Population-specific reliability and validity of select impairment measures for individuals with Parkinson's disease

The purpose of this dissertation was to investigate the reliability and validity of select measures of gait, balance, and coordination when used with individuals with Parkinson's disease (PD). These measures are often used in research settings in an attempt to capture and quantify motor performance in the target population. However, there is a lack of psychometric support for these measures, particularly when used with individuals with PD. The subjects included in this work were individuals with PD, either on or off their anti-Parkinsonian medications, or control subjects. The three specific procedures investigated were: 100% limits of stability test using the Balance Master, select gait parameters as measured by the GAITRite automated walkway, and one aspect of upper and lower extremity coordination ability, as measured by a neuromotor channel capacity system. Depending on the device, specific elements of outcome reliability and validity were investigated. The data were evaluated by either Pearson Product Moment Correlations or Analysis of Variance. For most of the analyses, testing individuals at their self-perceived peak medication state seemed to produce more reliable outcomes than while off medication. Relying on previously published work on the psychometric properties of measures using control subjects should not be extrapolated to individuals with Parkinson's disease. Finally, state-of-the-art outcome measures, like the ones investigated in this research, may produce detailed quantitative data, yet the results should not be correlated with more basic clinical outcomes that may, on the surface, appear to measure similar abilities.