- Email: [email protected]
The complementary role of the plantarflexors, hamstrings and gluteus maximus in the control of stance limb stability during gait
Gait and Posture 17 (2003) 264 /272 www.elsevier.com/locate/gaitpost
The complementary role of the plantarflexors, hamstrings and gluteus maximus in the control of stance limb stability during gait Ilse Jonkers a,*, Caroline Stewart b, Arthur Spaepen a a
Laboratory of Occupational Biomechanics and Ergonomics, Kinesiology Department, Faculty for Physical Education and Physiotherapy, Katholieke Universiteit Leuven, Tervuursevest 101, 3001 Leuven, Belgium b ORLAU, Robert Jones and Agnes Hunt and Orthopaedic Hospital, Oswestry, UK Received 12 May 2002; received in revised form 20 June 2002; accepted 11 July 2002
Abstract This paper focuses on the contributions of the gluteus maximus, biceps femoris, gastrocnemius and soleus in maintaining the stability of the stance limb in the sagittal plane during the mid-stance phase of gait. In the absence of any one of these muscles, the potential compensatory changes in muscle activation are explored, the aim being to restore stability to the stance limb. The investigation was carried out by integrating musculoskeletal modelling, forward simulation and optimization techniques. We concluded that maintenance of stance limb stability requires a subtle interplay of muscle activations. Weakness in a single muscle is unlikely to be adequately compensated for by increasing or decreasing the activation of one muscle alone. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Forward simulation; Gait; Kinesiology; Muscle function
1. Introduction The study of gait disorders in individual patients has been of interest for several decades. Studying gait abnormalities using a combination of computer-aided movement analysis, force plate measurements and electromyography (EMG) recordings has increased the understanding of abnormal gait patterns in different patient groups, leading to improved treatment outcomes [1 /9]. When movement analysis is combined with force plate measurement and lower limb EMG, muscle coordination patterns, derived from the surface EMG recordings, can be linked to the resulting movement patterns (kinematics) and joint moments of the lower limb (kinetics). The analysis provides a quantitative
Abbreviations: Sw_, swing; St_, stance; GMX, gluteus maximus; GM, gluteus medius; I, iliacus; AL, adductor longus; RF, rectus femoris ; BF, biceps femoris; ST, semitendinosus; V, vastus lateralis and medialis; G, gastrocnemius; S, soleus; TA, tibialis anterior; EMG, electromyography; DOF, degrees of freedom; Fl., flexion; Ext., extension; Ft. Rot, foot rotation. * Corresponding author. Tel.: /32-16-329105; fax: /32-16-329196 E-mail address: [email protected] (I. Jonkers).
description of the patient’s gait, allowing primary and secondary abnormalities to be identified along with coping responses or compensatory strategies [1]. The differentiation between coping responses and primary or secondary gait disorders is important for careful treatment planning [1]. Although compensatory strategies will not be the target of therapeutic interventions, it is important to understand their role in allowing the locomotor system to maintain ambulation. The analysis of compensatory muscle action often involves an examination of the function of bi-articular muscles. These muscles, through their action across two joints, theoretically have the potential to compensate for muscular deficiencies at either joint. The hamstring muscles, for example, have the potential to compensate for weak hip extensors, secondary to their role as knee flexors. Similarly, gastrocnemius (G) has the potential to compensate for weakness in the remaining plantarflexors (soleus) as well as the knee flexors (hamstrings). This reasoning may hold for isolated muscle actions, but extrapolation to functional activities such as walking is not always straightforward: The anatomical function of G, described as a knee flexor and ankle plantarflexor, seems to contradict its role in controlling
0966-6362/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 9 6 6 - 6 3 6 2 ( 0 2 ) 0 0 1 0 2 - 9
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
265
Fig. 1. (A) Model response after exclusion of GMX (COSTA /208) and the model response compensating for exclusion of GMX (COSTA /98). Dotted lines replicate position before optimization. (B) Model response after exclusion of G (COSTA /388) and the model response compensating for exclusion of G (COSTA /98). Dotted lines replicate position before optimization. (C) Model response after exclusion of S (COSTA /358) and the model response compensating for exclusion of S (COSTA /178). Dotted lines replicate position before optimization. (D) Model response after exclusion of BF (COSTA /218) and the model response compensating for exclusion of BF (COSTA /108). Dotted lines replicate position before optimization.
266
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
Fig. 1 (Continued)
Induces Induces Induces Restrains Restrains Induces (III) Induces (II) Restrains (II) Restrains (II) Stance Stance Stance Stance
G S BF GMX
Induces Induces Induces Induces
Induces (III)
Induces Induces (I)
Restrains (II) Restrains (III) Induces Restrains
Restrains (I) Restrains Induces (I) Induces (I)
Swing ankle dorsiflexion Swing knee flexion/extension Swing hip flexion Stance hip extension Stance knee extension Stance ankle dorsiflexion Foot rotation
Table 1 Schematic overview of the muscle function of individual stance limb muscles in the control of the joints of the kinematic chain, with indication of the major muscle actions
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
267
knee extension during walking. Dynamic analysis of muscle action during functional movements requires a relationship to be established between individual muscle force production and the movement sequence [10]. Muscle-driven forward simulation techniques can be used to establish these causal relationships between the muscle activation patterns, the muscle forces acting on the skeletal system, the generated joint moments and the resulting gait pattern. The influence of altered muscle force production and compensatory muscle actions can be linked directly to changes in movement patterns. The work described in this paper is an extension of previous studies [11,12]. These investigations sought to explore the feasibility of obtaining a ‘normal’ movement sequence during single limb stance, based on muscledriven forward simulation with initial segmental conditions and muscle activation sequences specified. This paper extends the analysis of a model response [13] by concentrating on the compensatory muscle activation changes in the stance limb which occur after exclusion of either gluteus maximus (GMX), hamstrings or plantarflexor muscle force. These compensatory changes are required to maintain stance limb stability in the sagittal plane only. In view of the crucial role of the plantarflexor muscles in controlling tibial advancement and therefore ankle and knee position in stance, the potential compensatory mechanisms for maintaining stance ankle and knee stability will be explored. The potential of the hamstrings muscles to compensate for weak GMX will be analyzed in detail, in particular considering their supposed role in introducing stance phase knee flexion at the same time as compensating for weak hip extensors. The adaptations in muscle activation patterns following optimization are discussed in the light of compensatory muscle activation, particularly with respect to maintaining stance limb stability in pathological gait. Despite the limitations of the present musculoskeletal model, the induced muscle activation patterns can provide insight into coordination strategies adopted to compensate for loss of muscle force.
2. Method A musculoskeletal model was defined using SIMM (Musculographics, Inc. [11], Fig. 1A /D). It consisted of seven segments (stance foot, shank, femur, pelvis, as well as swing femur, shank and foot) connected by pin joints. The model had seven degrees of freedom (DOF) that allowed: rotation at the metatarsal heads of the stance foot, stance and swing limb ankle dorsi- and plantarflexion, knee flexion/extension and hip flexion/extension. Anterior and posterior pelvic tilt followed as a consequence. Inertial parameters of the segments were taken to be those of an adult male (height 1.76 m and
268
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
Fig. 2. Changes in activation level of the stance limb muscles to compensate for exclusion of (1) GMX, (2) BF, (3) G and (4) S.
weight 75 kg) and the mass of the head, arms and trunk was located at a fixed point in the pelvis. Joint restraint moments were applied to prevent the model from achieving non-physiological joint angles, thus mimicking the behavior of ligament and bone contact forces at the extremes of the joint range of motion [11]. 11 muscles were included bilaterally: GMX, iliacus (I), gluteus medius (GM), adductor longus (AL), biceps femoris (BF), semitendinosus (ST), rectus femoris (RF), vastus lateralis and medialis (V), tibialis anterior (TA), G and soleus (S). The muscle lines of action and origin and insertion points were defined according to Delp et al. [14]. A Hill-based muscle model was used to calculate the individual muscle forces [10]. Inputs to the forward simulation included the initial joint configurations (angles and angular velocity) and the muscle activation sequences as reported in Jonkers et al. [11]. This study focuses on the model response and compensatory strategies obtained after exclusion of (1) plantarflexors (G and S), (2) hamstrings (BF) and (3) hip extensor muscle (GMX) of the stance limb. The active force production of each of these muscles was eliminated successively from the forward simulation. The changes in joint kinematics were quantified by calculating the average cumulative sum of the squared
differences between the calculated joint angles and the reference data set for all joint angles (COSTA) [2]. Optimization techniques were then applied to alter the activation levels of the remaining muscle groups to restore the reference kinematics (therefore, minimizing COSTA). The changes in muscle activation revealed are believed to represent the strategies available to compensate for the exclusion of the individual muscle function.
3. Results 3.1. Model response after exclusion of the individual muscle actions The analysis of the model response after exclusion of the individual muscle actions has been described in more detail elsewhere [13]. The most prominent findings relating to the muscle actions of the stance limb GMX, BF, S and G are given below. The actions of the stance limb ST and GM were not included in the present analysis, in view of the limited effect of excluding these muscles from the analysis on the model response in the sagittal plane.
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
269
Fig. 3. Overview of the success of the compensatory changes in muscle activation after exclusion of GMX, BF, G and S.
. Exclusion of the stance limb GMX prevented the initiation of third rocker and induced prolonged stance phase knee flexion. At the stance hip, flexion was prolonged so that no hip extension was initiated. At the swing hip, however, there was excessive hip flexion combined with premature knee extension before initial contact (Fig. 1A). The total change in model response (COSTA) equalled 208. Based on the percentage distribution of muscle action over the joints, the muscle function of GMX can be classified as (I) contributing to stance hip extension (36%), (II) limiting swing hip flexion (21%) and (III) inducing stance knee extension (16%). . Excluding G prevented the initiation of the heel rocker, whereas exclusion of S resulted in a substantial reduction in the forefoot rocker amplitude. Exclusion of either muscle caused excessive ankle dorsiflexion, prolonged stance knee flexion, and excessive hip extension. Excluding S resulted in a larger knee flexion compared with exclusion of G. At the swing limb, exclusion of either muscle group resulted in diminished hip flexion combined with increased knee flexion before initial contact (Fig. 1B and C). COSTA equalled 388 after exclusion of G and 358 after exclusion of S. Based on the percentage distribution of muscle action over the joints, the
muscle function of the stance G can be classified as (I) restraining stance hip extension (24%), (II) limiting stance ankle dorsiflexion (23%) and (III) inducing swing hip flexion (20%). The action of S is classified as (I) inducing stance knee extension (31%) and (II) restraining stance ankle dorsiflexion (29%). . Exclusion of BF substantially reduced the range of motion of the ankle during the second and third rocker. In addition, no hip extension was initiated during terminal stance. At the swing limb, increased hip flexion was observed, as well as premature swing knee extension, with a reversal towards flexion (Fig. 1D). COSTA after exclusion of BF equals 218. The action of BF can therefore be classified as (I) inducing stance hip extension (36%) and (II) restraining swing hip flexion (30%). A summary of the synergistic action of the muscle groups is presented in Table 1. 3.2. Compensatory changes in muscle activation . Compensatory changes in muscle activation after exclusion of GMX (Fig. 2) included increased activation of BF and G contributed to restoring control at the stance hip and knee. The decreased activation of
270
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
TA and, to a lesser extent, AL supported these actions further. . Compensatory changes in muscle activation after exclusion of G and S included increased activity in the synergistic muscles (Fig. 2), i.e. increased activity of S after exclusion of G, and increased G activity after exclusion of S. The observed increase in BF activity was beneficial in both cases as it had a positive effect on the initiation of foot rotation. Additional changes in muscle activity were introduced in both situations in the stance limb AL and TA. Increased activity of the AL muscle restrained stance phase hip extension. If the S muscle was excluded, additional stance limb I activity was also introduced. Decreased activity of the stance limb TA helped in restraining the amplitude of stance ankle dorsiflexion, as well as initiating stance knee extension. After excluding G, the activity of the GMX muscle was decreased to restore the control at the different joints. However, increased activity of GMX muscle was introduced after exclusion of S to control stance knee extension. This effect was augmented by a simultaneous increase in the activity of the stance limb RF and the vasti. . Compensatory changes in muscle activation after exclusion of BF (Fig. 2) included increased activity in G and GMX to restore control at the level of the stance hip. Decreased activation of TA was also introduced. 3.3. Success in restoring the model response The changes introduced to compensate for the action of GMX restored the model response to 55%, with good restoration of the stance knee and hip extension (average difference 18). After exclusion of G, the model response could be restored up to 76%, thus achieving good control across the joints. After exclusion of S, compensatory changes in muscle activation restored the model response up to 51%; however, some control was still lacking at the stance ankle and knee. The action of BF was compensated for 47%, with good restoration of stance hip extension but at the expense of control at the level of the stance knee. Fig. 3 summarizes these findings.
4. Discussion Based on the results of the sensitivity analysis and optimization sequences, it can be concluded that the joint kinematics observed during the single limb support phase of gait result from a multi-dimensional, wellcoordinated process in which balanced muscle action at the different joints of the model is essential. Nevertheless, the authors recognize that the inclusion of a
two-dimensional hip model oversimplifies the observed interactions [11 /13]. Extending the control at the hip to three dimensions would require the incorporation of additional muscle groups to achieve a three-dimensional, stabilizing equilibrium around the hip. This would increase the complexity of the optimization routines used in the present study. In the future, the use of patient-specific musculoskeletal models, incorporating anatomical joints and patient derived anatomical characteristics such as bony deformation, is to be expected. This study confirms that the stance limb BF muscle has the potential to contribute to stance hip extension in the presence of a weak GMX. There was no overlap in the functions of G and BF, with the exception of the effect on foot rotation. A similar action of GMX and the plantarflexor muscle group (either S or G) can be seen at the level of the distal stance limb joints. The analysis of muscle activation changes after exclusion of an individual muscle revealed that the muscle balance at the joint is mostly restored by increased activation of synergistic muscles and/or by decreased activation of antagonistic muscles. This principle can be detected in the muscle activation changes induced to compensate for the absence of muscle force in the muscles studied: . After exclusion of S, the increased activation of G and decreased activation of TA mimicked the action of the S muscle at the different joints. . A compensatory increase in activity of BF and decrease in activity of AL was observed after exclusion of GMX. Additional muscle activation changes were introduced simultaneously due to the influence of the excluded muscle action on the control of remote joints and the muscle’s action [13]. These control the secondary actions of the excluded muscle at different joints of the model and thereby contribute positively to the model response. When comparing the compensatory strategies of synergistic muscles, the secondary muscle activations introduced clearly demonstrated the specific action of each of the synergists in the model. When comparing the changes in muscle activation following exclusion of S and G, most changes were comparable, with the exception of GMX where there was an opposite response. This relates to the difference in the primary function of the two muscles. Increased activation of GMX contributed to the primary function of S, i.e. the control of stance ankle dorsiflexion and stance knee extension, whereas decreased activation of GMX contributed to the control of stance hip extension. Since muscle action is a multi-joint process [10,15], compensation for muscle weakness through synergistic muscle action is not straightforward. Contributing
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
factors include the relative distribution of the different muscles and possible secondary synergistic muscle actions. Additional changes in muscle activation are required to reproduce a similar movement sequence to compensate for muscle weakness by synergistic muscle groups. Furthermore, compensation for a deficient muscle action (e.g. after exclusion of S) may be possible at the expense of control loss at one joint of the kinematic chain. Within the present work, the optimization process attempted to restore the reference kinematics at all joints of the kinematic chain. Patients may compensate by deviating from normal kinematics. From the present analysis, it can be concluded the hamstrings, especially BF, could compensate for weak glutei. However, increased activation at the level of G was required simultaneously to mimic the function of the gluteus muscle in stabilizing the stance knee in extension. This may be difficult in pathological gait in the presence of inadequate force generating capacity of the plantarflexors. In these situations, e.g., after iatrogenic over lengthening of the plantarflexors, compensation for a weak GMX by using the hamstrings will always result in a failure to achieve adequate stance knee extension. Furthermore does the present analysis indicate that isolated weakness of a plantarflexor muscle can be compensated by increased activity of the synergistic muscle, either G or S. However, an additional increase in the activity of the hamstrings muscles and to some extent the knee extensors is required if the stance limb S is excluded, to restore the function of the plantarflexors at the more proximal joints. In clinical cases, where both S and G have been weakened through excessive Achilles tendon lengthening, the importance of sufficient hamstring and knee extensor force in maintaining stance limb stability is evident. If this procedure is carried out in a multi-level approach, combined with excessive lengthening of the hamstrings muscles, stability is threatened even further. Although isolated weakness of BF is rare in clinical practice (e.g. after a single level injection with botulinum toxin), this analysis indicates that a simultaneous increase of GMX and G is capable of recovering the stability of the stance limb in the model. This opens an important therapeutic window since it forces the use of GMX in a functional activity such as walking, thereby altering the coordination pattern of the child. If either of these two muscles fails to produce sufficient force, stance limb stability can no longer be guaranteed. Careful inspection of the force generating capacity of the plantarflexors, combined with intensive training of the mono-articular hip extensor muscles are crucial factors contributing to the success of inducing modified hip extension control in the stance phase of gait. The authors would like to emphasize that the studied compensatory actions, all relate to the stability of the
271
stance limb in the sagittal plane and should not be extrapolated towards the other planes of movement without further research. Based on the clinical experience, weakness of glutei may be encompassed by coronal plane compensations that are unlikely to be attributed to the action of the hamstrings group. Although muscle compensations may be attributed to an increased activation of synergists or decreased action of antagonists, the success of the compensation depends on the success of simultaneously re-establishing a balanced action at the different joints of the model.
5. Conclusion The sensitivity analysis demonstrates that most muscles contribute to the control of remote joints within the kinematic chain and show a specific relative distribution of their action. Consequently, the identified compensation strategies encompass a series of muscle activation changes to restore control at the different joints of the model. The clinical significance of this study is that gait requires a subtle interplay of muscle activations and compensation for muscle weakness is unlikely to occur by increasing or decreasing the activation of a single muscle.
References [1] Gage JR. Gait analysis in cerebral palsy. Clin Dev Med 1991;121:1 /206. [2] Perry J. Gait Analysis: Normal and Pathological Function. Thorofare: Slack, 1992. [3] Loizeau J, Allard P, Duhaome M, Landjerot B. Bilateral gait patterns in subjects fitted with a total hip prosthesis. Arch Phys Med Rehabil 1995;76:552 /7. [4] Boscaronio LF, Ounpuu S, Davis RB, Gage JR, DeLuca PA. Effect of selective dorsal rhizotomy on gait in children with cerebral palsy. J Pediatr Orthop 1993;13:174 /9. [5] Rose SA, DeLuca PA, Davis RB, Ounpuu S, Gage JR. Kinematic and kinetic evaluation of the ankle after lengthening of the gastrocnemius fascia in children with cerebral palsy. J Pediatr Orthop 1993;13:727 /32. [6] Ounpuu S, Muik E, Davis RB, Gage JR, DeLuca PA. Rectus femoris surgery in children with cerebral palsy. Part I. The effect of rectus femoris transfer location on knee motion. J Pediatr Orthop 1993;13:325 /30. [7] Sutherland DH, Davids JR. Common gait abnormalities of the knee in cerebral palsy. Clin Orthop 1993;288:139 /47. [8] Sutherland DH, Olshen R, Cooper L, Wyatt M, Leach J, Mubarak S, Schultz P. The pathomechanics of gait in Duchenne muscular dystrophy. Dev Med Child Neurol 1981;23:3 /22. [9] Sutherland DH, Santi M, Abel MF. Treatment of stiff-knee gait in cerebral palsy: a comparison by gait analysis of distal rectus femoris transfer versus proximal rectus release. J Pediatr Orthop 1990;10:433 /41. [10] Zajac FE, Gordon ME. Determining muscle’s force and action in multi-articular movement. Exerc Sport Sci Rev 1989;17:187 /230.
272
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
[11] Jonkers I, Spaepen A, Papaioannou G, Stewart C. An EMGbased, muscle-driven forward simulation of single support phase of gait. J Biomech 2002;35:609 /19. [12] Jonkers I, Stewart C, Spaepen A. The study of muscle action during single support and swing phase of gait: Clinical relevance of forward simulation techniques. Gait Posture 2001;14:124. [13] Jonkers I, Stewart S, Spaepen A. The study of muscle action during single support and swing phase of gait: Clinical relevance of forward simulation techniques, Gait Posture 2002, accepted.
[14] Delp SL, Loan PL, Hoy MG, Zajac FE, Topp EL, Rosen JM. An interactive graphics based model of the lower extremity to study orthopaedic surgical procedures. IEEE Trans Biomed Eng 1990;37:757 /67. [15] Zajac FE. Muscle co-ordination of movement: a perspective. J Biomech 1993;26(Suppl. 1):109 /24.
The complementary role of the plantarflexors, hamstrings and gluteus maximus in the control of stance limb stability during gait Ilse Jonkers a,*, Caroline Stewart b, Arthur Spaepen a a
Laboratory of Occupational Biomechanics and Ergonomics, Kinesiology Department, Faculty for Physical Education and Physiotherapy, Katholieke Universiteit Leuven, Tervuursevest 101, 3001 Leuven, Belgium b ORLAU, Robert Jones and Agnes Hunt and Orthopaedic Hospital, Oswestry, UK Received 12 May 2002; received in revised form 20 June 2002; accepted 11 July 2002
Abstract This paper focuses on the contributions of the gluteus maximus, biceps femoris, gastrocnemius and soleus in maintaining the stability of the stance limb in the sagittal plane during the mid-stance phase of gait. In the absence of any one of these muscles, the potential compensatory changes in muscle activation are explored, the aim being to restore stability to the stance limb. The investigation was carried out by integrating musculoskeletal modelling, forward simulation and optimization techniques. We concluded that maintenance of stance limb stability requires a subtle interplay of muscle activations. Weakness in a single muscle is unlikely to be adequately compensated for by increasing or decreasing the activation of one muscle alone. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Forward simulation; Gait; Kinesiology; Muscle function
1. Introduction The study of gait disorders in individual patients has been of interest for several decades. Studying gait abnormalities using a combination of computer-aided movement analysis, force plate measurements and electromyography (EMG) recordings has increased the understanding of abnormal gait patterns in different patient groups, leading to improved treatment outcomes [1 /9]. When movement analysis is combined with force plate measurement and lower limb EMG, muscle coordination patterns, derived from the surface EMG recordings, can be linked to the resulting movement patterns (kinematics) and joint moments of the lower limb (kinetics). The analysis provides a quantitative
Abbreviations: Sw_, swing; St_, stance; GMX, gluteus maximus; GM, gluteus medius; I, iliacus; AL, adductor longus; RF, rectus femoris ; BF, biceps femoris; ST, semitendinosus; V, vastus lateralis and medialis; G, gastrocnemius; S, soleus; TA, tibialis anterior; EMG, electromyography; DOF, degrees of freedom; Fl., flexion; Ext., extension; Ft. Rot, foot rotation. * Corresponding author. Tel.: /32-16-329105; fax: /32-16-329196 E-mail address: [email protected] (I. Jonkers).
description of the patient’s gait, allowing primary and secondary abnormalities to be identified along with coping responses or compensatory strategies [1]. The differentiation between coping responses and primary or secondary gait disorders is important for careful treatment planning [1]. Although compensatory strategies will not be the target of therapeutic interventions, it is important to understand their role in allowing the locomotor system to maintain ambulation. The analysis of compensatory muscle action often involves an examination of the function of bi-articular muscles. These muscles, through their action across two joints, theoretically have the potential to compensate for muscular deficiencies at either joint. The hamstring muscles, for example, have the potential to compensate for weak hip extensors, secondary to their role as knee flexors. Similarly, gastrocnemius (G) has the potential to compensate for weakness in the remaining plantarflexors (soleus) as well as the knee flexors (hamstrings). This reasoning may hold for isolated muscle actions, but extrapolation to functional activities such as walking is not always straightforward: The anatomical function of G, described as a knee flexor and ankle plantarflexor, seems to contradict its role in controlling
0966-6362/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 9 6 6 - 6 3 6 2 ( 0 2 ) 0 0 1 0 2 - 9
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
265
Fig. 1. (A) Model response after exclusion of GMX (COSTA /208) and the model response compensating for exclusion of GMX (COSTA /98). Dotted lines replicate position before optimization. (B) Model response after exclusion of G (COSTA /388) and the model response compensating for exclusion of G (COSTA /98). Dotted lines replicate position before optimization. (C) Model response after exclusion of S (COSTA /358) and the model response compensating for exclusion of S (COSTA /178). Dotted lines replicate position before optimization. (D) Model response after exclusion of BF (COSTA /218) and the model response compensating for exclusion of BF (COSTA /108). Dotted lines replicate position before optimization.
266
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
Fig. 1 (Continued)
Induces Induces Induces Restrains Restrains Induces (III) Induces (II) Restrains (II) Restrains (II) Stance Stance Stance Stance
G S BF GMX
Induces Induces Induces Induces
Induces (III)
Induces Induces (I)
Restrains (II) Restrains (III) Induces Restrains
Restrains (I) Restrains Induces (I) Induces (I)
Swing ankle dorsiflexion Swing knee flexion/extension Swing hip flexion Stance hip extension Stance knee extension Stance ankle dorsiflexion Foot rotation
Table 1 Schematic overview of the muscle function of individual stance limb muscles in the control of the joints of the kinematic chain, with indication of the major muscle actions
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
267
knee extension during walking. Dynamic analysis of muscle action during functional movements requires a relationship to be established between individual muscle force production and the movement sequence [10]. Muscle-driven forward simulation techniques can be used to establish these causal relationships between the muscle activation patterns, the muscle forces acting on the skeletal system, the generated joint moments and the resulting gait pattern. The influence of altered muscle force production and compensatory muscle actions can be linked directly to changes in movement patterns. The work described in this paper is an extension of previous studies [11,12]. These investigations sought to explore the feasibility of obtaining a ‘normal’ movement sequence during single limb stance, based on muscledriven forward simulation with initial segmental conditions and muscle activation sequences specified. This paper extends the analysis of a model response [13] by concentrating on the compensatory muscle activation changes in the stance limb which occur after exclusion of either gluteus maximus (GMX), hamstrings or plantarflexor muscle force. These compensatory changes are required to maintain stance limb stability in the sagittal plane only. In view of the crucial role of the plantarflexor muscles in controlling tibial advancement and therefore ankle and knee position in stance, the potential compensatory mechanisms for maintaining stance ankle and knee stability will be explored. The potential of the hamstrings muscles to compensate for weak GMX will be analyzed in detail, in particular considering their supposed role in introducing stance phase knee flexion at the same time as compensating for weak hip extensors. The adaptations in muscle activation patterns following optimization are discussed in the light of compensatory muscle activation, particularly with respect to maintaining stance limb stability in pathological gait. Despite the limitations of the present musculoskeletal model, the induced muscle activation patterns can provide insight into coordination strategies adopted to compensate for loss of muscle force.
2. Method A musculoskeletal model was defined using SIMM (Musculographics, Inc. [11], Fig. 1A /D). It consisted of seven segments (stance foot, shank, femur, pelvis, as well as swing femur, shank and foot) connected by pin joints. The model had seven degrees of freedom (DOF) that allowed: rotation at the metatarsal heads of the stance foot, stance and swing limb ankle dorsi- and plantarflexion, knee flexion/extension and hip flexion/extension. Anterior and posterior pelvic tilt followed as a consequence. Inertial parameters of the segments were taken to be those of an adult male (height 1.76 m and
268
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
Fig. 2. Changes in activation level of the stance limb muscles to compensate for exclusion of (1) GMX, (2) BF, (3) G and (4) S.
weight 75 kg) and the mass of the head, arms and trunk was located at a fixed point in the pelvis. Joint restraint moments were applied to prevent the model from achieving non-physiological joint angles, thus mimicking the behavior of ligament and bone contact forces at the extremes of the joint range of motion [11]. 11 muscles were included bilaterally: GMX, iliacus (I), gluteus medius (GM), adductor longus (AL), biceps femoris (BF), semitendinosus (ST), rectus femoris (RF), vastus lateralis and medialis (V), tibialis anterior (TA), G and soleus (S). The muscle lines of action and origin and insertion points were defined according to Delp et al. [14]. A Hill-based muscle model was used to calculate the individual muscle forces [10]. Inputs to the forward simulation included the initial joint configurations (angles and angular velocity) and the muscle activation sequences as reported in Jonkers et al. [11]. This study focuses on the model response and compensatory strategies obtained after exclusion of (1) plantarflexors (G and S), (2) hamstrings (BF) and (3) hip extensor muscle (GMX) of the stance limb. The active force production of each of these muscles was eliminated successively from the forward simulation. The changes in joint kinematics were quantified by calculating the average cumulative sum of the squared
differences between the calculated joint angles and the reference data set for all joint angles (COSTA) [2]. Optimization techniques were then applied to alter the activation levels of the remaining muscle groups to restore the reference kinematics (therefore, minimizing COSTA). The changes in muscle activation revealed are believed to represent the strategies available to compensate for the exclusion of the individual muscle function.
3. Results 3.1. Model response after exclusion of the individual muscle actions The analysis of the model response after exclusion of the individual muscle actions has been described in more detail elsewhere [13]. The most prominent findings relating to the muscle actions of the stance limb GMX, BF, S and G are given below. The actions of the stance limb ST and GM were not included in the present analysis, in view of the limited effect of excluding these muscles from the analysis on the model response in the sagittal plane.
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
269
Fig. 3. Overview of the success of the compensatory changes in muscle activation after exclusion of GMX, BF, G and S.
. Exclusion of the stance limb GMX prevented the initiation of third rocker and induced prolonged stance phase knee flexion. At the stance hip, flexion was prolonged so that no hip extension was initiated. At the swing hip, however, there was excessive hip flexion combined with premature knee extension before initial contact (Fig. 1A). The total change in model response (COSTA) equalled 208. Based on the percentage distribution of muscle action over the joints, the muscle function of GMX can be classified as (I) contributing to stance hip extension (36%), (II) limiting swing hip flexion (21%) and (III) inducing stance knee extension (16%). . Excluding G prevented the initiation of the heel rocker, whereas exclusion of S resulted in a substantial reduction in the forefoot rocker amplitude. Exclusion of either muscle caused excessive ankle dorsiflexion, prolonged stance knee flexion, and excessive hip extension. Excluding S resulted in a larger knee flexion compared with exclusion of G. At the swing limb, exclusion of either muscle group resulted in diminished hip flexion combined with increased knee flexion before initial contact (Fig. 1B and C). COSTA equalled 388 after exclusion of G and 358 after exclusion of S. Based on the percentage distribution of muscle action over the joints, the
muscle function of the stance G can be classified as (I) restraining stance hip extension (24%), (II) limiting stance ankle dorsiflexion (23%) and (III) inducing swing hip flexion (20%). The action of S is classified as (I) inducing stance knee extension (31%) and (II) restraining stance ankle dorsiflexion (29%). . Exclusion of BF substantially reduced the range of motion of the ankle during the second and third rocker. In addition, no hip extension was initiated during terminal stance. At the swing limb, increased hip flexion was observed, as well as premature swing knee extension, with a reversal towards flexion (Fig. 1D). COSTA after exclusion of BF equals 218. The action of BF can therefore be classified as (I) inducing stance hip extension (36%) and (II) restraining swing hip flexion (30%). A summary of the synergistic action of the muscle groups is presented in Table 1. 3.2. Compensatory changes in muscle activation . Compensatory changes in muscle activation after exclusion of GMX (Fig. 2) included increased activation of BF and G contributed to restoring control at the stance hip and knee. The decreased activation of
270
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
TA and, to a lesser extent, AL supported these actions further. . Compensatory changes in muscle activation after exclusion of G and S included increased activity in the synergistic muscles (Fig. 2), i.e. increased activity of S after exclusion of G, and increased G activity after exclusion of S. The observed increase in BF activity was beneficial in both cases as it had a positive effect on the initiation of foot rotation. Additional changes in muscle activity were introduced in both situations in the stance limb AL and TA. Increased activity of the AL muscle restrained stance phase hip extension. If the S muscle was excluded, additional stance limb I activity was also introduced. Decreased activity of the stance limb TA helped in restraining the amplitude of stance ankle dorsiflexion, as well as initiating stance knee extension. After excluding G, the activity of the GMX muscle was decreased to restore the control at the different joints. However, increased activity of GMX muscle was introduced after exclusion of S to control stance knee extension. This effect was augmented by a simultaneous increase in the activity of the stance limb RF and the vasti. . Compensatory changes in muscle activation after exclusion of BF (Fig. 2) included increased activity in G and GMX to restore control at the level of the stance hip. Decreased activation of TA was also introduced. 3.3. Success in restoring the model response The changes introduced to compensate for the action of GMX restored the model response to 55%, with good restoration of the stance knee and hip extension (average difference 18). After exclusion of G, the model response could be restored up to 76%, thus achieving good control across the joints. After exclusion of S, compensatory changes in muscle activation restored the model response up to 51%; however, some control was still lacking at the stance ankle and knee. The action of BF was compensated for 47%, with good restoration of stance hip extension but at the expense of control at the level of the stance knee. Fig. 3 summarizes these findings.
4. Discussion Based on the results of the sensitivity analysis and optimization sequences, it can be concluded that the joint kinematics observed during the single limb support phase of gait result from a multi-dimensional, wellcoordinated process in which balanced muscle action at the different joints of the model is essential. Nevertheless, the authors recognize that the inclusion of a
two-dimensional hip model oversimplifies the observed interactions [11 /13]. Extending the control at the hip to three dimensions would require the incorporation of additional muscle groups to achieve a three-dimensional, stabilizing equilibrium around the hip. This would increase the complexity of the optimization routines used in the present study. In the future, the use of patient-specific musculoskeletal models, incorporating anatomical joints and patient derived anatomical characteristics such as bony deformation, is to be expected. This study confirms that the stance limb BF muscle has the potential to contribute to stance hip extension in the presence of a weak GMX. There was no overlap in the functions of G and BF, with the exception of the effect on foot rotation. A similar action of GMX and the plantarflexor muscle group (either S or G) can be seen at the level of the distal stance limb joints. The analysis of muscle activation changes after exclusion of an individual muscle revealed that the muscle balance at the joint is mostly restored by increased activation of synergistic muscles and/or by decreased activation of antagonistic muscles. This principle can be detected in the muscle activation changes induced to compensate for the absence of muscle force in the muscles studied: . After exclusion of S, the increased activation of G and decreased activation of TA mimicked the action of the S muscle at the different joints. . A compensatory increase in activity of BF and decrease in activity of AL was observed after exclusion of GMX. Additional muscle activation changes were introduced simultaneously due to the influence of the excluded muscle action on the control of remote joints and the muscle’s action [13]. These control the secondary actions of the excluded muscle at different joints of the model and thereby contribute positively to the model response. When comparing the compensatory strategies of synergistic muscles, the secondary muscle activations introduced clearly demonstrated the specific action of each of the synergists in the model. When comparing the changes in muscle activation following exclusion of S and G, most changes were comparable, with the exception of GMX where there was an opposite response. This relates to the difference in the primary function of the two muscles. Increased activation of GMX contributed to the primary function of S, i.e. the control of stance ankle dorsiflexion and stance knee extension, whereas decreased activation of GMX contributed to the control of stance hip extension. Since muscle action is a multi-joint process [10,15], compensation for muscle weakness through synergistic muscle action is not straightforward. Contributing
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
factors include the relative distribution of the different muscles and possible secondary synergistic muscle actions. Additional changes in muscle activation are required to reproduce a similar movement sequence to compensate for muscle weakness by synergistic muscle groups. Furthermore, compensation for a deficient muscle action (e.g. after exclusion of S) may be possible at the expense of control loss at one joint of the kinematic chain. Within the present work, the optimization process attempted to restore the reference kinematics at all joints of the kinematic chain. Patients may compensate by deviating from normal kinematics. From the present analysis, it can be concluded the hamstrings, especially BF, could compensate for weak glutei. However, increased activation at the level of G was required simultaneously to mimic the function of the gluteus muscle in stabilizing the stance knee in extension. This may be difficult in pathological gait in the presence of inadequate force generating capacity of the plantarflexors. In these situations, e.g., after iatrogenic over lengthening of the plantarflexors, compensation for a weak GMX by using the hamstrings will always result in a failure to achieve adequate stance knee extension. Furthermore does the present analysis indicate that isolated weakness of a plantarflexor muscle can be compensated by increased activity of the synergistic muscle, either G or S. However, an additional increase in the activity of the hamstrings muscles and to some extent the knee extensors is required if the stance limb S is excluded, to restore the function of the plantarflexors at the more proximal joints. In clinical cases, where both S and G have been weakened through excessive Achilles tendon lengthening, the importance of sufficient hamstring and knee extensor force in maintaining stance limb stability is evident. If this procedure is carried out in a multi-level approach, combined with excessive lengthening of the hamstrings muscles, stability is threatened even further. Although isolated weakness of BF is rare in clinical practice (e.g. after a single level injection with botulinum toxin), this analysis indicates that a simultaneous increase of GMX and G is capable of recovering the stability of the stance limb in the model. This opens an important therapeutic window since it forces the use of GMX in a functional activity such as walking, thereby altering the coordination pattern of the child. If either of these two muscles fails to produce sufficient force, stance limb stability can no longer be guaranteed. Careful inspection of the force generating capacity of the plantarflexors, combined with intensive training of the mono-articular hip extensor muscles are crucial factors contributing to the success of inducing modified hip extension control in the stance phase of gait. The authors would like to emphasize that the studied compensatory actions, all relate to the stability of the
271
stance limb in the sagittal plane and should not be extrapolated towards the other planes of movement without further research. Based on the clinical experience, weakness of glutei may be encompassed by coronal plane compensations that are unlikely to be attributed to the action of the hamstrings group. Although muscle compensations may be attributed to an increased activation of synergists or decreased action of antagonists, the success of the compensation depends on the success of simultaneously re-establishing a balanced action at the different joints of the model.
5. Conclusion The sensitivity analysis demonstrates that most muscles contribute to the control of remote joints within the kinematic chain and show a specific relative distribution of their action. Consequently, the identified compensation strategies encompass a series of muscle activation changes to restore control at the different joints of the model. The clinical significance of this study is that gait requires a subtle interplay of muscle activations and compensation for muscle weakness is unlikely to occur by increasing or decreasing the activation of a single muscle.
References [1] Gage JR. Gait analysis in cerebral palsy. Clin Dev Med 1991;121:1 /206. [2] Perry J. Gait Analysis: Normal and Pathological Function. Thorofare: Slack, 1992. [3] Loizeau J, Allard P, Duhaome M, Landjerot B. Bilateral gait patterns in subjects fitted with a total hip prosthesis. Arch Phys Med Rehabil 1995;76:552 /7. [4] Boscaronio LF, Ounpuu S, Davis RB, Gage JR, DeLuca PA. Effect of selective dorsal rhizotomy on gait in children with cerebral palsy. J Pediatr Orthop 1993;13:174 /9. [5] Rose SA, DeLuca PA, Davis RB, Ounpuu S, Gage JR. Kinematic and kinetic evaluation of the ankle after lengthening of the gastrocnemius fascia in children with cerebral palsy. J Pediatr Orthop 1993;13:727 /32. [6] Ounpuu S, Muik E, Davis RB, Gage JR, DeLuca PA. Rectus femoris surgery in children with cerebral palsy. Part I. The effect of rectus femoris transfer location on knee motion. J Pediatr Orthop 1993;13:325 /30. [7] Sutherland DH, Davids JR. Common gait abnormalities of the knee in cerebral palsy. Clin Orthop 1993;288:139 /47. [8] Sutherland DH, Olshen R, Cooper L, Wyatt M, Leach J, Mubarak S, Schultz P. The pathomechanics of gait in Duchenne muscular dystrophy. Dev Med Child Neurol 1981;23:3 /22. [9] Sutherland DH, Santi M, Abel MF. Treatment of stiff-knee gait in cerebral palsy: a comparison by gait analysis of distal rectus femoris transfer versus proximal rectus release. J Pediatr Orthop 1990;10:433 /41. [10] Zajac FE, Gordon ME. Determining muscle’s force and action in multi-articular movement. Exerc Sport Sci Rev 1989;17:187 /230.
272
I. Jonkers et al. / Gait and Posture 17 (2003) 264 /272
[11] Jonkers I, Spaepen A, Papaioannou G, Stewart C. An EMGbased, muscle-driven forward simulation of single support phase of gait. J Biomech 2002;35:609 /19. [12] Jonkers I, Stewart C, Spaepen A. The study of muscle action during single support and swing phase of gait: Clinical relevance of forward simulation techniques. Gait Posture 2001;14:124. [13] Jonkers I, Stewart S, Spaepen A. The study of muscle action during single support and swing phase of gait: Clinical relevance of forward simulation techniques, Gait Posture 2002, accepted.
[14] Delp SL, Loan PL, Hoy MG, Zajac FE, Topp EL, Rosen JM. An interactive graphics based model of the lower extremity to study orthopaedic surgical procedures. IEEE Trans Biomed Eng 1990;37:757 /67. [15] Zajac FE. Muscle co-ordination of movement: a perspective. J Biomech 1993;26(Suppl. 1):109 /24.