Locomotor Training After Human Spinal Cord Injury: A Series of Case Studies.

Spinal cord injury (SCI) affects approximately 230,000 individuals in the United States, with 10,000 new cases yearly.[1] Despite the fact that the majority of people with SCI do not recover functional walking, rehabilitative strategies for ambulation beyond the use of orthotic and assistive devices have changed little over the last 20 years.[2-10] Conventional rehabilitation primarily provides compensatory strategies for accomplishing mobility and strengthening above the level of the lesion.[9,10] Recently, new approaches to facilitate locomotor recovery have been explored in humans using locomotor training that optimizes sensory information associated with locomotion.[11-18]

These studies using locomotor training in humans after SCI are based on extensive research in the control of locomotion in mammals. After a complete thoracic spinal cord transection, an adult cat can step independently with its hind limbs after several weeks of training that provides phasic sensory information associated with locomotion.[19-21] This ability to regain stepping has been attributed to oscillating neural circuits in the lumbosacral spinal cord that interact with phasic sensory input.[22-25] Initially during training, the spinally transected cat is suspended in the quadruped position over a treadmill. The cat is partially supported by a sling around the trunk, and manual assistance is provided to the hind limbs to ensure the rhythmic loading and unloading of the limbs[26] and appropriate limb kinematics.[27-29] After several weeks of training, the cat can begin to step independently with its hind limbs so that the trunk support and assistance in moving the hind limbs are no longer necessary. The neural circuits in the lumbosacral spinal cord caudal to the lesion respond to this peripherally mediated information and produce a coordinated, adaptable locomotor pattern in the absence of supraspinal influence.[19,20]

Harkema and colleagues (unpublished research) and other researchers[15,30] have demonstrated that the human lumbosacral spinal cord has the capability to respond to sensory information related to locomotion. Subjects with clinically complete SCI generated locomotor-like electromyographic (EMG) patterns after training (Harkema et al, unpublished research).[12-15] The technique used to evoke this response was similar to that used for cats[19,20]: providing manual assistance to the subject while he or she stepped with body weight support on a treadmill (BWST). The EMG and kinematic patterns of these subjects were modulated by limb load, kinematics, and speed of stepping (Harkema et al, unpublished research).[15,30] For example, the EMG amplitude of the ankle muscles was directly related to the load of the lower limb during stepping such that amplitude increased as load increased.[15] The EMG amplitude was furthermore correlated to the swing and stance phases of the step cycle.[15] Subjects with SCI could generate independent steps on the treadmill with locomotor training (Harkema et al, unpublished research).[12-16,18,31-33] Some subjects with SCI also developed the ability to walk overground, improved their overground walking speed and kinematics, and regained the ability to climb stairs.[16,18,31,32,34]

Edgerton et al[35] suggested that plasticity and motor learning in the spinal cord neural circuitry is dependent, first, on providing the specific sensory input associated with performance of a motor task and, second, on repetitive practice of that task. If the human spinal cord can learn by responding to specific sensory cues related to locomotion,[35-37] then understanding these mechanisms may lead to new approaches for rehabilitation of gait after neurologic injury. The purpose of this series of case studies is to describe locomotor training designed to incorporate sensory cues related to locomotion to improve stepping on a treadmill and walking overground by individuals after SCI.

Case Descriptions

Subjects

Four adults with SCIs were evaluated and trained at either the Functional Assessment Laboratory, Neurologic Rehabilitative Services, University of California at Los Angeles (n=2), or the Motor Behavior Laboratory, Department of Physical Therapy, University of Florida (n=2). We classified each subject's neurological level, lower-extremity motor score, and motor impairment level using the American Spinal Injury Association (ASIA) standards for neurological and functional classification (Tabs. 1 and 2).[38] Subjects signed an informed consent form approved by the institutional review board of the respective institution.

Table 1. American Spinal Injury Association (ASIA) Impairment Scale(a)

A=Complete. No sensory or motor function is preserved in the
sacral segments S4-S5.

B=Incomplete. Sensory but not motor function is preserved
below the neurological level and includes the sacral segments
S4-S5.

C=Incomplete. Motor function is preserved below the
neurological level, and more than half of key muscles below the
neurological level have a muscle grade less than 3.

D=Incomplete. Motor function is preserved below the
neurological level, and at least half of key muscles below the
neurological level have a muscle grade greater than or equal
to 3.

E=Normal. Sensory and motor function is normal.

(a) Reprinted with permission from Maynard FM Jr, et al, eds. International Standards for Neurological and Functional Classification of Spinal Cord Injury. Rev ed. Chicago, Ill: American Spinal Injury Association; 1996.

Table 2. Subject Description

                             Time                   ASIA
                             Since    ASIA(a)    Impairment
Subject                      Injury   Level of      Scale
No.       Sex      Age (y)   (mo)     Injury     Classification

1         Female     20       12         T5           A
2         Male       20        1         T5           C
3         Male       43        8         C6           D
4         Male       45        3         T9           D

(a) ASIA=American Spinal Injury Association Impairment Scale.[38]

Functional Assessments

Before and after completion of training, we obtained Functional Independence Measure (FIM)[39] scores and evaluated walking speed, distance covered while walking, balance, and handicap using assessments from the World Health Organization's International Classification of Impairments, Disabilities, and Handicaps[40] for each subject.

Locomotor disability. AS recommended by the ASIA standards for neurological and functional classification,[38] we used the locomotion subscales of the FIM to reflect disability (Tab. 3),[39] although this instrument was not developed for use with people with SCI or for measuring functional locomotion.

Table 3. Locomotion Subscale of the Functional Independence Measure[39]

Locomotion. Walk/Wheelchair. Includes walking, once in a standing
position, or, if using a wheelchair, once in a seated position, on
a level surface. Performs safely. Check the most frequent mode of
locomotion (walk ar wheelchair). If both are used about equally,
check both.

No Helper

7   Complete Independence--Subject walks a minimum of 50 m
    (1.50 ft) without assistive devices. Does not use a
    wheelchair. Performs easily.

6   Modified Independence--Subject walks a minimum of 50 m
    (150 ft) but uses a brace (orthosis) or prosthesis on leg,
    special adaptive shoes, cane, crutches, or walkerette; takes
    more than a reasonable amount of time, or there are safety
    considerations.

    If not walking, subject operates manual or motorized wheelchair
    independently for a minimum of 50 m (150 ft); turns around;
    maneuvers the chair to a table, bed, toilet; negotiates at least
    a percent grade; maneuvers on rugs and over doorsills.

5   Exception (Household Ambulation)--Subject walks only short
    distances (a minimum of 17 m [50 ft]) independently with or
    without a device. Takes more than reasonable amount of time,
    or there are safety considerations, or operates a manual or
    motorized wheelchair independently only short distances
    (a minimum of 17 m [50 ft]).

Helper

5   Supervision
    If walking, subject requires standby supervision, cueing,
    or coaxing to ga a minimum of 50 m (150 ft).

    If not walking, requires standby supervision, cueing, or
    coaxing to go a minimum of 50 m (150 ft) in wheelchair.

4   Minimal Contact Assistance--Subject performs 75% or more of
    locomotion effort to go a minimum of 50 m (150 ft).

3   Moderate Assistance-Subject performs 50% to 74% of locomotion
    effort to go a minimum of 50 m (150 ft).

2   Maximal Assistance--Subject performs 25% to 49% of locomotion
    effort to go a minimum of 50 m (150 ft). Requires assistance
    of 1 person only.

1   Total Assistance--Subject performs less than 25% of effort,
    or requires assistance of 2 people, or does not walk or wheel
    a minimum of 17 m (50 ft).

Locomotion: Stairs. Goes up and down 12 to 14 stairs (one flight)
indoors. Performs safely.

No Helper

7   Complete Independence--Subject goes up and down at least one
    flight of stairs without depending on any type of handrail
    or support. Performs easily.

6   Modified Independence--Subject goes up and down at least one
    flight of stairs but requires a side support or handrail, cane,
    or portable supports; takes more than a reasonable amount of
    time, or there are safety considerations.

5   Exception (Household Ambulation)--Subject goes up and down 4 to
    6 stairs independently with or without a device. Takes more
    than reasonable amount of time, or there are safety
    considerations.

Helper

5   Supervision--Subject requires standby supervision, cueing, or
    coaxing to go up and down one flight of stairs.

4   Minimal Contact Assistance--Subject performs 75% or more effort
    to go up and down one flight of stairs.

3   Moderate Assistance--Subject performs 50% to 74% of the effort
    to up and down one flight of stairs.

2   Maximal Assistance--Subject performs 25% to 49% of the effort to
    go up and down 4-6 steps.

1   Total Assistance--Subject performs less than 25% of effort, or
    requires assistance of 2 people, or does not go up and down
    4-6 stairs, or is carried.

Walking speed and distance covered in 2 minutes. We recorded gait speed for subjects who could walk overground independently (n=3). Subjects walked at a self-selected, comfortable speed (instructed as "walk at your natural pace") and at a fast speed (instructed as "walk as fast as you safely can") over a 7.3-m distance, and we recorded the time for the middle 3.6 m. We also measured the distance subjects (n=2) could walk in 2 minutes at a comfortable pace and covering as much ground as possible.[41]

Balance. We evaluated balance in one subject using the Berg Balance Test[42,43] and the Falls Efficacy Scale[44] because tests for balance have not been tested for validity for people with SCI.[45] The Berg Balance Test is used to assess balance performance during 14 sitting and standing tasks. An examiner scores each test item on a scale from 0 to 4 based on the time or distance requirements, supervision required, need for external support, or need for assistance from the examiner. The highest attainable score of balance function is 56. For the Falls Efficacy Scale, the subject self-rates the confidence with which he or she can perform 10 daily activities (eg, get dressed or undressed, hurry to answer the telephone) without falling on a scale of 1 to 10. The lower the percentage (10%), the greater the individual's confidence that the activities can be performed without falling.

Handicap. We assessed the degree of handicap using the physical function component of the Medical Outcomes Study (MOS) 36-Item Short-Form Health Survey (SF-36)[46] and the Craig Handicap Assessment and Reporting Technique (CHART)[47] in one subject. For the SF-36, the subject rates the degree to which his or her health currently limits the performance of certain daily activities (eg, climbing one flight of stairs, lifting or carrying groceries). The scores range from 0% to 100%, with the high percentage indicating that an individual's health does not at all limit performance of the activities. The CHART is an interview of 27 questions evaluating physical dependence, mobility, occupation, social integration, and economic self-sufficiency as a measure of a person's degree of handicap. In addition to these 2 assessments of handicap, this subject was also asked to identify 3 activities that he perceived as the most difficult to perform because of his SCI. The subject then rated each activity on a Likert-type 10-point scale, with a score of 10 representing the most difficult activity to perform.

Training Procedures

The locomotor training provided sensory cues and phasic information related to locomotion. These cues are critical for inducing stepping in animals with spinal transection[19,20] and in humans after an SCI.[12-15] Some of the sensory cues were: (1) generating stepping speeds approximating normal walking speeds (0.75-1.25 m/s[48])[20,30,49]; (2) providing the maximum sustainable load on the stance limbl[15,50]; (3) maintaining an upright and extended trunk and head[26,28]; (4) approximating normal hip, knee, and ankle kinematics for walking[15,28,29,51]; (5) synchronizing timing of extension of the hip in stance and unloading of limb with simultaneous loading of the contralateral limb[26,52,53]; (6) avoiding weight bearing on the arms and facilitating reciprocal arm swing[16]; (7) facilitating symmetrical interlimb coordination[54,55]; and (8) minimizing sensory stimulation that would conflict with sensory information associated with locomotion (eg, stimulation of extensor afferents during swing and flexor afferents during stance).[27,56-60] Locomotor training included step training on a treadmill and training for walking overground. Subjects wore lightweight tennis shoes with low-cut ankles and thin soles, slippers, or no footwear during training to maximize cutaneous stimulation from the sole of the foot during the stance phase. Orthoses were not used during locomotor training.

Step training on a treadmill. If a subject could not independently balance or maintain his or her full body weight load while stepping and executing appropriate limb kinematics, body weight support (BWS) was used in an effort to provide a safe and effective environment for locomotor training. A harness (Medical Harness(*)) worn by the subject and connected to an overhead-motorized or pneumatic lift (Neuro II([dagger])) suspended over the treadmill provided the BWS. The level of weight support was adjusted to maximize bilateral limb loading without the knee buckling during stance. If the subject, while step training with BWST, could not generate the appropriate limb and trunk kinematics associated with locomotion, manual assistance was provided. One trainer assisted each lower extremity. One hand of the trainer was placed on the anterior surface of the leg just below the patella to assist with knee extension during stance. The other hand was placed at the ankle or foot to assist with toe clearance during swing and heel-strike at initial stance. A third trainer stabilized the pelvis in order to maintain an upright trunk. The subjects were not permitted to support themselves by holding on to the treadmill side rails. Elastic side supports were attached diagonally to the subject and to side bars for balance when necessary but still allowed arm swing. Each subject was encouraged to swing his or her arms reciprocally with the lower extremities. When assistance for arm swing was necessary, one trainer manipulated 2 poles positioned horizontal to the ground and held by the subject. The trainers moved the poles forward and backward so that the contralateral arm moved forward with the ipsilateral leg, thereby producing a reciprocating arm swing (Fig. 1). Arm swing assistance was provided intermittently to achieve interlimb coordination, then removed when the subject could perform independently.

[Figure 1 ILLUSTRATION OMITTED]

To initiate stepping on the treadmill, subjects stood with their feet in a position simulating a stride, with one leg in an extended position near the mid-stance phase of the step cycle and bearing the majority of the body weight. As the treadmill speed was increased and carried the leg back to a position of terminal stance, the subjects shifted their body weight laterally and forward to the opposite leg as it reached initial contact. In this way, the body weight was supposed to quickly transfer to the forward leg, and the posterior limb could initiate swing. During stance, one trainer stabilized the knee to promote extension and the full loading of the limb while the trainer at the pelvis assisted in shifting of the body weight as needed. During leg swing, the trainer achieved toe clearance by increasing knee flexion with an anterior hold of the ankle or dorsum of the foot while being certain to avoid pressure or firm contact with the sole of the foot or the Achilles tendon. Pressure or stimulation of extensor afferents in these areas during swing may provide conflicting sensory information to the cues normally associated with swing and thus cause cessation of forward leg movement and transition to stance. The trainer facilitated knee flexion during swing by providing manual pressure on the medial hamstring tendon. Foot placement on the treadmill was monitored and assisted when necessary. Appropriate; simulation of leg movement through the swing and stance phases of gait was one key factor in eliciting stepping on the treadmill. Another factor was providing symmetric interlimb coordination. Particular attention was given by the trainers to achieving simultaneous heel-strike of one limb with lift off of the other limb in order to facilitate swing.

The trainers and the subject monitored the limb and joint kinematics during stepping and attained an optimal stepping pattern by adjusting the treadmill speed, BWS, and level of manual assistance. An optimal stepping pattern was achieved when it was spatially and temporally coordinated and most resembled normal walking. This pattern was then practiced repetitively during step training sessions. The duration of step training sessions was determined by (1) subject fatigue;, (2) maintenance of joint kinematics aligned with locomotion, and (3) maintenance of proper weight shift and limb loading patterns across the legs. A full-length mirror was placed in front of the subject for viewing and self-monitoring while training. Step training began by reaching what we considered normal walking speeds (optimized for each subject), and then focused on minimizing the manual assistance, decreasing the BWS, and increasing the duration to three 10-minute sessions. With a normal walking speed as a target,[48] training speeds were adjusted and selected based on achieving the optimal stepping pattern and often coincided with diminished need for manual assistance.

Training overground walking. We began locomotor training overground in conjunction with step training on the treadmill when a subject could stand independently while supporting at least 80% of his or her body weight and could independently generate the appropriate stepping kinematics on at least one limb. If a subject could not maintain balance at what we considered normal walking speed, assistance was provided using 1 of 3 different techniques. One approach was to have a trainer stand facing the subject with the subject holding the hands of the trainer. The trainer then walked backward, facilitating a reciprocating arm swing of the subject and offering enough support through the subject's arms to maintain balance but not substantially bear weight with the arms. Another trainer walked behind the subject, assisting with weight shifting and trunk extension at the pelvis, if needed. The second technique used to facilitate arm swing and provide balance was to have 2 trainers hold a pole (approximately 1.2 m x 1.9 cm [4 ft x 3/4 in]) on each side of the subject (Fig. 2). The poles were positioned horizontally (parallel to the floor) at the level of the greater trochanter in order for the subject to grip them during ambulation. The trainers then moved the poles forward and back in unison with the stepping of the legs. By having trainers hold the poles, the amount of weight bearing through the arms was minimized. The third approach was to have the subject use assistive devices. The approach used for each subject depended on comfort of the subject, availability of trainers, and locomotor ability of the subject.

[Figure 2 ILLUSTRATION OMITTED]

Assistive devices were selected based on principles of locomotor training such as maintaining proper limb and trunk kinematics and minimizing weight bearing by the arms. We found that long trekking poles or walking sticks provided suitable support for balance and appeared to facilitate weight bearing through the legs rather than the arms. We also adjusted the handle height of a rolling walker to achieve a level forearm with 90 degrees of elbow flexion in an effort to minimize upper-extremity weight bearing and forward trunk flexion. However, use of the rolling walker would not allow reciprocal arm swing during training overground and, therefore, was the least desirable approach we used during locomotor training.

To initiate overground walking, subjects stood with their feet in a position simulating a stride, with the majority of their body weight being supported by the limb positioned in terminal stance. They then quickly unloaded this limb and simultaneously loaded the contralateral limb, which initiated the swing phase ipsilaterally. As with step training, if we believed that an individual could not generate the appropriate limb kinematics associated with locomotion overground, manual assistance was provided. For example, a trainer initially assisted a subject in achieving knee flexion and toe clearance during swing by providing manual pressure on the hamstring tendon. As the subject's ability to achieve swing with toe clearance improved, the trainer no longer provided the manual assistance. Subjects were trained 3 to 5 times per week. The frequency of training, total number of sessions, and total time span for training of each subject are detailed in Table 4.

Table 4. Training Schedule

                      Total
Subject   Frequency   No. of     Total Time
No.       (no. of     Sessions   Span (wk)
          times/wk)

1             5          85          17
2             3          64          20
3             3          27          11
4             3          15           5

Outcomes

Tables 2 and 5 detail the results of training for each subject according to: (1) ASIA assessment, (2) step training on a treadmill, (3) training for overground walking, and (4) functional assessments.

Table 5. Results for Subjects 1, 2, 3, and 4 for American Spinal Injury Association (ASIA) Impairment Scale,[38] Functional Independence Measure (FIM),[39] Walking Performance, Balance, and Handicap Assessments(a)

                              Subject 1

                              Pre-        Post-
                              training    training

ASIA impairment               A           A
ASIA lower-limb motor score   0           0
FIM: locomotion
  Walk/wheelchair             6           6
  Stairs                      1           1
Walking speed
  Self-selected (m/s)
  Fast (m/s)
Distance covered during
  2-min walk (m)
Berg Balance Test[42,43]
Falls Efficacy Scale[44]
MOS SF-36 physical
  function
CHART

                              Subject 2

                              Pre-        Post-
                              training    training

ASIA impairment               C           D
ASIA lower-limb motor score   2           38
FIM: locomotion
  Walk/wheelchair             6           6
  Stairs                      1           6
Walking speed
  Self-selected (m/s)                     0.53
  Fast (m/s)                              0.75
Distance covered during
  2-min walk (m)                          67
Berg Balance Test[42,43]
Falls Efficacy Scale[44]
MOS SF-36 physical
  function
CHART

                              Subject 3

                              Pre-        Post-
                              training    training

ASIA impairment               D           D
ASIA lower-limb motor score   32          34
FIM: locomotion
  Walk/wheelchair             6           6
  Stairs                      1           6
Walking speed
  Self-selected (m/s)         0.09        0.33
  Fast (m/s)                  0.16        0.50
Distance covered during
  2-min walk (m)              12          68
Berg Balance Test[42,43]      30/56       43/56
Falls Efficacy Scale[44]      28%         16%
MOS SF-36 physical
  function                    10%         20%
CHART                         76%         89%

                              Subject 4

                              Pre-        Post-
                              training    training

ASIA impairment               D           D
ASIA lower-limb motor score   46          46
FIM: locomotion
  Walk/wheelchair             6           7
  Stairs                      6           7
Walking speed
  Self-selected (m/s)         0.6         1.6
  Fast (m/s)                  1.2         1.9
Distance covered during
  2-min walk (m)
Berg Balance Test[42,43]
Falls Efficacy Scale[44]
MOS SF-36 physical
  function
CHART

(a) MOS SF-36=Medical Outcomes Study 36-Item Short-Form Health Survey,[46] CHART=Craig Handicap Assessment and Reporting Technique.[47] A blank data cell indicates that the subject was not evaluated for that specific test.

Subject 1

ASIA assessment. Subject 1, a 20-year-old woman, had a traumatic SCI (T5, 1 year postinjury) and had completed an inpatient rehabilitation program. This subject's injury was classified, prior to training, as ASIA A with a lower-extremity motor score of 0/50. After completion of training, the subject's injury remained classified as ASIA A, and the lower-extremity motor score remained 0/50.

Step training on a treadmill. Subject 1 received 85 step training sessions on the treadmill. Prior to training, she was unable to stand or step independently on the treadmill. After step training, she consistently stepped with BWST (10% BWS) requiring only manual assistance for accurate foot placement. When what we believed were the optimal conditions were attained for limb loading, treadmill speed, and kinematics, the subject could produce 3 to 10 consecutive steps without manual assistance on at least one leg. After step training, she was also able to bear 90% of her body weight without manual assistance for 1 minute in an upright, stationary position with the harness and elastic straps to assist with stability.

Training for overground walking. Subject 1 did not receive training for overground walking.

Functional assessment of locomotor disability. Subject 1's pretraining scores were 6 (modified independence) for the FIM locomotion subscale walk/wheelchair item, with the wheelchair being the sole means of mobility, and 1 (total assistance) for the FIM locomotion subscale stairs item. After training, the FIM scores did not change.

Subject 2

ASIA assessment. Subject 2, a 20-year-old man, had a traumatic SCI (T5, 1 month postinjury) and was receiving conventional inpatient rehabilitation therapy in addition to locomotor training. His injury was classified prior to training as ASIA C with a lower-extremity motor score of 2/50. After completing training, the subject's ASIA impairment classification progressed to D, and his lower-extremity motor score had improved to 38/50.

Step training on the treadmill. Subject 2 received 64 step training sessions with BWST. During initial training, he required manual assistance to complete the lower-limb swing and maintain knee extension. After 3 weeks of training, he could step independently with the left leg and required minimal manual assistance to Complete the right leg swing. This stepping pattern was accomplished on the treadmill at a time when the subject had little or no voluntary control of his lower extremities. On completion of the 64 step training sessions, the subject required no BWS and was able to step independently on the treadmill.

Training for overground walking. Subject 2 received 44 sessions of training for overground walking. He initially used a rolling walker and required manual assistance to complete swing of the right leg. After 4 sessions, he stepped with the rolling walker without assistance. As his stepping with BWST improved, he also increased both speed and duration of overground walking. After 14 sessions of training for overground walking and 34 sessions of step training on a treadmill, he began overground walking with a cane. After completing locomotor training, he progressed to independent ambulation using a single-point cane.

Functional assessment of locomotor disability. Subject 2's pretraining FIM locomotion subscale walk/wheelchair score was 6 (modified independence) and his locomotion subscale stairs score was 1 (total assistance), with the wheelchair being his sole means of mobility. At the completion of training, his FIM locomotion subscale walk/wheelchair score remained 6 (modified independence), but his mode of locomotion changed from the wheelchair to walking full-time with a cane. Following training, his FIM locomotion subscale stairs score improved to 6 (modified independence), as he could climb stairs independently using a cane.

Functional assessment of walking speed and distance covered. As subject 2 was nonambulatory prior to training, walking performance overground was not assessed. After training, his overground, self-selected walking speed using a cane was 0.53 m/s and his fast speed was 0.75 m/s. After training, he walked 67 m in 2 minutes.

Subject 3

ASIA assessment. Subject 3, a 43-year-old man, had a traumatic SCI (C6, 8 months postinjury) and had completed an inpatient rehabilitation program. Prior to training, his injury was classified as ASIA D with a lower-extremity motor score of 32/50. Following completion of training, his ASIA impairment classification remained D, although his lower-extremity motor score increased to 34/50, with 2 muscles upgraded from a score of 1 (palpable or visible contraction) to 2 (active movement, full range of motion with gravity eliminated).

Step training on the treadmill. Subject 3 received 27 step training sessions. During early training, manual assistance at the pelvis and legs was provided to attain an upright trunk and facilitate lower-limb kinematics. Additionally, arm swing was assisted using the horizontal poles. After locomotor training, the subject progressed to requiring only intermittent assistance for upright posture, knee flexion, and toe clearance. He was then able independently to coordinate arm swing with the lower limbs. He progressed from requiring 45% BWS during his initial training to requiring 20% BWS at the completion of his sessions.

Training for overground walking. Subject 3 received 15 sessions of training for overground walking in conjunction with step training on the treadmill. Prior to training, he walked using a rolling walker. During training for overground walking, horizontal poles were used to facilitate reciprocal arm swing. Initiating walking from a stride position, maintaining an upright posture, and walking speed were emphasized. The subject's primary residual deficit during overground locomotion was limited toe clearance, bilaterally. After completing the combined step training and overground walking sessions, he walked full-time using forearm crutches.

Functional assessment of locomotor disability. Subject 3's initial FIM locomotion subscale walk/wheelchair score was 6 (modified independence) and his stairs score was 1 (total assistance). Prior to training, this subject used the wheelchair as the most frequent mode of mobility and could not ascend stairs. He was able to walk using a rolling walker, but he was not able to walk using a less restrictive assistive device such as forearm crutches. Following training, his primary means of mobility had improved from a wheelchair to walking with an FIM score of 6 (modified independence). His FIM stairs score improved to 6 (modified independence), as he required a handrail to climb stairs. Subject 3 had also progressed from using a rolling walker to using forearm crutches.

Functional assessment of walking speed and distance covered. Before training, subject 3 walked overground using a rolling walker at 0.09 m/s (self-selected, comfortable speed) with an asymmetrical stride. He walked at 0.16 m/s with the rolling walker when requested to walk as fast as possible. He also walked 12 m in 2 minutes using the rolling walker. He exhibited forward flexion of the trunk and weight bearing through the upper extremities throughout the gait cycle. He advanced his lower extremities by hip hiking with extended knees and plantar-flexed ankles. Foot contact occurred, with a plantar-flexed ankle resulting in toe-to-heel contact.

After locomotor training, the subject's overground walking speed using a rolling walker increased to 0.33 m/s at a self-selected speed and to 0.50 m/s at a fast walking speed. He was also able to walk with forearm crutches at a self-selected walking speed of 0.20 m/s and a fast walking speed of 0.27 m/s. He walked 68 m in 2 minutes using a rolling walker and 36.6 m in 2 minutes using forearm crutches. He was able to take multiple steps without the use of assistive devices, though very slowly and with balance difficulties.

Functional assessment of balance. Subject 3 initially had scores of 30/56 on the Berg Balance Test and 28% on the Falls Efficacy Scale. Following training, the Berg Balance Test score improved to 43/56 and the Falls Efficacy Scale score improved to 16%. Initial balance assessments indicated that the subject was at risk for falling[43] and had diminished confidence in performing certain everyday tasks without falling.[44] These improvements after training indicated a decrease in falls risk and a greater confidence in the subject's ability to perform everyday tasks without falling.[43,44]

Functional assessment of handicap. Before training, subject 3's SF-36 physical function score was 10%. After training, this score increased to 20%, indicating that he perceived that his health status after training did not limit his activities as greatly as it did prior to training.[46] His initial CHART score of 76% improved to 89% after training, also supporting a decrease in the degree of self-rated handicap.[47] After training, self-ratings also showed a decline in the degree of difficulty for walking, climbing stairs, and getting around in the home. Self-rating scores declined (1) from 10 to 3 for climbing stairs, (2) from 9 to 5 for walking, with clonus being an occasional problem, and (3) from 8 to 4 for getting around in the house and being able to go outside.

Subject 4

ASIA assessment. Subject 4, a 45-year-old man, had a decompression injury (T9, 3 months postinjury) and had completed inpatient rehabilitation. Prior to training, his injury was classified as ASIA D with a lower-extremity motor score of 46/50. After completing the training, his ASIA classification and lower-extremity motor score remained the same.

Step training on the treadmill. Subject 4 received 5 step training sessions with supervision, followed by 10 independent step training sessions at a health club. He did not require BWS or manual assistance during training, as he balanced independently, maintained his full body weight load, and achieved appropriate stepping kinematics on the treadmill. Following training, he stepped on the treadmill for 30 minutes at what we considered a normal walking speed of 1.3 m/s and maintained appropriate limb kinematics.

Training for overground walking. Subject 4 did not receive any training for overground walking.

Functional assessment of locomotor disability. Prior to training, subject 4 achieved scores of 6 (modified independence) on both the FIM locomotion subscale walk/ wheelchair and stair items. He was a full-time ambulator without assistive devices but required more than a reasonable amount of time to walk 45.7 m (150 ft). He required the use of a handrail for safety during stair climbing. After training, his FIM locomotion subscale walk/wheelchair and stair scores improved to 7 (complete independence). He thus walked at what we considered a normal speed and could safely go up and down a flight of stairs without depending on any type of handrail or support.

Functional assessment of walking speed. Subject 4 initially walked overground at 0.6 m/s (self-selected, comfortable speed) and at 1.2 m/s when requested to walk as quickly as possible. At the fast speed, he exhibited uncoordinated steps, exaggerated limb excursion, scissoring, and difficulty maintaining balance. After training, his overground, self-selected walking speed increased to 1.6 m/s and his fast speed increased to 1.9 m/s. He was then able to walk faster with no change in the coordination of his gait pattern. When he tried to run, his gait pattern deteriorated, demonstrating balance difficulties, step asymmetry, and uncoordinated limbs. At a 1-month follow-up evaluation after completion of training, his overground, self-selected walking speed was 1.2 m/s and his fast walking speed was 1.8 m/s. These speeds were retained at a 4-month follow-up evaluation, with a self-selected walking speed of 1.3 m/s and a fast walking speed of 1.9 m/s.

Discussion

All subjects improved their ability to step on the treadmill. One subject achieved the ability to walk overground, and 2 subjects improved their overground walking using locomotor training designed to simulate and produce phasic sensory information such as that related to locomotion. The subject with a chronic, clinically complete SCI (subject 1) achieved the ability to generate independent steps only on the treadmill with BWS, and she improved her ability to bear weight. The subject who initially was nonambulatory with an acute, incomplete SCI (subject 2) stepped independently on the treadmill and achieved the ability to walk overground independently with a cane. The subject with a chronic, incomplete SCI (subject 3), who prior to training ambulated only short distances within the home using a rolling walker, ambulated full-time using forearm crutches after locomotor training. Subject 3 initiated walking in his home and community in place of using the wheelchair and continually increased the amount of time he walked until the wheelchair was no longer necessary. His balance, distance walked, and perception of handicap also improved. The subject with a chronic, incomplete SCI (subject 4), who was initially a full-time ambulator with severely compromised overground walking speed, achieved what we considered a normal walking speed after locomotor training. These results suggest that locomotor training may increase the potential for recovery of walking in humans after an SCI.

The locomotor training was based on neuromuscular principles of locomotion from animal and human research studies (Harkema et al, unpublished research).[14,15,19-30] Given the response of the cases reported here, there appear to be several factors that are key to maximizing the locomotor capacity of individuals after SCI.

First, maximum weight bearing of the lower limbs is important during stance. Several animal studies have demonstrated that an increase in loading increases extensor motoneuron activity and facilitates locomotion; a response that can be mediated by the lumbosacral spinal cord in isolation of descending drive.[26,50] Similar results have been demonstrated in humans during locomotion.[15]

Second, when the speed of locomotion replicated normal walking speeds,[48] less manual assistance and greater independence while stepping were observed. Studies of cats with transected spinal cords have shown that the motor output from the lumbosacral spinal cord appropriately accommodates to increased treadmill speeds by enhancing extensor and flexor activity in an appropriately phased manner.[22,50] Similar results have been demonstrated in humans, where speeds nearing normal overground walking speeds increased lower-limb EMG activity[30] and improved stepping kinematics on a treadmill.[17] Given that normal walking speeds may not be attainable by subjects attempting to walk overground after an SCI, step training with BWST offers a safe and effective alternative, especially during the acute phase of injury. Step training with BWST provides an environment where normal walking speeds can be attained while providing appropriate sensory information of bilateral limb loading and kinematics to facilitate improved motor efferent patterns.

Third, by ensuring sufficient hip extension and unloading of the limb at the end of stance, we facilitated the swing phase of the step cycle. Several studies of cats not only demonstrated the relationship among hip extension, unloading of the limb, and the swing phase, but also demonstrated that this response is spinally mediated.[26,27,29,52] These sensory cues have also been demonstrated to be important in the initiation of swing in subjects with complete and incomplete SCIs (Harkema et al, unpublished research).

Fourth, weight bearing on the arms appeared to inhibit rhythmic stepping with the lower extremities, but a reciprocating arm swing, in a natural, coordinated form, facilitated stepping. Arm swing is an integral component of afferent input needed to facilitate lower-extremity motor output for walking.[17] Visintin and Barbeau[17] demonstrated that patterns of lower-limb EMG activity improved while individuals with an incomplete SCI walked on a treadmill with BWS without supporting part of their weight with parallel bars, as compared with walking while weight bearing through the arms on parallel bars.

The locomotor training we conducted is based on 2 assumptions. First, the spinal cord has the ability to respond to appropriate afferent information to generate stepping.[61] Second, activity-dependent plasticity occurs in the neural circuitry responsible for locomotion at both spinal and supraspinal levels (Harkema et al, unpublished research).[35-37,62] Using these principles as the foundation for locomotor rehabilitation hypothesizes that the nervous system adapts to specific activity and that recovery requires relearning the task of walking by providing the spinal cord with the appropriate sensory information.[35]

Conventional gait rehabilitation following SCI usually is designed to emphasize facilitation of recovery through strengthening and endurance training of the unaffected muscles and compensation for nonremediable deficits by using braces and assistive devices for support.[9,10] Together, these therapeutic strategies are designed to promote maximum functional capacity of muscles and to compensate for the absence of volitional lower-limb muscle contractions or for weakness. Given the generally accepted assumption that repair and recovery of the injured spinal cord is not possible,[63] successful mobility is dependent on learning new behavioral strategies,[64] requiring either a wheelchair and or bracing with assistive devices.

Repetitive practice using conventional gait rehabilitation may teach a compensatory mode of ambulation that may not take advantage of the plasticity of the neuromuscular system.[35] For instance, when using a walker, attaining hip extension may be compromised due to the forward flexion of the trunk while weight bearing through the arms. This posture likely attenuates hip extension during terminal stance and reduces lower-limb loading, thereby altering the sensory input that facilitates the swing phase.[29,65] Lower-extremity bracing that provides support for the lower extremities against gravity may limit normal joint range of motion. Weight bearing on the hands while using a walker or crutches allows for unweighting and forward progression of the lower extremities by a compensatory strategy; however, it inhibits lower-limb EMG activity.[17]

Ambulation following SCI and conventional therapy can be predicted by the manual muscle test scores of the lower limbs, the completeness of the lesion, and the neurological level of injury.[66,67] However, new evidence suggests that the degree of voluntary motor control of the lower extremities may not be a predictor of locomotor outcome.[31,32] This perspective was supported by our observations of subject 2, whose independent stepping ability preceded his ability to voluntarily contract his lower-extremity muscles. In addition, subject 1, who had no detectable voluntary movement below the spinal cord lesion, was able to generate steps independently on the treadmill, although she did not achieve overground walking. Together, these 2 observations further challenge the validity of predicting locomotor outcomes after an SCI solely based on voluntary motor control.

We presented the results of 4 individual case studies. Without a control group, factors other than the locomotor training may have contributed to the outcomes and affected the recovery of locomotion. The time since injury, for example, can influence recovery, as can the level and severity of injury and age at time of injury.[2,67-70] The probability of spontaneous recovery following injury decreases dramatically 6 months following an injury and is not expected I year postinjury.[67] All subjects in the case studies were trained within I year of injury, specifically, at 1, 3, 8, and 12 months postinjury. As the subjects in these case studies had injuries 1 year or less prior to training, spontaneous recovery may not be ruled out as relevant to the subjects' recovery of function.

The subjects' responses to training, however, suggest that spontaneous recovery may not have been the sole cause of changes in locomotor status. Three of the subjects completed their inpatient and outpatient physical therapy programs prior to participating in locomotor training. Physical rehabilitation programs are often terminated when physical and functional gains are minimal or have plateaued. In contrast, subject 1, for example, stepped on the treadmill without the return of voluntary motor control of the lower extremities. Subject 2 began stepping on the treadmill prior to the onset of voluntary motor control of the lower extremities. Subject 3's improvement in walking function was not paralleled by changes in lower-extremity muscle force. Subject 4, during his first training session, demonstrated immediate improvement in his step length, step symmetry, and right lower-limb kinematics while on the treadmill. Following the first session, his gait pattern while walking overground did not similarly improve, suggesting that the immediate change in walking while on the treadmill was specific to the training environment. These responses to training are explainable by the plasticity of the nervous system and its capacity to respond to locomotor-specific afferent input to generate stepping.[35] Spontaneous recovery may have an interactive effect with the activity-dependent plasticity of the nervous system on the outcomes achieved with training.

The level and severity of spinal cord injury may have affected our posttraining findings. The ASIA impairment and lower-limb motor scores varied widely among the 4 individuals presented in the case studies. Furthermore, all of our subjects were young to middle-aged adults. Neuroplasticity and spontaneous recovery likely vary with age at the time of injury and add an instrumental variable.[69,70] Investigating the influence of injury chronicity, level, and severity and of age at the time of injury on the outcomes of locomotor training is certainly warranted via controlled, experimental studies.

Summary

The purpose of this series of case studies was to describe the use of locomotor training designed to incorporate sensory cues related to locomotion in order to improve stepping on a treadmill and overground walking in adults with an SCI. The training was based on research indicating that the neuromuscular system can respond to appropriate sensory information associated with locomotion to generate stepping. The principles for training were developed from research of animal models of SCI and activity-dependent plasticity and then applied to the locomotor training of humans after SCI. Whether locomotor training is effective with people who have SCI at all neurological levels, all degrees of injury severity, and in both acute and chronic injuries is not known. Further studies are warranted to understand the control of locomotion and to develop locomotor training to optimize outcomes following an SCI and other neurologic deficits.

Future Studies

By understanding the role of the human spinal cord in the generation of human locomotion, principles and strategies can be developed to retrain the human nervous system for the task of locomotion after an SCI. Future studies to discern the critical sensory information that can be interpreted and integrated by the human spinal cord should provide important information in designing rehabilitation strategies and assistive devices. Whether locomotor training can improve recovery after human SCI as compared with conventional therapy should be evaluated in a randomized clinical trial. The impact of locomotor training on the quality of life and the degree of handicap in individuals after an SCI should be addressed. The application of locomotor training to other neurologic deficits associated with stroke,[71,72] cerebral palsy, Parkinson disease, and brain injury should be explored. If the spinal cord has a critical role in locomotion and can relearn to execute stepping with training driven by afferent information, and if supraspinal centers can also reorganize, then these approaches can potentially have a dramatic impact o:n the recovery of walking after neurologic injury.

(*) Robertson Harness, PO Box 90086, Henderson, NV 89009-0086.

([dagger]) Vigor Equipment Inc, 4915 Advance Way, Stevensville, MI 49127.

References

[1] Green BA, Eismont FJ, O'Heir JT. Pre-hospital management of spinal cord injuries. Paraplegia. 1987;25:229-238.

[2] Daverat P, Sibrac MC, Dartigues JF, et al. Early prognostic factors for walking in spinal cord injuries. Paraplegia. 1988;26:255-261.

[3] Frankel HL, Hancock DO, Hyslop G, et al. The value of postural reduction in the initial management of closed injuries of the spine with paraplegia and tetraplegia, I. Paraplegia. 1969;7:179-192.

[4] Burke DC, Burley HT, Ungar GH. Data on spinal injuries, part II: outcome of the treatment of 352 consecutive admissions. Aust NZ J Surg. 1985;55:377-382.

[5] Stover S, Find P. Spinal Cord Injury: The Facts and Figures. Birmingham, Ala: The University of Alabama at Birmingham; 1986.

[6] Knutsdottir S. Spinal cord injuries in Iceland 1973-1989: a follow-up study. Paraplegia. 1993;31:68-72.

[7] Guttman S. Spinal Cord Injuries: Comprehensive Management and Research. 2nd ed. Osney Mead, Oxford, England: Blackwell Scientific Publications; 1976.

[8] Ford J, Duckworth B. Physical Management for the Quadriplegic Patient. Philadelphia, Pa: FA Davis Co; 1974.

[9] Somers M. Spinal Cord Injury: Functional Rehabilitation. East Norwalk, Corm: Appleton & Lange; 1992.

[10] Atrice M, Gonter M, Griffin D, et al. Traumatic spinal cord injury. In: Umphred D, ed. Neurological Rehabilitation. St Louis, Mo: CV Mosby Co; 1995:484-534.

[11] Barbeau H, Blunt R. A novel interactive locomotor approach using body weight support to retrain gait in spastic paretic subjects. In Wernig A, ed. Plasticity of Motorneuronal Connections. New York, NY Elsevier Science Publishers; 1991:461-474.

[12] Dietz V, Colombo G, Jensen L. Locomotor activity in spinal man. Lancet. 1994;344:1260-1263.

[13] Dietz V, Colombo G, Jensen L, Baumgartner L. Locomotor capacity of spinal cord in paraplegic patients. Ann Neurol. 1995;37:574-582.

[14] Dobkin BH, Harkema SJ, Requejo PS, Edgerton VR. Modulation of locomotor-like EMG activity in subjects with complete and incomplete spinal cord injury. J Neurol Rehabil. 1995;9:183-190.

[15] Harkema SJ, Hurley SL, Patel UK, et al. Human lumbosacral spinal cord interprets loading during stepping. J Neurophysiol. 1997;77: 797-811.

[16] Visintin M, Barbeau H. The effects of body weight support on the locomotor pattern of spastic paretic patients. Can J Neurol Sci. 1989;16: 315-325.

[17] Visintin M, Barbeau H. The effects of parallel bars, body weight support and speed on the modulation of the locomotor pattern of spastic paretic gait: a preliminary communication. Paraplegia. 1994;32: 540-553.

[18] Stewart JE, Barbeau H, Gauthier S. Modulation of locomotor patterns and spasticity with clonidine in spinal cord injured patients. Can J Neurol Sci. 1991; 18:321-332.

[19] Lovely RG, Gregor RJ, Roy RR, Edgerton VR. Effects of training on the recovery of full-weight-bearing stepping in the adult spinal cat. Exp Neurol. 1986;92:421-435.

[20] Barbeau H, Rossignol S. Recovery of locomotion after chronic spinalization in the adult cat. Brain Res. 1987;412:84-95.

[21] de Leon RD, Hodgson JA, Roy RR, Edgerton VR. Locomotor capacity attributable to step training versus spontaneous recovery after spinalization in adult cars. J Neurophysiol. 1998;79:1329-1340.

[22] Forssberg H. Stumbling corrective reaction: a phase-dependent compensatory reaction during locomotion. J Neurophysiol. 1979;42: 936-953.

[23] Grillner S. Interaction between central and peripheral mechanisms in the control of locomotion. Prog Brain Res. 1979;50:227-235.

[24] Grillner S. Neurobiological bases of rhythmic motor acts in vertebrates. Science. 1985;228:143-149.

[25] Pearson KG, Rossignol S. Fictive motor patterns in chronic spinal cats. J Neurophysiol. 1991;66:1874-1887.

[26] Conway BA, Hultborn H, Kiehn O. Proprioceptive input resets central locomotor rhythm in the spinal cat. Exp Brain Res. 1987;68: 643-656.

[27] Andersson O, Grillner S, Lindquist M, Zomlefer M. Peripheral control of the spinal pattern generators for locomotion in cat. Brain Res. 1978;150:625-630.

[28] Andersson O, Grillner S. Peripheral control of the cat's step cycle, II: entrainment of the central pattern generators for locomotion by sinusoidal hip movements during "fictive locomotion." Acta Physiol Scand. 1983;118: 229-239.

[29] Grillner S, Rossignol S. On the initiation of the swing phase of locomotion in chronic spinal cats. Brain Res. 1978;146:269-277.

[30] Patel UK, Dobkin BH, Edgerton VR, Harkema SJ. The response of neural locomotor circuits to changes in gait velocity [abstract]. Soc Neurosci. 1998;24:2104.

[31] Wernig A, Muller S. Laufband locomotion with body weight support improved walking in persons with severe spinal cord injuries. Paraplegia. 1992;30:229-238.

[32] Wernig A, Muller S, Nanassy A, Cagol E. Laufband therapy based on "rules of spinal locomotion" is effective in spinal cord injured persons [published erratum appears in Eur J Neurosci. 1995;7:1429]. Eur J Neurosci. 1995;7:823-829.

[33] Barbeau H, Danakas M, Arsenault B. The effects of locomotor training in spinal cord injured subjects: a preliminary study. Restor Neurol Neurosci. 1993;5:81-84.

[34] Wernig A, Nanassy A, Muller S. Laufband (treadmill) therapy in incomplete paraplegia and tetraplegia. J Neurotrauma. 1999;16: 719-726.

[35] Edgerton VR, Roy RR, Hodgson JA, et al. A physiological basis for the development of rehabilitation strategies for spinally injured patients. J Am Paraplegia Soc. 1991;14:150-157.

[36] Dobkin BH. Neuroplasticity: key to recovery after central nervous system injury. West J Med. 1993;159:56-60.

[37] Muir GD, Steven JD. Sensorimotor stimulation to improve locomotor recovery after spinal cord injury. Trends in Neuroscience. 1997;20: 72-77.

[38] Maynard FM Jr, Bracken MB, Creasey G, et al. International Standards for Neurological and Functional Classification of Spinal Cord Injury. American Spinal Injury Association. Spinal Cord. 1997;35: 266-274.

[39] Functional Independence Measure: Guide for the Uniform Data Set for Medical Rehabilitation (Adult FIM), Version 4.0. Buffalo, NY: State University of New York at Buffalo; 1993.

[40] International Classification of Impairments, Disabilities, and Handicaps. Geneva, Switzerland: World Health Organization; 1980.

[41] Butland RJ, Pang J, Gross ER, et al. Two-, six-, and 12-minute walking tests in respiratory disease. Br Med J (Clin Res Ed). 1982;284: 1607-1608.

[42] Berg KO, Wood-Dauphinee SL, Williams JI, Gayton D. Measuring balance in the elderly: preliminary development of an instrument. Physiotherapy Canada. 1989;41:304-311.

[43] Berg KO, Wood-Dauphinee SL, Williams JI, Maki B. Measuring balance in the elderly: validation of an instrument. Can J Public Health. 1992;83 (suppl 2):S7-S11.

[44] Tinetti ME, Richman D, Powell L. Falls efficacy as a measure of fear of falling. J Gerontol. 1990;45:P239-P243.

[45] Ladouceur M, Pepin A, Norman KE, Barbeau H. Recovery of walking after spinal cord injury. Adv Neurol. 1997;72:249-255.

[46] Ware JE Jr, Sherbourne CD. The MOS 36-item short-form health survey (SF-36), I: conceptual framework and item selection. Med Care. 1992;30:473-483.

[47] Whiteneck GG, Charlifue SW, Gerhart KA, et al. Quantifying handicap: a new measure of long-term rehabilitation outcomes. Arch Phys Med Rehabil. 1992;73:519-526.

[48] Craik RL, Dutterer L. Spatial and temporal characteristics of foot fall patterns. In: Craik RL, Oatis CA, eds. Gait Analysis: Theory and Application. St Louis, Mo: Mosby-Year Book; 1995:143-158.

[49] Edgerton VR, Roy RR, Hodgson JA, et al. Potential of adult mammalian lumbosacral spinal cord to execute and acquire improved locomotion in the absence of supraspinal input. J Neurotrauma. 1992; 9(suppl 1):S119-S128.

[50] de Guzman CP, Roy RR, Hodgson JA, Edgerton VR. Coordination of motor pools controlling the ankle musculature in adult spinal cats during treadmill walking. Brain Res. 1991;555:202-214.

[51] Lovely RG, Gregor RJ, Roy RR, Edgerton VR. Weight-bearing hindlimb stepping in treadmill-exercised adult spinal cats. Brain Res. 1990;514:206-218.

[52] Duysens J, Pearson KG. Inhibition of flexor burst generation by loading ankle extensor muscles in walking cats. Brain Res. 1980;187: 321-332.

[53] Pearson KG, Duysens J. Function of segmental reflexes in the control of stepping in cockroaches and cats. In: Herman R, Grillner S, Stein PSG, Stuart DG, eds. Neural Control of Locomotion. New York, NY: Plenum Press; 1976:510-537.

[54] Berger W, Dietz V, Quintern J. Corrective reactions to stumbling in man: neuronal co-ordination of bilateral leg muscle activity during gait. J Physiol (Lond). 1984;357:109-125.

[55] Burke D, Dickson HG, Skuse NF. Task-dependent changes in the responses to low-threshold cutaneous afferent volleys in the human lower limb. J Physiol (Lond). 1991;432:445-458.

[56] Aniss AM, Gandevia SC, Burke D. Reflex responses in active muscles elicited by stimulation of low-threshold afferents from the human foot. J Neurophysiol. 1992;67:1375-1384.

[57] Fung J, Barbeau H. Effects of conditioning cutaneomuscular stimulation on the soleus H-reflex in normal and spastic paretic subjects during walking and standing. J Neurophysiol. 1994;72:2090-2104.

[58] Yang JF, Stein RB. Phase-dependent reflex reversal in human leg muscles during walking. J Neurophysiol. 1990;63:1109-1117.

[59] Abraham LD, Marks WB, Loeb GE. The distal hindlimb musculature of the cat: cutaneous reflexes during locomotion. Exp Brain Res. 1985;58:594-603.

[60] Duysens J. Reflex control of locomotion as revealed by stimulation of cutaneous afferents in spontaneously walking premammillary cats. J Neurophysiol. 1977;40:737-751.

[61] Edgerton VR, de Leon RD, Tillakaratne N, et al. Use-dependent plasticity in spinal stepping and standing. Adv Neurol. 1997;72:233-247.

[62] Hodgson JA, Roy RR, de Leon R, et al. Can the mammalian lumbar spinal cord learn a motor task? Med Sci Sports Exerc. 1994;26:1491-1497.

[63] Basso DM. Neuroplasticity of descending and segmental systems after spinal cord contusion. Neurology Report. 1998;2:48-53.

[64] Almli C, Finger S. Toward a definition of recovery of function. In: Finger S, LeVere T, Almli C, Stein D, eds. Brain Injury and Recovery: Theoretical and Controversial Issues. New York, NY: Plenum Press; 1988: 1-14.

[65] Andersson O, Forssberg H, Grillner S, Lindquist M. Phasic gain control of the transmission in cutaneous reflex pathways to motoneurones during "fictive" locomotion. Brain Res. 1978;149:503-507.

[66] Waters RL, Yakura JS, Adkins R, Barnes G. Determinants of gait performance following spinal cord injury. Arch Phys Med Rehabil. 1989;70:811-818.

[67] Waters RL, Adkins R, Yakura J, Vigil D. Prediction of ambulatory performance based on motor scores derived from standards of the American Spinal Injury Association. Arch Phys Med Rehabil. 1994;75: 756-760.

[68] Crozier KS, Graziani V, Ditunno JF Jr, Herbison GJ. Spinal cord injury: prognosis for ambulation based on sensory examination in patients who are initially motor complete. Arch Phys Med Rehabil. 1991;72:119-121.

[69] Penrod LE, Hegde SK, Ditunno JF Jr. Age effect on prognosis for functional recovery in acute, traumatic central cord syndrome. Arch Phys Med Rehabil. 1990;71:963-968.

[70] Burns SP, Golding DG, Rolle WA, et al. Recovery of ambulation in motor-incomplete tetraplegia. Arch Phys Med Rehabil. 1997;78:1169-1172.

[71] Visintin M, Barbeau H, Korner-Bitensky N, Mayo NE. A new approach to retrain gait in stroke patients through body weight support and treadmill stimulation. Stroke. 1998;29:1122-1128.

[72] Hesse S, Bertelt C, Jahnke MT, et al. Treadmill training with partial body weight support compared with physiotherapy in nonambulatory hemiparetic patients. Stroke. 1995;26:976-981.

AL Behrman, PT, PhD, is Assistant Professor, Department of Physical Therapy and University of Florida Brain Institute, University of Florida, Box 100154, Gainesville, FL 32510-0154 (USA) ([e-mail]). Address all correspondence to Dr Behrman.

SJ Harkema, PhD, is Assistant Professor, Department of Neurology and Brain Research Institute, University of California at Los Angeles, Los Angeles, Calif.

Both authors provided concept/idea, writing, data collection and analysis, subjects, and facilities/equipment. Dr Harkema provided project management, fund procurement, institutional liaisons, clerical support, and consultation (including review of manuscript before submission). Carlos Arnaiz, Janell Beres, Sheryl Flynn, PT, MHS, Uday Patel, Dorian Rose, PT, MS, Emily Taylor, PT, and Mary Thigpen, PT, MHS, assisted with locomotor training of the subjects. Michele Basso, PT, PhD, and Katherine Sullivan, PT, PhD, provided critical review of the manuscript.

This work was supported, in part, by research grants NS 36854, NS 16333, and MO1 RR00865-19 from the National Institutes of Health.