Gait & Posture 39 (2014) 859–864
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Effect of compensatory trunk movements on knee and hip joint loading during gait in children with different orthopedic pathologies Felix Stief a,*, Harald Bo¨hm b, Carsten Ebert c, Leonhard Do¨derlein b, Andrea Meurer a a
Orthopedic University Hospital Friedrichsheim gGmbH, Frankfurt/Main, Germany Orthopedic Hospital for Children, Aschau i. Chiemgau, Germany c Goethe-University Frankfurt, Department of Sports Medicine, Frankfurt/Main, Germany b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 29 July 2013 Received in revised form 13 November 2013 Accepted 24 November 2013
Ipsilateral trunk lean toward the affected stance limb has been identified as a compensatory mechanism to unload the hip joint. However, this altered gait pattern increases the lever arm around the knee joint by shifting the ground reaction vector more lateral to the knee joint center, which could be sufficient to deform the lateral compartment of the knee. The purpose of the present study was to show the effect of ipsilateral trunk lean on hip and knee joint moments in the frontal plane in 132 young patients with different orthopedic diagnosis. Linear correlations between ipsilateral trunk lean and the external knee and/or hip adduction moment were detected for patients with Legg–Calve´–Perthes disease (LCPD), arthrogryposis multiplex congenita, myelomeningocele, and unilateral cerebral palsy (CP). In contrast, children with bilateral CP did not show such a relationship due to an increased internal foot placement. In comparison to the hip joint, the effect of ipsilateral trunk lean in patients with LCPD is obviously more pronounced in the knee joint. The valgus thrust of the knee could initiate degenerative changes by placing altered loads on regions of the articular cartilage that were previously conditioned for different load levels. The results suggest that the ipsilateral trunk lean should not be considered and recommended as unloading mechanism for the hip joint on its own but also as a potential increased joint loading of the lateral knee compartment. Therefore, an acceptable therapy concept for limping patients should aim for an inconspicuous gait pattern with a reduced trunk movement. ß 2013 Elsevier B.V. All rights reserved.
Keywords: Gait analysis Lateral trunk lean Knee adduction moment Hip adduction moment Knee osteoarthritis
1. Introduction Ipsilateral trunk lean toward the affected stance limb (Fig. 1) is a useful substitute for inadequate hip abductors, joint contractures [1] and/or dysfunction of the affected hip [2]. This compensatory gait pattern was first described by Duchenne [2] in patients with paralyzed hip abductor muscles due to poliomyelitis. From a biomechanical point of view, body weight is moved toward the center of the hip joint, which reduces the demand on the weak hip abductor muscles. A recent study has presented a correlation between abductor strength and lateral trunk lean in 375 patients with cerebral palsy (CP) [3]. This suggests that weak hip abductors in patients with CP are accompanied by increased trunk lean to the ipsilateral side. In addition to neuromuscular considerations, the limping of the hip has been identified as an important compensatory mechanism
* Corresponding author at: Movement Analysis Lab, Orthopedic University Hospital Friedrichsheim gGmbH, Marienburgstraße 2, 60528 Frankfurt/Main, Germany. Tel.: +49 69 6705 862. E-mail address: [email protected] (F. Stief). 0966-6362/$ – see front matter ß 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gaitpost.2013.11.012
to unload the hip joint and to relief pain during walking [4–6]. According to SˇvehIı´k et al. [5] and Westhoff et al. [6], children with Legg–Calve´–Perthes disease (LCPD) adopted an ipsilateral trunk lean gait pattern with a reduced external hip adduction moment indicating a hip-unloading mechanism. There is also a growing base of knowledge on compensatory angular trunk movements during gait in subjects with osteoarthritis (OA) of the lower limb [7,8]. The ipsilateral trunk lean is considered as a non-invasive intervention for patients suffering from hip pain without a clear indication for surgery [4] or adult patients with medial compartment knee OA [9–11]. Hunt et al. [10] have shown that 13% of the variation in the maximum external knee adduction moment was explained by lateral trunk lean in patients with knee OA. The increased trunk lean reduced medial knee load in a dose–response relationship, with larger lean angles leading to greater reductions in the knee adduction moments [11]. Additionally, Mu¨ndermann et al. [12] and Van den Noort et al. [13] have demonstrated that lateral trunk lean has the potential of reducing the knee adduction moment during walking by up to 65% in healthy subjects. Therefore, it appears feasible that older persons [14] and patients with knee OA [15] do adopt a gait pattern that involves increased trunk sway.
860
F. Stief et al. / Gait & Posture 39 (2014) 859–864
knee OA. Previously published studies investigated the effect of compensatory trunk movements in adult patients [4,7,8,10,11,15] or healthy subjects [9,12–14]. Therefore, the major purpose of the present study was to show the correlation between ipsilateral trunk lean toward the affected stance limb and the knee/hip joint moment in a large group of young children with different pathologies. The second purpose of the present study was to determine kinematic and kinetic differences between patients with natural ipsilateral trunk lean (NTL) and patients with excessive ipsilateral trunk lean (ETL). We hypothesized that ETL might provoke a lateral shift of the ground reaction vector, which could increase lateral tibio-femoral compartment load at the knee joint. 2. Methods 2.1. Subjects Hip abductor weakness or joint contractures and the resulting ipsilateral trunk lean occur in many types of pathology. This approach was initiated by observations of young patients with a diagnosis of bilateral or unilateral CP, hip disorders (LCPD), congenital joint contractures (arthrogryposis multiplex congenita; AMC), and MMC who were routinely examined in two gait laboratories. A total of 132 patients participated in this multicenter study. Patients were able to walk without assistance or assistive devices. Exclusion criteria were: previous orthopedic surgery in the past two years and morbid obesity with a body mass index 30. Details on patients’ characteristics as well as the prevalence of ipsilateral trunk lean in each patient group are summarized in Table 1. Twenty typically developed children were recruited as control group (Table 1). None of the control subjects had previously been treated for any clinical spine or lower extremity conditions and none had any activity-restricting medical or musculoskeletal conditions. All subjects and their parents were thoroughly familiarized with the gait analysis protocol. The parents of children provided written informed consent to participate in this study, as approved by the local ethics committee and in accordance with the Helsinki Declaration. 2.2. Data collection
Fig. 1. Ipsilateral trunk lean toward the affected stance limb.
Besides the unloading mechanism for the hip joint and the medial knee compartment of the above described gait pattern, a secondary effect of the ipsilateral trunk lean is valgus of the knee as a result of the laterally displaced weight line. Having in mind that ambulatory load at the knee has been identified as a risk factor for the development and progression of knee OA [16], this could be sufficient to deform the lateral compartment of the knee in young patients with hip abductor weakness or joint contractures. Williams et al. [17] were the first to report that knee OA was present in 24% of their adult myelomeningocele (MMC) patient population. Nevertheless, there is a lack of research on the effect of trunk sway on knee joint loading in young patients without signs of
Three-dimensional gait analyses were performed in two different gait laboratories. Comparability between both laboratories was achieved by using the same Vicon motion capture system (VICON Motion Systems, Oxford, UK) operating at a sampling rate of 200 Hz. In both laboratories the level walkway was 15 m long and viewed by eight infrared cameras. AMTI force plates (Advanced Mechanical Technology, Inc., Watertown, MA, USA) were situated at the mid-point of the walkway to collect kinetic data at 1000 Hz. Prior to the gait analysis, for each subject anthropometric data (Table 1) were collected. The standardized Plug-in-Gait marker set was applied to determine joint centers and kinematic data. Reflective markers were placed on well-defined anatomical points as described in a previous investigation [18] to define the pelvis, thigh, shank, and foot segments. Six additional markers on the left/right shoulder (acromioclavicular joint), clavicle (on the jugular notch where the clavicles meet the sternum), sternum (xiphoid process), C7 (on the spinous process of the 7th cervical vertebra), and T10 (on the spinous process of the 10th thoracic vertebra) were attached to record trunk movements according to an enlarged model, which is part of the VICON software. The consistency between investigators, in particular regarding the marker placement, was achieved by the circumstance that both responsible investigators for the present study
F. Stief et al. / Gait & Posture 39 (2014) 859–864
861
Table 1 Study population characteristics (mean with standard deviation in parenthesis) and prevalence of excessive ipsilateral trunk lean. Variable
Patients
Controls
CP bilateral
CP unilateral
LCPD
AMC
MMC
n Sex, no. female/no. male Age, years Height, m Body mass, kg Body mass index, kg/m2 Prevalence ipsilateral trunk lean, %
37 9/28 8.6 (2.3) 1.32 (0.14) 31.0 (10.6) 17.4 (3.3) 54
31 13/18 8.9 (3.3) 1.37 (0.19) 35.2 (14.2) 17.9 (3.7) 26
27 8/19 6.1 (1.8) 1.18 (0.11) 22.9 (5.1) 16.3 (1.6) 30
18 6/12 8.8 (2.6) 1.35 (0.17) 31.5 (11.5) 16.7 (3.3) 61
19 7/12 9.0 (3.5) 1.31 (0.19) 33.4 (14.7) 18.4 (3.6) 58
20 9/11 9.3 (2.3) 1.40 (0.17) 32.7 (11.9) 16.2 (2.3) 0
CP, cerebral palsy; LCPD, Legg–Calve´–Perthes disease; AMC, arthrogryposis multiplex congenita; MMC, myelomeningocele. Prevalence of excessive ipsilateral trunk lean was characterized by maximum trunk lean toward the stance limb of greater than 2 standard deviations from typically developed children (controls).
were working together using the Plug-in-Gait marker set for more than two years in the same gait laboratory. 2.3. Data processing and statistical analysis
thorax obliquity [°]
Medio-lateral trunk sway was measured as the frontal plane thorax angle defined as the movement of the trunk in relation to the global system (Fig. 2). Negative values indicate a trunk lean toward the affected stance limb. The prevalence of ETL was characterized by maximum trunk lean toward the stance limb of greater than 2 standard deviations of the control group. Pelvis obliquity is defined as the movement of the pelvis in relation to the global system in the frontal plane. Positive values indicate a pelvic lift (Fig. 2). The foot endorotation is defined as the movement of the foot in relation to the global system in the transverse plane.
2 0 −2 −4 −6 −8
pelvis obliquity [°]
0
40
60
80
100
Controls LCPD NTL LCPD ETL
5 0 −5 0
Hip adduction moment [Nm/kg]
20
20
40
60
80
100
0.6
3. Results
0.4 0.2 0
−0.2
Knee adduction moment [Nm/kg]
Positive values indicate an internal foot placement. Further processing incorporating force plate data allowed external moments about the mathematically derived joint centers to be calculated (Fig. 2). After each acquisition session, 3D marker trajectories were reconstructed and missing frames were handled with a fill-gap procedure using the Vicon-Nexus software version 1.7.1 (VICON Motion Systems, Oxford, UK). Discrete variables of interest (Table 2) were calculated for at least three barefoot trials at a self-selected speed and then averaged for further analysis on the basis of complete marker trajectories and a clear foot-forceplatecontact. In the case of bilateral involvement (CP bilateral, AMC, MMC) measurements were performed only on the limb with greater maximum trunk lean toward the stance limb. This ensures that the statistical calculations were independent from each other. No significant differences were found in the discrete variables of interest comparing the left and right side in the control group. Therefore, only the right side was used for further analysis of the controls. Walking speed was assessed in a non-dimensional unit that corrects different leg lengths of the participants [19]. Means and standard deviations were automatically determined by a custom made algorithm in Matlab 7.12.0 (The MathWorks, Inc., Natick, MA). The normal distribution of the present sample tested with the Kolmogorov–Smirnov test could not be confirmed in all parameters of interest. Therefore, the nonparametric spearman correlation coefficient (r) was used to examine the association between ipsilateral trunk lean toward the affected stance limb (maximum value) and the maximum and mean knee and hip adduction moments. The significance level adopted in the present study was set at p 0.05. Differences between groups were tested for significance using a nonparametrical Mann–Whitney U-test.
0
20
0
20
40
60
80
100
40
60
80
100
0.4 0.2 0
time [% gaitcycle]
Fig. 2. Average thorax and pelvis movements in the frontal plane and the corresponding external hip and knee adduction moments for patients with unilateral LCPD (n = 27) that showed natural ipsilateral trunk lean (NTL) and excessive ipsilateral trunk lean (ETL). Mean values including standard deviation of healthy controls are shown in gray color.
The prevalence of ETL in the patient groups (Table 1) varied between 26% (CP unilateral) and 61% (AMC). Linear correlations between ipsilateral trunk lean and the hip adduction moment were detected for patients with AMC, MMC, and unilateral CP. Regarding the knee adduction moment, linear correlations were shown for patients with LCPD, AMC, and MMC. In contrast, children with bilateral CP did not show such a relationship neither with the hip nor with the knee joint moments (Table 3). Gait parameters of the patient groups and the controls are presented in Table 2. No significant differences were observed for pelvis obliquity between NTL and ETL in the patient groups. Significant differences were observed between NTL (10.1 10.28) and ETL (16.1 12.48) in bilateral CP patients for the mean endorotation of the foot during the stance phase of gait. The calculation of the maximum and mean hip joint moments in the frontal plane showed significant differences between NTL
F. Stief et al. / Gait & Posture 39 (2014) 859–864
862
Table 2 Mean with standard deviation in parenthesis for kinematic and kinetic gait parameters during the stance phase of gait. Walking speed, nondimensional
Patients
Angles (8)
Moments (Nm/kg)
Thorax
Pelvis
Foot
Hip
Max. obliquity
Max. obliquity
Mean endorotation
Max. adduction
Knee Mean adduction
Max. adduction
Mean adduction
CP bilateral
NTL (n = 17) ETL (n = 20)
0.45 (0.06) 0.40 (0.10)
5.8 (1.7)** 10.9 (2.9)**
7.7 (2.9) 7.1 (4.5)
10.1 (10.2)* 16.1 (12.4)*
0.90 (0.30) 0.80 (0.29)
0.34 (0.19) 0.30 (0.18)
0.52 (0.20) 0.40 (0.15)
0.17 (0.10) 0.11 (0.14)
CP unilateral
NTL (n = 23) ETL (n = 8)
0.45 (0.06) 0.46 (0.05)
1.5 (3.1)** 9.0 (1.4)**
4.0 (3.1) 3.6 (3.8)
1.0 (15.3) 0.4 (12.8)
0.89 (0.22) 0.77 (0.24)
0.41 (0.14) 0.32 (0.10)
0.42 (0.13) 0.34 (0.14)
0.18 (0.08)* 0.10 (0.06)*
LCPD
NTL (n = 19) ETL (n = 8)
0.48 (0.07) 0.44 (0.06)
4.3 (1.8)** 10.3 (3.5)**
3.7 (3.0) 2.2 (3.4)
2.8 (10.7) 7.3 (9.1)
0.60 (0.12) 0.51 (0.17)
0.35 (0.07) 0.30 (0.15)
0.29 (0.14)* 0.16 (0.08)*
0.12 (0.07)** 0.01 (0.06)**
AMC
NTL (n = 7) ETL (n = 11)
0.44 (0.07)* 0.31 (0.11)*
4.2 (1.5)** 16.0 (7.1)**
4.3 (4.0) 2.9 (6.9)
1.5 (4.8) 7.5 (16.5)
0.71 (0.16)* 0.49 (0.24)*
0.36 (0.11)* 0.11 (0.16)*
0.47 (0.23) 0.28 (0.22)
0.20 (0.10)* 0.03 (0.16)*
MMC
NTL (n = 8) ETL (n = 11)
0.42 (0.06) 0.41 (0.14)
4.9 (1.2)** 15.2 (7.1)**
6.0 (3.0) 9.2 (7.3)
2.1 (18.9) 8.2 (19.9)
0.79 (0.14)* 0.43 (0.31)*
0.40 (0.13)* 0.01 (0.29)*
0.43 (0.16) 0.30 (0.29)
0.18 (0.12)* 0.02 (0.14)*
Controls
NTL (n = 20)
0.49 (0.06)
3.2 (2.2)
5.9 (2.0)
7.0 (5.1)
0.73 (0.14)
0.36 (0.10)
0.47 (0.16)
0.18 (0.07)
CP, cerebral palsy; LCPD, Legg-Calve´-Perthes disease; AMC, arthrogryposis multiplex congenita; MMC, myelomeningocele; NTL, natural ipsilateral trunk lean; ETL, excessive ipsilateral trunk lean; n, number of subjects; ‘‘thorax obliquity’’ is defined as the movement of the trunk in relation to the global system in the frontal plane, negative values indicate a trunk lean toward the affected stance limb; ‘‘pelvis obliquity’’ is defined as the movement of the pelvis in relation to the global system in the frontal plane, positive values indicate a pelvic lift; foot endorotation is defined as the movement of the foot in relation to the global system in the transverse plane. Positive values indicate an internal foot placement. Significant differences are indicated in gray. * Significant difference between NTL and ETL at p 0.05. ** Significant difference between NTL and ETL at p 0.001.
and ETL in patients with AMC and MMC. The maximum hip joint moments in the ETL groups were reduced between 31% (AMC) and 46% (MMC). Regarding the knee joint moments, significant differences between NTL and ETL were shown for patients with LCPD, AMC, MMC, and unilateral CP. The maximum knee joint moments in the ETL groups were reduced by up to 45% (LCPD). No significant differences were observed between NTL and ETL for the hip and knee joint moments in patients with bilateral CP. Fig. 2 exemplarily shows the average thorax and pelvis movements in the frontal plane and the corresponding knee and hip joint adduction moments for patients with LCPD.
4. Discussion The purpose of the present study was to determine the contribution of ipsilateral trunk lean to a change in the hip joint loading and in the distribution of loads transferred through the medial and lateral knee compartments in young patients with an increased ipsilateral trunk lean. It is, to the best of our knowledge, the first study to have investigated the effect of trunk lean on the knee adduction moment during gait in a large group of young patients with different orthopedic pathologies.
The external hip adduction moment is used as the primary parameter for characterizing hip loading or unloading [20]. The excessive ipsilateral trunk lean incorporated by our ETL patients decreased the moment around the hip by shifting the ground reaction vector closer to the hip joint center. This suggests that the trunk lean toward the supporting limb in stance serves the purpose of minimizing the force demands placed on the weakened hip abductor muscles typically demonstrated in our patient groups. Furthermore, the acetabulum is shifted over the femoral head and therefore the weight bearing area of the hip joint is increased. These results are in accordance with previous studies showing that frontal plane compensatory movements of the trunk were associated with unloading of the hip joint [4–6,21]. Nevertheless, in the present study, a significant correlation between ipsilateral trunk lean and the hip adduction moment could only be shown in patients with AMC, MMC, and unilateral CP (Table 3). Compared to the knee joint moments, the effect of ipsilateral trunk lean on the frontal hip joint is obviously less pronounced in patients with LCPD (Fig. 2). Besides the unloading mechanism for the hip joint of the above described limping, this altered gait pattern may also have a negative effect on the knee joint. The ipsilateral trunk lean increases the lever arm around the knee joint by shifting the
Table 3 Spearman correlation coefficients (r) and p-values to ipsilateral trunk lean (maximum value) for the external maximum and mean knee and hip adduction moment during the stance phase of gait in the patient groups. Patients
Knee max. adduction moment (Nm/kg)
Knee mean adduction moment (Nm/kg)
Hip max. adduction moment (Nm/kg)
Hip mean adduction moment (Nm/kg)
r CP bilateral CP unilateral LCPD AMC MMC
0.33 0.21 0.31 0.51 0.46
p-Value
r
p-Value
r
p-Value
r
p-Value
0.055 0.302 0.116 0.032 0.049
0.31 0.29 0.47 0.60 0.63
0.061 0.104 0.014 0.010
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Effect of compensatory trunk movements on knee and hip joint loading during gait in children with different orthopedic pathologies Felix Stief a,*, Harald Bo¨hm b, Carsten Ebert c, Leonhard Do¨derlein b, Andrea Meurer a a
Orthopedic University Hospital Friedrichsheim gGmbH, Frankfurt/Main, Germany Orthopedic Hospital for Children, Aschau i. Chiemgau, Germany c Goethe-University Frankfurt, Department of Sports Medicine, Frankfurt/Main, Germany b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 29 July 2013 Received in revised form 13 November 2013 Accepted 24 November 2013
Ipsilateral trunk lean toward the affected stance limb has been identified as a compensatory mechanism to unload the hip joint. However, this altered gait pattern increases the lever arm around the knee joint by shifting the ground reaction vector more lateral to the knee joint center, which could be sufficient to deform the lateral compartment of the knee. The purpose of the present study was to show the effect of ipsilateral trunk lean on hip and knee joint moments in the frontal plane in 132 young patients with different orthopedic diagnosis. Linear correlations between ipsilateral trunk lean and the external knee and/or hip adduction moment were detected for patients with Legg–Calve´–Perthes disease (LCPD), arthrogryposis multiplex congenita, myelomeningocele, and unilateral cerebral palsy (CP). In contrast, children with bilateral CP did not show such a relationship due to an increased internal foot placement. In comparison to the hip joint, the effect of ipsilateral trunk lean in patients with LCPD is obviously more pronounced in the knee joint. The valgus thrust of the knee could initiate degenerative changes by placing altered loads on regions of the articular cartilage that were previously conditioned for different load levels. The results suggest that the ipsilateral trunk lean should not be considered and recommended as unloading mechanism for the hip joint on its own but also as a potential increased joint loading of the lateral knee compartment. Therefore, an acceptable therapy concept for limping patients should aim for an inconspicuous gait pattern with a reduced trunk movement. ß 2013 Elsevier B.V. All rights reserved.
Keywords: Gait analysis Lateral trunk lean Knee adduction moment Hip adduction moment Knee osteoarthritis
1. Introduction Ipsilateral trunk lean toward the affected stance limb (Fig. 1) is a useful substitute for inadequate hip abductors, joint contractures [1] and/or dysfunction of the affected hip [2]. This compensatory gait pattern was first described by Duchenne [2] in patients with paralyzed hip abductor muscles due to poliomyelitis. From a biomechanical point of view, body weight is moved toward the center of the hip joint, which reduces the demand on the weak hip abductor muscles. A recent study has presented a correlation between abductor strength and lateral trunk lean in 375 patients with cerebral palsy (CP) [3]. This suggests that weak hip abductors in patients with CP are accompanied by increased trunk lean to the ipsilateral side. In addition to neuromuscular considerations, the limping of the hip has been identified as an important compensatory mechanism
* Corresponding author at: Movement Analysis Lab, Orthopedic University Hospital Friedrichsheim gGmbH, Marienburgstraße 2, 60528 Frankfurt/Main, Germany. Tel.: +49 69 6705 862. E-mail address: [email protected] (F. Stief). 0966-6362/$ – see front matter ß 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gaitpost.2013.11.012
to unload the hip joint and to relief pain during walking [4–6]. According to SˇvehIı´k et al. [5] and Westhoff et al. [6], children with Legg–Calve´–Perthes disease (LCPD) adopted an ipsilateral trunk lean gait pattern with a reduced external hip adduction moment indicating a hip-unloading mechanism. There is also a growing base of knowledge on compensatory angular trunk movements during gait in subjects with osteoarthritis (OA) of the lower limb [7,8]. The ipsilateral trunk lean is considered as a non-invasive intervention for patients suffering from hip pain without a clear indication for surgery [4] or adult patients with medial compartment knee OA [9–11]. Hunt et al. [10] have shown that 13% of the variation in the maximum external knee adduction moment was explained by lateral trunk lean in patients with knee OA. The increased trunk lean reduced medial knee load in a dose–response relationship, with larger lean angles leading to greater reductions in the knee adduction moments [11]. Additionally, Mu¨ndermann et al. [12] and Van den Noort et al. [13] have demonstrated that lateral trunk lean has the potential of reducing the knee adduction moment during walking by up to 65% in healthy subjects. Therefore, it appears feasible that older persons [14] and patients with knee OA [15] do adopt a gait pattern that involves increased trunk sway.
860
F. Stief et al. / Gait & Posture 39 (2014) 859–864
knee OA. Previously published studies investigated the effect of compensatory trunk movements in adult patients [4,7,8,10,11,15] or healthy subjects [9,12–14]. Therefore, the major purpose of the present study was to show the correlation between ipsilateral trunk lean toward the affected stance limb and the knee/hip joint moment in a large group of young children with different pathologies. The second purpose of the present study was to determine kinematic and kinetic differences between patients with natural ipsilateral trunk lean (NTL) and patients with excessive ipsilateral trunk lean (ETL). We hypothesized that ETL might provoke a lateral shift of the ground reaction vector, which could increase lateral tibio-femoral compartment load at the knee joint. 2. Methods 2.1. Subjects Hip abductor weakness or joint contractures and the resulting ipsilateral trunk lean occur in many types of pathology. This approach was initiated by observations of young patients with a diagnosis of bilateral or unilateral CP, hip disorders (LCPD), congenital joint contractures (arthrogryposis multiplex congenita; AMC), and MMC who were routinely examined in two gait laboratories. A total of 132 patients participated in this multicenter study. Patients were able to walk without assistance or assistive devices. Exclusion criteria were: previous orthopedic surgery in the past two years and morbid obesity with a body mass index 30. Details on patients’ characteristics as well as the prevalence of ipsilateral trunk lean in each patient group are summarized in Table 1. Twenty typically developed children were recruited as control group (Table 1). None of the control subjects had previously been treated for any clinical spine or lower extremity conditions and none had any activity-restricting medical or musculoskeletal conditions. All subjects and their parents were thoroughly familiarized with the gait analysis protocol. The parents of children provided written informed consent to participate in this study, as approved by the local ethics committee and in accordance with the Helsinki Declaration. 2.2. Data collection
Fig. 1. Ipsilateral trunk lean toward the affected stance limb.
Besides the unloading mechanism for the hip joint and the medial knee compartment of the above described gait pattern, a secondary effect of the ipsilateral trunk lean is valgus of the knee as a result of the laterally displaced weight line. Having in mind that ambulatory load at the knee has been identified as a risk factor for the development and progression of knee OA [16], this could be sufficient to deform the lateral compartment of the knee in young patients with hip abductor weakness or joint contractures. Williams et al. [17] were the first to report that knee OA was present in 24% of their adult myelomeningocele (MMC) patient population. Nevertheless, there is a lack of research on the effect of trunk sway on knee joint loading in young patients without signs of
Three-dimensional gait analyses were performed in two different gait laboratories. Comparability between both laboratories was achieved by using the same Vicon motion capture system (VICON Motion Systems, Oxford, UK) operating at a sampling rate of 200 Hz. In both laboratories the level walkway was 15 m long and viewed by eight infrared cameras. AMTI force plates (Advanced Mechanical Technology, Inc., Watertown, MA, USA) were situated at the mid-point of the walkway to collect kinetic data at 1000 Hz. Prior to the gait analysis, for each subject anthropometric data (Table 1) were collected. The standardized Plug-in-Gait marker set was applied to determine joint centers and kinematic data. Reflective markers were placed on well-defined anatomical points as described in a previous investigation [18] to define the pelvis, thigh, shank, and foot segments. Six additional markers on the left/right shoulder (acromioclavicular joint), clavicle (on the jugular notch where the clavicles meet the sternum), sternum (xiphoid process), C7 (on the spinous process of the 7th cervical vertebra), and T10 (on the spinous process of the 10th thoracic vertebra) were attached to record trunk movements according to an enlarged model, which is part of the VICON software. The consistency between investigators, in particular regarding the marker placement, was achieved by the circumstance that both responsible investigators for the present study
F. Stief et al. / Gait & Posture 39 (2014) 859–864
861
Table 1 Study population characteristics (mean with standard deviation in parenthesis) and prevalence of excessive ipsilateral trunk lean. Variable
Patients
Controls
CP bilateral
CP unilateral
LCPD
AMC
MMC
n Sex, no. female/no. male Age, years Height, m Body mass, kg Body mass index, kg/m2 Prevalence ipsilateral trunk lean, %
37 9/28 8.6 (2.3) 1.32 (0.14) 31.0 (10.6) 17.4 (3.3) 54
31 13/18 8.9 (3.3) 1.37 (0.19) 35.2 (14.2) 17.9 (3.7) 26
27 8/19 6.1 (1.8) 1.18 (0.11) 22.9 (5.1) 16.3 (1.6) 30
18 6/12 8.8 (2.6) 1.35 (0.17) 31.5 (11.5) 16.7 (3.3) 61
19 7/12 9.0 (3.5) 1.31 (0.19) 33.4 (14.7) 18.4 (3.6) 58
20 9/11 9.3 (2.3) 1.40 (0.17) 32.7 (11.9) 16.2 (2.3) 0
CP, cerebral palsy; LCPD, Legg–Calve´–Perthes disease; AMC, arthrogryposis multiplex congenita; MMC, myelomeningocele. Prevalence of excessive ipsilateral trunk lean was characterized by maximum trunk lean toward the stance limb of greater than 2 standard deviations from typically developed children (controls).
were working together using the Plug-in-Gait marker set for more than two years in the same gait laboratory. 2.3. Data processing and statistical analysis
thorax obliquity [°]
Medio-lateral trunk sway was measured as the frontal plane thorax angle defined as the movement of the trunk in relation to the global system (Fig. 2). Negative values indicate a trunk lean toward the affected stance limb. The prevalence of ETL was characterized by maximum trunk lean toward the stance limb of greater than 2 standard deviations of the control group. Pelvis obliquity is defined as the movement of the pelvis in relation to the global system in the frontal plane. Positive values indicate a pelvic lift (Fig. 2). The foot endorotation is defined as the movement of the foot in relation to the global system in the transverse plane.
2 0 −2 −4 −6 −8
pelvis obliquity [°]
0
40
60
80
100
Controls LCPD NTL LCPD ETL
5 0 −5 0
Hip adduction moment [Nm/kg]
20
20
40
60
80
100
0.6
3. Results
0.4 0.2 0
−0.2
Knee adduction moment [Nm/kg]
Positive values indicate an internal foot placement. Further processing incorporating force plate data allowed external moments about the mathematically derived joint centers to be calculated (Fig. 2). After each acquisition session, 3D marker trajectories were reconstructed and missing frames were handled with a fill-gap procedure using the Vicon-Nexus software version 1.7.1 (VICON Motion Systems, Oxford, UK). Discrete variables of interest (Table 2) were calculated for at least three barefoot trials at a self-selected speed and then averaged for further analysis on the basis of complete marker trajectories and a clear foot-forceplatecontact. In the case of bilateral involvement (CP bilateral, AMC, MMC) measurements were performed only on the limb with greater maximum trunk lean toward the stance limb. This ensures that the statistical calculations were independent from each other. No significant differences were found in the discrete variables of interest comparing the left and right side in the control group. Therefore, only the right side was used for further analysis of the controls. Walking speed was assessed in a non-dimensional unit that corrects different leg lengths of the participants [19]. Means and standard deviations were automatically determined by a custom made algorithm in Matlab 7.12.0 (The MathWorks, Inc., Natick, MA). The normal distribution of the present sample tested with the Kolmogorov–Smirnov test could not be confirmed in all parameters of interest. Therefore, the nonparametric spearman correlation coefficient (r) was used to examine the association between ipsilateral trunk lean toward the affected stance limb (maximum value) and the maximum and mean knee and hip adduction moments. The significance level adopted in the present study was set at p 0.05. Differences between groups were tested for significance using a nonparametrical Mann–Whitney U-test.
0
20
0
20
40
60
80
100
40
60
80
100
0.4 0.2 0
time [% gaitcycle]
Fig. 2. Average thorax and pelvis movements in the frontal plane and the corresponding external hip and knee adduction moments for patients with unilateral LCPD (n = 27) that showed natural ipsilateral trunk lean (NTL) and excessive ipsilateral trunk lean (ETL). Mean values including standard deviation of healthy controls are shown in gray color.
The prevalence of ETL in the patient groups (Table 1) varied between 26% (CP unilateral) and 61% (AMC). Linear correlations between ipsilateral trunk lean and the hip adduction moment were detected for patients with AMC, MMC, and unilateral CP. Regarding the knee adduction moment, linear correlations were shown for patients with LCPD, AMC, and MMC. In contrast, children with bilateral CP did not show such a relationship neither with the hip nor with the knee joint moments (Table 3). Gait parameters of the patient groups and the controls are presented in Table 2. No significant differences were observed for pelvis obliquity between NTL and ETL in the patient groups. Significant differences were observed between NTL (10.1 10.28) and ETL (16.1 12.48) in bilateral CP patients for the mean endorotation of the foot during the stance phase of gait. The calculation of the maximum and mean hip joint moments in the frontal plane showed significant differences between NTL
F. Stief et al. / Gait & Posture 39 (2014) 859–864
862
Table 2 Mean with standard deviation in parenthesis for kinematic and kinetic gait parameters during the stance phase of gait. Walking speed, nondimensional
Patients
Angles (8)
Moments (Nm/kg)
Thorax
Pelvis
Foot
Hip
Max. obliquity
Max. obliquity
Mean endorotation
Max. adduction
Knee Mean adduction
Max. adduction
Mean adduction
CP bilateral
NTL (n = 17) ETL (n = 20)
0.45 (0.06) 0.40 (0.10)
5.8 (1.7)** 10.9 (2.9)**
7.7 (2.9) 7.1 (4.5)
10.1 (10.2)* 16.1 (12.4)*
0.90 (0.30) 0.80 (0.29)
0.34 (0.19) 0.30 (0.18)
0.52 (0.20) 0.40 (0.15)
0.17 (0.10) 0.11 (0.14)
CP unilateral
NTL (n = 23) ETL (n = 8)
0.45 (0.06) 0.46 (0.05)
1.5 (3.1)** 9.0 (1.4)**
4.0 (3.1) 3.6 (3.8)
1.0 (15.3) 0.4 (12.8)
0.89 (0.22) 0.77 (0.24)
0.41 (0.14) 0.32 (0.10)
0.42 (0.13) 0.34 (0.14)
0.18 (0.08)* 0.10 (0.06)*
LCPD
NTL (n = 19) ETL (n = 8)
0.48 (0.07) 0.44 (0.06)
4.3 (1.8)** 10.3 (3.5)**
3.7 (3.0) 2.2 (3.4)
2.8 (10.7) 7.3 (9.1)
0.60 (0.12) 0.51 (0.17)
0.35 (0.07) 0.30 (0.15)
0.29 (0.14)* 0.16 (0.08)*
0.12 (0.07)** 0.01 (0.06)**
AMC
NTL (n = 7) ETL (n = 11)
0.44 (0.07)* 0.31 (0.11)*
4.2 (1.5)** 16.0 (7.1)**
4.3 (4.0) 2.9 (6.9)
1.5 (4.8) 7.5 (16.5)
0.71 (0.16)* 0.49 (0.24)*
0.36 (0.11)* 0.11 (0.16)*
0.47 (0.23) 0.28 (0.22)
0.20 (0.10)* 0.03 (0.16)*
MMC
NTL (n = 8) ETL (n = 11)
0.42 (0.06) 0.41 (0.14)
4.9 (1.2)** 15.2 (7.1)**
6.0 (3.0) 9.2 (7.3)
2.1 (18.9) 8.2 (19.9)
0.79 (0.14)* 0.43 (0.31)*
0.40 (0.13)* 0.01 (0.29)*
0.43 (0.16) 0.30 (0.29)
0.18 (0.12)* 0.02 (0.14)*
Controls
NTL (n = 20)
0.49 (0.06)
3.2 (2.2)
5.9 (2.0)
7.0 (5.1)
0.73 (0.14)
0.36 (0.10)
0.47 (0.16)
0.18 (0.07)
CP, cerebral palsy; LCPD, Legg-Calve´-Perthes disease; AMC, arthrogryposis multiplex congenita; MMC, myelomeningocele; NTL, natural ipsilateral trunk lean; ETL, excessive ipsilateral trunk lean; n, number of subjects; ‘‘thorax obliquity’’ is defined as the movement of the trunk in relation to the global system in the frontal plane, negative values indicate a trunk lean toward the affected stance limb; ‘‘pelvis obliquity’’ is defined as the movement of the pelvis in relation to the global system in the frontal plane, positive values indicate a pelvic lift; foot endorotation is defined as the movement of the foot in relation to the global system in the transverse plane. Positive values indicate an internal foot placement. Significant differences are indicated in gray. * Significant difference between NTL and ETL at p 0.05. ** Significant difference between NTL and ETL at p 0.001.
and ETL in patients with AMC and MMC. The maximum hip joint moments in the ETL groups were reduced between 31% (AMC) and 46% (MMC). Regarding the knee joint moments, significant differences between NTL and ETL were shown for patients with LCPD, AMC, MMC, and unilateral CP. The maximum knee joint moments in the ETL groups were reduced by up to 45% (LCPD). No significant differences were observed between NTL and ETL for the hip and knee joint moments in patients with bilateral CP. Fig. 2 exemplarily shows the average thorax and pelvis movements in the frontal plane and the corresponding knee and hip joint adduction moments for patients with LCPD.
4. Discussion The purpose of the present study was to determine the contribution of ipsilateral trunk lean to a change in the hip joint loading and in the distribution of loads transferred through the medial and lateral knee compartments in young patients with an increased ipsilateral trunk lean. It is, to the best of our knowledge, the first study to have investigated the effect of trunk lean on the knee adduction moment during gait in a large group of young patients with different orthopedic pathologies.
The external hip adduction moment is used as the primary parameter for characterizing hip loading or unloading [20]. The excessive ipsilateral trunk lean incorporated by our ETL patients decreased the moment around the hip by shifting the ground reaction vector closer to the hip joint center. This suggests that the trunk lean toward the supporting limb in stance serves the purpose of minimizing the force demands placed on the weakened hip abductor muscles typically demonstrated in our patient groups. Furthermore, the acetabulum is shifted over the femoral head and therefore the weight bearing area of the hip joint is increased. These results are in accordance with previous studies showing that frontal plane compensatory movements of the trunk were associated with unloading of the hip joint [4–6,21]. Nevertheless, in the present study, a significant correlation between ipsilateral trunk lean and the hip adduction moment could only be shown in patients with AMC, MMC, and unilateral CP (Table 3). Compared to the knee joint moments, the effect of ipsilateral trunk lean on the frontal hip joint is obviously less pronounced in patients with LCPD (Fig. 2). Besides the unloading mechanism for the hip joint of the above described limping, this altered gait pattern may also have a negative effect on the knee joint. The ipsilateral trunk lean increases the lever arm around the knee joint by shifting the
Table 3 Spearman correlation coefficients (r) and p-values to ipsilateral trunk lean (maximum value) for the external maximum and mean knee and hip adduction moment during the stance phase of gait in the patient groups. Patients
Knee max. adduction moment (Nm/kg)
Knee mean adduction moment (Nm/kg)
Hip max. adduction moment (Nm/kg)
Hip mean adduction moment (Nm/kg)
r CP bilateral CP unilateral LCPD AMC MMC
0.33 0.21 0.31 0.51 0.46
p-Value
r
p-Value
r
p-Value
r
p-Value
0.055 0.302 0.116 0.032 0.049
0.31 0.29 0.47 0.60 0.63
0.061 0.104 0.014 0.010
