Jasmine Cash1, Shraddha Srivastava1,2, John Kindred1,2, Bryant Seamon1,2, Steven Kautz1,2, Mark Bowden1,2. 1Medical University of South Carolina, Charleston, SC. 2Ralph H. Johnson Veterans Affair Medical Center, Charleston, SC.

BACKGROUND: Current biomechanical measures of ankle dorsiflexion (DF) during post-stroke walking have limited potential for clinical translation and weak associations with clinical measures of walking function. The purpose of this study was to determine whether ankle angular velocity (AAV) and acceleration (AAA) have test-retest reliability and stronger linear associations with clinical measures of walking function in persons post-stroke compared to current biomechanical measures of ankle DF, peak foot clearance (PFC) and peak DF angles (PDA). METHODS: Retrospective analyses were performed using 62 chronic (>6 months) stroke individuals (18-85 years). Clinical assessments included: lower extremity Fugl-Meyer (FMA-LE), Dynamic Gait Index (DGI), and six-minute walk test (6MWT). Participants performed three 30 second trials at their self-selected walking speed (SSWS) on an instrumented split-belt treadmill. PFC, PDA, peak AAV, and peak AAA during the first half of paretic swing were calculated for each stride and averaged across strides and trials. Spearman’s correlation was used to quantify the relationship between clinical assessments and PFC, PDA, AAV, and AAA. An intraclass correlation coefficient (ICC) was used to quantify test-retest reliability of AAV and AAA with a two-way mixed effect model (n=23). RESULTS: AAV was positively associated with clinical measures (DGI, r=0.51; FMA-LE, r=0.56; 6MWT, r=0.59; SSWS, r=0.63)(p<0.001). Similar positive correlations were observed between AAA and clinical measures (DGI, r=0.57; FMA-LE, r=0.55; SMWT, r=0.69)(p<0.001), (SSWS, r=0.79)(p<0.001). PFC had a positive correlation (SSWS, r=0.70)(p<0.001), moderate correlations (DGI, r=0.46; 6MWT, r=0.57)(p<0.001), and a weak correlation (FMA-LE, r=0.37)(p=0.003) with clinical measures. DF had a weak positive correlation (FMA-LE, r=0.39; SMWT, r=0.30)(p<0.019) with clinical measures. AAV (ICC=0.968) and AAA (ICC=0.947) demonstrated excellent test-retest reliability (ICC 0.9 to 1). CONCLUSION: AAV and AAA can be used post-stroke as task-specific measures of ankle DF during walking, as they are valid and reliable measures, and demonstrated stronger correlations to clinical measures. PFC and PDA are confounded by other motor control deficits post-stroke, such as hip and knee contributions and soft tissue contracture respectively, possibly indicated by variability in correlation strength across clinical measures. Future work should validate other methods of obtaining AAA (i.e. with accelerometers) in a similar manner.

This document is currently not available here.