Abstract
Background and Purpose Step training on a treadmill is a common intervention for adult and pediatric patients with spinal cord injuries (SCI). Treadmill training has not been used as an intervention for infants and toddlers with SCI before walking onset. This case report describes the intervention and stepping behaviors on a treadmill and overground of a toddler after the surgical removal of a rare spinal tumor resulting in SCI.
Case Description The toddler presented with an inability to step on the left, rare stepping on the right, and an apparent lack of sensation in the lower extremities. After spinal tumor excision at 5.5 weeks of age, step training on a treadmill and overground occurred once per week from 15 to 35 months of age in addition to traditional physical therapy.
Outcomes Independent symmetrical stepping emerged both on and off the treadmill over 20 months. Improvements in the number and pattern of steps occurred with training. Walking speed increased, and milestones important to overground walking developed.
Discussion Independent steps developed during the intervention with little motor development of the lower extremities during the first year of life. Furthermore, improvements in stepping alternation, standing, and walking occurred despite no evidence of sensation in the lower extremities.
Walking, an important milestone in infant development, facilitates participation throughout life. After spinal cord injury (SCI), treadmill training can facilitate relearning of stepping alternation and walking in adults.1,2 However, for children who sustain SCI before learning to walk, there is a paucity of information regarding interventions for facilitating stepping. In this case report, we describe the response to treadmill training of a toddler with an in utero spinal tumor resulting in SCI and little to no sensation below the injury.
Independent mobility provides infants with the ability to interact with the environment, thus improving depth perception, exploration, and social and psychological skills.3–5 As infants learn to walk and move, they learn more about the world. Upright standing, independent steps, and walking experience are important factors in bone and muscle health6 as well as cardiovascular fitness and can facilitate appropriate social interactions and play in toddlers.7 In addition, dynamic interactions among postural control, muscle strength, and balance facilitate walking and are improved through the trial and error of early walking.8 Independent walking is a goal shared by families of children with motor disabilities.
In some methods for facilitating stepping, stepping alternation, and locomotion in very young children, treadmill training approaches with known efficacy in older children or adults have been adapted.9–11 In older children with cerebral palsy, treadmill training increased walking speed, endurance, and gross motor function.9,10 In adults and older children with SCI, locomotor training also improved walking speed, lower extremity strength, balance, and gross motor function.2,11–15 Therefore, treadmill training promotes some recovery when it is used to promote relearning or refining walking that has already developed. Whether treadmill training is effective when SCI occurs before walking has developed is unknown.
Training and stepping on a treadmill have been implemented and tested before walking has developed in populations of children other than those with SCI, with good results.16–19 The most well-studied pediatric populations are infants and toddlers with Down syndrome; in brief, children who were preambulatory and participated in an at-home treadmill training program walked 3 months earlier than those who received standard care.20 Augmenting treadmill training with higher doses induced even earlier walking onset and further improved stepping behaviors and motor milestones,21 whereas augmenting with orthoses had a detrimental effect on motor milestones.22 Treadmill experience also appeared to increase stepping with an alternating pattern in infants who were preambulatory and at risk for cerebral palsy23–25 and in infants with myelomeningocele.26,27 Furthermore, a home-based treadmill program for 1 toddler who was not ambulatory and had myelomeningocele resulted in ambulation with a walker.28
In pediatric populations, damage to the spinal cord occurs in 2 major forms: (1) traumatic injury, typically well after independent walking has been established, and (2) developmental defects, such as myelomeningocele, in which typical walking milestones are limited or absent. Spinal cord damage in children results in unique impairments important for the development and rehabilitation of stepping behaviors. Common characteristics include upper and lower motor neuron signs, marked movement asymmetries, and incomplete motor and sensory characteristics, including paralysis, weakness, varying tone, abnormal reflexes, and decreased or absent sensation.
The primary purpose of this case report is to describe the stepping behaviors of a toddler with an in utero spinal cord tumor resulting in SCI during an intensive treadmill training program. The goal of the program was to allow as much repetitive, alternating, and independent stepping practice as possible. We did this by manually facilitating stepping with normal kinematics and by trial and error during independent stepping attempts on and off the treadmill.
Patient History and Review of Systems
The toddler in this case report had an in utero spinal tumor (from T11 to L4) that was identified at birth (Fig. 1). His left leg had no spontaneous movement and reduced muscle bulk and tone. His right leg had some spontaneous movements of the hip, knee, ankle, and toes. Patellar, plantar grasp, Babinski, and anal wink reflexes were absent bilaterally. He had no response to painful stimulation until his high left groin and right upper thigh. One dose of chemotherapy was administered before he underwent T9 to L5 osteoplastic laminectomy with decompression of the spinal cord and excision of the intraspinal extradural neoplasm on day 40 of life. Nerve roots at T11 and T12 were resected, and the tumor was removed. Some spontaneous left hip flexion was observed after surgery. Traditional physical therapist services began at 3 months of age. His family's goal was for him to have an independent form of ambulation; therefore, at 15 months of age, he began the 20-month treadmill program described in this case report.
Sagittal T2-weighted magnetic resonance image of the tumor at spinal levels T11 to L4 (arrow) in a 1-day-old infant.
Clinical Impression
It is likely that the in utero tumor and surgical resection caused injury to the sensorimotor spinal system, resulting in the asymmetrical weakness, sensory deficits, and abnormal reflexes in the toddler's lower extremities and trunk. This unique patient had impairments similar to those in children with myelomeningocele and adults and children with SCI. Treadmill programs were used successfully with a child who was preambulatory and had myelomeningocele and with children and adults who had SCI.2,14,28 The family and toddler were very motivated to achieve a form of ambulation. Therefore, the toddler appeared to be a good candidate for an intensive treadmill program.
To better determine whether the use of a treadmill program would benefit the toddler, we planned further examination, including administration of the Bayley Scales of Infant Development III, Pediatric Evaluation of Disabilities Index, Modified Ashworth Scale, anthropometric measurements, reflex tests, and a treadmill trial.
Examination
An examination was completed when the patient was 10 to 12 months old to determine appropriateness for treadmill training. The Bayley Scales of Infant Development III, reflex tests, and a treadmill trial were conducted when the child was 10 months old. The Bayley Scales of Infant Development III, valid and reliable norm-referenced scales used to evaluate motor function in children who are 1 to 42 months old, were administered to evaluate the toddler's motor skills.7 His gross motor skills but not his fine motor skills were found to be delayed, as noted by scale scores of 1 and 9, respectively. The Achilles, patellar, and Babinski reflexes were absent bilaterally. When suspended over the treadmill, he stepped a few times independently with his right lower extremity and was unable to step on the left.
The Pediatric Evaluation of Disabilities Index, a reliable and valid clinical assessment instrument that samples key functional capabilities and performances in children from 6 months to 7.5 years of age, was administered when the child was 11 months old.29,30 Standard scores on the instrument are based on a mean of 50 and a standard deviation of 10, and scores have been found to correlate significantly with neurological level, ability to walk, and independence in activities of daily living in children with spina bifida.31 The toddler's standard scores of 49.2 for self-care, 38.4 for mobility, and 49.9 for social function indicated mobility delays.
The Modified Ashworth Scale and anthropometric measurements were administered when the child was 12 months old to evaluate spasticity and lower extremity symmetry.32 All lower extremity joints scored 0 on the Modified Ashworth Scale, with the exception of the right hip extensors, which scored 1. His lower extremity tone was generally hypotonic. Anthropometric measurements of the lower extremities revealed greater size on the right than on the left: 1.5 cm for length, 1.5 cm for the thigh circumference, and 2 cm for the calf circumference. These differences warranted evaluation for hip dislocation, which was ruled out by radiography.
Clinical Impression
Overall, the patient had profound sensorimotor loss in both lower extremities, including asymmetrical strength, growth, and antigravity movements, noted as worse on the left. Tone was generally hypotonic, with flaccid paralysis. We cautiously assumed that he had no sensation in either lower extremity on the basis of a lack of attention or physical response to stimulation; however, he was too young to verbally answer or understand questions about sensation.
The lack of sensation in either lower extremity was a concern because experimental evidence for SCI has shown that some sensation appears to be important for the recovery of stepping behaviors.33 The risk that the toddler might not respond to treadmill training was mitigated by his ability to take a few independent steps on the treadmill with his right lower extremity.
The evaluation confirmed that the child's impairments were similar to those found in patients with myelomeningocele and SCI. Given that the child demonstrated the ability to step with his right leg on the treadmill despite a loss of sensation and that treadmill training has been successfully used to improve ambulatory skills in patients with similar impairments, treadmill training continued to be a good option for this patient. It was still unknown whether a toddler who had not previously walked, had no sensation, and had very limited left leg mobility would be able to perform alternating steps. Therefore, the primary goal of the treadmill training was to improve stepping. The planned outcome measures included independent stepping frequency and pattern on the treadmill, overground static standing, and stepping in a walker.
Intervention
An intensive stepping-focused intervention was added to traditional physical therapy. The step training program occurred at The Ohio State University Medical Center, and traditional physical therapy was delivered at Nationwide Children's Hospital. The intensive stepping-focused intervention at The Ohio State University Medical Center, consisted of treadmill training once per week for 20 months and was combined with a home treadmill training program performed by the family 8 minutes per day, 5 days per week, for 16 months.
Treadmill training with a harness and body weight support included 2 primary components: stepping manually facilitated by trainers and unassisted, independent stepping attempts by the toddler. In previous work with treadmill training for infants, the moving belt of the treadmill provided an environment that afforded stepping.16,17,24,27,34 Manual facilitation generally is unnecessary because infants at various ages and with various diagnoses produce independent steps.16,17,23,24,26,27,35–37 In our patient, only sporadic steps on 1 side occurred. Therefore, we selected facilitation to provide experience with frequency of stepping and alternating steps. The overarching objective was to encourage the toddler to step as independently as possible. Given that learning to step depends on volitional, independent movement on the treadmill,11,20 the toddler was encouraged to produce unassisted steps on the treadmill. Facilitated overground locomotor training was added to the training program once independent alternating stepping emerged at 24 months of age.
The treadmill program was completed by 4 trainers with 5 years of experience in adult step training; they delivered a combination of facilitated, unassisted, and overground training in progressively longer bouts for 50 to 60 minutes per session. During facilitated stepping, 3 trainers provided manual assistance to facilitate the proper alignment of all body segments and alternating and symmetrical steps. Facilitation was provided at the pelvis to achieve rotation and at each leg to maintain proper limb kinematics during swing and stance. Initially, we alternated 1 minute of step training with bilateral facilitation and 1 minute without facilitation to encourage independent attempts to step. Typically, we delivered 6 to 10 bouts per session. The training progressed to unilateral facilitation with independent stepping of the opposite limb. The transition to unilateral training occurred when attempts to swing the left leg emerged. We used fewer bouts but increased bout duration to 5 minutes or more. Finally, we added bouts of independent stepping with both legs to develop alternating stepping and bouts of backward walking to improve active hip extension. We added 3 to 5 of these bouts per session and continued each bout until alternating stepping ceased. All independent training bouts included manual assistance at the pelvis for rotation.
Embedded within the training progression were a reduction of body weight support and higher treadmill speeds. These parameters were adjusted to deliver the lowest body weight support and the highest speed.
Between bouts, the toddler performed standing training with little or no body weight support. Standing bouts varied between double-limb stance and single-limb stance. The toddler also performed dynamic standing training by moving from sitting to standing and back to sitting with facilitation as needed. Age-appropriate games, songs, bubbles, and toys were used as motivators for participation and positive reinforcement throughout all parts of the treatment sessions.
The home treadmill training program consisted of approximately 8 minutes of both manual suspension over the treadmill without lower extremity facilitation and use of the WalkAble Pediatric LiteGait WK100 suspension system (LiteGait, Tempe, Arizona) with lower extremity facilitation by the family. After 16 months of home training, the family moved and no longer had access to the home treadmill equipment.
Outcome
Data Reduction and Analysis
Each weekly treadmill intervention session was digitally recorded and used for behavioral analysis. Before each training session, data from two 1-minute bouts of independent stepping were collected as the toddler was held over the treadmill by a trainer (Fig. 2). No facilitation occurred. These bouts served as our measure of stepping frequency and pattern when analyzed frame by frame (30 frames per second). A step was defined as the foot moving past the hip joint during the swing phase and moving at least 1.5 foot lengths in the sagittal plane.20,27 Each stepping pattern was classified as alternating, single, parallel, or double on the basis of previous literature.17 Specifically, an alternating step was initiated within 20% to 80% of the step cycle on the opposite leg and, as a result, was preceded or followed by a step with the opposite leg. When a step with 1 leg was not preceded or followed by a step with the opposite leg, it was a single step. In a parallel step, both feet initiated the swing phase at approximately the same time. A double step occurred during a sequence of alternating steps when a second step was taken with 1 leg without a second step being taken with the opposite leg; this pattern appears as a stutter step.20,24,38 The average number of steps per 1-minute bout of independent stepping and the percentage of each stepping pattern served as primary outcomes per session. To measure overground performance, we calculated gait speed from video recordings collected at a sagittal angle, a rate of 30 frames per second, and a measured distance. Progression of walking and standing was based on a chart review of traditional physical therapy sessions.
(A) Test for independent stepping on the treadmill without facilitation. (B) Facilitated step training.
Independent Stepping Frequency on the Treadmill
At 15 months of age, a low rate of independent stepping was observed, with only 10 steps per minute. The rate of independent stepping progressively increased with intensive treadmill training to 45 steps per minute by 35 months of age (Fig. 3A). Of note, only the right leg accounted for most of the stepping rate from 15 through 20 months of age because there were few or no independent steps on the left (Fig. 3B). Over the 20-month intervention period, stepping with the right and left legs increased, with a greater rate of improvement being observed for the left leg, suggesting improvements in symmetry and bilateral function (Fig. 3).
Stepping frequency. (A) Monthly means and standard deviations of independent steps taken on the treadmill during testing sessions. (B) Comparison of right and left steps that contributed to the average number of steps per minute. An increase in the number of steps was observed over time, with the greatest improvement being noted for the left lower extremity.
Stepping Pattern During Treadmill Testing Sessions
Of the 4 possible treadmill stepping patterns described in the literature for infants and toddlers,17,27,38 only single or alternating steps on the treadmill were taken by the toddler in this case report. An increase in the percentage of alternating steps and a matching decrease in the percentage of single steps over the 20-month intervention period were observed (Fig. 4). Importantly, at 30 months of age, a pattern of alternating stepping on the treadmill occurred more than 80% of the time. In sharp contrast, during the initial 6 months of training, alternating steps comprised fewer than 10% of the total steps. The improvement in alternating stepping coincided with gains in independent stepping on the treadmill and overground (Fig. 4, solid lines). “Meaningful attempts” at independent stepping with the left leg on the treadmill occurred after 6 months of facilitated and unassisted training.
Pattern of stepping. The percentage of alternating steps on the treadmill increased dramatically as the percentage of single steps decreased. Initially, a response to training occurred during facilitated stepping (solid purple line). Variability in stepping pattern coincided with less facilitation and more unassisted stepping during training (solid orange line). A stable alternating pattern emerged after several months of unassisted stepping and overground (solid green) training. From 21 months on, responsiveness to both unassisted stepping (solid orange line) and facilitated stepping (broken purple line) occurred on the treadmill and overground in the clinic.
Overground Stepping and Independent Standing
Table 1 shows the development and progression of overground walking during the period of intensive treadmill training. Performance began with static standing with arm support on the walker and reached independent walking with a reverse walker. The pattern of walking varied, as did gait speed, which was measured at 0.048 and 0.040 m/s at 31 and 35 months of age, respectively. Coinciding with changes in walking were gains in standing during functional tasks and independently (Tab. 2). During both walking and standing, reliance on assistance decreased over time, and independence in walking and standing emerged.
Walking Progression at Clinical Assessmentsa
Static Standing Progression at Clinical Assessmentsa
In summary, before the intervention, the patient produced few or no steps in the treadmill environment, and those that did occur required facilitation. The toddler displayed greater deficits in the left lower extremity, as observed by occasional but unsuccessful attempts to step with the left leg. Improvements noted during the treadmill training program included an increase in the total number of steps taken, more symmetrical stepping, a shift from a single-step pattern to an alternating stepping pattern, and the ability to take independent steps with a reverse walker. A video of the patient's progress through the treadmill training intervention over time is available below.
A video of the patient's progress through the treadmill training intervention over time.
Discussion
Intensive locomotor treadmill training in combination with traditional physical therapy and a home treadmill training program may be effective for a toddler with SCI after the removal of a spinal tumor. The intensive treadmill training program was designed to promote the development of walking despite a lack of sensation in and severe motor impairments of the legs. The toddler showed great improvements in stepping frequency, stepping pattern, standing, and walking. He developed the ability to ambulate approximately 6 m (20 ft) in his walker with standby assistance.
Infants learning to walk take hundreds of practice steps per day.39 Treadmill training provides a task-specific intervention option for step training for infants with motor disabilities. The belt speed, body weight support, and type of physical therapist facilitation can be modified to accommodate a variety of physical needs and maximize voluntary stepping. We used these techniques to promote consecutive stepping practice in a toddler who had SCI and had never walked. Beyond independent practice, treadmill training allows therapists to kinematically replicate normal stepping through hands-on guidance or self-initiated movements by the child. We noted an increase in the total number of steps taken, more symmetrical numbers of steps taken with the right and left legs, and a shift from unilateral to alternating stepping. Therefore, like treadmill training for adults with SCI, treadmill training for toddlers with SCI may promote stepping behaviors important for walking.
The present case report describes a novel stepping facilitation protocol for use in pediatrics. The rationale for simultaneous and unilateral facilitation in our intervention was 2-fold. First, previous work with treadmill training in pediatric populations highlighted the importance of independent and voluntary movements as a catalyst for improvement.11,20 Here, we sought to strengthen any spared spinal pathways and teach the nervous system how to produce alternating steps in a child who had never learned to walk through normal milestone progression. We reasoned that allowing as much toddler-initiated stepping as possible during the treadmill training would better engage the nervous system than a preponderance of manual facilitation. Second, facilitated stepping on a treadmill in adults and a child with incomplete SCI improved gait speeds, upper and lower motor skills, balance, and recovery of walking, even in patients who were not ambulatory before the intervention.11,15,40
Several strategies are known to be important for the initiation of gait. Distal strategies include ankle unloading to facilitate swing, and proximal hip strategies include hip extension and flexor stretching to facilitate swing.41,42 The injury in the toddler in the present case report extended from the thoracic spine to the lumbar spine; therefore, it is likely that he used limited hip strategy when performing voluntary movement of his legs during stepping and walking. In recent work on immediate stepping responses in infants with myelomeningocele, those with higher lumbar lesions stepped less than those with lower lumbar lesions.27 Although this result is not surprising, it highlights the fact that infants with higher lumbar lesions did demonstrate the ability to perform some steps. The present case report adds to that literature, as no infant with myelomeningocele above L1 was included in the analysis. In addition, it is likely that our treadmill training program contributed to the toddler's ability to walk independently with a reverse walker.
Stepping interventions are designed to improve the motor behavior of the legs and provide a foundational context for improvements in functional skills, mainly mobility. Two key time points were important in linking spontaneous stepping on the treadmill and walking with a reverse walker. The toddler first performed more alternating steps than single steps on the treadmill at 23 months of age (Fig. 4), and at the same age he also started alternating his lower extremities when taking steps with a reverse walker (Tab. 1). At 29 months of age, he first demonstrated the ability to walk using his reverse walker with standby assistance (Tab. 1); then at 30 to 35 months of age, he took more steps per minute on the treadmill (Fig. 3).
Interestingly, there was a period of inconsistency and variability in the number and pattern of steps from 22 to 29 months of age. Initially, the toddler's stepping behavior was stable in that it was characterized by a small number of steps and consisted of single steps. By the end of the intervention, the stepping behavior was characterized by a much larger number of steps and a stable alternating pattern. We describe these improvements in the middle months as a period of phase transition from a lower to a higher frequency of steps and from a single-step to an alternating pattern.17 The movement of the treadmill belt and manual facilitation may have provided sensory feedback and musculoskeletal challenges needed to improve hip strategy and ankle and foot placement during gait; they also allowed for repeated independent practice during unassisted stepping.
Unlike children and adults with SCI, the toddler in the present case report had never learned to walk. The step training had to facilitate initial learning and practice of a skill, not relearning or retraining. Therefore, several factors could be important. There is evidence for adults with SCI that the role of sensation, including proprioception, is vital for the production of task-specific motor outputs, including walking.43–45 Interneuronal networks (central pattern generators) at the lumbosacral cord level integrate sensory input arriving from the environment with supraspinal input through the corticospinal tracts to produce a motor output specific to the task and environment in which that task is performed.46 When the central nervous system was provided with repetitive, appropriate task-specific sensory input in adults with SCI (retraining), neurologic recovery was shown to occur.2,15,44,46 Because the child in the present case report had never learned to walk, the theoretical framework that is used to treat adults with SCI could not be used here. Whether the gains in motor learning reported here were related to the impact of afferent sensory input from the environment on the central nervous system or were related to the intensive, active initiation of repetitive alternating steps in a permissive treadmill environment or the constant encouragement to take independent steps with errors47 remains unclear and warrants further study.
Pediatric treadmill training typically occurs for 8 minutes 5 days per week, as in randomized controlled trials with toddlers who had Down syndrome and a case study of a toddler with myelomeningocele.20,28 Our treadmill training program involved facilitated stepping 50 to 60 minutes once per week to provide a concentrated dose of intensive stepping activity. The protocol was tolerated well by the toddler and his family.
Clinical Impressions
The myelomeningocele literature suggests that infants born with high lumbar lesions have a 20% chance of walking, infants born with low lumbar lesions have an 80% chance of walking, and infants born with sacral lesions have a 90% chance of walking.48 Given that the child in the present case report had a thoracic lesion, he likely had less than a 20% chance of walking. In addition, his apparent limited sensation also was a negative factor in the development of walking.33 Therefore, we believe that it is unlikely that he would have walked without the treadmill intervention. We acknowledge that cause and effect cannot be determined in a case report. On the basis of previous research on treadmill training in infants with motor impairments, we believe that the task-specific treadmill intervention targeting independent stepping was necessary for the toddler in the present case report because it provided him with exposure to stepping practice that he would not have had without exposure to the treadmill.11,20,28 Infants who are developing typically have a great deal of practice stepping, which aids in walking development.39 The treadmill intervention may have provided the environment for the toddler to practice stepping and develop the ability to walk with assistance.
Limitations
The child in the present case report had a complex diagnosis and a complex set of symptoms and was receiving multiple disciplines of care. During our intervention, he received 3 separate intensive orthopedic evaluations and other interventions, including bracing, surgical consultation, and brief rehabilitation. In addition, the traditional physical therapist services that he received 2 times per week included strengthening, postural control, motor planning, and progression of motor skills. These experiences likely contributed to his overall stepping performance.
Footnotes
Dr Heathcock, Dr Christensen, and Dr Basso provided concept/idea/project design. All authors provided writing, data collection, and data analysis. Dr Heathcock and Dr Basso provided project management. Dr Christensen provided the patient. Dr Christensen and Dr Basso provided facilities/equipment. Dr Heathcock provided institutional liaisons.
Pilot data from this case report appeared as a poster presentation at the Combined Sections Meeting of the American Physical Therapy Association; February 3–6, 2014; Las Vegas, Nevada.
Ms Bush was supported by a Summer Research Opportunities Program (SROP) undergraduate fellowship and an SROP graduate fellowship during the time of data collection and manuscript preparation.
- Received October 23, 2013.
- Accepted April 8, 2014.
- © 2014 American Physical Therapy Association
References
Issue highlights
