Background/Purpose: Falls are a common and costly complication following stroke. While residual walking and balance deficits contribute substantially to long-term disability, of greater concern for this population is the incidence of falls. Between 40% and 70% of individuals fall within their first year post-stroke. Individuals who fall have increased incidence of serious adverse outcomes, including fractures, depression, and mortality. A primary goal of stroke rehabilitation is to improve mobility in the presence of motor, balance, and visual-spatial deficit. The conundrum all stakeholders must face is that increased mobility may increase risk for falls, whereas limiting walking and general mobility will lead to a multitude of deficits associated with inactivity, including recurrent stroke. Because of these serious consequences, there is an urgent need in the rehabilitation of Veterans with stroke to both identify those who are at fall risk and develop intervention strategies that will reduce fall incidence while increasing mobility. Backward walking has recently emerged as both a potential predictor of fall incidence as well as an intervention modality to prevent future falls. However, it has not yet been assessed under the rigors of a controlled trial. We will conduct a randomized controlled trial with three specific aim to examine 1) the effectiveness of Backward Walking Training (BWTraining) early post- stroke in decreasing falls, 2) the timing of BWTraining delivery to increase Backward Walking speed (BWSpeed), and 3) the relationship between BWSpeed and prospective fall incidence. Study Design: This is a randomized single-blind controlled trial. Subjects: One-hundred twenty-eight individuals, 2 months post-stroke, will participate. Additional study criteria include: 1) ambulatoy with gait speed < 0.8 m/s; 2) community-dwelling; 3) Berg Balance Scale < 42; 4) absence of other neurological conditions; 5) discharged from physical therapy services; 6) stable cardiac status; 7) absence of lower extremity orthopedic impairments or pain that limits gait ability. Methods: Following baseline assessment, participants will be randomized to BWTraining at 2 months (Immediate group) or 1-year post-stroke (Delayed group). The intervention consists of 18 sessions (3X/week for 6 weeks) of BWTraining: 20-30 minutes of step training using a Body Weight Support and Treadmill system (BWST) followed by 15 minutes of overground gait training delivered by a physical therapist-led team. Participants will walk backward overground with assistance as warranted to transfer stepping skills from the BWST environment to overground. The Immediate group will be followed prospectively for 1-year and compared to the Delayed group to determine the effectiveness of BWTraining in decreasing falls (Specific Aim #1). BWSpeed post-intervention will be compared between groups to assess the timing effect of BWTraining (Specific Aim #2). The Delayed group will be followed prospectively for 1-year prior to BWTraining, to determine if BWSpeed at 2-months is a predictor of fall incidence in the first year post-stroke (Specific Aim #3). Outcome Measures: Fall incidence 2-months to 1-year post-stroke is our primary outcome measure. Secondary measures include forward and backward walking gait speed, dynamic balance measured by the Functional Gait Assessment, fall self-efficacy measured by the Activities-Balance Confidence Scale and gait kinematics. Data Analysis Plan: The total number of falls per patient month for the immediate and delayed intervention groups will be calculated, and using a nonparametric ratio estimate, point and interval estimates for the average hazard for each group will be determined. The ratio of the two individual group estimates, 95% confidence limits and P-values will be obtained. Forward stepwise ordinal logistic regression will be used to determine the independent relationship between BWSpeed at 2-months post-stroke and 1-year fall incidence of those in the Delayed intervention group.

Public Health Relevance

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. Contributing to these costs is the incidence of falls Falls are a costly complication for Veterans with stroke as they lead to an increased incidence of fractures, depression, and mortality. New strategies are needed to help Veterans post-stroke regain their ability to safely walk without increasing their risk of falling as well as readily idetify those who are a fall risk. This study addresses both of these needs as it will 1) investigate a new treatment approach, backward walking training, to determine if it will decrease fall incidence in the first year post-stroke and 2) determine if backward walking speed early after a stroke can identify those that are at risk for future falls.

National Institute of Health (NIH)
Veterans Affairs (VA)
Non-HHS Research Projects (I01)
Project #
Application #
Study Section
Musculoskeletal Health & Function (RRD2)
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Veterans Health Administration
United States
Zip Code