- Email: [email protected]
Comparison of a dynamic and a hinged ankle–foot orthosis by gait analysis in patients with hemiplegic cerebral palsy
Gait and Posture 15 (2002) 18 – 24 www.elsevier.com/locate/gaitpost
Comparison of a dynamic and a hinged ankle–foot orthosis by gait analysis in patients with hemiplegic cerebral palsy Jacqueline Romkes *, Reinald Brunner Laboratory for Gait Analysis, Children’s Uni6ersity Hospital (UKBB), Burgfelderstrasse 101, CH-4055 Basel, Switzerland Received 30 January 2001; received in revised form 30 June 2001; accepted 28 August 2001
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 fror the h-AFO to the proximal calf included a heel– toe gait pattern, reduced plantarflexion, increased step and stride length and reduced power absorption. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Gait analysis; Cerebral palsy; Ankle–foot orthosis; AFO; Equinus
1. Introduction 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 AFO’s 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
* Corresponding author. Tel.: + 41-61-685-5345; fax: +41-61-6855012. E-mail address: j – [email protected] (J. Romkes).
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 d-AFO 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.
2. 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 11.99 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 26.99 6.3 years, mean height of 174.39 6.5 cm, and a mean weight of 69.89 12.3 kg. The normal reference data were collected during barefoot walking.
0966-6362/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 9 6 6 - 6 3 6 2 ( 0 1 ) 0 0 1 7 8 - 3
J. Romkes, R. Brunner / Gait and Posture 15 (2002) 18–24
19
Table 1 Patient information Patient identification
Sex
Age (years)
Diagnosis
Contractures (°)
1 2 3 4 5 6 7 8 9 10 11 12 Total: 12
M M M M F M F M M M M F M: 9, F: 3
10.1 9.4 14.8 7.3 7.3 22.8 14.8 8.4 9.2 18.9 10.7 9.0 Mean: 11.9 94.9
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: – K: – – K: – – – K: – –
5° 5°
10°, A: 5°
5°, A: 10°
M, male; F, female; K, knee flexion contracture; A, ankle plantarflexion contracture.
Two types of AFOs were used in this study. The posterior part of the h-AFO (Fig. 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 orthosis was a d-AFO constructed according to Hylton’s method [4] (Fig. 2). 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 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
Fig. 1. Hinged ankle – foot orthosis with blocked plantarflexion.
20
J. Romkes, R. Brunner / Gait and Posture 15 (2002) 18–24
Fig. 2. Dynamic ankle – foot orthosis.
cycle as reported previously by Abel et al. [6]. Values were normalised to kg bodyweight and moments were expressed as internal moments. Statistical analysis was performed using a one-way analysis of variance (ANOVA) for repeated measurements to test the means of the conditions within the patient group. The twotailed 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 B 0.05. An improvement meant that the values came closer to normal reference values. Only data of the hemiparetic body side were considered.
3.3. 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 Figs. 3 and 4. Fig. 3 also includes the plantarflexor moment and power generation curves for this patient. Relevant mean values are summarised in Tables 3–6.
3.4. Ankle The majority of the significant differences between the two conditions is shown in Table 3. Ankle plantarflexion at initial foot contact was 18.5910.3°
3. Results
3.1. 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.
Table 2 Temperospatial parameters Barefoot
d-AFO
h-AFO
1.15 90.26 1.13 90.23 0.57 90.13 118.4 9 3.8
1.23 9 0.24 1.27 9 0.22 † 0.65 9 0.13 † 117.5 9 3.5
1.26 9 0.25 1.319 0.22† 0.669 0.12† 120.1 9 3.6‡
3.2. Temperospatial parameters
Velocity (m/s) Stride length (m) Step length (m) Cadence (steps per min) Stance phase (% GC) Double support (% GC)
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).
Values are mean 9 1 S.D. 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 (PB0.05); ‡, mean of this parameter tested significantly different from the mean of the d-AFO condition (PB0.05).
59.6 9 1.9
60.3 9 2.1
60.4 9 1.6
21.7 92.5
23.2 9 3.2
23.3 9 3.3
J. Romkes, R. Brunner / Gait and Posture 15 (2002) 18–24
21
values were no longer significantly different from the normal reference value of 12.99 4.1°. The h-AFO changed the plantarflexion angle at toeoff to 2.79 5.9° dorsiflexion, this being significantly different from the normal value (10.998.4° plantarflexion) and the barefoot (17.19 11.3°) and d-AFO (11.49 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 (18.1912.1°) while it was higher barefoot (26.79 12.6°) and lower (0.79 3.5° dorsiflexion) with the h-AFO.
Fig. 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 91 S.D. taken from six trials in each condition. Curves are normalised to 100% gait cycle. Stance is separated from swing by the vertical line.
for the group of hemiplegic CP patients and 2.39 3.3° of dorsiflexion for the group of normal subjects. Although the d-AFO improved this angle significantly (8.0 9 7.3° plantarflexion), the h-AFO (3.79 4.0° dorsiflexion) produced a more normal value. Individual values of the nine patients are plotted in Fig. 5. The peak ankle dorsiflexion in stance was 4.99 8.8° for barefoot walking. Peak dorsiflexion increased significantly when the patients wore the d-AFO to 12.69 11.0° and to 16.19 6.3° with the h-AFO. Both mean
Fig. 4. Sagittal plane motion of pelvis, hip and knee for a typical patient (no. 6) in the study population. The curves are mean values 9 1 S.D. taken from six trials under each condition. Curves are normalised to 100% gait cycle. Stance is separated from swing by the vertical line.
J. Romkes, R. Brunner / Gait and Posture 15 (2002) 18–24
22 Table 3 Ankle sagittal plane kinematic data Variable
Barefoot
d-AFO
h-AFO
Normal
Initial contact Dorsiflexion
−18.59 10.3*
−8.09 7.3*†
3.7 9 4.0†‡
2.3 9 3.3
Stance phase Peak dorsiflexion % gait cycle
4.99 8.8* 33.5 915.0
12.6 911.0† 42.5 911.9†
16.1 9 6.3† 41.2 9 11.7†
12.9 9 4.1 42.4 97.2
Toe off Plantarflexion % gait cycle
17.1 9 11.3 59.59 1.8
11.49 12.8 60.392.2
−2.795.9*†‡ 60.2 9 2.1
10.9 9 8.4 60.1 91.9
Swing phase Peak plantarflexion % gait cycle
26.7 912.6* 76.29 14.6*
18.19 12.1† 67.89 2.5*
−0.79 3.5*†‡ 69.2 9 6.1*
14.9 9 7.6 64.6 92.5
Plantar- and dorsiflexion values are mean 9 1 S.D. 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 (PB0.05). Table 4 Knee sagittal plane kinematic data Variable Initial contact Knee flexion Stance phase Peak extension % gait cycle Swing phase Peak flexion % gait cycle
Barefoot
26.79 8.6*
d-AFO
h-AFO
3.6 95.9*
Normal
3.8 9 6.5*
11 9 3.4
−13.79 13.2 41.09 5.9
−12.9910.9 44.8 91.8*†
−12.99 10.2 44.3 9 3.5*†
−6.79 2.8 39.29 3.9
64.59 3.5 76.09 2.4*
67.49 4.2*† 75.3 9 3.20
67.2 9 5.1*† 73.3 9 3.0‡†
61.9 9 1.7 73.0 9 2.2
Values for knee flexion and extension are mean 9 1 S.D. 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 (PB0.05).
3.5. Knee
3.6. Hip
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 11.093.4°. In barefoot walking this value was 26.79 8.6°, for the d-AFO 23.695.9°, and for the h-AFO 23.89 6.5°. The differences were not significant. 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 91.8%, h-AFO 44.393.5%) compared with the normal reference (39.293.9%) and barefoot condition (41.0 9 5.9%). Peak flexion in swing phase increased compared with the normal (61.991.7°) and barefoot condition (64.5 93.5°) when patients wore the orthoses (d-AFO: 67.494.2°, h-AFO: 67.29 5.1°).
Hip flexion at initial contact under all three conditions (barefoot: 46.39 6.1°, d-AFO: 47.49 5.4°, h-
Table 5 Hip sagittal plane kinematic data Variable Initial contact Hip flexion Stance phase Peak extension % gait cycle
Barefoot
46.3 96.1*
d-AFO
47.4 95.4*
−4.69 8.4* −2.99 7.2 52.3 9 2.2
52.5 91.5
h-AFO
Normal
47.6 9 6.0* 39.9 9 5.1 −2.19 6.8
2.8 9 5.8
53.29 2.0
51.4 9 2.3
Hip flexion and extension values are mean 91 S.D. 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 (PB0.05).
J. Romkes, R. Brunner / Gait and Posture 15 (2002) 18–24
23
Table 6 Kinetic data Moments (Nm/kg) and powers (W/kg) (mean 91 S.D.)
Barefoot
d-AFO
h-AFO
Normal
Ankle Ankle Ankle Ankle
1.12 9 0.29 0.82 90.18* 2.31 90.74* 0.91 90.36*
1.08 9 0.26 0.92 9 0.30*† 1.78 90.72*† 0.89 9 0.23*
1.16 90.33 1.08 9 0.22*†‡ 1.50 90.84*† 0.87 90.42*
NR 1.37 9 0.14 0.49 9 0.21 2.29 90.66
plantarflexor moment, 0–30% GC plantarflexor moment, 30–65% GC power absorption, 0–30% GC power generation, 30–65% GC
Values are normalised 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 (PB0.05).
AFO: 47.6 9 6.0°) was higher than the normal (39.99 5.1°) (Table 5).
3.7. 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 1.1290.29 Nm/kg, with d-AFOs 1.08 90.26 Nm/kg, and with h-AFOs 1.1690.33 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 0.4990.21 W/kg. In the patients, peak values reached 2.3190.74 W/kg while walking barefoot. This value was significantly reduced to 1.7890.72 W/kg with the d-AFO and 1.5090.84 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 1.3790.14 Nm/kg. For the patients this value was 0.82 9 0.18 Nm/kg in barefoot walking, which increased significantly to 0.9290.30 Nm/kg with the d-AFO and 1.089 0.22 Nm/kg with the h-AFO. The patients generated significantly less ankle power during this period than the normal range of 2.299 0.66 W/kg. For barefoot walking this value was 0.919 0.36 W/kg and did not change when wearing either the d-AFO (0.89 90.23 W/kg) or the h-AFO (0.879 0.42 W/kg).
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].
4. 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 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
Fig. 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 91 S.D. reference values.
24
J. Romkes, R. Brunner / Gait and Posture 15 (2002) 18–24
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], 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 Ounpuu 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 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 fu¨ r Bewegungssto¨ rungen.’ 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, O0 unpuu 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] O0 unpuu S, Gage JR, Davis RB. Three-dimensional lower extremity joint kinetics in normal pediatric gait. J Pediatr Orthop 1991;11:341 – 9. [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 Biomechan 1992;25:347 – 57.
Comparison of a dynamic and a hinged ankle–foot orthosis by gait analysis in patients with hemiplegic cerebral palsy Jacqueline Romkes *, Reinald Brunner Laboratory for Gait Analysis, Children’s Uni6ersity Hospital (UKBB), Burgfelderstrasse 101, CH-4055 Basel, Switzerland Received 30 January 2001; received in revised form 30 June 2001; accepted 28 August 2001
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 fror the h-AFO to the proximal calf included a heel– toe gait pattern, reduced plantarflexion, increased step and stride length and reduced power absorption. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Gait analysis; Cerebral palsy; Ankle–foot orthosis; AFO; Equinus
1. Introduction 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 AFO’s 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
* Corresponding author. Tel.: + 41-61-685-5345; fax: +41-61-6855012. E-mail address: j – [email protected] (J. Romkes).
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 d-AFO 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.
2. 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 11.99 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 26.99 6.3 years, mean height of 174.39 6.5 cm, and a mean weight of 69.89 12.3 kg. The normal reference data were collected during barefoot walking.
0966-6362/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 9 6 6 - 6 3 6 2 ( 0 1 ) 0 0 1 7 8 - 3
J. Romkes, R. Brunner / Gait and Posture 15 (2002) 18–24
19
Table 1 Patient information Patient identification
Sex
Age (years)
Diagnosis
Contractures (°)
1 2 3 4 5 6 7 8 9 10 11 12 Total: 12
M M M M F M F M M M M F M: 9, F: 3
10.1 9.4 14.8 7.3 7.3 22.8 14.8 8.4 9.2 18.9 10.7 9.0 Mean: 11.9 94.9
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: – K: – – K: – – – K: – –
5° 5°
10°, A: 5°
5°, A: 10°
M, male; F, female; K, knee flexion contracture; A, ankle plantarflexion contracture.
Two types of AFOs were used in this study. The posterior part of the h-AFO (Fig. 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 orthosis was a d-AFO constructed according to Hylton’s method [4] (Fig. 2). 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 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
Fig. 1. Hinged ankle – foot orthosis with blocked plantarflexion.
20
J. Romkes, R. Brunner / Gait and Posture 15 (2002) 18–24
Fig. 2. Dynamic ankle – foot orthosis.
cycle as reported previously by Abel et al. [6]. Values were normalised to kg bodyweight and moments were expressed as internal moments. Statistical analysis was performed using a one-way analysis of variance (ANOVA) for repeated measurements to test the means of the conditions within the patient group. The twotailed 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 B 0.05. An improvement meant that the values came closer to normal reference values. Only data of the hemiparetic body side were considered.
3.3. 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 Figs. 3 and 4. Fig. 3 also includes the plantarflexor moment and power generation curves for this patient. Relevant mean values are summarised in Tables 3–6.
3.4. Ankle The majority of the significant differences between the two conditions is shown in Table 3. Ankle plantarflexion at initial foot contact was 18.5910.3°
3. Results
3.1. 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.
Table 2 Temperospatial parameters Barefoot
d-AFO
h-AFO
1.15 90.26 1.13 90.23 0.57 90.13 118.4 9 3.8
1.23 9 0.24 1.27 9 0.22 † 0.65 9 0.13 † 117.5 9 3.5
1.26 9 0.25 1.319 0.22† 0.669 0.12† 120.1 9 3.6‡
3.2. Temperospatial parameters
Velocity (m/s) Stride length (m) Step length (m) Cadence (steps per min) Stance phase (% GC) Double support (% GC)
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).
Values are mean 9 1 S.D. 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 (PB0.05); ‡, mean of this parameter tested significantly different from the mean of the d-AFO condition (PB0.05).
59.6 9 1.9
60.3 9 2.1
60.4 9 1.6
21.7 92.5
23.2 9 3.2
23.3 9 3.3
J. Romkes, R. Brunner / Gait and Posture 15 (2002) 18–24
21
values were no longer significantly different from the normal reference value of 12.99 4.1°. The h-AFO changed the plantarflexion angle at toeoff to 2.79 5.9° dorsiflexion, this being significantly different from the normal value (10.998.4° plantarflexion) and the barefoot (17.19 11.3°) and d-AFO (11.49 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 (18.1912.1°) while it was higher barefoot (26.79 12.6°) and lower (0.79 3.5° dorsiflexion) with the h-AFO.
Fig. 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 91 S.D. taken from six trials in each condition. Curves are normalised to 100% gait cycle. Stance is separated from swing by the vertical line.
for the group of hemiplegic CP patients and 2.39 3.3° of dorsiflexion for the group of normal subjects. Although the d-AFO improved this angle significantly (8.0 9 7.3° plantarflexion), the h-AFO (3.79 4.0° dorsiflexion) produced a more normal value. Individual values of the nine patients are plotted in Fig. 5. The peak ankle dorsiflexion in stance was 4.99 8.8° for barefoot walking. Peak dorsiflexion increased significantly when the patients wore the d-AFO to 12.69 11.0° and to 16.19 6.3° with the h-AFO. Both mean
Fig. 4. Sagittal plane motion of pelvis, hip and knee for a typical patient (no. 6) in the study population. The curves are mean values 9 1 S.D. taken from six trials under each condition. Curves are normalised to 100% gait cycle. Stance is separated from swing by the vertical line.
J. Romkes, R. Brunner / Gait and Posture 15 (2002) 18–24
22 Table 3 Ankle sagittal plane kinematic data Variable
Barefoot
d-AFO
h-AFO
Normal
Initial contact Dorsiflexion
−18.59 10.3*
−8.09 7.3*†
3.7 9 4.0†‡
2.3 9 3.3
Stance phase Peak dorsiflexion % gait cycle
4.99 8.8* 33.5 915.0
12.6 911.0† 42.5 911.9†
16.1 9 6.3† 41.2 9 11.7†
12.9 9 4.1 42.4 97.2
Toe off Plantarflexion % gait cycle
17.1 9 11.3 59.59 1.8
11.49 12.8 60.392.2
−2.795.9*†‡ 60.2 9 2.1
10.9 9 8.4 60.1 91.9
Swing phase Peak plantarflexion % gait cycle
26.7 912.6* 76.29 14.6*
18.19 12.1† 67.89 2.5*
−0.79 3.5*†‡ 69.2 9 6.1*
14.9 9 7.6 64.6 92.5
Plantar- and dorsiflexion values are mean 9 1 S.D. 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 (PB0.05). Table 4 Knee sagittal plane kinematic data Variable Initial contact Knee flexion Stance phase Peak extension % gait cycle Swing phase Peak flexion % gait cycle
Barefoot
26.79 8.6*
d-AFO
h-AFO
3.6 95.9*
Normal
3.8 9 6.5*
11 9 3.4
−13.79 13.2 41.09 5.9
−12.9910.9 44.8 91.8*†
−12.99 10.2 44.3 9 3.5*†
−6.79 2.8 39.29 3.9
64.59 3.5 76.09 2.4*
67.49 4.2*† 75.3 9 3.20
67.2 9 5.1*† 73.3 9 3.0‡†
61.9 9 1.7 73.0 9 2.2
Values for knee flexion and extension are mean 9 1 S.D. 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 (PB0.05).
3.5. Knee
3.6. Hip
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 11.093.4°. In barefoot walking this value was 26.79 8.6°, for the d-AFO 23.695.9°, and for the h-AFO 23.89 6.5°. The differences were not significant. 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 91.8%, h-AFO 44.393.5%) compared with the normal reference (39.293.9%) and barefoot condition (41.0 9 5.9%). Peak flexion in swing phase increased compared with the normal (61.991.7°) and barefoot condition (64.5 93.5°) when patients wore the orthoses (d-AFO: 67.494.2°, h-AFO: 67.29 5.1°).
Hip flexion at initial contact under all three conditions (barefoot: 46.39 6.1°, d-AFO: 47.49 5.4°, h-
Table 5 Hip sagittal plane kinematic data Variable Initial contact Hip flexion Stance phase Peak extension % gait cycle
Barefoot
46.3 96.1*
d-AFO
47.4 95.4*
−4.69 8.4* −2.99 7.2 52.3 9 2.2
52.5 91.5
h-AFO
Normal
47.6 9 6.0* 39.9 9 5.1 −2.19 6.8
2.8 9 5.8
53.29 2.0
51.4 9 2.3
Hip flexion and extension values are mean 91 S.D. 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 (PB0.05).
J. Romkes, R. Brunner / Gait and Posture 15 (2002) 18–24
23
Table 6 Kinetic data Moments (Nm/kg) and powers (W/kg) (mean 91 S.D.)
Barefoot
d-AFO
h-AFO
Normal
Ankle Ankle Ankle Ankle
1.12 9 0.29 0.82 90.18* 2.31 90.74* 0.91 90.36*
1.08 9 0.26 0.92 9 0.30*† 1.78 90.72*† 0.89 9 0.23*
1.16 90.33 1.08 9 0.22*†‡ 1.50 90.84*† 0.87 90.42*
NR 1.37 9 0.14 0.49 9 0.21 2.29 90.66
plantarflexor moment, 0–30% GC plantarflexor moment, 30–65% GC power absorption, 0–30% GC power generation, 30–65% GC
Values are normalised 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 (PB0.05).
AFO: 47.6 9 6.0°) was higher than the normal (39.99 5.1°) (Table 5).
3.7. 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 1.1290.29 Nm/kg, with d-AFOs 1.08 90.26 Nm/kg, and with h-AFOs 1.1690.33 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 0.4990.21 W/kg. In the patients, peak values reached 2.3190.74 W/kg while walking barefoot. This value was significantly reduced to 1.7890.72 W/kg with the d-AFO and 1.5090.84 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 1.3790.14 Nm/kg. For the patients this value was 0.82 9 0.18 Nm/kg in barefoot walking, which increased significantly to 0.9290.30 Nm/kg with the d-AFO and 1.089 0.22 Nm/kg with the h-AFO. The patients generated significantly less ankle power during this period than the normal range of 2.299 0.66 W/kg. For barefoot walking this value was 0.919 0.36 W/kg and did not change when wearing either the d-AFO (0.89 90.23 W/kg) or the h-AFO (0.879 0.42 W/kg).
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].
4. 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 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
Fig. 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 91 S.D. reference values.
24
J. Romkes, R. Brunner / Gait and Posture 15 (2002) 18–24
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], 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 Ounpuu 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 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 fu¨ r Bewegungssto¨ rungen.’ 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, O0 unpuu 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] O0 unpuu S, Gage JR, Davis RB. Three-dimensional lower extremity joint kinetics in normal pediatric gait. J Pediatr Orthop 1991;11:341 – 9. [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 Biomechan 1992;25:347 – 57.