Jump squats are a popular addition to most strength and conditioning programs for athlete of all kinds to improve explosive power. And most researchers and coaches know that there is an optimal load for power output for jump squats and that it is usually around 0% of back squat 1RM (no load – just bodyweight).
However, while the total overall leg power may be maximal at 0% of 1RM, that doesn’t mean that hip, knee and ankle joint powers are all maximal at the same point.
Let’s take a look at a study that investigates exactly this question.
The study: The effects of load on system and lower-body joint kinetics during jump squats, by Moir, Gollie, Davis, Guers and Witmer, in Sports Biomechanics, 2012
(Too much detail? Skip to the practical implications)
What’s the background?
Power is a key determinant of sporting performance that varies depending upon the exercise, the number of repetitions and sets, the recovery periods and the loads used relative to 1RM.
In respect of the load used, researchers have generally found that this differs widely according to the exercise. For traditional resistance exercises, Siegel (2002) reported maximal power output with loads of 50 – 70% of 1RM for the squat and 40 – 60% of 1RM for the bench press. Similarly, Cormie (2007) found that the optimal loads were 0% of 1RM for the jump squat, 56% of 1RM for the squat, and 80% of 1RM for the power clean.
In measuring power, most studies focus on the power exerted upon the external load, which is measured by reference to the displacement characteristics of the body and the barbell.
However, the individual joints also have their own power outputs and these may not change in direct proportion to the external power output, just as the relative contribution of various joint torques to a barbell exercise change with increasing loads (see further the effect of load affects joint torques in lunges and squats).
What did the researchers do?
The researchers wanted to see whether the power output at the hip, knee, and ankle joints were similarly affected by changes in the external load during jump squats. So they recruited 12 resistance-trained males who had previously been participating regularly in resistance training programs during the previous year and who were involved in sports including football, soccer and basketball.
The researchers recorded various measurements from the subjects during two testing sessions. In the first session, the subjects performed a 1RM parallel back squat. In the second session, the subjects performed jump squats with loads equivalent to 0%, 12%, 27%, 42%, 56%, 71%, and 85% of their 1RM back squat with 2-3 minutes of rest between sets.
During the tests, the researchers measured ground reaction forces using two force plates. They also measured the movements of the barbell and joints using a three dimensional (3D) motion analysis system that was designed to monitor the movements of 16 retro- reflective markers positioned over various key anatomical landmarks.
The researchers reported that the average 1RM back squat of the subjects was 181.8 ± 40.4kg. They compared this to the average bodyweight and found that it represented 1.81 ± 0.32 times bodyweight. The subjects were therefore considered relatively well-trained although by no means were they strength athletes.
The researchers reported that the average jump height reduced significantly as load increased, as shown in the chart below. This was expected and was in line with previous studies.
External power output
The researchers reported that the average external power output reduced significantly as load increased, as shown in the following chart. This was also expected and was in line with previous studies.
Internal (joint) torques
The researchers found that the moment at the hip, knee and ankle joints all increased significantly with load. They did not note any significant differences in the increases of the hip, knee and ankle torques with increasing load. All joint torques appeared to increase similarly, as shown in the chart below.
This result was in contrast to recent investigations by Bryanton (2012) and Lorenzetti (2012), who both found that as squat load increased, the hip and ankle torques increased with greater rapidity than the knee torques.
Internal (joint) power outputs
The researchers found that there was a significant quadratic trend for power at the hip joint in that it increased with the load up to a maximum at the 42% of 1RM condition, after which values decreased. The researchers found that the power at the knee and ankle joints decreased significantly as the load increased according to linear trends.
Since the power outputs of the various joints are very different, this is quite difficult to see on a chart comprising the absolute figures. I have therefore expressed the figures as percentages of the 0% of 1RM power output at each joint. While this isn’t a hugely scientific way of presenting the data, it does make it much easier to see the difference in the trends. The hip power follows a curved path up to 42%, while the knee and ankle powers decrease linearly.
The chart looks a little messy at first glance but if you mentally separate out the hip power (the darkest set of lines) then you can see that the knee and ankle powers behave similarly and just decrease with increasing load.
What did the researchers conclude?
The researchers concluded that the power output of each of the lower-body joints does not alter in proportion to the external power output during jump squats. The researchers also concluded that while the power output at the knee and ankles joints decreased with increasing load, the power output at the hip increased with increasing load up to 42% 1RM.
The researchers also conclude that using loads at specific percentage of 1RM could lead to preferential development of either hip or knee/ankle joint power, depending on the load used. This can be seen in the following chart.
The above chart shows that the relative contribution of the powers at the hip and knee change depending on the load used. At 0% of 1RM, the knee power is much more significant than the hip power. At 42% of 1RM, the two joints appear to contribute similarly. As loads increase further, the hip joint power decreases more quickly than the knee joint power and the relative contribution of the knee joint power increases again.
So training at 42% of 1RM in the jump squats would appear to maximize hip extension power while training at 0% of 1RM will likely emphasize the relative contribution of the knee joint power and train this aspect of the leg musculature more effectively.
As noted above, the point at which optimal power differs widely depending on the exercise and therefore it is likely that joint power will also differ depending on the exercise. Therefore, this study was limited to an analysis of the jump squat only and different results might be obtained with hex-bar jump squats, Olympic lift variations and other explosive lifts.
What are the practical implications?
For all-round lower-body power development:
It may be superior to train jump squats with a range of loads instead of a single optimal load. By training with a single load, it is likely that the hip joint power is not being trained to the fullest extent. At least two loads are likely preferable, such as one with 0% of 1RM and one with c. 40% of 1RM.
For improving an athlete’s vertical jump:
Athletes tend to have either a hip-dominant or a knee-dominant jumping style. Therefore, allocating the correct type of jump squat load to help the athlete improve power in their preferred jumping style could be central to improving their vertical jump performance.
For developing power for a specific sport:
It may be important to identify the power requirements of the sport before selecting the training load for jump squats. For example, given the relatively higher level of hip torque involved in maximal parallel squats than knee joint torque, developing hip joint power with jump squat loads at c. 40% may be more beneficial for powerlifters than developing knee joint power with 0% of 1RM. However, again, this likely depends on the individual’s squatting style.