The primary objective of this study is to achieve successful walking skills using exoskeletal walking devices over the course of 36 sessions in 3 months at specific velocities and distances in people with chronic SCI who are wheelchair dependent for community mobility. The secondary objectives are to determine if this amount of exoskeletal walking is effective in improving bowel function and body composition in the same patient population. The exploratory objectives are to address additional questions concerning the retention or non-retention of the positive changes, the effects of the increased physical activity from this intervention on vagal tone, orthostatic tolerance, lipid profile, total testosterone, estradiol levels, and quality of life (QOL).

A Phase III randomized clinical trial (RCT) will be performed using a crossover design and employing an exoskeletal-assisted walking intervention. The experimental arm will be compared to a usual activities (UA) arm, as the control, in 64 persons with chronic SCI (>6 month post injury) who are wheelchair-dependent for outdoor mobility in the community. The WALK arm will consist of supervised exoskeletal-assisted walking training, three sessions per week (4-6 h/week) for 36 sessions for their second 12-week period. The UA arm will consist of identification of usual activities for each participant, encouragement to continue with these activities and attention by study team members throughout the 12-week UA arm. These activities will be recorded in a weekly log. The investigators hypotheses are that 1) this exoskeletal intervention will be successful in training ambulatory skills in this patient population, 2) the exoskeletal intervention will be better than a control group in improving body composition, bowel function, metabolic parameters and quality of life in the same population.

Study Design Research design: A Phase III randomized clinical trial (RCT) will be performed using a crossover design and employing an exoskeletal-assisted walking intervention. The experimental arm will be compared to a usual activities (UA) arm, as the control, in 64 persons with chronic SCI (>6 month post injury) who are wheelchair-dependent for outdoor mobility in the community. Eligible participants will be randomized (within site) to one of two groups for 12 weeks (three months): Group 1 (n=32) will receive exoskeletal-assisted walking (WALK) first for 12 weeks then crossover UA for a second 12 weeks; Group 2 (n=32) will receive UA first for 12 weeks then cross-over to the WALK arm for 12 weeks of training. The WALK arm will consist of supervised exoskeletal-assisted walking training, three sessions per week (4-6 h/week) for 36 sessions for their second 12-week period. The UA arm will consist of identification of usual activities for each participant, encouragement to continue with these activities and attention by study team members throughout the 12-week UA arm. These activities will be recorded in a weekly log. A fixed answer format will be used to capture this information.

Rationale for intervention to be studied: This research design has several advantages. Group 1 will serve as the intervention follow-up to assess retention or non-retention of change due to the intervention on the outcome variables. Group 2 will serve as a lead-in to assess stability of the outcome variables prior to the intervention. Additionally, because of tremendous variability in the SCI population and difficulty with case-control matching, the cross-over design will help to control for variability between participants because each participant will serve as their own control. Veterans and nonveterans with SCI generally do not receive further structured rehabilitation once they have completed the acute and sub-acute phases of their rehabilitation. As such, the usual lifestyle activities have been chosen as the control for this study. Dr. Spungen has discussed the optimal control for an exoskeletal-assisted walking program. She has consulted with the Chiefs of several SCI Services in both the VA and non-VA rehabilitation hospitals. A case-controlled, matched RCT study would be strongest with two interventions: exoskeletal-assisted walking compared with any other form of physical activity such as robotically assisted body weight supported treadmill training (Lokomat), body weight supported treadmill training, arm ergometry etc. However, there are two considerations that make this type of study not feasible. One is that it would be extremely difficult to case-control match for independent variables within site, and the second is that most participants would likely not agree to participate if there was a 50% chance they would not get the exoskeletal-assisted walking arm. In order to control for an attention effect, participants in the UA arm will keep weekly activity logs and be contacted by phone calls from the study team to receive attention and encouragement to continue participation in the UA of their choice. In order to control for attention aspects of this type of study, participants will also make in person twice monthly visits to their sites during the UA arm. During these visits, study team staff will review their UA logs and administers the 10Q BFS and BSS.

Because the achievement of improved secondary medical outcomes is likely dependent on the ability to walk in the exoskeletons, the primary outcome will be achievement of mobility skills in the exoskeletal devices. It has not been demonstrated in a large sample that people with SCI can use the device as described by the pilot data. Secondary outcomes will be change in fat mass and bowel function. Exploratory outcomes will be change in cardiovascular and autonomic measures and quality of life surveys [using the newly validated SCI quality of life (SCI-QOL) tool for bowel, bladder and five emotional domains]21,50.

Description of walking in the ReWalk: The ReWalk exoskeletal system requires the user to actively shift their body appropriately over their legs in order to maintain balance on the stance leg allowing the swing leg to clear the floor. This is accomplished with coordinated movements involving the upper extremity muscles in the trunk, arms, shoulder, back, neck, and head for the body to maintain balance dynamically as the system assists the legs to move in a normal walking pattern. Balance is maintained with the use of Loftstrand crutches. A mode selector, which is worn on the wrist, is used to select "walk mode" once the person is stable during standing. Following walk mode selection, the user shifts onto the left leg and slightly forward so that the right leg is offloaded. The tilting/leaning action triggers the device to swing the right leg forward and the user weight shifts so that their trunk is over the right leg as they step onto the right foot. This will then initiate the left leg to swing forward; the use weight shifts onto it and continuous walking is achieved if this coordinated weight shifting and stepping onto each leg is continued. The unit will stop walking in two ways. If the user does not shift their weight onto the front foot, then the ReWalk will time out after two seconds and return to the standing position. The second way is if the person does not shift their weight appropriately to have the swing leg clear the floor. This will cause the leg to have some additional external force that the ReWalk will sense and cause the system to return to standing mode. The stepping and walking gait pattern of the ReWalk is very similar to normal walking51. The gait pattern settings in the ReWalk can be preset in order to increase or decrease the step length, speed of the step and/or the amount of foot clearance the person will have during walking, permitting a velocity of 0.10 to 0.80 m/s.

In order to successfully accomplish walking with the ReWalk, considerable coordination between the upper extremity muscles in the trunk, arms, shoulder, back, neck, and head is needed in conjunction with dynamic standing balance. A specific pattern of trunk positioning, crutch placement and weight shifting is needed. To prepare for walking, the user must attain a stable standing position with the crutches placed slightly in front of him/her. To begin walking, the user must shift their weight onto the left foot to unweight the right foot. The right foot steps forward first, and this is initiated when the user slightly leans forward, by bringing the crutches forward, away from the stepping leg to allow it to swing forward and clear the floor. As the stepping leg moves forward and extends toward the ground, the user shifts weight forward, stepping onto that leg and then sequentially brings the crutches forward for the next step. Immediately after the weight is shifted, the user must straighten his/her posture with their body perpendicular to the ground before leaning forward to initiate the next step. This is similar to a normal walking pattern. As the leg contralateral to the initial step leg swings forward, the participant must shift the crutches away from that leg to allow it to step forward. This process is repeated for continued walking.

Description of walking in the Ekso: The users are progressed through a series of steps to participate in over ground ambulation. The first method, called FirstStep. This is when the trainer triggers the step by using a controller while verbally cueing the participant that the device is taking a step. The second method, called ActiveStep is when the user to initiates the stepping action by pressing a button on a pair of instrumented crutches for each step. The last method, called ProStep, is when the user shifts their weight lateral and forward onto the front foot, until sensors indicate when an appropriate shift has been performed. The stepping pattern can be adjusted to change the step time and step length. The trajectory of the foot allows for foot clearance similar to a marching pattern. The device has the ability for provide some abduction and plantar/dorsi flexion positioning. This device has the ability to alter the stepping, which would allow the user to take faster or slower steps with a longer or shorter stride length, thus permitting a walking velocity of 0.10 to 0.45 m/s. Stopping is achieved by holding a stationary position and by not activating sensors. The Ekso allows for varying robot-assist levels that can be decrease or increased bilaterally. The robot assistance can operate will full or variable assist, dependent on the level of function of the individual.

Inclusion Criteria:

• Males and females, between 18-70 years old
• Traumatic or non-traumatic paraplegia >6 months in duration
• SCI motor deficit at any level
• Unable to ambulate faster than 0.17 m/s on level ground with or without an assistive device and are wheelchair-dependent for community mobility
• Height 160 to 190 cm (63-75 in or 5'3" to 6'3" ft)
• Weight <100 kg (<220 lb)
• Able to hold the crutches
• Able to sign informed consent.

Exclusion Criteria:

• Diagnosis of neurological injury other than SCI including:
• Multiple sclerosis, Stroke, Cerebral Palsy, Amyotrophic lateral sclerosis, Traumatic Brain injury, Spina bifida, Parkinson's disease, or
• Other neurological condition that the study physician considers in his/her clinical judgment to be exclusionary
• Severe concurrent medical disease, illness or condition
• Recent lower extremity fracture within the past 2 years
• DXA results indicating a t-score below -3.5 at the femoral neck or the total proximal femur bone and knee BMD <0.60 gm/cm2
• Diagnosis of heterotropic ossification of the lower extremities which affect range of motion or proper measurement of BMD measurements
• Significant contractures defined as flexion contracture limited to 35º at the hip and 20º at the knee
• Untreated hypertension (SBP>140, DBP>90 mmHg)
• Symptomatic orthostatic hypotension with standing that does not resolve after attempts at upright posture that were made over several days, and standing by the participant is deemed to pose a health risk, as determined by a physician, because of symptomatic orthostatic hypotension
• Systemic or peripheral infection
• Atherosclerosis, congestive heart failure, or history of myocardial infarction;
• Trunk and/or lower extremity pressure ulcers
• Severe spasticity (defined by an Ashworth score of >4.0 or clinical impression of the study physician or physical therapist)
• Significant contractures defined as flexion contracture limited to 25º at the hip and knee
• Diagnosis of heterotropic ossification of the lower extremities which affect range of motion or proper measurement of BMD measurements
• Psychopathology documentation in the medical record or history of that may conflict with study objectives
• Pregnancy and/or lactating females
• Brain injury with score on mini-mental status examination less than 26
• Diagnosis of coronary artery disease that precludes moderate to intense exercise
• Deep vein thromboses in lower extremities of less than 6 months duration
• Other illness, that the study physician considers in his/her clinical judgment to be exclusionary.