Myelomeningocele (MMC), the most prevalent form of spina bifida, is a congenital neural tube defect that occurs when the neural tube does not close properly, leaving vertebral elements incompletely formed, neural structures exposed, and potentially impacting brain development. With the introduction of folic acid food fortification between 1997 and 1999, the incidence of spina bifida in Canada has decreased dramatically, from 0.86 to 0.4 per 1,000 births.1 Children with MMC present with varying degrees of physical ability according to the neurological level of the lesion; however, outcomes such as ambulation are often associated with numerous other factors, including the presence of lever arm dysfunction (related to orthopedic deformity such as calcaneovalgus, tibial torsion, or coxa valga), muscle paresis, impaired balance responses, and lower limb spasticity.2 Although a variety of orthotic options exist to address complex biomechanical impairments, the process of orthotic prescription and evaluation is often circumscribed by its dependence on observational gait analysis and a lack of evidence-based guidelines.
Gait in children with low lumbar and sacral level MMC is generally characterized by excessive lower limb flexion (i.e., crouch gait), increased pelvic rotation and obliquity, trendelenburg gait, and compensatory upper limb motion.3–6 For these individuals, altered gait mechanics have been reported to contribute to increased energy costs, pain, progressive orthopedic deformity, impaired balance, and mobility restrictions.2,7 Gait characteristics tend to progressively approach more normal levels in groups of children with successively lower lesion levels,8 although the most significant energy costs and gait alterations seem to be associated with weak or absence of gluteus medius, gastrocnemius, and soleus muscle function.5,7,8 Although these muscles, which are important for internal extension moment generation and stance-phase stability,9 are primarily innervated by the L4 to S1 spinal nerves, orthotic intervention and/or walking aids enable many children with lesions at or above S1 to be community ambulators.2
Ankle-foot orthoses (AFOs) are often recommended for children with lumbosacral lesions, to compensate for poor internal ankle plantarflexion moment generation (because of paresis) and inadequate foot lever function (because of paresis, orthopedic foot deformities, etc.). During stance phase, these impairments can contribute to excessive lower limb flexion and result in an inefficient gait pattern, which may be associated with increased energy cost and anteromedial knee joint stress.6 In this situation, the ground reaction force (GRF) may move anterior to the ankle joint and posterior to the knee joint early in loading response, increasing the external dorsiflexion and knee flexion moments. However, by transferring the GRF to the anterior tibial region, an AFO creates a plantarflexion moment at the ankle, inducing a knee extension moment from mid-stance through terminal stance and restoring the plantarflexion-knee extension couple.10–12
Ground reaction AFOs (GRAFOs) have been recommended for children with MMC to augment the deficient plantarflexion-knee extension couple.12 In contrast to a typical AFO with a pretibial strap, a GRAFO features a solid pretibial shell to more effectively redirect the GRF vector and slow tibial progression as the center of pressure moves distally under the foot during stance. Contraindications for GRAFO use include hip, knee, and/or ankle flexion contractures and moderate to severe pes valgus,13,14 although successful application has been reported in a child with MMC and 20° knee flexion contractures.10 Indications include excessive knee flexion and/or dorsiflexion during weightbearing (e.g., the stance phase of gait) and quadriceps strength of fair minus (3−/5) or better.14,15 Because of the effect of deformity-related lever arm dysfunction on the plantarflexion-knee extension couple, appropriate alignment of the foot and knee has been cited as a prerequisite for GRAFO use.13 However, the boundaries of what constitutes “appropriate alignment” have not been defined.
For children with MMC, the use of AFOs to compensate for muscle paresis and lever arm dysfunction is supported by conflicted evidence and lacks rigorous scientific evaluation (for a review see Ref. 14). In general, the reported improvements primarily affect sagittal plane lower limb kinematics, such as decreased stance phase ankle dorsiflexion and knee flexion and reduced prolonged knee extensor activity.10,16–18 Some of the conflicting outcomes with respect to ankle power generation, kinetics, and motion control reported in the literature19 may be related to factors such as the materials used to fabricate the orthoses (stiffness, resilience, etc.). Other reported benefits of AFO use include increases in stride length, velocity (because of decreased double support time), and hip flexion at initial contact, along with decreased energy consumption.19 Because AFOs are able to limit excessive stance phase dorsiflexion, they may be most beneficial to the gait mechanics of children with L4 and L5 lesions.16
The purpose of this case study was to compare the gait kinematics associated with GRAFOs, articulated AFOs (AAFOs), and barefoot walking for an adolescent with lumbosacral MMC. The primary goal for this client's orthotic management was a reduction in crouch gait (i.e., increased hip and knee extension and less ankle dorsiflexion during mid-stance to terminal stance), toward more normal values. Because visual gait observations have been shown to be discrepant with three-dimensional instrumented gait analysis, and therefore not reliable as the only source of gait data,20 we chose to augment our routine clinical gait assessment by including three-dimensional gait analysis. It was hypothesized that the improved biomechanical stability provided by the GRAFOs would lead to improved sagittal plane kinematic gait parameters (i.e., less stance phase ankle dorsiflexion and more knee extension to improve crouch gait) in comparison with either the AAFOs or the barefoot walking.
The participant was an active 14-year-old girl with lumbosacral MMC. She was a community ambulator with the aid of forearm crutches and AFOs and could walk short distances without these aids. She had started to use crutches for community ambulation after undergoing surgical tethered cord release 2 years earlier.
Physical assessment was conducted by the first author, a physical therapist with 13 years of experience. Lower limb muscle strength was graded on a 0 to 5 scale using standard manual muscle testing procedures.21 The participant's results are presented in Table 1. Physical examination did not reveal any lower limb contractures. Her thigh foot angle22 was 45° on the right side and 20° on the left, suggesting external tibial torsion that was more severe on the right. Normal values for the thigh foot angle range from 5° internal to 30° external, with a mean of 10° external.23 In standing, she demonstrated pes planus and excessive calcaneovalgus that was again more severe on the right.
She walked with crouch gait and a compensated trendelenburg gait pattern (with increased frontal plane hip and trunk motion and increased pelvic obliquity and rotation), with increased lumbar lordosis and upper limb motion. During stance phase, excessive ankle dorsiflexion, knee flexion, and hip flexion were noted, indicative of crouch gait. Excessive stance phase hip external rotation accompanied apparent valgus thrust at the knee, and she seemed to use further hip external rotation and trunk extension to progress her swing leg. Compared with the barefoot and AAFO conditions, the GRAFOs seemed to worsen the external orientation of her foot progression angle on the right leg but diminished it on the left leg (so that it was closer to normal). Otherwise, coronal and transverse plane kinematics did not seem noticeably different when comparing the barefoot condition to either of the orthosis conditions.
After approval by the institutional research ethics board, written parental consent and verbal child assent to videotape the testing and to present the results were obtained.
During the gait analysis, the participant wore two different pairs of custom-made AFOs. The AAFOs (Figure 1A), which she had worn for almost a year, had dual action ankle joints with dorsiflexion-assist that allowed 15° of plantarflexion and prevented dorsiflexion. They were custom fabricated from 5-mm polypropylene. She was fitted with the GRAFOs (Figure 1B) 2 weeks before the gait evaluation to allow an adjustment period. The GRAFOs positioned her ankle in neutral dorsiflexion/plantarflexion and were fabricated from 6-mm polypropylene.
Three-dimensional gait analysis was performed using a 6-camera optical tracking system (Vicon Motion Systems, Centennial, CO). Fifteen retroreflective markers were placed on the lower limbs and pelvis according to the Helen Hayes Hospital marker set24 (Figure 2). All markers were secured with adhesive tape or straps to minimize the movement of the markers on the skin. The infrared cameras recorded the data at a rate of 60 frames per second.
The participant performed six walking trials, at a self-selected speed, in each of three conditions: GRAFOs, AAFOs, and barefoot. For each trial, she walked a distance of 5 m. For the AFO trials, in which shoes were worn, care was taken to carefully place the ankle and foot markers over the estimated locations of the appropriate anatomical landmarks. In total, 18 trials were recorded, although three trials (one from each condition) were lost because of technical difficulties. Crutches were not used during any of the trials.
Motion data was processed using VICON Peak Motus (version 9.2; Centennial, CO) software. For kinematic data analysis, one gait cycle was extracted from each of the five useable trials for each condition (i.e., GRAFO, AAFO, and barefoot). The sagittal plane kinematic data were then ensemble averaged for the right and left ankle, knee, and hip and visually compared with normative data that were collected in the same laboratory. Peak sagittal plane range of motion (ROM) achieved during stance phase was calculated for the right and left ankle, knee, and hip and then compared with normative values. Step length and velocity were calculated to supplement the interpretation of kinematic data. For each condition, the mean of five gait cycles was calculated.
To quantify the magnitude of each orthosis' effect on the variables of interest (peak ROM for each joint, step length, and velocity), Cohen's d was calculated using the standardized mean difference formula, which has been recommended for single subject designs.25 This measure of effect size was calculated by subtracting the average for the orthosis condition from the average for the baseline (i.e., barefoot) condition and dividing by the standard deviation of the baseline condition.25 This method was used to compare the effectiveness of each orthosis relative to barefoot gait and the GRAFO relative to the AAFO, for the right and left legs.
SAGITTAL PLANE KINEMATICS
Overall, sagittal plane kinematic data (Figure 3) and peak ROM values during the stance phase (Figure 4) illustrate that left lower limb joint motion was closest to normal when the GRAFOs were worn and deviated farthest from normal in the barefoot and AAFO conditions. Although results were generally similar for the GRAFOs and AAFOs on the right leg, both orthoses resulted in small kinematic improvements compared with barefoot gait at the hip and knee. Table 2 presents the results of effect size calculations for peak ROM comparing each orthosis with barefoot gait, and the estimates of effect size (Cohen's d) for the GRAFO relative to the AAFO are included in the text.
At the ankle, the AAFOs allowed excessive dorsiflexion (27° on the left and 15° on the right), whereas the GRAFOs controlled this motion better than the AAFO, particularly on the left side (left d = 29.5; right d = 1.0). Compared with barefoot walking, both orthoses reduced foot drop during the swing phase.
Stance phase knee extension was best facilitated by the GRAFO on the left leg (d = 6.2, compared to AAFO), where peak motion approached normal values (i.e., approximately 7°). In contrast, the left knee remained flexed by more than 20° at peak stance phase extension in the AAFO and barefoot conditions. On the left leg, the GRAFO decreased knee flexion (d = 6.5), and with the AAFO, knee flexion was increased (d = −3.7), compared with barefoot walking. On the right, peak mid-stance flexion was greatest in the barefoot condition, and the orthoses limited this value slightly compared with barefoot gait (d = 5.5 for the GRAFO; d= 5.0 for AAFO). The participant achieved the least swing phase knee flexion while barefoot (<40°) and the most in the AAFOs (>45°).
The right hip achieved similar amounts of peak hip extension in stance in the two orthotic conditions, extending almost to 0° (the GRAFO being somewhat less effective than the AAFO; d = −0.6). In the barefoot condition, the right hip extended less than with either orthosis to approximately 6° of hip flexion. On the left leg, the GRAFOs allowed the hip to extend approximately 10° further than the AAFOs (d = 4.1). That is, the left hip stayed more flexed in the AAFO condition (8°) and extended past 0° in the GRAFOs. In fact, the only condition in which the hip extended past neutral (to approximately −3°) was on the left leg, with the GRAFOs.
The participant's average gait velocity (Figure 5) was faster in the GRAFOs (mean = 0.51 m/s) compared with the AAFOs (d = 0.46). She walked slightly faster in the AAFOs (mean = 0.494 m/s) than in the barefoot condition (mean = 0.488 m/s).
Mean step length (Figure 6) was generally shorter on the left leg compared with the right and shorter in barefoot than with either of the orthoses. Right step length was longer with the GRAFOs (mean = 52.5 cm) than with the AAFOs (mean = 49.2 cm; d = 1.7). There was no difference between orthoses on the left (mean = 44 cm).
The purpose of this case study was to compare the kinematic characteristics of gait associated with AAFOs, GRAFOs, and barefoot walking for an adolescent with lumbosacral MMC. Observational gait analysis provided a general overview of the gait pattern, identifying problematic features such as crouch gait posture, whereas the instrumented gait analysis facilitated identification of more specific details that helped to quantify the effects of each orthosis. Thus, in this context, these methods were viewed as complementary. Overall, the results illustrate the differing effects of 1) the two orthoses in comparison with barefoot walking and 2) the degree of lever arm dysfunction associated with external tibial torsion (which was greater on the right than on the left).
EFFECT OF GRAFOs ON SAGITTAL PLANE KINEMATICS
For the participant in this study, GRAFOs were associated with improved sagittal plane kinematics throughout the gait cycle and more normal peak extension in stance phase at the left ankle, knee, and hip. Other authors have reported reductions in stance phase knee flexion and ankle dorsiflexion with the use of GRAFOs by children with MMC10 and cerebral palsy.26 Observations such as these illustrate the potential to affect more proximal components of the kinematic chain by addressing the position and stability of the foot and ankle joints.
Although improvements in sagittal plane kinematics have also been noted with the use of conventional solid AFOs (with a posterior wall),16 this case study supports the notion that GRAFOs (with a pretibial shell) may be more effective in controlling stance phase ankle dorsiflexion. The difference in peak mid-stance dorsiflexion between barefoot and AAFO conditions was unexpected but likely related to the participant's step length, which was longer in the AAFO condition than barefoot. That is, her shorter step length while barefoot (which presumably functioned to maintain the GRF close to the ankle joint in an attempt to compensate for a diminished plantarflexion-knee extension couple) is likely to be associated with less dorsiflexion on the stance leg. In contrast, the longer step length observed in the AAFO condition was associated with the inability of the orthosis to adequately control her ankle motion compared with the GRAFO. Although technically articulated, the AAFOs were designed to limit dorsiflexion past neutral, so they can be considered similar to conventional solid AFOs in this respect. The amount of dorsiflexion achieved during the stance phase while wearing the AAFOs suggests some ankle motion within the orthosis and/or flexibility due to its design or materials used. Regardless, the reduction in stance phase knee flexion that was observed with the GRAFO (left > right) compared with the other two conditions suggests that this orthosis improved control at the ankle and restored the plantarflexion-knee extension couple most effectively.
SIGNIFICANCE OF ALTERED SAGITTAL PLANE KINEMATICS DURING THE STANCE PHASE
Qualitative observations of the kinematic data and peak flexion values for all joints during the stance phase indicate that sagittal plane flexion was generally greater on the right leg than it was on the left and greater in the barefoot and AAFO conditions than with the GRAFOs. These findings suggest that the degree of lever arm dysfunction, musculoskeletal stress, and energy costs were greatest on the right leg and in the barefoot and AAFO conditions.
When the lower limbs are excessively flexed during gait, continuous knee extensor activity must occur, consistent with an increased internal knee extensor moment.16 Electromyographic studies have confirmed that prolonged rectus femoris activity occurs during stance phase in this group of children but returns to more normal levels with AFO use.18 Similarly, examination of kinetic data has also shown a reduction in internal knee extensor moment with GRAFO use, an effect that is likely associated with the restoration of the plantarflexion-knee extension couple.10 Although this case study did not examine joint kinetics, the current literature would suggest that the improved left lower limb kinematics associated with the GRAFO condition may be associated with reduced internal knee and hip extension moments and improved gait efficiency.
EFFECT OF EXTERNAL TIBIAL TORSION
Excessive transverse plane pelvic and hip motion during gait may contribute to the development of external tibial torsion in children with MMC.27 When present, external tibial torsion positions the vertical component of the GRF vector posteriorly and laterally with respect to the knee flexion-extension axis, thereby shortening the knee flexion/extension axis lever arm and further impairing the plantarflexion-knee extension couple.13,27 When the ankle and knee flexion-extension axes are positioned in a typical relationship with respect to one another, the pretibial shell of the GRAFO is positioned to preferentially address sagittal plane function; however, external tibial torsion diminishes the effect of the sagittal plane correction. This is because the GRF vector that is normally applied posteriorly against the anterior tibia to extend the knee is diminished and redirected to increase the valgus force vector. Lim et al.12 have suggested that this risk of increased valgus knee stress becomes significant in individuals with lumbosacral MMC who have thigh foot angles >20°.
Vankoski et al.27 suggested that for children with more than 20° of external tibial torsion, an AFO may not effectively impact knee motion or knee extensor activity enough to significantly improve gait efficiency. With growth and time, these individuals are potentially at higher risk of knee arthrosis and deterioration or cessation of ambulation. Therefore, children with more than 20° external tibial torison may be good candidates for derotation osteotomy27,28 to optimize long-term ambulation outcomes. Conservative interventions such as higher level bracing may also be useful, and crutches may help to decrease the transverse plane knee joint stresses associated with AFO use, tibial torsion, and altered gait mechanics and distribute joint loads to the upper limbs.27 Strengthening programs that target the hip and knee extensors, abductors, and plantarflexors warrant exploration as well.18
RECOMMENDATIONS FOR GRAFO PRESCRIPTION
The findings provide further information about the effects of GRAFOs on gait kinematics in individuals with tibial torsion, supporting and clarifying previous recommendations13 that rotational deformities should be considered when prescribing a GRAFO. Although the “appropriate alignment” required for GRAFO use has not been defined, the results of this case study suggest that there may be an acceptable limit, beyond which the orthosis is not able to facilitate knee extension. There are likely to be multiple relevant factors, however, including lesion level, orthopedic deformity, and plantarflexor and hip abductor muscle strength. On the basis of these results, we suggest that an individual with a 20° external tibial torsion deformity may be considered an appropriate candidate for GRAFO use; however, with more severe deformities, such as those approximating or exceeding 45°, the GRAFO's biomechanical effectiveness may be impaired.
This case study demonstrates the successful application of a GRAFO to improve sagittal plane, lower limb joint kinematics in the presence of 20° external tibial torsion. The results also highlight the potential for less effective restoration of the plantarflexion-knee extension couple in the presence of more severe transverse plane deformities such as external tibial torsion of 45° or more. Based on our results, GRAFOs were recommended for this particular individual, and further modifications were carried out on the right GRAFO in an attempt to improve her gait mechanics.
As children with MMC grow, ambulation may become more challenging and lead to deterioration of walking ability over time,29 making gait efficiency and reduction of musculoskeletal strain a high priority for medical management. There is a need for higher levels of evidence in this field and the use of functional and participation outcome measures to support better clinical decision making regardless of whether three-dimensional quantitative gait analysis is available. However, when possible, the clinical application of three-dimensional gait analysis can provide valuable information to the evaluation process, helping to clarify more subtle visual observations.
The authors thank Kerri Staples and Marla Simpson for their helpful feedback during the preparation of this article, Ali Bell for her invaluable statistical advice, and the participant and her parents for their contributions to this project.
1.De Wals P, Tairou F, Van Allen MI, et al. Reduction in neural-tube defects after folic acid fortification in Canada. N Engl J Med 2007;357:135–142.
2.Bartonek A, Saraste H. Factors influencing ambulation in myelomeningocele: a cross-sectional study. Dev Med Child Neurol 2001;43:253–260.
3.Duffy CM, Hill AE, Cosgrove AP, et al. Three-dimensional gait analysis in spina bifida. J Pediatr Orthop 1996;16:786–791.
4.Fabry G, Molenaers G, Desloovere K, Eyssen M. Gait analysis in myelomeningocele: possibilities and applications. J Pediatr Orthop B 2000;9:170–179.
5.Gutierrez EM, Bartonek A, Haglund-Akerlind Y, Saraste H. Kinetics of compensatory gait in persons with myelomeningocele. Gait Posture 2005;21:12–23.
6.Õunpuu S, Thomson JD, Davis RB, DeLuca PA. An examination of the knee function during gait in children with myelomeningocele. J Pediatr Orthop 2000;20:629–635.
7.Duffy CM, Hill AE, Cosgrove AP, et al. The influence of abductor weakness on gait in spina bifida. Gait Posture 1996;4:34–38.
8.Gutierrez EM, Bartonek A, Haglund-Akerlind Y, Saraste H. Characteristic gait kinematics in persons with lumbosacral myelomeningocele. Gait Posture 2003;18:170–177.
9.Arnold AS, Anderson FC, Pandy MG, Delp SL. Muscular contributions to hip and knee extension during the single limb stance phase of normal gait: a framework for investigating the causes of crouch gait. J Biomech 2005;38:2181–2189.
10.Freeman D, Orendurff M, Moor M. Case study: improving knee extension with floor-reaction ankle-foot orthoses in a patient with myelomeningocele and 20° knee flexion contractures. J Prosthet Orthot 1999;11:63–73.
11.Kirtley C. Clinical Gait Analysis. Toronto: Elsevier Limited; 2006.
12.Lim R, Dias L, Vankoski S, et al. Valgus knee stress in myelomeningocele: a gait-analysis evaluation. J Pediatr Orthop 1998;18:428–433.
13.Gage JR. The Treatment of Gait Problems in Cerebral Palsy. London: Mac Keith Press; 2004.
14.Mazur JM, Kyle S. Efficacy of bracing the lower limbs and ambulation training in children with myelomeningocele. Dev Med Child Neurol 2004;46:352–356.
15.Seymour R. Prosthetics and Orthotics: Lower Limb and Spinal. Baltimore, MD: Lippincott Williams & Wilkins; 2002.
16.Thomson JD, Õunpuu S, Davis RB III, DeLuca PA. The effects of ankle-foot orthoses on the ankle and knee in persons with myelomeningocele: an evaluation using three-dimensional gait analysis. J Pediatr Orthop 1999;19:27–33.
17.Hullin MG, Robb JE, Loudon IR. Ankle-foot orthoses function in low level myelomeningocele. J Pediatr Orthop 1992;12:518–521.
18.Park BK, Song HR, Vankoski SJ, et al. Gait electromyography in children with myelomeningocele at the sacral level. Arch Phys Med Rehabil 1997;78:471–475.
19.Duffy CM, Graham HK, Cosgrove AP. The influence of ankle-foot orthoses on gait and energy expenditure in spina bifida. J Pediatr Orthop 2000;20:356–361.
20.Kawamura CM, de Morais Filho M, Barreto M, et al. Comparison between visual and three-dimensional gait analysis in patients with spastic diplegic cerebral palsy. Gait Posture 2007;25:18–24.
21.Clarkson HM, Gilewich GB. Musculoskeletal Assessment: Joint Range of Motion and Manual Muscle Strength. Baltimore, MD: Williams and Wilkins; 1989.
22.Staheli LT. Torsional deformity. Pediatr Clin North Am 1977;24:799–811.
23.Staheli LT, Corbett M, Wyss C, King H. Lower-extremity rotational problems in children. J Bone Joint Surg 1985;67:39–47.
24.Vaughan CL, Davis BL, O'Connor JC. Dynamics of Human Gait. 2nd ed. Cape Town, South Africa: Kiboho Publishers; 1999.
25.Olive ML, Smith BW. Effect size calculations and single subject designs. Educ Psychol 2005;25:313–324.
26.Rogozinski BM, Davids JR, Davis RB III, et al. The efficacy of the floor-reaction ankle-foot orthosis in children with cerebral palsy. J Bone Joint Surg Am 2009;91:2440–2447.
27.Vankoski S, Michaud S, Dias L. External tibial torsion and the effectiveness of the solid ankle-foot orthoses. J Pediatr Orthop 2000;20:349–355.
28.Dunteman RC, Vankoski SJ, Dias LS. Internal derotation osteotomy of the tibia: pre- and postoperative gait analysis in persons with high sacral myelomeningocele. J Pediatr Orthop 2000;20:623–628.
29.Williams JJ, Graham GP, Dunne K, Menelaus MB. Late knee problems in myelomeningocele. J Pediatr Orthop 1993;13:701–703.
KEY INDEXING TERMS: myelomeningocele; ground reaction ankle-foot orthoses; external tibial torsion; crouch gait; case report; gait analysis