Sprint acceleration mechanics and performance: the interest of very-heavy sled training

Co-written with Matt R Cross

This post will describe the reasoning behind our general interest, and the eventual use of very heavy sled (VHS) sprint training in our recent study. This project (and all the associated references) was designed following our review of literature on the topic of sled training for sprint performance, and the ideas of two Master students I’ve had the pleasure of supervising this year: Matt Cross (Auckland University of Technology) and Satya Vesseron (University of Nice). In the interest of maintaining ease of reading I won’t list all the references used to support my statements - however they are all available in the papers mentioned above for those of you who want to dig deeper into the science behind VHS.

Two blog posts by George Petrakos (Glasgow Warriors) have tackled the issue pretty well:

https://www.freelapusa.com/resisted-sled-sprint-training-part-1-methods-of-sled-load-prescription/

https://www.freelapusa.com/programming-for-resisted-sled-sprint-training/

As you can read in these posts and the present one, by VHS we mean sled loading of more than 30% of body mass (BM), up to loads around 100-120%BM… you should give it a try, just for fun. By the way, even if we should express resistance as a friction force (see the recent work of Matt Cross here), for more clarity and comparison to the extisting literature, we’ll express “load” as the total mass pulled by the athlete (sled mass + additional mass).

Why does VHS training ‘make sense’, mechanically and practically speaking?

Even if my personal background and passion is athletics and sports training, I think that one method of innovation in this field is not to design hypotheses and studies according to what is or has been done in the field. Even if very successful coaches have brought some innovative training methods and still do, my point as a researcher is to start from the theoretical side of things, and ask the question: “what does make sense, mechanically speaking?” Regardless of whether or not this is what is actually done in practice, including at very high levels.

Innovation is not doing what has always been done better, it is doing something else. Something that makes more sense, rather than simply following tradition.

The traditional way of using sleds (at least in the scientific literature and for most coaches especially in athletics) is to keep the load comparatively light (typically 10-15%BM, or ~10% decrement in maximum velocity) so that the maximal running speed reached at the end of the acceleration is not altered too much, and the overall running pattern is maintained. The associated argument is something like: “If sled training induces a slow running speed and modifies the running pattern and technique, it is therefore non-specific; and non-specific is not good”. By such a logic, only free sprinting should be used to train for sprint performance, and consequently any other exercise (e.g. gym- or weightlifting-based) should be judged as non-specific and thus counterproductive. My point here is that there should be a separation between specific and non-specific exercises, based on the training goal. That is, we consider VHS as a strength and conditioning exercise, not a sprint-specific exercise.

The most interesting part of this project is how it all started. A few weeks after publication of our review in Sports Medicine - which showed that almost no studies had tested the effect of VHS training - a Master student approaches me and says: “Prof., I am a soccer coach, and I think VHS could be a very good training stimulus for my amateur players”. Wow. Except for my lessons about force-velocity profiling in sprinting, this student had not read the above-mentioned blog posts or our review. Before giving him my opinion, I asked him: “why do you think so?”. He answered that basically, his players had no access to a gym to work on their lower limb strength, and no time for specific training sessions to work on transferring that strength into a better sprint acceleration technique and performance. His main argument was that VHS is both a cost- and time-effective way of overloading the athlete, training both lower limb strength (i.e. general capacity) and the technical ability to apply this force effectively into the ground (i.e. horizontally-oriented force). After all, the only tools required are a solid sled and harness, 100kg of weightlifting plates, and access to a field for sprinting.

At the same time, another Masters student I was supervising (Matt Cross from Auckland) was working on a very similar topic, and his preliminary findings were showing that much heavier sprints than currently recommended showed promise for the development of maximal power. The discussions with these two students led to the following biomechanical arguments in favor of VHS, regardless of the traditional thoughts of most coaches:

Argument 1: When sprinting maximally against a VHS you still move damn slowly!

Haha nice one, my 7-yr old daughter knows it, no PhD needed here. But the point here is not that you run slow, it’s that because you run slow, your horizontal force output is much higher than when you run faster (according to the force-velocity curve, applied to sprinting). For full details on this F-V thing, please check our recent synthesis. So VHS, more than light sleds and much more than unloaded sprinting, is theoretically a way of targeting development of the “force” side of your horizontal F-V profile. That’s how human muscles work, too much speed = not much force, and vice versa. In fact, our understanding of how much load is required to overload these capacities is not very well understood – as we’ll discuss later in this post. So if my F-V profile shows low maximal force output (F0), light sleds or unloaded sprinting is likely a poor way of developing these capacities. Furthermore, it makes sense to expect that the higher the load, the higher the force output during acceleration, as shown in loaded jumping, cycle sprinting, or treadmill sprinting.

Argument 2: Sprinting with a VHS allows you to incline your body and push forward much more, compared with lighter sleds or during unresisted sprinting.

The higher the resistance, the greater the possible forward orientation, and vice versa. If we assume a good relationship between the overall incline of the body and the orientation of the resultant ground reaction force (GRF) vector, VHS is a very useful way of providing this incline. As an example, in the picture below the athlete is pulling a load equivalent to ~120%BM. There is no way he could apply force onto the ground with such a horizontally-oriented angle of propulsion with lighter loads.

Athlete pulling a load equal to 120% of his own body mass...definitely forward-oriented push

Argument 3: hip extensors work

Although no scientific evidence has been brought to our knowledge, our practice, anecdotal evidence, and athlete’s comments tend to show that hip extensors, mainly glutes, are particularly stimulated by the VHS modality. It makes sense mechanically (more crouched position, higher load against lower limb extension). Almost systematically, subjects who started to familiarize with VHS reported muscle pain after their first training sessions (glutes especially, hamstring to a lesser extent). In the video below, a PhD student I’m supervising performed a full session of VHS (about 10 sprints of 20-m in total, using loads of 40 to 100%BM). Our phone conversation 2 days after this session was fun: Although he is practicing weightlifting and resistance training, he suffered major DOMS at the glutes…

Argument 4: When sprinting with a VHS you push slowly (thus hard) and forward for a longer period.

This is the main argument behind VHS in my opinion, and combines the previous 3 points: Sprinting using VHS allows you to create and maintain conditions of high force, high forward lean, and high muscular activity (anecdotal), only seen for single instances during unloaded sprinting, for an extended period of time. As seen in the figure below, during an unresisted sprint acceleration, maximal power is typically reached within the first 2 seconds (for a single footstrike), and everything that happens afterward is a low-force/low-power/high-speed exercise, in the horizontal direction. In addition, during unresisted sprinting and with a lighter load (or a parachute), the body cannot stay forward inclined for a long time so you are forced to get vertical pretty quickly. As you can see in the video above, in a VHS context, you can apply more horizontally-oriented GRF, for a much longer period/distance. The result is the accumulated cumulated overload per sprint is much higher than using unresisted or lightly loaded sprinting exercise.

Our computation of Usain Bolt's velocity, horizontal force and power outputs during the World Record acceleration phase

Why would highly cited previous papers and “training recommendations” suggest it doesn’t make sense?

VHS training is not popular in athletics, compared to light sled training. I’ve heard some coaches use very heavy resistances during pulling exercises, in rugby for instance, but sled training has hitherto mostly been performed with light loads (e.g. 10 to 20%BM at most). Some coaches even talk about heavy loads when the total mass pulled is around 20kg, i.e. about 25% of the athlete’s body mass. I guess one of the argument is that when using heavier loads during this “sprint” exercise, the running speed is too much altered, and one may not consider it a sprint exercise anymore. As far as we can tell, the argument in the sports science academic literature originated with a couple ofstudies in the early to late 2000’s that tested changes in kinematics with resisted sprinting loads and recommended that “The lighter load is likely best for use in a training program“. Collectively, their results showed that the use of loads that induce >10% decrease in running speed (approximately), significantly altered the sprint running pattern (mainly joint kinematics). Consequently, these papers suggested the ‘optimal’ loading for training was one that did not change acute running technique, for fear that this would likely result in long term negative adaptations in these ‘technical’ factors, and decrease performance. On this basis, the result was a recommendation against heavy loads. For instance, a 2008 study using loads that induced a ~10% decrease in maximal velocity reports: “The athlete should use a high load so as to experience a large training stimulus, but not so high that the device induces substantial changes in sprinting technique“. According to the citation record, these articles appear to have steered the direction of research and practice, with researchers following suit with these recommendations and generally steering clear of VHS, or using tentatively ‘heavy loads’ of up to 42.6% BM – unsurprisingly, those sticking to the ‘guidelines’ and selecting loading protocols ‘non-significantly different’ to unloaded sprinting have typically found no differences between resisted sprinting and free sprinting training intervention outcomes...

Our main point here is that VHS should not be seen as an exercise ‘specific’ to unloaded sprinting, but instead it should be seen as a horizontal strength or power based exercise that may result in beneficial adaptations in determinant factors of sprint performance. The points that follow were key in our approach towards experimentally testing the hypothesis that VHS would be an effective training stimulus for sprint-specific, horizontally-oriented force production.

Turning points…

The first trigger for this project was our publications showing the importance of a forward-oriented GRF (mechanical effectiveness) for sprint acceleration performance. We showed the importance of this mechanical effectiveness in both low and elite level athletes in two cross-sectional studies, but did not perform associated training studies to answer the basic question: How can athletes develop their mechanical effectiveness and horizontal force production for an improved acceleration performance?

The second trigger was the publication of two timely studies by Naoki Kawamori and his colleagues. The first one showed that, during the second step of a sprint acceleration, pulling a sled load of 30%BM (higher load than in previous studies) induced a significantly higher ratio of force (the index of mechanical effectiveness, computed as the ratio between the horizontal component of the GRF and the resultant GRF over the step, see picture). The ratio of force was 28% on average in the control (unresisted sprinting) condition versus 31% in the “classical” 10% BM sled condition and 39% in the 30% BM condition… clearly a more effective ground push. In addition, no difference in sprint kinetics were observed between the 10% BM condition and the control condition, suggesting (as highlighted above) an insufficient mechanical overload.

Kawamori et al.’s second paper was a training intervention, that compared two groups training (8 weeks, 2 sessions/week) with sled loads that induced 10% (light load, about 13%BM for these subjects) versus 30% (heavy load, about 43%BM) decrease in maximal sprint velocity. Although mechanical effectiveness was not measured in this study, the conclusions were clear: “Heavy- and light-load weighted sled towing were equally effective in improving sprint acceleration ability over 10 m, but only the use of heavy load improved 5-m sprint performance. Therefore, it is conclusive that coaches and athletes should abandon the myth regarding the optimal training load of weighted sled towing (i.e., 10% rule) and should explore the use of heavier external resistance for weighted sled towing.“

Our review of literature published in 2015 by George Petrakos overall concluded that more studies were needed to clarify the effects of sled training (especially heavy: 20-30%BM and very heavy: >30%BM loads) on sprint acceleration performance. Note that we are talking about performance (sprint time/speed) here, not sprint technique or running pattern. My opinion is that what gives you an advantage in most sports (soccer, rugby, etc.) is being fast, not having a “good” technique. By the way, what is ‘good technique’? A technique that makes you fast, first and foremost. So if the training intervention makes you faster, it is effective.

The last trigger was the Masters work of Matt Cross, which aimed to assess the mechanics underlying resisted sled sprinting. The work showed that not only did sled sprinting exist on the same Fv spectrum as unloaded sprinting (confirming our initial ‘Argument 4’), but the loads that maximized horizontal power output were ~70-100%BM. Note that based on these results, some athletes may need to go higher than ~100%BM to work in the ‘force end’ of the Fv spectrum! Obviously these loads are much greater than those previously studied. In fact, while not published in the thesis, our calculations suggest that sprinting using this ‘optimal loading’ scheme presents an acute response in power output ~3x greater than traditional ~10%BM loading.

So based on all these points, and according to our theoretical framework around the sprint force-velocity-power profiling, we hypothesized that VHS training would mainly result in improved mechanical effectiveness and thus maximal horizontal force and power outputs. In our view of individualized sprint training, VHS could be an effective way to train that specific part of the profile, for athletes whose individual profile shows a deficit in this particular part of the sprint.

"VHS" means Very Heavy Sled...

What we did, and what our study says

Briefly (all details in the paper), 20 amateur soccer players were assigned to a control group and a VHS group, after familiarization with heavy loaded sled sprinting. Subsequently, each group performed 2 sessions of ten 20-m sprints for 8 weeks. The control group performed unresisted sprints, while the VHS group performed sprints against sled loads of 80%BM (so about 60kg on average). Except for the 4 players in the control group who could not perform the study or the post-testing (personal and professional issues), all players completed the program safely. The results showed that overall, VHS was an effective stimulus for improving mechanical effectiveness (i.e. more horizontally-oriented GRF during the early phase of acceleration post-training) and in turn the maximal horizontal force and power outputs. In addition, the increase in sprint performance tended to be higher in the VHS group, especially at 5m. The improvements in sprint performance for the control group were negligible.

The results from this pilot study basically confirm the theoretical arguments listed above, and highlight the interest of VHS training in this specific context.

For a cool video summary of this study, see below

What the study does not say: limitations and future studies

Here is a list of arguments for those who focus more on what studies did not do/say, unfortunately too many people out there.

Mainly, this pilot study needs confirmation and replication with other populations (higher level athletes in soccer or other sports). Our subjects were amateur soccer players, so we don’t know if this method will be effective and useful in other populations. There were not many subjects involved, and we would hope for more statistical power in future studies on the topic. The load, although heavy, was not great enough to work on the maximal force area of the F-V curve, as Matt Cross’ work shows. Rather, these athlete were working more in the zone of maximal power. Nevertheless, this provided a sufficient overload to improve force output in these subjects. Other force dominant cohorts, highly trained rugby players for instance, will likely require much greater loading schemes (100%BM or more?). These subjects were not trained for resistance or weightlifting, so they might have been good responders to the VHS stimulus… well that was exactly our point! As highlighted early in this post, part of the value of VHS is it provides a good mix between strength and mechanical effectiveness stimuli.

What about the “10% rule” and the alteration of the running “technique”

This study clearly shows the interest of VHS training to improve mechanical effectiveness and horizontal force output during the early acceleration phase. Although we did not directly compare our results with those of 10%BM loads, our conclusion based on the existing literature is that this 10%BM load is likely not heavy enough to induce such improvements. So you may maintain unchanged sprinting technique when pulling sleds with <10%BM, but you may also not improve your mechanical effectiveness and horizontal force output; hence, your performance is also unchanged. So if your running technique is acutely altered during the VHS sessions, but your running mechanics longitudinally improve following training, the net advantage in terms of mechanical effectiveness and performance is pretty clear.

Note that we also believe that training with light loads (eg 10%BM) may be effective, but to develop other parts of the sprint mechanical profile. In particular, we think that it may be useful to train the ability to produce high amounts of horizontal GRF at higher running speeds, which corresponds to the right part of the sprint force-velocity profile. We have several studies that will tackle these areas – stay posted!

Conclusion

People who see the topic as “light loads are useful to improve sprint mechanics and performance” versus “very heavy loads are better than light loads” are missing the point, in our opinion. This is a shortcut, and an over simplified ‘black and white’ view of things. The best answer (sorry for that…) to the question “is loaded sled training useful, and with what load” is: IT DEPENDS.

It all depends on the goal of training (e.g. the sport, aim of periodization), the athlete’s individual mechanical profile and the specific weaknesses. If maximal horizontal force/power output (and especially the ratio of force) is to be developed, then YES, VHS is useful, and light loads are likely not or less effective for your immediate goals. If the needs of the athlete are on the force output at higher speeds, then YES, lighter loads may be interesting and useful. Our own practice shows that within a team/group context, some athletes will clearly show different sprint mechanical profiles, different levels of effectiveness, and so different training needs. So the solution is not “sleds or not sleds” nor “VHS or not VHS”, it all depends on what you want to improve. Sprint performance is not just “being slow vs being fast”.

Determine the important factors for your own sporting context, each athlete’s mechanical sprint profile, compare athletes, monitor over time and then decide on an individual basis (who needs what) the best way to target each athlete’s needs.

Very-Heavy Sled training, Santa approved…

Optimal FV profile in sprint swimmers: more than just another marginal gain?

Unless you were too busy chasing Pokemons, you’ve noticed that the Men’s Olympic title on the 50-m freestyle was a 0.01s story. Relatively speaking, for a 21.40s effort, this is a 0.05% difference. Put that into another metric, the difference between gold and silver was about one inch, about the distance between your middle and little fingers.

  

  

We all know elite sport is a world of “marginal gains”, and among the many possible ways French swimmer Florent Manaudou and his staff could have chased this 0.05% marginal gain, my point here is that the starting block push performance could have been improved using the optimal force-velocity approach we’ve developed over the last years (see here for a reminder). Of course, many other factors could have been considered. One is having longer fingers of course. But hey, I have a hammer so sorry, I see nails everywhere…

For this story (any resemblance to reality is pure coincidence) I will take my own personal FV profile and imagine I’m an Olympic Games finalist chasing for a 0.05% margin in the Final. Note that my points here are based on scientifically validated methods and research-based evidence, not on my coach (or my own) « gut feeling and experience ».

Sprint start push-off performance in swimming

Swim start can be considered as an inclined single all-out push-off during which the swimmer must produce as much impulse as possible, so that when taking off, his body as the highest velocity possible to enter the water (in addition to an optimal diving angle). Many papers are available on this topic, an example with this review. Depending on the studies, the overall angle of push is about 25-30° (we’ll take 30° in our analysis). What all studies agree on is the fact that 50-m sprint start take-off velocity is a determinant of final performance.

So should I improve my take-off velocity through training, then my block phase performance will be better and all other things being equal, my final time will also be better. Remember I need a 0.05% improvement.

How to improve push-off performance through “optimized FV training”

Our approach based on individual assessment of the force-velocity profile and the computation of the individual optimal force-velocity profile (see here for a practical summary) shows that in theory, should my training be specifically programmed to reduce my force-velocity deficit (i.e. the gap between my actual and optimal profiles), then, my performance (take-off velocity or jump height in the case of vertical jump) should increase. This theoretical statement has been confirmed very clearly by experimental results (paper submitted, see the slides of a Congress presentation here). Briefly, in this study, subjects whose training loads and program were individually tailored to reduce the FV deficit all improved jump height, by extremely large magnitudes (by 14% on average for subjects with a 30-60% force deficit, which is ≈300 times the 0.05% I need) whereas only half of the controls did, and by lower margins. All control subjects followed the same program, working on all the spectrum of loads, from very high force to very high velocity, independently from their individual FV imbalance. A “one-size-fits-all” approach.

So below is my FV profile. The interpretation is pretty simple: I have a deficit in force (my actual FV profile (black curve) is 35% below what would be my optimal FV profile for a swimming start (pink dashed curve). Note that my actual FV profile was computed using Samozino’s method knowing jump height and total mass in a 5 or 6 loaded jumps protocol (see previous episodes and our IJSPP paper for details). It only took 20min and the scientifically validated iPhone app “MyJump”. The optimal FV profile is computed as the profile maximizing take-off velocity for a given maximal power output. Since the equation to calculate the optimal profile includes the push-off angle, we obtain different optimal profiles for vertical jumping (90°) and swimming sprint start (30°). Although MyJump only considers 90° vertical jump optimal profile (update in progress) the equation to compute optimal profile for any angle has been published here. So my objective here is to reduce this 35% deficit in force and shift my actual FV profile closer to the optimal for a 30° push-off.

The training program we used on our scientific experiment in such a case was pretty simple: heavy loads, maximal strength work, for 9 weeks with 2 sessions a week. In my opinion, any other training dose (sessions per week, total duration) would result in positive outcomes provided maximal lower limb strength is stimulated. More research is needed here to better design training programs (duration, content, etc…) but what we know for now is that even small training inputs resulted in substantial changes in the FV profile (decrease in the FV imbalance) and in turn increased push-off velocity. This resulted in our study in higher jump height, but in the case of a block phase, the final outcome would be an improved take-off velocity. Just do it.

On this basis, how can improve MY push-off performance

So my personal testing shows a force deficit, I will have to program a strength training that includes a majority of very heavy strength exercises for my lower limbs, and stimulate my maximal force as much as possible. Of course, taking into account all the other training components and performance determinants in a balanced and harmonious manner. Training is not an all/nothing world.

Then, in order to check my personal adaptations to the training program, I can do the FV testing (about 20min) every two or four weeks to check whether my responses follow the training predictions. This approach is dynamic and follows my training-induced adaptation. I may also program the duration of this specific program: short-term high dose or long-term more progressive dose, or any other combination…

Counterarguments and reasons for not doing it

Although at first sight this analysis and the training prescription make sense to me (but i'm totally biased), it might not be the opinion of my coach and S&C staff…here are counterarguments I may have to face, and my responses.

1/ wait, “Doctor”: what do you know about swimming? Who are you coaching?

Response: *facepalm*

2/ sorry, we usually do not work on lower limbs maximal strength. Swimmers don’t like it. We think it destroys their swimming legs technique, plus they prefer upper body strength work.

Response: *double facepalm* + no evidence on the long-term interference between legs strength training and swimming “technique”. However, evidence clearly show that strength work in the long term induces improvements in running economy, through a more efficient running “technique”…well ok running is different from swimming.

3/ you already have a pretty good start, why trying to improve it? You’d better work on your weaknesses.

Response: the thing here is that my start is ok, but according to the FV profile diagnosis, the margin for improvement is large and this could become an even stronger asset. Training philosophy here: should we work on our strengths and/or weaknesses…

4/ wait, your study does not include elite swimmers, right? So what if the results you obtained in trained soccer or rugby players differed in elite swimmers?

Response: I agree, but what if they did not? I would love to see such studies.

Important notes

First, our personal, unpublished observations show that most young swimmers we tested, and some elite sprint swimmers actually show a force deficit such as mine, so this approach may very likely apply to their case. Second, a recent MSc work I followed also shows this kind of force deficit in trained swimmers.

Conclusion

Swimming, and especially the shortest Olympic distance, is (now) a sport in which technological improvements are very unlikely or very small (once the body is shaved and piercings removed…), so a major part of performance improvements will likely result from training content and periodization (plus of course anatomical/genetical and pharmacological factors…). This is the reason why we think a better start performance (and in turn a better overall performance in the 50-m events at least) could result from monitoring the FV profile and most importantly from putting the conclusions of this diagnosis into practice and adapt (even a little bit) the training load accordingly. A diagnosis without a corresponding treatment is a loss of time.

Alternative solution: let your fingers grow!