Learn more about robotics and rehabilitation at the CABRR website.

From 2002 to 2007, NRH headed up a multi-center clinical trial that compared the use of roboticassisted treadmill Lokomat® gait training to conventional gait training in sub-acute stroke patients. The study was sponsored by the National Institute on Disability and Rehabilitation Research (NIDRR) and conducted in partnership with the Rehabilitation Institute of Chicago.

A total of 63 patients were enrolled in the study, in which each subject received twentyfour, 90-minute sessions of gait training.

Thirty-three subjects were trained on the Lokomat, while 30 subjects received gait-specific therapy with a physical therapist. Outcome measures were evaluated at baseline, after 12 and 24 sessions, and at a three-month follow-up, and included self-selected over-ground walking speed, distance covered in six minutes, balance, lower limb function, overall health and well-being, and activities of daily living.

The original research hypothesis was that Lokomat trained subjects would experience greater gains in walking ability than those trained using conventional gait training approaches because the volume of steps practiced within a single training session is greater. The device promotes symmetric stepping patterns, and these patterns are consistent. But results indicate that this was not the case.

Subjects in the conventional therapy group experienced greater gains in walking speed than those trained on the Lokomat. Specifically, average improvements in walking speed for subjects in the conventional group were 200 percent greater than improvements demonstrated by the Lokomat subjects, while improvements in distance walked over six minutes were 66 percent greater in the conventional group compared to the Lokomat group. Most other measures were not statistically different between the two groups.

Recent additional research studies at NRH that ran in parallel with the multi-center study are helping to explain the results.

When walking in the Lokomat, the pelvis and trunk are highly constrained so that subjects do not practice trunk stabilization, a behavior essential in over-ground walking. Furthermore, because the device resides on the outside of the subject’s legs, arm swing is not possible in most individuals with upper limb impairments.

Finally, we have found that while subjects demonstrate near normal movement patterns, since they are guided by the Lokomat, the forces and muscle activation patterns they produce are often inappropriate for the task. Without an “error signal,” patients do not make corrections to their control strategies.

While the results of this clinical study favor conventional gait training over the Lokomat, they should only be interpreted for sub-acute stroke patients who fit the criteria of the study. For example, subjects included only patients with cortical and subcortical strokes (none with brainstem or cerebellar strokes), and the participants were within six months of their stroke, had walking speeds between 0.1-0.6 m/s and had minimal cognitive deficits.

Perhaps some of the stated limitations with the device for sub-acute stroke patients would actually be benefits for individuals in the more acute stages of their injuries. Until further studies are completed in different patient populations, the appropriateness of Lokomat training should be considered on a case-by-case basis.