Research Summary: Speech Movement Variability in People Who Stutter

NEW!! This work has recently been recognised with the ASHA editor's award for best paper in the Speech section of the Journal of Speech, Language and Hearing Research. You can watch their summary video here. 

I’m excited to share some results with you! This work is from the vocal tract MRI session during the first visit to Oxford (the baseline session). You can listen to me chat about these results in this video (transcript available below video) or read the full paper here.

Transcript_vtMRI_video.docx

We aimed to understand how variable the movements of the speech organs (e.g. lips, tongue) are during fluent speech in people who stutter and people who are typically fluent. I.e if I say the same word over and over, do I produce very similar movements each time I say the word, or do I produce slightly different, or variable, movements each time I say the word? This can tell us about how tightly controlled the speech movement system is.

To do this, we recorded images of the vocal tract whilst participants were speaking using Magnetic Resonance Imaging. This is the first time that vocal tract MRI has been used to look at the speech movements of people who stutter. This method means that we can look at the whole vocal tract (from lips, through the mouth to the voice box) at the same time.

Participants were asked to repeat made-up words: Mab, Mabshibe, Mabfieshabe, Mabshaytiedoib, and Mabteebeebee.

We measured the movements of the lips, the tongue and the velum (the soft bit at the top-back of the mouth).

We found that people who stutter made more variable movements compared with people who are typically fluent during fluent utterances of the pseudowords. The amount of variability did not change as the pseudowords got more complex.

The severity of someone’s stutter did not predict how variable the speech movements were. So, according to our data, it’s not true that greater severity leads to more variable (fluent) speech – there must be something else that contributes to the variability that we recorded!

Variability, as measured here, can tell us about general differences in the control of speech movements between people who stutter and people who are typically fluent but it cannot reveal the reason for the variability. We predict that variability results from altered communication within the central nervous system, for example between the brain area responsible for moving the speech muscles (motor cortex) and the brain area responsible for sensing this movement (sensory cortex).

In addition, we found that people who stutter had longer utterance durations compared with people who are typically fluent. This time, the difference between groups increased as the complexity of the pseudowords increased. This suggests that people who stutter slow down their speech to a larger degree than people who are typically fluent – particularly when the pseudowords are more complex.

Overall, this work shows that people who stutter have greater variability compared with people who are typically fluent, even during fluent utterances. This suggests that the brain co-ordinates speech movements differently for people who stutter, however the exact cause of this variability is unknown.

I hope you like our work – if you’d like to chat about this work further, please email us or use the contact form. A huge THANK YOU to all those who took part in the study!!


Written by Charlie Wiltshire, Spring 2023