A Prediction-Based Analysis of Hypokinetic Dysarthria in an
“Off” versus “On” Medication State

J Liss, J Josephs, L Parsons, C Adler, J Caviness


There is a lack of consensus in the literature regarding the effects of dopaminergic therapy on speech production in Parkinson’s disease (PD).  It is hypothesized that this may be due, in part, to methodologies and designs that do not accommodate the variable manifestation of hypokinetic dysarthria across patients with PD.  The purpose of this study was to evaluate a prediction-based analysis method that takes into consideration each speaker’s pattern of speech deficit in the “off” medication state to make predictions about the amount and direction of change expected in the “on” state.  This study examined the speech characteristics of four patients with idiopathic PD whose speech showed perceptible improvement in a medicated “on” state, using acoustic measures of speaking rate, articulatory precision, articulatory working space, and vocal integrity.  Although some classes of speech characteristics (e.g., vocal integrity) did show change for all of the targeted speakers from the “off” to “on” state, the overall patterns of change were quite speaker-specific, and related to the perceived speech patterns in the “off” state.  The results support a prominent role of perceptual data in making decisions about acoustic analysis.  The utility and limitations of prediction-based assessment and preliminary interpretations are discussed.