Optimizing Gait Rehabilitation for Veterans With Non-traumatic Lower Limb Amputation (GEM)

The population of older Veterans with non-traumatic lower limb amputation is growing. Following lower limb amputation, asymmetrical movements persist during walking and likely contribute to disabling sequelae including secondary pain conditions, poor gait efficiency, impaired physical function, and compromised skin integrity of the residual limb. This study seeks to address chronic gait asymmetry by evaluating the efficacy of two error-manipulation gait training programs to improve gait symmetry for Veterans with non-traumatic lower limb amputation. Additional this study will evaluate the potential of error-manipulation training programs to improve secondary measures of disability and residual limb skin health. Ultimately, this study aims to improve conventional prosthetic rehabilitation for Veterans with non-traumatic amputation through gait training programs based in motor learning principles, resulting in improved gait symmetry and lower incidence of long-term disability after non-traumatic lower limb amputation.

The focus of this study is to improve walking symmetry in order to optimize walking ability and reduce disability for Veterans with non-traumatic lower limb amputation (LLA). Over 80% of current LLAs are non-traumatic, resulting from complications of pathologies, such as diabetes mellitus and peripheral artery disease. Despite current declines in total amputation rate among Veterans, the population with non-traumatic LLA is growing. For example, from 2000 to 2004 the relative amputation rate decreased by 34%. However, due to an increase in the number of Veterans with diabetes during the same period, the population of Veterans with diabetes and initial LLA increased by 23%. Following LLA, excessive gait asymmetry is common. Asymmetric gait characteristics are of critical importance as excessive asymmetry may increase the severity of disability experienced by people with non-traumatic LLA and contribute to secondary pain conditions (low back pain and osteoarthritis), poor gait efficiency, declines in physical performance, and compromised skin integrity of the residual limb. Compared to individuals with traumatic LLA, the poor gait performance of those with non-traumatic LLA is further compounded by older age, lower premorbid function, presence of comorbidities, frequent wound development, and delayed healing in the residual limb. While improving gait symmetry is a goal of conventional prosthetic rehabilitation, persistence of gait asymmetry for years after LLA highlights the ineffectiveness of current rehabilitation practices in achieving this goal. As a means of improving gait symmetry, this study aims to determine the efficacy of error-manipulation gait training using two approaches (error-augmentation and error-correction) compared to current standard-of-care in a three-arm randomized controlled trial. Error-manipulation gait training intervention will be delivered in eight training sessions (2x/week, 4 weeks) with 54 Veterans (18 per intervention group, 18 in control group) who have non-traumatic, unilateral, transtibial LLA. It is unclear which, if either, form of error-manipulation gait training is efficacious for improving persistent gait symmetry in Veterans with non-traumatic LLA. Error-augmentation gait training is a promising and novel intervention that involves exaggerating an existing movement error to force the neuromuscular system to correct the error. While this form of gait training improved gait symmetry in small studies of individuals with chronic stroke or traumatic amputation, it has yet to be evaluated in older Veterans with non-traumatic LLA. In contrast, error-correction training involves reducing movement errors by overcorrecting for asymmetry. Both error-augmentation and error-correction gait training are based upon motor learning principles of distributed practice, task specificity, and feedback. Each of these error-manipulation interventions have potential advantages over traditional gait training following LLA which involves repeated bouts of walking with minimal feedback on movement quality and is often unsupervised. Therefore, the primary aim of this study is to determine the efficacy of error-manipulation gait training to improve gait symmetry. A secondary aim is to evaluate signals of efficacy for improved secondary measures of physical function. Lastly, this study will explore changes to residual limb skin health and prosthesis socket fit following error-manipulation gait training. The unique use of motor learning principles in error-manipulation gait training to improve gait symmetry addresses the problem of chronic gait asymmetry following non-traumatic LLA. The results of this study will advance rehabilitation knowledge and provide necessary evidence for the clinical translation of gait training protocols based in motor learning principles for the at-risk population of Veterans with non-traumatic LLA.

Outcome assessors will be blinded to group assignment and participants will be reminded at the time of testing to not disclose any aspects of the training sessions. The Principal Investigator will also remain blinded to group assignment. A co-investigator, who is not involved in participant testing/intervention, will manage randomization and maintenance of group assignment codes.

A 4-week, 8 session, treadmill-based gait training program, with error-augmentation of step asymmetry delivered on a split-belt treadmill. Each training session will adhere to the same schedule. During the training blocks on the treadmill, the belt under the limb with the shorter step length will be set at 3/4 of the pre-intervention over-ground self-selected walking speed while the belt under the limb with the longer step length will be set to 1/2 of the fast belt speed (2:1 ratio between belts).

A 4-week, 8 session, treadmill-based gait training program, with error-correction of step asymmetry delivered with an auditory metronome signal while walking on a treadmill. During each training block, the metronome will be set to overcorrect stance time asymmetry through use of asymmetrical metronome tones, 2:1 ratio.

A 4-week, 8 session, treadmill-based supervised walking program. The active comparator group will participate in a supervised treadmill walking program of the same frequency and duration, to the two experimental groups.

Split-belt treadmill training to enhance between-limb asymmetry during treadmill walking, which is intended to force compensation and correction of step asymmetry during treadmill walking. The error-augmentation gait training program is delivered in 8 sessions over 4 weeks.

Metronome will be set to cue participants to overcorrect between-limb step asymmetry during treadmill walking, through use of asymmetrical metronome tones in a 2:1 ratio. The error-correction gait training program is delivered in 8 sessions over 4 weeks.

An active comparator intervention, in which a supervised treadmill walking program is delivered without attempt to correct step asymmetry. The supervised walking program is delivered in 8 sessions over 4 weeks.

Measurement of step length symmetry by overground walking on GAITRite mat and kinetic and kinematic measurements using 3-D motion capture system.

Change from baseline (pre-intervention) to intervention end, persisting at 12 weeks after intervention end.

Change from baseline (pre-intervention) to intervention end, persisting at 12 weeks after intervention end.

Change from baseline (pre-intervention) to intervention end, persisting at 12 weeks after intervention end.

Self-reported assessment of disability using a scale from 0 to 100, with 0 representing no disability, and 100 representing full disability.

Change from baseline (pre-intervention) to intervention end, persisting at 12 weeks after intervention end.

Change from baseline (pre-intervention) to intervention end, persisting at 12 weeks after intervention end.

Measurement of limb-socket interface by rating of prosthetic socket fit using a scale from 0 to 10, with 0 representing the most uncomfortable socket fit and 10 representing the most comfortable socket fit.

Change from baseline (pre-intervention) to intervention end, persisting at 12 weeks after intervention end.

Inclusion Criteria: Unilateral, non-traumatic, transtibial amputation Diabetes mellitus and/or peripheral artery disease Able to ambulate in the community without assistive device Step length asymmetry during walking (>1.0 asymmetry index.) 6 months to 10 years since amputation Exclusion Criteria: Unstable heart condition Including unstable angina, uncontrolled cardiac dysrhythmia, acute myocarditis, hypertension, and acute pericarditis Acute systemic infection Active cancer treatment Traumatic or cancer related amputation etiology

Participant data will be made available at time of study completion, per VA recommendations and approval.

Kline PW, So N, Fields T, Juarez-Colunga E, Christiansen CL. Error-Manipulation Gait Training for Veterans With Nontraumatic Lower Limb Amputation: A Randomized Controlled Trial Protocol. Phys Ther. 2021 Nov 1;101(11):pzab192. doi: 10.1093/ptj/pzab192.