Presented at the 12th International Congress of Parkinson’s Disease and Movement Disorders, Chicago, IL, June 22–26 June 2008

Objective: To evaluate the acoustic characteristics such as fundamental frequency, jitter, shimmer, intensity, formants etc. in the speech produced by patients with Parkinson’s disease (PD) in Greek comparing with a matched control group.

Background: Impairment of speech can affect almost 89% of patients with Parkinson’s disease (PD) (Logeman et al, 1978). The speech of PD patients is characterized as hypokinetic dysarthria and an extensive range of perceptual acoustic and physiological studies have been devoted to its analysis. These speech difficulties are related only to phonation in 28% of the cases, or in combination with articulation, speech fluency and prosody. Regarding phonation, this is characterized with reduced pitch due not only to reduction in fundamental frequency but also to reduced ability to alter fundamental frequency according to speech demands. Also the reduced intensity observed in PD patients is related to reduced respiratory support.

Method: Seven female patients with PD (mean age= 68.3 years), part of a larger trial, were recruited in the Department of Neurosurgery, Medical School, University of Athens, Greece, to participate in our study. We measured the acoustic characteristics of speech in these patients in comparison with seven age-, gender-, and education-matched controls. The assessment protocol consisted of the Greek version of Frenchay Dysarthria Assessment (FDA), for articulatory functioning, sustained phonation, reading a list of words, reading texts, and a one minute monologue describing the cookie theft picture of BDAE, for acoustic measures and speech rate. The word list they read consisted of 40 words. The first ten words were chosen to contain the five Greek vowels /a, i, o, u, e/ and plosive sounds in the following conditions: open and closed syllables, stressed and unstressed. So for each vowel we collected data in the following conditions: open syllable stressed, open syllable unstressed, closed syllable stressed, closed syllable unstressed. The remaining 30 words were chosen to start with the six plosive consonant sounds /p,b,t,d,k,g/ followed by the five vowels to measure voice onset time. Speech and voice samples were recorded to facilitate the analysis. The speech analyzer program was used for the recording and Praat for the analysis of the recordings. For the statistical analysis the paired t-test was used. Two trained speech & language therapists carried out the analyses. All patients gave informed consent to participate in the study.

Results: The results demonstrated significant differences in several speech and phonation related variables.
Articulatory and speech functions: In relation to articulatory movements there were observed significant differences between the two groups only in the tongue movements: protrusion (t=3.5, p<0.05), lateral movements (t=2.8, p<0.05) and diadochokinetic rate (t=3.1, p<0.05).
Acoustic measures: Regarding the maximum phonation time there were significant differences between the two groups in the production of both the voiced sound – sustained /a/ (t=2.95, p<0.05) – and the voiceless sound – sustained /s/ (t=2.8, p<0.05). No significant differences were observed in jitter, shimmer, speaking and reading pitch. In relation to the acoustic measures which were evaluated with our protocol including vowel duration in stressed and unstressed syllable, vowel pitch, F1, F2, F3, and voice onset time, significant differences were observed only for the production of the vowel /u/ in stressed and unstressed syllables for the variables: vowel duration within word (t=2.8, p<0.05), F1 (t=3.1, p<0.05) and F2 (t=2.9, p<0.05).

Conclusions: Our findings suggest that the nature of the speech impairment observed in patients with PD seems to be related to phonatory and articulatory variables of speech. From the articulation variables the observed differences were in the tongue movements and especially related to the production of the back sounds, which is also confirmed with the differences in the acoustic properties of the back vowel /u/. Differences were also observed in maximum phonation time for both voiceless and voiced sounds indicating reduced respiratory support. Our data are in agreement with previous reported studies in other languages.

References
Logeman et al (1978). Frequency and occurrence of vocal tract dysfunctions in the speech in a large sample of Parkinson patients. Journal of Speech & Hearing Disorders, 43, 47–57.