Approximately 15,000 Veterans are hospitalized for stroke each year with new cases costing an estimated $111 million for acute inpatient, $75 million for post-acute inpatient, and $88 million for follow-up care over 6 months post-stroke. The investigators have previously established the effectiveness of a backward walking training program to improve gait and balance in post-stroke Veterans. To best serve Veterans in this era of personalized medicine, there is a current need to determine the appropriate training dose as well as which post-stroke Veterans would most benefit. This study addresses both needs as it will 1) test responses to two different doses (18 vs. 27 sessions) of backward walking training and 2) assess brain activity, measured by magnetic resonance imaging, before and after training intervention to determine its ability to predict rehabilitation response as well as brain mechanisms of behavioral change.

This prospective, single-blind, randomized controlled trial will enroll individuals between 2- and 4-months post-stroke and is designed to address the study's three Specific Aims (See Figure below). The study will be approved by an ethics review board and all participants will provide written informed consent. Baseline pre-intervention assessment (Assessment A) will consist of clinical gait and balance assessments and resting state Functional Connectivity MRI and functional MRI. Brain imaging at Assessment A will test the hypotheses that rs-FC and fMRI can predict BWTraining intervention response (Specific Aim 2). Following Assessment A, participants will be randomized to receive 1) 18 sessions or 2) 27-sessions of BWTraining 3x/week for six or nine weeks respectively. Participants in the 18-session group will wait 3 weeks following randomization to begin training such that Assessment B will take place at approximately the same time post-stroke for all participants. This will assure that all participants have had the same duration of time post-stroke after study enrollment to experience spontaneous recovery. Gait and balance measures at Assessment B will be compared to Assessment A to test Specific Aim #1. MRI brain measurements at Assessment B will be compared to those at Assessment A to test the hypotheses of exploratory Aim #3. Assessment C, at six weeks post-intervention will evaluate short-term and Assessment D, at 6-months post-intervention will evaluate long-term retention gains in gait and balance. The investigators will recruit 54 individuals post-stroke to allow up to a 10% attrition rate and still provide a sufficient participant pool (n=48) to address the Specific Aims.

Inclusion Criteria:

• Berg Balance Scale < 45
• Self-selected 10 meter gait speed < 0.8 m/s
• Diagnosis of unilateral stroke
• > 2 months < 4 months post-stroke
• Able to ambulate at least 10 feet with maximum 1 person assist
• Medically stable

Exclusion Criteria:

• Presence of neurological condition other than stroke

• Serious cardiac conditions (hospitalization for myocardial infarction or heart surgery within 3 months, history of congestive heart failure, documented serious and unstable cardiac arrhythmias, hypertrophic cardiomyopathy, severe aortic stenosis, angina or dyspnea at rest or during activities of daily living)

  • Anyone meeting New York Heart Association criteria for Class 3 or Class 4 heart disease will be excluded

• Severe arthritis or orthopedic problems that limit passive ranges of motion of lower extremity (knee flexion contracture of -10 degrees, knee flexion ROM < 90 degrees, hip flexion contracture > 25 degrees, and ankle plantar flexion contracture > 15 degrees)

• Severe hypertension with systolic greater than 200 mmHg and diastolic greater than 110 mmHg at rest

• Pain upon ambulation

• Receiving physical therapy services for mobility and/or gait

• Living in a skilled nursing facility

• Any MRI contraindication, including but not limited to the presence of metal, MR sensitive implanted medical devices, or claustrophobia

Outcome Assessors will be masked to group assignment. Outcome Assessors will not deliver the intervention and will not be present in the building when the intervention is being delivered. Participants will be instructed to not reveal their group assignment to the outcome assessor.