BWSTT in Chronic SCI
Discussion
BWSTT vs. Physical Therapy (Usual Care)
Most research that we found testing BWSTT vs. ‘usual care’ (i.e., some combination of range of motion training, stretching, strengthening, and/or some overground walking training) found that BWSTT was superior for walking-based outcomes. The most common improvement was in lower limb muscle strength (i.e., LEMS scores), but some studies also found improvements in functional walking (i.e., SCIM-III mobility scores or WISCI-II scores). Midik et al. (2020) found that participants in usual care group and the BWSTT group both improved with regular physiotherapy, but the improvement in the LEMS scores and SCIM-III mobility scores was significantly higher in the RAGT group at the end of the fifth week and at three months (p=0.017; p=0.038). In a relatively large RCT, Tarnacka et al. (2023; N=105) found that people with incomplete SCI who performed RAGT plus conventional PT had significantly more improvements in motor scores [2.58 (SE 1.21, p < 0.05)] and functional walking [WISCI II – 3.07 (SE 1.02, p < 0.01)] than those who did conventional PT and balance training.
Conversely, two studies by Piira et al. (2019a; 2019b) comparing Lokomat training to usual care found that all participants improved LEMS scores with no differences between groups in either walking speed or walking endurance.
BWSTT vs. OGT
When comparing BWSTT vs. OGT in people with SCI, the results are mixed, suggesting that neither method is superior to the other, but that the decision to use BWSTT may be more based on partially helping people to walk or providing a safer method to do so.
In a larger RCT, Field-Fote and Roach (2011; N=64) enrolled participants into a 5 days/week, 12 week walking training program, using either: treadmill-based training with manual assistance (TM), treadmill-based training with stimulation (TS), overground training with stimulation (OG), or treadmill-based training with robotic assistance (LR). For speed, they found no significant between-group differences; however, distance gains were greatest with OGT. They calculated effect sizes of each training method and found that they were largest for the OG group for speed and distance (d=0.43 and d=0.40, respectively). Authors suggested that a robotic training method that requires more active participation would possibly yield different results. In one small RCT, Wu et al. (2018) found that gains in 6MWT distance were greater for the robotic training group than that for treadmill-only training group (P = 0.03), although gains in self-selected and fast walking speeds were not significantly different between the two groups (P = 0.06; P = 0.12 respectively).
BWSTT vs. Other Types of BWSTT
There are multiple variations in how BWSTT or robotic gait training can be delivered; for example, the amount of support can be a smaller or larger percentage of body-weight, a BWSTT-assistance or -resistance mode can be used, or a hybrid assistive limb (HAL) wearable exoskeleton that assists walking and lower limb movements via real-time actuator controls may be combined with BWSTT (Kubota et al. 2018).
Three small RCTs (Lam et al. 2015; Wu et al 2016; Wu et al. 2012) tested different variations of BWSTT where one group used a robotic training method with resistance, and the other condition used robotic training with assistance. All three studies reported that both Lokomat-resistance and Lokomat-assistance groups improved walking speed (10MWT) and walking distance (6MWT), but that there were no between-group differences. Wu et al. (2012) hypothesized that robotic resistance training would be more effective for higher functioning patients, based on their results that people with better initial walking capacity had greater gains in walking speed over the course of their trial.
Multiple studies have shown walking training benefits using a hybrid assistive limb (HAL) wearable exoskeleton in people with chronic SCI. HAL uses power units and biofeedback sensors mounted on and around hip and knee joints to detect bioelectric signals and movement information to assist with (Koda et al. 2023). Although the HAL exoskeleton is not a walking aid for use during daily activities, it could represent a temporary training tool to improve functional mobility without the device in patients with SCI. Four pre-post trials from the same clinical setting reported significant improvement in gait performance following a program of 3-5 days per week of treadmill- progressing to overground-based BWST (Behrman et al. 2012; Buehner et al. 2012; Harkema et al. 2012; Lorenz et al. 2012).
Which Walking Training Type is Best?
Several systematic reviews with or without meta-analysis have been conducted to establish the effectiveness of different gait training interventions for people with SCI and the results have been mixed as to whether there is a ‘best’ walking training method (Aguirre-Güemez et al. 2019; Arroyo-Fernández et al. 2024; Huang et al. 2024; Mehrholz et al. 2017; Nam et al. 2017; Wan et al. 2024; Wall et al. 2015; Yang et al. 2022; Zhang et al. 2022). Systematic reviews and meta-analyses conducted by Mehrholz et al. (2017), Cheung et al. (2017), Nam et al. (2017), and Alashram et al. (2021) have indicated that neither BWSTT nor robot-assisted BWSTT increased walking ability, strength, or independence more than OGT and other forms of physiotherapy in patients with SCI. However, other systematic reviews and meta-analyses have shown positive results regarding these effects of RAGT (e.g., Lokomat) compared to different therapeutic interventions (such as conventional therapy, no intervention, or strength training, among others) (Aguirre-Güemez et al. 2019; Arroyo-Fernández et al. 2024; Fang et al. 2020).
In a network meta-analysis including 15 RCTs (and 497 participants), Yang et al. (2022) compared the effectiveness of three strategies (BWSTT, RAGT and BWSOGT) for ambulatory improvements. RAGT was found to be significantly more favorable than the control intervention (i.e., versus conventional gait training, such as sit-to-stand, weight shifting, walking, turning, and stand-to-sit), whereas BWSTT and BWSOGT provided similar effects and did not result in significant differences compared with the control interventions (Yang et al. 2022). An RCT by Niu et al. (2014) showed that Lokomat training in participants with lower walking capacity (i.e., longer TUG, lower 10MWT and shorter 6MWT) at baseline did not show significant improvements, in contrast with participants with a higher walking capacity who improved walking speed over the 4-week training period. It is reasonable to expect that walking training therapy works better and faster for people who are already better at walking.
There has been some specificity established by research in walking training intervention types; it would depend on what skills you are hoping your patient will develop. In a network meta-analysis, Zhang et al. (2022) showed that wearable EAW were most likely to be superior for improving walking speed (10MWT; the probability ranking first: EAW=89%; Lokomat=47%) but that Lokomat training was mostly likely to improve functional walking (WISCI-II: probability of ranking first: Lokomat=73%; EAW=63%).
Who For? How Much? How Long?
As with most research in SCI, there is a great degree of variability in what is tested, to the point where it is difficult to determine what is evidence-based and how you would build a program for your patient. For example, the duration of walking training in studies for people with SCI has a large range. In all studies we found, treatment session times ranged from 60 (Lucareli et al. 2011) to 600 (Behrman et al. 2012; Buehner et al. 2012; Lorenz et al. 2012; Harkema et al. 2012) min per week, and treatment duration lasted between 5 (Midik et al. 2020) and 24 (Piira et al. 2019b) weeks. In a systematic review and meta-analysis, Wan et al. (2024) found 8 studies testing RAGT and its effects on improving lower limb muscle strength; they found that studies that implemented the intervention for longer than 6 weeks found significant improvements in LEMS, but not for those who tested shorter durations of RAGT training.
Another variable in walking training for people with SCI is what percentage of body-weight support is most ideal. A systematic review by Ettema et al. (2024) found that in 33 studies involving 156 people with SCI the BWS levels ranged from 17-78% and the median level was 30%, suggesting it is the most common starting point in RAGT studies in neurological populations. We found one small RCT (N=20) that randomly assigned 20 males with incomplete SCI to receive BWSTT twice per week for 6 weeks at either 30% or 40% body-weight support (El Semary & Daker 2019). They found that people in the 40% group had superior improvements in walking speed (89.36% vs. 23.84%, p = 0.001), step length (17.23% vs. 0.89%, p = 0.001), stride length (51.81% vs. 13.66% p = 0.001), and in cadence (16.07% vs. 4.69%, p = 0.009).
Partially because SCI is known as a ‘low-frequency’ condition, it can be difficult to get enough participants into one study to test its effects on people with different levels of injury. In a prospective controlled trial, Grasmücke et al. (2017) tested BWSTT using the HAL exoskeleton in four different subgroups of people with SCI: incomplete paraplegia, incomplete tetraplegia, complete paraplegia, and complete tetraplegia (though all participants needed to have some residual motor function of the hip and knee extensor and flexor muscle groups to trigger and control the exoskeleton). After 12 weeks of 5 times per week BWSTT, all participants’ walking ability increased, as evidenced by improvements in walking speed (mean 47% faster 10MWT), in walking endurance (mean 50% greater distance in 6MWT), and 43% of participants were less dependent on walking aids than before starting training. Surprisingly, there were no significant differences in walking outcomes between all subgroups of completeness or SCI level of injury.
Other prospective controlled trials using HAL found that, although people with all SCI types and classifications can benefit from walking training, that those with incomplete injuries at lower levels and with lower AIS classifications of severity tend to improve more. Okawara et al. (2020) classified participants into ‘low’ and ‘high’ walking groups, and there were significant differences between participants in the high/low groups at enrollment; the low group had more people with cervical level injuries and people classified as AIS B. In the high group, there was a significant improvement in 10MWT time (134.0 to 88.3 s, p = 0.01), speed (0.26 to 0.34 s/m, p < 0.01) and number of steps (44.8 to 36.5 steps, p = 0.05) (this decrease in number of steps indicates an extension of step length). Additionally, no participants in the low group in this study were able to complete the 10MWT at any time point. Buehner et al. (2012) showed that people with AIS C improved significantly after walking training in LEMS, but participants in AIS D significantly improved in LEMS, 6MWT and 10MWT. Harkema et al. (2012) also found that people with AIS D walked significantly further than those with AIS C.
A standard protocol for locomotor training has been developed for people with SCI by the NeuroRecovery Network (NRN) designed to be applied to people who have nonprogressive SCI above the 11th thoracic level with an AIS grade of C or D (Morrison et al. 2012). The standardized locomotor training occurs 3-5 times per week, addressing three components – step/stand retraining (using body-weight support and manual facilitation with physical therapists), overground walking training (evaluating the transfer of step/stand training to level ground walking, improvement in posture and walking skills), and community integration (instruction helping the patient perform daily activities in home and community environment). It is recommended to continue with the program as long as progress is being made in any of the three components. The NRN protocol may be a useful starting point for anyone with SCI when beginning walking training in rehabilitation.
Conclusions
There is level 1 evidence (from 2 RCTs: Piira et al. 2019a; Piira et al. 2019b) that BWSTT with manual assistance or Lokomat® training does not improve walking or LEMS more than usual care in patients with chronic and motor incomplete SCI.
There is level 1 evidence (from 1 RCT: Field-Fote & Roach 2011) that different strategies for implementing body weight support gait retraining (i.e., including BWSTT with various assists and FES-assisted overland training) all yield improved ambulatory outcomes and strength in people with chronic, incomplete SCI, except for robotic-assisted treadmill training, which showed little change in walking speed. It is recommended that therapists choose a body weight support gait retraining strategy based on available resources (Field-Fote & Roach 2011).
There is level 1 evidence (from 1 RCT: Alexeeva et al. 2011) that different locomotor interventions (structured rehab program individualized for each participant; BWS ambulation on a fixed track; and BWS ambulation on a treadmill) provide significant improvements in maximal walking speed and muscle strength, regardless of training method, in patients with chronic and incomplete SCI.
There is level 1 evidence (from 1 RCT: Labruyère & van Hedel 2014) that four weeks of a strength training program provides larger improvements in 10MWT at maximal speed compared to a RAGT with Lokomat in patients with incomplete and chronic SCI.
There is level 2 evidence (from 6 prospective controlled/cohort trials: Buehner et al. 2012; Behrman et al. 2012; Grasmücke et al. 2017; Lorenz et al. 2012; Okawara et al. 2020; Sawada et al. 2021) and level 4 evidence (from 3 pre-test/post-test studies: Harkema et al. 2012; Jansen et al. 2017b; Winchester et al. 2009) that BWSTT (with different approaches) is effective for improving ambulatory function in people with chronic and incomplete SCI.
