Precise exercise prescription to improve walking capacity in people with chronic stroke

Abstract

The residual effects of a stroke impact 795,000 people in the United States each year, and it is estimated stroke prevalence will increase by nearly 50% by 2035. Following stroke, a vast majority of individuals have significant difficulty walking which impairs their ability to independently complete household and community activities. The inability to independently walk leads many with chronic stroke to fall into a cycle of physical inactivity, which exacerbates the already elevated risk of secondary complications such as secondary stroke, cardiovascular disease, and/or mortality. Stroke-related disability leads to exorbitant economic health care costs, estimated to be $103.5 billion annually. ☐ Moderate-to-high intensity walking exercise interventions have shown promise in being an effective tool to improve walking speed and walking endurance in people with chronic stroke. However, current exercise guidelines for people with chronic stroke have imprecise recommendations for multiple parameters of exercise dose (volume, intensity, duration, etc.). This imprecise exercise dosing in combination with assuming all individuals with chronic stroke present to rehabilitation with the same characteristics, leaves it up to chance which individuals with chronic stroke will achieve improvements in their walking capacity and which will not. As the United States is faced with significant increases in stroke-related disability costs in the next decade, it is imperative there is precise exercise prescription to improve walking function in people with chronic stroke. ☐ Using advanced analytical techniques and a methodical and rigorous study design, this dissertation aimed to improve the precision of exercise dosing to optimize walking capacity outcomes in people with chronic stroke. To do this, this dissertation considered the heterogeneity of the chronic stroke population and the complexity of the exercise dose-response relationship. All clinical and training data used for this dissertation is from the Promoting Recovery Optimization of Walking Activity in Stroke randomized control trial (NCT 02835313). ☐ In Aim 1, we found there are homogeneous groups of people with chronic stroke who have vastly different clinical profiles. Despite significantly differing on measures of walking endurance, walking speed, balance self-efficacy, comorbidity burden, depressive symptoms, body mass index, and daily step activity, there were no differences between groups on the exercise dose attained across a fast-walking treadmill intervention. The overall exercise dose was quantified several different ways, yet the results remained the same: there were individuals in each group who attained higher doses of exercise and individuals who attained lower doses of exercise. This result was interesting as it has generally been thought that individuals with lower measures of walking capacity or physical health at baseline are likely to be unable to engage in similar amounts of moderate-to-high intensity exercise interventions. ☐ In Aim 2, it was found that exercise dose did not mediate the relationship between a group’s clinical profile and their change in walking speed or walking endurance. As there was no relationship between the clinical profile (the predictor) and exercise dose (the mediator), no mediation can be present. This result, in combination with individuals across the entire sample attaining various exercise doses agnostic to clinical profile, led to a modified analysis. Following a pooling of the entire sample, we tested the strength of 2 different metrics of exercise intensity to predict a change in walking speed or walking endurance after accounting for several important covariates. Surprisingly, heart rate was not a significant predictor of walking capacity outcomes, but training speed was. This is important as heart rate has nearly always been used to quantify exercise intensity in walking interventions in people with chronic stroke. These results support previous work suggesting that training speed may be a more precise metric of exercise intensity to use in walking exercise interventions in people with chronic stroke. ☐ In Aim 3, we found that exercise intensity, quantified as training speed, was not a significant predictor of who would attain a small or moderate meaningful change in walking speed or walking endurance. This result is important as researchers and clinicians often use minimal clinically important differences (MCID) to interpret if the outcome of their intervention(s) were meaningful to their patient. It may be that one parameter of exercise dose may not be sufficient to predict a dichotomized result of a small or moderate change in a walking capacity outcome. However, there is no consensus on the best approach to determining an MCID for rehabilitation measures, and uncertainty about whether these thresholds accurately assess a meaningful change in walking endurance in people with chronic stroke. This is an important consideration as it raises methodological concerns in using MCIDs for interpretation of responses to interventions. Further work is needed to determine if there is a metric with improved prediction accuracy to determine which individuals with chronic stroke will meet an MCID in walking speed or walking endurance. ☐ The overarching goal of this dissertation was to improve the precision of exercise prescription to optimize walking capacity outcomes in people with chronic stroke. Exercise prescription is a complex and nuanced process influenced by individual variability, the numerous ways to manipulate exercise dose, and the countless factors which may impact one’s ability to engage in exercise on a specific day or across time. The results of this dissertation demonstrate that when a specific outcome or goal of an exercise intervention is determined for a specific population, we can further refine exercise prescription while accounting for individual factors.