Acute effects of progressive sled loading on resisted sprint performance and kinematics

Original article written by Fernando Pareja-Blanco, Lucas A Pereira, Tomás T Freitas and colleagues

Background

It’s common knowledge among our profession that speed is one of (if not the) most important physical attribute, for team and court sport athletes. Naturally then, one of the most sought-after pieces of information is how we, as practitioners, can make our athletes faster. Whilst sprinting itself serves as a very useful strategy to improve how fast an athlete runs ^[1], great interest is also placed on other methods such as resisted and assisted sprinting. For resisted sprinting, weighted sled towing is a common method used by practitioners ^[2] with the addition of external load resulting in a greater demand for athletes to apply horizontal forces – also thought to be an important factor influencing acceleration performance ^[3].

Despite the interest in this training modality, there does not appear to be any over-arching consensus on optimal loads to be used (to the best of my knowledge). Some thoughts exist whereby heavier loads (e.g., > 20% of body mass) may transfer to greater improvements in acceleration (due to the increased demand for horizontal force application), whereas lighter loads (e.g., < 10% of body mass) may provide a more appropriate overload stimulus for maximal velocity running ^[4]. Further to this, the addition of load will also impact the orientation of the body and subsequently kinematics as well. For example, unsurprisingly, previous research has shown that resisted sprinting using loads of ~13% and ~32% of body mass, results in increased trunk lean and hip flexion ^[5], whilst concurrently reducing running velocity as expected. Other studies have also reported notable changes in sprint technique from resisted sprinting, such as: reductions in flight time and stride length, and increases in ground contact time ^[6,7]. Despite these expected changes in sprinting with the use of additional load, it doesn’t appear that enough supporting evidence exists for any certain recommendations to be made as to which loads are most advantageous to use in practice and was therefore, the primary aim of this study.

What the authors did

Ten rugby union players from the Brazilian National Team undertook the following testing on a single day: i) unloaded 20-m sprint test, and ii) 20-m resisted sprints in a randomized order, with loads of 20, 40, 60, and 80% of body mass. A standardized warm-up protocol was employed, whereby each athlete undertook: i) 10-minutes of jogging at a moderate pace, ii) 3-minutes of dynamic stretches, and iii) sub-maximal sprinting efforts (although the volume and perceived level of intensity of these was not described).

Four pairs of photocells were used, enabling sprint velocity from 0-5 m, 5-10 m and 10-20 m to be recorded. A total of 10 sprints were completed by each player, with 2 sprints completed at each load. From a measurement perspective for kinematics, the Noraxon MyoMotion motion analysis system was used, with 8 inertial sensors placed on the athlete’s shoe, tibia, thighs, pelvis and lower thorax, with kinematic data (in degrees) reported for the hip, knee and ankle joints and stride length. 

What the authors found

Changes in Velocity

There is quite a lot of information to unpack here, so I will start with changes in velocity relative to the unloaded sprint trial (0% body mass [BM]), with mean, standard deviation (SD) and % drop-offs in velocity, presented in the table below. As a brief summary, and unsurprisingly, any additional load resulted in a drop-off in velocity, with these reductions naturally increasing as more external load was pulled. When % drop-offs are viewed across the different distances of the sprint, it appears they are reasonably consistent for each load. For example, when using 20% of body mass, drop-offs in velocity range from 18.5-20.4%, across the three measured distances (0-5, 5-10 and 10-20 m).

Changes in Kinematics

When looking at changes in kinematics, data was presented graphically, so providing clear-cut mean and SD data is harder to do. However, the authors did report % changes for each load condition across the 4 measured variables (stride length, hip, knee and ankle kinematics). Given the volume of comparisons across three different sprint distances, I have presented this in sub-sections, purely for ease of understanding.

Stride Length (cm)

Reductions in stride length were evident across all distances and these reductions progressively increased as load increased. Consequently, I have provided a range of % drop-offs, noting that they followed a linear trend as external load increased. For the 0-5 m distance, reductions as a % relative to the unloaded trial ranged from: -14.8% (20 BM) to -35.2% (80 BM). For the 5-10 m distance, reductions ranged from -12.7% (20 BM) to -31.8% (80 BM). Finally, for the 10-20 m distance, reductions ranged from -11.3% (20 BM) to -45.0% (80 BM).

Hip Kinematics (°)

The amount of hip flexion also followed a linear trend, with the smallest % increases seen for the 20% body mass condition and largest for the heaviest sled loads used (80 BM). Thus, for the 0-5 m distance, % increases ranged from 16.6% (20 BM) to 36.6% (80 BM). For the 5-10 m distance, % increases ranged from 27.5% (20 BM) to 58.9% (80 BM). Finally, for the 10-20 m distance, % increased ranged from 33.6% (20 BM) to 76.3% (80 BM). Unlike changes in stride length where % changes were relatively consistent across all sprint distances, the lower and upper boundaries for hip kinematics clearly show that substantially larger changes were evident, as the latter part of the 20 m sprint trial was assessed. Also of note, individual variation for this metric was much greater than for stride length, which seemed to show much greater consistency across the group.

Knee Kinematics (°)

Interestingly, changes in knee flexion did not follow the same trend as the hip joint, with any % changes across loads being somewhat more subtle. Specifically, for the 0-5 m distance, % reductions ranged from -1.6% (20 and 40 BM) to -2.5% (80 BM). For the 5-10 m distance, % reductions ranged from -1.4% (20 BM) to -5.5% (80 BM). Finally, for the 10-20 m distance, % reductions ranged from -1.8% (20 BM) to -8.8% (80 BM).

Ankle Kinematics (°)

In total contrast to the knee joint, changes in dorsiflexion at the ankle were very large, relative to the unloaded sprint condition, but did again follow a linear trend for each distance. For 0-5 m, % increases ranged from 350.1% (20 BM) to 554.2% (80 BM). For the 5-10 m distance, % increases ranged from 65.9% (20 BM) to 165.7% (80 BM). Finally, for the 10-20 m distance, % increases ranged from 60.5% (20 BM) to 146.1% (80 BM). It’s important to point out why the % increases appear to be so much larger for this metric and the answer largely lies in the raw data.

For example, mean values typically ranged from -10° (0 BM) to approximately +5° (80 BM). Thus, when dealing with small raw numbers, and always comparing back to baseline, the relative % values become somewhat inflated when compared to the raw numbers for knee and hip flexion, where values ranged from approximately 30° to 70°. If this isn’t clear, let me briefly provide a simpler example. If I want to calculate the relative % difference in jump height from athlete A who jumps 30 cm on their left leg and 33 cm on their right leg, the equation is simply: (33-30)/33*100 = 9.1%. However, if I retain the same absolute difference (i.e., 3 cm), but athlete B’s jump height scores now become 20 cm and 23 cm, the relative % value changes to 13.0%. Thus, as the raw data (jump height) got smaller, the relative % difference increased, which is why we see smaller absolute numbers in ankle kinematics, produce much larger relative % increases when compared back to: i) the unloaded sprint condition, and ii) knee and hip kinematic variables.

As a concluding statement to the large volume of data associated with this study, findings showed that sprint kinematics are generally heavily disrupted when loads are ≥ 40% BM, whilst loads of 20% BM seem to be adequate for promoting improvements in sprint ability, without causing drastic changes in technique.

Limitations

There were no major limitations pointed out by the author team, so I have concluded a couple myself. Firstly, despite the sample being rugby union players of the highest standard in Brazil and the authors undertaking a detailed look at both sprint performance and technique, the sample size was very small (n = 10). The authors did a great job of providing individual data for kinematic variables; however, this wasn’t provided for sprint velocity, which may have added some additional further transparency in the data. Secondly, the present study provides data for a single time point only. Having previously acknowledged that some of the kinematic variables showed much greater within-group variation in scores compared to sprint velocity, a test-retest or repeated measures design would be a good idea for future research in this space, to determine how consistent some of the more variable kinematic metrics are.

What this means for practitioners

In light of the concluding statement to the ‘what did they find’ section, and based off the authors’ suggestions, it seems that using loads of 20% BM are appropriate if aiming to elicit positive adaptations for sprint speed, without causing large alterations in kinematics. Whilst heavier loads (≥ 40% BM) may be beneficial for amplifying horizontal force production relative to lighter loads (which may well be required for some athletes), practitioners may wish to consider that the concurrent alterations in sprint technique may have the potential to be counter-productive to sprint performance, if they become evident during unloaded sprinting as well.

Recommended resources

Live – Applying & monitoring resisted sprint training – Hakan Andersson, Jason Hettler and George Petrakos