Acoustic Measures Capture Speech Dysfunction in Spinocerebellar Ataxia

Spinocerebellar ataxias (SCA) are hereditary cerebellar degenerative disorders with a common feature of dysarthria, involving impaired phonatory and articulatory control of speech, thereby affecting social communication. In this study, we investigated whether acoustic measures could objectively measure speech dysfunction and identify common indicators across SCA types 1, 2, and 3.

We recorded speech from 24 SCA patients and 24 age-matched controls during connected speech and sustained vowel tasks. Speech was assessed using the Acoustic Voice Quality Index (AVQI) and sub-metrics (cepstral peak prominence smoothed [CPPS], harmonics-to-noise ratio [HNR], and shimmer). Jitter was used to quantify dysfunction in sustained vowels, and formant analysis was used to examine articulatory control. We also evaluated whether these measures capture speech dysfunction severity.

AVQI was the most consistent indicator of speech dysfunction across SCA subgroups, showing a significant increase compared to controls. SCA patients exhibited abnormalities in AVQI sub-metrics (CPPS, HNR, and shimmer) and increased jitter during sustained vowels. Furthermore, formant analysis revealed articulation differences linked to changes in F2, F3, and F4 during connected speech, and F2 and F4 during sustained vowels. Additionally, AVQI correlates with the Scale for Assessment and Rating of Ataxia (SARA) speech sub-scores, whereas CPPS and jitter can differentiate between mild and moderate stages of ataxic speech.

Our findings show that SCA patients exhibit speech dysfunction across both connected speech and sustained vowel tasks, as AVQI stands out as a key indicator, whereas CPPS and jitter track ataxic speech severity. Our findings support using acoustic measures for objectively assessing speech dysfunction and thereby informing disease monitoring and guiding future targeted therapies.