Electrophysiological evidence for a general auditory prediction deficit in adults who stutter

Abstract

We previously found that stuttering individuals do not show the typical auditory modulation observed during speech planning in nonstuttering individuals. In this follow-up study, we further elucidate this difference by investigating whether stuttering speakers' atypical auditory modulation is observed only when sensory predictions are based on movement planning or also when predictable auditory input is not a consequence of one's own actions. We recorded 10 stuttering and 10 nonstuttering adults' auditory evoked potentials in response to random probe tones delivered while anticipating either speaking aloud or hearing one's own speech played back and in a control condition without auditory input (besides probe tones). N1 amplitude of nonstuttering speakers was reduced prior to both speaking and hearing versus the control condition. Stuttering speakers, however, showed no N1 amplitude reduction in either the speaking or hearing condition as compared with control. Thus, findings suggest that stuttering speakers have general auditory prediction difficulties.

Keywords: Auditory cortex; Movement planning; Sensorimotor integration; Speech; Stuttering.

Fig. 1

Experimental procedures for delayed-speaking (A), delayed-hearing (B), and silent reading (C) conditions. A probe auditory stimulus (1 kHz; 40 ms duration; 75 dB SPL) was presented in 40% of trials (tone trials) of each condition (D). No auditory probe was presented in the remaining trials (no-tone trials). For each subject and in each condition, the average evoked potential for no-tone trials was subtracted from the average evoked potential for tone trials (E). This procedure provided the best estimate of the auditory cortex’s response to the probe tones.

Fig. 2

Top row: Three regions of interests (ROIs), each consisting of six electrodes (AEPs from the six electrodes in each ROI were averaged). Middle row: Grand average AEPs for the nonstuttering group from the Left ROI (left column), Central ROI (middle column), and Right ROI (right column) in the delayed-speaking (green), delayed-hearing (orange), and silent reading (black) conditions. Bottom row: Grand average AEPs for the stuttering group in the same conditions as above for the nonstuttering group. AEPs of the nonstuttering group show a distinct attenuation of N1 amplitude in the delayed-speaking and delayed-hearing conditions relative to the silent reading control condition.

Fig. 3

Group average N1 amplitude (A) and P2 amplitude (C) for stuttering and nonstuttering groups during the delayed-speaking condition (green), delayed-hearing condition (orange), and silent reading condition (black). Subject distribution with regard to the amount of amplitude modulation in the delayed-speaking and delayed-hearing conditions vs. the silent reading condition for N1 and P2 are shown as boxplots (B and D, respectively; vertical axes are adjusted so that greater reductions in amplitude are toward the top of the plots). There was a statistically significant Group × Condition interaction for N1 amplitude, indicating significant modulation in delayed-speaking and delayed-hearing conditions for the nonstuttering group but not the stuttering group. The Group × Condition interaction for P2 was not statistically significant. All data averaged across three ROIs. Error bars indicate standard errors. Asterisks indicate p < .001.