How deep should you squat? How many reps should you? These existential questions have troubled lifters for many years, although the current generation seems to have greater difficulty with the issue than those preceding it.
This study did not quite make the cut for our monthly review. I thought it was worth reviewing anyway, because it sheds a little light on the debate between partial and full range-of-motion squats and how they could be used in strength and conditioning programs.
The study: Effects of changing from full range of motion to partial range of motion on squat kinetics, by Drinkwater, Moore and Bird, in Journal of Strength and Conditioning Research, 2012
The confusion over squat depth
By way of background to their study, Drinkwater et al. explain that the squat and more specifically squat depth are poorly understood in the fitness industry. There is confusion regarding what a “deep squat” implies and even the concept of “parallel” seems to puzzle some people. So to avoid any bamboozlement in the following discussion of this study about parallel and partial squats, I am going to use the following definitions:
- Deep or full squat – this is where the hips are well below the horizontal plane of the knees. From what I have seen, not many people squat like this unless they do a lot of Olympic weightlifting or are seriously old school.
- Parallel squat – this is where the center of the hip joint only goes as far as being parallel to knees. This is not to be confused with (a) the bottom of the thigh being parallel with the knees, which is significantly higher, nor (b) 90 degrees of knee flexion, which is also significantly higher unless you are box squatting or have really weird anthropometrics involving femurs that are 150% as long as everybody else’s. A lot of powerlifters squat to this depth or slightly below in training.
- Partial squat – this is where the knees are flexed to only 90–120 degrees, the upper end of which is the standard squatting depth in commercial gyms.
Of course, this may not be the terminology you prefer. And that’s fine. I just need to define my own terms so that the use of them in this article is clear. Let’s move on.
A brief history of squat depth and knee health concerns
Drinkwater et al. note that the idea that deep squats could be detrimental to knee health originated with a study by Karl Klein in 1961 (in press). The researchers note that Klein was worried by the high shear forces at the knee he observed during deep squats and also by an increase in ligament laxity in weightlifters that used them.
However, Drinkwater et al. note that these findings were soon extrapolated by many in the industry, who confused deep squats with parallel squats (hence the terminology discussion above) and soon any squat deeper than a slight knee bend was criticized as being dangerous. Additionally, as Schoenfeld (2008) has pointed out, even deep squats have since been cleared of their bad reputation (see his article for more references).
Nevertheless, Drinkwater et al. note that this confusion and consequent tide of criticism led to a wave of partial squats being performed throughout the industry and in gyms up and down the country. This was unfortunate, as for beginners, full range-of-motion (ROM) movements have generally been found to be more beneficial for strength and size gains, as was recently noted by Ronei (2011).
Benefits of partial ROM lifting
However, this does not mean that the partial squat is not a useful exercise for advanced lifters. After all, Mookerjee (1999) found that partial lifts could be used to improve the performance of a whole lift in subjects whose performance in that lift had previously stalled.
Additionally, there are sound theoretical grounds for using partial lifts. Frost (2010) explained that the concentric phase of heavy (i.e. greater than 70-80% of 1RM) lifts tends to have four sections. At first, there is an acceleration phase, which is followed by a deceleration phase called the “sticking region”.
This is then followed by a second acceleration phase, which is the recovery, and a final deceleration phase to the finishing point. The sticking region is a point in the joint ROM that is weaker than the other parts, either for reasons of physiology or mechanical advantage, and it can be trained specifically using partial ROM lifts.
Though the partial squat doesn’t strengthen the sticking region, it does allow for heavier loads to be lifted in the top range of the movement, theoretically leading to increased strength in that specific range of motion.
Limitations of partial ROM lifting
But of course, there is no perfect lifting method that has only benefits and no limitations. Some of the of the key limitations of partial ROM lifting are:
- Partial lifts only operate within a small range of the length-tension curve of the prime movers, whereas these muscles change length significantly during the full ROM lift.
- Partial lifts only operate within a small ROM and therefore the motor learning that occurs appears to apply only to that ROM. This means that significant strength is only gained through neural adaptations for that specific ROM and slightly either side. However, Kubo (2006) found that some neural gains are made as a result of partial lifts throughout the whole joint ROM.
- Partial lifts do not activate the prime movers in the same pattern as full lifts, although, as Clark (2012) has noted, we do not currently know how squat depth affects EMG muscle activity of the prime movers.
- Full ROM has been shown to be better for the vertical jump than partial squats, as Hartmann (2012) recently reported.
What did the researchers do?
Drinkwater et al. compared high bar, shoulder-with stance parallel and partial squats to 120 degrees knee flexion at similar and different weights (67% and 83% of 1RM). The weights were all lifted at a self-selected speed.
The researchers recruited 10 male recreational rugby players and had them perform four separate squat workouts, of 4 sets of 10 reps with 90 seconds rest, being parallel – 67%, parallel – 83%, partial – 67% and partial – 83%.
While the subjects performed the lifts, the researchers measured the displacement, velocity and forces applied to the barbell throughout.
Drinkwater et al. observed that the 1RM parallel squat mean was 148.8g, and the 1RM partial squat to 120 degrees of knee flexion was 270.8kg. So the rugby players were not untrained, by any means, and they had certainly spent a little time under the bar.
As expected, the partial squat at 83% produced the greatest peak force, followed by the partial squat at 67%, the parallel squat at 83% and the parallel squat at 67%, as shown in the chart below:
Again, as expected, the peak velocity results show the exact opposite trend as the force results. This is understandable, as the heavier weight makes it harder to accelerate.
So far, so good. What is very interesting, however, is that at these intensities, peak power follows the force trend reasonably closely and is not particularly affected by the reducing velocity. Obviously, because of the opposing velocity trend, the line is not as steep but nevertheless it is still a strong trend for increasing power with increasing weight.
Finally, the researchers found that total concentric work per repetition was highest in the parallel squat with the heavier load. In fact, the partial squats were pretty poor for doing work and neither of them was as high as either of the parallel squats.
This last chart spoils the nice consistent pattern we were getting. And that is interesting, as we will see in the next section, where we analyze these results a bit more closely.
What do the researchers conclude?
Researchers generally like to find variants of lifts that produce “the most” of a particular variable, as this tends to indicate that a quality (i.e. speed, force, power, etc.) is being trained most effectively. Of course, this falls down slightly where other considerations are important (such as a sports-specific ROM or a sports-specific velocity or movement pattern).
However, it is an interesting starting position and certainly at least as good as any other. Drinkwater et al. note that the following methods are maximal in these respects:
- Force – partial 83% (i.e. high percentage of 1RM)
- Speed – parallel 67% (i.e. low percentage of 1RM)
- Power – partial 83% (i.e. high percentage of 1RM)
- Work – parallel 83% (i.e. high percentage of 1RM)
They therefore conclude that partial squats with high loads and parallel squats with both high and low loads can be effective training modalities depending on goal. However, they suggest that training partial squats with low percentages of 1RM is not useful for developing any specific goal.
Sadly, of course, partial squats with moderate loads (i.e. 8-12RM) are probably the most commonly performed squat variant in all the gyms in the world. And they are definitely not “the most” at anything. Such is life.
What are the limitations?
As noted above, this study has a number of limitations:
- The recommendation for athletic training does not consider whether a sports-specific ROM or a sports-specific velocity or movement pattern is required.
- The recommendation for athletic training does not consider whether the non-ballistic method to the ballistic method would be better for velocity-specific training, nor does it compare the isoinertial (constant resistance) method to the variable method (bands or chains) for the same purpose.
- The study does not cover a wide range of percentages of 1RM and substantial differences would be expected at much lower percentages of 1RM, where power output would be higher because of the higher velocities obtained.
- While work done is an important consideration for body composition goals, particularly fat loss, it is not necessarily the panacea when it comes to hypertrophy, which can be caused by a wide range of factors, including the degree of muscle activity, volume, muscle damage and blood flow restriction (i.e. pump). The researchers did not record EMG muscle activity during the different squat variants, which would have been interesting. Previous studies, such as Caterisano (2002), which have looked at the EMG muscle activity to different depths, have used the same loading for all squat types and therefore have been flawed.
What are the practical implications?
A combination of partial squats with high loads and parallel squats with low loads but a focus on speed could be a potent combination for developing both speed and strength qualities for athletes.
For body composition programs requiring either fat loss or muscle gain, work output should be maximized, which suggests that deeper squats for heavy weights should be preferred. Although lighter squats can be performed for more reps, more sets of the heavier squats easily make up the difference.
For more information just like this, please subscribe to our monthly research review service. For only $10 per month, you get between 45 – 50 study reviews every month, delivered directly to your inbox…