In-vivo Pelvic and Hip Joint Kinematics in Patients with Cam Femoroacetabular Impingement Syndrome: A Dual Fluoroscopy Study

Abstract

Femoroacetabular impingement syndrome (FAIS) may alter the kinematic function of the hip, resulting in pain and tissue damage. Previous motion analysis studies of FAIS have employed skin markers, which are prone to soft tissue artifact and inaccurate calculation of the hip joint center. This may explain why the evidence linking FAIS with deleterious kinematics is contradictory. The purpose of this study was to employ dual fluoroscopy (DF) to quantify in-vivo kinematics of patients with cam FAIS relative to asymptomatic, morphologically normal control participants during various activities. Eleven asymptomatic, morphologically-normal controls and seven patients with cam-type FAIS were imaged with DF during standing, level walking, incline walking, and functional range of motion activities. Model-based tracking calculated the kinematic position of the hip by registering projections of three-dimensional computed tomography models with DF images. Patients with FAIS stood with their hip extended (mean [95% confidence interval], −2.2 [−7.4, 3.1]°, flexion positive), whereas controls were flexed (5.3 [2.6, 8.0]°;p=0.013). Male patients with cam FAIS had less peak internal rotation than the male control participants during self-selected speed level-walking (−0.2 [−6.5, 6.1]° vs. −9.8 [−12.2, −7.3]°; p=0.007) and less anterior pelvic tilt at heel-strike of incline (5°) walking (3.4 [−1.0, −7.9]° vs. 9.8 [6.4, 13.2]°; p=0.032). Even during submaximal range of motion activities, such as incline walking, patients may alter pelvic motion to avoid positions that approximate the cam lesion and the acetabular labrum.

Introduction

Cam-type femoroacetabular impingement syndrome (FAIS) has been identified as an etiological factor in the development of hip osteoarthritis (OA) and represents a common cause of hip pain in young and otherwise healthy adults.¹ Cam FAIS morphology is described as a femoral head asphericity with reduced head-neck offset over the anterosuperior and anterolateral regions of the femoral head-neck junction and is thought to be more prevalent in males.² In addition to this morphology, patients with cam FAIS often report pain and reduced range of hip motion.^3,4 It is hypothesized that the cam lesion abuts abnormally with and begins to pivot about the acetabular rim and labrum after contact. This abnormal abutment may result in compensatory pelvic motion which may reduce impingement even during low range of motion activities.⁵ With high ranges of flexion and internal rotation, increased translations within the joint may occur as the aspherical femoral head begins to act as a cam within the acetabulum.^6,7 This hypothesis is supported by damage patterns observed in patients with cam FAIS.¹

Many studies have attempted to quantify the kinematics of cam FAIS to assess the relationship between abnormal femoral morphology and limited range of hip joint and pelvic motion; however, results to date have been somewhat inconsistent. The investigation of kinematics during gait has identified a decreased range of motion in one or more planes of motion in patients with cam FAIS.^8–11 Similarly, limited peak ranges of motion have been reported for patients with cam FAIS in all planes of motion,¹² but these conclusions have not been consistent across investigations.^9,13 In addition to reduced hip range of motion, several studies have identified reduced pelvic range of motion in patients with cam FAIS, primarily in the frontal plane.^10,14 However, the relationship between altered hip joint and pelvic motion patterns in patients with cam FAIS has yet to be well defined and the role morphological differences, including sex-based differences, play in hip function remains poorly understood.

To combat these issues, several studies have used computer simulations or cadaveric specimens to better understand motion restrictions in the setting of cam FAIS using participant-specific morphology of the femur and pelvis. Herein, high range of motion activities, such as the anterior impingement exam, can be recreated using a computer model in which direct impingement between the femur and pelvis serves as the cause of reduced range of motion.^15–17 However, these simulations do not incorporate pelvic motion and assume direct impingement between the femoral neck and acetabular bone which do not represent in-vivo motion patterns.⁷

Studying kinematic movement relative to the underlying three-dimensional anatomy (i.e. arthrokinematics) could identify the pathomechanics of FAIS. We previously developed and validated a dual fluoroscopy (DF) system to quantify in-vivo hip arthrokinematics, and found it to be accurate within 0.5 mm and 0.6°.¹⁸ DF has been used to measure arthrokinematics during clinical exams for patients with FAIS and during weight-bearing activities of daily living for control participants.^6,19 However, this technology has not been applied to compare hip arthrokinematics between patients with cam FAIS and controls during weight-bearing activities.

The purpose of this study was to employ DF to quantify in-vivo kinematics of patients with cam FAIS relative to asymptomatic, morphologically normal control participants during standing, weight-bearing activities of daily living, and unweighted functional activities. We hypothesized that patients with cam FAIS would have reduced peak hip joint angles and pelvic motion during gait activities.

Participants and Methods

Seven patients with cam FAIS were recruited from the clinic of an orthopaedic surgeon (SKA; Table 1). Diagnosis of cam FAIS was determined based on patient reported symptoms, positive clinical examinations (i.e. anterior impingement exam), and confirmation of cam morphology on radiographic images in the modified false profile, frog-leg lateral, and anteroposterior positions.²⁰ All patients were scheduled for femoral osteoplasty at the time of dynamic imaging. Eleven control participants were used for comparison (Table 1). These control participants had no history of hip pain. Controls were screened with an anterior-posterior (AP) radiograph prior to study inclusion.^19,21–23 The AP radiograph provided general screening of hip pathology, including acetabular dysplasia, Legg-Calvé-Perthes disease, and acetabular retroversion. Alpha angle measurements were obtained from the AP radiograph as well as from oblique-axial reformatted magnetic resonance images (MRI) that were acquired for another study that utilized the same control population.²¹ Here, an alpha angle larger than 55.5° resulted in exclusion from the control group (Table 1).²⁴ All patients and control participants were recreationally active, had no previous history of lower limb surgery, a body mass index (BMI) less than 30 kg/m², a lateral center edge angle between 20° and 45°, and no radiographic evidence of OA or other anatomical abnormalities, other than cam morphology for the patients. Each participant provided informed consent for this Institutional Review Board approved study. Participants were then imaged with computed tomography (CT) and DF to capture in-vivo hip kinematics. Since there was predominance for male patients, with only two female patients, the five male patients with cam FAIS were analyzed separately to remove sex as a confounding factor (Table 1).

Table 1:

Participant demographics, represented as mean (standard deviation) as applicable.

Demographic	Cam FAIS Patients	Control Participants	PVal	Male Cam FAIS Patients	Male Control Participants	PVal
Sex (n)	5 M, 2 F	6 M, 5 F	NA	5 M	6 M	NA
Age (years)	29 (7)	23 (2)	0.114	29 (6)	23 (1)	0.045
Height (cm)	179.1 (10.1)	173.3 (10.4)	0.448	183.6 (7.0)	182.0 (2.9)	0.613
Mass (kg)	78.9 (15.2)	63.8 (10.9)	0.033	84.0 (15.4)	71.4 (7.2)	0.106
BMI (kg/m2)	24.4 (3.2)	20.9 (1.8)	0.003	24.8 (3.7)	21.3 (1.6)	0.067
AA, Radiograph (°)	61 (15)	39 (4)	<0.001	59 (15)	40 (4)	0.015
AA, MRI (°)	61 (5)	40 (5)	<0.001	63 (4)	40 (6)	<0.001

M, Male; F, Female; NA, not applicable; BMI, Body Mass Index; AA, Alpha Angle; MRI, magnetic resonance imaging

The CT and DF imaging protocols have been previously described.^18,23 Briefly, CT images of the hip and distal femur were acquired with a SOMATOM Definition 128 CT scanner (Siemens AG, Munich, Germany). For the proximal femur and pelvis, images were acquired at 120 kVp, 1.0 mm slice thickness, and 200 to 400 mAs with variable fields of view due to participant size.²⁵ For the distal femur, images were acquired at 120 kVP, 3.0 mm slice thickness, and 150 mAs. The femur and pelvis were segmented and reconstructed from CT images (Amira, v5.6, FEI, Hillsboro, OR, USA).

Each participant performed activities of daily living, including standing with feet at hip-width and pointed forward, level walking at a standardized speed (1.3 m/s), level and incline (5°) walking at a self-selected speed, internal and external rotational pivots to end range of motion, a functional star-arc maneuver,²⁶ and unassisted abduction to approximately 45°. For the rotational pivots, study participants were individually positioned to result in the proper positioning at their end range of motion. The feet were placed parallel, roughly hip width apart, and generally symmetric. However, if necessary for data collection, the foot of the imaged hip may have been more anterior for external rotation or posterior for internal rotation. The star-arc maneuver, which is commonly used for functional hip joint center assessments in gait analysis, included five positions of hip flexion-extension and abduction-adduction followed by circumduction performed in a continuous manner. Each participant performed dynamic activities on an instrumented treadmill (Bertec Corporation, Columbus, OH, USA) with the hip of interest positioned in the combined field of view of the custom DF system (Radiological Imaging Services, Hamburg, PA, USA), which consisted of two pairs of x-ray emitters and image intensifiers mounted on independent bases and arranged with an overlapping field of view (Figure 1). Images were captured at 100 Hz while fluoroscopy settings ranged from 78–100 kVp and 1.9–3.2 mAs with camera exposures of 4.5–7.0 ms.

Figure 1:

Dual fluoroscopy of the left hip of a representative male participant during level walking on an instrumented treadmill. Image intensifiers (II) are positioned on the far side of the participant, while the beam emitters are in the foreground.

CT voxel intensities within each bone were used as input to model-based markerless tracking of the DF images.²⁷ Here, projections of each bone were manipulated in six degrees-of-freedom to simultaneously align with each frame from DF. The spatial position of each bone was tracked using bony landmarks over the length of each trial.¹⁸ Consistent with many gait analysis studies, the femur was tracked using landmarks of the center of the femoral head and the two femoral condyles, while the pelvis used bilateral landmarks of the anterior superior iliac spine (ASIS) and posterior superior iliac spine (PSIS). These landmarks, as well as the acetabular center, were identified semi-automatically in MATLAB (v9.3.0, MathWorks, Natick, MA, USA) using surface curvature generated in PostView (v2.1, University of Utah, Salt Lake City, UT).¹⁸ The acetabular joint center and femoral head center were found by fitting the articulating surfaces to spheres, the femoral condyles were identified as the medial and lateral apex of the condylar surfaces, the ASIS was isolated as the node furthest from the best-fit plane of the region identified using surface curvature, and the PSIS was selected based on the morphology of the spines of pelvic curvature.

For each activity, two trials were captured when possible, but activity level, allotted DF time, and image quality limited the ability to capture and analyze activities from some participants (Table 2). For the walking trials at a self-selected speed, the same treadmill speed was used for level and incline walking trials based on the preferred walking speed of the participant.^19,22 For all walking trials, a single gait cycle was evaluated. For rotational pivots and abduction, only the position of maximum range of motion was evaluated herein. For the functional star-arc activity, the five positions of the star (1, flexion; 2, flexion-abduction; 3, abduction; 4, extension-abduction; and 5, extension) and the range of motion during circumduction were evaluated while the participant balanced on the contralateral limb. Gait data was time-normalized from heel-strike to heel-strike; data from the functional star-arc activity was time-normalized based on the five peaks associated with each of the positions of the star and the start and end of circumduction to allow for comparison across participants. Kinematic data for the controls was previously published,^19,21–23 while all activities for the patients with cam FAIS were included herein for the first time.

Table 2:

Participants for each activity, listed by sex, with the basis for exclusion for each identified activity.

Activity	Cam FAIS Patients	Control Participants
Neutral Stance	5 M, 2 F	6 M, 5 F
Level Walk	5 M, 2 F	6 M, 5 F
Standardized Level Walk	4 M (IQ), 1 F (AL)	6 M, 5 F
Incline Walk	5 M, 2 F	6 M, 5 F
Rotation	5 M, 2 F	6 M, 5 F
Functional Star-Arc	4 M (IQ), 2 F	5 M (DF), 5 F
Abduction	4 M (IQ), 2 F	4 M (2 DF), 5 F

IQ, image quality: AL, activity level; DF, allotted DF time.

For this study, kinematic data were presented as raw joint angles between the local coordinate systems of the femur and the pelvis as well as angles relative to the standing position. Representation of hip kinematics relative to the standing position was considered so as to minimize the effect of anatomic variability on kinematic calculations; this is a reasonable approach given that model-based markerless tracking of DF images represents kinematic motion relative to the underlying hip anatomy from reconstructed CT images.¹⁹ This approach provided data on both in-vivo joint motion and the relative relationship between static (i.e. standing) and dynamic motions. Joint translation data were evaluated in the local coordinate system of the pelvis, relative to the static neutral position, such that values represented the relative movement between the femoral head center and the acetabular center. Pelvic tilt and obliquity were calculated respectively relative to the projection of the vertical axis of the lab onto the sagittal and coronal planes of the local coordinate system of the pelvis. The lateral axis was projected onto the horizontal plane of the lab to measure pelvic rotation; all pelvic rotation angles were represented relative to the average pelvic rotation during the level walking activity at a self-selected speed as this represented a consistent definition of neutral pelvic rotation that was independent of positioning within the dual-fluoroscopy system. Joint angles, joint translations, and pelvic rotation angles were calculated using a custom script in MATLAB.

Unless otherwise noted, all data were presented as mean [95% confidence interval]. Student’s t-tests were used to evaluate mean differences between patients with cam FAIS and control participants for both the entire cohort and only for the male participants, specifically (female patients were not analyzed separately due to the sample size). The Holm-Bonferroni method was used to correct for multiple comparisons, where corrections were applied across directions of motion. For time series data, the Benjamini and Hochberg method of false discovery rate was used to correct for non-independence. P-values greater than 1.0 after correction for multiple comparison were set to 1.0. Correlations were assessed between hip joint angles and pelvic rotation angles using the Pearson correlation coefficient which was used to qualitatively interpret the absolute value of R-values as negligible (0.00–0.30), weak (0.31–0.50), moderate (0.51–0.70), strong (0.71–0.90), or very strong (0.91–1.00).²⁸ All statistics were completed in MATLAB.

Results

Alpha angles on both the AP film and MR images were larger in the patient group; this was true when males and females were analyzed together and when only males were analyzed (Table 1). Patients had larger mass and BMI compared to the control group (Table 1). Male patients were significantly older and had larger BMI values compared to males in the control group (Table 1).

Standing

Patients with FAIS stood with their hip extended (−2.2 [−7.4, 3.1]°, flexion positive), whereas controls were flexed (5.3 [2.6, 8.0]°, p=0.013). Generally, both patients with cam FAIS and control subjects stood in slight abduction (patients, 0.8 [−3.1, 4.7]°; controls, 3.1 [1.5, 4.7]°) and internal rotation (patients, −4.0 [−14.7, 6.7]°; controls, −8.0 [−12.5, −3.4]°). Male patients with FAIS stood with their hip extended (−4.6 [−9.9, 0.6]°, flexion positive), whereas controls were flexed (5.3 [0.5, 10.0]°, p=0.015). Male patients with cam FAIS stood with neutral abduction-adduction (−0.1 [−5.7, 5.6]°), while male control subjects stood in slight abduction (3.6 [1.3, 5.9]°). For rotation, male patients with cam FAIS stood in a more neutral position (2.3 [−2.8, 7.4]°), while male control patients stood in slight internal rotation (−6.8 [−14.5, 1.0]°). All of the control participants stood in a position of anterior pelvic tilt (9.7 [6.5, 13.0]°, anterior tilt positive), while two males of the seven patients with cam FAIS stood in a position of posterior pelvic tilt which brought the mean for the FAIS group into less anterior tilt (5.0 [−2.6, 12.5]°), but group differences were not significant. For standing, there was a very strong correlation between flexion angle and pelvic tilt for patients with cam FAIS (r=0.948, p=0.003) and a strong correlation for control participants (r=0.720, p=0.038).

Gait Metrics and Hip Kinematics

No significant differences were observed in preferred walking speed or cadence during any of the gait activities. In particular, patients with cam FAIS preferred a walking speed of 1.29 [1.13, 1.45] m/s at 125 [118, 132] steps/min for level walking and 122 [113, 131] steps/min for incline walking; controls preferred a walking speed of 1.29 [1.22, 1.37] m/s at 124 [118, 131] steps/min for level walking and 120 [114, 126] steps/min for incline walking. Similarly, no significant differences were found in cadence or preferred walking speed for the male populations. The male patients with cam FAIS preferred a walking speed of 1.31 [1.09, 1.52] m/s at 124 [115, 132] steps/min for level walking and 123 [109, 136] steps/min for incline walking, while male control participants preferred a walking speed of 1.29 [1.14, 1.44] m/s at 119 [110, 127] steps/min for level walking and 115 [107, 123] steps/min for incline walking.

In general, patients with cam FAIS had increased peak extension, external rotation, and posterior pelvic tilt as compared to controls during level walking at a standardized speed as well as level and incline walking at a self-selected speed, but differences were not significant (Figure 2). During level walking, male patients with cam FAIS had less peak internal rotation (−0.2 [−6.5, 6.1]°) than male controls (−9.8 [−12.2, −7.3]°, p=0.007), but did not have greater peak external rotation (11.8 [2.9, 20.7]°) compared to male controls (1.9 [−3.8, 7.6]°, p=0.083). No differences were observed between patients with cam FAIS and control participants with regards to translations of the femur (Figure 2).

Figure 2:

Peak joint angles, joint translations, and pelvic rotation angles during self-selected speed level walk (top), standardized speed level walk (middle), and self-selected speed incline walk (bottom). The symbols indicate the mean, while the vertical bars represent the 95% confidence interval.

Compared to controls, patients with cam FAIS appeared in more external rotation and less flexion with less anterior pelvic tilt throughout the gait cycle, but these differences were not statistically significant (Figure 3). During inclined walking, male patients with cam FAIS had less anterior pelvic tilt at heel-strike (3.4 [−1.0, 7.9]°) than male controls (9.8 [6.4, 13.2]°, p=0.032).

Figure 3:

Hip joint angles and pelvic rotation angles for control participants and patients with cam FAIS during self-selected speed level walking. The solid line indicates the mean, while the semi-transparent band represents the 95% confidence interval.

During the functional star-arc activity, patients with cam FAIS had greater range of motion in abduction (33.9 [30.1, 37.6]°) when compared to controls (26.0 [22.9, 29.1]°, p=0.007). The same trend was observed when evaluating the male patients with cam FAIS (33.1 [26.9, 39.4]°) relative to the male controls (24.9 [20.9, 28.9]°, p=0.032).

No differences were observed in hip joint angles at the position of interest for the abduction activity or for the rotational pivots (Figure 4). The male patients with cam FAIS and control participants showed similar joint angles during the abduction and rotational motions (Figure S-1).

Figure 4:

Hip joint angles during abduction (left) and at the maximum range of motion for the internal (center) and external rotational (right) pivot for patients with cam FAIS and control subjects. The horizontal bars indicate the mean, while the vertical bars represent the 95% confidence interval.

Hip Kinematics Relative to Standing

Overall, patients with FAIS had less relative posterior pelvic tilt during the standardized speed level walk compared to controls (−1.0 [−3.8, 1.7]° vs. −4.4 [−6.0, −2.8]°; p=0.050). Relative to their standing position, male patients with FAIS patients had greater peak abduction compared to male controls during the self-selected speed level (7.8 [4.9, 10.6]° vs. 2.0 [0.6, 3.4]°; p=0.002) and incline walk (7.6 [3.4, 11.9]° vs. 2.3 [1.0, 3.6]°; p=0.017). Male patients with cam FAIS patients also were in greater relative abduction than male control participants at heel-strike (0.9 [−3.3, 5.1]° vs. −3.6 [−4.9, −2.2]°; p=0.045) and toe-off (6.8 [3.0, 10.6]° vs. 1.4 [−0.1, 3.0]°; p=0.012) of self-selected speed level walk and toe-off of standardized speed level walk (6.6 [2.4, 10.7]° vs. 0.8 [−2.3, 4.0]°; p=0.043).

Relative to standing and compared to control participants, patients with cam FAIS were in more abduction (27.3 [22.5, 32.1]° vs. 18.9 [14.3, 23.6]°; p=0.046), with more anterior pelvic tilt (6.6 [1.8, 11.4]° vs. 1.7 [0.2, 3.2]°; p=0.030) and upward obliquity (13.1 [10.5, 15.6]° vs. 10.2 [9.0, 11.5]°; p=0.033) during position 3 (abduction) of the functional star-arc activity. Patients with FAIS were also in less extension during position 5 (extension) when compared to controls (−2.1 [−7.9, 3.7]° vs. −9.6 [−12.5, −6.7]°; p=0.024).

Dynamic Pelvic Motion

Pelvic tilt was significantly different from standing for the control participants at the loading response, terminal stance to pre-swing, and terminal swing of level gait. However, significant differences in pelvic tilt were only observed during incline gait for patients with cam FAIS, predominantly during terminal stance and the swing phase (Figure 5). During both level and incline walking, dynamic obliquity was significantly different from pelvic obliquity in standing for less of the gait cycle for patients with cam FAIS when compared to control participants, who had significant differences during loading response, pre-swing, and terminal swing (Figure 5). Differences between pelvic rotation during standing and gait were found for both patients with cam FAIS and control participants during mid- to terminal swing of incline gait (Figure 5).

Figure 5:

Percent of the gait cycle that the pelvic position was significantly different than during stance of level walk, standardized level walk, and incline walk. Solid bars are shown for the specific percentage of the gait cycle which was significantly different from stance for control participants (blue) and patients with cam FAIS (red), while the total percentage of the gait cycle is shown on the right.

When analyzing males alone, the only differences in pelvic position relative to standing were in pelvic obliquity. Specifically, male patients with cam FAIS had significant differences in obliquity from standing for 14% of the gait cycle (59–72% gait) during self-selected speed level walk, while male control participants had significant differences for 16% of the gait cycle (58–73% gait) during incline walk.

While no significant differences in pelvic motion were observed during the functional star-arc activity, patients with FAIS had consistent patterns of initially increasing and then decreasing anterior pelvic tilt, upward obliquity, and external rotation during the arc (circumduction) portion of the activity; data from control participants did not clearly demonstrate the same trends (Figure 6).

Figure 6:

Hip joint angles and pelvic rotation angles for control participants and patients with cam FAIS during the functional star-arc activity. The five star positions are labeled numerically: 1, flexion; 2, flexion-abduction; 3, abduction; 4, extension-abduction; and 5, extension. The arc (circumduction) portion of the activity is indicated by a light gray vertical band. The solid line indicates the mean, while the semi-transparent band represents the 95% confidence interval.

Correlations between Hip Joint Angles and Pelvic Motion

Abduction range of motion was strongly correlated to the range of pelvic obliquity during all gait activities for both participant groups, though not all correlations were significant (Table 3); when only male participants were evaluated, patients with cam FAIS had very strong correlations between abduction range of motion and the range of pelvic obliquity during both speeds of level gait, while very strong correlations were found for control participants during incline gait (Table S-1). Flexion range of motion was moderately correlated to the range of pelvic tilt during self-selected speed level gait for control participants (Table 3) and strongly correlated for male control participants (Table S-1). Range of hip rotation angle and pelvic rotation were strongly correlated for cam patients during incline gait (Table 3).

Table 3:

Correlation between joint angle range of motion and range of pelvic motion.

Activity	Correlation	Group	Flexion - Tilt	Abduction - Obliquity	Joint Rotation - Pelvic Rotation
Level Walk	Heel-Strike	Cam	0.931 *	−0.890 *	−0.216
		Control	0.526	−0.356	−0.004
	Toe-Off	Cam	0.976 *	−0.802	0.143
		Control	0.796 *	−0.381	0.294
	Range of Motion	Cam	0.363	0.832 *	0.200
		Control	0.660 *	0.747 *	0.462
Standardized Level Walk	Heel-Strike	Cam	0.958 *	−0.871	−0.530
		Control	0.693	−0.647	−0.232
	Toe-Off	Cam	0.959 *	−0.696	0.668
		Control	0.656	−0.659	−0.204
	Range of Motion	Cam	0.471	0.870	0.265
		Control	0.457	0.735 *	0.573
Incline Walk	Heel-Strike	Cam	0.744	−0.801	−0.180
		Control	0.575	−0.650	−0.536
	Toe-Off	Cam	0.656	−0.499	0.625
		Control	0.632	−0.710 *	−0.240
	Range of Motion	Cam	−0.738	0.734	0.824 *
		Control	0.117	0.867 *	0.376

^*Correlations significant at the p<0.05 level.

For heel-strike and toe-off of level gait, patients with cam FAIS had stronger correlations between abduction and obliquity than control participants during level gait at a self-selected speed, while the differences were less clear for standardized level or incline gait (Table 3). For the analysis of males, the correlation between abduction and obliquity was strong or very strong during toe-off and heel-strike of both level and incline gait for both participant groups, except for at toe-off of level gait for male control participants and toe-off of incline gait for patients with cam FAIS (Table S-1). Flexion angle was very strongly correlated to pelvic tilt during heel-strike and toe-off of both self-selected and standardized speeds of level gait for patients with cam FAIS and strongly correlated during toe-off at self-selected speed for control participants (Table 3). While the sagittal plane correlations remained similar for both cohorts of male participants, male control participants also had a very strong correlation between flexion angle and pelvic tilt at toe-off of standardized level gait (Table S-1). For male participants, hip rotation angle and pelvic rotation were very strongly correlated for control participants at heel-strike of standardized speed level walking and incline walking (Table S-1).

Discussion

We employed DF to quantify in-vivo kinematics, including femoral head translations, hip joint angles, and pelvic motion of patients with cam FAIS relative to asymptomatic, morphologically normal control participants during standing, level walking, incline walking, and an unweighted functional activity. Patients with cam FAIS stood with more hip extension, but additional kinematic differences between patients and controls were not observed. When only the male FAIS and control participants were evaluated, we found patients had reduced peak internal rotation during the self-selected speed level walk and reduced anterior pelvic tilt at heel-strike during incline walk. While overall range of motion was not different between patients and controls, we observed increased abduction in patients with cam FAIS during gait relative to their standing position when compared to controls. Hip joint angles in patients with cam FAIS were more strongly correlated to pelvic motion in the sagittal and coronal planes than controls. No differences in femoral head translation were found between the two groups. Overall, our results suggest that kinematic differences between patients with FAIS and morphologically normal controls are subtle, with the most striking of differences observed with regards to motion of the pelvis. Additional research with a larger sample size is needed to confirm our findings.

To our knowledge, this is the first study to evaluate in-vivo kinematics of patients with cam FAIS during weight-bearing activities of daily living using DF. Consistent with previous studies evaluating kinematics and range of motion in patients with cam FAIS using skin marker motion analysis, we observed no differences in cadence between patients with cam FAIS and controls.¹⁰ Similar to previous studies reporting reduced peak internal rotation,^8,11,12 we also observed reduced peak internal rotation during level walking in the male patients with cam FAIS relative to the male control participants.

Previous research that employed skin marker motion analysis reported reduced sagittal range of motion or reduced peak extension in patients with FAIS during weight-bearing activities,^8,10–14 yet we did not consistently observe these findings. One source for this discrepancy may be in the definition of the hip joint coordinate system. Specifically, DF uses anatomical landmarks visible on 3D CT reconstructions, whereas traditional motion capture methods calculate a coordinate system according to the spatial position of markers adhered to the skin; as we showed, this approach can produce errors in the estimation of the hip joint center that exceed 20 mm.^22,23 Further, soft tissue artifact can cause erroneous measurements of hip joint kinematics during dynamic loading for all anatomic planes, including rotation about the sagittal and coronal plane.¹⁹ It has also been shown that participant-specific pelvic bony anatomy is necessary to measure in-vivo pelvic tilt accurately.²⁹ Though we cannot definitively attribute the discrepancies between our results and previous studies to the limitations and errors of skin marker motion capture, we are confident that hip kinematics reported herein are accurate given that errors for DF are submillimeter and subdegree.¹⁸

We did observe that patients with cam FAIS had stronger correlations between hip joint kinematics and pelvic motion in the sagittal and coronal planes when compared to control participants. We theorize this is because patients with cam FAIS are maintaining stricter muscle control to avoid positions that cause impingement and pain.³⁰ Future studies could analyze electromyography data with DF-based kinematics to study this concept in more detail. Recent research found altered pelvifemoral rhythm in patients with FAIS;⁵ our findings of stronger correlations for the cam FAIS group also suggest that pelvic motion plays an important role in the disease. However, additional research is needed. Future motion analysis studies evaluating this relationship should consider inclusion of a control group that has cam morphology but no symptoms of the disease; analysis of this group may identify the specific role(s) structural deformities play in motion patterns without pain acting as a confounding factor. Similarly, the analysis of patients treated for cam FAIS with a femoral osteoplasty would identify whether the removal of cam morphology realigns pelvic motion and kinematics to that of control participants.

Relative to standing, abduction was increased in patients with cam FAIS relative to controls at time points of gait and the functional star-arc activity, which is contrary to the classic description of reduced abduction in patients with cam FAIS.^{8: 9} We also observed increased abduction range of motion during the functional star-arc activity for patients with cam FAIS. However, it is important to note the ranges of motion for these activities were submaximal. Thus, it is unlikely that abduction was limited by mechanical conflict between the acetabulum and femur. While not significant, patients with cam FAIS had trends of reduced peak downward obliquity (Figure 2), which could manifest as reduced pelvic obliquity range of motion as has previously been observed during level gait.¹⁰ Interestingly, we found no significant differences in hip joint range of motion, even during the high range of motion rotational pivots. This finding may be the result of joint mobility limitations, as both participant cohorts had large inter-participant variability indicating that bony morphology may not be the only factor responsible for changes in range of motion.

Previous reports of patients with FAIS, hip OA, and lower back pain have reported reduced range of pelvic motion.^10,14,31,32 While we did not explicitly identify any reductions in the range of pelvic motion, we did find that when compared to control participants, patients with cam FAIS had reduced pelvic tilt during level gait relative to their standing position and an altered relationship between hip joint and pelvic motion relative to the control participants. Interestingly, the differences between pelvic motion in control participants and patients with cam FAIS were inconsistent between level and incline gait, which may indicate that patients with FAIS move with altered pelvic motion as a result of altered stability strategies during more challenging or irregular tasks. This is supported by our observations for incline walking or circumduction (Figure 6), where patients had increased pelvic motion that may be the result of a compensation mechanism to avoid positions that approximate the cam lesion and the acetabular labrum. It should be noted that the differences in motion patterns between level and incline gait were observed with only a 5° incline, which is approximately that of many wheelchair ramps (1:12 slope ratio) and often encountered in daily living.

Our study findings are important, as they provide calculations of in-vivo hip joint and pelvic motion of patients with FAIS during activities of daily living and functional range of motion activities relative to participant-specific morphology. Our results represent active motion patterns that are free from errors associated with soft tissue artifact. Further, we did not have to assume generic morphology or motion patterns when evaluating kinematics, as we directly measured in-vivo bone motion based on anatomical landmarks specific to each individual. Given the use of participant-specific morphology and in-vivo kinematics, we still did not find consistent reductions in peak joint angles or range of motion, indicating that the effect of FAIS morphology on range of motion during daily activities is minimal. Nevertheless, altered pelvic motion may contribute to symptoms of FAIS through an overall alteration in hip joint kinematics, motion patterns, and, inherently, muscle recruitment patterns.³⁰

This study was not without limitations. Only a limited number of participants were recruited and analyzed, which also resulted in an unbalanced gender distribution in the two groups. For this reason, a separate analysis was completed evaluating only male participants, as most patients with cam FAIS were male. While we still found significant differences in gait patterns relative to hip joint and pelvic motion between patients and control participants, additional significant, yet possibly subtle, differences in pelvic tilt, external rotation, and flexion may have been observed with a larger sample size. Thus, our results should be viewed in the context of a preliminary study. Age, mass, and BMI were significantly larger for patients when compared to control participants. With our modest sample size, it was not feasible to evaluate age and BMI as confounding factors in the statistical analysis. However, all study participants were recreationally active and were of young age (less than 40 years). Thus, we believe it unlikely that age or BMI contributed to observed differences in hip kinematics between the two groups. We utilized an AP radiograph to screen for deformities because this view is ubiquitous in the clinical evaluation of young adults with hip pain. However, our group recently discovered that radiographic measurements of the AP film have limited ability to quantify the severity of cam lesions.³³ Nevertheless, alpha angles measured on oblique MRI reformats of the control group were less than 55.5°, which provides confidence that the control group had normally developing hip anatomy. Only a single gait cycle or activity was analyzed for each participant. However, we believe the high accuracy and low bias of DF somewhat obviates the need to average results across multiple trials and gait cycles.¹⁸ Our approach to quantifying hip kinematics exposed participants to ionizing radiation from DF and CT; the estimated dose was 10.72 mSv when considering the combined DF and CT exposure. This equates to 21% of the annual exposure limit for a radiation worker, or nearly three years of background radiation in Salt Lake City, Utah, which is situated at around 4,200 feet.³⁴ The reader should note that the annual dose limit for the general public from licensed operation is 1.0 mSv,³⁵ but this limit excludes background radiation and voluntary participation in medical research studies. Additionally, it is possible that participant kinematics were altered slightly to perform activities within the field of view and footprint of the DF system. However, this effect would have been similar across participants in both cohorts. Finally, hip joint translations were quantified relative to sphere-fits of the acetabulum and femoral head, however the latter may not be appropriate, as even non-pathological femoral heads are not spherical.³⁶

In conclusion, our findings agree with those previous studies that concluded that the kinematic effects of cam FAIS are subtle during weight-bearing activities and highlight the possible importance of pelvic motion in patient populations with hip pain. Future studies should investigate altered pelvic motion relative to femur-labrum approximation to understand whether these differences are compensatory strategies to minimize pain.