The effect of experimentally induced gluteal muscle weakness on joint kinematics, reaction forces, and dynamic balance performance during deep bilateral squats

Squatting is a common daily activity and fundamental exercise in resistance training and closed kinetic chain programs. The aim of this study was to investigate the effects of an experimentally induced weakness of the gluteal muscles on joint kinematics, reactions forces (JRFs), and dynamic balance performance during deep bilateral squats in healthy young adults. Ten healthy adults received sequential blocks of (1) branch of the superior gluteal nerve to the tensor fasciae latae (SGNtfl) muscle, (2) superior gluteal nerve (SGN), and (3) inferior gluteal nerve (IGN) on the dominant right leg. At the control condition and following each block, the participants were instructed to perform deep bilateral squats standing on two force plates. Hip, knee, ankle, and pelvis kinematics did not differ significantly following iatrogenic weakness of gluteal muscles. The most important finding was the significant differences in JRFs following SGN and IGN block, with the affected hip, patellofemoral, and ankle joint demonstrating lower JRFs, whereas the contralateral joints demonstrated significantly higher JRFs, especially the patellofemoral joint which demonstrated an average maximum difference of 1.43 x body weight compared with the control condition. When performing a deep bilateral leg squat under SGN and IGN block, the subjects demonstrated an increased center of pressure (CoP) range and standard deviation (SD) in mediolateral compared with the control condition. These results imply that squat performance changes significantly following weakness of gluteal muscles and should be considered when assessing and training athletes or patients with these injuries.

Squatting is a common daily activity and fundamental exercise in resistance training and closed kinetic chain programs, [1][2][3] as it can effectively improve lower body strength, sprint speed, and agility in athletes. 4It has biomechanical and neuromuscular similarities to several athletic movements and it is therefore included in many sports routines, designed to improve athletic performance. 5,6The deep bilateral squat is a sitting posture with dorsiflexed ankles, deeply flexed knee, and hip.The optimal performance is a parallel alignment of hip, knee, and ankles with no mediolateral (ML) movement, while both heels remain on the ground at all times. 7Incorrect squatting execution could induce increased compressive and shear forces at the hip, knee, and ankle joints. 7,8veral conditions such as chronic hip muscle weakness/ insufficiency, 9 fatigue, 10 pain, 11 and osteoarthritis 12 might influence squat performance 13 and therefore, should be considered when assessing and training patients with these pathologies.In the presence of lower extremity muscle weakness, abnormal movements, and compensation mechanisms might be observed in an attempt to complete the task. 14Since the gluteus maximus is the primary muscle involved in squatting motions, 15,16 and hip abductors (gluteus medius, minimus, and tensor fascia latae) are the main contributors to ML pelvic stability and balance, 17,18 a weakness of the gluteal muscles could significantly affect the squat performance.
The gluteus maximus is innervated from the inferior gluteal nerve (IGN) and does not play an essential role in posture.It remains relaxed in the standing position and it is activated during running, climbing, standing from a sitting position, and rising during squats.Gluteus maximus weakness is observed through prolonged sitting, 19 anterior pelvic tilt, 20 overactivity of hip flexors, 21 pain, 22 and damage to the IGN (usually iatrogenic). 23Moreover, it is demonstrated that patients with symptomatic femoroacetabular impingement (FAI) might have weaker hip extension muscle strength compared with subjects with asymptomatic FAI. 24Hip abductors are innervated from the superior gluteal nerve (SGN) and are important for the stabilization of the trunk in ML directions during gait (especially stance phase), 25 single-leg standing, 26 and landing. 27p abductor weakness is commonly observed in patients with greater trochanteric pain syndrome, 28 L5 radiculopathy, 29 SGN injury, 30 hip osteoarthritis, 31 and developmental dysplasia of the hip. 32e aim of this study was to investigate the effects of an experimentally induced weakness of the gluteal muscles, through sequential nerve blocks of the superior and inferior gluteal nerves, on joint kinematics, reactions forces (JRFs), and dynamic balance performance during deep bilateral squats in healthy young adults.
The study hypothesis was that the subjects would demonstrate significantly altered squat performance (JRFs, and dynamic balance performance) following the nerve blocks compared with the control condition (without a nerve block).

| Participants
The study was approved by the ethical state committee (BASEC-Nr.:2021-00340)and each patient provided written informed consent before participation.Ten healthy adults (five females) with an average age of 24.7 ± 2.7 (22-29)

| Nerve blocks
Sequential blocks of the (1) branch of the SGN to the tensor fasciae latae (SGNtfl) muscle, (2) SGN, and (3) IGN were performed on the dominant right leg in the lateral decubitus position using ultrasound (SonoSite SII, FUJIFILM SonoSite, Inc) with a 6-to 13-MHz linear probe from an experienced anesthesiologist (J. A. and U. E.) at the same day.
The ultrasound probe was placed transversally in the gluteal region with the leg in hip and knee joint flexion up to 90°. 33The SGNtfl was identified on the lateral part of the leg in the triangle formed by the gluteus medius, gluteus minimus, tensor fasciae latae muscle, the SGN superior to the M. piriformis, and the IGN inferior to the M. piriformis. 34 in-plane or out-of-plane technique depending on accessibility, a short bevel needle (UPC 90-150, RM Temena GmbH, Felsberg, D) connected to a nerve stimulator (Stimuplex HNS 11 ® ; B. Braun Melsungen AG) with the setting 0.5 mA current intensity, 0.1 ms impulse duration, and 2 Hz impulse frequency for dual guidance was advanced until its tip is positioned near the nerve or in the plane between the respective muscles. 35Mepivacaine 1% 5-10 mL was used for each of the nerve blocks.The duration of action of mepivacaine 1% for a peripheral nerve block in human volunteers is 125-150 min. 36e success of the nerve block was confirmed clinically as reduced strength against resistance of the affected muscles but also electromyographically by an experienced neurologist certified in clinical neurophysiology (J.R).Needle electromyograpy (EMG) recordings (Dantec Keypoint software 2.0) were performed before and approximately 5 min following each nerve block in the corresponding, innervated muscles (i.e., tensor fascia latae, gluteus medius/minimus, and gluteus maximus).This was done to document that the effect of the previous block had dissipated and to assess spontaneous muscle activity, which helped exclude pre-existing neurogenic pathology.Interference patterns during maximal voluntary contraction were compared between the two-time points.
Successful paralysis was defined as a complete disappearance of EMG activity obtained during maximal voluntary contraction.To rule out an inadvertent weakness of the remaining muscles of the pelvic girdle due to diffusion of the anesthetic agent, additional EMG recordings of adjacent muscles were obtained.With this approach, it was confirmed that all participants received successful nerve blocks.

| Full-body movement analysis
Participants were equipped with 56 retroreflective skin markers placed by an experienced movement scientist (M.G.) according to a previously described marker configuration. 37They were then asked to perform a series of activities of daily living including one-leg stance, lounges, gait, stairs, and deep bilateral squats before and following the three nerve blocks.For this study, only data on deep bilateral squats were considered.The participants were instructed to perform squats standing on two force plates (9260AA) (Kistler Group) starting from the upright position (0%) with their hip, knee, and ankles parallel, to the sitting posture (50%) and again in the standing position (100%) (Figure 1).The recording frequency of the motion caption and force plates were 200 and 1000 Hz, respectively.Cameras and force plates were time synchronized (Vicon Nexus).All activities were repeated until six valid trials were collected.Three-dimensional marker positions were tracked using a 27-camera optical motioncapturing system (Vicon UK) at a sampling frequency of 200 Hz.

| Data collection and processing
Motion capture data were preprocessed using commercial software (Nexus version 2.8.1;Vicon UK), involving marker reconstruction and labeling, gap filling, and filtering of the marker trajectories, as well as the setting of temporal events (i.e., endpoint sitting posture and endpoint of standing posture) for the identification of the relevant data sections.Postprocessing was conducted with a custom-built MATLAB routine (R2019a, MathWorks, Inc).
The rotational kinematics (i.e., flexion/extension, abd-/adduction, internal/external rotation) of, bilateral hips, knees, ankles, and pelvis were calculated.A positive value indicated a flexion, abduction, and external rotation of the affected joint.A previously validated and published musculoskeletal modeling 38 was performed using OpenSim (Version 4.0, SimTK) to calculate the JRFs of bilateral hip, knee, and ankle joints in anteroposterior (AP), ML, and proximal-distal (PD) direction.A positive value indicated an anterior, inferior, and lateral F I G U R E 1 Study set-up.Participants were equipped with 56 retroreflective skin markers and were asked to perform deep bilateral squats standing on two force plates starting from the upright position (A) with their hip, knee, and ankles parallel, to the sitting posture (B) and again in the standing position, before and following the three nerve blocks.Three-dimensional marker positions were tracked using a 27-camera optical motion-capturing system.direction of the force vector.Balance measures included the center of pressure (CoP) path range, standard deviation (SD), and velocity in the ML and AP directions.CoP path range indicated the maximum range of movement during squatting in each axis, with higher values implying that the participant swayed further. 39The SD of CoP represented the variability about the average position of the CoP coordinate, with a higher value indicative of a less clustered movement pattern (or less controlled) behavior. 40CoP path velocity described the speed at which the center of mass was moving, with a higher value suggesting a more rapid and potentially unstable movement. 39

| Statistical analysis
A two-way repeated measured analysis of variance (ANOVA) and Bonferroni post hoc analyses were applied to compare the average (all subjects with six valid trials each) rotational kinematics (pelvis and bilateral hip, knee, and ankle joints), JRFs and CoP range and SD (in both ML and AP direction) between each state (control, SGNtfl, SGN, and IGN block).

| RESULTS
Due to an incomplete block of the IGN according to the EMG, this block was repeated with additional 5 mL mepivacaine in two volunteers.The subsequent EMG control confirmed a complete block of the IGN.

| Joint kinematics
No significant differences were observed in bilateral joint kinematics between the SGNtfl block and the control condition.Following the SGN block, bilateral hip, knee, ankle, and pelvis kinematics were similar to the control condition (Figure 2).Interestingly, no pelvis drop to the contralateral side was observed.Following the IGN block, at the maximal sitting position (50%), the pelvis was more flexed (average: 24 ± 15°) compared with the control condition (average: 17 ± 12°), p < 0.001 (Figure 2).Otherwise, no significant differences in bilateral joint kinematics were observed.The kinematics of the contralateral, unaffected hip, knee, and ankle remained similar to the control condition following each block.with a peak at 40%-60% (average maximum difference 0.7 × BW) (Figure 3).At the contralateral leg, the JRFs of the hip in the PD direction were significantly higher compared with the control condition at the beginning of the squat (about 10%-30% of the movement, average difference: 0.96 × BW) (Figure 3).The affected knee demonstrated significantly lower JRFs at the AP direction, more prominent at about 30%-70% of the movement (average maximum difference 0.96 × BW), whereas the contralateral knee was significantly higher in the AP direction during the sit-down phase of the movement (about 15%-40%, average maximum difference 1.43 × BW).The affected ankle demonstrated significantly lower JRFs in the PD direction, most prominent at about 30%-60% of the movement (average maximum difference 0.2 × BW), whereas the JRFs of the contralateral ankle were significantly higher in the PD direction during the mid-sit-down phase (about 10%-30%, average maximum difference 0.35 × BW) and the majority of sit-up phase (about 65%-100% average maximum difference 0.23 × BW) of the movement.

| Joint reaction forces
Following the IGN block, the JRFs of the hip at the affected leg in the PD direction were significantly lower than the control condition until about 70%, with the peak at 40%-60% of the movement (average maximum difference 0.66 × BW) and thereafter were similar to the control condition, whereas the contralateral leg demonstrated significantly higher JRFs in the PD direction at the majority of the movement (about 10%-90%, with a peak average difference of 0.57 × BW at about 40%-60% of the movement) (Figure 3).The JRFs of both the affected and contralateral knee and ankle demonstrated similar patterns to the SGN block.

| Dynamic balance performance
No significant differences were observed in the CoP range and SD between the SGNtfl block and control condition, whereas a slightly higher CoP velocity, was observed following the SGNtfl block (Table 1).During deep bilateral squats under SGN and IGN blocks, the CoP range and SD were significantly higher in ML direction, whereas all the other parameters were similar to the control condition (Table 1).

| DISCUSSION
The present study investigated the effects of an experimentally induced weakness of the gluteal muscles, through sequential blocks of the superior and IGN, on joint kinematics, JRFs, and dynamic F I G U R E 3 Summary of the hip, knee ankle joint reaction forces (JRF).A positive value indicated an anterior, inferior, and lateral direction of the force vector.AP, anteroposterior; BW, body weight; IGN, inferior gluteal nerve; JRF, joint reaction force; ML, mediolateral; PD, proximaldistal; SGNtfl, superior gluteal nerve to facial latae; SGN, superior gluteal nerve.balance performance during deep bilateral squats in healthy young adults.Hip, knee, ankle, and pelvis kinematics (other than a slight anterior pelvic tilt following IGN block) did not differ significantly following iatrogenic weakness of the gluteal muscles.More interestingly, a pelvic drop was not observed.The most important finding was the significant differences in JRFs following SGN and IGN blocks, with the affected hip, knee, and ankle demonstrating lower JRFs, whereas the contralateral joints demonstrated significantly higher JRFs than the control condition.When performing a deep bilateral leg squat under SGN and IGN blocks, the subjects demonstrated an increased CoP range and SD in ML directions suggesting that they swayed further and had less controlled behavior compared with the control condition.
Considering the complex interplay of the lumbo-pelvic-hip complex during the deep bilateral squat, understanding squat biomechanics in the normal state but also during muscle weakness/ fatigue is of fundamental importance for both achieving optimal muscular development and reducing training-related injuries.
Hemmerich et al. 41 described the three-dimensional kinematics of the hip, knee, and ankle joints in activities of daily life that required a higher range of motion such as squats, and concluded that an average maximum 95 ± 27°of hip flexion, 157 ± 6°of knee flexion and 38 ± 6°o f the ankle flexion were required to perform bilateral deep squats.Similar kinematics, except from lower knee flexion ankle (average 120 ± 14°) were observed in the control condition (before nerve blocks) in this study.Interestingly, hip, knee, and ankle kinematics did not differ following SGN and IGN block suggesting that compensatory changes have occurred.Following IGN block the pelvis demonstrated an increased anterior pelvic tilt which can be explained by the weakness of the gluteus maximus muscle.However, following the SGN block, no lateral pelvic drop was observed suggesting that patients with acute weakness of the hip abductors could compensate adequately and perform deep bilateral squats with similar kinematics as in the normal state.As no superficial electromyography was performed it remains unclear which muscles increased their activity to compensate for the iatrogenic weakness of hip abductors and gluteus maximus.However, it can be assumed that core activation might have contributed to the compensatory changes observed, as several studies highlighted the enhanced activity of hip muscles under core activation conditions (reduced lateral pelvic tilt and quadratus lumborum activity). 42,43Furthermore, as the upper portion of the gluteus maximus is predominantly functioning as a hip abductor and external rotator, 44,45 it may have contributed to the unaltered kinematics observed following the SGN block.
Weakness of gluteal muscles might alter the magnitude and direction of JRFs of the hip due to compensatory mechanisms, specifically an increase of muscle forces in other synergistic muscles. 46Tateuchi et al. 9 using individualized, scaled musculoskeletal models of 10 healthy volunteers, simulated mild (15%) and severe (30%) muscle weakness during squats and reported an increase of the anterior hip JRF by 18.3% during a severe weakness of the inferior gluteus maximus and 27.8% during a severe weakness of inferior gluteus maximus/deep external rotators of the hip, due to compensatory increase in quadriceps, hamstrings, and superior gluteus maximus, whereas the medial hip JRF decreased in the inferior gluteus maximus/deep external rotators weakness model.In this study, no change of the hip JRFs in AP and ML direction was observed, but a decrease of the superior JRF following SGN and IGN block in the affected hip and superior overloading of the contralateral unaffected hip.These above-mentioned differences might be attributed to the fact that the entire inferior and superior portion of the gluteus maximus was weakened through the IGN block or due to the experimental model, as different simulated and actual in vivo compensatory changes might have occurred.At the affected knee the JRFs decreased in the AP direction, whereas at the contralateral knee increased, suggesting an overload of the patellofemoral joint, especially at the sit-down phase of the squats.In a recent metanalysis, Pereira et al. 47 reported that all squat exercises cause tension overload in the patellofemoral joint, especially with a knee flexion between 60°and 90°.In accordance with the literature, knee flexion of 60°-90°was observed between 15% and 75% of the movement in this study, and the JRFs at the patellofemoral joint increased from 1.3 × BW to 5.1 × BW (392% increase) at the normal state.Following SNG as an IGN block, the JRFs at the contralateral patellofemoral joint increased with an average maximum difference of 1.43 x BW (133% increase) compared with the control condition.
T A B L E 1 Summary of the dynamic balance performance.At the affected ankle, JRFs decreased in the PD direction, whereas at the contralateral ankle increased implying a superior overload of the ankle joint.These results imply that a weakness of both hip abductors and gluteus maximus could lead to an overloading of the contralateral hip, patellofemoral, and ankle joint.Physicians and physiotherapist training patients with gluteal muscle weakness should be aware of those changes and focus on strengthening the core muscles such as the internal/external oblique to prevent overloading of the contralateral lower limb joints. 42,43cent evidence suggests that a translation of CoP location might have an important influence on muscle activity.Kitamura et al., 48  The present study should be interpreted in light of its potential limitations.The most obvious drawback was that acute muscle weakness was stimulated with a nerve block, which might not represent the condition of patients with chronic muscle weakness/insufficiency, pain, or fatigue which might develop other compensation mechanisms than the ones observed in the current study.However, this is the first study to report joint kinematics, JRFs, and dynamic balance performance following an acute, iatrogenic muscle weakness and provided significant insights on the squat performance changes and compensatory mechanisms.Certainly, further studies are needed to evaluate the compensation mechanisms in chronic states.Although a previously validated musculoskeletal modeling was used to predict the JRFs during squatting, the calculated JRFs could differ from the in vivo JRFs due to calculation errors.Finally, surface electromyography was not performed during squats and therefore the exact muscle groups which were compensatory activated following nerve blocks could not be reported.
In conclusion, the current study is the only one available in the literature investigating hip, knee, ankle, and pelvis kinematics, JRFs, and dynamic balance performance during deep bilateral squats at the normal state and following sequential blocks of the superior and IGN.
Although joint kinematics did not differ significantly following iatrogenic weakness of gluteal muscles, significant differences were observed in JRFs with overloading of the contralateral hip, patellofemoral, and ankle joints.Furthermore, an increased CoP range and SD in the ML direction were observed, suggesting that the subjects swayed further and had less controlled behavior compared with the control condition.The results of this study imply that squat performance changes significantly following a gluteal muscle weakness and should be considered when assessing and training athletes or patients with these injuries.
years, height of 1.75 ± 0.5 (1.63-1.86)m, body weight (BW) of 66.5 ± 12.7 (61-95) kg, and body mass index (BMI) of 21.4 ± 2.7 (16.4-27.4)kg/m 2 participated in the pre study.All the patients were students at a local university and had a Tegner score of 7 points.Inclusion criteria comprised healthy, adult volunteers with no history of hip or low back pain, previous hip surgery, neuromuscular disorders, and pregnancy.
No significant differences were observed regarding JRFs following the SGNtfl block and control condition on either leg.However, the JRFs were significantly different following the SGN and IGN block at both affected and contralateral legs.Specifically, following the SGN block at the affected leg, the JRFs of the hip at PD direction were significantly lower than the control condition at the beginning of the squats (average difference 0.45 x BW), then were similar at about 15%-25% of the movement and remained less than the control condition throughout the movement F I G U R E 2 Summary of the hip, knee ankle, and pelvis kinematics.A positive value indicated a flexion, abduction, and external rotation of the affected joint.IGN, inferior gluteal nerve; SGNtfl, superior gluteal nerve to facial latae; SGN, superior gluteal nerve.DIMITRIOU ET AL.| 167