By Austin Baraki
Introduction
Today we’ll be discussing a paper published in the Journal of Sports Sciences in April of this year (see here) analyzing the biomechanics of the vertical jump, power clean, and jump squat. The vertical jump is often used as a diagnostic test to evaluate an athlete’s ability to generate power; that is, how much work (Force x Distance) is done per unit of time. Vertical jump performance is determined by many factors, one of which is a genetic predisposition for high neuromuscular efficiency, but part of an athlete’s ability to produce power is also trainable within their genetic potential. Training the “fast lifts” (e.g. power clean) is a very effective way of converting basic strength into power, as evidenced by improvements in vertical jump performance after training them. However, the precise mechanism by which this “carryover” occurs has not been clearly described in the literature.
An anectodal showcase of strength, which is a general adaptation, being applied specifically. Aside from his weightlifting prowess, Hamman also plays basketball and golf. Despite his 350 pounds (160 kilograms) frame, he can hit a golf ball 350 yards (320 m), do a standing double back flip, and leap vertically three feet (0.9 m). -source (Wikipedia)
The principle of specificity states that the training adaptation will be specific to the training stress; in other words, a Specific Adaptation to Imposed Demands (SAID). In many contexts this makes intuitive sense; in order to improve one’s squat, one must squat. In order to swim faster, a swimmer must swim, and in order to run faster, a runner must run. However, this idea does have limitations. Strength is the most general physical adaptation from which all other physical attributes can be built; conversely, as stated in Practical Programming, 3rd ed., “the absence of sufficient strength limits the development of all other athletic parameters.” Comparing two powerlifters, runners, or swimmers who train with equivalent volume and intensity, with all else equal the stronger athlete will win. So, in order to train most effectively, general adaptations should not be overlooked in favor of exclusively sportspecific training; an athlete should develop a general (i.e. nonspecific) base of strength, and then practice their sport for the “specificity effect”. There is a reason why football players don’t just scrimmage every day yearround in order to get better at football, and why Crossfitters don’t (or shouldn’t) just do metcons every day yearround in order to get better at Crossfit; offseason is a time to get generally bigger, stronger, and faster in order to then be more effective at their particular task.
It should therefore be clear that although the vertical jump is a skill that needs to be practiced in order to improve, it could certainly improve by other means as well. Specific practice would improve neuromuscular coordination by way of familiarizing the athlete with the movement pattern. But we would also expect basic, nonspecific strength to be one of these determinants as well, because more strength means more force to propel them upwards. But jumping is also necessarily a rapid movement; one cannot “jump slowly” in the same way one cannot perform a power clean slowly. Therefore, this is a situation where an athlete’s ability to generate force rapidly (i.e. “rate of force development”) is critical. Finally, it is understood that standing vertical jump performance is significantly influenced by genetic predisposition (which influences things like muscle fiber makeup and neuromuscular efficiency for motor unit recruitment), and is therefore limited in how much it can be improved by training for a given individual.
So, we can postulate that vertical jump performance is determined by :
1) practical skill with the movement;
2) general strength base;
3) rate of force development; and
4) genetic endowment (which partially influences #3)
Now, the principle of specificity would hold that one should train the vertical jump by jumping vertically. One could also use weighted vertical jumps to simulate the kinematics of the goal movement as closely as possible while providing a progressive overload stimulus for further adaptation. These methods have been well established in the literature. But as discussed previously, since the power clean is another method known to improve vertical jump performance, what is the precise mechanism by which this happens? Are the kinematics and kinetics of the power clean similar enough to those of the vertical jump that principle of specificity applies here? Or are there other mechanisms at play? These are the questions the authors sought to investigate.
Study Methods
The authors recruited 10 healthy collegeaged football bros and 10 healthy collegeaged volleyball babes for their study. These athletes were already familiar with the movements and had performed them “on a routine basis for at least 2 years” (exactly how “routinely” is left unspecified). The subjects were studded with EMG electrodes (devices used to measure electrical muscle activity) over the muscle bellies of the medial gastrocs, vastus lateralis, rectus femoris, biceps femoris, and gluteus medius of one leg. They were also fitted with electronic goniometers (devices used to measure reference angles) across the ankle joint, knee joint, and hip joint of the same leg. After a brief standardized warmup, the athletes were cycled through the three exercises a few times while standing on a force plate. Fortunately, the study describes in detail with pictures exactly how the athletes were instructed to perform the exercises:
1) Vertical jump: athletes performed a downward “countermovement” to whatever arbitrary point felt strongest for them while keeping their arms at their sides, then launched upwards with a swing of the arms overhead.
2) “Power cleans” used a 70% 1RM weight for each individual; they were apparently instructed to start from a standing position, lower the bar to midshin height without touching the floor, then perform the power clean as explosively as possible. Although this sounds more like a lowhang power clean rather than a true power clean starting from the floor, the authors explain that this technique was used to mimic the “countermovement” used in the vertical jump and jump squat to generate a stretch reflex.
3) Jump squat used the same weight (70% of 1RM power clean) carried in the back squat position (unspecified, but presumably highbar) and athletes were to perform the same style of “countermovement” as in the vertical jump trial, although this time the hands would obviously be holding the barbell on the back rather than down at the sides.
Data for multiple trials was measured from the start of the upward phase of each exercise (i.e. the instant when any of reference joints began extending by goniometry) until the end of extension (when all reference joints were maximally extended, or “triple extension”).
1) The force plate data provided information about ground reaction forces. These data allowed for calculation of parameters such as maximum force, maximum rate of force development (RFD), power generation, and the precise time points when these maxima occurred during the upward phase of each exercise.
2) The electromyographic (EMG) data showed the maximum raw signal emanating from the reference muscles (indicating peak muscle activation) and the time point when these maxima occurred.
3) The goniometric (joint angle) data was also crunched and plotted temporally as a “percentage” of the upward phase (i.e. time 0 = 0%, time at “triple extension” = 100%)
Results
| Movement | Force (Newtons) | RFD* (Newtons/S) | Power (Watts) |
| Power Clean | 2411 | 17254 | 3532 |
| Jump Squat | 2234 | 3836 | 3772 |
| Vertical Jump | 1770 | 3517 | 4384 |
*RFD= Rate of Force Development
1) Results from force plate measurements during the upward phase of each movement are shown in the table above. The takeaway: Power cleans generated significantly greater maximum force, as well
as much higher maximum rate of force development (RFD), than did jump squats or vertical
jumps. The authors hypothesize that the power clean’s greater force production and RFD result from the “double knee bend” phase generating a brief stretchshortening cycle while the joints are in an optimized position. Interestingly, when compared to the data from the upward phase of each movement, the “countermovement” (downward dip before exploding upwards) for vertical jump and jump squat seem to generate higher RFD (9465 N/s and 7920 N/s, respectively) than the upward movements themselves, although these are still blown away by the power clean’s peak RFD.
Vertical jumps showed the greatest power output compared to power cleans and squat jumps, and this point is less surprising (and less significant) for a few reasons. Consider that Power = Force x Distance / Time, and that the results show vertical jumps producing less force than the other exercises. Therefore, the only mathematical way for the vertical jump to produce significantly more power despite less force is for it to have resulted in a greater distance moved, and/or been completed in a shorter time period. And although the complete raw data are not provided in this paper, they did plot the results of one typical study participant whose vertical jump upward phase was completed in less than half the time of his power clean’s upward phase. This alone would be more than enough to account for the difference in calculated power, but then consider that a plain unweighted vertical jump also results in a higher jump height (distance) than a weighted squat jump or a power clean. Therefore the increased power output of the vertical jump is primarily accounted for by moving a greater distance in a shorter period of time, despite producing significantly less force output.
It is known that maximum power output occurs at significantly lower intensities compared to 1RM effort, and this is what makes the vertical jump a good test for an athlete to demonstrate power output. But this should not be confused with the best way to train for power output, which requires an overload stress to drive an adaptation to produce more force in less time. Logically, the best way to train for power output would be to increase force production (i.e. strength), and improve one’s rate of force development. And what exercise happened to absolutely dwarf its competitors in terms of rate of force development…? The power clean. So get your deadlift up, folks, and don’t forget to do your power cleans.
2) EMG data showed that for most of the studied muscles (vastus medialis, rectus femoris, gastrocnemius, gluteus medius), maximal activation occurred later in the power clean (during the second pull) and earlier in the jumping movements. The only exception was the biceps femoris muscle (which, being one of the hamstrings, was predictably engaged earlier during the first pull). The data also showed greater activation of the rectus femoris in the jumps, while the biceps femoris showed greater activation in the power clean. These data are much less interesting because of the significant limitations and unreliability of EMG data; plus, relative activation of individual muscles is not a particularly useful parameter for our purposes in an athletic context. However, the significant differences in EMG patterns between the power clean and jumps suggests that specificity is not the major contributor in the carryover between them.
3) Goniometer data showed significantly earlier ankle and knee extension in the power clean, followed by a brief phase of flexion (the “double knee bend”) before the final rapid hip extension. This should make sense considering how the power clean is performed as compared to a freestanding jump. Hip extension in the clean also occurred significantly later than in either of the jumping movements. In short, all three joints exhibited statistically significant kinematic differences between the power clean and jumping movements.
Discussion
Considering force production parameters, muscular activation patterns, and joint kinematics involved in these exercises, the data suggest that power cleans and jumping movements are significantly different enough that the principle of specificity does not adequately explain the training carryover between power clean and vertical jump performance. To quote from the paper (emphasis mine): “In order for a training exercise to facilitate an improvement in performance in a specific sport movement, such as vertical jumping, the exercise must stimulate a trainable feature of the neuromuscular system beyond the level that can be achieved when executing the sport movement” (i.e. progressive overload!). This study suggests that peak force production and rate of force development are these aspects of the neuromuscular system that are uniquely stimulated by training the power clean, accounting for its training effect on vertical jump performance more so than any kinetic or kinematic specificity. Of course, another benefit of the barbell power clean its its incrementally loadable nature to continue driving progress for a long time. One limitation of this particular study is its relatively small sample size of 20 athletes, but the results are certainly convincing. Ideas suggested for future study include studying other “fast lifts” like the snatch, comparing these lifts from the hang position versus the floor, and studying the effect of the “countermovement” phase in power output.
To close with another quote from the paper: “While muscle coordination is important in jumping performance, if it is achieved in a separate facet of training, an athlete’s strength and conditioning programme need not focus on kinematics. Instead, exercises should be prescribed based on an understanding of the specific motor ability that is being trained” (emphasis mine).
Brilliantly put. Get strong using the best overall exercises to get you strong, then practice your particular task or sport separately. The results of this study show that conflating the two is clearly suboptimal.
Citation
Mackenzie SJ, Lavers RJ, Wallace BB. A biomechanical comparison of the vertical jump, power clean, and jump squat. Journal of Sports Sciences. 2014 Aug;32(16):157685
About the Author:
