Three dimensional kinematics of visually classified lower extremity movement patterns during a single leg squat among people with chronic hip joint pain.

Abstract

Objective:

Describe the proportional occurrence of visually determined, lower extremity movement patterns (dynamic knee valgus [DKVal], neutral [NEU], dynamic knee varus [DKVar]) during a single leg squat (SLSquat) among patients with chronic hip joint pain (CHJP). Compare 3D hip and pelvic kinematics among the categories and determine whether within-session movement pattern changes are possible among those who demonstrate DKVal or DKVar.

Design:

Cross-sectional.

Setting:

Movement science laboratory.

Participants:

36 patients with CHJP (18 to 40 years).

Main Outcome Measures:

Visually-based classification of lower extremity movement and 3D kinematic angles of hip and pelvis during SLSquat, performed under usual (self-selected) and modified (therapist instruction) conditions.

Results:

Based on visual appraisal, 14 patients demonstrated DKVal, 22 demonstrated NEU and none demonstrated DKVar. Those with DKVal demonstrated greater hip adduction (23.5±5.7° vs. 16.0±5.7°, p<0.001) and internal rotation (7.4±7.1° vs. 1.6±7.0°, p=0.023) than those with NEU. Compared to the usual condition, the DKVal group demonstrated significant decrease in hip adduction (23.5±5.7° vs. 20.9±5.8°, p=0.001) and internal rotation (7.4±7.1° vs. 5.3±7.8°, p=.050) in the modified condition.

Conclusions:

Patients with CHJP demonstrated 2 movement patterns, DKVal and NEU. Compared to NEU, those with DKVal demonstrated greater hip adduction and internal rotation motion and were able to make small modifications to their movement pattern with therapist instruction.

INTRODUCTION

Poor lower extremity movement patterns demonstrated during daily and athletic tasks have been associated with musculoskeletal pain, dysfunction and injury. Dynamic knee valgus, also referred to as medial collapse, has been associated with: noncontact anterior cruciate ligament (ACL) injuries (Hewett et al, 2005); patellofemoral pain (Bley et al, 2014; Nakagawa, Moriya, Maciel, and Serrao, 2012; Powers, 2003; Salsich, Graci, and Maxam, 2012); iliotibial band syndrome (Noehren, Davis, and Hamill, 2007); and femoroacetabular impingement (Diamond et al, 2017; Kumar et al, 2014). Dynamic knee varus or varus thrust, has been associated with early and established knee osteoarthritis (Fukutani et al, 2016; Lo, Harvey, and McAlindon, 2012). Recent studies have shown that strategies to modify poor lower extremity movement patterns may result in reduced symptoms (Graci and Salsich, 2014; Harris-Hayes et al, 2018; Noehren, Scholz, and Davis, 2011; Salsich et al, 2018) and improved function (Harris-Hayes et al, 2018; Noehren, Scholz, and Davis, 2011; Salsich et al, 2018), suggesting that poor movement patterns are modifiable treatment targets for existing musculoskeletal disorders and potentially injury prevention. Therefore, classifying lower extremity movement patterns may provide an important component in the clinical evaluation of musculoskeletal disorders and may assist in determining the best treatment approach for patient subgroups.

Visual assessment of lower extremity movement patterns demonstrated during the performance of daily tasks may be a useful clinical tool to classify poor movement patterns. Previously, we described a clinical method using visual assessment of the lower extremity during a single leg squat, and assessed the tester reliability in identifying poor movement patterns of the lower extremity (Harris-Hayes et al, 2014). Briefly, we used the observed change in the knee frontal plane projection angle (FPPA), represented by the angulation between the lower leg and thigh, during the performance of the single leg squat. Three categories of lower extremity movement were defined: neutral (NEU), dynamic knee valgus (DKVal) and dynamic knee varus (DKVar). This classification method is inexpensive, easy to administer in a clinical or athletic setting and can be performed, reliably by both expert and novice clinicians after a short training period (Harris-Hayes et al, 2014). Our previous work showed that we could identify the three categories of lower extremity movement patterns among asymptomatic participants (Harris-Hayes et al, 2014), however little is known about the type and frequency of movement patterns demonstrated by people with chronic hip joint pain (CHJP).

Common hip joint disorders that contribute to CHJP include femoroacetabular impingement, developmental dysplasia of the hip and labral tears. The literature related to CHJP has focused primarily on the presence of abnormal bony morphology with minimal attention given to the presence of poor movement patterns. However, it is likely that both contribute to altered stresses on the hip joint structures, potentially leading to injury and subsequent pain. Preliminary work using 3D kinematic analysis has suggested that excessive movement of the hip joint in the frontal and transverse plane may be associated with diminished articular cartilage integrity (Kumar et al, 2014) and lower hip-specific function among those with CHJP (Harris-Hayes et al, 2018), however little has been reported on the clinical assessment of lower extremity movement among people with CHJP. Visual assessment of hip adduction and rotation during a single leg squat can be challenging, because the pelvic landmarks often become obscured during the performance of the task. Given the interregional dependence between the knee and hip joint, it is possible that a simple assessment of the knee FPPA change during a functional task may serve as a proxy for the hip joint motion in the frontal and transverse planes. We therefore sought to determine if our proposed categories existed in people with CHJP and if the categories represented quantifiable differences among the subgroups in the 3D hip joint kinematics. Additionally, we wanted to know if a poor movement pattern could be modified with simple instruction, suggesting a potential treatment approach for those with poor movement patterns.

The primary purpose of our study was to describe the proportional occurrence of DKVal, NEU and DKVar movement patterns among our sample of patients with CHJP and to compare the 3D hip joint kinematics among the three movement pattern categories. We hypothesized that those with DKVal would demonstrate greater hip adduction and hip internal rotation compared to those with DKVar or NEU pattern. Those with DKVar would demonstrate greater hip abduction and hip external rotation compared to those with DKVal or NEU pattern. Because we were interested in modifying poor movement patterns as a potential treatment strategy, a secondary purpose was to determine the change in 3D kinematics of the hip when those with a poor movement pattern were provided specific cues to mimic the neutral pattern. Based on our clinical experience, we expected those who initially demonstrated a poor movement pattern, DKVal or DKVar, would be able to modify the movement pattern after receiving instruction.

METHODS

The participants in this study were a subset of a larger cohort parent study designed to assess potential risk factors for CHJP. We recruited people, aged 18-40, with deep hip joint or anterior groin pain. To be included in the study, their pain had to be present greater than 3 months, rated at least 3/10 on a numerical pain scale and be reproduced with the Hip Flexion, Adduction, Internal Rotation (FADIR) Impingement Test, indicating the hip joint as the source of pain (MacDonald, Garbuz, and Ganz, 1997). Exclusion criteria were body mass index (BMI) greater than 30 kg/m², previous hip fracture or surgery, or neurological involvement that would influence coordination or balance during testing. We excluded those with higher BMI to limit the effect of soft tissue artefact on the 3D kinematic measures. The study was approved by the Human Research Protection Office at Washington University School of Medicine and all participants signed an informed consent statement.

Prior to testing, all participants completed questionnaires including demographic information, Modified Harris Hip Score (MHHS) (Byrd and Jones, 2000) and UCLA Activity Scale (Amstutz et al, 1984). The MHHS is a patient-reported outcome measure to assess hip-specific pain and function. The UCLA is a self-report measure of physical activity level. Following questionnaire completion, participants completed a 5 minute warm-up using a treadmill or stationary bike and participated in a focused clinical examination to confirm the presence of intra-articular hip pain using the FADIR test. The FADIR test has been shown to have high sensitivity for detecting intra-articular hip disorders (Martin and Sekiya, 2008; Narvani et al, 2003).

We captured three-dimensional hip, pelvis and knee kinematic data during SLSquat movements using an 8-camera motion capture system (Vicon Nexus, Los Angeles, LA) sampling at 120 Hz. Each movement trial was concurrently recorded using a digital camera (Cannon Powershot, Cannon, Inc, USA). Retro-reflective markers were adhered over bony landmarks of the pelvis and lower extremity. Tracking markers included rigid 4-marker clusters secured laterally on the distal thigh and distal shank, and individual markers placed bilaterally over anterior superior iliac spine, posterior superior iliac spine, calcaneus, peroneal tuberosity and the cuboid. Calibration markers were placed bilaterally on the greater trochanter, medial and lateral femoral epicondyle, medial and lateral malleolus, and first and fifth metatarsal heads. A standing calibration trial was collected first, followed by the SLSquat trials.

Each participant performed two conditions of SLSquat, usual followed by modified. The participant began with their feet hip-width apart and arms across their chest. They were instructed to shift their weight to the painful limb, bend their opposite knee behind them and squat as they normally would. The examiner also demonstrated the movement to the participant. During the demonstration the examiner stood to the side of the participant to avoid biasing the movement. Three trials of the usual condition were then collected. Next, the modified condition was performed. The participant’s visually classified movement pattern was determined after the laboratory testing session, therefore we asked all participants to perform the modified condition. The examiner cued the participant to perform the SLSquat as before. However, during these trials, the examiner stated, “The following cues may or may not pertain to you. We are giving the same instruction to all study participants to see how each person responds. During the performance of these trials, keep your trunk upright and avoid side bending or twisting, keep your knee over the foot, and do not allow the knee to go in (medial) or out (lateral).” The examiner demonstrated correct and incorrect performance of the modified condition. For each condition, the participants practiced before formal testing commenced. Three trials were then collected. For a trial to be valid, the participant must achieve approximately 60° of knee flexion and maintain their balance. If the participant did not reach at least 60° of knee flexion, as judged visually by the examiner, they were instructed to increase the depth of the squat. It was considered a loss of balance if the participant: 1) placed their untested limb on the ground before completion of the movement; 2) demonstrated extraneous upper extremity movement; 3) demonstrated a trunk lean that resulted in excessive motion of the untested limb; or 4) moved the stance limb by sliding, hopping or twisting the foot.

We used Visual 3D software (C-Motion Inc, Germantown, MD) to process kinematic data. Marker trajectories were low-pass filtered using a fourth-order Butterworth filter with a 6-Hz cutoff frequency. A 6-degrees-of-freedom model was used for the pelvis and thigh. The pelvis was defined using the Codamotion model (Charnwood Dynamics Ltd, Leicestershire, UK). The thigh was defined by a functional hip joint center (Schwartz and Rozumalski, 2005) and the femoral epicondyle markers. The shank was defined by the femoral epicondyle markers and malleoli markers. For the hip and knee, we used the Cardan sequence of flexion/extension, adduction/abduction, internal rotation/ external rotation and for the pelvis we used internal/rotation, adduction/abduction and flexion/extension (Baker, 2011). Hip angles were defined as thigh relative to pelvis, knee angles were defined as shank relative to the thigh and the pelvis angles were calculated relative to the lab coordinate system. The primary variables of interest were frontal, and transverse plane angles of the hip joint at the point of maximal hip flexion, denoting the maximal depth of the SLSquat. Frontal and transverse plane angles of the pelvis and knee were analyzed for descriptive purposes. Average 3D angles for each participant were included in the group mean. Using the described methods to assess 7 asymptomatic participants, our within-session reliability was excellent for all hip joint variables (ICC_3,3: > 0.94). Standard error of measurements (SEMs) for hip flexion, adduction and internal rotation was 2.3°, 0.93°, 0.66° respectively.

Referencing methods previously reported (Harris-Hayes et al, 2014), the examiner (MHH) observed the digital videos, without the assistance of measurement software, to classify the sample into NEU, DKVal and DKVar groups. To mimic clinical practice, the videos were observed at normal speed, without using the slow motion feature or pausing the video. To determine the classification, the frontal plane projection angle (FPPA) of the knee was visually assessed at the initiation of descent and at the maximum depth of the SLSquat. We defined DKVal as a difference between the initial and final FPPA greater than 10° with the knee moving toward the midline of the body, DKVar as a 10° difference with the knee moving away from the midline of the body and NEU if the difference was not greater than 10° (Figure 1). Expert clinicians using the described method to assign movement pattern categories during a SLSquat demonstrated high intertester reliability (weighted kappa 0.90, 95% CI 0.77-1.00) and categories assigned using visual assessment were in excellent agreement with objective 2D measures of the FPPA (Harris-Hayes et al, 2014). To achieve this level of reliability and agreement, training included review of a written procedures manual and practice observing 8-10 videos with immediate feedback (Harris-Hayes et al, 2014).

Figure 1.

Visual assessment for group classification.

FPPA difference between starting (A,C,E) and ending (B,D,F) positions was used to classify movement pattern. FPPA difference < 10° was defined the neutral movement pattern (A,B). FPPA difference > 10° was defined the dynamic knee valgus (C,D) movement pattern. FPPA difference < 10° was defined the dynamic knee varus (E,F) movement pattern.

No participants demonstrated the DKVar pattern. Therefore, statistical analysis was performed using the two observed subgroups, DKVal and NEU. The Shapiro-Wilk test was used to confirm normal distribution of data, and Levene’s test was used to confirm equality of variance. For between-group analysis, we used independent-samples t-test and the Mann-Whitney U test for continuous and ordinal data (i.e. UCLA; Average, and Worst pain), respectively. For within-group analysis, we used a paired-samples t-test to compare kinematics in each group during the usual versus modified SLSquat. We analyzed the data using SPSS 21.0 (IBM Corporation, Armonk, New York, USA) with the threshold for significance set at p ≤ 0.05.

RESULTS

We enrolled 40 participants. Three participants did not provide permission for digital videos and videos of one participant were lost to technical malfunction. Among the remaining 36 participants, we classified 14 as DKVal and 22 as NEU using visual assessment. The two groups were demographically similar, however, more of those in the DKVal group reported higher pain intensity when asked to rate their “worst pain in the previous week” (Table 1).

Table 1.

Descriptive Data for Dynamic Knee Valgus (DKVal) and Neutral (NEU) Groups.

	DKVal N=14	NEU N=22	P Value
Age, years*	28.0 ± 4.1	28.2 ± 5.6	0.895
Sex	0M: 14F	6M: 16F	----
BMI, kg/m2*	24.4 ± 3.5	24.2 ± 2.8	0.870
MHHS, %*‡	75.9 ± 8.9	80.6 ± 13.7	0.268
UCLA†§	8 (4 – 10)	9.5 (3 – 10)	0.460¶
Pain duration, years*	4.5 ± 3.5	2.8 ± 3.2	0.163
Average pain†ǁ	4 (1 – 8)	3 (1 – 8)	0.438¶
Worst pain†ǁ	7 (3 – 10)	5 (2 – 10)	0.034¶

Abbreviations: DKVal, Dynamic knee valgus; NEU, neutral; MHHS, Modified Harris Hip Score; UCLA, University of California, Los Angeles activity scale

^*Values are means ± SD.

^‡100 = no disability.

^†Values are medians (range).

^§UCLA =; 1 = totally dependent, 10 = regularly participates in impact sports.

^¶P value based on Mann-Whitney U test.

^ǁSelf-reported pain intensity in the previous week, per Numerical Pain Scale, 0=no pain, 10=worst imaginable.

Figure 2 illustrates between-group differences in 3D hip and pelvic angles for the usual condition. Table 2 provides groups comparisons for all kinematic variables. The DKVal group demonstrated significantly greater hip adduction (23.5±5.7° vs.16.0±5.7°, p < 0.001), hip internal rotation (7.4±7.1° vs. 1.6±7.0°, p = 0.023) and less knee internal rotation (−0.2±5.1° vs. 5.6±6.1°, p = 0.006) compared to the NEU group. Time to complete the motion, and all other angles did not differ between groups.

Figure 2.

Mean hip and pelvis angles at maximum squat depth in the usual condition.

Error bars indicate 1 SD. Pelvic obliquity: (+) hip hike / (−) hip drop on non-weight bearing limb. Pelvic rotation: (+) pelvis rotating away / (−) rotating toward the weight bearing limb. *Indicates significant differences between those classified with Dynamic Knee Valgus and Neutral movement patterns, p ≤ 0.05.

Table 2.

Group comparisons of 3D kinematics between participants who demonstrated DKVal and NEU movement patterns during Usual condition.

	DKVal N=14	NEU N=22	Mean Difference (95% CI)	Effect size	P Value
Hip Flexion°*	69.3 ± 10.4	69.0 ± 16.6	0.3 (−9.8, 10.4)	0.0	0.951
Hip Adduction°*	23.5 ± 5.7	16.0 ± 5.7	7.5 (3.6, 11.4)	1.1	<0.001
Hip Internal Rotation°*	7.4 ± 7.1	1.6 ± 7.0	5.8 (0.86, 10.7)	0.8	0.023
Pelvic anterior tilt°*ǁ	34.9 ± 6.9	33.9 ± 12.0	0.9 (−5.4, 7.3)	0.1	0.766
Pelvic Obliquity°*ǁ	−6.8 ± 5.5	−5.9 ± 4.9	−0.9 (−4.5, 2.6)	0.2	0.592
Pelvic Rotation°*	−3.4 ± 4.4	−1.5 ± 6.0	−1.9 (−5.7, 1.9)	0.4	0.316
Knee Flexion°*	65.7 ± 8.9	67.6 ± 6.6	−1.9 (−7.2, 3.3)	0.3	0.459
Knee Adduction°*	2.2 ± 4.2	4.8 ± 4.6	−2.7 (−5.8, 0.5)	0.6	0.091
Knee Internal Rotation°*	−0.2 ± 5.1	5.6 ± 6.1	−5.9 (−5.7, 0.4)	0.9	0.006
Time (s)	2.9 ± 0.9	3.1 ± 1.1	−0.3 (−1.0, −0.5)	-	0.486

Abbreviations: DKVal, Dynamic knee valgus; NEU, neutral; Time, Time to complete motion; s, seconds

^*Values are means ± SD assessed at peak hip flexion.

^ǁPelvic obliquity: (+) hip hike / (−) hip drop on non-weight bearing limb. Pelvic rotation: (+) pelvis rotating away / (−) rotating toward the weight bearing limb.

Table 3 provides within-group comparisons for usual and modified conditions among those with DKVal. Compared to the usual condition, the DKVal group demonstrated a significant decrease in hip adduction (23.5±5.7° vs. 20.9±5.8°, p = 0.001) and internal rotation (7.4±7.1°vs. 5.3±7.8°, p = 0.050) in the modified condition. Time to complete the motion, and all other angles did not differ between conditions (Table 3). The NEU group demonstrated no differences between the modified and usual conditions (all p > 0.129).

Table 3.

Within-group comparison of 3D kinematics between the usual and modified performance of the single leg squat among participants who demonstrated a DKVal.

	Usual N=14	Modified N=14	Mean Difference (95% CI)	Effect size	P Value
Hip Flexion°*	69.3 ± 10.4	65.1 ± 13.8	4.2 (−2.2, 10.5)	0.3	0.177
Hip Adduction°*	23.5 ± 5.7	20.9 ± 5.8	2.5 (1.3, 3.7)	0.4	0.001
Hip Internal Rotation°*	7.4 ± 7.1	5.3 ± 7.8	2.0 (−.0, 4.0)	0.3	0.050
Pelvic anterior tilt°*ǁ	34.9 ± 6.9	33.2 ± 6.9	1.7 (−1.2, 4.6)	0.2	0.234
Pelvic Obliquity°*ǁ	−6.8 ± 5.5	−6.1 ± 4.5	−0.7 (−2.2, 0.8)	0.1	0.347
Pelvic Rotation°*	−3.4 ± 4.4	−2.0 ± 3.6	−1.4 (−3.0, 0.2)	0.3	0.087
Knee Flexion°*	65.7 ± 8.9	63.2 ± 8.8	2.5 (−0.6, 5.5)	0.3	0.105
Knee Adduction°*	2.2 ± 4.2	2.3 ± 4.2	−0.2 (−1.6, 1.3)	0.0	0.794
Knee Internal Rotation°*	−0.2 ± 5.1	0.9 ± 5.9	−1.1 (−3.3, 1.1)	0.2	0.294
Time (s)	2.9 ± 0.9	3.4 ± 1.4	−0.5 (−1.0, −0.1)	-	0.090

Abbreviations: DKVal, Dynamic knee valgus; Time, Time to complete motion; s, seconds

^*Values are means ± SD assessed at peak hip flexion.

^ǁPelvic obliquity: (+) hip hike / (−) hip drop on non-weight bearing limb. Pelvic rotation: (+) pelvis rotating away / (−) rotating toward the weight bearing limb.

DISCUSSION

We used our previously reported methods to classify lower extremity movement patterns in patients with CHJP and determined that patients with CHJP demonstrated two of the three proposed patterns, NEU and DKVal. In our sample, patients in the DKVal category demonstrated greater hip adduction and internal rotation than those in the NEU category, indicating that our visual assessment methods using the change in FPPA of the knee can identify different movement patterns of the hip joint. Not surprisingly, patients with DKVal demonstrated less knee internal rotation compared to those with NEU pattern. Also, we noted we that with instruction, those who demonstrated DKVal were able to modify their 3D hip joint kinematics within one session, supporting the theory that poor movement patterns are modifiable and may be an important treatment target. Our findings suggest that visual assessment of lower extremity movement patterns may be a useful evaluation tool to identify subgroups based on movement patterns that may guide treatment strategies for patients with CHJP.

Given our clinical observations and our previous published work (Harris-Hayes et al, 2014), we expected to observe all three movement pattern categories among our patients with CHJP, however we noted that none of the patients with CHJP demonstrated a DKVar pattern. We are hesitant to state that the DKVar is not present among patients with CHJP, however our findings suggest that the occurrence may be low in this young population. In our previous study (Harris-Hayes et al, 2014), we classified lower extremity movement patterns among asymptomatic people and found only three of the 30 participants demonstrated DKVar. To our knowledge no other classification system has included all three movement pattern categories. Although DKVar has been identified and associated with lower extremity mechanical disorders (Fukutani et al, 2016; Lo, Harvey and McAlindon, 2012), this pattern has not been the focus of previously published classification systems using visual appraisal (Ageberg et al, 2010; Chmielewski et al, 2007; Whatman, Hume, and Hing, 2013). The previous studies (Ageberg et al, 2010; Chmielewski et al, 2007; Whatman, Hume, and Hing, 2013) have focused primarily on differentiating poor movement quality from good movement quality. The definition of good movement quality varies. Ageberg et al. (2010) visually classified 25 participants as either having knee-medial-to-foot indicating poor movement quality or knee-over-foot position indicating good movement quality during a single leg squat. In their report, they stated that one participant, whose knee moved lateral to the foot, was excluded because they did not meet the criteria for either good movement quality or poor movement quality. Motion where the knee moves lateral to the foot would be consistent with our DKVar category. Whatman, Hume, and Hing (2013) rated movement quality as a “yes” or “no” to the question, “Does the patella move medial to the 2^nd toe?” This would suggest that a movement where the patella moved lateral to the foot may have been classified as good movement quality. Chmieleski et al. (2007) defined good movement quality as “no deviation from neutral alignment” and poor movement quality as “movement out of a neutral position and/or segment oscillation”, indicating that deviations consistent with DKVal and with DKVar would be classified into the category of poor movement quality. The low proportional occurrence of the DKVar movement pattern may be due to the task performed. We assessed the SLSquat and quantified joint angles when the knee was in its most flexed position. A movement pattern similar to DKVar has been reported during gait when the knee is in a relatively less flexed position (Barrios, Crossley, and Davis, 2010; Chang et al, 2013) Nevertheless, DKVar has been previously observed in a small group of people. Our classification system expands on the previous work by including the DKVar category. Although the occurrence may be low, the DKVar movement pattern may represent a distinct movement pattern that should be differentiated from the DKVal and NEU categories.

Sophisticated, 3D kinematic testing remains the gold standard to quantify intersegmental motion during movement tasks. However, visual assessment may be sufficient to subgroup patients during clinical assessment or screening, where kinematic testing is time- and cost-prohibitive. To our knowledge, no other studies describe visual assessment of lower extremity movement patterns among people with CHJP. Previous studies have used similar, but slightly different methods to classify lower extremity movement patterns during a single leg squat among healthy, asymptomatic participants (Ageberg et al, 2010; Chmielewski et al, 2007; Whatman, Hume, and Hing, 2013). Despite assessing asymptomatic participants, movement patterns consistent with our categories of DKVal and NEU were observed, and between-group differences were similar to those of our symptomatic cohort. In our sample, patients with DKVal demonstrated 7.5° greater hip adduction and 5.8° greater hip internal rotation compared to those in the NEU category. These differences have large effect sizes, 1.1 and 0.8 respectively, suggesting our visual classifications using the knee FPPA can differentiate between movement patterns of the hip joint. Ageberg et al. (2010) compared asymptomatic people who demonstrated the knee-medial-to-foot to those with knee-over-foot position during a single leg squat and found the former demonstrated greater hip internal rotation (5.8°). Also using the knee-foot-position criteria, Whatman, Hume, and Hing (2013) classified lower extremity movement among 23 young, healthy athletes. They reported greater hip adduction (6.0°) and hip internal rotation (4.0°) among those rated as having poor versus good movement quality (Whatman, Hume, and Hing, 2013) Although visual appraisal criteria to distinguish movement categories differed slightly among the previous studies and our own, angle differences between movement categories were comparable (Harris-Hayes et al, 2014).

Excessive hip adduction (Bley et al, 2014; Diamond et al, 2017; Kumar et al, 2014; Nakagawa, Moriya, Maciel, and Serrao, 2012; Noehren, Davis, and Hamill, 2007) and internal rotation (Bley et al, 2014; Nakagawa, Moriya, Maciel, and Serrao, 2012) motion have been implicated in a number of lower extremity musculoskeletal disorders, including CHJP. Based on our study, however, we are unable to make definitive statements about the relationship among our proposed categories of movement patterns and musculoskeletal pain or injury. It is interesting to note that patients with DKVal in our study reported higher pain intensity compared to NEU. Although this difference was statistically significant, we are cautious to state a direct relationship exists between pain intensity and the movement pattern, due to our small sample. The relationship between pain and movement patterns is complex. Clearly in some instances, the presence of pain may result in a poor movement pattern to avoid further tissue irritation and pain provocation. However, poor movement patterns also exist in the absence of pain. Previous studies (Ageberg et al, 2010; Chmielewski et al, 2007; Whatman, Hume, and Hing, 2013), including our own (Harris-Hayes et al, 2014) have reported that a DKVal movement pattern exists even among those who are asymptomatic, suggesting that pain alone may not explain the presence of poor movement patterns. We believe a poor movement pattern may be a risk factor for developing CHJP, but that other factors such as poor muscle performance or abnormal bony morphology may also play a role. Large, prospective studies are needed to better understand the relationship among movement pattern classification, pain conditions and risk for future injury. Our relatively simple classification methods could facilitate this type of investigation by providing an inexpensive, reliable tool that may be used in the clinical and athletic setting.

Our study shows that it is possible to improve a lower extremity movement pattern within one session. The change in motion was greater than the SEM (1°) of our methods, however would be considered small; a 3° reduction in hip adduction (effect size of 0.4) and 2° in internal rotation (effect size of 0.3). Standard deviations of the joint angles, particularly for internal rotation are relatively large suggesting large variability in the patients’ motion and in the patients’ ability to change their motion. This variability may suggest differences in muscle performance capabilities across participants and may provide assistance in determining the most effective treatment approach. Smaller changes may indicate poor muscle performance limiting the patient’s ability to make changes in movement within this short time frame. Therefore, muscle strengthening exercises may be necessary to improve muscle performance. Those who are able to modify the movement pattern in one session have sufficient force production capability in the hip musculature; therefore, frequent practice of the improvement movement pattern during daily tasks may be the most effective approach.

Recently, we reported on a movement pattern training approach to optimize biomechanics among patients with CHJP. We found that after completion of six-weeks of treatment, patients improved their movement pattern by reducing their hip adduction motion. Additionally, the patient’s ability to reduce hip adduction after treatment was associated with improvements in pain and function (Harris-Hayes et al, 2018). Similar approaches have been used in the treatment of patellofemoral pain syndrome. Previous work by Salsich et al. (2018) and Noehren, Scholz, and Davis (2011) has shown that with treatment focused on changing lower extremity movement patterns, people with patellofemoral pain syndrome can also decrease hip adduction and internal rotation. Importantly, participants in each of these studies reported significant improvements in pain and function after treatment. Our data, along with previous treatment studies, suggests that poor movement patterns can be identified with visual assessment and may represent a modifiable target for treatment for patients with CHJP.

Nevertheless, our study has limitations. Using video recordings to determine visually classified movement patterns may not directly reflect current clinical practice, however videos could be easily implemented in the clinic using readily available equipment such smartphones or tablets. Our study did not account for potential bony differences between groups. Previous work has shown that abnormal bony structure often exists in people with CHJP (Clohisy et al, 2011; Ito, Leunig, and Ganz, 2004) however the relationship between bony structure and specific movement patterns is unknown. These nuances are the focus of future work. Related to structural abnormalities, our proposed classifications are based on a change in the knee FPPA during the motion, therefore a person with an alignment of excessive knee valgus that does not change during the movement would be classified as neutral. For treatment decisions, structural factors would need to be considered in addition to movement assessment. More women than men participated in our study. While our sample is too small to draw conclusions, no men demonstrated a DKVal movement pattern, leading us to consider the potential for kinematic differences between sexes. Others have reported lower extremity kinematic differences between men and women (Graci, Van Dillen, and Salsich, 2012; Sakaguchi et al, 2014) and that is possibly the case here. Larger samples are needed to draw out these differences. Given our exclusion criteria, our methods may not be generalizable to those with BMI greater than 30kg/m².

We did not consider pelvic motion when visually classifying lower extremity movement patterns. We recognize that pelvic motion contributes to the 3D values of hip joint motion and is a limitation of our visual method. Our goal was to assess a relatively simple screening technique that may be used by those who may be novice to movement assessment. For a number of subjects, the pelvic landmarks, in particular, the anterior superior iliac spine (ASIS) were obscured at the lowest depth of the squat, therefore we focused our visual assessment on the knee FPPA. Interestingly, we found no differences in pelvic motion between the two groups. This was surprising, given the SLSquat requires relatively more pelvic control than activities that have double limb support. Other tasks such as a stair descent may provide better visualization of pelvic motion during a functional task. Finally, the specific portion of the SLSquat analyzed may differ slightly between the 3D kinematic assessment and visual assessment. The maximal hip flexion position was selected because this position is proposed to contribute to mechanical impingement between the femur and acetabulum. For 3D kinematic assessment, we were able to identify the specific point of maximum hip flexion. For the visual assessment, the examiner was trained to assess knee FPPA at the maximum depth of the squat, which served as a proxy for maximum hip flexion.

CONCLUSIONS

Patients with CHJP demonstrated two distinct lower extremity movement patterns; DKVal and NEU. No participant demonstrated DKVar. Using our methods of classifying movement during a SLSquat, we noted between-group differences in hip joint motion, specifically hip adduction and internal rotation motion. Compared to those with neutral movement, participants who demonstrated DKVal had greater hip adduction and internal rotation. Those with DKVal were also able to reduce hip adduction and internal rotation with instructions within a single session, however these changes were small. Classifications based on lower extremity movement patterns may provide an important assessment tool and may assist in determining individualized treatment approaches.