Sensorimotor adaptation of voice fundamental frequency in Parkinson's disease

Abstract

Objective: This study examined adaptive responses to auditory perturbation of fundamental frequency (fo) in speakers with Parkinson's disease (PD) and control speakers.

Method: Sixteen speakers with PD and nineteen control speakers produced sustained vowels while they received perturbed auditory feedback (i.e., fo shifted upward or downward). Speakers' pitch acuity was quantified using a just-noticeable-difference (JND) paradigm. Twelve listeners provided estimates of the speech intelligibility for speakers with PD.

Results: Fifteen responses from each speaker group for each shift direction were included in analyses. While control speakers generally showed consistent adaptive responses opposing the perturbation, speakers with PD showed no compensation on average, with individual PD speakers showing highly variable responses. In the PD group, the degree of compensation was not significantly correlated with age, disease progression, pitch acuity, or intelligibility.

Conclusions: These findings indicate reduced adaptation to sustained fo perturbation and higher variability in PD compared to control participants. No significant differences were seen in pitch acuity between groups, suggesting that the fo adaptation deficit in PD is not the result of purely perceptual mechanisms.

Significance: These results suggest there is an impairment in vocal motor control in PD. Building on these results, contributions can be made to developing targeted voice treatments for PD.

Fig 1. Average adaptive responses by group.

An adaptive “shift-up” (left) and “shift-down” (right) perturbation was applied to the f_o of auditory feedback with a maximum perturbation of 100 cents or -100 cents respectively (schematized in upper panel). Mean adaptive responses in cents are plotted as the mean across five-trial blocks for control speakers (dark blue line) and speakers with Parkinson’s disease (light pink line) with shading indicating 95% confidence intervals.

Fig 2. Individual adaptive responses by group.

Individual adaptive responses in cents for “shift-up” (left) and “shift-down” (right) perturbations (schematized in upper panel). Responses are plotted as the mean across five-trial blocks for fifteen adaptive responses from control speakers (upper) and fifteen adaptive responses from speakers with Parkinson’s disease (lower).

Fig 3. Baseline f_o variability compared to the degree of compensation.

Relationship between baseline f_o variability (cents) and degree of compensation (cents) for speakers with PD (upper panel) and for control speakers (lower panel) during the “shift-up” (blue, closed circles) and “shift-down” (orange, open circles) conditions. Lines of best fit are plotted for the “shift-up” (blue solid line) and the “shift-down” (orange solid line) conditions.