Transcranial direct current stimulation over left dorsolateral prefrontal cortex facilitates auditory-motor integration for vocal pitch regulation

Abstract

Background: A growing body of literature has implicated the left dorsolateral prefrontal cortex (DLPFC) in the online monitoring of vocal production through auditory feedback. Specifically, disruption of or damage to the left DLPFC leads to exaggerated compensatory vocal responses to altered auditory feedback. It is conceivable that enhancing the cortical excitability of the left DLPFC may produce inhibitory influences on vocal feedback control by reducing vocal compensations.

Methods: We used anodal transcranial direct current stimulation (a-tDCS) to modulate cortical excitability of the left DLPFC and examined its effects on auditory-motor integration for vocal pitch regulation. Seventeen healthy young adults vocalized vowel sounds while hearing their voice pseudo-randomly pitch-shifted by ±50 or ±200 cents, either during (online) or after (offline) receiving active or sham a-tDCS over the left DLPFC.

Results: Active a-tDCS over the left DLPFC led to significantly smaller peak magnitudes and shorter peak times of vocal compensations for pitch perturbations than sham stimulation. In addition, this effect was consistent regardless of the timing of a-tDCS (online or offline stimulation) and the size and direction of the pitch perturbation.

Conclusion: These findings provide the first causal evidence that a-tDCS over the left DLPFC can facilitate auditory-motor integration for compensatory adjustment to errors in vocal output. Reduced and accelerated vocal compensations caused by a-tDCS over left DLPFC support the hypothesis of a top-down neural mechanism that exerts inhibitory control over vocal motor behavior through auditory feedback.

Keywords: auditory feedback; left dorsolateral prefrontal cortex; speech motor control; top–down modulation; transcranial direct current stimulation.

Figure 1

Grand-averaged voice f_o responses to pitch perturbations of ±50 cents (left panel) and ±200 cents (right panel) when active (red solid lines) or sham (blue solid lines) a-tDCS over the left DLPFC was applied during the FAF task. Highlighted areas represent the standard errors of the mean vocal responses, while the vertical dash lines indicate the onset of pitch perturbations.

Figure 2

Violin plots illustrating the magnitudes (A,B) and latencies (C,D) of vocal responses to pitch perturbations of ±50 cents and ±200 cents when active (red) and sham (blue) a-tDCS over the left DLPFC was applied during the FAF task. The shape of the violin shows the kernel density estimate of the data. The white dots and box plots represent the medians and ranges from first to third quartiles of the data sets. The red and blue dots represent the individual vocal responses for active and sham a-tDCS over the left DLPFC. The asterisks indicate significant differences across the conditions.

Figure 3

Grand-averaged voice f_o contours in response to pitch perturbations of ±50 cents (left panel) and ±200 cents (right panel) when active (red solid lines) or sham (blue solid lines) a-tDCS was applied over the left DLPFC prior to the FAF task. Highlighted areas represent the standard errors of the mean vocal responses, while the vertical dash lines indicate the onset of pitch perturbations.

Figure 4

Violin plots illustrating the magnitudes (A,B) and latencies (C,D) of vocal responses to pitch perturbations of ±50 cents and ±200 cents when active (red) or sham (blue) a-tDCS was applied over the left DLPFC prior to the FAF task. The white dots and box plots represent the medians and ranges from first to third quartiles of the data sets. The red and blue dots represent the individual vocal responses for active and sham a-tDCS over the left DLPFC. The asterisks indicate significant differences across the conditions.

Figure 5

Violin plots illustrating the magnitudes (A) and latencies (B) of vocal responses to pitch perturbations when active or sham a-tDCS was applied over the left DLPFC prior to (offline) or during (online) the FAF task. The white dots and box plots represent the medians and ranges from first to third quartiles of the data sets. The red and blue dots represent the individual vocal responses for online and offline a-tDCS over the left DLPFC. The asterisks indicate significant differences across the conditions.