Objective: We characterized creaky voice in Parkinson’s disease (PD) and evaluated the association of this acoustic phenomenon with motor and cognitive status.

Background: Creaky voice (“vocal fry”) has a ‘rough’ voice quality and is associated with a low rate of vocal fold vibration [1]. While hypokinetic dysarthria and related acoustic characteristics have undergone detailed study in PD [2], creaky voice has been understudied and may be an informative measure of laryngeal motor speech control. Automated algorithms to detect creaky voice are available [3,4], but have not been validated in PD.

Method: Subjects with PD (n = 26) were given a reading task (Rainbow Passage) and a picture description task (Cookie Theft picture), and assessed with MDS-UPDRS (with calculation of PD subtypes), MoCA and a cognitive battery to determine cognitive status [5]. Creak probabilities were measured in MATLAB, and creak% were calculated [6].  Using SPSS, we assessed correlation of log transformed creak% and clinical variables with Pearson’s correlations. Backward linear regression models were performed, and the best fit model determined by adjusted R-square, ANOVA, and collinearity diagnostics.

Results: Median creak% was 3.11% (range 17.86%) for the reading task, and 2.50% (range 13.38%) for the picture description task. There were no differences in creak by sex, PD subtype, or cognitive diagnosis. Creak on one task was highly correlated with creak on the other task (r = 0.7, p < 0.001). In the reading task, MDS-UPDRS Part II was significantly associated with creak after controlling for age and sex (β= 0.89, SE = 0.34, p = 0.017, R2 = 0.329). A model including MDS-UPDRS part III, age and sex yielded similar results with improved R2 = 0.468. In the picture description task, creak was more variable between participants and regression models did not determine any significant predictors of creak.

Conclusion: Creaky voice can be detected by automated algorithm in PD and occurs to a similar degree in different speech tasks. Creaky voice was associated with motor severity and may be a marker of vocal fold hypokinesis. Future studies should further explore creaky voice as a speech marker and assess patterns of creaky voice in PD compared to the healthy population.

References: [1] Keating, P., Garellek, M., & Kreiman, J. (2015). Acoustic properties of different kinds of creaky voice. The Scottish Consortium for ICPhS 2015 (Ed.). Proceedings of the 18th International Congress of Phonetic Sciences, Glasgow, UK the University of Glasgow.
[2] Ramig, L., Bonitati, C., Lemke, J., & Horii, Y. (1994) Voice treatment for patients with Parkinson’s disease: development of an approach and preliminary efficacy data. Journal of Medical Speech-Language Pathology, 2, 191-209
[3] Ishi, C. T., Sakakibara, K., Ishiguro, H., & Hagita, N. (2008) A Method for Automatic Detection of Vocal Fry. IEEE Transactions on Audio, Speech, and Language Processing, 16 (1), 47-56. https://doi.org/10.1109/TASL.2007.91079
[4] Kane, J., Drugman, T., & Gobl, C. (2013). Improved automatic detection of creak. Computer Speech & Language. 27(4). 1028–1047. https://doi.org/10.1016/j.csl.2012.11.002.
[5] Cholerton, B. A., Zabetian, C. P., Wan, J. Y., Montine, T. J., Quinn, J. F., Mata, I. F., Chung, K. A., Peterson, A., Espay, A. J., Revilla, F. J., Devoto, J., Watson, G. S., Hu, S. C., Leverenz, J. B., & Edwards, K. L. (2014). Evaluation of mild cognitive impairment subtypes in Parkinson’s disease. Movement disorders : official journal of the Movement Disorder Society, 29(6), 756–764. https://doi.org/10.1002/mds.25875
[6] Murton, O. M., Hillman, R. E., Mehta, D. D., Semigran, M., Daher, M., Cunningham, T., Verkouw, K., Tabtabai, S., Steiner, J., Dec, G. W., & Ausiello, D. (2017). Acoustic speech analysis of patients with decompensated heart failure: A pilot study. The Journal of the Acoustical Society of America, 142(4), EL401. https://doi.org/10.1121/1.5007092

« Back to 2022 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/association-of-creaky-voice-with-motor-severity-in-parkinsons-disease-using-an-automated-algorithm/