Dopaminergic Treatment Effects on Dysarthric Speech: Acoustic Analysis in a Cohort of Patients With Advanced Parkinson's Disease

Abstract

Importance: The effects of dopaminergic treatment on speech in patients with Parkinson's disease (PD) are often mixed and unclear. The aim of this study was to better elucidate those discrepancies. Methods: Full retrospective data from advanced PD patients before and after an acute levodopa challenge were collected. Acoustic analysis of spontaneous monologue and sustained phonation including several quantitative parameters [i.e., maximum phonation time (MPT); shimmer local dB] as well as the Unified Parkinson's Disease Rating Scale (UPDRS) (total scores, subscores, and items) and the Clinical Dyskinesia Rating Scale (CDRS) were performed in both the defined-OFF and -ON conditions. The primary outcome was the changes of speech parameters after levodopa intake. Secondary outcomes included the analysis of possible correlations of motor features and levodopa-induced dyskinesia (LID) with acoustic speech parameters. Statistical analysis included paired t-test between the ON and OFF data (calculated separately for male and female subgroups) and Pearson correlation between speech and motor data. Results: In 50 PD patients (male: 32; female: 18), levodopa significantly increased the MPT of sustained phonation in female patients (p < 0.01). In the OFF-state, the UPDRS part-III speech item negatively correlated with MPT (p = 0.02), whereas in the ON-state, it correlated positively with the shimmer local dB (p = 0.01), an expression of poorer voice quality. The total CDRS score and axial subscores strongly correlated with the ON-state shimmer local dB (p = 0.01 and p < 0.01, respectively). Conclusions: Our findings emphasize that levodopa has a poor effect on speech acoustic parameters. The intensity and location of LID negatively influenced speech quality.

Keywords: Parkinson's disease; acoustic analysis; dysarthria; dyskinesias; levodopa; speech.

Figure 1

Correlation between intensity and location of LID and acoustic speech parameters. (A) Significant positive correlation between Clinical Dyskinesia Rating Scale (CDRS) total score (x-axis) and shimmer local DB (y-axis) in the ON condition. (B) Significant positive correlation between axial CDRS subscore (x-axis) and noise-to-harmonics ratio (y-axis) in the ON condition. (C) Statistically positive correlation between axial CDRS subscore (x-axis) and shimmer local DB (y-axis) in the ON condition. These results highlight that patients with higher intensity of levodopa-induced dyskinesia, particularly in the axial region, had a poorer voice quality.