Post-stroke Perturbation Training

Proactive and Reactive Perturbation Training to Reduce Falls and Improve Gait Stability in People With Chronic Stroke

Following a stroke, many individuals have a high risk of falls, which can negatively influence quality of life. Unfortunately, current treatments have not effectively addressed this problem. This study investigates whether two methods of delivering mechanical perturbations during walking have the potential to improve post-stroke walking balance and reduce real-world fall incidence.

A novel device will be used to apply controlled mediolateral perturbations to participants' trunk as they walk on a treadmill, requiring reactive responses to avoid a loss of balance. All perturbations will have a duration of 200 ms, and will be applied immediately after onset of the swing phase with either the paretic or non-paretic leg. If in a given training session, a participant does not experience any losses of balance, the perturbation magnitude will be increased in the next training session. Perturbations will begin at 3% body weight (BW), and will progress in increments of 3% BW up to a maximum of 15% BW. The structure of the Reactive perturbation session will be identical to that used for Proactive perturbation training sessions, including overground trials, treadmill trials without perturbations, and Perturbed and Catch trials in which Perturbations are delivered.

A novel device will be used to perturb the relationship between pelvis motion and step width by applying mediolateral forces to participants' legs. Participants will first perform a 3-minute overground walking trial at their self-selected speed, in which they do not interact with the perturbation device. Participants will then perform a 3-minute treadmill walking trial at their self-selected speed, in which they again do not interact with the perturbation device. Participants will then perform a series of alternating 3-minute Perturbed and Catch trials, all at their self-selected speed. For the Perturbed trials, perturbations will be delivered in each step throughout the trial. For the Catch trials, perturbations will cease for the final minute of walking. A total of 5 Perturbed and 5 Catch trials will be performed. Finally, participants will perform a 3-minute overground walking trial at their self-selected speed, in which they do not interact with the perturbation device.

A fall will be defined as an event in which a participant loses their balance and comes to rest on the ground, floor, or lower level. Falls will be measured during the 12-week periods preceding and following the intervention, using 2-week calendars on postcards to be sent to the investigators.

Participants will be instructed to walk overground along a 10-meter path at the speed they would normally use around their house of the store. Speed will be measured for the middle 6-meters of the path.

We will calculate the partial correlation between the pelvis displacement at the start of each step, and step width at the end of the step, accounting for pelvis velocity.

We will calculate the partial correlation between the pelvis displacement at the start of each step, and the mean magnitude of gluteus medius activity during the first half of the swing phase, accounting for pelvis velocity.

We will calculate the partial correlation between the pelvis velocity at the start of each step, and the magnitude of the change in mediolateral center of pressure location under the stance foot during the step, accounting for pelvis displacement.

We will calculate the partial correlation between the pelvis velocity at the start of each step, and the mean magnitude of peroneus longus muscle activity during the single support stance phase of this step, accounting for pelvis displacement.

We will calculate the partial correlation between the pelvis displacement at the start of each step, and the time integral of the total ground reaction force under the trail leg during the double support phase preceding this step, accounting for pelvis velocity.

We will calculate the partial correlation between the pelvis displacement at the start of each step, and the mean magnitude of medial gastrocnemius activity during the double support phase preceding this step, accounting for pelvis velocity.

We will identify the largest mediolateral magnitude perturbation that participants can experience without a loss of balance.

We will identify the largest mediolateral magnitude perturbation that participants can experience without a loss of balance.

We will calculate the absolute difference in the range of frontal plane angular momentum during a step, relative to an unperturbed step. This will be calculated for the perturbed step and the subsequent recovery step.

We will calculate the absolute difference in the range of frontal plane angular momentum during a step, relative to an unperturbed step. This will be calculated for the perturbed step and the subsequent recovery step.

Inclusion Criteria: Experience of a stroke at least 6 months prior to participation Self-reported experience of a fall in the previous year, and/or a fear of falling Gait speed of at least 0.2 m/s Ability to walk on a treadmill without a cane or walker Provision of informed consent Exclusion Criteria: Evidence of cerebellar damage Resting blood pressure higher than 220/110 mm Hg History of unstable cardiac arrhythmias, hypertrophic cardiomyopathy, severe aortic stenosis, angina or dyspnea at rest or during activities of daily living Preexisting neurological disorders or dementia Legal blindness or severe visual impairment History of DVT or pulmonary embolism within 6 months Uncontrolled diabetes with recent weight loss, diabetic coma, or frequent insulin reactions Orthopedic injuries or conditions (e.g. joint replacements) in the lower extremities with the potential to alter the gait pattern

De-identified participant data consisting of the clinical assessments and biomechanical walking data collected in this study will be shared through the IPCSR website.