Вы находитесь на странице: 1из 144

Katholieke Universiteit Leuven Group Biomedical Sciences Faculty of Kinesiology and Rehabilitation Sciences Department of Biomedical Kinesiology

Sciences Department of Biomedical Kinesiology GAIT IN HEMIPLEGIC CEREBRAL PALSY EFFECTS OF ANKLE FOOT

GAIT IN HEMIPLEGIC CEREBRAL PALSY

EFFECTS OF ANKLE FOOT ORTHOSES

Jacqueline Romkes

Jury:

Promotores:

Prof. Dr. J. Duysens Prof. Dr. med. R. Brunner

Chair:

Prof. Dr. J. Lefevre

Secretary:

Prof. Dr. K. Desloovere

Jury members:

Prof. Dr. G. Molenaers Dr. J. Harlaar

Leuven, 30.09.2008

Doctoral thesis in Biomedical Kinesiology

The production of this thesis was supported by the Foundation for Biomechanical Research “ProMotio”, Basle, Switzerland.

Biomechanical Research “ProMotio”, Basle, Switzerland. ISBN: 978-90-9023213-3 Cover design: Jacqueline

ISBN:

978-90-9023213-3

Cover design:

Jacqueline Romkes

Printed by:

PrintPartners Ipskamp B.V., Enschede

© J. Romkes, Basle, 2008 All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage or retrieval system, without written permission from the author.

Contents

Chapter 1.

General introduction

5

Chapter 2.

An electromyographic analysis of obligatory (hemiplegic cerebral palsy) and voluntary (normal) unilateral toe walking

19

Chapter 3.

Comparison of a dynamic and a hinged ankle-foot orthosis by gait analysis in patients with hemiplegic cerebral palsy

41

Chapter 4.

Changes in muscle activity in children with hemiplegic cerebral palsy while walking with and without ankle- foot orthoses

57

Chapter 5.

Influence of an ankle foot orthosis on muscle activity during gait in children with hemiplegic cerebral palsy

73

Chapter 6.

Changes in gait and EMG when walking with the Masai Barefoot Technique

91

Chapter 7.

General discussion

105

Summary

121

Samenvatting (Dutch summary)

127

Acknowledgements

133

About the author and list of publications

137

Chapter

1

General introduction

Chapter 1

Gait can be defined as a person’s manner of walking characterized by periods of loading (stance phase) and unloading (swing phase) of the lower limbs. The gait cycle contains a sequence of events that usually begins with floor contact (0 %) with one foot and ends with the next contact of the same foot (100 %), which will be the initial contact of the next cycle. In normal symmetrical gait, a stance phase accounts for approximately 60 %, and a swing phase for approximately 40 % percent. 1,2 As Duysens et al. 3 described: ‘Somehow the central nervous system is able to coordinate which joint has to be moved, how far and at what time. Such movements can only be made properly if a set of biomechanical requirements are met using a pattern of electrical signals sent along the nerves to activate the appropriate set of muscles.’ Therefore, the muscle activity of the lower extremities used in walking need to be well coordinated to provide support, dynamic balance, propulsion, and foot clearance. 4 These gait tasks are temporally coupled to particular phases of the gait cycle and, as a result, the timing of lower extremity muscle activity during healthy human walking follows characteristic patterns. 1 One of the most distinctive features of healthy human walking is the heel strike at the onset of stance. This feature is crucial for the development of an adequate gait cycle. In pathology it is often seen that heel strike is affected. One of the most common gait abnormalities seen in patients with hemiplegic cerebral palsy (CP) is an initial foot contact with the toes on the affected limb. The aim of this thesis is to investigate this unilateral toe walking gait pattern in children with spastic hemiplegic CP. Several research questions concerning this topic are asked. The first question is: How does a toe walking gait pattern in patients with hemiplegic CP differ from voluntary unilateral toe walking by neurological healthy subjects? It is hypothesized that common deviations for both patients and the healthy control subjects can be observed due to the biomechanical constraints of the toe walking gait pattern. A common treatment measure to correct the toe walking gait pattern is to provide the patient with an ankle foot orthosis (AFO) on the affected lower leg. Two research questions concerning AFOs are: How does an AFO biomechanically influences the gait pattern? Can an AFO modulate the muscular activation pattern with a change of gait pattern?

Cerebral palsy (CP)

The patients participating in the scientific studies described in this thesis suffer

6

General introduction

from CP which is a condition caused by lesions to the central nervous system. Several definitions have been proposed in the literature over the years. 5-10 The crucial components of all definitions are that the impairment is of the brain, produces disorders of movement or posture, occurs during early development (in utero, during birth, or early after birth in timing), is non-progressive and, although non-progressive, the symptoms and signs may vary as the child matures. 11 The peripheral manifestations of CP depend on the magnitude and location of the insult that causes the damage to the central nervous system. The severity ranges from subtle motor impairment to involvement of the whole body. In ambulatory patients with CP we can discriminate between primary and secondary gait deviations. Primary deviations are directly related to the lesion in the central nervous system and can be due to the altered muscle tone, balance, strength, and selective motor control of different muscle groups. Secondary deviations are defined as those occurring as a compensation for another primary gait deviation. 12,13

Disease classification

Patients with CP can be classified according to topographical distribution and neurological dysfunction. Apart from that, classifications on the basis of abilities and limitations in gross motor function and gait have been developed. This thesis focuses on patients with the spastic hemiplegic type of CP. Hemiplegia refers to a topographical distribution of one sided body involvement with relative sparing of the contralateral limbs. Usually the arm is more severely involved than the leg. Spasticity refers to the neurological dysfunction. Spasticity is a key characteristic of CP and the most common. It is a form of hypertonia due to pyramidal tract pathology. A generally accepted definition is provided by Lance 14 : ‘Spasticity is a motor disorder characterized by a velocity dependent increase in tonic stretch reflexes (muscle tone) with exaggerated tendon jerks, resulting from hyper excitability of the stretch reflex as one component of the upper motor neuron syndrome.’ The rapid rise of resistance is often referred to as ‘spastic catch’ that is felt on clinical examination. Studying the hemiplegic type of CP is especially interesting since one body side functions normal or near normal and hence allows comparison with an internal control.

Prevalence of cerebral palsy

Overall prevalence rates of CP in developed countries range from 1.9 to 2.5 per

7

Chapter 1

1000 live births. 15-23 The majority of studies exclude children who sustained brain damage resulting in CP after the neonatal period (28 days after birth). The prevalence of CP, however, is not evenly distributed over birth weight groups and prevalence increases with decreasing birth weight. Whereas reported prevalence for CP is 1.2 (range 1.1-1.3) per 1000 infants at a birth weight > 2500 grams, this increases to 9.7 (range 6.2-11.4) at a birth weight of 1500-2499 grams, to 62.9 (range 39.0-81.0) at a birth weight < 1500 grams, and to 75.4 (range 48.5-99.5) at a birth weight < 1000 grams. 15-17,19-23 More boys than girls are reported with an overall male excess ratio of 1.33:1. 23 It is common that CP is accompanied by other disabilities including intellectual impairment (47 % with IQ < 70), hearing and visual impairment (14 %), sensory impairment (21 %), or epilepsy/seizures (33 %). 17,20,22,23 The likelihood of epilepsy, extrapyramidal abnormalities, and severe cognitive impairment is generally higher among those with quadriplegia than among those with diplegia or hemiplegia. In all, spastic diplegic and hemiplegic CP subtypes account for the vast majority, 43 % and 30.1 %. 16-18,20,21,23 The type of CP differs among different birth weight groups, with the lightest (< 1000 grams) and heaviest (> 2500 grams) groups showing fewer cases of diplegia and more cases of hemiplegia. 18,20,21 Cases of dyskinesia (9.1 %) and simple ataxia (5.1 %) are reported to be less prevalent. Overall, between 18 % and 33.4 % of the children with CP are unable to walk. 17,20,21,23 Children with diplegia or hemiplegia, however, nearly always walk. Wren et al. 24 investigated the prevalence of specific gait abnormalities in ambulatory children with CP. In this study 492 patients (of which 115 with hemiplegia) who were referred for an instrumented gait analysis were included and fourteen specific gait abnormalities (e.g. equinus, crouch, stiff knee, rotational malalignment, and others) were examined. The prevalence of each gait abnormality and the influence of CP subtype, age, and previous surgery on the likelihood of having each of these gait abnormalities were correlated and documented. The most common gait abnormality for hemiplegic CP patients without previous surgery with a prevalence of 70 % was equinus gait (i.e. increased ankle plantarflexion at initial contact). For the hemiplegic patients with previous surgery, crouch (74 %), stiff knee (71 %), and excessive hip flexion (58 %) were more prevalent than equinus (48 %).

8

General introduction

Gait classification

Classifying patients into categories or groups can be useful for diagnostic purposes, prognostic reasons, to help in establishing treatment strategies, to streamline communication, to facilitate clinical decision making, and for research purposes. Although gait and posture are variable in children who have CP, there are certain patterns that can be identified and recognized. There are, however, often changes as the child gets older and as the result of intervention. 25 The term ‘gait classification’ refers to a system that allows the allocation of gait patterns into groups that can be differentiated from one another based on a set of defined variables. 26 Classification scales for gait patterns in children with hemiplegic CP have been described by a number of authors but the one most widely used and frequently quoted in the literature was reported by Winters et al. 27 This scale subdivides the gait patterns into four groups based on the sagittal plane joint angles derived from three dimensional gait analysis (table 1). The scale accounts for severity from a distal to proximal involvement. For this thesis, patients were selected according to the Winters et al. gait classification.

Table 1: Classification of hemiplegic gait according to Winters and colleagues 27

Type I

Ankle equinus in the swing phase of gait due to a relative overactivity of the triceps surae compared to the anterior tibial musculature.

Type II

Plantarflexion throughout stance and swing from either static or dynamic contracture of the triceps surae. Knee is often forced into slight hyperextension in middle or late stance.

Type III

Findings of type II with reduced range of knee flexion/extension (< 45°). Reference is common to sub-types IIIa: reduced knee extension during stance, and IIIb: hyperextension in stance and reduced flexion in swing.

Type IV

Findings of type III with additional involvement of the hip musculature. Reduced hip range of motion (< 35°).

Orthotic treatment for toe walking

and

pharmacotherapeutic (e.g. botulinum toxin type A treatment), ankle foot

Amongst

the

various

choices

such

as

surgery,

physical

therapy

9

Chapter 1

orthoses (AFOs) are a common treatment measure to correct the toe walking gait pattern and prevent ankle plantar flexor contractures in children with hemiplegic CP. Most frequently applied AFOs are the solid AFO, hinged AFO, dynamic AFO and posterior leaf-spring. All these orthoses prevent plantar flexion motion but the amount of dorsiflexion depends on the specific type. 28 A disadvantage of the solid AFO (figure 1A) is that it completely restricts the dorsiflexion movement of the ankle. As such it limits the normal forward movement of the tibia over the weight bearing foot which leads to an early heel rise in the stance phase of gait. 29,30 Therefore the hinged or articulated polypropylene AFO with a plantar flexion stop has been increasingly recommended by clinicians to decrease equinus positioning (figure 1B). 31 Unlike the solid AFO, the hinged AFO allows the tibia to move forward over the weight bearing foot during stance resulting in more normal ankle dorsiflexion. 29,32 Most children with hemiplegic CP will benefit from the use of an AFO but its role will differ depending on the type. 33 According to the Winters et al. 27 classification those patients with type I hemiplegia (the mildest form) require an orthosis primarily to assist the dorsiflexors in preventing foot drop during swing. However, those with type II hemiplegia require a device capable of controlling the equinus through stance and swing. Individuals with type IIIa require a device that

and swing. Individuals with type IIIa require a device that Figure 1. Two types of ankle

Figure 1. Two types of ankle foot orthoses (AFOs) with A) a solid (i.e. stiff) AFO and B) a hinged AFO to allow for dorsiflexion.

10

General introduction

will facilitate a heel contact and the development of an external knee extension moment. Those with type IIIb require a device that will resist the excessive external knee extension moment, which is pushing the knee into hyperextension. The proximal involvement in those with type IV hemiplegia may limit what can be achieved by orthoses. Individuals with type III and IV may often also have significant fixed contractures of the Achilles tendon, which cannot be controlled by an AFO. Furthermore, the AFO also has the potential to support deformities of the foot. 33 This thesis will focus on the gait and muscular changes due to orthoses worn by patients of the hemiplegic types I and II. It will be investigated how an AFO biomechanically influences the gait pattern in patients and if an AFO can modify muscular activation patterns with a change of gait pattern.

Instrumented gait analysis

The term gait analysis can mean many things to different people, from a brief observation to sophisticated computerized measurements. 2 The work for this thesis uses instrumented three dimensional gait analysis approaches to examine and understand the pathological toe walking gait pattern as observed in children with spastic hemiplegic CP and the biomechanical and muscular changes that occur with the use of an AFO. A variety of devices are used to measure joint kinematics, kinetics, or electrical activity of muscles. A video-based motion analysis system analyses the kinematic part of locomotion. For calculating joint kinematics, sets of markers (typically small spheres covered with reflective tape) are placed on the skin in relation to anatomical or bony landmarks to define body segments (figure 2). These segments are made of a bony part and by soft tissues (skin layers, muscle tissue). Bony segments are considered to be rigid bodies according to Classical Mechanics. 34 Although the soft tissues around the bony segments are deformable, most literature regards the entire body segment to be rigid and ignores this deformability. If two or more cameras detect a marker and the position and orientation of these cameras are known then it is possible to calculate the three dimensional position of that marker. Joint centres and segmental coordinate systems are then defined to calculate angular kinematics. Joint kinematics is the description of the relative movement between two contiguous bony segments, the proximal and distal. 34 Studying kinetics of the lower limbs are undertaken by integrating a force plate system with the motion analysis system. For calculation of kinetic

11

Chapter 1

parameters, the masses of the limb segments and the location of their centre of gravity need to be known. Typically, published data from the literature are taken and modified to suit the subject’s anthropometry. With the measured ground reaction force by the force plates during the stance phase of gait and the joint centre locations as determined through the kinematics, joint moments and powers are calculated. Although useful information about muscle action during gait is obtained from joint moments and powers, only the net moment created by all of the forces crossing the joint are obtained. The contribution of single muscles cannot be determined without additional information. 35 Therefore, electrical activity of single muscles is recorded through electromyography (EMG). Surface EMG recorded during voluntary muscle action (e.g. gait) can be considered as a signal where the contributions of all active motor units are intermingled in a so-called interference pattern. The signal characteristics are largely dependent on the properties of the contributing motor units, their firing patterns and their interdependence. 36 In biomechanics there are several applications in which surface EMG is used. It can

several applications in which surface EMG is used. It can Figure 2. Data collection and analysis

Figure 2. Data collection and analysis process for instrumented gait analysis

12

General introduction

be used to determine the activation timing of one or multiple muscles, which is when the excitation to a muscle begins and ends. As such, the activation patterns can give an overview of the coordination between multiple muscle groups. In case of pathology, for example, co-contraction between agonist and antagonists can be identified. Other applications of surface EMG in biomechanics are: investigation of the muscle force / EMG signal relationship and the use of the EMG signal as a fatigue index. 37 The information extracted from EMG signals are primarily amplitude, timing and frequency or conduction velocity. A time-frequency analysis, based on wavelets, to process EMG signals is increasingly being reported in the literature. 38,39 This wavelet transform allows more detailed analysis and better resolution of changes over time than any other processing technique. On the microscopic level the analysis tries to resolve the interplay of motor units or muscle fibre types which requires short time resolutions. For running, for example, the EMG signals were simultaneously resolved for amplitude, timing and frequency content using non-linearly scaled wavelets of specified resolution for each gait cycle by Von Tscharner. 40,41 He combined the wavelet-based analysis with principle component analysis and pattern recognition methods to extract macroscopic (timing, intensity and frequency distribution) and microscopic (motor units, fibre types) aspects from the surface EMG signals. In research of CP patients this analysis technique was only recently introduced in the literature. 42-44

Outline of the thesis

Children with spastic hemiplegic CP walk differently compared to normal children with toe walking being the most common gait deviation. Chapter 2 addresses the first research question namely how the unilateral toe walking gait pattern in patients with hemiplegic CP differs from voluntary unilateral toe walking in neurological healthy subjects. Furthermore, it is questioned whether it is possible to discriminate between primary deviations in muscle activation patterns due to the neurological disorder of hemiplegic CP itself and secondary (i.e. compensatory) deviations that follow due to the biomechanics of the altered gait pattern. It is unknown what muscle patterns are needed for unilateral toe walking in neurologically healthy subjects. This chapter, therefore, shows the results of a study on changes in muscle activity, kinematics, and kinetics when healthy subjects mimic a hemiplegic walking pattern. Results of the healthy subjects are compared to those of the hemiplegic CP patients. In this way it could

13

Chapter 1

be speculated which components of the muscle activity pattern of the patient were pathological and which parts were due to the biomechanics of unilateral toe walking. In the following chapters the question is addressed as to how toe walking can be affected by the use of an AFO. By prescribing a patient an orthosis, one attempts to modify the way this patient controls gait. The effect of this treatment can only be understood in the context of the questions of how normal gait is controlled, how it can be influenced, and how these processes are altered in patients with hemiplegic CP. One way to investigate changes in the gait pattern by orthotic intervention is to compare selected parameters between walking conditions with and without orthoses within the patient population. Such studies are described in chapters 3, 4, and 5. Chapter 3 addresses the question how AFOs biomechanically influence the gait pattern. It describes the comparison of two types of orthoses used by children with hemiplegic CP. Since both orthoses have the same aim, that is enhance gait function and promote a heel-toe gait pattern, the study was performed to evaluate and compare both types of orthoses. Outcome helped to optimize orthotic management and understand how they influence gait. Chapter 4 describes a study that investigated activity of specific leg muscles when children with spastic hemiplegic CP walked with a hinged AFO. Muscle activity was compared in the affected limb when the children walked barefoot and with the AFO and footwear. The study was performed to determine if children with hemiplegic CP can change the muscle activity pattern with a change of gait pattern, despite the brain being damaged. In Chapter 5, similarly to chapter 4, the activity patterns of selected leg muscles in healthy and hemiplegic CP children with and without a hinged AFO were studied. The EMG signals, however, were analysed with a processing technique using non-linearly scaled wavelets to investigate the time-frequency content in both the affected and unaffected leg. In all chapters so far, the primary interest was on toe walking. However, other patient groups have more problems with the use of heel strike than of toe walking. The reason is that a heel landing leads to a rather brisk loading of the limb. One has therefore speculated that the use of damping material would be beneficial for such patients. However, one first has to know how such damping affects the basic walking pattern. Chapter 6 therefore compares the kinematic and EMG data in

14

General introduction

healthy subjects walking with normal shoes and with a special therapeutic shoe of which the heel is replaced by a soft cushioning material. This therapeutic shoe has become a product often used for therapy by physiotherapists and is sold by regular sport stores and orthopaedic shoe-makers. It is important to realize the changes that occur in the walking pattern and muscle activation. Chapter 7 generally discusses the results of this thesis in the light of its original aim. It emphasizes the limitations and possibilities of the applied methods and directions for future research.

References

1. Perry J. Gait analysis: Normal and pathological function. Thorofare, NJ, USA: SLACK Incorporated, 1992.

2. Kirtley C. Clinical gait analysis: theory and practice. Edinburgh; New York: Churchill Livingstone Elsevier: 2006.

3. Duysens J, Van de Crommert HWAA. Neural control of locomotion; Part 1: The central pattern generator from cats to humans. Gait Posture 1998;7:131-41.

4. Den Otter AR, Geurts ACH, Mulder T, Duysens J. Gait recovery is not associated with changes in the temporal patterning of muscle activity during treadmill walking in patients with post-stroke hemiparesis. Clin Neurophysiol 2006;117:4-15.

5. Freud S. Die infantile Cerebrallähmung. In: Nothnagel H, editor. Specielle Pathologie und Therapie, Bd IX, Teil III. Vienna: Holder 1897;1-327.

6. Balf CL, Ingram TTS. Problems in the classification of cerebral palsy in childhood. Br Med J

1955;16:163-6.

7. Bax MCO. Terminology and classification of cerebral palsy. Dev Med child Neurol

1964;6:295-7.

8. Mutch L, Alberman E, Hagberg B, Kodama K, Velickovic Perat M. Cerebral palsy epidemiology: where are we now and where are we going? Dev Med Child Neurol

1992;34:547-51.

9. Cans C. Surveillance of cerebral palsy in Europe: a collaboration of cerebral palsy surveys and registers. Surveillance of Cerebral Palsy in Europe (SCPE). Dev Med Child Neurol.

2000;42:816-24

10. Bax M, Goldstein M, Rosenbaum P, Leviton A, Paneth N, Dan B, Jacobsson B, Damiano D. Proposed definition and classification of cerebral palsy Dev Med Child Neurol 2005;47:571-6.

11. Hutton JL, Pharoah POD. Life expectancy in severe cerebral palsy. Arch Dis Child

2006;91:254-8.

15

Chapter 1

12.

Goodman MJ, Menown JL, West JM, Barr KM, Van der Linden DW, McMulkin ML. Secondary gait compensations in individuals without neuromuscular involvement following a unilateral imposed equinus constraint. Gait Posture 2004;20:238-244.

13.

Davids JR, Foti T, Dabelstein J, Bagley A. Voluntary (normal) versus obligatory (cerebral palsy) toe-walking in children: a kinematic, kinetic, and electromyographic analysis. J Pediatr Orthop

1999;19:461-469.

11.

Hoffer MM. Basic considerations and classifications of cerebral palsy. American Academy of Orthopedic Surgeons Instructional Course Lectures 1976;25:96-106.

12.

Minear WL. A classification of cerebral palsy. Pediatrics 1956;18:841.

13.

Gage JR. Gait analysis in cerebral palsy. London: Mac Keith Press; 1991.

14.

Lance JW (1980). Pathophysiology of spasticity and clinical experience with baclofen. In:

Feldman RG, Young RR, Koella WP (eds.). Spasticity: Disordered Motor Control. Chicago, IL: Year Book Medical Publishers 1980; pp. 185-220

15.

Colver AF, Gibson M, Hey EN, Jarvis SN, Mackie PC, Richmond S. Increasing rates of cerebral palsy across the severity spectrum in north-east England 1964-1993. Arch Dis Child Fetal Neonatal Ed 2000;83:F7-F12.

16.

Hagberg B, Hagberg G, Beckung E, Uvebrant P. Changing panorama of cerebral palsy in Sweden. VIII. Prevalence and origin in the birth year period 1991–94. Acta Paediatr

2001;90:271-7.

17.

Dolk H, Parkes J, Hill N. Trends in the prevalence of cerebral palsy in Northern Ireland, 1981-1997. Dev Med Child Neurol 2006;48:406-12.

18.

Jessen C, Mackie P, Jarvis S. Epidemiology of cerebral palsy. Arch Dis Child Fetal Neonatal Ed

1999;80:F158.

19.

Meberg A, Broch H. Etiology of cerebral palsy. J Perinat Med 2004;32:434-9.

20.

Mongan D, Dunne K, O’Nuallain S, Gaffney G. Prevalence of cerebral palsy in the West of Ireland 1990-1999. Dev Med Child Neurol 2006;48:892-5.

21.

Pharaoh POD, Cooke T, Johnson MA, King R, Mutch L. Epidemiology of cerebral palsy in England and Scotland, 1984-9. Ach Dis Child Fetal Neonatal Ed 1998;79:F21-F25.

22.

Winter S, Autry A, Boyle C, Yeargin-Allsopp M. Trends in the prevalence of cerebral palsy in a population-based study. Pediatrics 2002;110:1220-5.

23.

Surveillance of Cerebral Palsy in Europe. Prevalence and characteristics of children with cerebral palsy in Europe. Dev Med Child Neurol 2002;44:633-40.

24.

Wren TAL, Rethlefsen S, Kay RM. Prevalence of specific gait abnormalities in children with cerebral palsy: Influence of cerebral palsy subtype, age, and previous surgery. J Pediatr Orthop

2005;25:79-83.

16

General introduction

25. Rodda J, Graham HK. Classification of gait patterns in spastic hemiplegia and spastic diplegia: a basis for a management algorithm. European Journal of Neurology 2001;8(Suppl. 5):98-108.

26. Dobson F, Morris ME, Baker R, Graham HK. Gait classification in children with cerebral palsy:

A systematic review. Gait Posture 2007;25:140-152.

27. Winters TF, Gage JR, Hicks R. Gait patterns in spastic hemiplegia in children and young adults.

J Bone Joint surg (Am) 1987;69:437-441.

28. Desloovere K, Molenaers G, Van Gestel L, Huenaerts C, Van Campenhout A, Callewaert B, Van de Walle P, Seyler J. How can push-off be preserved during use of an ankle foot orthosis in children with hemiplegia? A prospective controlled study. Gait Posture 2006;142-51.

29. Carmick J. Managing equinus in a child with cerebral palsy: merits of hinged ankle foot orthoses. Dev Med Child Neurol 1995;37:1006-19.

30. Abel MF, Juhl GA, Vaughan CL, Damiano DL. Gait assessment of fixed ankle-foot orthoses in children with spastic diplegia. Arch Phys Med Rehabil 1998;79:126-33.

31. Knutson L, Clark D. Orthotic devices for ambulation in children with cerebral palsy and myelomeningocele. Phys Ther 1991;71:947-60.

32. Brunner R, Meier G, Ruepp T. Comparison of a stiff and a spring-type ankle-foot orthosis to improve gait in spastic hemiplegic children. J Pediatr Orthop B 1998;18:719-26.

33. Thompson NS, Taylor TC, McCarthy KR, Cosgrove AP, Baker RJ. Effect of a rigid ankle–foot orthosis on hamstring length in children with hemiplegia. Dev Med Child Neurol 2002;44:51-7.

34. Cappozzo A, Della Croce U, Leardini A, Chiari L. Human movement analysis using stereophotogrammetry Part 1: theoretical background. Gait Posture 2005;21:186-96.

35. Sutherland DH. The evolution of clinical gait analysis part 1: kinesiological EMG. Gait Posture

2001;14:61-70.

36. Stegeman DF, Blok JH, Hermens HJ, Roeleveld K. Surface EMG models: properties and applications. J Electromyogr Kinesiol 2000;10:313-26.

37. DeLuca CJ. The use of surface electromyography in biomechanics. J Appl Biomech

1997;13:135-63.

38. Von Tscharner V. Intensity analysis in time-frequency space of surface myoelectric signals by wavelets of specified resolution. J Electromyogr Kinesiol 2000;10:433-45.

39. Karlsson JS, Östlund N, Larsson B, Gerdle B. An estimation of the influence of force decrease on the mean power spectral frequency shift of the EMG during repetitive maximum dynamic knee extension. J Electromyogr Kinesiol 2003;13:461-8.

40. Von Tscharner V, Goepfert B, Nigg BM. Changes in EMG signals for the muscle tibialis anterior while running barefoot or with shoes resolved by non-linearly scaled wavelets.

J Biomech 2002;36:1169-76.

17

Chapter 1

41. Von Tscharner V, Goepfert B. Gender dependent EMGs of runners resolved by time/frequency and principal pattern analysis. J Electromyogr Kinesiol 2003;13:253-72.

42. Lauer RT, Stackhouse C, Shewokis PA, Smith BT, Orlin M, McCarthy JJ. Assessment of wavelet analysis of gait in children with typical development and cerebral palsy. J Biomech

2005;38:1351-7.

43. Wakeling J, Delaney R, Dudkiewicz I. A method for quantifying dynamic muscle dysfunction in children and young adults with cerebral palsy. Gait Posture 2007;25:580-9.

44. Lauer RT, Stackhouse CA, Shewokis PA, Smith BT, Tucker CA, McCarthy J. A time-frequency based electromyographic analysis technique for use in cerebral palsy. Gait Posture 2007;26:420-7.

18

Chapter

2

An electromyographic analysis of obligatory (hemiplegic cerebral palsy) and voluntary (normal) unilateral toe walking

Jacqueline Romkes & Reinald Brunner

Gait and Posture 2007;26:577-586

Chapter 2

Abstract

This study compares lower extremity muscle activation during gait in patients with hemiplegic cerebral palsy (CP) with healthy subjects mimicking the patients. The purpose was to understand the differences between obligatory toe walking as observed in hemiplegic CP gait and voluntary toe walking. The results contribute

to a better understanding by distinguishing between primary deviations in muscle

activity as a direct consequence of the underlying neurological pathology of hemiplegic CP and secondary, compensatory deviations due to the biomechanics of toe walking. Surface electromyographic (EMG), kinematic and kinetic data were compared between a group of 12 hemiplegic CP patients and a group of 10 healthy subjects walking normally and when mimicking hemiplegic gait. Integrated 3D gait analysis was performed with simultaneous EMG recordings of the medial gastrocnemius, tibialis anterior, rectus femoris, and semitendinosus muscles bilaterally. The EMG pattern of the toe walking leg in the mimicking subjects was modified in gastrocnemius and tibialis anterior as compared with normal gait. This

modified muscle activation pattern showed strong similarities to that of the patients and therefore can be regarded, at least in part, as activity required for toe walking.

A possible primary gait abnormality was observed in the rectus femoris where the

patients showed a burst of EMG activity in mid-swing phase in contrast to the mimicking subjects who showed normal EMG during this phase.

20

Introduction

Obligatory versus voluntary toe walking

Patients with hemiplegic cerebral palsy (CP) often demonstrate a toe walking gait pattern. This pattern is characterized by distinct kinematic, kinetic, and electromyographic (EMG) changes relative to normal heel-toe gait. 1 It is important to distinguish between primary gait deviations which are a direct consequence of the underlying neuromuscular disorder, and secondary deviations which are compensatory. Primary gait deviations should be regarded as the focus of treatment. 2 CP results from an injury to the developing central nervous system, which can occur in utero, during delivery, or during the first 2 years of life. 1 Hemiplegic CP is typically associated with single hemisphere injury and there is involvement of predominantly one body side with relative sparing of the contralateral limbs. 1 Characteristic signs of CP are movement disorders, spasticity, muscle weakness, ataxia, and rigidity. 3 A widely recognized classification system for hemiplegia based on joint kinematic data during gait was proposed by Winters et al. 4 who classified the patterns into four types. A principle finding in type I (mildest form) is equinus in swing secondary to a relative overactivity of the triceps surae compared to the anterior tibial musculature. The mean feature of type II is ankle plantarflexion throughout the gait cycle, in both swing and stance from either static or dynamic contracture of the triceps surae. The knee is often forced into slight hyperextension in middle or late stance. Patients with type III show limited swing phase knee motion in addition to the gait anomalies of type II. Patients can either show reduced knee extension during stance or hyperextension in stance and reduced flexion in swing. The patients in group IV show all features of type III with additional involvement of the hip musculature. In types III and IV the musculature in the proximal part of the lower extremity is more involved than in types I and II. The damage to the central nervous system appears to be least in the latter two groups. 4 Toe walking is an obligatory gait pattern for persons with calf muscle spasticity or primitive control that prevents heel contact with the ground. 5 Excessive plantarflexion of the ankle may be present in swing and/or stance phase. Anterior tibialis muscle weakness, dysfunction, or premature gastrocnemius and soleus muscle contraction causes swing phase equinus with resultant dragging of the foot, tripping, and proximal joint kinematic compensations. 6 Treatment of toe

21

Chapter 2

walking includes physical therapy, orthotics, local medication (botulinum toxin-A) and surgery. Patients with type I hemiplegia usually need orthotic treatment (ankle foot orthosis) but do not require a lengthening of the plantarflexor muscles whereas in patients with type II both may be necessary if the calf muscles are shortened. 4 Studies investigating bilateral toe walking in healthy subjects showed premature activation of the calf muscles starting in the preceding initial swing phase when tibialis anterior EMG activity ceased much earlier in toe walking compared to heel- toe gait. 5 Similar activity patterns were observed in EMG studies in patients with equinus gait. 7-9 The aim of the current study was to understand the differences seen in EMG between obligatory unilateral toe walking as observed in a hemiplegic CP gait and voluntary unilateral toe walking. The results should contribute towards a better understanding by distinguishing between primary deviations in muscle activity as a direct consequence of the underlying neurological pathology of hemiplegic CP and deviations due to the biomechanics of toe walking.

Methods

Subjects

The study group consisted of 12 children (8 boys, 4 girls; age 12.3 ± 4.1 years) diagnosed with hemiplegic CP (9 right, 3 left). Inclusion criteria were: diagnosis of hemiplegic CP, between 8 years and 18 years of age, initial toe-contact during barefoot walking, type I or II classification for hemiplegic CP as described by Winters et al. 4 , passive knee extension deficit less than 5°, no contractures or signs of spasticity at hip level. The children had not received Botulinum toxin-A treatment in the previous 6 months nor any orthopaedic surgery in the past. Passive range of motion (RoM) of the hemiplegic side in dorsiflexion ranged from -10° to 20° with the knee flexed and from -20° to 10° with knee extended. Four patients did not reach an angle of 0° at the foot. At the knee, two patients had 5° of flexion contracture but no patients had any hip contracture. Spasticity was evaluated according to the modified Ashworth scale which for the plantarflexors ranged between a score of 2 (seven patients) and 4 (two patients). 10 In two children the Ashworth scale was 2 for knee extensors and flexors, while all others were scored as 1 (i.e. normal). Scores at the hip were 1 for all children. Clinical muscle tests showed reduced strength against manual resistance in all patients for both plantarflexors and dorsiflexors. According to the Winters et al. 4 classification, six

22

Obligatory versus voluntary toe walking

children were type I (mildest form with plantarflexion throughout swing) and six children as type II (plantarflexion over entire gait cycle). Classification was based on the collected sagittal plane kinematic data. The patients showed a leg-length discrepancy varying between 0 cm and 3 cm, with the hemiplegic leg being always shorter. Reference data were collected in a group of 10 healthy adult subjects (5 males, 5 females; age 29.5 ± 3.3 years) without a history of neurologic or orthopaedic deficits. They had no significant leg-length discrepancy. We were ethically not allowed to perform the study with children.

Gait analysis

Kinematic and kinetic data were collected by 3D gait analysis using a six-camera, 120 Hz movement analysis system (VICON 460, Oxford Metrics Ltd., UK) and two force platforms (Kistler Instrumente AG, Winterthur, Switzerland). All subjects walked barefoot at a self-selected speed. The healthy subjects were first tested during normal walking and were then shown a sagittal plane video of one of the patients with a type II right hemiplegic gait. The same video was shown to all healthy subjects. The subjects were asked to mimic the gait pattern seen in the video. The subjects first practiced to obtain a reproducible pattern, while care was taken by the investigators to ensure that knee and ankle positions were correct. Fifteen passive reflective markers (diameter 14 mm) were fixed to specific landmarks bilaterally on the subject’s legs and pelvis according to the protocol of Kadaba et al. 11 Height, weight, leg length, width of ankles and knees, and tibial torsion were measured clinically for appropriate anthropometric scaling. Kinematic and kinetic data were analysed in the sagittal plane and normalized over the gait cycle using the Polygon software (Oxford Metrics Ltd., UK). Kinetic data were normalized with respect to body weight. The spatiotemporal parameters investigated were duration of stance, single support, and double support expressed as percentage of gait cycle. Kinematic parameters included peak flexion and extension values during the gait cycle as well as the RoM for pelvis, hip, knee and ankle. For the knee and ankle, the angles at initial foot contact were also calculated. Kinetic parameters included peak joint moment and power values during stance for the hip, knee and ankle joints. The value for each parameter was taken from six individual trials for each of the conditions and averaged within subjects.

23

Chapter 2

EMG measurements

Simultaneously with the 3D gait analysis, surface EMG was recorded. Bipolar Ag/AgCl surface electrode pairs (10 mm diameter, 22 mm inter-electrode spacing) were placed on the clean shaved skin bilaterally overlying the tibialis anterior, medial gastrocnemius, rectus femoris, and semitendinosus muscles following the SENIAM 12 guidelines. The ground electrode was placed over the tibial tuberosity. Raw EMG signals were visually inspected for artefacts and noise. The electrodes were connected to single differential amplifiers with a band path of 10-700 Hz (Biovision AG, Wehrheim, Germany). The data were collected at a sampling frequency of 2520 Hz using a Zebris system (Zebris, Tübingen, Germany). The EMG signal was full wave rectified and the moving average was calculated with a time constant of 39.6 ms. For each subject the data of six gait cycles were analysed and each data set was interpolated to the smallest number of data points for one cycle. In order to investigate changes in EMG, the amplitude was normalized to the average value of each cycle for the respective muscle and subject. The data were then divided into nine intervals according to the gait phases as described by Perry 13 : initial contact and loading response (0-10 %), mid-stance I (10-20 %), mid-stance II (20- 30 %), terminal stance I (30-40 %), terminal stance II (40-50 %), pre-swing (50-60 %), initial swing (60-73 %), mid-swing (73-87 %) and terminal swing (87-100 %). An average of the six gait cycles was calculated for each subject, muscle, and interval. The MATLAB software package (The MathWorks Inc.) was used.

Statistical analysis

To determine if there were differences between the three conditions within each side (toe walking or non-toe walking side), the group data (kinematics, kinetics, intervals of EMG) were analysed using multiple one-way-analysis of variance (ANOVA) with Tukey’s multiple comparison test. Paired t-tests were applied where differences between the toe walking and non-toe walking side within a group were analyzed. The SPSS 13.0 statistical software was used. The level of significance was set at P < 0.05.

Results

Spatiotemporal, kinematic and kinetic parameters

Mimicking did not change any of the spatiotemporal parameters compared to

24

Obligatory versus voluntary toe walking

normal gait. The patients, however, did show significant differences to both normal and mimicking (table 1). There were no significant differences in kinematics between the patients’ hemiplegic leg and the controls’ toe walking leg except for the increased pelvic RoM in the patients (table 2, figure 1). Peak hip flexion over the gait cycle significantly increased in the mimicking controls but was not different from the patients. Differences were larger for the patient’s non-hemiplegic leg (table 3, figure 1). Pelvic and hip RoMs as well as peak hip flexion were increased, and there was more knee flexion at initial contact and reduced knee extension during stance. Knee RoM was decreased for both the mimicking controls and patients compared to normal gait whereas peak knee flexion was only decreased in the mimicking controls but not in the patients. Unilateral toe walking produced significantly different internal moment patterns at the ankle as compared to normal gait. Normal gait generated a small internal dorsiflexion moment (0.22 ± 0.08 Nm/kg at 4 % gait cycle), which then reversed at 10 ± 3 % gait cycle into an increasing plantarflexion moment. With unilateral toe walking, the gait cycle commenced with an increasing internal plantarflexion moment (table 4, figure 2). The kinetic data on the toe walking side did not show significant differences between the patients and the mimicking controls except for the internal peak plantarflexion moment at 0-30 % gait cycle which was slightly higher for the mimicking controls (table 4). At the knee, a reduction of peak knee power absorption from 0 % to 20 % gait cycle, a reduction of internal peak plantarflexion moment in terminal stance, and a reduction in peak ankle power generation in terminal stance was observed (figure 3). The only significant difference observed for the non-toe walking leg was a reduced internal knee flexion moment during terminal stance for the patients (table 5).

EMG data

The results for the EMG data of the toe walking and non-toe walking leg are shown in figure 4 with significant differences (P < 0.05) in activation level per gait phase indicated with asterisks. For the toe walking leg, the EMG patterns for the gastrocnemius muscle were similar both in the patients and in the mimicking controls. Whereas level of activity was low for the entire swing phase, initial contact, and loading response in normal gait, activity was significantly higher for the mimicking subjects for 73-100 % (i.e. mid and terminal swing) and 0-10 % gait cycle (i.e. initial contact and loading response). For the patients, activity was

25

Chapter 2 26 Table 1. Spatiotemporal parameters Parameter Toe walking side Non-toe walking side Healthy
Chapter 2
26
Table 1. Spatiotemporal parameters
Parameter
Toe walking side
Non-toe walking side
Healthy right
Patient
Mimic
Healthy left
Patient
Mimic
Stance phase (% GC)
61.4
± 0.9
59.6
± 3.5 *
61.5
± 1.2
61.1
40.6
± 0.9
± 1.3
Single support (% GC)
39.4
22.2
36.3
23.3
± 1.8 *, **
± 1.6
63.8
40.8
± 1.6 **
61.1
± 1.9
± 1.8
± 3.7
± 1.7
Double support (% GC)
± 2.1
± 4.9
39.6
21.9
± 2.7
20.8
± 2.2
23.1
± 4.5
39.3
21.9
± 2.7
Data are reported as group mean ± 1 standard deviation.
* Significant
difference
between
the normal
toe walking
and non-toe
side to (paired
t-test
0.05)
**
Patients differ
significantly
from
and mimicking
gait walking
(referred
the same
side, P < ANOVA
P < 0.05)

Obligatory versus voluntary toe walking

Obligatory versus voluntary toe walking Figure 1. Sagittal plane kinematic movement curves for pelvis, hip, knee,

Figure 1. Sagittal plane kinematic movement curves for pelvis, hip, knee, and ankle (group mean ± 1 standard deviation).

27

Chapter 2

Table 2. Sagittal plane kinematic parameters for the toe walking leg

Parameter

Normal right

Patient

Mimic

Pelvis

RoM Average RoM Peak flex Peak ext RoM Angle at IC Peak flex Peak ext RoM DF at IC Peak DF Peak PF

1.92 ± 0.81 14.29 ± 3.72 46.31 ± 3.70 41.52 ± 3.75 4.79 ± 4.59 59.74 ± 4.20 8.66 ± 5.27 62.13 ± 3.39 -2.39 ± 4.44 33.04 ± 4.24 1.13 ± 2.14 14.75 ± 3.96 18.28 ± 5.82

 

8.04 ± 5.20 13.87 ± 6.37 45.45 ± 8.76 45.09 ± 6.75 0.36 ± 11.02 52.64 ± 13.89 21.59 ± 8.98 60.18 ± 4.59 -7.54 ± 11.49 30.14 ± 11.12 -18.10 ± 7.99 4.40 ± 11.15 25.73 ± 12.42

* ^

2.32 ± 0.66 15.59 ± 4.04 51.11 ± 4.94 49.05 ± 4.45 2.06 ± 4.25 52.49 ± 4.98 24.84 ± 8.31 63.39 ± 6.48 -10.90 ± 5.45 27.30 ± 4.25 -18.74 ± 5.67 -1.12 ± 4.86 28.42 ± 4.64

 

Hip

*

Knee

** ^

Ankle

** ^

** ^

Data are reported as group mean ± 1 standard deviation. * significantly different from normal; ** significantly different from patient; ^ significantly different from mimicking. RoM: range of motion; flex: flexion; ext: extension; IC: initial contact; DF: dorsiflexion; PF: plantarflexion.

increased compared to normal for 87-100 % (i.e. terminal swing) and 0-10 %. All three groups showed higher levels of activation throughout stance. For the non-toe walking side, EMG activity was significantly higher during pre-swing (50-60 %) in the patients. For the tibialis anterior muscle EMG was also similar in the patients and mimicking controls. One of the differences observed on the toe walking side between the two groups was in pre-swing (50-60 %) where activity was significantly increased for the patients whereas the mimicking subjects did not differ from normal. There was a difference in activity in terminal swing (87-100 %) between all three groups. Midstance (10-30 %) activity for the non-toe walking side was increased in the patients. Observed changes in the rectus femoris muscle on the toe walking side were more complex. Until pre-swing (0-50 %) the EMG showed a similar pattern for all

28

Obligatory versus voluntary toe walking

three groups as visually observed from the figure. However, there were differences in amplitude level. In pre-swing and initial swing, the activity in the mimicking group decreased whereas activity for the patients increased. During midswing, the mimicking subjects showed equal activity (no significant difference) as in normal gait, whereas the patients showed an abnormal pattern with significantly increased activity. On the non-toe walking side, the patients showed increased activity in the second half of midstance (20-30 %) and in midswing (73-87 %). Activation pattern of the semitendinosus muscle showed changes in amplitude level throughout stance for all three groups. On the toe walking side, the patients showed less variation in amplitude around the mean activity than in mimicking and healthy gait.

Table 3. Sagittal plane kinematic parameters for the non-toe walking leg

Parameter

Normal left

Patient

Mimic

Pelvis

RoM

1.99 ± 0.72

8.66 ± 5.25 * ^

2.38 ± 0.64

Average

14.05 ± 3.59

14.06 ± 6.41

15.68 ± 4.05

Hip

RoM

46.31 ± 4.55

53.45 ± 8.70 *

^

45.98 ± 3.11

Peak flex

41.24 ± 2.82

49.51 ± 6.68 *

^

40.80 ± 4.46

Peak ext

5.08 ± 4.20

3.94 ± 7.99

5.19 ± 3.57

Knee

RoM

61.30 ± 5.34 ** ^

54.43 ± 5.69

53.45 ± 7.04

Angle at IC

7.45 ± 4.44

17.90 ± 5.18 *

^

8.98 ± 5.24

Peak flex

61.43 ± 3.70

64.98 ± 6.42

53.29 ± 7.03 * **

Peak ext

-0.13 ± 2.54

10.54 ± 5.44 * ^

0.16 ± 1.51

Ankle

RoM

30.57 ± 5.53

30.05 ± 3.50

28.45 ± 5.80

DF at IC

0.21 ± 1.93

0.21 ± 5.05

2.79 ± 1.74

Peak DF

13.71 ± 2.65

16.87 ± 5.85 ^

11.12 ± 2.58

Peak PF

16.86 ± 5.82

13.19 ± 5.64

17.33 ± 6.19

Data are reported as group mean ± 1 standard deviation. * significantly different from normal; ** significantly different from patient; ^ significantly different from mimicking. RoM: range of motion; flex: flexion; ext: extension; IC: initial contact; DF: dorsiflexion; PF: plantarflexion.

29

Chapter 2

Table 4. Kinetic parameters for the toe walking leg

Parameter

Normal right

Patient

Mimic

Hip

Peak HE moment stance

1.14 ± 0.28

1.12 ± 0.42

0.96 ± 0.33

Peak HF moment stance

0.38 ± 0.34

0.39 ± 0.24

0.54 ± 0.17

Peak power abs. stance

0.63 ± 0.22

0.29 ± 0.23

0.35 ± 0.36

Peak power gen. Stance

1.13 ± 0.25

1.31 ± 0.61

1.09 ± 0.42

Knee

Peak KE moment 0-20 % GC

0.36 ± 0.09

0.11 ± 0.26

0.22 ± 0.30

Peak KF moment 30-60 % GC

0.50 ± 0.18

0.32 ± 0.17

0.31 ± 0.09

Peak power abs. 0-20 % GC

0.80 ± 0.32 **, ^

0.43 ± 0.35

0.31 ± 0.23

Peak power gen. 30-60 % GC

0.60 ± 0.36

0.34 ± 0.24

0.42 ± 0.30

Ankle

Peak PF moment 0-30 % GC

NR

0.97 ± 0.22

1.23 ± 0.19 **

Peak PF moment 30-60 % GC

1.65 ± 0.12 **, ^

1.07 ± 0.26

1.29 ± 0.24

Peak power abs. 0-30 % GC

0.49 ± 0.19 ^

1.23 ± 0.93

2.00 ± 0.89

Peak power gen. 30-60 % GC

4.71 ± 0.73 **, ^

2.29 ±0.69

2.87 ± 0.89

Data are reported as group mean ± 1 standard deviation. Moments are internal joint moments in Nm/kg and powers in W/kg. ** significantly different from patient; ^ significantly different from mimicking. HE: hip extension; HF: hip flexion; abs.:

absorption; gen.: generation; PF: plantarflexion; GC: gait cycle.

Discussion

The aim of the current study was to understand the differences between obligatory unilateral toe walking as observed in hemiplegic CP gait and voluntary toe- walking. The results should contribute to a better understanding by distinguishing between primary deviations in muscle activity as a direct consequence of the underlying neurological pathology of hemiplegic CP and deviations due to the biomechanics of toe walking. The gastrocnemius muscle is biarticular and influences both ankle and knee function. In normal gait this muscle progressively increases its activity during stance until peaking in terminal stance. Activity rapidly declines then, until it ceases with pre-swing, as the other foot contacts the floor. Its function in mid stance, together with the soleus muscle, is to restrain the passive dorsiflexion motion as the body advances over the foot in response to momentum from the

30

Obligatory versus voluntary toe walking

swing limb and forward fall of body weight. 13 In terminal stance the gastrocnemius muscle action stabilizes the dorsiflexing ankle and allows heel-rise. 13 Activity patterns of the gastrocnemius for the hemiplegic patients in this study are in agreement with earlier studies reported in the literature on EMG in CP with equines gait. 7-9 This study, however, showed that unilateral voluntary toe walking of healthy subjects results in an activation pattern very similar to that of the patients. Both groups showed early onset of activity in the terminal swing phase continuing throughout stance on the toe walking side. This activity pattern can be explained by the kinematic and kinetic data. With an initial toe contact instead of a the same EMG signal for patients (both types I and II hemiplegia) and mimicking subjects, we cannot distinguish between possible spastic activity, of the calf muscles and activity due to the biomechanical needs of toe walking. Furthermore, some patients had an ankle-contracture where the neutral joint position could not

Table 5. Kinetic parameters for the non-toe walking leg

Parameter

Normal left

Patient

Mimic

Hip

Peak HE moment stance

1.16 ± 0.24

1.36 ± 0.49

1.08 ± 0.39

Peak HF moment stance

0.42 ± 0.35

0.61 ± 0.27

0.54 ± 0.19

Peak power abs. stance

0.60 ± 0.21

0.56 ± 0.49

0.40 ± 0.30

Peak power gen. Stance

1.13 ± 0.23

1.40 ± 0.69

0.94 ± 0.36

Knee

Peak KE moment 0-20 % GC

0.40 ± 0.14

0.56 ± 0.44

0.57 ± 0.26

Peak KF moment 30-60 % GC

0.52 ± 0.16

0.27 ± 0.14 *, ^

0.59 ± 0.12

Peak power abs. 0-20 % GC

0.83 ± 0.33

1.15 ± 1.04

1.61 ± 0.74

Peak power gen. 30-60 % GC

0.61 ± 0.30

0.38 ± 0.34

0.65 ± 0.25

Ankle

Peak PF moment 0-30 % GC

NR

NR

NR

Peak PF moment 30-60 % GC

1.64 ± 0.12

1.48 ± 0.28

1.62 ± 1.39

Peak power abs. 0-30 % GC

0.51 ± 0.18

0.55 ± 0.26

0.59 ± 0.21

Peak power gen. 30-60 % GC

4.64 ± 0.74

4.51 ± 1.52

3.69 ± 0.94

Data are reported as group mean ± 1 standard deviation. Moments are internal joint moments in Nm/kg and powers in W/kg. * significantly different from normal; ^ significantly different from mimicking. HE: hip extension; HF: hip flexion; abs.:

absorption; gen.: generation; PF: plantarflexion; GC: gait cycle, NR: no relevant parameter.

31

Chapter 2

Chapter 2 Figure 2. Internal joint moment data for the hip, knee and ankle for the

Figure 2. Internal joint moment data for the hip, knee and ankle for the toe walking and non-toe walking side.

32

Obligatory versus voluntary toe walking

Obligatory versus voluntary toe walking Figure 3. Joint power data for the hip, knee and ankle

Figure 3. Joint power data for the hip, knee and ankle for the toe walking and non-toe walking side.

33

Chapter 2

Chapter 2 Figure 4. EMG data (group mean ± 1 standard deviat ion) over 1 gait

Figure 4. EMG data (group mean ± 1 standard deviation) over 1 gait cycle for normal gait (black), hemiplegic CP patients (dark grey) and mimicking controls (light grey) for both the toe walking and non-toe walking sides. Statistical differences (ANOVA with P < 0.05) are indicated with asterisks for each gait phase at the top of the graphs. Toe-off is indicated with a solid vertical line around 60 %.

34

Obligatory versus voluntary toe walking

be reached. It was also not possible to differentiate between these two types of patients because the EMGs were the same. However, we can state that the observed activity pattern of the gastrocnemius muscle is related to toe walking since it was not present in normal heel-toe gait. Therefore, since we observe it in healthy subjects, we expect it to be also present in toe walking CP patients. The activity can therefore not be considered as purely pathological. The gastrocnemius activity in terminal swing anticipates the plantarflexion at initial contact and helps to prepare the foot. At the knee, the plantarflexed position of the foot in hemiplegic gait limits knee extension at initial contact, as the tensed gastrocnemius muscle does not allow the knee to fully extend during the initial contact. 14 For the tibialis anterior muscle, both patients and mimicking subjects showed the same EMG pattern throughout the entire gait cycle. In normal gait, the tibialis anterior shows a large burst of EMG activity from initial contact to the end of loading response which controls foot placement. At initial contact the ankle is in neutral position and the forefoot is tilted upward. 13 The tibialis anterior helps to control plantarflexion at the end of loading response as indicated by the small internal dorsiflexion moment. A second burst of activity is seen starting in pre- swing and lasting for the whole swing phase. In pre-swing the ankle rapidly plantarflexes and the onset of tibialis anterior activity helps to decelerate the rate of foot fall. 13 Peak activity is reached in initial swing phase just after toe-off. Ankle motion now changes to dorsiflexion towards neutral position to allow floor clearance. In the patients and mimicking subjects the first burst of activity at initial stance phase was very low or absent. This can be explained by its function in this part of the gait cycle. In the case of initial toe-contact, activity at this phase is not required as shown by the internal plantarflexion moment. As the movement pattern at initial contact and loading response is reversed, the muscle activity of the agonist and antagonist muscles is also reversed. This is also the case at the end of swing where tibialis anterior activity is reduced due to the equinus position of the foot. The well-balanced phasic activity of the gastrocnemius, soleus and tibialis anterior muscles proves a well coordinated activity in the mimicking subjects and patients. The only possible primary deviation in the EMG activity for this group of patients could be observed in the rectus femoris muscle. Activity seen in the EMG at initial stance, loading response and terminal swing is regarded as crosstalk from the vasti muscles which control knee flexion as proposed by the work of Nene et al. 15,16 For the hemiplegic legs of all patients, the rectus femoris was active during

35

Chapter 2

mid-swing, which is consistent with the pattern of CP gait reported by others. 17 Whereas the EMG for the gastrocnemius and tibialis anterior muscles showed similar patterns in the patients and mimicking subject, EMG of the rectus femoris clearly differed between the two groups. None of the mimicking subjects showed the typical midswing activity for hemiplegia but had a normal quiet EMG instead. The rectus femoris is a biarticular muscle and anatomically it acts as an extensor of the knee. It simultaneously causes hip flexion especially during preswing. It is well documented that the rectus femoris is a causative factor of crouched gait or stiff- knee gait in children with CP. 18 The overactivity of the rectus femoris restricts knee flexion during swing and can subsequently limit foot clearance. 18 Although all patients in this study showed pathological rectus femoris activity during swing, a limitation of peak knee flexion was not present. This was also true for the two patients with 5° of knee flexion contracture. This rectus femoris activity may be regarded as primary as it was absent in the mimicking subjects. The semitendinosus muscle is part of the hamstrings and in normal gait begins contracting in the late swing phase (80 % gait cycle). It reaches its peak activity in terminal swing. At this point it prevents excessive knee extension and also decelerates hip flexion. 13 Activity continues into initial stance phase and together with the vasti it assists in hip and knee stability. 13 Although for all three groups the activity profiles of the toe walking leg showed the same pattern, the patients seem to have a reduced level of activity which is more centred around the mean activity to which the data were normalized. The method used in this study only refers to EMG profiles investigating activation patterns. No consideration has been given as to why the equinus gait developed in the patients or that muscle strength was reduced in all patients as observed in the clinical examination. For example, the gastrocnemius muscle of a patient may show a similar dynamic EMG pattern to that of mimicking subjects, the muscle-tendon-complex, however, can be short and strength reduced. Rose and McGill 19 suggested that weakness of both the gastrocnemius and tibialis anterior muscles in CP appears to result, in part, from the inability of the damaged motor pathways to provide sufficient excitatory drive to the motor neuron pool to fully activate all available motor units. Other sources of CP-related weakness may lie within the morphology of single muscle fibres and whole muscle, as well as muscle structural changes. 20 A weakness of the study was that the mimicking was not perfect. The

36

Obligatory versus voluntary toe walking

differences seen in spatiotemporal parameters, leg length discrepancy, kinematics and kinetics between the patients and mimicking subjects, may have influenced the EMG data. For example the decrease in peak knee flexion in swing for the mimicking non-toe walking leg, can be explained by the toe walking leg being functionally longer due to the equinus. Therefore, less knee flexion is needed to clear the swinging leg. A further possible source of error may arise from the age difference between the groups. However, a study by Õunpuu et al. 21 investigating 31 normal children by gait analysis showed data that were similar to that of normal adults and to adult data published by Kadaba et al. 11 Sutherland et al. 22 concluded that a mature gait pattern is established by 7 years of age but time-distance variables vary with age and stature. Since we were ethically not allowed to perform the study with children, we had to use adults. In conclusion, the results of this study suggest that in hemiplegic obligatory toe walking, activity of the rectus femoris muscle during mid-swing phase could be considered as a primary deviation due to the neurological condition. This phenomenon of mid-swing activity was absent in healthy subjects mimicking the toe walking patients. Further, muscle activation patterns of gastrocnemius and tibialis anterior as observed in the patients showed clear similarities to the patterns in the unilateral toe walking mimicking subjects. The increased loading response activity of the medial gastrocnemius muscle and reduced activity of the tibialis anterior muscle during unilateral toe walking compared to normal walking was present, at least in part, to meet the increased internal plantarflexion moment demands. It is possible that, in hemiplegic patients, a compensation for the altered gait in combination with muscle weakness and/or spasticity is responsible for the EMG activity observed.

Acknowledgement

The authors like to acknowledge the Foundation for Movement Disorders (Stiftung für Bewegungsstörungen) for their financial support.

References

1. Gage JR. Gait analysis in cerebral palsy. Oxford: MacKeith Press; 1991.

2. Davids JR, Foti T, Dabelstein J, Bagley A. Voluntary (normal) versus obligatory (cerebral palsy) toe walking in children: a kinematic, kinetic, and electromyographic analysis. J Pediatr Orthop

1999;19:461-9.

37

Chapter 2

3. Koman LA, Paterson Smith B, Shilt JS. Cerebral palsy. Lancet 2004;363:1619-31.

4. Winters TF, Gage JR, Hicks R. Gait patterns in spastic hemiplegia in children and young adults.

J Bone Joint Surg 1987;69-A:437-41.

5. Perry J, Burnfield JM, Gronley JK, Mulroy SJ. Toe walking: muscular demands at the ankle and knee. Arch Phys Med Rehabil 2003;84:7-16.

6. Aiona MD, Sussman MD. Treatment of spastic diplegia in patients with cerebral palsy: part II.

J Pediatr Orthop B 2004;13:S13-38.

7. Perry J, Hoffer MM, Giovan P, Antonelli D, Greenberg R. Gait analysis of the triceps surae in cerebral palsy. J Bone Joint Surg 1974;56-A:511-20.

8. Brunt D, Scarborough N. Ankle muscle activity during gait in children with cerebral palsy and equinovarus deformity. Arch Phys Med Rehabil 1988;69:115-7.

9. Higginson JS, Zajac FE, Neptune RR, Kautz SA, Burgar CG, Delp SL. Effect of equines foot placement and intrinsic muscle response on knee extension during stance. Gait Posture

2006;23:32-6.

10. Bohannon RW, Smith MB. Interrater reliability of a modified Ashworth scale of muscle spasticity. Phys Ther 1987;67(2):206-7.

11. Kadaba MP, Ramakrishnan HK, Wooten ME. Measurement of lower extremity kinematics during level walking. J Orthop Res 1990;8:383-92.

12. Hermens HJ, Freriks B, Merletti R, Stegeman DF, Blok J, Rau G, Disselhorst Klug C, Hägg G. European recommendations for surface electromyography, results of the SENIAM project. The Netherlands: Roessingh Research and Development; 1999.

13. Perry J. Gait analysis: normal and pathological function. Thorofare: SLACK Inc.; 1992.

14. Hullin MG, Robb JE, Loudon IR. Gait patterns in children with hemiplegic spastic cerebral palsy. J Pediatr Orthop 1996;5(B):247-51.

15. Nene A, Mayagoitia R, Veltink P. Assessment of rectus femoris function during initial swing phase. Gait Posture 1999;9:1-9.

16. Nene A, Byrne C, Hermens H. Is rectus femoris really a part of quadriceps? Assessment of rectus femoris during gait in able-bodied adults. Gait Posture 2004;20:1-13.

17. Patikas D, Wolf S, Döderlein L. Electromyographic evaluation of the sound and involved side during gait of spastic hemiplegic children with cerebral palsy. Eur J Neurol 2005;12:691-9.

18. Chambers H, Lauer A, Kaufman K, Cardelia JM, Sutherland D. Prediction of outcome after rectus femoris surgery in cerebral palsy: the role of cocontraction of the rectus femoris and vastus lateralis. J Pediatr Orthop 1998;18:703-11.

19. Rose J, McGill KC. Neuromuscular activation and motor-unit firing characteristics in cerebral palsy. Dev Med Child Neurol 2005;47:329-36.

38

Obligatory versus voluntary toe walking

20. Stackhouse SK, Binder-Macleod SA, Lee SCK. Voluntary muscle activation, contractile properties, and fatigability in children with and without cerebral palsy. Muscle Nerve

2005;31:594-601.

21. Õunpuu S, Gage JR, Davis RB. Three-dimensional lower extremity joint kinetics in normal pediatric gait. J Pediatr Orthop 1991;11:341-9.

22. Sutherland DH, Olshen RA, Biden EN, Wyatt MP. The development of mature walking. London, UK: Mac Keith Press; 1988.

39

Chapter

3

Comparison of a dynamic and a hinged ankle-foot orthosis by gait analysis in patients with hemiplegic cerebral palsy

Jacqueline Romkes & Reinald Brunner

Gait and Posture 2002;15:18-24

Chapter 3

Abstract

We studied the effect of a dynamic ankle-foot orthosis (d-AFO) on gait in 12 hemiplegic cerebral palsy patients. Sagittal plane kinematic and kinetic data of walking with the d-AFO were compared with walking barefoot, walking in a hinged ankle-foot orthosis (h-AFO) with a plantarflexion block and normal values. All patients had excessive plantarflexion and initial toe contact when walking barefoot. The d-AFO did not improve gait significantly whereas the h-AFO did. The benefits of controlling plantarflexion by a longer lever arm from the h-AFO to the proximal calf included a heel-toe gait pattern, reduced plantarflexion, increased step and stride length and reduced power absorption.

42

Introduction

A dynamic versus a hinged orthosis

The primary brain lesion in cerebral palsy (CP) and the secondary alterations in the locomotor apparatus can cause an energy inefficient gait 1 which can often be improved with an orthosis. These may improve function by helping to prevent deformity, support normal alignment and mechanics and allow a more normal range of motion. 2 An ankle-foot orthosis (AFO) is generally prescribed in patients with hemiplegic CP to prevent excessive plantarflexion that is one cause of toe walking. The AFO aims to correct the foot-shank angle in swing to improve pre- positioning of the foot at initial contact and allow a heel strike. Several designs of AFOs are available for hemiplegics. The dynamic AFO (d-AFO) has been reported as influencing abnormal joint motions through changes in the spastic reflexes and underlying muscle tone by tone-reducing features. 3,4 Another reported benefit of a d-AFO is to enable maximum midline stability and movement control while permitting freedom of movement. 4 We have used a hinged AFO (h-AFO) 8 that blocks ankle plantarflexion and allows free ankle dorsiflexion during stance for hemiplegic children. The purpose of this study was to compare the effect of the dAFO on gait in patients with spastic hemiplegic CP and to compare its performance with barefoot walking, walking in an h-AFO with normal values using sagittal plane kinematic and kinetic data.

Methods

Twelve patients (three females and nine males) diagnosed with hemiplegic CP (nine right and three left body side involvement) were enrolled in this study (Table 1). The mean age was 11.9 ± 4.9 years. All patients were community ambulators who did not use walking aids but did wear an h-AFO on the involved side. A control group of ten healthy subjects was included in the study. The group (seven females and three males) had a mean age of 26.9 ± 6.3 years, mean height of 174.3 ± 6.5 cm, and a mean weight of 69.8 ± 12.3 kg. The normal reference data were collected during barefoot walking. Two types of AFOs were used in this study. The posterior part of the h-AFO (figure 1) extended to just below the knee and its flat foot-plate extended to the tip of the toes to provide control of the ankle. The orthosis blocked ankle plantarflexion, but allowed free dorsiflexion through the hinge. The second type of

43

Chapter 3

orthosis was a d-AFO constructed according to Hylton’s method (figure 2). 4 This orthosis was trimmed just above the tip of the malleoli. The foot-plate, designed for tone reduction, included the medial longitudinal arch, the peroneal notch, the metatarsal arch and the space under and between the toes. Three orthotists were involved in the study and patients had at least 4 weeks to adapt to each type of orthosis. Each patient had their gait analysed wearing both types of orthoses and barefoot. Gait was assessed using a three-dimensional, six-camera, 50 Hz VICON 370 motion measurement system (Oxford Metrics Ltd., Oxford, UK) and two forceplates (Kistler Instrumente AG, Winterthur, Switzerland). The VICON Clinical Manager software was used for calculating and plotting temperospatial parameters, sagittal plane joint motion data, and kinetic data. This system incorporated infrared sensitive solid-state cameras for locating and tracking fixed retro-reflective markers through space. The markers were spheres (diameter

Table 1. Patient information

Patient

identification

Sex

Age (years)

Diagnosis

Contractures (°)

1

M

10.1

Hemi right Hemi left Hemi left Hemi right Hemi left Hemi right Hemi right Hemi right Hemi right Hemi right Hemi right Hemi right Right: 9, left: 3

A: 5°

2

M

9.4

3

M

14.8

K: 5°

4

M

7.3

5

F

7.3

6

M

22.8

K: 10°,A:5°

7

F

14.8

8

M

8.4

9

M

9.2

10

M

18.9

K: 5°,A:10°

11

M F M: 9, F: 3

10.7

12

9.0

Total: 12

Mean: 11.9 ± 4.9

M, male; F, female; K, knee flexion contracture; A, ankle plantarflexion contracture.

44

1
1

A dynamic versus a hinged orthosis

2
2

Figures 1 and 2. Hinged ankle-foot orthosis with blocked plantarflexion (left) and a dynamic ankle-foot orthosis (right).

25 mm) covered with retro-reflective tape affixed with double-sided tape to specific landmarks bilaterally of the subject’s legs according to the marker protocol of Davis et al. 5 Anthropometric measures of height, weight, leg length, widths of the ankles and knees, tibial torsion and femoral anteversion were taken for appropriate anthropometric scaling. A video tape recording was taken simultaneously from the front and side of the patient. The initial foot contact was classified from the video as toe, foot-flat, or heel. Patients were asked to walk at a self-selected speed along a 10 m walkway without being informed about the positions of the forceplates. Testing continued until a minimum of six trials with clear forceplate data were collected for each testing condition. Collection of data for all three conditions was taken on the same day.

Data of six trials under each condition were averaged for each patient. All data were expressed in percentage of gait cycle. The kinetic data were split up into two parts from 0 to 30 % and from 30 to 65 % of the gait cycle as reported previously by Abel et al. 6 Values were normalized to kg bodyweight and moments were expressed as internal moments. Statistical analysis was performed using a one-way

45

Chapter 3

analysis of variance (ANOVA) for repeated measurements to test the means of the conditions within the patient group. The two-tailed t-test for differences between two independent variables was used to compare the group of patients and the normal subject group. Statistical significance was set at P < 0.05. An improvement meant that the values came closer to normal reference values. Only data of the hemiparetic body side were considered.

Results

Video

The sagittal video showed that none of the patients had a heel-toe gait pattern on the hemiparetic side when walking barefoot. The h-AFO produced a heel-toe gait in all 12 patients while only four patients (numbers 4, 7, 11, 12) showed a heel-toe gait wearing the d-AFO.

Temperospatial parameters

Stride length and step length increased significantly when compared with barefoot walking when using the d-AFO and h-AFO but there was no difference between the two orthoses (table 2).

Table 2. Temperospatial parameters

 

Barefoot

d-AFO

h-AFO

Velocity (m/s) Stride length (m) Step length (m) Cadence (steps/min) Stance phase (% GC) Double support (% GC)

1.15 ± 0.26 1.13 ± 0.23 0.57 ± 0.13 118.4 ± 3.8 59.6 ± 1.9 21.7 ± 2.5

1.23 ± 0.24 1.27 ± 0.22 † 0.65 ± 0.13 † 117.5 ± 3.5 60.3 ± 2.1 23.2 ± 3.2

1.26 ± 0.25 1.31 ± 0.22 † 0.66 ± 0.12 † 120.1 ± 3.6 ‡ 60.4 ± 1.6 23.3 ± 3.3

Values are mean ± 1 SD; GC, gait cycle; d-AFO, dynamic ankle-foot orthosis; h-AFO, hinged ankle–foot orthosis; †, mean of this parameter tested significantly different from the mean of the barefoot condition (P < 0.05); ‡, mean of this parameter tested significantly different from the mean of the d-AFO condition (P < 0.05).

46

A dynamic versus a hinged orthosis

Kinematics

The joint motion of ankle, knee, hip and pelvis in the sagittal plane of one typical patient (no. 6) in this study is shown in figures 3 and 4. Figure 3 also includes the plantarflexor moment and power generation curves for this patient. Relevant mean values are summarised in tables 3-6.

Ankle

The majority of the significant differences between the two conditions is shown in table 3. Ankle plantarflexion at initial foot contact was 18.5 ± 10.3° for the group of hemiplegic CP patients and 2.3 ± 3.3° of dorsiflexion for the group of normal subjects. Although the d-AFO improved this angle significantly (8.0 ± 7.3 plantarflexion), the h-AFO (3.7 ± 4.0°dorsiflexion) produced a more normal value. Individual values of the nine patients are plotted in figure 5. The peak ankle dorsiflexion in stance was 4.9 ± 8.8° for barefoot walking. Peak dorsiflexion

Table 3. Ankle sagittal plane kinematic data

Variable

Barefoot

d-AFO

h-AFO

Normal

Initial contact

 

Dorsiflexion

-18.5 ± 10.3 *

-8.0 ± 7.3 *†

3.7 ± 4.0 †‡

2.3 ± 3.3

Stance phase

 

Peak dorsiflexion

4.9 ± 8.8 *

12.6 ± 11.0 †

16.1 ± 6.3 † 41.2 ± 11.7 †

12.9 ± 4.1 42.4 ± 7.2

%

gait cycle

33.5 ± 15.0

42.5 ± 11.9 †

Toe off

 

Plantarflexion

17.1 ± 11.3

11.4 ± 12.8

-2.7 ± 5.9 *†‡ 60.2 ± 2.1

10.9 ± 8.4 60.1 ± 1.9

%

gait cycle

59.5 ± 1.8

60.3 ± 2.2

Swing phase

 

Peak plantarflexion

26.7 ± 12.6 *

18.1 ± 12.1 †

-0.7 ± 3.5 *†‡ 69.2 ± 6.1 *

14.9 ± 7.6 64.6 ± 2.5

%

gait cycle

76.2 ± 14.6 *

67.8 ± 2.5 *

Plantar- and dorsiflexion values are mean ± 1 SD in degrees; d-AFO, dynamic ankle-foot orthosis; h-AFO, hinged ankle-foot orthosis; *, mean of this condition tested significantly different from the normal mean value; †, mean of this condition is significantly different from the barefoot condition; ‡, means of d-AFO and h-AFO conditions are significantly different (P < 0.05).

47

Chapter 3

Table 4. Knee sagittal plane kinematic data

Variable

Barefoot

d-AFO

h-AFO

Normal

Initial contact

 

Knee flexion

26.7 ± 8.6 *

23.6 ± 5.9 *

23.8 ± 6.5 *

11 ± 3.4

Stance phase

 

Peak extension

-13.7 ± 13.2

-12.9 ± 10.9

-12.9 ± 10.2 44.3 ± 3.5 *†

-6.7 ± 2.8 39.2 ± 3.9

%

gait cycle

41.0 ± 5.9

44.8 ± 1.8 *†

Swing phase

 

Peak flexion

64.5 ± 3.5

67.4 ± 4.2 *†

67.2 ± 5.1 *† 73.3 ± 3.0 ‡†

61.9 ± 1.7 73.0 ± 2.2

%

gait cycle

76.0 ± 2.4 *

75.3 ± 3.20

Values for knee flexion and extension are mean ± 1 SD in degrees; d-AFO, dynamic ankle-foot orthosis; h-AFO, hinged ankle-foot orthosis; *, mean of this condition tested significantly different from the normal mean value; †, mean of this condition is significantly different from the barefoot condition; ‡, means of d-AFO and h-AFO conditions are significantly different (P < 0.05).

increased significantly when the patients wore the d-AFO to 12.6 ± 11.0° and to 16.1 ± 6.3° with the h-AFO. Both mean values were no longer significantly different from the normal reference value of 12.9 ± 4.1°. The h-AFO changed the plantarflexion angle at toe-off to 2.7 ± 5.9° dorsiflexion, this being significantly different from the normal value (10.9 ± 8.4° plantarflexion) and the barefoot (17.1 ± 11.3°) and d-AFO (11.4 ± 12.8°) conditions. Plantarflexion angle at toe-off was normal under the barefoot and d-AFO conditions. Further on in the gait cycle, peak plantarflexion in swing phase was normal with the d-AFO (18.1 ± 12.1°) while it was higher barefoot (26.7 ± 12.6°) and lower (0.7 ± 3.5° dorsiflexion) with the h-AFO.

Knee

The kinematic data at the knee (table 4) showed a mean knee flexion angle at initial foot contact that was, for the patient group under all three conditions, significantly higher than the normal value of 11.0 ± 3.4°. In barefoot walking this value was 26.7 ± 8.6°, for the d-AFO 23.6 ± 5.9°, and for the h-AFO 23.8 ± 6.5°. The differences were not significant.

48

A dynamic versus a hinged orthosis

Although peak knee extension during stance revealed no significant differences, the timing in the gait cycle was significantly delayed for the orthoses (d-AFO 44.8 ± 1.8%, h-AFO 44.3 ± 3.5%) compared with the normal reference (39.2 ± 3.9%) and barefoot condition (41.0 ± 5.9%). Peak flexion in swing phase increased compared with the normal (61.9 ± 1.7°) and barefoot condition (64.5 ± 3.5°) when patients wore the orthoses (d-AFO:

67.4 ± 4.2°, h-AFO: 67.2 ± 5.1°).

Hip

Hip flexion at initial contact under all three conditions (barefoot: 46.3 ± 6.1°, d-AFO: 47.4 ± 5.4°, h-AFO: 47.6 ± 6.0°) was higher than the normal (39.9 ± 5.1°) (table 5).

Table 5. Hip sagittal plane kinematic data

Variable

Barefoot

d-AFO

h-AFO

Normal

Initial contact Hip flexion Stance phase Peak extension % gait cycle

46.3 ± 6.1 *

47.4 ± 5.4 *

47.6 ± 6.0 *

39.9 ± 5.1

-4.6 ± 8.4 * 52.3 ± 2.2

2.9 ± 7.2 52.5 ± 1.5

-2.1 ± 6.8 53.2 ± 2.0

2.8 ± 5.8 51.4 ± 2.3

Hip flexion and extension values are mean ± 1 SD in degrees; d-AFO, dynamic ankle-foot orthosis; h-AFO, hinged ankle-foot orthosis; * mean of this condition tested significantly different from the normal mean (P < 0.05).

Kinetic parameters

The patients showed a high peak internal plantarflexor moment under all three conditions in the first 30 % of the gait cycle (table 6). This peak had a maximum value of 1.12 ± 0.29 Nm/kg, with d-AFOs 1.08 ± 0.26 Nm/kg, and with h-AFOs 1.16 ± 0.33 Nm/kg (no significant differences between conditions). Such an internal plantarflexor moment during first half of stance phase was absent in the normal subjects and values were therefore only compared within the patient group. Peak ankle power absorption in this part of the gait cycle was significantly higher than the normal reference of 0.49 ± 0.21 W/kg. In the patients, peak values reached

49

Chapter 3

2.31 ± 0.74 W/kg while walking barefoot. This value was significantly reduced to 1.78 ± 0.72 W/kg with the d-AFO and 1.50 ± 0.84 W/kg with the h-AFO, the values between d-AFO and h-AFO not being significant. In the second half of stance (30-65 % gait cycle), the peak internal plantarflexor moment was significantly lower for the patients compared with the normal values of 1.37 ± 0.14 Nm/kg. For the patients this value was 0.82 ± 0.18 Nm/kg in barefoot walking, which increased significantly to 0.92 ± 0.30 Nm/kg with the d-AFO and 1.08 ± 0.22 Nm/kg with the h-AFO. The patients generated significantly less ankle power during this period than the normal range of 2.29 ± 0.66 W/kg. For barefoot walking this value was 0.91 ± 0.36 W/kg and did not change when wearing either the d-AFO (0.89 ± 0.23 W/kg) or the h-AFO (0.87 ± 0.42 W/kg).

Table 6. Kinetic data

Moments (Nm/kg) and powers (W/kg)

Barefoot

d-AFO

h-AFO

Normal

1.12

± 0.29

1.08 ± 0.26

1.16 ± 0.33

NR

0.82

± 0.18*

0.92 ± 0.30*†

1.08 ± 0.22*†‡

1.37 ± 0.14

2.31

± 0.74*

1.78 ± 0.72*†

1.50 ± 0.84*†

0.49 ± 0.21

0.91

± 0.36*

0.89 ± 0.23*

0.87 ± 0.42*

2.29 ± 0.66

Ankle plantarflexor moment: 0-30 % GC Ankle plantarflexor moment: 30-65 % GC Ankle power absorption: 0-30 % GC Ankle power generation: 30-65 % GC

Values (mean ± 1 SD) are normalized to body weight and the moments are internal moments; GC, gait cycle; d-AFO, dynamic ankle-foot orthosis; h-AFO, hinged ankle-foot orthosis. * mean of this condition tested significantly different from the normal mean value; † mean of this condition is significantly different from the barefoot condition; ‡ means of d-AFO and h-AFO conditions are significantly different (P < 0.05).

Discussion

The purpose of this study was to evaluate the effect of the two types of AFOs on gait function in patients with hemiplegic CP. We found that the h-AFO changed toe

50

A dynamic versus a hinged orthosis

walking on the hemiparetic side into a heel-toe gait pattern in all patients. The h-AFO successfully controlled excessive ankle plantarflexion in swing phase that correctly pre-positioned the foot for an initial heel contact. The d-AFO is claimed to improve midline biomechanical stability, enhance deep sensory information and through both of these avenues, greater movement options but in our study did not improve gait function as effective as the h-AFO. In normal gait the ground reaction force (GRF) travels from underneath the heel along the plantar surface of the foot to the toes. This produces an increasing internal plantarflexor moment during stance phase as the distance from the GRF vector to ankle joint centre increases. A power generation burst is seen in late stance phase where ankle plantarflexion motion (angular velocity) is high just before toe-off. In hemiplegics, spasticity, abnormal muscle activation patterns, and muscle tightness disrupt the normal interaction of the ankle and foot with the supporting surface. 6 These patients have a toe strike that is reflected by a high plantarflexion moment in early stance. 6,7 With weight acceptance, the ankle initially shows a dorsiflexion movement that is interrupted by premature active plantarflexion as reflected by the power absorption spike in early stance. Plantarflexion persists throughout the remainder of stance. The power generation peak in late stance is reduced as shown by a decreased ankle movement and reduced ankle plantarflexor moment. 6 The AFO is commonly prescribed for hemiplegics to control the excessive plantarflexion to improve gait. This study showed that the h-AFO was superior to the d-AFO for gait correction. The h-AFO reduced ankle plantarflexion angle at initial foot contact to the normal reference value thus producing a physiological heel contact in all patients. This in turn produced more physiological forces that improved the foot shank position throughout the gait cycle. These positive changes increased the stride and step length. Increased dorsiflexion by the h-AFO in particular did not lead to an increase in knee flexion as might have been anticipated from the two joint action of gastrocnemius. The internal ankle plantarflexor moment in early stance remained unchanged and the restricted ankle motion reduced the power absorption burst and could not be improved because of the loss of motion. Abel et al. 6 found similar results when investigating the effects of AFOs on equinus in spastic diplegia. While the positive effect of the h-AFO on gait has been shown previously 8 ,

51

Chapter 3

Chapter 3 Figure 3. Ankle kinematic and kinetic data curves (dorsi- (+) and plantarflexion ( )

Figure 3. Ankle kinematic and kinetic data curves (dorsi- (+) and plantarflexion ( ) motion; ankle plantar- (+) and dorsiflexor ( ) moment; ankle joint power generation (+) and absorption ( )) for a typical patient (no. 6) in this study. The curves are mean values ± 1 SD taken from six trials in each condition. Curves are normalized to 100% gait cycle. Stance is separated from swing by the vertical line.

52

A dynamic versus a hinged orthosis

A dynamic versus a hinged orthosis Figure 4. Sagittal plane motion of pelvis, hip and knee

Figure 4. Sagittal plane motion of pelvis, hip and knee for a typical patient (no. 6) in the study population. The curves are mean values 1 ± SD taken from six trials under each condition. Curves are normalized to 100% gait cycle. Stance is separated from swing by the vertical line.

53

Chapter 3

Chapter 3 Figure 5. Ankle dorsi- (+) plantarflexion ( ) angles at first foot contact under

Figure 5. Ankle dorsi- (+) plantarflexion ( ) angles at first foot contact under the three conditions of walking (barefoot, dynamic ankle-foot orthosis (d-AFO), hinged ankle-foot orthosis (h-AFO) for the 12 patients. The solid horizontal line and the two dotted horizontal lines are the normal mean ± 1 SD reference values.

the main interest was whether the d-AFO would behave similarly. The differences between the two types were the tone-reducing foot-plate and lower posterior shell

of the d-AFO. In a third of the patients a heel-toe gait pattern was achieved with

the d-AFO and no further improvements were seen when they wore the h-AFO. However, the majority of patients did not benefit as much from the d-AFO. The tone-reducing foot-plate which has been reported as applying deep pressure and precise touch to increase sensory awareness, in combination with the reduced

height of the posterior shell failed to control excessive plantarflexion during swing.

A similar study by Crenshaw et al. 3 came to the same conclusion, i.e. that the

positive benefits of an AFO could be attributed directly to the physical restriction

of ankle. The reduced lever-arm of the d-AFO could not be compensated by the

tone-reducing foot-plate. While some patients achieved active control on dorsi- plantarflexion at the ankle when wearing the d-AFO, the majority still benefited from an external control of the ankle motion by the long lever arm of the h-AFO. The comparison of kinematic and kinetic data of this group of patients, mainly children, to normal reference data may be one cause of error. A study by Õunpuu et

al. 9 investigating 31 normal children by gait analysis showed that data were similar

to that of normal adults and to adult data published by Kadaba et al. 10 and Bresler

54

A dynamic versus a hinged orthosis

and Frankel. 11 Sutherland et al. 12 concluded that a mature gait pattern is established by 7 years of age but time-distance variables vary with age and stature. Therefore, kinematic and kinetic data of the patient group were compared with normal adult reference data whereas the spatial parameters were not. A further possible source of error in this study may arise from the assumption that the motion of the skin covering the bones is identical to the motion of the bone. 13 An even greater error, however, results from placement of markers on the orthoses and shoes because the anatomical landmarks must be estimated. Nevertheless we conclude that the d- AFO was not as effective as an h-AFO in improving gait in hemiplegic patients.

Acknowledgements

This project was partially supported by the ‘Stiftung für Bewegungsstörungen’. The authors wish to express their thanks to the three orthotists Thomas Ruepp, Thomas Glauser and Andreas Reinhard and to the physical therapists of practice Jordi & Team for their enthusiastic assistance with data collection and processing.

References

1. Gage JR. Gait Analysis in Cerebral Palsy. London, UK: Mac Keith Press, 1991.

2. Knutson LM, Clark DE. Orthotic devices for ambulation in children with cerebral palsy and myelomeningocele. Phys Ther 1991;71:947-60.

3. Crenshaw S, Herzog R, Castagno P, Richards J, Miller F, Michaloski G, Moran E. The efficacy of tone-reducing features in orthotics on the gait of children with spastic diplegic cerebral palsy. J Pediatr Orthop 2000;20:210-6.

4. Hylton N. Dynamic Orthotic Concepts: Background and Experiences. Dortmund, Germany:

Verlag Orthopaedie-Technik, 2000.

5. Davis RB III, Ounpuu S, Tyburski D, Gage JR. A gait analysis data collection and reduction technique. Hum Move Sci 1991;10:575-87.

6. Abel MF, Juhl GA, Vaughan CL, Damiano DL. Gait assessment of fixed ankle-foot orthoses in children with spastic diplegia. Arch Phys Med Rehabil 1998;79:126-33.

7. Perry J. Gait Analysis: Normal and Pathological Function. Thorofare (NJ): Slack Inc, 1992.

8. Brunner R, Meier G, Ruepp T. Comparison of a stiff and a spring-type ankle-foot orthosis to improve gait in spastic hemiplegic children. J Pediatr Orthop 1998;18:719-26.

9. Õunpuu S, Gage JR, Davis RB. Three-dimensional lower extremity joint kinetics in normal pediatric gait. J Pediatr Orthop 1991;11:341-9.

55

Chapter 3

10. Kadaba MP, Ramakrishnan HK, Wooten ME. Measurement of lower extremity kinematics during level walking. J Orthop Res 1990;7:849-60.

11. Bresler B, Frankel JP. The forces and moments in the leg during level walking. Trans ASME

1950;72:27-36.

12. Sutherland DH, Olshen RA, Biden EN, Wyatt MP. The Development of Mature Walking. London, UK: Mac Keith Press, 1988.

13. Lafortune MA, Cavanagh PR, Sommer III HJ, Kalenak A. Three-dimensional kinematics of the human knee during walking. J Biomech 1992;25:347-57.

56

Chapter

4

Changes in muscle activity in children with hemiplegic cerebral palsy while walking with and without ankle-foot orthoses

Jacqueline Romkes, Anna K. Hell & Reinald Brunner

Gait and Posture 2006;24:467-474

Chapter 4

Abstract

We compared the electromyographic (EMG) signals of lower extremity muscle groups in 10 children with hemiplegic cerebral palsy (CP) while walking barefoot and in a hinged ankle-foot orthosis (HAFO). All children had excessive plantarflexion and initial toe-contact on the affected side when walking barefoot, a typical gait pattern for hemiplegic patients. The patients walked with a physiological heel-toe gait pattern when wearing the HAFO. The peak activity of the tibialis anterior muscle was reduced by 36.1 % at initial contact and loading response phase and by 57.3 % just after toe-off when using a HAFO. The decrease in activity was thought to result from the change in gait pattern from a toe-gait to a heel-toe gait as well as the use of a HAFO. The HAFO also slightly decreased muscle activity in the proximal leg muscles mainly during swing phase, improved stride length, decreased cadence, improved walking speed, increased peak hip flexion, improved kinematics in loading response phase at the knee, and reduced the excessive ankle plantarflexion.

58

Introduction

Muscle activity with ankle foot orthoses

A

typical finding in the gait pattern of patients with hemiplegic cerebral palsy (CP)

is

toe walking on the affected side. Previous studies involving three-dimensional

gait analysis 1-3 showed that an ankle-foot orthosis (AFO) could successfully correct the excessive plantarflexion angle and thereby improve the children’s gait pattern. Data on changes in muscular function, apart from changes in joint angles, joint forces, and moments or energy consumption are required to investigate the effects

of AFOs on gait.

A widely recognized classification system for hemiplegia based on joint kinematic data was proposed by Winters et al. 4,5 and classifies the patterns into four

different types. A principle finding in type I (the mildest form) is equinus in swing secondary to a relative over activity of the triceps surae compared to the anterior tibial musculature. The main feature of type II is ankle plantarflexion throughout the gait cycle, in both swing and stance from either static or dynamic contracture of the triceps surae. The knee is often forced into slight hyperextension in middle or late stance. Patients with type III show, in addition to the gait anomalies of type II, limited knee motion in the swing phase of gait. Patients can either show reduced knee extension during stance or hyperextension in stance and reduced flexion in swing. The patients in group IV show all features of type III with involvement of the hip musculature. Equinus may be present in swing or stance phase. Anterior tibialis weakness, dysfunction, or premature gastrocsoleus contraction causes swing phase drop foot with resultant dragging of the foot, tripping, and proximal joint kinematic compensations. 6 Compensatory mechanisms, however, can cause secondary problems and alter the gait pattern further. Children with hemiplegia can benefit from the use of an AFO to improve their gait pattern. If an orthosis has been correctly designed, constructed and applied to the patient, it may act to resist the abnormal pattern of motion and hence restore more normal function. 7 The use of AFOs has remained a standard treatment because of their usefulness in preventing equinus contractures. 8 Type I hemiplegic patients require an AFO primarily to assist the dorsiflexors in preventing foot drop during swing. To improve gait, the AFO aims at correcting the foot-shank angle in swing phase to improve pre-positioning of the foot at initial contact and allow heel- strike. 3 However, those with type II hemiplegia require an AFO capable of

59

Chapter 4

controlling the equinus through stance and swing. The most common orthotic solution is the solid AFO. The effect of the design can be optimized by the addition of ankle articulations permitting free dorsiflexion, i.e. a hinged AFO (HAFO), and by the addition of some shoe adaptations. 2,7 Dynamic electromyography (EMG) provides valuable information with respect to muscle activity. Surface electrodes applied to the skin overlying the muscles provide an indirect measure of muscle-generated potentials. Errors in neurological control, muscle weakness, voluntary substitutions and obligatory posturing lead to abnormal EMG during walking. 1 Investigation of EMG data on the improvement of gait function using AFOs has been limited. Radtka et al. 9 compared the effects of two types of AFOs on the gait of 10 children with CP four of whom were hemiplegic. The orthoses increased stride length, decreased cadence, and decreased excessive ankle plantarflexion. No changes in lower extremity muscle timing were found compared to barefoot walking. The study of Radtka et al. 9 looked at the timing of muscle activity and did not investigate change in level of activity. Muscle recruitment of spastic muscles in children with CP is characterized by disordered onsets and offsets of muscle contraction. Loss of selective control may result in abnormal co-activation of agonist-antagonist muscles and the presence of co-contraction due to poor selective control may be the reason for the disordered timing of muscle activity. 7,10,11 Knowledge of the change in activation of different muscle groups should help to understand further the function of AFOs. The purpose of this study was to detect changes in EMG activity of the tibialis anterior muscle in patients with hemiplegic CP when walking with and without a HAFO. We wished to determine if the pattern of muscle activation changed when the movement pattern at the ankle joint level was changed by a HAFO. We also considered the influence of a HAFO on selected upper leg muscles.

Methods

Patients

The study group consisted of 10 children (four girls and six boys) with hemiplegic CP (six right sided, four left sided, mean age 9.7 ± 1.6 years, weight 30.2 ± 6.2 kg, and height 135.2 ± 7.9 cm). None had had orthopaedic surgery. The passive range of ankle motion on the hemiparetic side in direction dorsiflexion ranged from 0° to 20° with the knee flexed and from -10° to 10° with the knee extended. Spasticity

60

Muscle activity with ankle foot orthoses

was evaluated according to the modified Ashworth scale that ranged between a score of 2 and 3 for the plantarflexors. 12 In two children the Ashworth scale was 2 for knee extensors and knee flexors; all others were scored as 1. Ashworth scores at the hip level were 1 for all children. Clinical muscle tests showed reduced strength against manual resistance in all patients for both plantarflexor and dorsiflexor muscles. According to the Winters et al. 5 classification, eight children were categorized as type I and two as type II. On the affected side, all children demonstrated excessive ankle plantarflexion during gait and initial foot contact with the toes. Only children whose gait pattern changed from initial toe-contact while walking barefoot to initial heel-contact while walking with a HAFO were included in this study. All children could walk independently without walking aids but wore a HAFO on the involved body side on average for 8.1 ± 2.7 h/day. A typical HAFO worn by the children is shown in figure 1. The posterior part of the HAFO extended to just below the knee and the footplate reached to the tip of the toes to provide control of the ankle. The orthosis blocked ankle plantarflexion, but allowed at least 20° dorsiflexion through a hinge. Generally, the HAFOs were fitted by the orthotist so that the tibial axis was positioned perpendicular to the

so that the tibial axis was positioned perpendicular to the Figure 1. Hinged ankle-foot orthosis with

Figure 1. Hinged ankle-foot orthosis with blocked plantarflexion.

61

Chapter 4

floor surface to avoid a forward or backward lean. A control group of 10 healthy children (six girls, four boys mean age 8.7 ± 0.8 years, mean height 133.4 ± 7.7 cm, and mean weight 28.4 ± 6.6 kg) provided comparative kinematic data. The patients and controls did not differ statistically in age, height and weight. The normal reference data were collected during barefoot walking.

Study protocol

The study protocol consisted of 3D gait analysis and lower extremity muscle activity measured under two conditions, barefoot and wearing a HAFO with shoe wear. All subjects walked at a self-selected walking speed. Testing continued until a minimum of six trials with clear data sets had been collected under each test condition.

Gait analysis

3D gait analysis was performed using a six-camera, 50 Hz movement analysis system (VICON 370, Oxford Metrics Ltd., UK). Fifteen 25 mm diameter retro- reflective markers were affixed to the subjects according to the marker protocol of Davis et al. 13 (figure 2). When walking with the HAFO and shoes, the heel and toe markers were placed on the shoes at the positions best projecting the anatomical landmarks. All other markers remained at the same positions throughout the testing protocol. Height, weight, leg length, widths of the ankles and knees, and tibial torsion were measured clinically for appropriate anthropometric scaling. Joint angle data were expressed as percentage of gait cycle using Polygon software (Oxford Metrics Ltd.). Specific points and values selected from the movement curves were compared between conditions.

EMG measurements

Surface EMG was recorded simultaneously with the 3D gait analysis. Bipolar Ag/AgCl surface electrode pairs with an electrode diameter of 10 mm and an inter electrode spacing of 22 ± 1 mm were placed on the clean shaven skin overlying the vastus medialis, vastus lateralis, long head of the biceps femoris, semitendinosus, and tibialis anterior muscles of the hemiparetic leg. The SENIAM 14 guidelines for electrode placements in surface EMG were followed. The reference ground electrode was placed overlying the tuberosity of the tibia. All electrodes were left in position between conditions and were connected to single differential amplifiers with a band path of 10-700 Hz (Biovision AG, Wehrheim, Germany). The data

62

Muscle activity with ankle foot orthoses

Muscle activity with ankle foot orthoses Figure 2. Patient ready for data collection. were collected at

Figure 2. Patient ready for data collection.

were collected at a sampling frequency of 2500 Hz using a Zebris system (Zebris, Tübingen, Germany). EMG signals were visually inspected for artefacts and noise. Initial foot contact and toe-off were determined using a force plate (Kistler Instrumente AG, Winterthur, Switzerland) embedded in the floor. The following ipsilateral initial foot contact was determined with a small 1-directional accelerometer (Biovision, Wehrheim, Germany) attached to the heel placed vertically. Force plate and accelerometer data were sampled at the same frequency as the EMG data. Determining the second foot strike for EMG analysis with an accelerometer proved to be more precise than using the 50 Hz kinematic data and equally good as using a force plate. The electromyographic data of six trials were first resampled according to the

63

Chapter 4

trial with the shortest data length, then full wave rectified for each patient. The data were then time normalized by dividing data for 1 gait cycle into 16 equally spaced intervals ( 1– 16). Root mean square values for each muscle signal were calculated for each of these time intervals. The maximum value of the EMG was calculated for all trials of each patient when walking barefoot and all other EMG data were expressed as percentage of the average maximum value over the six trials. The MATLAB software package (The MathWorks Inc.) was used.

Data analysis

Data of six trials under each condition were averaged for each patient. Paired t-tests were applied to the kinematic parameters and EMG data of the impaired body side to determine changes between the two conditions during walking barefoot and with a HAFO. The level of significance was set at P < 0.05.

Results

Temporal-spatial parameters

Table 1 shows the mean values of the temporal-spatial outcome measures for the patients walking barefoot and walking with a HAFO. With a HAFO, patients walked significantly faster by increasing their step and stride length. Cadence was decreased. Stance phase duration, single support time and double support time did not show any significant alterations.

Joint angles

Figure 3 shows the mean sagittal plane kinematic data for walking with and without a hinged AFO compared to data from healthy children. All patients had an initial toe contact when walking barefoot due to excessive ankle plantarflexion (15.0 ± 9.7° at initial contact) followed by a movement in the direction of dorsiflexion. When wearing a HAFO, this angle was within the normal range (-0.1 ± 6.3°) and a physiological heel strike was followed by a movement in the direction of plantarflexion. Patients showed a significantly increased knee flexion position (22.8 ± 8.9°) at initial foot contact compared to the control group(10.0 ± 3.3°) that did not alter with a HAFO (21.0 ± 7.3°). Although the increased angle at initial contact could not be reduced with the HAFO, the loading response phase improved by becoming more dynamic and a flexion-extension pattern could be observed. A significant increase in hip flexion at initial contact

64

Table 1. Temporal-spatial parameters

Muscle activity with ankle foot orthoses

Mean ± 1 SD

Patients barefoot

Patients hinged AFO

P value*

Walking speed (m/sec)

1.05 ± 0.18

1.14 ± 0.14

0.018

Cadence (steps/min)

119.4 ± 7.5

112.5 ± 7.3

0.006

Stride length (m)

1.06 ± 0.15

1.22 ± 0.11

< 0.001

Step length (m)

0.52 ± 0.07

0.61 ± 0.06

< 0.001

Stance (%)

57.5 ± 2.2

57.6 ± 1.9

N.S.

Single limb support (%)

38.8 ± 3.0

38.4 ± 4.2

N.S.

Double limb support (%)

17.7 ± 2.8

18.2 ± 3.1

N.S.

* paired student t-test with P < 0.05; N.S. = not significant; Values are for the hemiparetic leg

was seen in the patient group when wearing a HAFO (barefoot: 44.8 ± 6.7°; HAFO: 48.1 ± 5.4°). However, both values were not significantly different from the control group (44.5 ± 4.5°). There were no significant differences at the level of the pelvis or in other selected kinematic parameters.

EMG data

Data of the muscle activities are shown in figure 4. Most pronounced changes were noted in the tibialis anterior muscle. Compared to barefoot walking, the hemiplegic children showed an average decrease in peak tibialis anterior muscle activity of 36.1 ± 20.6 % at the first 12.5 % (average of 1 and 2) of the gait cycle, i.e. initial contact and loading response phase. Individual EMG profiles of the tibialis anterior muscle for all patients are shown in figure 5. Nine out of the 10 patients showed a reduction of activity varying between 13.6 % and 57.5 % whereas one subject stayed equal (reduction of 0.7 %). Peak swing phase activity showed, on average, a reduction of 57.3 ± 20.2 % when walking with the HAFO. In all the patients activity was reduced varying from 30.1 % to 87.5 %. Slight significant changes during swing phase could be detected in the semitendinosis, biceps femoris, vastus medialis and lateralis muscles. On all occasions the average activity over the whole group of patients was reduced when walking with an HAFO compared to barefoot walking. Results were consistent within the group. A reduction of EMG was also found for the rectus femoris during swing phase.

65

Chapter 4

Chapter 4 Figure 3. Kinematic curves of pelvis, hip, knee, and ankle for the sagittal plane.

Figure 3. Kinematic curves of pelvis, hip, knee, and ankle for the sagittal plane. Curves are averages with 1 SD for 10 patients. The grey area is the mean with 1 SD of a control group of nine healthy children. The vertical lines represent toe-off.

66

Muscle activity with ankle foot orthoses

Discussion

The main aim of this study was to investigate tibialis anterior muscle activity when patients with hemiplegic CP walked with and without a HAFO. Results showed that a change of movement pattern (toe-gait to heel-toe gait) when wearing a HAFO resulted in a decrease of peak activity of the tibialis anterior muscle at initial stance, loading response, and initial swing phase. In healthy persons, the tibialis anterior muscle shows a large burst of EMG activity from initial contact to the end of loading response phase which controls foot placement in gait. At initial contact, the ankle is in a neutral position and the forefoot is tilted upward. The tibialis anterior provides support of the foot by a dorsiflexion pull. 15 During loading response the tibialis anterior, together with the long toe extensors, decelerates the rate of ankle plantar flexion as the foot drops to the floor. 15 A second burst of activity is seen starting in pre-swing lasting for the whole swing phase. In pre- swing the ankle rapidly plantar flexes and the onset of tibialis anterior helps to decelerate the rate of foot fall. 15 Peak activity is reached in initial swing phase just after toe off. Ankle motion is now changed to dorsiflexion towards neutral position to allow floor clearance by the toes. The children with hemiplegic CP walking barefoot with an initial toe-contact on the affected side, showed, similar to healthy persons, a first burst of electromyographic activity at initial foot contact and loading response and a second burst with peak activity after toe-off during swing phase. However, in the hemiplegic CP patients this first burst of activity at initial stance phase was lower compared to the second burst. While making initial contact with the toes instead of the heel, a greater internal plantarflexor moment is generated around the ankle joint 3 and the role of tibialis anterior activity in this phase can be associated with stabilisation and control of the ankle position. The second burst of activity shows normal function at pre-swing but activity diminishes in mid-swing in contrast to healthy subjects, in which this second burst is followed by further EMG activity to secure foot clearance during the rest of swing phase. Thus, no activity of the tibialis anterior muscle is present during the end of swing phase. With a HAFO, the children achieved a physiological heel-toe gait pattern. Peak tibialis anterior muscle activity decreased by 36.1 % at initial foot contact and by 57.3 % just after toe-off. The HAFO controls the foot position relative to the shank and prevents a foot-drop and the dorsiflexors are no longer required as their function has been replaced by the orthosis. This is in agreement with a study by Geboers et al. 16 who presented data

67

Chapter 4

Chapter 4 Figure 4. Average electromyographic data for 10 subject s (mean ± 1 SD) with

Figure 4. Average electromyographic data for 10 subjects (mean ± 1 SD) with and without walking with a hinged ankle-foot orthosis. The vertical lines represent the moment of toe-off. The grey boxes on the horizontal axis represent the areas of significance (paired t-test with P < 0.05).

of healthy adult subjects wearing a stiff AFO when walking on a treadmill. They reported that the first burst of tibialis anterior muscle activity, i.e. first 15 % of the gait cycle, decreased by 20 % when walking with an AFO. Activity of the tibialis anterior muscle did not change during the gait cycle as a whole. However, in this study 16 the healthy subjects walked in a heel-toe gait pattern both with and without

68

Muscle activity with ankle foot orthoses

an AFO and the AFO did not have a hinge to allow ankle dorsiflexion. This study 16 showed that the use of an AFO immediately reduced muscle activity of the ankle dorsiflexors in healthy people. Patients in the present study and normal individuals seem to reduce muscle activity when this is no longer necessary. If healthy individuals adopt a different gait pattern by voluntary toe walking, EMG will change accordingly. Davids et al. 17 compared kinematic, kinetic, and EMG data of healthy children during voluntarily toe walking to children with CP walking in an obligatory toe walking gait pattern. Significant deviations common to both normal and CP toe walking groups were determined to be due, at least in part, to the biomechanical constraints associated with a toe walking gait pattern. The reduction of tibialis anterior muscle activity in this study could be partly related to the change of movement pattern, i.e. toe walking to heel-toe gait pattern and partly to the use of a HAFO in that the HAFO prevents plantarflexion and hence reduces the need to dorsiflex the foot. Orthoses may be prescribed to prevent a deformity that is anticipated or might increase without treatment 8 and to allow improved gait function. All children in this study were experienced users, having worn a HAFO for at least several years. None had fixed contractures of the Achilles tendon and calf muscles which might result from wearing a HAFO. While the aims of an AFO in children with hemiplegic CP are well defined, their effectiveness in the long term is less certain. It is possible that a decrease in muscle activity of the dorsiflexors with the prolonged use of an AFO could lead to loss of strength and have a negative effect. In this group of hemiplegic children, the benefits of achieving: (1) a more normal gait pattern (i.e. increased walking speed and step and stride length and improved kinematics) and (2) preventing secondary compensation mechanisms and equinus contracture, were more important than a potential loss of muscle strength of their tibialis anterior muscle. In the present study, kinematic changes observed at the ankle joint were as expected and in agreement with a previous study comparing the same type of orthosis in a similar group of patients. 3 Other sagittal plane kinematic data showed a significant increase in hip flexion at initial contact when walking with a HAFO. Although within group values were significant, these were not significantly different to normal values. At the level of the knee, loading response phase improved with a HAFO and showed a more dynamic movement pattern. These findings can be the result of increased walking speed as reported by

69

Chapter 4

Chapter 4 Figure 5. Individual EMG profiles of the tibialis anterior muscle of 10 patients walking

Figure 5. Individual EMG profiles of the tibialis anterior muscle of 10 patients walking with and without a hinged ankle-foot orthosis (AFO). Each trace is an average of six trials for one patient.

Van der Linden et al. 18 With increasing walking speed, peak hip flexion and extension increased over the gait cycle in a group of healthy children. In the present study, a higher walking speed was observed when walking with a HAFO due to larger steps, which can explain the increased peak hip flexion in early stance. Changes in knee dynamics in loading response can also be influenced by the increased walking speed. 18 Although small significant changes were seen as a reduction of activity in upper leg muscle groups, these did not translate into relevant functional effects. A limitation to this study is that the three markers representing the foot had to be placed on the shoes but these were placed as correctly as possible to represent the long axis of the foot. The associated measurement errors are likely to be greater compared to those from skin markers placement. EMG data of the gastrocnemius and soleus muscles was not obtained and might have given some further insights.

Conclusion

This study showed that the use of a HAFO reduced activity of the tibialis anterior muscle during walking at initial contact, loading response, pre-swing and initial swing phase. Wearing the HAFO also improved stride length, decreased cadence, improved walking speed, increased hip flexion at initial contact, improved knee kinematics at loading response phase, and reduced the excessive ankle plantarflexion at initial contact and during swing phase. The decrease in activity of

70

Muscle activity with ankle foot orthoses

the tibialis anterior muscle is thought to result from the change in gait pattern from a toe-gait to a heel–toe gait pattern as well as the use of a HAFO itself which substitutes for the tibialis anterior muscle.

Acknowledgement

The authors like to acknowledge the Foundation for Movement Disorders (Stiftung für Bewegungsstörungen) for their financial support of this study.

References

1. Abel MF, Juhl GA, Vaughan CL, Damiano DL. Gait assessment of fixed ankle-foot orthoses in children with spastic diplegia. Arch Phys Med Rehabil 1998;79:126-33.

2. Brunner R, Meier G, Ruepp T. Comparison of a stiff and a spring-type ankle–foot orthosis to improve gait in spastic hemiplegic children. J Pediatr Orthop 1998;18:719-26.

3. Romkes J, Brunner R. Comparison of a dynamic and a hinged ankle–foot orthosis by gait analysis in patients with hemiplegic cerebral palsy. Gait Posture 2002;15:18-24.

4. Gage JR. Gait analysis in cerebral palsy. London: Mac Keith Press; 1991.

5. Winters TF, Gage JR, Hicks R. Gait patterns in spastic hemiplegia in children and young adults.

J Bone Joint Surg [Am] 1987;69A:437-41.

6. Aiona MD, Sussman MD. Treatment of spastic diplegia in patients with cerebral palsy: part II.

J Pediatr Orthop B 2004;13:S13-38.

7. Condie DN. Orthotic function: its biomechanical basis. Orthot Prosthet World 2000;18–22.

8. Knutson LM, Clark DE. Orthotic devices for ambulation in children with cerebral palsy and Myelomeningocele. Phys Ther 1991;71:947-60.

9. Radtka SA, Skinner SR, Dixon DM, Johanson ME. A comparison of gait with solid, dynamic, and no ankle–foot orthoses in children with spastic cerebral palsy. Phys Ther 1997;77:395-409.

10. Damiano D. Reviewing muscle cocontraction: is it a developmental, pathological or motor control issue? Phys Occup Ther Pediatr 1993;12:213-35.

11. Crenna P, Inverno M. Objective detection of pathophysiological factors contributing to gait disturbance in supraspinal lesions. In: Fedrizzi E, Avanzini G, Crenna P, editors. Motor development in children. London: John Libbey; 1994. p. 103-18.

12. Bohannon RW, Smith MB. Interrater r