Walking phenotypes are different between people, to the point that we can recognize people by the way they
walk. Despite these evident individual differences, researchers will obtain averages across a specific
population, such as healthy individuals or individuals with walking impairments due to musculoskeletal or
neurological injury, to derive group-level descriptors of walking function. These averages, which eliminate
important between-individual differences, inform rehabilitation research studies that prescribe the same
intervention for all participants. Precision Medicine is an innovative approach put forth by the NIH to consider
individual patients’ biological, environmental, and lifestyle differences to inform the prescription of treatment.
Despite the stronghold and success that precision medicine has had in pharmaco-genomics and oncology, it
has yet to be implemented to fine-tune interventions for walking behaviors. Our goal is to identify individual-
specific walking phenotypes and their underlying joint and muscle level impairments to effectively guide clinical
decision-making long-term. To achieve this goal, we will use data analysis pipelines that use machine learning
to leverage the wealth of clinical and laboratory data used to characterize function after injury. This approach
will allow us to identify the specific muscles and joints responsible for each walking phenotype, which can
serve as rehabilitation intervention targets. Our KL2 funded project took the first step in this direction: using
stroke survivors as a model of pathological walking behavior, we identified four distinct walking phenotypes,
which point at deficits in either walking speed, balance, propulsion, or shock absorption. What remains to be
determined are the individual joints and muscles responsible for these distinctive walking phenotypes (Aim 1),
whether these impairments can be detected early after injury to develop treatment strategies in the early
stages (Aim 2), and whether clinical measures within the International Classification of Functioning, Disability,
and Health, can be used to draw inference on mechanisms of impairment (Aim 3), allowing easy
implementation of our findings in clinical settings. We will achieve these aims via secondary analyses of
longitudinal data collected as part of our KL2 project to identify the joint and muscle impairments that
characterize each walking phenotype. We will harness our results in an R01 proposal to assess the effects of
generalized vs. phenotype-specific prescription of walking interventions, using the intervention targets
identified here. Identification of individualized intervention targets could improve the efficacy of walking
interventions and more generally improve mobility, and associated participation, health, and well-being, all
aligned with the NIH mission.
