Stuttering, defined as a disfluency in speech, affects some 40 million people worldwide. Researchers led by Ranit Sengupta and Sazzad Nasir at Northwestern University recently uncovered new evidence indicating anomalies in neural timing as a cause of stuttering.
A unique approach was used for this investigation, involving the use of electroencephalogram (EEG) patterns recorded during real-time distortion of participants’ voices while speaking. Recordings obtained from a group of 8 adults with stuttering (AWS) were compared with a group of 9 fluently speaking adults (FA) and revealed signs of disruptions in the coordination of neural activity of AWS.
Disrupted Neural Communication
The experiment was motivated by previous evidence supporting the theory that stuttering is, at least in part, caused by deficits in communication between cortical areas subserving the integration of sensory and motor information involved in speech.
Speech production requires coordinated activity between these areas in order to align the motoric goals of speech – the movements that comprise speech – and auditory goals defined by prediction of speech sounds expected as a result of vocal movements.
The theory stipulates that AWS have a deficit in sensory-motor integration. That is, during speech the neural representation of the vocal sound doesn’t correspond with the motor sequence executed. The incongruence is registered as an error and potentially causes them to stop making the sound they were making and start again, resulting in stuttering.
Lead author Ranit Sengupta says:
“Perhaps, what is happening, at least among some stutterers, is a disruption in the information flow between the sensory domain that generates the auditory goal of the target word, and the motor domain that is responsible for generating the movements, resulting in the repetition, blockage, protraction of syllables. This is why we hypothesized why sensorimotor integration to be an underlying cause of stuttering.”
Through headphones, participants in both the AWS and FA groups heard a modified version of their voice while repeating the word “head” that made it sound more like “hid.” Such alteration causes unconscious modification of vocal movements in what is thought to be an online attempt to adapt to and correct for the odd, incongruent auditory outcome of their vocal motor actions.
As participants adapt to the alteration, they learn produce a shift towards pronouncing “had” to make the alteration sound more like “head.”
The authors elected to use this approach because it requires participants to learn a new correspondence, or ‘mapping’ between their vocal gestures and the auditory feedback produced, requiring effective communication between sensory and motor networks in the brain.
Neural activity was simultaneously recorded using EEG while participants experienced the auditory modification of their voice. Adults with stuttering did not adapt their voice to the alteration whereas a group of fluently speaking adults demonstrated significant adaptation.
The results revealed key differences between the groups in the temporal pattern of neural signals occurring during speech production.
AWS showed less synchrony between different frequency bands – a phenomenon known as ‘phase coherence’ – over the course of the training period. This decrease in phase coherence over the training is significant because previous evidence supports phase coherence as a marker of communication between the motor and auditory-sensory networks.
Thus, the findings implicate communication breakdown between cortical networks in adults with stuttering resulting from a deficit in activity coordination.
In sum, Sazzad says,
“What people haven’t looked at is how these different frequency bands interact, and what is being communicated across the different bands of the pools of neurons through phase coherence. This is something we have taken on, I believe for the first time.”
Image: GerryShaw CC BY-SA 3.0, via Wikimedia Commons. Rat brain cultures stained for coronin 1a, found in microglia here in green, and alpha-internexin, in red, found in neuronal processes. Antibodies and image generated by EnCor Biotechnology Inc.