Intrinsic and Extrinsic Feedback Generate Similar Propulsion but Distinct Biomechanical Strategies During Split-Belt Walking

Hamad K. Bin Shuwayyi^1,2*William Ray Reed³Chen Lin⁴Jonathan D Hill⁵Ann M Varghese⁵Ya-Yu Liang³Christopher P Hurt^1,3*

• ¹Rehabilitation Sciences Program, University of Alabama at Birmingham, Birmingham, United States
• ²Department of Physical Therapy, Prince Mohammed Bin Abdulaziz Hospital, Riyadh Second Health Cluster, Riyadh, Saudi Arabia
• ³Department of Physical Therapy, University of Alabama at Birmingham, Birmingham, United States
• ⁴Louisiana State University Health Shreveport, Shreveport, United States
• ⁵Department of Biomedical Engineering, University of Alabama at Birmingham, Birmingham, United States

Background: Forward propulsion during walking is generated by different joints and biomechanical mechanisms depending on environmental and task demands. Although propulsion can be modulated by feedback, it is unclear whether extrinsic and intrinsic feedback generate similar propulsion or promote different joint-level strategies during split-belt walking. The purpose of this study was to investigate strategies used by non-impaired individuals to generate greater propulsion using different feedback to reach targeted levels of propulsion force during split-belt walking Methods: Fifteen young adults walked on a split-belt treadmill with the dominant leg on the fast belt at their comfortable walking speed and the non-dominant leg on the slow belt at half speed. They performed trials with extrinsic via visual feedback of propulsive force (targeting 5% and 10% body weight) and with intrinsic feedback via a backward resistive force at the center of mass (5% and 10% body weight). Primary outcome was propulsion accuracy, measured as average propulsion error relative to target levels. Secondary analyses examined explanatory biomechanical variables related to propulsion generation. Outcomes were analyzed using two-way repeated measures ANOVA with Bonferroni correction. Results: Participants achieved similar target propulsion across feedback types (p = 0.66). However, biomechanical strategies differed. Visual feedback increased trailing limb angle (TLA) at 5% (p = 0.0011) and 10% (p < 0.0001) and increased ankle moment at 5% (p = 0.0005) and 10% (p < 0.0001). In contrast, backward resistive force increased (BRF) hip moment at 5% (p = 0.0018) and 10% (p < 0.0001), and hip power at both 5% and 10% (p < 0.0001). Ankle power did not differ between feedback types at 5% (p = 0.0754) but was greater under BRF at 10% (p < 0.0001). Conclusion: While both feedback types generate similar propulsion to achieve different target levels during split-belt treadmill walking, they engaged distinct biomechanical strategies. Our results indicate that participants increased TLA and ankle moment under visual feedback. However, they increased hip moment and hip power under BRF, with ankle power adjustments depending on the target level. The findings highlight motor abundance in gait and suggest tailoring rehabilitation strategies in populations with impaired propulsion.