Adaptation Strategies to Split-Belt Treadmill Walking in Healthy Young Adults

Abstract

Human gait must be highly adaptable in order to maintain dynamic stability as we move through varying external environments. Split-belt treadmills provide an opportunity to create challenging locomotor patterns and have recently been used to evaluate adaptation patterns during walking. Healthy young adults (HYA) adapt to split-belt treadmill walking (SBTW) by adjusting spatiotemporal patterns asymmetrically to match the task demands of the treadmill; however, while a basic adaptation pattern has been established among HYA, it remains unknown whether more specific adaptation strategies exist. PURPOSE: To characterize the different individual spatiotemporal adaptation strategies HYA use during SBTW. METHODS: Ten HYA (23 ± 4 yrs) walked on a split-belt treadmill, first with the belts moving together at 50% of the participant's self-selected "fast" velocity for two minutes (tied) and then with the dominant limb belt sped up to the "fast" velocity (split) for 10 minutes. Kinematic data of the last 30 seconds of each trial were collected using a 12-camera Vicon Nexus motion capture system (Vicon Nexus, Oxford, UK). Stride length and swing/stance ratio were calculated from manually-labeled gait events. Z-scores were calculated for the bilateral symmetry change (BSC) of these two parameters from the symmetrical condition to the asymmetrical condition. BSC was calculated as the difference in the percent change of bilateral symmetry (fast limb parameter divided by the slow limb parameter) from the tied condition to the split condition. The five fastest walkers (self-selected "fast" walking speed > 1.4 m/s) and five slowest walkers (self-selected "fast" walking speed ≤ 1.4 m/s) were separated into two groups and independent-samples t-tests were run to analyze between-group differences for BSC of swing/stance ratio and BSC of stride length. RESULTS: The slower walkers showed lower z-scores for BSC of swing/stance ratio (-.75 ±.44 vs.75 ±.81, p =.004) and higher z-scores for BSC of stride length (.54 ±.40 vs. -.80 ±.65, p =.006) when compared to the faster walkers. CONCLUSION: There appear to be at least two different spatiotemporally-driven adaptation approaches to SBTW and these may be related to walking speed.