Digital voice biomarkers for Parkinson’s disease: A study on sustained vowel analysis in the Russian population
Advanced Neurology · 11.11.2025 · 5 (2) · 025180042
DOI: 10.36922/an025180042 ↗Kurz gefasst
Fachartikel in der Zeitschrift Advanced Neurology, online erschienen am 11.11.2025. Verfasst von Ildar Khasanov und 1 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref).
Abstract (Originalsprache)
Voice abnormalities are common in Parkinson’s disease (PD), but the extent to which language-robust acoustic markers capture PD dysphonia in real-world clinical recording conditions and whether they are confounded by sex, language background, or medication state remains uncertain. This study aims to quantify PD-controlled differences in sustained-vowel acoustics in a Russian cohort, evaluate sex and language effects (Russian, Russian–Tatar bilinguals, and exploratory Tatar subgroup), and assess the robustness to clinical covariates. Cross-sectional data from the BRAINPHONE project were analyzed (n = 201; PD = 109; controls = 92). Participants produced sustained/aː/vowels in routine clinics (≥16 kHz, 32-bit.,wav). Acoustic features included perturbation (jitter and shimmer), cepstral/noise measures (cepstral peak prominence; harmonic-to-noise ratio; glottal-to-noise excitation ratio), and pitch metrics. Group contrasts used the Mann–Whitney U test and false discovery rate (FDR). Robust models adjusted for age, sex, and language; prespecified interactions probed diagnosis × sex/language. Spearman correlations related acoustics to Movement Disorder Society-Unified Parkinson’s Disease Rating Scale III, Hoehn and Yahr, disease duration, and medication variables. PD showed higher perturbation and lower cepstral/noise measures than controls (all q≤0.01 effects were consistent in females and males and replicated in Russian monolinguals and Russian–Tatar bilinguals with the Tatar monolingual subgroup being directionally similar. Covariate-adjusted models retained significant PD effects. Acoustic–clinical correlations were small (|ρ|≤~0.21) and did not survive FDR. In real-world clinical recordings of sustained vowels, CPPS, GNE, and shimmer provide robust, language-tolerant, medication-insensitive markers of PD dysphonia, supporting use as a complementary digital biomarker for telemedicine and longitudinal monitoring.
Zitieren
APA
Khasanov, I., & Khasanova, D. (2025). Digital voice biomarkers for Parkinson’s disease: A study on sustained vowel analysis in the Russian population. Advanced Neurology, 5(2), 025180042. https://doi.org/10.36922/an025180042
Vancouver
Khasanov I, Khasanova D. Digital voice biomarkers for Parkinson’s disease: A study on sustained vowel analysis in the Russian population. Advanced Neurology. 2025;5(2):025180042. doi:10.36922/an025180042
BibTeX
@article{Khasanov2025,
author = {Khasanov, Ildar and Khasanova, Diana},
title = {Digital voice biomarkers for Parkinson’s disease: A study on sustained vowel analysis in the Russian population},
journal = {Advanced Neurology},
year = {2025},
volume = {5},
number = {2},
pages = {025180042},
doi = {10.36922/an025180042}
}
EUVIMED