., Kreighbaum E., Mann R., Crakes J.. Del Mar. California: Academic Publishers, pp. 17–33, 1984.
16. Ross AA, Leveritt M, Riek S. Neural influences on sprint running. Sports Med 31:409–425, 2001.
17. Salo A, Bezodis I, Batterham AM, Kerwin DG. Elite sprinting: Are athletes individually step frequency or step length reliant? Med Sci Sports Exerc 43:1055–1062, 2011.
18. Van Der Walt WH, Wyndham CH. An equation for prediction of energy expenditure of walking and running. J Appl Physiol 34:559–563, 1973.
19. Van Inghen Schenau GJ. Optimization of sprinting performance in running, cycling, and speed skating. Sports Med 17:259–275, 1994.