Stuttering is associated with atypical structural and functional connectivity in sensorimotor brain areas, in particular premotor, motor, and auditory regions. It remains unknown, however, which specific mechanisms of speech planning and execution are affected by these neurological abnormalities. To investigate pre-movement sensory modulation, we recorded 12 stuttering and 12 nonstuttering adults’ auditory evoked potentials in response to probe tones presented prior to speech onset in a delayed-response speaking condition vs. no-speaking control conditions (silent reading; seeing nonlinguistic symbols). Findings indicate that, during speech movement planning, the nonstuttering group showed a statistically significant modulation of auditory processing (reduced N1 amplitude) that was not observed in the stuttering group. Thus, the obtained results provide electrophysiological evidence in support of the hypothesis that stuttering is associated with deficiencies in modulating the cortical auditory system during speech movement planning. This specific sensorimotor integration deficiency may contribute to inefficient feedback monitoring and, consequently, speech dysfluencies.
Stuttering is a disorder of speech fluency associated with abnormal brain activation in a widespread network of pre-motor, motor, and sensory regions ( Braun et al., 1997, Chang et al., 2009, De Nil et al., 2000, De Nil et al., 2003, Fox et al., 1996, Neumann et al., 2003 and Watkins et al., 2008). Across individual stuttering subjects, the involvement of specific brain regions appears to vary considerably ( Ingham et al., 2013 and Wymbs et al., 2013). Nevertheless, several structural brain abnormalities have been reported, and these abnormalities include atypical white matter in pathways suggested to connect speech motor and auditory regions ( Chang et al., 2008, Chang et al., 2011, Cykowski et al., 2010, Foundas et al., 2003, Sommer et al., 2002 and Watkins et al., 2008). Accordingly, some of the most prominent contemporary theoretical views of stuttering suggest that the disorder may result from deficits in specific processes of sensorimotor integration that are critical for both early speech motor learning and mature speech motor control ( Beal et al., 2010, Brown et al., 2005, Cai et al., 2012, Chang et al., 2011, Daliri et al., 2013, Hickok et al., 2011, Kell et al., 2009, Liotti et al., 2010, Max, 2004 and Watkins et al., 2008).
One aspect of sensorimotor integration that may be of particular theoretical importance in this regard is the central nervous system’s (CNS) prediction of the sensory consequences (or, more generally, movement outcomes) of planned motor commands. In a recent study ( Daliri, Prokopenko, Flanagan, & Max, 2014), we found that individuals who stutter accurately predict specific movement consequences in a ballistic reaching task (i.e., arm movements completed without relying on afferent feedback) in which those consequences could be fully compensated through anticipatory adjustments during movement planning. However, based on our overall theoretical framework ( Max, 2004), stuttering individuals may be more likely to have difficulties with appropriately using such predictions to successfully “prime” task-relevant sensory systems for their subsequent role in (a) closely monitoring afferent inputs for online feedback control while (b) simultaneously preventing feedback-based motor responses that are undesirable during self-generated voluntary movements (note that the latter part of this hypothesis overlaps with ideas proposed by Zimmermann, 1980). Limited evidence consistent with this hypothesis was already provided by McClean (1996) who demonstrated that, as compared with fluent speakers, stuttering adults show less attenuation of mechanically-evoked lip muscle reflexes prior to the onset of speech (with lip muscle activity measured in speech trials vs. no speech trials). When participating sensory systems (auditory, somatosensory) are insufficiently modulated in terms of their response to self-generated afferent inputs, the triggered motor responses may interfere with, and disrupt, ongoing movements. To date, however, it remains completely unknown (a) whether stuttering individuals’ atypical sensorimotor responses at speech onset are in fact due to a lack of central modulation of sensory neural systems, and, if so, (b) whether stuttering individuals show a lack of pre-speech sensory modulation in the auditory cortical regions that have been implicated in several, although not all, neuroimaging studies (see above).
Here, we addressed both these questions directly by using electroencephalographical (EEG) data and auditory evoked potential analyses to investigate, in stuttering vs. nonstuttering adults, the modulation of auditory cortical activity in response to probe tones presented prior to speaking (i.e., during speech movement planning) and in control conditions without preparation for motor activity.1 Using an experimental paradigm that we previously developed for work with typically fluent speakers (Max, Daniels, Curet, & Cronin, 2008), we recorded long latency auditory evoked potentials (LLAEPs) in response to auditory stimuli presented during the delay phase of a delayed-response speaking task (seeing a word on a monitor, silently reading the word, and saying it aloud after a go signal), a silent reading task (seeing a word and silently reading it), and a seeing task (seeing nonlinguistic symbols). We also recorded the same LLAEPs in a standard eyes-closed rest condition to compare both groups in terms of basic auditory processing in the absence of an active task, and to verify, through comparison with these reference data, the validity of data processing and analysis procedures used in the three active tasks. Analyses focused on the amplitude and latency of the LLAEP components N1 and P2. 2 We hypothesized that if stuttering is associated with a lack of modulation of auditory cortical regions prior to speech onset, the stuttering group would fail to show the typical within-subject N1 amplitude attenuation that we have previously documented for normally fluent speakers ( Max et al., 2008).
