top of page

Critical review of: The lumbar and sacrum movement patterns during the back-squat exercise.

In their study; The Lumbar and Sacrum Movement Pattern During the Back Squat Exercise; McKean, Dunn and Burkett (2010) looked to quantify and assess the differences in how the sacrum and the lumbar spine moved during the high bar back squat at two different widths, Narrow squat (NS) and wide squat (WS) and across gender. This is an interesting study and is an area that still needs more research due to the small number of participants. The authors tried to quantify the movement of the lumbar spine and the sacrum during the back squat. The data collected shows that once a subject is loaded the sacrum and lumbar spine change from their initial unloaded state, with gender and stance altering what happens during the movement and in which sequence the angles of the lumbar spine and sacrum adjust. The study was unfunded and seemed to have deficiencies in certain areas it required to ensure that the data collected was as accurate as possible.

The study reads very well and from initial observation appears to be a well thought out study performed to the best of the ability of the authors, it is hard to establish if this is a final year student project or whether the author is working as a researcher, due to the crossover of the lead authors educational timeline and the release date of this research. From November 2009 – June 2012 the lead author was in a position of post doctoral research fellow on a scholarship at the University of the Sunshine Coast and his education timeline runs up to June 2009. This may account for lack of constraints and control applied to the subjects. In the abstract there is also a miss-spelling of Sacrum which is a big oversight for a researcher to make. During this period all the authors were also involved in another squat movement study, released in June 2010, that has a lot of similarities to this study, which was released in October 2010. With the candidate selection process and the number of candidates it suggests that they collected the data for both studies at the same time.

In this study it appears a few areas have been overlooked and left to drift in the set-up, monitoring and execution. The data was collected with a 3 Dimensional Magnetic Tracking Device at 120 Hz and analysed using the Bland and Altman method. The authors state that the Three-dimensional magnetic tracking had been previously validated in a treadmill gait running study Mills, Morrison, Lloyd and Barrett (2007). The authors provided no other justifications to why this was the best equipment and methods for data gathering in this study and how the running study transferred to a squat movement.

The readings the authors collected were maximum lumbar angle; Timing of maximum lumbar angle; Maximum sacrum angle; and the time it occurred; maximum lumbar flexion angle; and the time when this occurred. Tracked in the decent and ascent phases of the back squat in both positions and both genders. No where do they mention any limitations to the study, but they do touch briefly on the limitations and the complexity of squat mechanics. The NS was established as being equal to the anterior superior iliac spine (ASIS) and WS being established as twice the width of the ASIS. Thirty subjects divided into 18 men and 12 women with a relative mean of 123% 1 Rep max (1RM) for men and 93% 1RM for the females. The subjects used were currently following a strength training programme at least twice a week for 12 months including the back squat exercise and were eliminated from the study if they had not met the required criteria. The candidate pool in which the authors chose to use were in their final year of study or working as an accredited personal trainer and freely volunteered to partake in this research study. There was no control group, however, with it being an observational study with no experimental manipulations in place, this would make having a control group unnecessary. They stated in the requirements that to take part in this research study participants must have been following a strength and conditioning programme for at least 12 months including the back-squat exercise. Does this make the subjects experienced enough or skillful enough to perform a back squat experiment to the degree of control required for lumbar spine and sacrum control? In the study Abdoli-Eramaki, Agababova, Janabi, Pasko, and Damecour (2019) they found that “The trained group experienced less loading on L5/S1” of which that would indicate that the control required in the lumbar and sacrum region would require more experience. This could have a material effect on the data collected, using more experienced strength trained athletes could produce a more consistent data source.

The qualification process for becoming a personal trainer has a large degree of variance in the standard and length of training and can range from 6 weeks to 2 years. Nowhere in the paper do they discuss who the candidates are training with and if it is an accredited awarding body. If the candidates have undergone a 6-week training course have they procured 6 weeks formal training and 46 weeks self-taught with regards to correct lifting techniques? A more extensive screening and selection process for candidates or a more specific demographic of athletes should be considered for future experiments, taking into consideration the ability and training experience of the athlete rather than if they have been going to the gym for a year. This area of the study demands a closer inspection, and further discussion to selecting appropriate candidates is needed.

Raudys and Jain (1991) observed that small study groups can produce problems in the design and data gathering of pattern analysis and much larger sample sizes are required to reduce error in statistical pattern analysis. In this study the data collected was analysed as an average, with such a small group you would only need a few of the participants to have poor technique to generate misleading data. Another solution to this would be to have a larger experiment with a greater number of participants in the study. The study ended up using 18 male subjects and 12 female subjects, this may have been due to availability of suitable candidates, is doing a comparison study with an uneven split on genders going to alter the results?

For a unfunded study it was very well thought out in reference to what data they were going to collect, gathering a 3 dimensional analysis of timing and range of movement in the lumbar-sacrum region, the 8 anatomical contact points had previously been validated in Morrissey, Harman, Frykman, and Han (1998). and validated against standardised reference measures. They did manage to quantify the movement of the lumbar sacrum during the back squat, however it was a very small study and the reliability of the data in questioned could be skewed if a couple of the participants techniques were array. In the paper figure 1 and figure 2 were photographs of one of the participants in the study connected to the 3-dimensional tracking device, loaded at the bottom of the decent phase in figure 1 and unloaded standing tall in figure 2. Although the participant seems to demonstrate a good back squat to a suitable depth, on closer inspection it is obvious that the participant is wearing what appears to be running trainers during a weighted squat study. Is not having a solid platform going to un-stabilise the participant and possibly create a forward lean in the squat and therefore influence the outcome of test? Barefoot squat tests have been performed against squat shoes and they found that heel raised squats in barefoot did not alter trunk or lower extremity biomechanics in the back squat exercise Charlton, Hammond, Cochrain, Hatfield and Hunt (2017). This control measure would have been easy to implement and would have taken away the variable of different shoe densities of the participants.

In figure 2, the participant is photographed without a load on her back. If we are to assume the pictures are an accurate representation of how the test was performed then the participant is stood with an anterior pelvic tilt and therefore creating a more lordotic curvature to the lumbar spine, rather than standing in a neutral setup. There are no mentions in the material to cues to stand in neutral for the initial unloaded readings, this could explain the initial data that recorded a kyphotic response to the bar being loaded on the back. They touched briefly on the alignment of the pelvis, from a previous study Delitto and Rose (1992) that found having an anterior pelvic tilt increased trunk muscle activation in a squat lifting and lowering technique. This was an industrial style lift, that allowed the heels to lift of the floor and the weight to be loaded to the front of the body. The authors noted “Which is quite different from a strength training deep back squat technique” but made no mention of cueing their participants to use this method, or any mention of cueing participants to find the correct alignment before the readings were taken.

Participants performed 8 back squats with an added load of 50% Body weight (BW), this weight was selected as a Common weight employed by beginners and women and allowed multiple sets of 8 repetitions without fatigue Donnelly, Berg and Fiske (2006). wanting to stay away from maximal loads to prevent nonvoluntary technique changes and to allow comparison to previous studies. 3 squats were analysed, and participants were blinded to what 3 squats were measured, although the first and last squat are not included in the selection process. There are no rationales in the material for what the reason they discounted the first and last repetition. Even at a submaximal load wouldn’t the first repetition be a good gage to assess the data, as the athlete will be mentally turned on and fresh, possibly performing the repetition to the best of their athletic ability?

In the study Nuzzo, McCaulley, Cormie, Cavil and McBride (2008) they found that performing squats at loads >90% 1RM created the greatest back extensor muscles (L1-L5) and trunk activation, With the load increasing and the activation and stability increasing would the lumbar and sacrum be more stable if a greater load was applied to the participants to increase the core and extensor activation? This area of study would be a useful tool to recognise what happens to the movement of the spine as greater loads are applied to the back squat exercise?

Squat width was one of the aims of the study, so a constraint on leg position was placed on the participants. However, participants where allowed to angle their feet how they felt best and to squat to a depth where they felt comfortable, although they were encouraged to straighten their feet for the narrow stance. In the Baechle and Earl (2008, p. 350) they talk in their squat guidelines to turn the feet out slightly. Why wasn’t there an official control placed on the foot position? Why wasn’t there a control placed on the depth of the squat? Will a variety of squat depths effect the amount of pressure that goes through the lumbar and sacrum? Bryanton, Kennedy, Carey and Chiu (2012) found in their study that despite load, the depth of the squat was the lead contributor to the relative muscular effort and the force applied. The deeper you descend into the squat, is it going to force the lumbar spine into a kyphotic posture or force more flexion through the sacrum? This factor being overlooked could affect the data received quite significantly due to the varying depths that the participants could have been hitting during the study. A NSCA accredited coach overlooked the study, reportedly all participants broke parallel, but this still leaves a very large variance in the range of movement that may have been achieved during the study. The authors of the study report that all NSCA guidelines were followed for the squat, despite the fact that they prompted the participants to point their feet forward in the squat against NSCA guidelines.

In the Baechle and Earl (2008, p. 250) it states that 1-5 minutes should be given between tests to avoid fatigue and up to 10 minutes for more maximal weight assessments, in this study they gave 120 seconds which is on the lower end of the guidelines for testing. In a study that requires the participants to perform a movement with precise control, surely a longer rest period would be advised to produce optimal performance and reliable data.

They found that there were differences in lumbar and sacrum movement across genders and across the different stances. But both genders had a significantly reduced maximum lumbar angle and range of movement with the wide stance compared to the narrow squat. The timing at which male and female participants hit maximum flexion was different and there was a difference across the varied stances. Lumbar and sacrum starting angles for male and female participants differed loaded and unloaded and in the varying stances. The authors declare within the discussion “sacrum and lumbar flexion angles from the unloaded to loaded start positions suggest that the human body has a means of determining the ideal posture of the spine, thus allowing it to achieve the optimal position to manage the additional load”. Partially backed up from Martin and Nelson (1986) where they found participants in their study moved their trunk forward to align the weight loaded and the body to the centre of gravity. This area needs more investigation and could have a positive implication on athletic development in the future.

Looking at the practical applications of this research, with differences of timing sequence from male to female genders, males in the decent reached maximum angle in the order of sacrum, lumbar angle, lumbar flexion and for the ascent phase it was just reversed being the same for both stance widths. The female participants on the other hand had a completely different maximum angle sequence and it varied between the stances and what phase of the movement they’re in. The researcher’s state that the reason for the difference in the movement for females is unknown and a topic for future studies. They go on to state that “screening protocols for squatting movement between men and women must be questioned”. This is a very strong statement to make considering the size of the study and potentially the experience of the participants that took part in the study. Further research would be an excellent idea to quantify the data across a broader selection of athletes and experience to give coaches more effective guidelines for the future. There are a few scenarios in this paper were the authors make strong statements, they stated that kyphotic posture in the lumbar spine during the back squat movement is normal and shouldn’t be a factor to stop an athlete from performing the back squat during the movement. This is an immense piece of information but doesn’t seem to be backed up from any other studies. The authors, earlier in the discussion go as far as to say “The change of lumbar curve is marginal and we suggest it would be difficult to view with the naked eye”, this point needs some clarification, as they have contradicted themselves in the same paper. If a coach can see kyphosis of the lumbar spine during the back squat of an athlete is this too kyphotic as it can be seen or is it fine to carry on with the exercise without a risk of injury?

In previous research, they found that male participants have greater lumbar support than women due to the additional soft tissue in male subjects, and female athletes would greatly benefit from additional strength training of the surrounding muscles of the lumbar spine. Gaining trunk and specifically lumbar strength, coaches should see an improvement in female squat performance. Could you argue that increased training, in any capacity, from general preparation phase to specific spinal and trunk activation training, would elicit a positive training response and thus increase the squat performance of a female athlete?

This is a very thought-provoking study with a clear objective which they successfully managed to collect the data for. Some of the faults in this study, potentially, could have been avoided with a greater emphasis on participant recruitment and more specific demographic selection to find experienced strength trained participants. Understanding the sacrum and lumbar movement during loaded back squat exercise for specific gender and stance will help with training guidelines in the future. This study could be expanded on to include greater loads, to help quantify if increased loads would help the stabilisation of the sacrum and lumbar. It would be very interesting to see if the data from this study would be affected if more specific constraints were placed on what is expected from the participants from set up, foot position and depth. Overall, this is a very comprehensive study and the data collected should be considered in the future for gender specific athletic programming.


Abdoli-Eramaki, M., Agababova, M., Janabi, J., Pasko, E., & Damecour, C. (2019). Evaluation and comparison of lift styles for an Ideal lift among individuals with different levels of training. Applied Ergonomics, 78, 120-126.

Baechle, T. R., & Earle, R. W. (2008). Essentials of strength training and conditioning (3rd ed.). Champaign, IL: Human Kinetics.

Bryanton, M. A., Kennedy, M. D., Carey, J. P., & Chiu, L. Z. F. (2012). Effect of squat depth on relative muscular effort in squatting. The Journal of Strength and Conditioning Research, 26(10),2820-2828.

Charlton, J. M., Hammond, C. A., Cochrane, C. K., Hatfield, G. L., & Hunt, M. A. (2017). The effects of a heel wedge on hip, pelvis and trunk biomechanics during squatting in resistance trained individuals. The Journal of Strength & Conditioning Research, 31(6),1678-1687.

Delitto, R. S., & Rose, S. J. (1992). An electromyographic analysis of two techniques for squat lifting and lowering. Physical Therapy, 72(6), 438-448.

Donnlly, D. V., Berg, W. P., & Fiske, D. M. (2006). The effect of the direction of the gaze on the kinematics of the squat exercise. The Journal of Strength and Conditioning Research, 20(1),145-150.

Martin, P., & Nelson, R. (1986). The effect of carried loads on the walking patterns of men and women. Ergonomics, 29(10), 1191-1202.

Morrissey, M. C., Harman, E. A., Frykman, P. N., & Han, K, H. (1998). Early phase differential of slow and fast barbell squat training. The American journal of Sports Medicine, 26(2),221-230.

Nuzzo, J. L., McCaulley, G. O., Cormie, P., Cavill, M. J., & McBride, J. M. (2008). Trunk muscle activity during stability ball and free weight exercises. The Journal of Strength and Conditioning Research, 22(1), 95-102.

Raudys, S. J., & Jain, A. K. (1991). Small sample size effects in statistical pattern recognition: recommendations for practitioners. IEEE Transactions on Pattern Analysis and Machine Intelligence, 13(3), 252-264.

201 views0 comments


bottom of page