Abstract
BACKGROUND: Postural instability can occur in the later-stages of Parkinson's disease (PD). The clinical pull-test is scored on a 0-4 scale on the Unified Parkinson's disease rating scale (UPDRS), with postural instability scored 2 or higher. This ordinal scale does not adequately track progression in early-PD or predict development of postural instability. RESEARCH QUESTION: To develop a test that quantifiably measured the backward stepping response on the pull-test in early-PD. METHODS: Participants (35 controls and 79 PD participants) were prospectively enrolled in this study. Participants stepped backwards with each shoulder pull at four strengths on an instrumented gait mat. Four spatiotemporal parameters (reaction-time, step-back-time, step-back-distance, step-back-velocity) were quantified using Protokinetics Movement Analysis Software. Spatiotemporal pull-test parameters were compared to standard PD measures using linear regression and correlation coefficients. Repeated measures analysis was used to determine group differences in pull-test parameters. In a subset of participants repeated testing was performed and Bland-Altman plots were used to determine reproducibility of the pull-test parameters. RESULT: Step-back-distance and step-back-velocity were inversely related to motor UPDRS and freezing of gait questionnaire scores. PD participants had shorter step-back-distance than controls adjusted for age and sex. Repeat assessments in 16 participants, on average 0.7 years apart, showed good agreement on most of the quantified parameters. SIGNIFICANCE: The backward stepping response in PD participants was quantifiable, reproducible, and related to disease severity and could be used to quantify progression towards postural instability in early-PD.