Automated speech analysis in early untreated Parkinson's disease: Relation to gender and dopaminergic transporter imaging
Abstract
Background: The mechanisms underlying speech abnormalities in Parkinson's disease (PD) remain poorly understood, with most of the available evidence based on male patients. This study aimed to estimate the occurrence and characteristics of speech disorder in early, drug-naive PD patients with relation to gender and dopamine transporter imaging.
Methods: Speech samples from 60 male and 40 female de novo PD patients as well as 60 male and 40 female age-matched healthy controls were analyzed. Quantitative acoustic vocal assessment of 10 distinct speech dimensions related to phonation, articulation, prosody, and speech timing was performed.
All patients were evaluated using [123]I-2b-carbomethoxy-3b-(4-iodophenyl)-N-(3-fluoropropyl) nortropane single-photon emission computed tomography and Montreal Cognitive Assessment. Results: The prevalence of speech abnormalities in the de novo PD cohort was 56% for male and 65% for female patients, mainly manifested with monopitch, monoloudness, and articulatory decay.
Automated speech analysis enabled discrimination between PD and controls with an area under the curve of 0.86 in men and 0.93 in women. No gender-specific speech dysfunction in de novo PD was found.
Regardless of disease status, females generally showed better performance in voice quality, consonant articulation, and pauses production than males, who were better only in loudness variability. The extent of monopitch was correlated to nigro-putaminal dopaminergic loss in men (r = 0.39, p = 0.003) and the severity of imprecise consonants was related to cognitive deficits in women (r = -0.44, p = 0.005).
Conclusions: Speech abnormalities represent a frequent and early marker of motor abnormalities in PD. Despite some gender differences, our findings demonstrate that speech difficulties are associated with nigro-putaminal dopaminergic deficits.