Kinematic Alterations Due to Different Loading Schemes in Early Acceleration Sprint Performance From Starting Blocks

Abstract

Maulder, PS, Bradshaw, EJ, and Keogh, JWL. Kinematic alterations due to different loading schemes in early acceleration sprint performance from starting blocks. J Strength Cond Res 22(6): 1992-2002, 2008-The purpose of this study was to examine the changes to block start and early acceleration sprint kinematics with resisted sled towing. Ten male sprinters performed 12 sprints (four each of unresisted and approximately 10 and 20% body mass [BM]) for 10 m from a block start. Two-dimensional high-speed video footage (250 Hz) of the starting action and the first three steps of each sprint were recorded to enable the sagittal sprinting kinematic parameters to be obtained using APAS motion analysis software. The overall results of this study indicated that early acceleration sprint performance from starting blocks decreases with increasing load during resisted sled towing. A load of approximately 10% BM had no “negative” effect on sprint start technique or step kinematic variables measured in this study (with the exception of one variable) and was also within the “no greater than 10% decrease in speed” limits suggested by Jakalski. Towing a load of approximately 20% BM increased the time spent in the starting blocks and induced a more horizontal position during the push-off (drive) phase. The approximately 20% BM load also caused the sprinters to shorten their initial strides (length), which may have resulted from the decreased flight distances. Such results suggest that the kinematic changes produced by the 10% BM load may be more beneficial than those of the 20% BM load. Future training studies will, however, need to investigate how these acute changes in sprinting technique impact on long-term adaptations in sprinting performance from resisted sprinting.

Introduction

Superior execution of the start and early acceleration phase (e.g., 0-10 m) is critical for achieving a performance edge over the competition in the short sprint events of track and field (^2,9,10,22). A successful sprint start requires the development of large horizontal forces at a high rate while in the blocks (¹⁰), resulting in a swift movement toward the initial steps (^20,21). The primary goal of the initial steps is to generate a rapid horizontal sprint running velocity, which is the product of the length and frequency of the athlete's steps (^6,11,14). Faster athletes have been identified to employ higher step frequencies with similar step lengths to those of their slower counterparts (²⁴). Interestingly, it is believed that a greater training emphasis is placed on methods that attempt to increase step length (²⁵), possibly because step frequency is more difficult to improve than step length (³). Both step frequency and step length can be increased through greater muscular force outputs of the hip extensor, knee, and ankle musculature (²⁹); therefore, on- and off-track resistance training underpins the conditioning program of the competitive sprint athlete (⁵). Identifying training strategies that are appropriate for improving horizontal force production in the starting blocks, the length and especially the frequency of the steps, and, consequently, overall sprint performance, may assist coaches and physical trainers in the task of training sprinters.

Resisted sprinting is one of the most popular means of improving sprinting speed (^19,25,30) and is considered the most appropriate training technique to specifically improve the strength of the muscles that are fundamental to sprint performance (^27,30). Commonly used resistive apparatus are parachutes, tires, and weighted/resisted sleds (⁸), with research generally focusing more on the use of resisted sled towing (^19,25,29,30). It is known that resisted sled towing causes acute alterations in sprint kinematics of the early acceleration phase (^18,19,25). Resisted sled towing has been reported to acutely decrease stride frequency and stride length, whereas kinematic parameters such as stance time, trunk, and hip angles have been reported to acutely increase when performing sprints with a sled (^18,19,23). Although these kinematic alterations seem somewhat arbitrary in the short term, the suggested long-term benefits from including resisted sled towing are a faster start performance (^23,28), a faster acceleration-phase performance (^29,30), an increased step frequency (³⁰), an increased step length (⁴), an increase in muscular force output of the lower body (²⁷), and the development of specific recruitment patterns that target the fast-twitch muscle fibers used in sprinting (¹⁹).

An issue to consider when prescribing resisted sled loading is whether the individual's sprint technique is being positively or negatively altered when the athlete is sprinting under resistance. If the load is too light, an insufficient stimulus to enhance sprint performance may arise, but if the resistance is too great, it may interfere with running mechanics, resulting in a loss of specificity (²⁵). The amount of technique disruption that is acceptable when training a sprinter with resistance is not well understood. Jakalski (¹⁵), for example, has suggested that athletes should not be slowed down more than 10%, because of the changes in ground-contact dynamics. Further, Mouchbahani et al. (²³) have proposed that the resistance should cause an increase in power output by increasing neural stimulation, but it should not cause a change in the pattern of muscle activity.

It is still unclear which resistive loads are best to use during resisted sled towing. The level of resistance imposed on the sprinter can be expressed as either an absolute load (e.g., 5 kg) or relative to the individual's body mass (BM; e.g., 5% BM). Absolute loading schemes have been regularly employed in the coaching and physical training literature (^18,23,27,30). However, the limitation of prescribing absolute loads as a guideline for coaches is that the athlete's individual anthropometry (e.g., stature and mass), physical strength, and current sprint performance capabilities are not considered (²⁵). The effect of resisted sled towing with a 10-kg load may, for example, enhance the performance of one athlete but be detrimental for another. Guidelines that recommend loads relative to BM (e.g., 15% BM) may be more appropriate, because these loads can be more adequately generalized across athletes irrespective of their stature or mass. However, there seems to be no consensus in the literature about the optimal training load. For example, Mouchbahani et al. (²³) have suggested the use of a 5-10% BM load when performing resisted towing. Similar recommendations have been proposed by Murray et al. (²⁵), who advocate the use of 10% BM loads. Lockie et al. (¹⁹) have suggested that both approximately 13 and 32% BM loads may improve sprinting performance. The lighter load may enhance hip flexion during the leg drive phase, resulting in improvements to the length and rate of the steps but with minimal disruption to technique. The approximately 32% BM was suggested to be better for developing the upper-body action during accelerative sprinting. Although these guidelines offer some insight into loading schemes to employ for certain technical aspects of sprint running, they are still unclear about the appropriate magnitudes of the loads to impose and whether different loads may be more useful for altering specific parts of the sprinting motion. It was also unknown how the standard, anthropometric, and physiological capacities of the sprinters as well as sprint distance and starting position may influence this relationship.

Resisted sled towing has been promoted as a useful tool for improving sprint start performance (^23,28), but no studies have examined the effects of this training modality on sprint start kinematics. The identification of the kinematic alterations to sprint start technique that result from resisted sled loading, if any, would provide useful information on how this training tool may be used for improving sprint start performance. The purpose of this study was to examine the changes in sprint performance, block start, and early acceleration sprint kinematics with resisted sled loading in an attempt to gain more insight into the most appropriate loads to prescribe as resistance when training. It was hypothesized that resistive loads would significantly alter sprint performance measures and sprint start and early-acceleration-phase kinematic parameters (i.e., total block time, push-off angle, step length) in trained track sprinters. This study was designed to provide coaches, athletes, and trainers with information that will help ascertain the use of various resisted sled towing loads during sprint start training.

Methods

Experimental Approach to the Problem

In this study, we used a cross-sectional approach to examine the acute effects of two different resistive sled loading protocols on sprint start and early acceleration kinematics in male track sprinters. Towing loads of approximately 10 and 20% BM were selected based on those similarly employed in the literature and also on what was frequently being used by the coaches of the athletes recruited for this study (^19,25). Each athlete performed unresisted and resisted sprints conditions in random order in the same testing session. The sprint start and early-acceleration-phase kinematics were selected based on deterministic modeling of these particular events provided in the literature (^14,21,24).

Subjects

Ten male track sprinters at a national and regional competitive level (mean ± SD: age 20 ± 3 years; height 1.82 ± 0.06 m; mass 76.7 ± 7.9 kg; 100-m personal best: 10.87 ± 0.36 seconds [10.37-11.42 seconds]) participated in this study. At the time of testing, the fastest and slowest tested athlete were slower than the world-record 100-m time of 9.78 seconds by 6% (0.59 seconds) and 17% (1.64 seconds), respectively. The athletes were in the competition phase of their domestic season, with five of the athletes preparing to compete in some pre-Olympics 2004 events held in Europe and two athletes preparing for the 2004 World Junior Athletics Championships. Sprint starts (both resisted and body weight only) were a regular part of the athlete's training program at the time of testing. Each participant gave informed consent in writing to participate in this study before testing. Ethical approval was obtained for all testing procedures from the Auckland University of Technology ethics committee.

Procedures

Testing was conducted at an International Association of Athletics Federations-accredited athletic stadium with a Mondo track surface. Each athlete completed his own individual warm-up under the supervision of his coach. The athletes were then asked to perform 12 × 10-m sprints, comprising four sprints in each of the three experimental conditions. The three conditions were unresisted sprinting and resisted sprinting with approximately 10 and 20% BM. A metal sled weighing 7 kg was employed in this study. Weight plates were added to the mass of the sled to reach the appropriate testing load for each participant. However, the exact towing mass could not be allocated to some of the participants because of the increments in weight plates. Three participants towed a greater or lesser mass than that required for the approximately 10% BM condition. Two participants were ± 0.1 kg (± 0.1-0.2% BM); the third participant, who weighed less than 70 kg, had to tow a sled of 11.1% BM because of the mass of the sled. For the approximately 20% BM condition, two participants towed a greater mass than their equivalent of approximately 20% BM + 0.1 kg (0.1% BM). A nylon rope 30 m in length was used to connect the athlete to the sled via a waist harness. The rope length of 30 m was selected because it produced a relatively horizontal (within 1-2°) angle of pull and would minimize any bouncing of the sled during the sprints. The 30-m rope length also allowed a sufficient deceleration distance after the 10-m sprint so that the sled would not crash into the starting blocks. All experimental conditions were performed in block randomized order. The placement of the starting blocks was set according to the preference of the individual athlete. An experienced starter was used to provide standard starting commands to the athletes. The sprints were separated by a 2- to 3-minute rest period to ensure sufficient recovery. Athletes performed the sprints in tight-fitting clothing and track spike shoes. The two fastest trials for each of the three conditions per subject were averaged and used in the data analysis.

Data Collection

Figure 1 provides a schematic representation of the setup procedures used during the testing session. Time from the start signal to the 10-m line was collected using a SWIFT timing light system. The timing lights consisted of a dual-beam-modulated Visible RED light sensor/reflector setup collecting at 4 MHz ± 80 Hz. A microphone attached to a wooden start clapper was connected to the timing light handset. Timing was initiated when the appropriate sound threshold was broken. Because sprint running from a block start involves body movements that occur predominantly in the sagittal plane, a two-dimensional protocol was considered satisfactory for the present study. The set position, starting action (leaving the starting blocks), and initial acceleration (first two to three steps from the starting blocks) were filmed with two Fastcam PCI 1000 cameras operating at 250 Hz with a shutter speed of 1/500 seconds. The cameras were synchronized in the Photron Motion Tools software required to operate the two Fastcam PCI 1000 cameras. The cameras were placed perpendicular to the running direction, with overlapping fields, giving a sagittal view of the athlete for approximately three full running steps. Each field was approximately 3 m with a crossover of approximately 1.5 m or more. The first camera registered the set position, starting action, and one full step, and the second camera captured the movement of the athlete during the remaining two steps. Both cameras were positioned 13 m from the athlete and elevated to the athlete's approximate hip height of 1.1 m. Three marker strips were placed in the field of view so that one was visible in the overlapping view and toward the outer edge of each camera. These three markers ran across the lane with a strip placed parallel to the lane's long axis in the lane center. These markers were used to calculate the measures of horizontal displacement. A 1.7-m-tall rod fitted with a spirit level was filmed pre and post testing session at each of the three marker strips to enable the calculation of vertical displacement measures. The starter stood in a face-on position to the camera so that they were visible in the first camera's field of view. The start clapper held by the starter was held straight up and face-on to the camera so both sides of the clapper would be visible on the video footage when forced together to initiate the start response of the athlete.

Figure 1: Equipment setup used during data collection.

Data Analyses

High-speed video footage collected from both cameras was analyzed frame by frame to identify the x and y coordinates of the athlete's joints using a kinematic analysis system (Ariel Performance Analysis System). Digitizing began from the moment the starter's clapper closed until five frames post step 3 takeoff. Eighteen points of the body were digitized: apex of the head, seventh cervical vertebra, glenohumeral joints, elbows, wrists, third metacarpophalangeal joints, hips, knees, ankles, and distal ends of the feet (¹⁶). From these 18 points, human body segments were modeled. The segments included trunk (shoulder to hip), head, upper arms, forearms, hands, thighs, shanks, and feet. The data were smoothed using a digital low-pass filter with a cutoff frequency of 8 Hz for all x and y coordinates (²⁶).

In considering what kinematic variables were of interest in this study, a “deterministic model” of 10-m sprint performance (see Figure 2) was adapted from the work of Hunter et al. (¹⁴). The model is split into the block start and running phase, which are then divided into subcomponents such as start velocity, active block time, step length, and step frequency. The models then further specify the determinants of these subcomponents. It is important to note that all start and finish points of the movement phases (e.g., ground contact) and time measures were determined visually from the video footage. Additionally, all absolute angles were measured from the distal end of a segment going in a counterclockwise direction from the horizontal plane. The performance start time initiated by the “starter's signal” was the first visual frame where the two sides of the clapper could be seen fully joined together.

Figure 2: Determinants of 10-m sprint performance (adapted from Hunter et al. [14]).

The performance measures of interest in this study are defined below:

10-m sprint performance: the total time taken (as measured by the timing lights) from the starter's signal (clapper) to breaking the timing light beam at the 10-m finish line.

10-m sprint velocity: mean velocity of the athlete during the 10-m race distance using the time data from the timing lights.

Start velocity: the horizontal velocity of the body's center of gravity (CG) at the first frame the athlete was no longer fully in contact with the starting blocks.

Start acceleration: the horizontal acceleration of the body's CG at the first frame the athlete was no longer fully in contact with the starting blocks. This measure was calculated using the change in “start velocity” between the starter's signal and leaving the blocks divided by the change in time between the starter's signal and leaving the blocks.

The sprint start kinematic parameters during the block start phase calculated in this study were

First movement time: the time between the “starter's signal” and the moment of first noticeable movement as seen on the video footage (generally the head).

Active block time: the time between the frame with the first noticeable movement as seen on the video footage and first frame the athlete was no longer fully in contact with the starting blocks.

Total block time: the time between the “starter's signal” and first frame the athlete was no longer fully in contact with the starting blocks as seen on the video footage.

Block takeoff angle: the angle, measured relative to the horizontal, between the line passing through the most front part of the foot that was at the front of the starting blocks and the CG at the first frame the athlete was no longer fully in contact with the starting blocks as seen on the video footage.

Block takeoff trunk angle: the angle, measured relative to the horizontal, between the line passing through the hip and shoulder (trunk segment) of the side of the body in which the athlete's front foot was in the start blocks at the first frame the athlete was no longer fully in contact with the starting blocks as seen on the video footage.

Kinematic variables were determined for the first three steps from the starting blocks. Each step was split into two major phases consisting of a contact phase (ground contact) and flight phase (time in the air). The contact phase was defined as the moment of ground touchdown of the foot to the moment the same foot takes off (takeoff) from the ground. The flight phase was defined as the moment of takeoff of one foot to the moment of touchdown of the opposite foot. Touchdown was estimated as the first frame as seen on the video footage where the athlete's foot made contact with the track surface. Takeoff was estimated from the first frame where one foot had fully left the ground/starting blocks until the first frame where the opposite foot made contact with the track as seen on the video footage. The variables during each step calculated from the x and y coordinates in this study were

Step length: the horizontal distance between the point of touchdown of one foot to that of the following touchdown for the opposite foot (see Figure 3).

Figure 3: Sprint gait phases of interest (block takeoff, step touchdown, and takeoff) and a representation of step length, stance distance, and flight distance.

Step frequency: steps taken per second. Step frequency was calculated as the inverse of step time (1 / step time), where step time was estimated as the time between touchdown of one foot and touchdown of the opposite foot.

Contact time: the time of the contact phase.

Flight time: the time of the flight phase.

Contact distance: the horizontal distance the CG traveled during the contact phase (see Figure 3).

Flight distance: the horizontal distance the CG traveled during the flight phase (see Figure 3).

Touchdown trunk angle: the angle, measured relative to the horizontal, between the line passing through the hip and shoulder (trunk segment) of the touchdown leg.

Takeoff trunk angle: the angle, measured relative to the horizontal, between the line passing through the hip and shoulder (trunk segment) of the touch-off leg.

Takeoff angle: the angle, measured relative to the horizontal, between the line passing through the most front part of the contact foot and the CG during takeoff.

Statistical Analyses

Means and standard deviations were calculated for each of the dependent and independent measures for each condition. The two fastest trials for each experimental condition were averaged for an individual subject mean, with subject means then being averaged for each condition to provide a group mean. Within-subject reliability of the dependent measures for each test condition was evaluated using coefficients of variation (CVs). A repeated-measures analysis of variance was used to determine whether there was a significant effect of sprint load on the sprinting kinematics. A Bonferroni adjustment was used to investigate the significant within-subject effects, with the level of significance being chosen as p ≤ 0.05. Cohen effect sizes (d) were also calculated to quantify the magnitude of the differences between conditions. In accordance with the revised effect size magnitudes of Drinkwater et al. (⁷) for sport science research, effect sizes were defined as trivial (< 0.2), small (0.2-0.6), moderate (0.6-1.2), or large (> 1.2). All statistical procedures were performed using SPSS for Windows 12.0.

Results

Sprint Performance Measures

Sprint performance measures can be observed in Table 1. Resisted sprint towing had significantly large effects (ES = 1.71-5.34, p = 0.000-0.003) on all sprint performance measures assessed in this study. Sprint time for 10 m during both loaded conditions was significantly different from the unresisted condition. Ten-meter sprint time was longer by approximately 8% (0.16 seconds) for the approximately 10% BM loaded condition and 14% (0.28 seconds) for the approximately 20% BM loaded condition. Velocity for 10 m, start velocity, and start acceleration decreased significantly with increasing load. Start velocity dropped by about 8% (0.3 m·s⁻¹) for the approximately 10% BM loaded condition and 15% (0.5 m·s⁻¹) for the approximately 20% BM loaded condition. Start acceleration decreased to a greater extent than that of start velocity by about 14% (1.1 m·s⁻²) and 23% (1.8 m·s⁻²) as a result of the approximately 10 and 20% BM loaded conditions, respectively. Coefficients of variation ranging between 1.70 and 5.82% were calculated for all sprint performance measures for all test conditions excluding start acceleration, which had the greatest variability (7.60-10.62%).

Table 1: Mean ± SD sprint performance measures (N = 10).

Sprint Start Kinematics

Sprint start kinematic variables can be observed in Table 2. Resisted sprinting with a load of approximately 10% BM had insignificant trivial to moderate effects on all of the sprint start kinematic variables measured in this study (ES = −1.00 to 0.90, p = 0.101-1.000). However, a load of approximately 20% BM caused significant acute alterations to total block time (ES = 1.08, p = 0.030) and block takeoff angle (ES = −1.31, p = 0.013). Total block time increased by approximately 9% (0.04 seconds), whereas block takeoff angle decreased by approximately 9% (4°) during resisted sprinting with a load of approximately 20% BM. Coefficients of variation ranging between 3.90 and 12.16% were calculated for all sprint start kinematics for all test conditions, excluding block takeoff trunk angle, which had the greatest variability (7.60-10.62%).

Table 2: Mean ± SD sprint start kinematics (N = 10).

Step Length and Step Frequency

Step length from the starting blocks and for the first three steps showed no significant changes during resisted sprinting with a load of approximately 10% BM (see Table 3). Step length between steps 2 and 3 and between steps 3 and 4 significantly decreased by approximately 12% (0.15 m) and 11% (0.16 m), respectively, during resisted sprinting with a load of approximately 20% BM. Progressive step lengths were generally the same length as the subsequent step across all testing conditions (see Figure 4). However, the length between steps 2 and 3 was significantly longer for the unresisted condition (p = 0.006; 19%) and the approximately 20% BM loaded condition (p = 0.048; 18%) compared with the subsequent length between steps 1 and 2. Step frequency showed no significant changes for any of the steps during resisted sprinting with a load of approximately 10 or 20% BM (see Table 3). Coefficients of variation ranging between 2.88 and 16.18% were calculated for step length and step frequency for all steps for all test conditions. The 20% BM condition elicited greater variability for these step kinematics compared with the other test conditions.

Table 3: Mean ± SD step length and step frequency (N = 10).

Figure 4: Changes in step length for all steps across all testing conditions.

Contact and Flight Phase Kinematics

Contact and flight phase kinematics can be observed in Table 4. Loads of approximately 10 and 20% BM had no significant effects on contact time, flight time, or contact distance for any of the steps examined in this study (ES = −1.02 to 1.10, p = 0.053-1.000). The flight distance from the starting blocks to step 1 for trials with an approximately 10% BM load was significantly shorter by approximately 26% (0.06 m) compared with unresisted sprints. A load of approximately 20% BM significantly shortened the flight distance from the starting blocks to step 1, steps 2 to 3, and steps 3 to 4 by approximately 36, 27, and 27%, respectively, when compared with unresisted sprints. Coefficients of variation ranging between 7.93 and 20.56% were calculated for contact time and contact distance, whereas flight time and flight distance exhibited larger variability (12.44-48.31%) during all steps under all experimental conditions.

Table 4: Mean ± SD contact and flight phase kinematics (N = 10).

As the sprinters progressed through their sprinting steps, flight times and flight distances had generally the same durations as the subsequent steps across all testing conditions (see Figures 5 and 6). However, the flight time between steps 1 and 2 was significantly longer for the unresisted condition (p = 0.002; 40%) and the approximately 20% BM loaded condition (p = 0.011; 43%) compared with the subsequent flight time between leaving the starting blocks and step 1. The flight distance between steps 2 and 3 was significantly longer for the approximately 10% BM condition (p = 0.022; 71%) and the approximately 20% BM loaded condition (p = 0.048; 92%) compared with the subsequent flight distance between steps 1 and 2.

Figure 5: Changes in flight time for all steps across all testing conditions.

Figure 6: Changes in flight distance for all steps across all testing conditions.

Step Angular Kinematics

Sprinting with loads of approximately 10 and 20% BM had no significant effects on touchdown trunk angle, takeoff trunk angle, or takeoff angle for any of the steps examined in this study (see Table 5). Furthermore, there were no significant differences for any of these angular kinematics between the steps. Coefficients of variation ranging between 10.73 and 29.83% were calculated for all step angular kinematics for all test conditions excluding takeoff angle, which had the lowest variability (3.61-5.96%).

Table 5: Mean ± SD step angular kinematics (N = 10).

Discussion

The purpose of this study was to examine the changes to sprint performance, block start, and early acceleration sprint kinematics with resisted sled towing. It was hypothesized that the resisted sled loads would alter sprint performance and sprint kinematics. According to the results of our study, our hypothesis was somewhat supported in that sprint performance measures, start kinematics, and a variety of acceleration kinematics were significantly altered by using resisted sled loads of 10 and 20% of BM.

The results of this study indicate that sprint time and sprint velocity for 10 m became slower with an increase in load towed. Similar results have been reported for a variety of athletes over a variety of distances (^18,19,25). Jakalski (¹⁵) has suggested that running speed should not be slowed down more than 10%, because greater resistances may result in changes to running technique and, consequently, reduce training specificity. In the present study, a load of approximately 10% BM decreased 10-m velocity by 8%. A load of approximately 20% BM was revealed to slow the athlete down by 15% across a distance of 10 m, which suggests that this load may not be appropriate to train with according to Jakalski (¹⁵).

Sprint start performance measures (start velocity and start acceleration) were also slowed down with the addition of resistance. The athlete's start velocity while towing a load of approximately 10% BM was the only start variable to fall within the < 10% decrease guidelines of Jakalski (¹⁵). This study revealed a load of approximately 20% BM to lead to a 15% slower start velocity compared with unresisted sprinting. It may be inappropriate to use Jakalski's (¹⁵) 10% decrease guide when focusing on the block start “only,” because this study found no significant differences in any of the sprint start kinematics measured while towing a load of approximately 10% BM. This possibly suggests that a load of approximately 10% BM may not provide a sufficient training stimulus for improving the block start if the goal of training is to “overload” a key determinant of sprint start performance (e.g., active block time, total block time, block takeoff angle). Towing a load of approximately 20% BM did, however, acutely alter total block time (longer period of time spent in blocks) and block takeoff angle (more horizontal push off the blocks). Taking into consideration that the aim of the block start is to activate the correct sequence of muscular activation so that maximal horizontal force production occurs while leaving the blocks in the shortest possible time (^10,12), it seems somewhat conflicting that a load of approximately 20% BM, which causes an athlete to spend more time in the blocks, would be of benefit. Spending more time in the blocks may be beneficial for training the athlete to produce a larger amount of force or impulse (compared with a start with no resistance)-both qualities that have been identified as critical success factors to start performance (^10,21). However, this is only speculation at this stage, because no kinetics were measured during the block start in this study to confirm these thoughts.

When coaching the sprint start, there is a technical emphasis on leaving the starting blocks in a more horizontal position. Resisted sled towing with an approximately 20% BM load caused the athletes to adopt a more (9%) horizontal push-off or drive angle out of the starting blocks than when unresisted, possibly because of the greater inertia restricting the ability of the athlete to move vertically. Hoster and May (¹³) have stated that the drive angle during block takeoff should be as low (horizontal) as possible. If the angle of takeoff is shifted closer to the horizontal, it is likely that an increase in step length would occur providing the takeoff velocity remains the same as a greater proportion of this resultant velocity is horizontal not vertical. Increases in the length of the first steps out of the starting blocks have been advocated as part of an optimal start (¹⁷). The findings of the current study suggest that a load of approximately 20% BM or possibly more may be appropriate to use during resisted sled training to increase the horizontal drive out of the blocks, especially for athletes who propel themselves in more of a vertical direction out of the starting blocks. However, whether this “better” acute horizontal position with resistance will lead to a more horizontal drive out of the blocks in the long term is unknown.

Once the athlete has left the starting blocks, the athlete attempts to increase his or her step length and step frequency to complete the sprint distance in the shortest time possible. Hence, a great deal of coaching emphasis is placed on improving step length and/or step frequency under the contention that improving one of these factors will lead to faster sprint running performance. In the current study, resisted sled towing led to shorter steps being performed at a step frequency similar to that employed during unresisted sprinting. The slower sprint times under resisted conditions discussed earlier were possibly attributable to the decreased step lengths because no significant differences were found across any of the loaded conditions for step frequency. This finding supports those of Murray et al. (²⁵), but it is in contrast to the findings of other authors (^18,19). Reductions in step frequency of approximately 2-6% have been reported in male field sport athletes and female track sprinters (^18,19) when performing resisted sprinting. Perhaps male track sprinters (as used in this study) are not predisposed to reductions in step frequency while performing resisted sled towing, especially from a block start.

Shorter step lengths occurred during the approximately 20% BM condition, with no significant differences in any of the first four step lengths being revealed while towing a load of approximately 10% BM. The latter step lengths (steps 2-3 [12%] and steps 3-4 [11%]) from the starting blocks seemed to decrease significantly more with the approximately 20% BM load compared with sprinting with no resistance. A longer step length during unresisted sprints is a suggested training effect of resisted sled towing (^3,4). The current study has revealed that a “heavy” load caused shorter step lengths during resisted sled towing, which is in accordance with the findings of other studies (^8,18,19,25). However, it is still unclear whether repeated use of resisted sprinting will lead to greater step length during resisted sprinting and whether such positive changes will also occur in normal unresisted sprinting.

Contact time and flight time are considered the main determinants of step frequency (¹⁴). Coincidentally, it was no surprise that contact time and flight time were revealed to show no significant alterations during resisted sled towing, because no significant change occurred for step frequency in this study. Longer contact times and shorter flight times have been reported for resisted sprints (^18,19,23). The fact that contact time was not significantly changed in this study may be a result of the sprinters having a sufficient strength base to produce sufficient propulsive force under the loaded conditions without impacting on sprint technique. However, this is speculative because no strength measures or propulsive forces were measured in this study to support the former statement.

The flight distance (of the CG) from the starting blocks to step 1 was revealed to be significantly shorter for loads of approximately 10 and 20% BM by approximately 26 and 36%, respectively, in the present study. Furthermore, flight distances shortened by approximately 27% from steps 2 to 3 and steps 3 to 4, respectively, when compared with unresisted sprints. Flight distance has been suggested to be a key determinant of step length (¹⁴). Therefore, it is not surprising that step length was reduced with load, which is possibly synonymous with the reductions in flight distance, especially for steps 2 to 3 under the approximately 20% BM condition. Interestingly the flight distance between steps 2 and 3 was significantly longer for the approximately 10% BM condition (71%) and the approximately 20% BM loaded condition (92%) compared with the subsequent flight distance between steps 1 and 2 (see Figure 6). If an athlete's flight distances between steps are relatively similar as the athlete progresses through the initial steps, the results of this study suggest that resisted sled towing may be beneficial for improving flight distance, particularly from steps 2 to 3, which may lead eventually to a longer step length. Whether this increase in flight distance will benefit acceleration from the starting blocks with training requires further research.

Resisted sled towing had no significant effects on trunk angle at either touchdown or takeoff in this study, which is in contrast to the results of Lockie et al. (¹⁹). The trunk angle values reported by Lockie et al. (¹⁹) indicate that their test athletes were in more of a vertical position throughout the initial steps, with increases in load significantly causing the athletes to adopt a more horizontal position compared with the athletes in this study. This may have been a result of the standing start used by the athletes of Lockie et al. (¹⁹) because the trunk angles presented in this paper are similar to the trunk angles reported by Atwater (¹) for male sprinters who started from blocks. It may be possible that resisted sled towing causes trunk angles to be significantly closer to horizontal during sprints performed from a standing start but that it allows trunk angles to remain relatively unchanged during sprints from a block start because the trunk angles are already reasonably close to horizontal. Whether this statement is true requires future research that examines the effects of resisted sled towing on kinematic parameters (e.g., trunk angle) during a variety of start positions (e.g., standing stationary start, block start, running start).

Practical Applications

The overall results of this study have shown that early acceleration sprint performance from starting blocks, whether it be 10-m sprint time or block start velocity, decreases with increasing load during resisted sled towing. A load of approximately 10% BM had no significant effect on sprint start technique or step kinematic variables measured in this study (with the exception of flight distance from the blocks to step 1). Towing the approximately 10% BM load was also within the “no greater than 10% decrease in speed” limits suggested by Jakalski (¹⁵), whereas a load of approximately 20% BM was not. Resisted sled towing with an approximately 20% BM did cause the male sprinters to stay in the starting blocks longer and drive out of the starting blocks in a more horizontal position, which may be beneficial to block start performance in the long term. The approximately 20% BM load caused reductions in step length in the latter steps tested, which might have resulted from the decreased flight distances in the corresponding steps. Coaches of track sprinters should consider resisted sled towing with a load of approximately 10% BM as a training tool for improving their athlete's sprint start and early acceleration performance because it was shown in this study to cause no significant changes to start and running technique while possibly allowing a training effect to take place. However, there may be some benefits from training with the approximately 20% BM load, despite the significant “negative” changes in technique it causes. There is still a great need for a better understanding on the effects of resisted sled towing, particularly the effects this type of training has on the kinetics of sprint running. Most importantly, studies are required that examine the long-term effects of resisted sled towing on sprint performance and technique. Furthermore, future study is needed to justify the recommendations of Jakalski (¹⁵) examining whether training with a load that causes an acute decrease in speed by less than 10% is more beneficial than training with a load that will decrease speed by greater than 10%.

Acknowledgments

The authors would like to thank Dr. Joe Hunter, Mr. Jamie Denton, and Mr. Mike Smith for their assistance with data collection.