The effect of learning due to voice assessment in acoustic analysis of vocal tremor in patients with Parkinson’s disease

Abstract

For years researchers have questioned the influence of learning from baseline assessments. In a 3‐yr study of 14 Parkinson’s patients who received no speech treatment, investigators studied phonatory variability and the effect of learning due to voice assessment. Statistically significant constant decline in weighted mean values were found to exist over time on acoustic variables of maximum performance. A strong learning experience as a result of baseline assessment was also documented. However, measures of maximum performance are notoriously variable, potentially unstable and have been linked to effort, motivation, and learning. Therefore, the present study was designed to examine the effect of learning reflected among the same group of 14 subjects but on a more sensitive, less motivationally dependent phonatory variable. Amplitude‐ and frequency‐demodulated outputs and measures of frequency and level (percent) of low‐frequency vocal tremor components in sustained phonation were analyzed. Growth curve analysis models were constructed to simultaneously estimate a constant rate of decline of the variables over time (excluding the influence of baseline values) and to estimate the effect of learning due to baseline assessments. Preliminary data from this study enhance the understanding of the progressive nature of Parkinson’s disease among untreated subjects, and further define which acoustic measures of voice are more sensitive to the effect of learning. [Work supported in part by DC 00976 from the National Institute on Deafness and Other Communication Disorders; Organized Research Grant, Northern Arizona University.]