Vowel formant characteristics in Huntington's disease

Abstract

The speech of Huntington's disease (HD) has been described with a variety of terms including slow, hesitant, and staccatolike, yet virtually no data have been presented to quantify these characteristics. Acoustic analysis of the vowels in HD speech may provide information about the nature of the speech output. In this study, two questions were posed. (1) Are the vowels of HD subjects different from normal speakers as determined by the formant cluster patterns? (2) Are HD speakers capable of reaching vowel formant targets when more complex speech utterances are utilized? Ten subjects with HD ranging in age from 26 to 58 years old were studied. Each subject was asked to produce four vowels in isolation and in CVC syllables in which the initial consonant was either /d/ or /g/ and the final consonant was always /d/. Measurements of F1 and F2 were made for all vowels from wideband spectrograms. The results indicate that the isovowel targets, especially for F2, for HD speakers, were different from normal F2 values. When compared to the isovowels, the formants in the CVC context reflected centralization. The results suggest that formant trajectories for the vowels in HD speech fail to reach normal isovowel formant targets and that as the stimulus complexity increases, approximation to normal vowel formant frequencies decreases. [Work supported by the Veterans Administration.]