- Email: [email protected]
Sound side joint contact forces in below knee amputee gait with an ESAR prosthetic foot
Gait & Posture 58 (2017) 246–251
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Full length article
Sound side joint contact forces in below knee amputee gait with an ESAR prosthetic foot
MARK
Mohammad Taghi Karimia,c, Firooz Salamib, Amir Esrafiliand,e, Daniel W.W. Heitzmannb, ⁎ Merkur Alimusajb, Cornelia Putzb, Sebastian I. Wolfb, a
Musculoskeletal Research Center, Isfahan University of Medical Sciences, Isfahan, Iran Clinic for Orthopaedics and Trauma Surgery, Heidelberg University Hospital, Schlierbacher Landstrasse 200a, 69118 Heidelberg, Germany c Rehabilitation Sciences Research Center, Shiraz University of Medical Sciences, Shiraz Iran d Department of Biomedical Engineering, Amirkabir University of Technology, Tehran, Iran e Department of Applied Physics, University of Eastern Finland, Kuopio, Finland b
A R T I C L E I N F O
A B S T R A C T
Keywords: Below knee amputee Prosthesis Joint contact force Osteoarthritis Gait
The incidence of knee and hip joint osteoarthritis in subjects with below knee amputation (BK) appears significantly higher compared to unimpaired subjects, especially in the intact side. However, it is controversial if constant higher loads on the sound side are one of the major factors for an increased osteoarthritis (OA) incidence in subjects with BK, beside other risk factors, e.g. with respect to metabolism. The aim wasto investigate joint contact forces (JCF) calculated by a musculoskeletal model in the intact side and to compare it with those of unimpaired subjects and to further elucidate in how far increased knee JCF are associated with increased frontal plane knee moments. A group of seven subjects with BK amputation and a group of ten unimpaired subjects were recruited for this study. Gait data were measured by 3D motion capture and force plates. OpenSim software was applied to calculate JCF. Maximum joint angles, ground reaction forces, and moments as well as time distance parameters were determined and compared between groups showing no significant differences, with some JCF components of knee and hip even being slightly smaller in subjects with BK compared to the reference group. This positive finding may be due to the selected ESAR foot. However, other beneficial factors may also have influenced this positive result such as the general good health status of the subjects or the thorough and proper fitting and alignment of the prosthesis.
1. Introduction
prosthesis [13–16]. Sharma et al. showed that there is a correlation between magnitude of knee varus joint moment and OA symptoms based on radiologic classification [17]. Therefore, an increase in incidence of knee OA in the intact side of BK amputees may be due to an increase in applied loads. If this is the case then it should be aimed at reducing loads by proper fitting, alignment of the prosthesis and an appropriate choice of prosthetic components [18]. In the research done by Beyart et al., it was shown that the peaks of vertical components of ground reaction force (GRF) were higher in the intact limb of subjects with BK compared to the contralateral side and to normal subjects [14], respectively. Engsberg et al. also showed that subjects with BK had a greater rate of loading in their sound side [19] (eleven able bodied’ children and four children with BK amputation participated in this study). In contrast, Hurley et al. showed that the forces acting across the joints of the contralateral side in adults with BK were not significantly higher than those in subjects without disability [15].
The number of subjects with amputation, especially in the lower limb, is increasing for several reasons such as rising incidence of diabetes mellitus, average age and social deprivation [1,2]. Although various prosthetic components have been designed for subjects with lower limb amputation to restore their abilities while walking, they still may have problems with their prosthesis including: Asymmetry in forces applied on the lower limbs, high energy consumption and also pain during walking [3–6]. Furthermore, subjects with unilateral BK are at risk of low back pain, knee and hip joint OA, especially in their intact limbs. Based on the results of previous studies, the incidence of knee and hip joint OA is high in intact sides of below knee amputees [7–12]. Subjects with BK typically employ compensatory mechanisms due to a lack of power generation in their ankle joint, which itself may increase the loads applied in the joints of the sound side, and due to limitations in the ankle kinematics (i.e. reduced dorsiflexion) of the ⁎
Corresponding author. E-mail address: [email protected] (S.I. Wolf).
http://dx.doi.org/10.1016/j.gaitpost.2017.08.007 Received 28 October 2016; Received in revised form 2 August 2017; Accepted 6 August 2017 0966-6362/ © 2017 Elsevier B.V. All rights reserved.
Gait & Posture 58 (2017) 246–251
M.T. Karimi et al.
As can be seen from the above mentioned studies, it is controversial whether the GRF of the contralateral side differ significantly from those in matched reference subjects without disability. This may be due to differing quality of prosthetic fitting, choice of prosthetic components, their alignment or any comorbidity associated with the amputation as e.g. proximal contractures and variation of the feet in the above mentioned studies. It has been shown that subjects with BK with a flex foot had a decrease in the peaks of GRF in the sound side [20]. Fey et al. showed that analyzing inter-segmental joint forces and moments did not give any reason for higher incidence of knee OA, since none of the knee kinetics parameters investigated showed higher values than in control knees. They also claimed that intact knee varus moment and force impulses in subjects with amputation were smaller than those of normal matched subjects [21]. Therefore, the high incidence of knee OA in subjects with BK may not be due to an increase in varus moment, as the results of most of the studies did not support this conclusion. Recently there is increasing awareness, that next to joint moments it might be joint contact forces acting on the knee that may play an important role in knee osteoarthritis [23–25]. In conventional 3D-gait modeling joint loads are assessed only indirectly by monitoring joint moments [26]. However, joint loads become directly available in the context of instrumented 3D gait analysis with the use of musculoskeletal software such as Anybody (Aarlborg, Denmark) or OpenSim (Stanford University, California, USA). Therefore, and due to conflicts in the literature in this regard, it was aimed to determine specifically joint contact forces (JCF) in the intact side of subjects with BK and compare these to reference subjects to possibly find a more sensitive indicator for early OA in these subjects. We therefore specifically hypothesized in this work that a) the joint kinematics of the sound side in subjects with BK is different to age matched reference subjects leading to b) larger knee joint contact forces in this group and c) that increased joint contact forces are associated with increased coronal plane knee moments.
Levels were described, for instance by Gailey et al. [27]. Details on the participants with a below knee amputation can be found in Table 1. The subjects of the RG group were healthy volunteers from the staff and students of the center. An ethical approval was obtained from the ethical committee of University Hospital. Moreover, each subject was asked to sign a consent form before data collection. It should be emphasized that all of the involved subjects who participated in this study used a Vari-Flex© foot (Ossur HF, Iceland). Otherwise, a Vari-Flex foot was fitted to their prosthesis by in-house prosthesists. The alignment of the prosthesis was adjusted based on the work of Blumentritt et al. and evaluated by using LASAR posture (Otto Bock, Duderstadt, Germany) [31]. The authors gave clear alignment recommendations for trans-tibial prostheses in this study, which are the following: In the sagittal plane the vertical GRF component, or the LASAR Posture laser-line should be approximately 15 mm in front of the compromise knee axis and cross the foot at one third of the foot length, measured from the heel. In the frontal plane, the vertical GRF/laser-line should be tangential to the lateral edge of the patella of the residual limb and cross the prosthetic foot through the medial aspect mimicking the big toe. All prostheses facilitated in this study were aligned according to these recommendations. Blumentritt and colleagues were able to prove that these recommendations lead to a significant decrease in knee joint coronal plane loads on the involved side [31] which were followed in this study but not documented. Based on the results of various studies, the structure of the prosthetic foot and its mechanical properties influence the gait performance of the user and also the loads applied to the body. Therefore, and in order to delete the different effects of various feet on the gait parameters of the users, it was decided that only one type of foot would be used in this study. The Vari-Flex foot was selected for this study since it can be regarded as a typical energy storing and returning (ESAR) foot. Moreover, the users were allowed to walk with the new foot for at least two weeks before data collection.
2. Methods
2.1. Equipment
A group of seven male adults with BK all of the left leg (age: 39.4 ± 12 years, height: 1.84 ± 0.07 m) and an age and height matched group of ten reference subjects (2 female, 8 male age: 36.6 ± 12 years, height: 1.81 ± 0.07 m) without gait disability (RG) were recruited for this study. The BK group was recruited from the outpatient clinic for prosthetics as well at the in house prosthetic workshop of the Clinic for Orthopedic Surgery and Traumatology Heidelberg. The main criteria to select the subjects were non vascular induced BK amputation (e.g. trauma and tumor), use of the current prosthesis for at least 6 months, no pain or pressure sores of the residual limb, ability to ambulate independently without use of any assistive devices (e.g. crutches or canes), no proximal contracture or other neuromuscular diseases influencing their standing and walking abilities. In this study, participants with BK with a K-Level 3–4 according to Medicare Functional Classification (MFCL) were included. MFCL K-
A motion analysis system with 12 cameras (M-Cams, Vicon Motion system Limited, Oxford, UK) was used to record the motion of the subjects. The ground reaction force (GRF) was recorded by use of two Kistler force plates.
2.2. Parameters The peak GRF in anteroposterior (gait) direction, mediolateral and vertical directions, range of motions of ankle, knee, and hip, the peak moments acting on the above mentioned joints and specifically the peak joint contact forces (JCF) in three dimensions were evaluated in this study as a possible key factor for early OA in these subjects. Moreover, the spatiotemporal gait parameters were reported in this study.
Table 1 Patient and prosthesis characteristics. Manufacturers: 1Össur hf; Reykjavik; Iceland, 2Blatchford Group, Basingstoke; UK, 3Otto Bock Healthcare; Duderstadt; Germany. ID
height [m]
mass [kg]
Age [y]
time since amputation [y]
cause of amputation
suspension method
50621 52605 52644 53430 53516 53534 54003 Mean SD Max Min
187 183 191 187 183 168 190 184.1 7.8 191.0 168.0
105 73 89 94.3 69 83 95 86.9 12.8 105.0 69.0
48 24.5 55.9 43.6 23.7 35.9 44.2 39.4 12.0 55.9 23.7
11 8 32 2 4 26 4 12.4 11.8 32.0 2.0
tumour tumour bone abscess tumour tumour trauma trauma
Synergy Wave1/pin Synergy Wave1/pin Synergy Wave1/pin Synergy Wave1/pin Seal in X51/valve Synergy Wave1/pin Seal in X51/valve
247
habitual prosthetic foot lock lock lock lock lock
Variflex1 Variflex XC1 Variflex XC1 Flex-Foot Assure1 Reflex Shock1 Variflex1 Variflex XC1
Gait & Posture 58 (2017) 246–251
M.T. Karimi et al.
2.3. Procedure
Table 2 The mean values of spatiotemporal gait parameters and ground reaction force components (GRF) of normal and below knee amputees, Fx: antero-posterior component, FY: vertical component, FZ: mediolateral component, (N/BW = Newton/body weight, the numbers 1 and 2 refer to the first and second force peaks).
The subjects were asked to walk at a comfortable speed along a 10 m walkway to collect five successful trials. The successful trials were determined based on full placement of the foot (both right and left side) on the force plates. Force plate data were collected at 1080 Hz and optical marker capture was performed at 120 Hz. Data was filtered by Butterworth low pass filter with cut off frequency of 10 Hz. The marker placement used in this study was based on the Plugin Gait model based on the work of Kadaba et al. [32]. Inverse dynamics kinetic parameters beside joint contact forces (e.g. knee moment) and kinematics (e.g. pelvic kinematics) were determined with the Plugin Gait Model (Vicon, Oxford, United Kingdom). Using Mokka, (powered by the open-source library Biomechanical ToolKit (BTK), Copyright© Arnaud Barré, http://biomechanical-toolkit. github.io/mokka/), the data output of the software by Vicon were converted to be used in OpenSim. The generic model 2392, a lower extremities musculoskeletal model with 23 ° of freedom and 92 muscles from OpenSim (Version 3.3) was selected to model gait. Using static trial data, all subject data were scaled aiming to reach less than 0.02 m difference between virtual and real marker position. Using walking trials, inverse kinematics and inverse dynamics tools and a residual reduction algorithm (RRA) tool were applied to adjust mass and mass center location. All muscle excitations were computed using the computer muscle control tool (CMC) [33]. The last step was the calculation of JCF. The term JCF refers to point load transmission through the model segments applied to the joint center (Eq. (1)) JCFdist + JCFprox + Fmus + GRF = Mseg.Aseg
Parameters
Normal Subjects
Amputees
P-value of difference
Cadence (steps/min) Speed (m/s) Stride length (m) Fx1 (N/BW) Fx2 (N/BW) FY1 (N/BW) FY2 (N/BW) FZ (N/BW)
111.84( ± 5.65) 1.45( ± 0.12) 1.55( ± 0.09) 0.21( ± 0.02) 0.23( ± 0.03) 1.16( ± 0.06) 1.17( ± 0.11) 0.07( ± 0.02)
110.3( ± 10.1) 1.37( ± 0.18) 1.49( ± 0.11) 0.20( ± 0.04) 0.23( ± 0.03) 1.22( ± 0.11) 1.09( ± 0.11) 0.082( ± 0.02)
0.72 0.35 0.25 0.34 0.89 0.26 0.16 0.21
mean values of groups was evaluated by two sample t-tests (p = 0.05). All statistical analysis was done with Minitab software 17 (Minitab Inc., Penn State University, USA). Finally, we ran Pearson correlation analysis between maximum values of knee total JCF and coronal plane knee moments determined by the Plugin Gait Model.
3. Results The mean values of spatiotemporal gait parameters and ground reaction force comparisons are shown in Table 2. There were no significant differences between the average group values of spatiotemporal gait parameters (stride length, cadence and gait speed) between groups BK and RG (p > 0.05). The mean values of the first peak of vertical component of GRF were 1.22 ± 0.11 and 1.16 ± 0.06 (N/ BW), for BK and RG, respectively (p = 0.26). To avoid any misconception of the data and its interpretation, we want to clearly state that the following parameters are exclusively calculated for the sound side of the BK group. There were no differences between ankle, knee, and hip joint range of motion. One subject did in fact show hip osteoarthritis in the involved side, however, without functional restrictions for normal walking and without any peculiarities in the gait pattern, pain was reported only after walking distances exceeding 5 km. For more information on ranges of motion of ankle, knee, hip, and pelvis see Table 4 in the Appendix. The range of antero-posterior pelvic tilt,
(1)
where JCFdist is the compressive joint contact force (N) applied to the distal joint, JCFprox is the joint force (N) applied to the proximal joint, GRF is the ground reaction force (N), Fmus are the muscle forces (N) applied on the distal joint, M is the mass matrix of the segment, and A is a 6 dimensional vector of rotational and translational acceleration of the segment [34,35]. Fig. 1 shows the procedure used in this study to determine joint contact forces. The peaks of GRF and JCF components were normalized to body weight (BW) and the peaks of the moments were normalized to body mass (kg). Normal distribution of these parameters was confirmed by use of the ShapiroWilk test and subsequently the difference between the
Fig. 1. The procedure used in OpenSim software to determine joint contact forces.
248
Gait & Posture 58 (2017) 246–251
M.T. Karimi et al.
side moments of hip and knee were higher in the RG, these differences were not significant. Interestingly, the peaks of joint contact forces of hip and knee did not differ significantly from those in RG. Moreover, the peaks of most of the components were even slightly smaller. This fact might be due to the slightly slower walking speed, but again subjects with BK did not walk significantly slower (see Table 2 with TSP results). Contrarily, the antero-posterior and vertical components of JCF of the ankle in the BK group were slightly smaller. Therefore, the data from this study does not support a particular increase in JCF in subjects with BK. As can be seen from the results, GRF in the sound side of subjects with BK did not differ substantially from those in the RG group. However, we found slightly increased hip and knee joint moments in the BK group. The results of this research differed from the results of studies done by Beyaert, and Engsberg et al. [14,19]. However, it supports the results of the studies done by Hurley et al. and Lemaire et al. [15,16]. The difference in the results may be due to the type of prosthetic feet used in the aforementioned studies. It has been shown that energy storing and returning prosthetic feed are able to reduce loads onto the sound side [29,30]. However, socket design and fitting as well as prosthetic alignment may also be responsible. On the contrary, the feet used in all studies were partly ESAR and partly non ESAR feet and this could not totally explain the difference. In the research performed by Beyaert, 15 subjects with various foot types were recruited (mostly with Re-flex, Flex-walk and SACH foot) [14]. Moreover, in the study by Engsberg, children with BK were recruited using mostly SACH and Seattle feet [19]. They mentioned that intact limbs were subject to greater loading, impulse and time of loading. The subjects participating in the study by Hurley et al. used mostly Seattle and Flex feet where the Flex feet could be categorized as ESAR while Seattle feet are not typical energy returning feet containing carbon fiber leaf springs [15]. In Lemaire et al., elderly subjects with BK mostly used SACH, Seattle, multiflex and flex-foot [16]. The difference between the aforementioned studies related to age of the participants and the type of selected feet. For example, the walking speed of elderly subjects is reduced compared to the speed of young adults and there is no doubt that reduced walking speed leads to decreased peak loads [16]. The comparison of the results of this study to findings of other studies confirmed that the forces and moments applied on the joints of the intact side in subjects with BK did not differ from those of normal subjects. Specifically, the joint contact force is a combination of all forces applied on a joint while walking including ground reaction force, muscle force, ligament and surrounding tissue forces of a joint [35]. In some studies the varus moment was used as an indicator for the loads acting on a joint, however, with development of musculoskeletal software such as OpenSim it is possible to derive joint contact forces [17,22,34] that may be a much more sensitive indicator for early OA. The results of JCF analysis performed in this study showed that JCF components were not larger in subjects with BK compared to a reference group, but some components tended to be even smaller. The slight correlation between JCF and knee coronal moment in the BK group may show that there is a relationship between increased JCF and increased coronal plane moments, but further studies are needed to have a solid conclusion. It should be emphasized that all subjects in the current study walked with a Vari-Flex foot (Össur HF, Reykjavik, Iceland). However, the effects of other prosthesis components such as suspension system, type of socket prosthetic alignment and rehabilitation, e.g. physical education in this group of BK may have been particularly beneficial. Based on the available literature, an increase in the loads applied on a joint may be due to misalignment of a limb, use of compensatory strategy used by subjects as a result of lack of ankle joint power and a change in walking speed [16,36]. However, the mean values of spatiotemporal gait parameters of the amputees in this study were almost the same as that of the normal subjects (Table 2). Although participants with BK walked slightly slower we did not find a statistical difference to
Table 3 The mean values of JCF components of normal and BK amputees, Fx: antero-posterior component, Fy: Vertical component and Fz: mediolateral component (the numbers 1 and 2 refer to the first and second force peaks). Parameters (N/BW)
Normal Subjects
Amputees
P-Value of difference
Ankle JCF Fx1 Ankle JCFFx2 Ankle JCF FY1 Ankle JCF FY2 Ankle JCF FZ Knee JCF Fx1 Knee JCF Fx2 Knee JCF FY1 Knee JCF FY2 Knee JCF FZ Hip JCF Fx1 Hip JCF Fx2 Hip JCF FY1 Hip JCF FY2 Hip JCF FZ
1.26( ± 0.31) 3.25( ± 0.21) 5.30( ± 0.85) 8.05( ± 0.64) 0.36( ± 0.20) 1.74( ± 0.40) 1.47( ± 0.19) 3.87( ± 0.31) 4.21( ± 0.49) 0.34( ± 0.11) 1.51( ± 0.45) 2.53( ± 0.66) 5.33( ± 0.51) 4.51( ± 0.65) 1.37( ± 0.17)
1.07( ± 0.26) 2.78( ± 0.48) 5.13( ± 0.78) 7.10( ± 0.95) 0.24( ± 0.11) 1.63( ± 0.52) 1.56( ± 0.14) 4.07( ± 0.64) 4.13( ± 0.34) 0.30( ± 0.17) 1.25( ± 0.34) 2.65( ± 0.50) 5.29( ± 0.59) 4.37( ± 0.83) 1.41( ± 0.20)
0.23 0.05 0.68 0.04 0.16 0.67 0.32 0.47 0.71 0.61 0.21 0.68 0.89 0.73 0.71
however, was significantly higher in BK (P < 0.05) whereas the range of pelvic obliquity was smaller (6.96 ± 1.44° in BK compared to 10 ± 2.4° in RG, respectively (p = 0.006)). The group average maximum value of ankle joint moment in BK was less than that of RG; however, the difference was not significant (see Table 5 in the Appendix). The results of joint contact force (JCF) analysis for BK and RG are shown in Table 3. The group average maximum values of antero-posterior component of the JCF were smaller in BK, however, differences were only significant for second peaks of antero-posterior and vertical components (p < 0.05). Regarding knee joint contact forces, some components showed slightly higher values. In contrast, the group average maximum value of the first peak of antero-posterior force, the second peak of vertical component of knee JCF and also the mediolateral force was slightly higher in BK compared to those in RG. Also, the hip joint contact force components were evaluated in this study. As can be seen from Table 3, there was no significant difference between BK and RG. Although the group average maximum value of some components of the joint contact force were smaller, the difference was not significant. Fig. 2 illustrates the knee joint flexion/extension JCF components for a sample BK amputee and a reference subject. The correlation between maximum total knee JCF (4.37 ± 0.39 N/BW) and maximum knee coronal moment derived from Plugin Gait Model (0.704 ± 0.169 Nm/kg) in BK is 0.65 (p = 0.08) and for BG is 0.63 (p = 0.3). 4. Discussion Current literature supports the finding that the incidence of hip and knee pain and OA in prosthesis users is higher than that in the normal population [10]. The main question posed here is which parameters influence the wear of the joints of the non-involved side in subjects with lower limb amputation and to specifically monitor JCF, which may be a major factor for early OA that has not yet been investigated in subjects with BK. Increased joint wear may be due to an increase in loads acting on the joints in the presence of compensatory mechanisms applied by the subjects to overcome the lack of ankle joint power [16]. However, the results of various studies are controversial about whether the loads (forces and moments) applied on hip and knee joints in the intact side of subjects with BK amputation were higher than those in subjects without gait disability. Therefore, the aim of this study was to check the loads acting on the intact side of subjects with BK and specifically to monitor JCF in these subjects. The results of this study confirmed that the peaks of ground reaction force components of the sound side in subjects with BK did not differ significantly from RG, Table 2. Although most components of the sound 249
Gait & Posture 58 (2017) 246–251
M.T. Karimi et al.
Fig. 2. Sample data of antero-posterior (a), vertical (b), and medio-lateral component (c) of the knee joint contact force in for BK subject (Sound side) and a reference subject.
OA. Information on the level of inflammatory markers or results of a typical questionnaire used in the OA population, e.g. the WOMAC score, are missing for the BK participants investigated here and are a limitation of the study.
the RG. Although they showed typical asymmetries between the sound and the involved side, most of the joints kinematic and kinetic parameters of the sound side showed no significant differences compared to normal subjects (Table 4 and 5 in the Appendix). Therefore, it can be summarized that the selection of prosthesis components including an ESAR prosthetic foot, the overall physical condition of participants with BK and possibly the prosthetic fitting and alignment was quite beneficial in this group of participants with BK. However, we do not have detailed information on socket fit or alignment to quantify that it was truly well performed. As a consequence, JCF in the sound side of these subjects was not higher compared to those in reference subjects without disability. In the absence of this finding, it should be emphasized that the output of this study does not support or approve any mechanisms that increases joint contact forces which could explain the increased incidence of OA in the BK population. For elucidating such mechanisms, it might be more effective for future studies to specifically select patients showing frontal plane gait deviations with their prosthesis to confirm or refute the hypothesis that increased coronal plane knee joints go hand in hand with higher joint contact forces. Lastly, it has been shown that the onset and course of OA can only partly be referred to the loading situation of the limb. Metabolic changes, e.g. increased inflammatory markers, promote the incidence and negative course of
5. Conclusion It turned out that the group of seven BK amputees showed hardly any differences in joint kinematics and kinetics in level walking when compared to an age matched group of controls. This may be explained by an overall positive clinical situation of the subjects, a beneficial prosthesis including proper socket fit, selection of ESAR prosthetic feet (in this case a Vari-Flex© ESAR foot) which had been shown to have beneficial impact on sound side loads [29,30], and good prosthesis alignment as checked via a LASAR posture and following the guidelines of Blumentritt and colleagues [28]. Consequently, the JCF in the sound side as calculated via OpenSim also did not differ significantly. Hence, in this patient cohort we could not find any biomechanical indicator for a possible early onset of osteoarthritis, except for the one subject who did in fact show osteoarthritis but in the hip of the involved side possibly partly due to the condition 32 years after amputation. Nevertheless, the approach to use a musculoskeletal model for 250
Gait & Posture 58 (2017) 246–251
M.T. Karimi et al.
knee amputees, Gait Posture 28 (2) (2008) 278–284. [15] G.R. Hurley, R. McKenney, M. Robinson, M. Zadravec, M.R. Pierrynowski, The role of the contralateral limb in below-knee amputee gait, Prosthet. Orthot. Int. 14 (1) (1990) 33–42. [16] E.D. Lemaire, F.R. Fisher, D.G. Robertson, Gait patterns of elderly men with transtibial amputations, Prosthet. Orthot. Int. 17 (1) (1993) 27–37. [17] L. Sharma, D.E. Hurwitz, E.J. Thonar, J.A. Sum, M.E. Lenz, D.D. Dunlop, et al., Knee adduction moment, serum hyaluronan level, and disease severity in medial tibiofemoral osteoarthritis, Arthritis Rheum. 41 (7) (1998) 1233–1240. [18] C.M. Powers, L. Torburn, J. Perry, E. Ayyappa, Influence of prosthetic foot design on sound limb loading in adults with unilateral below-knee amputations, Arch. Phys. Med. Rehabil. 75 (7) (1994) 825–829. [19] J.R. Engsberg, A.G. Lee, J.L. Patterson, J.A. Harder, External loading comparisons between able-bodied and below-knee-amputee children during walking, Arch. Phys. Med. Rehabil. 72 (9) (1991) 657–661. [20] R.D. Snyder, C.M. Powers, C. Fontaine, J. Perry, The effect of five prosthetic feet on the gait and loading of the sound limb in dysvascular below-knee amputees, J. Rehabil. Res. Dev. 32 (4) (1995) 309–315. [21] N.P. Fey, R.R. Neptune, 3D intersegmental knee loading in below-knee amputees across steady-state walking speeds, Clin. Biomech. 27 (4) (2012) 409–414. [22] D.C. Morgenroth, A.D. Segal, K.E. Zelik, J.M. Czerniecki, G.K. Klute, P.G. Adamczyk, et al., The effect of prosthetic foot push-off on mechanical loading associated with knee osteoarthritis in lower extremity amputees, Gait Posture 34 (4) (2011) 502–507. [23] R.H. Miller, A.Y. Esterson, J.K. Shim, Joint contact forces when minimizing the external knee adduction moment by gait modification: a computer simulation study, Knee 22 (6) (2015) 481–489. [24] H. Marouane, A. Shirazi-Adl, M. Adouni, Alterations in knee contact forces and centers in stance phase of gait: a detailed lower extremity musculoskeletal model, J. Biomech. 49 (2) (2016) 185–192. [25] G.T. Harding, M.J. Dunbar, C.L. Hubley-Kozey, W.D. Stanish, J.L. Wilson, Obesity is associated with higher absolute tibiofemoral contact and muscle forces during gait with and without knee osteoarthritis, Clin. Biomech. 31 (2016) 79–86. [26] S.L. Delp, F.C. Anderson, A.S. Arnold, P. Loan, A. Habib, C.T. John, et al., OpenSim: open-source software to create and analyze dynamic simulations of movement, IEEE Trans. Biomed. Eng. 54 (11) (2007) 1940–1950. [27] R.S. Gailey, K.E. Roach, E.B. Applegate, B. Cho, B. Cunniffe, S. Licht, M. Maguire, M.S. Nash, The amputee mobility predictor: an instrument to assess determinants of the lower-limb amputee's ability to ambulate, Arch. Phys. Med. Rehabil. 83 (5) (2002) 613–627. [28] S. Blumentritt, T. Schmalz, R. Jarasch, Die Bedeutung des statischen Prothesenaufbaus für das Stehen und Gehen des Unterschenkelamputierten, Der Orthop. 30 (3) (2001) 161–168. [29] A.D. Segal, K.E. Zelik, G.K. Klute, D.C. Morgenroth, M.E. Hahn, M.S. Orendurff, P.G. Adamczyk, S.H. Collins, A.D. Kuo, J.M. Czerniecki, The effects of a controlled energy storage and return prototype prosthetic foot on transtibial amputee ambulation, Hum. Mov. Sci. 31 (4) (2012) 918–931. [30] D.C. Morgenroth, A.D. Segal, K.E. Zelik, J.M. Czerniecki, G.K. Klute, P.G. Adamczyk, M.S. Orendurff, M.E. Hahn, S.H. Collins, A.D. Kuo, The effect of prosthetic foot push-off on mechanical loading associated with knee osteoarthritis in lower extremity amputees, Gait Posture 34 (4) (2011) 502–507. [31] S. Blumentrttt, T. Schmalz, R. Jarasch, M. Schneider, Effects of sagittal plane prosthetic alignment on standing trans-tibial amputee knee loads, Prosthet. Orthot. Int. 23 (December (3)) (1999) 231–238. [32] M.P. Kadaba, H.K. Ramakrishnan, M.E. Wootten, J. Gainey, G. Gorton, G.V. Cochran, Repeatability of kinematic, kinetic, and electromyographic data in normal adult gait, J. Orthop. Res. (1989) 7. [33] D.G. Thelen, F.C. Anderson, S.L. Delp, Generating dynamic simulations of movement using computed muscle control, J. Biomech. 36 (3) (2003) 321–328. [34] K.M. Steele, M.S. DeMers, M.H. Schwartz, S.L. Delp, Compressive tibiofemoral force during crouch gait, Gait Posture 35 (April (4)) (2012) 556–560. [35] J.P. Walter, N. Korkmaz, B.J. Fregly, M.G. Pandy, Contribution of tibiofemoral joint contact to net loads at the knee in gait, J. Orthop. Res. 33 (July (7)) (2015) 1054–1060. [36] D.A. Winter, S.E. Sienko, Biomechanics of below-knee amputee gait, J. Biomech. (1988) 21.
directly deriving JCF appears promising when monitoring factors for early osteoarthritis. Future studies should therefore concentrate on subjects presenting with compensatory movements in gait to study their influence on JCF, perhaps also in gait conditions other than level walking. Conflicts of interest All prosthetic feet used in this study were provided by Össur hf, Reykjavik, Iceland. Össur hf was not involved in writing the manuscript or interpretation of data. Acknowledgement We thank Saskia Ellis for careful English language editing. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.gaitpost.2017.08.007. References [1] K. Ziegler-Graham, E.J. MacKenzie, P.L. Ephraim, T.G. Travison, R. Brookmeyer, Estimating the prevalence of limb loss in the United States: 2005 to 2050, Arch. Phys. Med. Rehabil. 89 (3) (2008) 422–429. [2] P.W. Moxey, P. Gogalniceanu, R.J. Hinchliffe, I.M. Loftus, K.J. Jones, M.M. Thompson, P.J. Holt, Lower extremity amputations—a review of global variability in incidence, Diabet. Med. 28 (10) (2011) 1144–1153. [3] A.S.O.D.C. Soares, E.Y. Yamaguti, L. Mochizuki, A.C. Amadio, J.C. Serrão, Biomechanical parameters of gait among transtibial amputees: a review, Sao Paulo Med. J. 127 (2009) 302–309. [4] A.H. Vrieling, H.G. van Keeken, T. Schoppen, E. Otten, J.P.K. Halbertsma, A.L. Hof, et al., Gait initiation in lower limb amputees, Gait Posture 27 (3) (2008) 423–430. [5] N. Aslani, S. Noroozi, K.S. Yee, A.O.Z. Chao, C. Maggs, Simulation of gait asymmetry and energy transfer efficiency between unilateral and bilateral amputees, Sports Eng. 19 (3) (2016) 163–170. [6] E. Isakov, O. Keren, N. Benjuya, Trans-tibial amputee gait: time-distance parameters and EMG activity, Prosthet. Orthot. Int. 24 (3) (2000) 216–220. [7] D.C. Morgenroth, A.C. Gellhorn, P. Suri, Osteoarthritis in the disabled population: a mechanical perspective, PM & R 4 (5, Suppl) (2012) S20–S27. [8] A. Kusljugic, S. Kapidzic-Durakovic, Z. Kudumovic, A. Cickusic, Chronic low back pain in individuals with lower-limb amputation, Bosn. J. Basic Med. Sci. 6 (2) (2006) 67–70. [9] R. Gailey, K. Allen, J. Castles, J. Kucharik, M. Roeder, Review of secondary physical conditions associated with lower-limb amputation and long-term prosthesis use, J. Rehabil. Res. Dev. 45 (1) (2008) 15–29. [10] P.A. Struyf, C.M. van Heugten, M.W. Hitters, R.J. Smeets, The prevalence of osteoarthritis of the intact hip and knee among traumatic leg amputees, Arch. Phys. Med. Rehabil. 90 (3) (2009) 440–446. [11] M.J. Burke, V. Roman, V. Wright, Bone and joint changes in lower limb amputees, Ann. Rheum. Dis. 37 (3) (1978) 252–254. [12] J. Kulkarni, J. Adams, E. Thomas, A. Silman, Association between amputation, arthritis and osteopenia in British male war veterans with major lower limb amputations, Clin. Rehabil. 12 (4) (1998) 348–353. [13] S. Tanamas, F.S. Hanna, F.M. Cicuttini, A.E. Wluka, P. Berry, D.M. Urquhart, Does knee malalignment increase the risk of development and progression of knee osteoarthritis? A systematic review, Arthritis Rheum. 61 (4) (2009) 459–467. [14] C. Beyaert, C. Grumillier, N. Martinet, J. Paysant, J.M. Andre, Compensatory mechanism involving the knee joint of the intact limb during gait in unilateral below-
251
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Full length article
Sound side joint contact forces in below knee amputee gait with an ESAR prosthetic foot
MARK
Mohammad Taghi Karimia,c, Firooz Salamib, Amir Esrafiliand,e, Daniel W.W. Heitzmannb, ⁎ Merkur Alimusajb, Cornelia Putzb, Sebastian I. Wolfb, a
Musculoskeletal Research Center, Isfahan University of Medical Sciences, Isfahan, Iran Clinic for Orthopaedics and Trauma Surgery, Heidelberg University Hospital, Schlierbacher Landstrasse 200a, 69118 Heidelberg, Germany c Rehabilitation Sciences Research Center, Shiraz University of Medical Sciences, Shiraz Iran d Department of Biomedical Engineering, Amirkabir University of Technology, Tehran, Iran e Department of Applied Physics, University of Eastern Finland, Kuopio, Finland b
A R T I C L E I N F O
A B S T R A C T
Keywords: Below knee amputee Prosthesis Joint contact force Osteoarthritis Gait
The incidence of knee and hip joint osteoarthritis in subjects with below knee amputation (BK) appears significantly higher compared to unimpaired subjects, especially in the intact side. However, it is controversial if constant higher loads on the sound side are one of the major factors for an increased osteoarthritis (OA) incidence in subjects with BK, beside other risk factors, e.g. with respect to metabolism. The aim wasto investigate joint contact forces (JCF) calculated by a musculoskeletal model in the intact side and to compare it with those of unimpaired subjects and to further elucidate in how far increased knee JCF are associated with increased frontal plane knee moments. A group of seven subjects with BK amputation and a group of ten unimpaired subjects were recruited for this study. Gait data were measured by 3D motion capture and force plates. OpenSim software was applied to calculate JCF. Maximum joint angles, ground reaction forces, and moments as well as time distance parameters were determined and compared between groups showing no significant differences, with some JCF components of knee and hip even being slightly smaller in subjects with BK compared to the reference group. This positive finding may be due to the selected ESAR foot. However, other beneficial factors may also have influenced this positive result such as the general good health status of the subjects or the thorough and proper fitting and alignment of the prosthesis.
1. Introduction
prosthesis [13–16]. Sharma et al. showed that there is a correlation between magnitude of knee varus joint moment and OA symptoms based on radiologic classification [17]. Therefore, an increase in incidence of knee OA in the intact side of BK amputees may be due to an increase in applied loads. If this is the case then it should be aimed at reducing loads by proper fitting, alignment of the prosthesis and an appropriate choice of prosthetic components [18]. In the research done by Beyart et al., it was shown that the peaks of vertical components of ground reaction force (GRF) were higher in the intact limb of subjects with BK compared to the contralateral side and to normal subjects [14], respectively. Engsberg et al. also showed that subjects with BK had a greater rate of loading in their sound side [19] (eleven able bodied’ children and four children with BK amputation participated in this study). In contrast, Hurley et al. showed that the forces acting across the joints of the contralateral side in adults with BK were not significantly higher than those in subjects without disability [15].
The number of subjects with amputation, especially in the lower limb, is increasing for several reasons such as rising incidence of diabetes mellitus, average age and social deprivation [1,2]. Although various prosthetic components have been designed for subjects with lower limb amputation to restore their abilities while walking, they still may have problems with their prosthesis including: Asymmetry in forces applied on the lower limbs, high energy consumption and also pain during walking [3–6]. Furthermore, subjects with unilateral BK are at risk of low back pain, knee and hip joint OA, especially in their intact limbs. Based on the results of previous studies, the incidence of knee and hip joint OA is high in intact sides of below knee amputees [7–12]. Subjects with BK typically employ compensatory mechanisms due to a lack of power generation in their ankle joint, which itself may increase the loads applied in the joints of the sound side, and due to limitations in the ankle kinematics (i.e. reduced dorsiflexion) of the ⁎
Corresponding author. E-mail address: [email protected] (S.I. Wolf).
http://dx.doi.org/10.1016/j.gaitpost.2017.08.007 Received 28 October 2016; Received in revised form 2 August 2017; Accepted 6 August 2017 0966-6362/ © 2017 Elsevier B.V. All rights reserved.
Gait & Posture 58 (2017) 246–251
M.T. Karimi et al.
As can be seen from the above mentioned studies, it is controversial whether the GRF of the contralateral side differ significantly from those in matched reference subjects without disability. This may be due to differing quality of prosthetic fitting, choice of prosthetic components, their alignment or any comorbidity associated with the amputation as e.g. proximal contractures and variation of the feet in the above mentioned studies. It has been shown that subjects with BK with a flex foot had a decrease in the peaks of GRF in the sound side [20]. Fey et al. showed that analyzing inter-segmental joint forces and moments did not give any reason for higher incidence of knee OA, since none of the knee kinetics parameters investigated showed higher values than in control knees. They also claimed that intact knee varus moment and force impulses in subjects with amputation were smaller than those of normal matched subjects [21]. Therefore, the high incidence of knee OA in subjects with BK may not be due to an increase in varus moment, as the results of most of the studies did not support this conclusion. Recently there is increasing awareness, that next to joint moments it might be joint contact forces acting on the knee that may play an important role in knee osteoarthritis [23–25]. In conventional 3D-gait modeling joint loads are assessed only indirectly by monitoring joint moments [26]. However, joint loads become directly available in the context of instrumented 3D gait analysis with the use of musculoskeletal software such as Anybody (Aarlborg, Denmark) or OpenSim (Stanford University, California, USA). Therefore, and due to conflicts in the literature in this regard, it was aimed to determine specifically joint contact forces (JCF) in the intact side of subjects with BK and compare these to reference subjects to possibly find a more sensitive indicator for early OA in these subjects. We therefore specifically hypothesized in this work that a) the joint kinematics of the sound side in subjects with BK is different to age matched reference subjects leading to b) larger knee joint contact forces in this group and c) that increased joint contact forces are associated with increased coronal plane knee moments.
Levels were described, for instance by Gailey et al. [27]. Details on the participants with a below knee amputation can be found in Table 1. The subjects of the RG group were healthy volunteers from the staff and students of the center. An ethical approval was obtained from the ethical committee of University Hospital. Moreover, each subject was asked to sign a consent form before data collection. It should be emphasized that all of the involved subjects who participated in this study used a Vari-Flex© foot (Ossur HF, Iceland). Otherwise, a Vari-Flex foot was fitted to their prosthesis by in-house prosthesists. The alignment of the prosthesis was adjusted based on the work of Blumentritt et al. and evaluated by using LASAR posture (Otto Bock, Duderstadt, Germany) [31]. The authors gave clear alignment recommendations for trans-tibial prostheses in this study, which are the following: In the sagittal plane the vertical GRF component, or the LASAR Posture laser-line should be approximately 15 mm in front of the compromise knee axis and cross the foot at one third of the foot length, measured from the heel. In the frontal plane, the vertical GRF/laser-line should be tangential to the lateral edge of the patella of the residual limb and cross the prosthetic foot through the medial aspect mimicking the big toe. All prostheses facilitated in this study were aligned according to these recommendations. Blumentritt and colleagues were able to prove that these recommendations lead to a significant decrease in knee joint coronal plane loads on the involved side [31] which were followed in this study but not documented. Based on the results of various studies, the structure of the prosthetic foot and its mechanical properties influence the gait performance of the user and also the loads applied to the body. Therefore, and in order to delete the different effects of various feet on the gait parameters of the users, it was decided that only one type of foot would be used in this study. The Vari-Flex foot was selected for this study since it can be regarded as a typical energy storing and returning (ESAR) foot. Moreover, the users were allowed to walk with the new foot for at least two weeks before data collection.
2. Methods
2.1. Equipment
A group of seven male adults with BK all of the left leg (age: 39.4 ± 12 years, height: 1.84 ± 0.07 m) and an age and height matched group of ten reference subjects (2 female, 8 male age: 36.6 ± 12 years, height: 1.81 ± 0.07 m) without gait disability (RG) were recruited for this study. The BK group was recruited from the outpatient clinic for prosthetics as well at the in house prosthetic workshop of the Clinic for Orthopedic Surgery and Traumatology Heidelberg. The main criteria to select the subjects were non vascular induced BK amputation (e.g. trauma and tumor), use of the current prosthesis for at least 6 months, no pain or pressure sores of the residual limb, ability to ambulate independently without use of any assistive devices (e.g. crutches or canes), no proximal contracture or other neuromuscular diseases influencing their standing and walking abilities. In this study, participants with BK with a K-Level 3–4 according to Medicare Functional Classification (MFCL) were included. MFCL K-
A motion analysis system with 12 cameras (M-Cams, Vicon Motion system Limited, Oxford, UK) was used to record the motion of the subjects. The ground reaction force (GRF) was recorded by use of two Kistler force plates.
2.2. Parameters The peak GRF in anteroposterior (gait) direction, mediolateral and vertical directions, range of motions of ankle, knee, and hip, the peak moments acting on the above mentioned joints and specifically the peak joint contact forces (JCF) in three dimensions were evaluated in this study as a possible key factor for early OA in these subjects. Moreover, the spatiotemporal gait parameters were reported in this study.
Table 1 Patient and prosthesis characteristics. Manufacturers: 1Össur hf; Reykjavik; Iceland, 2Blatchford Group, Basingstoke; UK, 3Otto Bock Healthcare; Duderstadt; Germany. ID
height [m]
mass [kg]
Age [y]
time since amputation [y]
cause of amputation
suspension method
50621 52605 52644 53430 53516 53534 54003 Mean SD Max Min
187 183 191 187 183 168 190 184.1 7.8 191.0 168.0
105 73 89 94.3 69 83 95 86.9 12.8 105.0 69.0
48 24.5 55.9 43.6 23.7 35.9 44.2 39.4 12.0 55.9 23.7
11 8 32 2 4 26 4 12.4 11.8 32.0 2.0
tumour tumour bone abscess tumour tumour trauma trauma
Synergy Wave1/pin Synergy Wave1/pin Synergy Wave1/pin Synergy Wave1/pin Seal in X51/valve Synergy Wave1/pin Seal in X51/valve
247
habitual prosthetic foot lock lock lock lock lock
Variflex1 Variflex XC1 Variflex XC1 Flex-Foot Assure1 Reflex Shock1 Variflex1 Variflex XC1
Gait & Posture 58 (2017) 246–251
M.T. Karimi et al.
2.3. Procedure
Table 2 The mean values of spatiotemporal gait parameters and ground reaction force components (GRF) of normal and below knee amputees, Fx: antero-posterior component, FY: vertical component, FZ: mediolateral component, (N/BW = Newton/body weight, the numbers 1 and 2 refer to the first and second force peaks).
The subjects were asked to walk at a comfortable speed along a 10 m walkway to collect five successful trials. The successful trials were determined based on full placement of the foot (both right and left side) on the force plates. Force plate data were collected at 1080 Hz and optical marker capture was performed at 120 Hz. Data was filtered by Butterworth low pass filter with cut off frequency of 10 Hz. The marker placement used in this study was based on the Plugin Gait model based on the work of Kadaba et al. [32]. Inverse dynamics kinetic parameters beside joint contact forces (e.g. knee moment) and kinematics (e.g. pelvic kinematics) were determined with the Plugin Gait Model (Vicon, Oxford, United Kingdom). Using Mokka, (powered by the open-source library Biomechanical ToolKit (BTK), Copyright© Arnaud Barré, http://biomechanical-toolkit. github.io/mokka/), the data output of the software by Vicon were converted to be used in OpenSim. The generic model 2392, a lower extremities musculoskeletal model with 23 ° of freedom and 92 muscles from OpenSim (Version 3.3) was selected to model gait. Using static trial data, all subject data were scaled aiming to reach less than 0.02 m difference between virtual and real marker position. Using walking trials, inverse kinematics and inverse dynamics tools and a residual reduction algorithm (RRA) tool were applied to adjust mass and mass center location. All muscle excitations were computed using the computer muscle control tool (CMC) [33]. The last step was the calculation of JCF. The term JCF refers to point load transmission through the model segments applied to the joint center (Eq. (1)) JCFdist + JCFprox + Fmus + GRF = Mseg.Aseg
Parameters
Normal Subjects
Amputees
P-value of difference
Cadence (steps/min) Speed (m/s) Stride length (m) Fx1 (N/BW) Fx2 (N/BW) FY1 (N/BW) FY2 (N/BW) FZ (N/BW)
111.84( ± 5.65) 1.45( ± 0.12) 1.55( ± 0.09) 0.21( ± 0.02) 0.23( ± 0.03) 1.16( ± 0.06) 1.17( ± 0.11) 0.07( ± 0.02)
110.3( ± 10.1) 1.37( ± 0.18) 1.49( ± 0.11) 0.20( ± 0.04) 0.23( ± 0.03) 1.22( ± 0.11) 1.09( ± 0.11) 0.082( ± 0.02)
0.72 0.35 0.25 0.34 0.89 0.26 0.16 0.21
mean values of groups was evaluated by two sample t-tests (p = 0.05). All statistical analysis was done with Minitab software 17 (Minitab Inc., Penn State University, USA). Finally, we ran Pearson correlation analysis between maximum values of knee total JCF and coronal plane knee moments determined by the Plugin Gait Model.
3. Results The mean values of spatiotemporal gait parameters and ground reaction force comparisons are shown in Table 2. There were no significant differences between the average group values of spatiotemporal gait parameters (stride length, cadence and gait speed) between groups BK and RG (p > 0.05). The mean values of the first peak of vertical component of GRF were 1.22 ± 0.11 and 1.16 ± 0.06 (N/ BW), for BK and RG, respectively (p = 0.26). To avoid any misconception of the data and its interpretation, we want to clearly state that the following parameters are exclusively calculated for the sound side of the BK group. There were no differences between ankle, knee, and hip joint range of motion. One subject did in fact show hip osteoarthritis in the involved side, however, without functional restrictions for normal walking and without any peculiarities in the gait pattern, pain was reported only after walking distances exceeding 5 km. For more information on ranges of motion of ankle, knee, hip, and pelvis see Table 4 in the Appendix. The range of antero-posterior pelvic tilt,
(1)
where JCFdist is the compressive joint contact force (N) applied to the distal joint, JCFprox is the joint force (N) applied to the proximal joint, GRF is the ground reaction force (N), Fmus are the muscle forces (N) applied on the distal joint, M is the mass matrix of the segment, and A is a 6 dimensional vector of rotational and translational acceleration of the segment [34,35]. Fig. 1 shows the procedure used in this study to determine joint contact forces. The peaks of GRF and JCF components were normalized to body weight (BW) and the peaks of the moments were normalized to body mass (kg). Normal distribution of these parameters was confirmed by use of the ShapiroWilk test and subsequently the difference between the
Fig. 1. The procedure used in OpenSim software to determine joint contact forces.
248
Gait & Posture 58 (2017) 246–251
M.T. Karimi et al.
side moments of hip and knee were higher in the RG, these differences were not significant. Interestingly, the peaks of joint contact forces of hip and knee did not differ significantly from those in RG. Moreover, the peaks of most of the components were even slightly smaller. This fact might be due to the slightly slower walking speed, but again subjects with BK did not walk significantly slower (see Table 2 with TSP results). Contrarily, the antero-posterior and vertical components of JCF of the ankle in the BK group were slightly smaller. Therefore, the data from this study does not support a particular increase in JCF in subjects with BK. As can be seen from the results, GRF in the sound side of subjects with BK did not differ substantially from those in the RG group. However, we found slightly increased hip and knee joint moments in the BK group. The results of this research differed from the results of studies done by Beyaert, and Engsberg et al. [14,19]. However, it supports the results of the studies done by Hurley et al. and Lemaire et al. [15,16]. The difference in the results may be due to the type of prosthetic feet used in the aforementioned studies. It has been shown that energy storing and returning prosthetic feed are able to reduce loads onto the sound side [29,30]. However, socket design and fitting as well as prosthetic alignment may also be responsible. On the contrary, the feet used in all studies were partly ESAR and partly non ESAR feet and this could not totally explain the difference. In the research performed by Beyaert, 15 subjects with various foot types were recruited (mostly with Re-flex, Flex-walk and SACH foot) [14]. Moreover, in the study by Engsberg, children with BK were recruited using mostly SACH and Seattle feet [19]. They mentioned that intact limbs were subject to greater loading, impulse and time of loading. The subjects participating in the study by Hurley et al. used mostly Seattle and Flex feet where the Flex feet could be categorized as ESAR while Seattle feet are not typical energy returning feet containing carbon fiber leaf springs [15]. In Lemaire et al., elderly subjects with BK mostly used SACH, Seattle, multiflex and flex-foot [16]. The difference between the aforementioned studies related to age of the participants and the type of selected feet. For example, the walking speed of elderly subjects is reduced compared to the speed of young adults and there is no doubt that reduced walking speed leads to decreased peak loads [16]. The comparison of the results of this study to findings of other studies confirmed that the forces and moments applied on the joints of the intact side in subjects with BK did not differ from those of normal subjects. Specifically, the joint contact force is a combination of all forces applied on a joint while walking including ground reaction force, muscle force, ligament and surrounding tissue forces of a joint [35]. In some studies the varus moment was used as an indicator for the loads acting on a joint, however, with development of musculoskeletal software such as OpenSim it is possible to derive joint contact forces [17,22,34] that may be a much more sensitive indicator for early OA. The results of JCF analysis performed in this study showed that JCF components were not larger in subjects with BK compared to a reference group, but some components tended to be even smaller. The slight correlation between JCF and knee coronal moment in the BK group may show that there is a relationship between increased JCF and increased coronal plane moments, but further studies are needed to have a solid conclusion. It should be emphasized that all subjects in the current study walked with a Vari-Flex foot (Össur HF, Reykjavik, Iceland). However, the effects of other prosthesis components such as suspension system, type of socket prosthetic alignment and rehabilitation, e.g. physical education in this group of BK may have been particularly beneficial. Based on the available literature, an increase in the loads applied on a joint may be due to misalignment of a limb, use of compensatory strategy used by subjects as a result of lack of ankle joint power and a change in walking speed [16,36]. However, the mean values of spatiotemporal gait parameters of the amputees in this study were almost the same as that of the normal subjects (Table 2). Although participants with BK walked slightly slower we did not find a statistical difference to
Table 3 The mean values of JCF components of normal and BK amputees, Fx: antero-posterior component, Fy: Vertical component and Fz: mediolateral component (the numbers 1 and 2 refer to the first and second force peaks). Parameters (N/BW)
Normal Subjects
Amputees
P-Value of difference
Ankle JCF Fx1 Ankle JCFFx2 Ankle JCF FY1 Ankle JCF FY2 Ankle JCF FZ Knee JCF Fx1 Knee JCF Fx2 Knee JCF FY1 Knee JCF FY2 Knee JCF FZ Hip JCF Fx1 Hip JCF Fx2 Hip JCF FY1 Hip JCF FY2 Hip JCF FZ
1.26( ± 0.31) 3.25( ± 0.21) 5.30( ± 0.85) 8.05( ± 0.64) 0.36( ± 0.20) 1.74( ± 0.40) 1.47( ± 0.19) 3.87( ± 0.31) 4.21( ± 0.49) 0.34( ± 0.11) 1.51( ± 0.45) 2.53( ± 0.66) 5.33( ± 0.51) 4.51( ± 0.65) 1.37( ± 0.17)
1.07( ± 0.26) 2.78( ± 0.48) 5.13( ± 0.78) 7.10( ± 0.95) 0.24( ± 0.11) 1.63( ± 0.52) 1.56( ± 0.14) 4.07( ± 0.64) 4.13( ± 0.34) 0.30( ± 0.17) 1.25( ± 0.34) 2.65( ± 0.50) 5.29( ± 0.59) 4.37( ± 0.83) 1.41( ± 0.20)
0.23 0.05 0.68 0.04 0.16 0.67 0.32 0.47 0.71 0.61 0.21 0.68 0.89 0.73 0.71
however, was significantly higher in BK (P < 0.05) whereas the range of pelvic obliquity was smaller (6.96 ± 1.44° in BK compared to 10 ± 2.4° in RG, respectively (p = 0.006)). The group average maximum value of ankle joint moment in BK was less than that of RG; however, the difference was not significant (see Table 5 in the Appendix). The results of joint contact force (JCF) analysis for BK and RG are shown in Table 3. The group average maximum values of antero-posterior component of the JCF were smaller in BK, however, differences were only significant for second peaks of antero-posterior and vertical components (p < 0.05). Regarding knee joint contact forces, some components showed slightly higher values. In contrast, the group average maximum value of the first peak of antero-posterior force, the second peak of vertical component of knee JCF and also the mediolateral force was slightly higher in BK compared to those in RG. Also, the hip joint contact force components were evaluated in this study. As can be seen from Table 3, there was no significant difference between BK and RG. Although the group average maximum value of some components of the joint contact force were smaller, the difference was not significant. Fig. 2 illustrates the knee joint flexion/extension JCF components for a sample BK amputee and a reference subject. The correlation between maximum total knee JCF (4.37 ± 0.39 N/BW) and maximum knee coronal moment derived from Plugin Gait Model (0.704 ± 0.169 Nm/kg) in BK is 0.65 (p = 0.08) and for BG is 0.63 (p = 0.3). 4. Discussion Current literature supports the finding that the incidence of hip and knee pain and OA in prosthesis users is higher than that in the normal population [10]. The main question posed here is which parameters influence the wear of the joints of the non-involved side in subjects with lower limb amputation and to specifically monitor JCF, which may be a major factor for early OA that has not yet been investigated in subjects with BK. Increased joint wear may be due to an increase in loads acting on the joints in the presence of compensatory mechanisms applied by the subjects to overcome the lack of ankle joint power [16]. However, the results of various studies are controversial about whether the loads (forces and moments) applied on hip and knee joints in the intact side of subjects with BK amputation were higher than those in subjects without gait disability. Therefore, the aim of this study was to check the loads acting on the intact side of subjects with BK and specifically to monitor JCF in these subjects. The results of this study confirmed that the peaks of ground reaction force components of the sound side in subjects with BK did not differ significantly from RG, Table 2. Although most components of the sound 249
Gait & Posture 58 (2017) 246–251
M.T. Karimi et al.
Fig. 2. Sample data of antero-posterior (a), vertical (b), and medio-lateral component (c) of the knee joint contact force in for BK subject (Sound side) and a reference subject.
OA. Information on the level of inflammatory markers or results of a typical questionnaire used in the OA population, e.g. the WOMAC score, are missing for the BK participants investigated here and are a limitation of the study.
the RG. Although they showed typical asymmetries between the sound and the involved side, most of the joints kinematic and kinetic parameters of the sound side showed no significant differences compared to normal subjects (Table 4 and 5 in the Appendix). Therefore, it can be summarized that the selection of prosthesis components including an ESAR prosthetic foot, the overall physical condition of participants with BK and possibly the prosthetic fitting and alignment was quite beneficial in this group of participants with BK. However, we do not have detailed information on socket fit or alignment to quantify that it was truly well performed. As a consequence, JCF in the sound side of these subjects was not higher compared to those in reference subjects without disability. In the absence of this finding, it should be emphasized that the output of this study does not support or approve any mechanisms that increases joint contact forces which could explain the increased incidence of OA in the BK population. For elucidating such mechanisms, it might be more effective for future studies to specifically select patients showing frontal plane gait deviations with their prosthesis to confirm or refute the hypothesis that increased coronal plane knee joints go hand in hand with higher joint contact forces. Lastly, it has been shown that the onset and course of OA can only partly be referred to the loading situation of the limb. Metabolic changes, e.g. increased inflammatory markers, promote the incidence and negative course of
5. Conclusion It turned out that the group of seven BK amputees showed hardly any differences in joint kinematics and kinetics in level walking when compared to an age matched group of controls. This may be explained by an overall positive clinical situation of the subjects, a beneficial prosthesis including proper socket fit, selection of ESAR prosthetic feet (in this case a Vari-Flex© ESAR foot) which had been shown to have beneficial impact on sound side loads [29,30], and good prosthesis alignment as checked via a LASAR posture and following the guidelines of Blumentritt and colleagues [28]. Consequently, the JCF in the sound side as calculated via OpenSim also did not differ significantly. Hence, in this patient cohort we could not find any biomechanical indicator for a possible early onset of osteoarthritis, except for the one subject who did in fact show osteoarthritis but in the hip of the involved side possibly partly due to the condition 32 years after amputation. Nevertheless, the approach to use a musculoskeletal model for 250
Gait & Posture 58 (2017) 246–251
M.T. Karimi et al.
knee amputees, Gait Posture 28 (2) (2008) 278–284. [15] G.R. Hurley, R. McKenney, M. Robinson, M. Zadravec, M.R. Pierrynowski, The role of the contralateral limb in below-knee amputee gait, Prosthet. Orthot. Int. 14 (1) (1990) 33–42. [16] E.D. Lemaire, F.R. Fisher, D.G. Robertson, Gait patterns of elderly men with transtibial amputations, Prosthet. Orthot. Int. 17 (1) (1993) 27–37. [17] L. Sharma, D.E. Hurwitz, E.J. Thonar, J.A. Sum, M.E. Lenz, D.D. Dunlop, et al., Knee adduction moment, serum hyaluronan level, and disease severity in medial tibiofemoral osteoarthritis, Arthritis Rheum. 41 (7) (1998) 1233–1240. [18] C.M. Powers, L. Torburn, J. Perry, E. Ayyappa, Influence of prosthetic foot design on sound limb loading in adults with unilateral below-knee amputations, Arch. Phys. Med. Rehabil. 75 (7) (1994) 825–829. [19] J.R. Engsberg, A.G. Lee, J.L. Patterson, J.A. Harder, External loading comparisons between able-bodied and below-knee-amputee children during walking, Arch. Phys. Med. Rehabil. 72 (9) (1991) 657–661. [20] R.D. Snyder, C.M. Powers, C. Fontaine, J. Perry, The effect of five prosthetic feet on the gait and loading of the sound limb in dysvascular below-knee amputees, J. Rehabil. Res. Dev. 32 (4) (1995) 309–315. [21] N.P. Fey, R.R. Neptune, 3D intersegmental knee loading in below-knee amputees across steady-state walking speeds, Clin. Biomech. 27 (4) (2012) 409–414. [22] D.C. Morgenroth, A.D. Segal, K.E. Zelik, J.M. Czerniecki, G.K. Klute, P.G. Adamczyk, et al., The effect of prosthetic foot push-off on mechanical loading associated with knee osteoarthritis in lower extremity amputees, Gait Posture 34 (4) (2011) 502–507. [23] R.H. Miller, A.Y. Esterson, J.K. Shim, Joint contact forces when minimizing the external knee adduction moment by gait modification: a computer simulation study, Knee 22 (6) (2015) 481–489. [24] H. Marouane, A. Shirazi-Adl, M. Adouni, Alterations in knee contact forces and centers in stance phase of gait: a detailed lower extremity musculoskeletal model, J. Biomech. 49 (2) (2016) 185–192. [25] G.T. Harding, M.J. Dunbar, C.L. Hubley-Kozey, W.D. Stanish, J.L. Wilson, Obesity is associated with higher absolute tibiofemoral contact and muscle forces during gait with and without knee osteoarthritis, Clin. Biomech. 31 (2016) 79–86. [26] S.L. Delp, F.C. Anderson, A.S. Arnold, P. Loan, A. Habib, C.T. John, et al., OpenSim: open-source software to create and analyze dynamic simulations of movement, IEEE Trans. Biomed. Eng. 54 (11) (2007) 1940–1950. [27] R.S. Gailey, K.E. Roach, E.B. Applegate, B. Cho, B. Cunniffe, S. Licht, M. Maguire, M.S. Nash, The amputee mobility predictor: an instrument to assess determinants of the lower-limb amputee's ability to ambulate, Arch. Phys. Med. Rehabil. 83 (5) (2002) 613–627. [28] S. Blumentritt, T. Schmalz, R. Jarasch, Die Bedeutung des statischen Prothesenaufbaus für das Stehen und Gehen des Unterschenkelamputierten, Der Orthop. 30 (3) (2001) 161–168. [29] A.D. Segal, K.E. Zelik, G.K. Klute, D.C. Morgenroth, M.E. Hahn, M.S. Orendurff, P.G. Adamczyk, S.H. Collins, A.D. Kuo, J.M. Czerniecki, The effects of a controlled energy storage and return prototype prosthetic foot on transtibial amputee ambulation, Hum. Mov. Sci. 31 (4) (2012) 918–931. [30] D.C. Morgenroth, A.D. Segal, K.E. Zelik, J.M. Czerniecki, G.K. Klute, P.G. Adamczyk, M.S. Orendurff, M.E. Hahn, S.H. Collins, A.D. Kuo, The effect of prosthetic foot push-off on mechanical loading associated with knee osteoarthritis in lower extremity amputees, Gait Posture 34 (4) (2011) 502–507. [31] S. Blumentrttt, T. Schmalz, R. Jarasch, M. Schneider, Effects of sagittal plane prosthetic alignment on standing trans-tibial amputee knee loads, Prosthet. Orthot. Int. 23 (December (3)) (1999) 231–238. [32] M.P. Kadaba, H.K. Ramakrishnan, M.E. Wootten, J. Gainey, G. Gorton, G.V. Cochran, Repeatability of kinematic, kinetic, and electromyographic data in normal adult gait, J. Orthop. Res. (1989) 7. [33] D.G. Thelen, F.C. Anderson, S.L. Delp, Generating dynamic simulations of movement using computed muscle control, J. Biomech. 36 (3) (2003) 321–328. [34] K.M. Steele, M.S. DeMers, M.H. Schwartz, S.L. Delp, Compressive tibiofemoral force during crouch gait, Gait Posture 35 (April (4)) (2012) 556–560. [35] J.P. Walter, N. Korkmaz, B.J. Fregly, M.G. Pandy, Contribution of tibiofemoral joint contact to net loads at the knee in gait, J. Orthop. Res. 33 (July (7)) (2015) 1054–1060. [36] D.A. Winter, S.E. Sienko, Biomechanics of below-knee amputee gait, J. Biomech. (1988) 21.
directly deriving JCF appears promising when monitoring factors for early osteoarthritis. Future studies should therefore concentrate on subjects presenting with compensatory movements in gait to study their influence on JCF, perhaps also in gait conditions other than level walking. Conflicts of interest All prosthetic feet used in this study were provided by Össur hf, Reykjavik, Iceland. Össur hf was not involved in writing the manuscript or interpretation of data. Acknowledgement We thank Saskia Ellis for careful English language editing. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.gaitpost.2017.08.007. References [1] K. Ziegler-Graham, E.J. MacKenzie, P.L. Ephraim, T.G. Travison, R. Brookmeyer, Estimating the prevalence of limb loss in the United States: 2005 to 2050, Arch. Phys. Med. Rehabil. 89 (3) (2008) 422–429. [2] P.W. Moxey, P. Gogalniceanu, R.J. Hinchliffe, I.M. Loftus, K.J. Jones, M.M. Thompson, P.J. Holt, Lower extremity amputations—a review of global variability in incidence, Diabet. Med. 28 (10) (2011) 1144–1153. [3] A.S.O.D.C. Soares, E.Y. Yamaguti, L. Mochizuki, A.C. Amadio, J.C. Serrão, Biomechanical parameters of gait among transtibial amputees: a review, Sao Paulo Med. J. 127 (2009) 302–309. [4] A.H. Vrieling, H.G. van Keeken, T. Schoppen, E. Otten, J.P.K. Halbertsma, A.L. Hof, et al., Gait initiation in lower limb amputees, Gait Posture 27 (3) (2008) 423–430. [5] N. Aslani, S. Noroozi, K.S. Yee, A.O.Z. Chao, C. Maggs, Simulation of gait asymmetry and energy transfer efficiency between unilateral and bilateral amputees, Sports Eng. 19 (3) (2016) 163–170. [6] E. Isakov, O. Keren, N. Benjuya, Trans-tibial amputee gait: time-distance parameters and EMG activity, Prosthet. Orthot. Int. 24 (3) (2000) 216–220. [7] D.C. Morgenroth, A.C. Gellhorn, P. Suri, Osteoarthritis in the disabled population: a mechanical perspective, PM & R 4 (5, Suppl) (2012) S20–S27. [8] A. Kusljugic, S. Kapidzic-Durakovic, Z. Kudumovic, A. Cickusic, Chronic low back pain in individuals with lower-limb amputation, Bosn. J. Basic Med. Sci. 6 (2) (2006) 67–70. [9] R. Gailey, K. Allen, J. Castles, J. Kucharik, M. Roeder, Review of secondary physical conditions associated with lower-limb amputation and long-term prosthesis use, J. Rehabil. Res. Dev. 45 (1) (2008) 15–29. [10] P.A. Struyf, C.M. van Heugten, M.W. Hitters, R.J. Smeets, The prevalence of osteoarthritis of the intact hip and knee among traumatic leg amputees, Arch. Phys. Med. Rehabil. 90 (3) (2009) 440–446. [11] M.J. Burke, V. Roman, V. Wright, Bone and joint changes in lower limb amputees, Ann. Rheum. Dis. 37 (3) (1978) 252–254. [12] J. Kulkarni, J. Adams, E. Thomas, A. Silman, Association between amputation, arthritis and osteopenia in British male war veterans with major lower limb amputations, Clin. Rehabil. 12 (4) (1998) 348–353. [13] S. Tanamas, F.S. Hanna, F.M. Cicuttini, A.E. Wluka, P. Berry, D.M. Urquhart, Does knee malalignment increase the risk of development and progression of knee osteoarthritis? A systematic review, Arthritis Rheum. 61 (4) (2009) 459–467. [14] C. Beyaert, C. Grumillier, N. Martinet, J. Paysant, J.M. Andre, Compensatory mechanism involving the knee joint of the intact limb during gait in unilateral below-
251