The aim of this study is to compare the effectiveness of 6-weeks of reactive balance training (REACT) with and without neuromuscular electrical stimulation (NMES) to paretic lower limb muscles on biomechanical, clinical, neuromuscular and neuroplastic outcomes of reactive balance control. This project is a Phase-I study and incorporates a double-blinded, randomized controlled trial design.

Methods: Forty-six individuals with chronic stroke will be recruited and screened for determining their eligibility for the study. Once enrolled, they will be randomized into either of the two groups: intervention group (23 participants) and control group (23 participants). Both groups will undergo series of pre-training assessments which includes a postural disturbance in the form of a slip- or trip-like perturbations and walking tests in laboratory environment. After the pre-training assessment, individuals will undergo 6-weeks of training (2 hour per session, 2 sessions per week). The intervention group will receive NMES with the REACT training and the control group will receive ShamNMES. NMES will be applied to the different muscle groups of the paretic lower limb using an advanced software which is able to synchronize muscle activation with the time of perturbation onset and according to the phases of gait. After training, both groups will again be tested on all the assessments performed pre training.

This study will help us understand the immediate therapeutic and mechanistic effects of REACT+NMES and inform stroke rehabilitation research and clinical practice. Our study will provide foundational evidence for future use of NMES to implement clinically applicable neuromodulation adjuvants to reactive balance training, which could be leveraged for designing more effective future interventions for fall-risk reduction.

1.0 Background/Scientific Rationale Interventions such as conventional balance and exercise training constitute a major part of stroke rehabilitation and improve volitional balance control and gait in people with chronic hemiparetic stroke (PwCHS). However, they seldom target reactive balance (compensatory postural responses such as stepping) that forms the first line of defense while recovering from a balance loss. Reactive balance in PwCHS is affected by deficits in perturbation-evoked neuromuscular and biomechanical responses especially during gait. Further, previous research has shown that stability and adaptions to repeated perturbations is more affected on paretic compared to non-paretic limb. Thus, paretic limb deficits are postulated to be key contributors of falls in ambulatory PwCHS. Perturbation-based reactive balance training (REACT) is widely recognized as an intervention that reduces falls by improving fall-resisting skills. In the past five years, there is a 3-fold increase in perturbation training research in PwCHS (mostly low impairment). Thus, limited evidence exists for PwCHS with severe motor impairment who might not show similar tolerance or learning abilities.

Complementing REACT with interventions known to facilitate paretic limb performance and motor learning (i.e., neuromuscular electrical stimulation, NMES) can improve therapeutic effects of REACT and hence its clinical translation for PwCHS and other populations that could benefit from fall-risk reduction. While it is established that REACT programs and NMES can induce motor learning in behavioral variables, there is limited evidence on neuroplastic changes and exact neural mechanisms underlying these behavioral changes (especially during REACT). Similar to the precision medicine approach, modifiable causative factors, contributors, and mediators to falls must be targeted when designing effective falls prevention interventions that reduce training times and/or facilitate the inclusion of persons with high impairment.

This project aims to describe whether a specific pattern of lower limb muscle stimulation could modify the recovery response after an unexpected perturbation in the form of a slip and/or trip in individuals with stroke. Also, this study aims to examine the effectiveness of 6-weeks of reactive balance training (REACT) with and without neuromuscular electrical stimulation (NMES) to paretic lower limb muscles on biomechanical, clinical, neuromuscular and neuroplastic outcomes of reactive balance control.

2.0 Objectives/Aims

The specific aims of this study are below:

Aim 1: To examine effects of synchronous REACT+NMES on reactive balance control and clinical outcomes among people with chronic stroke with moderate to severe motor impairment.

H1.1: REACT+NMES will induce greater improvement in biomechanical outcomes of reactive balance (reactive stability, limb support) resulting in fewer laboratory falls post-training than REACT+ShamNMES (at 6 weeks).

H1.2: The improvements in reactive balance control in REACT+NMES will translate to greater improvement in clinical outcomes of balance (mini-BEST test), gait (10m walk test) and falls-efficacy (Activities specific Balance Confidence scores) than REACT. H1.3: The improvements in reactive balance control will also translate to reduced falls during overground gait-slips after REACT+NMES.

Aim 2: To examine neuromuscular and neuroplastic effects of REACT+NMES in PwCHS with moderate to severe motor impairment.

H2.1. REACT+NMES will induce greater neuromuscular (muscle synergy #s and activations) and neuroplastic (perturbation evoked potentials- PEPs) changes post-training than REACT (at 6 weeks).

H2.2: Baseline PEP amplitude and training-induced neuroplastic changes in PEP's will correlate with the training-induced improvements in biomechanical and neuromuscular responses.

3.0 Research Design This study trial employs a two-arm, double-blinded, randomized controlled trial (RCT) design This study will examine efficacy and feasibility of REACT-NMES intervention compared to REACT+ShamNMES among PwCHS with moderate to severe motor impairment (Aim 1 and 2) A sample size of 46 chronic stroke survivors will be enrolled, undergo initial screening and pre-training assessment, and then randomized into two groups (intervention and control). Next, both groups will undergo 6-weeks of in-lab reactive balance training (2x/week, total 12 sessions). After training, participants in both groups will undergo a post-training assessment, which will include all the tests performed in the pre-training assessment.

Study overview: All participants will undergo the following procedures.

• Session 1 (Week 1): Initial screening (2 hours)
• Session 2 (Week 2): Pre-test (total 4 hours)
• Session 3-15 (Week 3-8): Training sessions (2 hours/session, 2 times/week for 6 weeks) (total: 24 hours)
• Session 16 (Week 9): Post-test (total 4 hours)

Inclusion Criteria:

• Age group: 18-90 years.
• Presence of hemiparesis.
• Onset of stroke (> 6 months).
• Ability to walk at least for 2 minutes on the treadmill with or without ankle foot orthosis.
• Can understand and communicate in English.
• Cognitively and behaviorally capable of complying with the regimen (Mini-Mental State Examination > 25/30).
• No history or recent use (i.e., past 6 weeks) of any Neuromuscular electrical stimulation device to leg muscles during walking (e.g., Bioness, Walkaide).

Exclusion Criteria:

• Subjects will not proceed with the test if any of the following occurs at baseline measurement: 1) HR > 85% of age-predicted maximal heart rate (HRmax) (HRmax = 220 - age), 2) systolic blood pressure (SBP) > 165 mmHg and/or diastolic blood pressure (DBP) > 110 mmHg during rest, or 3) oxygen saturation (measured by pulse oximeter) < 95% during rest.
• Body weight of more than 250 lbs.
• Spasticity (Ashworth scale > 2).
• Loss of protective sensations on the paretic leg (indicated by inability to perceive the 5.07/10 g on Semmes-Weinstein Monofilament) or inability to feel the NMES.
• Severe osteoporosis (indicated by T score < -2)
• Cognitive impairment (indicated by Mini-Mental State Exam score<25)
• Global Aphasia (indicated by <71% on the Mississippi Aphasia Screening Test).
• Subjects with Chedoke McMaster Leg Assessment Scale score (> 4).

Subjects will be blinded to group existence and assignment and outcome assessors will be blinded to group assignment