BWSTT in Chronic SCI

Author Year; Country
Research Design
Sample Size
Methods Outcome

Labruyere et al. 2014; Switzerland
Randomized cross-over open label clinical pilot study
Level 1
N= 9

Population: 9 individuals- 5 males and 4 females; SCI ranging from C4 to T11; mean age= 59 ± 11y; months post injury= 50 ± 56m.
Treatment: Participants with a chronic iSCI were randomized to groups 1 or 2. Group 1 received 16 sessions of RAGT (45 min each) within 4 weeks followed by 16 sessions of strength training (45 min each) within 4 weeks. Group 2 received the same interventions in reversed order. Data were collected at baseline, between interventions after 4 weeks, directly after the interventions, and at follow-up 6 months after the interventions. Pain was assessed repeatedly throughout the study.
Outcome Measures: 10 MWT at preferred and maximal speed, walking speed under different conditions, balance, strength, 2 questionnaires that evaluate risk of falling and pain.
  1. There were no significant differences in changes in scores between the 2 interventions, except for maximal walking speed (10MWT), which improved significantly more after strength training than after RAGT.
  2. Pain reduced after both interventions.

Nooijen et al. 2009; USA
Level 1

Population:  All participants had motor-incomplete spinal cord injuries and were at least 1-year post injury; Group 1: mean age = 38.15; T11-C3; Group 2: mean age = 39.47; T9-C4; Group 3: mean age = 41.64; T6-C4; Group 4: mean age = 44.33; L2-C6.
Treatment: 12-week training period. All BWSTT: Group 1 = treadmill with manual assistance; Group 2 = treadmill with peroneal nerve stimulation; Group 3 = overground with peroneal nerve simulation; Group 4 = treadmill with assistance from Lokomat.
Outcome Measures: Cadence, step length, stride length, symmetry index, intralimb coordination, timing of knee extension onset within the hip cycle; all compared to non-disabled controls.
  1. No significant between-SCI group differences. Pooled data were then used to assess the effects of training.
  2. Training significantly improved: cadence, step length and stride of both the stronger and weaker legs.
  3. After training, participants were able to take more steps per minute
  4. There was an interaction effect between step and stride lengths. Post hoc analyses revealed Group 3 had a significantly larger gain compared to group Group 4.
  5. No training effects found on symmetry or coordination.
  6. After training gait outcome measures were more similar to able-bodied controls than they were before training.

Musselman et al. 2009; Canada
RCT with crossover
Level 1

Population: 2 male and 2 female participants, age 24-61, level of injury C5-L1, all AIS-C.
Treatment: All participants received 3 months of BWSTT, then participants underwent 3 months of BWSTT and 3 months of skill training in random order.
Outcome Measures: mEFAP; 10MWT; 6MWT; BBS; ABC.
  1. Improvement of mEFAP with skill training in all participants (average improvement 731.5); improvement also seen with BWSTT in 2 of 4 participants (-1379 and -731 respectively); gains were maintained after training (statistical test for significant was not done)
  2. Results for the 10m and 6-min walk tests improved more with skill training (average 0.10m/s) compared to BWSTT (average 0.02m/s); again, tests for significance was not done
  3. Minor improvements in Berg Balance Scale (9, 0, 10 and 5 points for participants 1, 2, 3 and 4 respectively), and no improvement for ABC

Field-Fote et al. 2005; USA
Level 1
N= 27

Population: 27 males and females; age 21-64 yrs; with incomplete SCI; C3-T10 lesion level; >1 yr post-injury.
Treatment: Randomized to 4 gait training strategies, 45-50 min, 5X/week, 12 weeks: 1) manual BWSTT (n=7); 2) BWSTT+FES (common peroneal nerve) (n=7); 3) BWS overground + FES (n=7); 4) BWS Lokomat (robotic gait device) (n=6).
Outcome measures:  Walking speed over 6 m (short-bout) and 24.4 m (long bout). 
  1. No significant differences between pre- and post-intervention walking speed in the manual BWSTT or BWS Lokomat groups.
  2. However, there was a tendency for participants with initially slower walking speeds (<0.1 m/s) to have a greater percent increase in walking speed (57% to 80%) compared to those with initially faster walking speeds (-19% to 5%)

Lucareli et al. 2011; Brazil
Level 1


Population: 14 males and 10 females with incomplete SCI; mean age 31.5; mean YPI 9.8.
Treatment: Group A – treadmill gait training with body weight support + conventional physiotherapy; Group B – conventional physiotherapy; both groups underwent 30 semi-weekly sessions lasting 30 min each.
Outcome Measures: Spatial temporal gait variables and angular gait variables. 
  1. Group B showed no within group differences for spatial-temporal gait measures. Group A showed within group improvements in gait speed (47%), step length (17%), and cadence (16%).
  2. There were no statistically significant improvements for Group B for any measure.
  3. Group A showed a significantly greater range of motion after intervention compared to Group B for maximum hip extension during stance and maximum plantar flexion during pre-swing. There were no significant group differences after treatment in other angular gait variables.
Effect Sizes: Forest plot of standardized mean differences (SMD ± 95%C.I.) as calculated from pre- and post-intervention data.

Yang et al. 2014; Canada
Level 1
N= 22

Population: 22 participants; 16 males and 6 females; Level of injury between C2 and T12; mean age= 48 ± 13y; years post injury= 5.7 ± 10.5y.
Treatment: Twenty-two participants, ≥7 months post injury, were randomly allocated to start with Precision or Endurance Training. Each phase of training was 5 times per week for 2 months, followed by a 2-month rest.
Outcome Measures: Walking speed- 10 MWT, distance- 6 MWT, skill, confidence- Activities specific balance confidence scale, depression- Centre for Epidemiologic Studies- Depression Scale (before training and monthly afterwards), WISCI-II, SCI-FAP.
  1. Both forms of training led to significant improvements in walking, with Endurance Training inducing bigger improvements in walking distance than Precision Training, especially for high-functioning walkers who had initial walking speeds >0.5 m/s.
  2. The largest improvements in walking speed and distance occurred in the first month of Endurance Training, with minimal changes in the second month of training.
  3. In contrast, improvements in walking skill occurred over both months during both types of training. Retention of over ground walking speed, distance, and skill was excellent for both types of training.

Gorman et al. 2016; USA

Population: 18 individuals chronic motor incomplete spinal cord injury between C4 and L2; >1 y post-injury.
Treatment: Participants were randomized to Robotic-Assisted Body-Weight Supported Treadmill Training (RABWSTT) or a home stretching program (HSP) 3 times per week for 3 months. Those in the home stretching group were crossed over to three months of RABWSTT following completion of the initial three-month phase.
Outcome Measures: Peak VO2 was measured during both robotic treadmill walking and arm cycle ergometry: twice at baseline, once at six weeks (mid-training) and twice at three months (post-training). Peak VO2 values were normalized for body mass.
  1. The RABWSTT group improved peak VO2 by 12.3% during robotic treadmill walking (20.2 ± 7.4 to 22.7 ± 7.5 ml/kg/min, P = 0.018)
  2. Peak VO2 during robotic treadmill walking and arm ergometry showed statistically significant differences.
Effect Sizes: Forest plot of standardized mean differences (SMD ± 95%C.I.) as calculated from pre- and post-intervention data.

Lam et al. 2014; Canada
Level 1
N= 15


Population: 15 individuals- 9 males and 6 females; chronic motor incomplete SCI; 5 AIS C and 10 AIS D; age range= 26-63y; years post injury> 1y;
Treatment: Participants were randomly allocated to BWSTT with Lokomat resistance (Loko-R group) or conventional Lokomat-assisted BWSTT (controls). Training sessions were 45 minutes, 3 times/wk for 3 months.
Outcome Measures: Skilled walking capacity (SCI-FAP), 10 MWT, 6 MWT. Timed Up and Go Test (TUG). All outcome measures were measured at baseline, post=training, and 1 and 6 months follow up. 
  1. Training was well tolerated by both groups, although participants in Loko-R tended to report higher levels of perceived exertion during training.
  2. Participants in the Loko-R group performed significantly better in the SCI-FAP compared with controls at posttraining and in follow-up assessments.
  3. Both groups showed improvements in walking speed (10MWT) and distance (6MWT) with training, but there were no between-group differences.
Effect Sizes: Forest plot of standardized mean differences (SMD ± 95%C.I.) as calculated from pre- to post-intervention data and pre-intervention to retention/follow-up data

Niu et al. 2014; USA
Single centre, unblinded, randomized study
Level 2
N= 40

Population: 40 individuals- 27 males and 13 females; spastic hypertonia in lower extremities
Treatment: Each participant was assigned either to the control or intervention (Lokomat training) group according to a permuted block randomization design. All participants were injured within their cervical or upper thoracic (superior to T10) vertebrae. Each participant received a one-hour training session three times per week for four consecutive weeks; as it took 15-20 mins to set up the participant, the gait training lasted up to 45 mins per session.
Outcome Measures: 10 MWT, 6MWT, Time up and Go (TUG), isometric torque resulting from MVC, Modified Ashworth Score (MAS), EMG, Walking Index for Spinal Cord Injury (WISCI II)
  1. The baseline (i.e. pre-training) measures of MVC torque (Td and Tp) could predict the differential treatment response, i.e., participants with high Tp and Td were more likely to have both high walking capacity and receive significant benefit from Lokomat training.
  2. Lokomat training in participants with low walking capacity did not show significant improvements. By contrast, participants with a high walking capacity at baseline presented a consistent linear trend in time for both speed and functional balance over the 4-week training period.
Gorassini et al. 2009; Canada
Prospective Controlled Trial
Level 2
Population: 17 participants with incomplete SCI, mean (SD) age 43.8(16.5), injury level C3-L1, and 6 AB controls. Participants were divided into 2 groups: those who improved in walking ability (responders, n=9, 4 AIS-C, 5 AIS-D) and those who did not (nonresponders, n=8, 7 AIS-C, 1 AIS-C).
Treatment: BWSTT, on average for mean (SD) 3.3(1.3) days/week for 14(6) weeks.
Outcome Measures: EMG; WISCI II.
  1. Responders had an average WISCI II increase of 4.6pts, compared to no increase in the nonresponders.
  2. The amount of EMG activity increased significantly after training in responders, whereas no change was observed in nonresponders.

Wu et al. 2012; USA
Repeated assessment with crossover
Level 2

Population: 10 participants with chronic SCI (8M 2F); mean (SD) age: 47(7); mean (SD) DOI: 5.8(3.8) yrs; level of injury: C2-T10.
Treatment: Group 1: BWSTT with 4 wks assistance training, then 4 weeks resistance training. Group 2:  BWSTT with 4 wks resistance training, then 4 wks assistance training. Resistance provided by a cable-driven robotic locomotor training system.  Sessions were 45 minutes, 3x/wk x 8 weeks.
Outcome Measures: Primary: self-selected and fast walking speed, 6MWT, BBS. Secondary: muscle strength tests, WISCI II, Physical SF-36, Activities-specific Balance Confidence Scale.
  1. A significant improvement in both walking speed (increased from 0.67(0.20) at baseline to 0.76(0.23) m/s post-intervention) and balance (increased from 42(12) to 45(12)) was observed after robotic treadmill training.
  2. Following robotic training, stride length, step length, and cadence during self-selected walking significantly improved.
  3. There was no significant difference in walking functional gains after resistance versus assistance training, although resistance training was more effective for higher functioning patients.
Behrman et al. 2012; USA
Prospective Cohort
Level 2
Population: 95 participants with SCI (75M, 20F); <1 yr (n=47), 1-3 yrs (n=24), ³3 yrs (n=24) since injury; level of injury: T11 or above; Mean (SD) age: 43(17); median time since injury: 1 year; 31 AIS C, 64 AIS D.
Treatment: At least 20 sessions of the NRN Locomotor Training Program consisting of manual-facilitated BWS standing and stepping on a treadmill and overground.  Training consisted of 1hr of treadmill training, 30 minutes overground assessment, and 15-30 minutes of community reintegration.  Frequency: 5 days/wk for non-ambulators, 4 days/wk for ambulators with pronounced assistance, 3 days/wk for independent walkers. Patients split into phases 1-3 depending on level of ability (higher ability = higher phase).
Outcome Measures: ISNSCI AIS, BBS, 6MWT, 10MWT.
  1. For those who enrolled in phase 1 and were still classified phase 1 after NRN training, no change was seen in BBS, 6MWT or 10MWT scores.
  2. For those who enrolled in phase 1 and were classified phase 2 after NRN training, mean change scores were 1 for BBS, 10 for 6MWT and 0 for 10MWT.
  3. For those enrolled at Phase 1 and classified as Phase 3 after NRN training, mean change scores were 38.5 for BBS, 265.5 for MWT and 0.7 for 10MWT.
  4. For those enrolled in Phase 2 and classified as Phase 2 after training, mean change scores were 7 for BBS, 46 for 6MWT and 0.1 for 10MWT.
  5. For those enrolled in Phase 2 and classified as Phase 3 after training, mean change scores were 15 for BBs, 82.3 for 6MWT and 0.3 for 10MWT.

Buehner et al. 2012; USA
Prospective cohort
Level 2

Population: 225 participants with chronic incomplete SCI (167M, 58F); mean (SD) age=42.5 (15.9); Median DOI=2.45; 57 AIS C, 167 AIS D.
Treatment: NRN Locomotor Training Program.  Training consisted of 1hr of treadmill training, 30 minutes overground assessment, and 15-30 minutes of community reintegration.  Frequency: 5 days/wk for non-ambulators, 4 days/wk for ambulators with pronounced assistance, 3 days/wk for independent walkers.
Outcome Measures: LEMS, pinprick, light touch, 10MWT, 6MWT, BBS.
  1. Significant gains occurred in LEMS scores (Pretraining: 31.85 (13.98); Posttraining: 38.61 (12.29)) but not in sensory scores.
  2. Although 70% of participants showed significantly improved gait speed after locomotor training, only 8% showed AIS category conversion.
  3. Significant gains in gait speed (72%), ambulation distance (74%) and balance (43%) occurred after NRN training regardless of initial AIS classification.

Lorenz et al. 2012; USA
Level 2

Population: 337 participants with SCI (255M, 82F); mean (SD) age: 40 (17); 99 AIS C, 238 AIS D.
Treatment: At least 20 sessions of the NRN Locomotor Training Program. Training consisted of 1hr of treadmill training, 30 minutes overground assessment, and 15-30 minutes of community reintegration.  Frequency: 5 days/wk for non-ambulators, 4 days/wk for ambulators with pronounced assistance, 3 days/wk for independent walkers.
Outcome Measures: BBS; 6MWT; 10MWT.
  1. There was significant improvement on each outcome measure and significant attenuation of improvement over time.
  2. Patients varied significantly across groups defined by recovery status and AIS grade at enrollment with respect to baseline performance and rates of change over time.
  3. Time since SCI was a significant determinant of the rate of recovery for all measures.

Wernig et al. 1995; Germany
Case Control
Level 3

Population: Study 1: 44 participants with chronic paraplegia or tetraplegia. Study 2: 53 participants with chronic paraplegia or tetraplegia.
Treatment: Study 1: BWSTT: 30-60 min, 5x/wk, 3-20 wks (median 10.5 wks). Study 2: 29 participants underwent BWSTT (as in Study 1) versus 24 historical controls that underwent conventional rehabilitation.
Outcome measures: Wernig Walking Capacity Scale.
  1. Study 1: 25/33 initially non-ambulatory could walk after BWSTT. Results of 11 initially ambulatory participants unclear. At 6 months post-training, 18/21 ambulatory patients maintained abilities or improved endurance.
  2. Study 2: 14/18 initially non-ambulatory participants could walk after BWSTT. Only 1/14 initially non-ambulatory in the historical controls learned to walk. Other specific improvements in initially ambulatory participants in either the BWSTT or historical control groups were not clearly described.

Benito Penalva et al. 2010; Spain
Case control
Level 3

Population: 29 motor incomplete SCI patients (24 males, 5 females, mean age 47; Group A < 3 months post-injury (n=16), Group B > 3 months post-injury (n = 13) and 13 healthy volunteers (10 males, 3 females, mean age 32) with pre-test only
Treatment: Gait training using either the Lokomat or Gait Trainer GT1 (based on availability of the system), 20-45 minutes per sessions (5 days a week for 8 weeks).
Outcome Measures: LEMS, WISCI II, 10MWT, H reflex modulation by TMS.
  1. After gait training, there was a significant improvement in LEMS, WISCI and 10MWT for both group A and B, with a significantly greater improvement in 10MWT for group A versus group B.
  2. After gait training, Group A showed significantly greater H reflex facilitation with TMS at 20 ms than Group B (170.7 + 10.2% vs. 125.3 + 5.6%), with no significant differences at 50 and 80 ms.

Fleerkotte et al. 2014; Netherlands
Pre-Post Test
Level 4

Population: 10 individuals- 4 males and 6 females; motor incomplete chronic SCI; 1 AIS B, 5 AIS C, 4 AIS D; mean age= 48.75 ± 11.3y; months post injury= 46.75 ± 41.03
Treatment: Participants participated in an eight-week training program. Participants trained three times per week, for a maximum of 60 minutes per session. The training period was divided in two four-week periods, with one week scheduled for clinical tests in between. During training sessions, rest intervals were introduced if required by the participant or suggested by the therapist. The first training session was used to 1) fit the LOPES to the participant, 2) let participants get used to walking in the device and 3) select their preferred walking speed.
Outcome Measures: 10-meter walking test (10MWT), the Walking Index for Spinal Cord Injury (WISCI II), the six-meter walking test (6MWT), the Timed Up and Go test (TUG), Lower Extremity Motor Scores (LEMS), spatiotemporal, kinematics measures.
  1. Participants experienced significant improvements in walking speed (0.06 m/s, p = 0.008), distance (29 m, p = 0.005), TUG (3.4 s, p = 0.012), LEMS (3.4, p = 0.017) and WISCI after eight weeks of training with LOPES.
  2. At the eight-week follow-up, participants retained the improvements measured at the end of the training period.
  3. Significant improvements were also found in spatiotemporal measures and hip range of motion.

Stevens et al. 2015; USA
Pre-Post Test
Level 4
N= 11

Population: 7 males and 5 females; average age 47.7y; >1y post injury; AIS C and D
Treatment: Participants completed 8 weeks (3 × /week) of UTT. Each training session consisted of three walks performed at a personalized speed, with adequate rest between walks. Body weight support remained constant for each participant and ranged from 29 to 47% of land body weight. Increases in walking speed and duration were staggered and imposed in a gradual and systematic fashion.
Outcome Measures:  Lower-extremity strength (LS), balance (BL), preferred and rapid walking speeds (PWS and RWS), 6-minute walk distance (6MWD), and daily step activity (DSA).
  1. Participants improved in leg strength (57%), balance (39%), preferred walking speed (34%), rapid walking speed (61%), 6-minute walk distance (82%), and DSA (121%) following UTT.

Aach et al. 2014; Germany
Level 4

Population: 6 males and 2 females; mean age 48 ± 9.43 years; years post injury= 97.2 ± 88.4 months; chronic stage of traumatic SCI; incomplete and complete SCI AIS A-D.
Treatment: The participants underwent a BWSTT five times per week using the HAL exoskeleton.
Outcome Measures: Walking distance, speed, time, 10m walk test (10MWT), timed-up and go test (TUG test), 6-minute walk test (6MWT), the walking index for SCI II (WISCI II), AIS with the lower extremity motor score (LEMS), spinal spasticity (Ashworth scale), and the lower extremity circumferences.
  1. Highly significant improvements of HAL-associated walking time, distance, and speed were noticed.
  2. Significant improvements have been especially shown in the functional abilities without the exoskeleton for over-ground walking obtained in the 6MWT, TUG test, and the 10MWT, including an increase in the WISCI II score of three patients.
  3. Muscle strength (LEMS) increased in all patients accompanied by a gain of the lower limb circumferences.

Yen et al. 2013; USA
Post Test
Level 4

Population: 12 participants; traumatic motor incomplete SCI; ASIA D; injuries ranging from C1-T7; mean age= 48 years; years post injury= 5 years.
Treatment: Each person participated in one data collection session, about 2.5h long. We recorded each participant’s maximum voluntary isometric contraction (MVC).  A robotic system provided resistance during the swing phase of gait. The data collection session consisted of three resistance load conditions: light, medium, and heavy.
Outcome Measures: MVC, EMG, stride length, swing time (with and without robotic system)
  1. An increase in the resistance load tended to cause a significant increase in kinematic error size for swing time (p<.001) and stride length (p<.001)
  2. After the robotic system was removed, the aftereffect resulted in a significant increase in stride length for the light (p=.02), medium (p.01) and heavy loads (p=.01)
  3. After the robotic system was removed, the aftereffect resulted in an increase in swing time was observed but the increase was not significant (p>.6)

Sczesny-Kaiser et al. 2015; Germany
Pre-Post Test
Level 4
N= 11

Population: 11 individuals- 7 males and 4 females; traumatic SCI with incomplete or complete paraplegia; mean age= 46.9 ± 2.7y; years post injury= 8.8 ± 2.1y.
Treatment: Eleven SCI patients took part in HAL® assisted BWSTT for 3 months. Each patient was scheduled for a 30min training session 5 times a week for 12weeks, as previously described by our group. Paired-pulse somatosensory evoked potentials (PpSEP) were conducted before and after this training period, where the amplitude ratios (SEP amplitude following double pulses – SEP amplitude following single pulses) were assessed and compared to eleven healthy control participants.
Outcome Measures: 10 MWT, 6 MWT Timed up and Go Test (TUG), Lower Extremity Motor Score (LEMS).
  1. After training, there was a significant increase in 10MWT speed from 0.25 ± 0.05 m/s to 0.5 ± 0.07 m/s (p=.001) and a significant increase in 6MWT from 86 ± 20.86 m to 149.73 ± 20.32 m (p<.001).

Varoqui et al. 2014; USA
Level 4
N= 30

Population: 30 individuals; ambulatory chronic incomplete SCI; mean age= 50.80 ± 2.12y; years post injury= 11.80 ± 2.54y
Treatment: 15 iSCI participants performed twelve 1-hour sessions of Lokomat training over the course of a month. The voluntary movement was qualified by measuring active range of motion, maximal velocity peak and trajectory smoothness for the spastic ankle during a movement from full plantar-flexion (PF) to full dorsi-flexion (DF) at the patient’s maximum speed. Dorsi- and plantar-flexor muscle strength was quantified by isometric maximal voluntary contraction (MVC). Clinical assessments were also performed using the Timed Up and Go (TUG), the 10-meter walk (10MWT) and the 6-minute walk (6MWT) tests. All evaluations were performed both before and after the training and were compared to a control group of fifteen iSCI patients.
Outcome Measures: Active range of motion, maximal velocity peak and trajectory smoothness from full plantar-flexion to full dorsi-flexion at patient’s maximum speed, maximal voluntary contraction (MVC), Timed up and Go (TUG), 10 MWT, 6 MWT, Modified Ashworth Scale (MAS)
  1. For the training group, the 10MWT resulted in a significant increase in mean gait speed of 13.4 ± 2.8% after training (P < 0.05).
  2. For the Control group, there was no significant difference in 10MWT (P = 0.36).

Knikou 2013; USA
Level 4

Population: 14 participants with chronic SCI (10M 4F); 21-55 yrs old; 0.5-11 yrs post-injury; 1 AIS A, 1 AIS B, 4 AIS C, 8 AIS D.
Treatment: All participants received BWS robot-assisted step training with a robotic exoskeleton system (Lokomat). Each participant was trained 1h/day, 5 days/wk.
Outcome Measures: WISCI II; 6MWT; number of sit-to-stand repetitions completed within 30s; TUG; EMG measurements.
  1. BWS robotic-assisted step training reorganized the soleus H-reflex in a functional manner during assisted stepping in people with clinically complete, motor incomplete and motor complete SCI.
  2. Training changed the amplitude and onset of muscle activity during stepping, decreased the step duration, and improved gait speed.
  3. For the AIS C and AIS D group, distance walked in the 6MWT increased after BWS training but not significantly. For the AIS D group, TUG time decreased after BWS training, but again, not significantly.

Harkema et al. 2012; USA
(subacute and chronic)
Level 4

Population: 96 individuals (148 male, 48 female) with incomplete SCI; mean age 41±15 yrs; YPI- <1 yrs (n=101), 1-3 yrs (n=43), >3 yrs (n=52).
Treatment: Locomotor training with three components: (1) 1 hour of step training in the body-weight support on a treadmill environment, followed by 30 minutes of (2) overground assessment and (3) community integration.
Outcome Measures: BBS, 6MWT, and 10MWT.
  1. 168 (86%) patients (66 of 66 AIS grade C, 102 of 130 AIS grade D) scored lower than 45, the reported threshold for risk for falls for the BBS-Patients with AIS grade C SCI had significantly lower scores at enrolment than those with AIS grade D classification – Patients with AIS grade D SCI walked significantly farther than those with AIS grade C SCI
  2. Scores on the BBS significantly improved by an average of 9.6 points.
  3. 6MWT distances and 10MWT speeds of all patients significantly improved by an average of 63m and 0.20m/s, respectively

Yang et al. 2011; Canada
Level 4

Population: 14 males, 5 females; mean age 44±13; >7 months post-injury (mean 5.8±8.9 years); AIS C or D
Treatment: 1 hour/day, 5 days/week of BWSTT until parameters did not progress for 2 weeks (minimum 10 weeks total, mean=18 weeks).
Outcome Measures: 10MWT, WISCI-II, LEMMT, BBS, EMG measurements (tibialis anterior, soleus, quadriceps, hamstrings), movement at the knee and ankles
  1. After training, 17/19 participants improved in duration of walking in a session (mean (SD) 15(11) min), 16/19 improved in treadmill speed (0.14(0.11) m/s), and 16/19 improved in their ability to support their own body weight (18(19)% decrease in body weight support).
  2. 13 participants responded to the treatment; 9 showed improvements of >1 m/s (exceeding the smallest real difference in overground walking speed) and 4 showed improvements <1 m/s but greater WISCI-II scores.
Stevens, 2010; USA
Level 4
Population: 11 participants with incomplete SCI (7M 4F); 23-64 yrs old; 1-28 years post-injury; 9 AIS C, 2 AIS D; all able to walk at least 10 meters with or without an assistive device.
Treatment: People participated in an underwater treadmill training exercise program for 8 weeks. Week 1 consisted of 3 5-minute walks, with scheduled increases in walking speed (10% increase biweekly) and duration (up to 8 minute walks) over the following weeks. Each participant completed 24 training sessions in 8 weeks.
Outcome Measures: lower limb strength (dynamometry); BBS; WISCI II; 10MWT; 6MWT; daily step activity.
  1. Repeated-measures ANOVA demonstrated that participants exhibited significant relative improvements in leg strength (57%), balance (39%), preferred walking speed (34%), rapid walking speed (61%), 6-minute walk distance (82%) and daily step activity (121%) following underwater treadmill training. Effect sizes for these 6 variables ranged from 0,50-0.84, indicating that the magnitude of the training effect was large.
  2. Prior to training, the average difference in strength between the stronger and lower legs was 33%, whereas after training, a 22% mean difference in strength between the stronger and weaker legs was detected.
  3. Relative gains in muscle strength ranged from 32% for the knee flexors to 95% for the hip flexors

Winchester et al. 2009; USA
Level 4

Population: Mean (SD) age = 38.3(13.6); 22 male; 23 participants with tetraplegia, 7 with paraplegia; mean (SD) time since injury = 16.3(14.8) months.
Treatment: Locomotor training, including: robotic assisted BWSTT, manually assisted BWSTT, and over ground walking. 3 times per week for 3 months.
Outcome Measures: WISCI II and 10MWT.
  1. 22 participants showed improvement in walking speed; 8 showed no change post-training.
  2. Pre-training, 16 participants could not walk. Post-training, 5 remained unable to ambulate, 7 recovered ambulation but needed assistance, and 4 recovered independent ambulation.
  3. Step-wise regression analysis showed that time post-injury, voluntary bowel and bladder voiding, functional spasticity, and walking speed before training were the strongest predictors of post-training overground walking speed.

Effing et al. 2006; Netherlands
Level 4

Population: 3 males; age 45-51 yrs; participant diagnosis were AIS C and D; C5-C7 lesion level; 29-198 months post-injury.
Treatment: BWSTT: 30 min, 5x/wk,12 wks.
Outcome measures: Wernig Walking Capacity Scale, gait speed over 7m.
  1. Gait improvements in all participants, indicated either by faster gait speed or higher score in Walking Capability Scale.

Hicks et al. 2005; Canada
Level 4

Population: 14 males and females; age 20-53 yrs; 2 participants with diagnosis of AIS B and 12 participants with diagnosis of AIS C; C4-L1 lesion level; 1.2-24 yrs post-injury.
Treatment: BWSTT: <45 min, 3x/wk, 144 sessions (12 months).
Outcome measures: Wernig Walking Capacity Scale.
  1. 6/14 participants improved in walking capacity, but only 3 maintained improvements at 8 months post-training.
  2. 3/10 initially non-ambulatory participants could walk (with assistance) post-training.

Thomas and Gorassini 2005; Canada
Level 4

Population: Age 29-78 yrs; 4 participants with diagnosis of AIS C and 2 participants with diagnosis of AIS D; C5-L1 lesion level; 2-28 yrs post-injury.
Treatment: BWSTT: < 60 min, 3-5X/week, 10-23 weeks.
Outcome measures: 10MWT, 6MWT, WISCI II.
  1. 5/6 participants improved WISCI II score. Overall significant improvements in 6MWT and 10MWT and improvements correlated with the increase in corticospinal connectivity.

Wirz et al. 2005; Switzerland
Level 4

Population: Age range =16-64 yrs; 9 participants with diagnosis of AIS C and 11 participants with diagnosis of AIS D; C3-L1 lesion level; 2-17 yrs post-injury
Treatment:  BWSTT: <45 min, 3-5x/wk, 8 wks.
Outcome measures:  WISCI II, 10MWT, 6MWT.
  1. 2/20 participants improved WISCI II scores.
  2. Overall increase in 10MWT of mean (SD) 0.11(0.10) m/s (56% improvement).
  3. 15/16 participants improved in 6MWT.

Protas et al. 2001; USA
Level 4

Population: 3 males; age 34-48 yrs; Participant diagnosis was AIS C and D; T8-T12 lesion level; 2-13 yrs post-injury
Treatment: BWSTT: 20 min, 5x/wk, for 12 wks.
Outcome measures: Garrett Scale of Walking, Assistive Device Usage Scale, Orthotic Device Usage Scale, gait speed (5m), gait endurance (5 minutes).
  1. All participants showed an increase in gait speed and endurance.
  2. All participants showed improvement, indicated by the Garrett Scale of Walking or the type of assistive or orthotic devices used.

Wernig et al. 1998; Germany
Level 4

Population: 35 males and females; age 19-70 yrs; C4-T12 lesion level; 1-15 yrs post-injury.
Treatment: BWSTT: 30-60 minutes, 5x/wk, 8-20 wks.
Outcome measures: Wernig Walking Capacity Scale.
  1. 20/25 initially non-ambulatory improved to walking with aids.
  2. 2/10 ambulatory patients improved functional class, but all improved speed and endurance.
  3. At follow-up (0.5-6.5 years later) all ambulatory patients remained ambulatory, with changes only in functional class.


As shown in Table 9, there have been 19 pre-post studies, 9 RCT (Musselman et al. 2009Nooijen et al. 2009Field-Fote et al. 2005Field-Fote & Roach 2011Lucareli et al. 2011Gorman et al. 2016Lam et al. 2014Labruyere et al. 2014) 5 prospective Controlled Trial (Gorassini et al. 2008) and 2 case-control studies (Wernig et al. 1995Benito Penalva et al. 2010) that altogether studied 812 persons with complete and incomplete SCI, with chronicity ranging from 1 to 28 years post-injury (although years of chronicity was not specified in Field-Fote et al. 2011 study). Treatment intensity ranged from 45 to 300 minutes per week, and treatment duration lasted between 3 and 48 weeks. Based on the stated primary outcome measure of each study where data was available, about 70% of all participants across these studies showed some improvement following treatment (Musselman et al. 2009Gorassini et al. 2009Hicks et al. 2005Yang et al. 2011Winchester et al. 2009Protas et al. 2001Thomas & Gorassini et al. 2005Effing et al. 2006Wernig et al. 1995). In the Harkema et al. 2012 study, 88% of patients had responded to locomotor training treatment, but this study included participants that had been injured less than one year.

All studies generally show improvements in overground walking capacity, whether locomotor training was provided with a treadmill or performed over ground, body-weight support, or involved other variations on walk-based therapies (e.g. over ground training with obstacles, robot-applied resistance. Alternative gait retraining therapies or modified approaches to BWSTT for chronic SCI are being introduced (Musselman et al. 2009; Stevens 2010; Wu et al. 2012; Lam et al. 2014Yang et al. 2014). Musselman et al. (2009) and Yang et al (2014) compared BWSTT with over ground ‘precision’ skilled walking training. The skilled walking training consisted of task-specific practice (without body weight support) of various gait tasks, such as stair climbing, obstacle crossing, and walking along sloped surfaces. BWSTT was better than precision over ground training in improving walking distance. Surprisingly, both training groups were comparable in improving walking skill. Wu et al. (2012) demonstrated a new cable-driven robotic device to apply resistance against leg movements during BWSTT. Participants were randomized (in a cross-over design) to receive robotic resistance or assistance BWSTT. Although there were no significant differences in outcomes between the two modalities, there was some indication that robotic resistance enabled greater gains in over ground walking speed in people who tended to have better initial ambulatory capacity; conversely, robotic assistance seemed to enable greater gains in walking speed in those who were initially slower walkers. More recently, Lam et al. (2014) showed that training with Lokomat-applied BWSTT with resistance yielded better improvements in skilled walking function that were retained even 6 months post-intervention, vs. Lokomat-assisted BWSTT.


There is level 1b evidence from 1 RCT (Field-Fote & Roach 2011) that different strategies for implementing body weight support gait retraining all yield improved ambulatory outcomes in people with chronic, incomplete SCI, except for robotic-assisted treadmill training which showed little change in walking speed. It is recommended that therapists may choose a bodyweight support gait retraining strategy based on available resources (Field-Fote & Roach 2011).

There is level 4 evidence from pre-test/post-test studies (Behrman et al. 2012Buehner et al. 2012Harkema et al. 2012Lorenz et al. 2012Winchester et al. 2009Hicks et al. 2005Wirz et al. 2005Thomas and Gorassini 2005Protas et al. 2001Wernig et al. 1998) that BWSTT is effective for improving ambulatory function in people with chronic, incomplete SCI.