Sensorimotor Oscillations Prior to Speech Onset Reflect Altered Motor Networks in Adults Who Stutter

Abstract

Adults who stutter (AWS) have demonstrated atypical coordination of motor and sensory regions during speech production. Yet little is known of the speech-motor network in AWS in the brief time window preceding audible speech onset. The purpose of the current study was to characterize neural oscillations in the speech-motor network during preparation for and execution of overt speech production in AWS using magnetoencephalography (MEG). Twelve AWS and 12 age-matched controls were presented with 220 words, each word embedded in a carrier phrase. Controls were presented with the same word list as their matched AWS participant. Neural oscillatory activity was localized using minimum-variance beamforming during two time periods of interest: speech preparation (prior to speech onset) and speech execution (following speech onset). Compared to controls, AWS showed stronger beta (15-25 Hz) suppression in the speech preparation stage, followed by stronger beta synchronization in the bilateral mouth motor cortex. AWS also recruited the right mouth motor cortex significantly earlier in the speech preparation stage compared to controls. Exaggerated motor preparation is discussed in the context of reduced coordination in the speech-motor network of AWS. It is further proposed that exaggerated beta synchronization may reflect a more strongly inhibited motor system that requires a stronger beta suppression to disengage prior to speech initiation. These novel findings highlight critical differences in the speech-motor network of AWS that occur prior to speech onset and emphasize the need to investigate further the speech-motor assembly in the stuttering population.

Keywords: beta suppression; beta synchronization; developmental stuttering; magnetoencephalography; speech preparation.

Figure 1

Task schematic and time course of two successive trials. Task sequence (A) includes a fixation (+) of alternating length (1 or 2 s), stimulus sentence (S, 0.5 s), a blank screen (B, 0.5 s) and the cue to speak (<))), 3 s). Lip EMG signal (B) and voice signal (C) are taken from one participant. The response of the previous trial and the succeeding trial are shown for the chosen participant.

Figure 2

CIVET-generated surface images with imposed Synthetic Aperture Magnetometry (SAM) localization of beta and alpha suppression (blue) during speech preparation (stimulus-locked) compared between fluent speakers (FS) and adults who stutter (AWS). The time at which peak magnitude was observed relative to stimulus presentation is indicated. The bilateral cuneus is displayed for the beta suppression only, but localizations were the same for the observed alpha suppression in that region (see Table 2).

Figure 3

Stimulus-locked time-frequency representation (TFR) plots of virtual sensors extracted from the bilateral mouth motor cortex (BA6) and auditory-sensory cortices (BA13/41) compared between AWS and FS. The corresponding task sequence time-line is displayed at the top for orientation and includes the fixation (+), stimulus sentence (S, 0.5 s), a blank screen (B, 0.5 s) and the cue to speak (<))), 3 s). Suppression (blue) is apparent following the stimulus presentation (S).

Figure 4

Time-courses of beta (15–25 Hz) and alpha (8–13 Hz) suppression in the bilateral mouth motor cortex (BA6) and auditory regions (BA41, 13), respectively, compared across AWS and FS. All windows and time-points of interest are indicated on the top plot. Speech preparation (PREP) was defined from stimulus presentation to the speech cue (0–1 s) in the stimulus-locked dataset (A). Speech execution (EXEC) was defined from the speech onset to participant specific beta event-related desynchronization (ERD) offset time in the speech-locked dataset (B). A window of beta synchronization (event-related synchronization, ERS) is indicated preceding the stimulus presentation, as well as the corresponding ERS onset and offset times (A, top). ERD onset time is marked for the purpose of the latency analysis in Section “Group Comparisons of Alpha and Beta Power Suppression Latencies”.

Figure 5

Integrated beta suppression during speech preparation (PREP) and execution (EXEC). Note. AWS group showed stronger left beta suppression during speech execution once the outlier was removed (*).

Figure 6

(A) Integrated beta ERS power in the left and right mouth motor cortex (BA6) preceding stimulus onset. (B) Significant linear relationship between the integrated beta ERS in the left and right BA6 of AWS (top). No significance found in FS (B, bottom).

Figure 7

Linear regression of stuttering severity on alpha suppression in left primary auditory cortex (BA41).

Figure 8

Onset of beta suppression in the left and right mouth motor cortex (BA6), relative to stimulus presentation. Task sequence time-line is displayed for orientation.