Developing an innovative, objective, child-friendly, clinical measurement to assess speech motor skills in children who stutter

Stuttering is a neurodevelopmental disorder with an incidence rate of 5 to 11%
in preschool years (3), and a prevalence of about 1% in adults. Both the
(probably) multifactorial cause of stuttering, and the inter-individual
differences in the pathway to unassisted recovery are still unknown. Moreover,
intra-individual differences in fluency cannot yet be explained either (4).
Last decades, ample research has been done on cognitive-linguistic functions,
temperament and social-emotional behaviors of children who stutter, and their
impact on the development of stuttering (5-7). Group differences were found,
but these factors could never fully explain inter- and intra-individual
differences. Furthermore, recent research studying the underlying genetic and
neurological patterns of stuttering show a strong genetic involvement (2). What
these patterns exactly cause, however, is not yet known. As stutters can be
defined as breakdowns in the speech motor system, it is not surprising that
many researchers have studied the control of speech motor movements in people
who stutter, i.e., the ability to coordinate movements of tongue, lips, jaw and
vocal folds. The speech motor processes underlying stuttering have been widely
documented in adults, showing more variable, slower, and physiologically
different speech motor movements with a worse relative timing than people who
do not stutter, also during perceptually fluent speech (e.g., 8). Only few
studies have assessed the speech motor dynamics of stuttering near its onset,
i.e., in pre-school aged children showing that children who stutter have more
variable articulatory movements than children who do not stutter (9). So far,
research has almost exclusively considered speech motor control as normal or
disabled, comparing persons who stutter and persons who do not stutter on a
group level. In general, these studies show an overlap between cases and
controls. Moreover, no studies in children have investigated SMS controlling
for compensatory behaviors.
Research theory: In contrast to these previous studies, we propose to study
control of speech motor movements as a continuous outcome, from extremely weak
to extremely strong control. This proposal is based on the Speech Motor Skills
(SMS) theory, which explains stuttering as a result of *an innate limitation in
the speech motor control system to prepare and perform complex motor actions in
the presence of cognitive, linguistic, emotional and speech motor
influences* (10). People who persist in stuttering are presumed to have SMS in
the lower end of a continuum, while people who do not stutter are distributed
across the more skilled end of the continuum. Children who recover from
stuttering are presumed to have only slightly weaker or immature SMS. The SMS
theory assumes that the speech motor control system of people who stutter is
not abnormal as such (like with dysarthria). It predicts that this system
becomes more unstable (variable) with increasing complexity on the motor level
(e.g., speaking with higher speed or smaller movements), the
cognitive-linguistic level (e.g., telling a complex storyline) or on the
social-emotional level (e.g., speaking when highly aroused or tensed). In these
situations, people with low SMS are more likely to stutter, unless
speed-accuracy tradeoffs or motor control solutions (strategies) are being used
as a compensation for their weaker SMS.
Clinical knowledge gap: To study SMS, until now, only invasive and
time-consuming methods were available, such as Electro Magnetic Articulography
(EMA), using tongue, lips and jaw electrodes. These methods are expensive and
unsuitable to use in preschool children or in a clinical setting. Consequently,
SMS of children who stutter can now only be rated with a perceptual scoring
system, the Oral Motor Assessment Scale (1), that has insufficient inter-rater
reliability. On top of this, it consists of only three probe words, which do
not cover the full SMS complexity range. Lastly, it can neither discriminate
between low skill and adaptation. The lack of a tool to objectively measure the
full range of SMS is a major gap in diagnosing and treating stuttering, leaving
speech therapists unable to adequately expose the core underlying skill of
children who stutter.

The first aim is to determine an objective acoustic measure to define SMS by
comparing acoustic features of probe words to a perceptual score of speech
motor movement characteristics and stuttering frequency judged by the expert
(i.e., the current golden standard).
The second aim is to ascertain whether the acoustic variables of SMS can
differentiate between children who do and who do not stutter, in children who
recovered from stuttering and children who persist in stuttering.
o ascertain whether the acoustically measured SMS of groups (i.e. children who
do and who do not stutter, in children who recovered from stuttering and
children who persist in stuttering) differs for low and high levels of word
complexity on a) the motor and b) the cognitive-linguistic level, when
correcting for lip and jaw movement adaptations.

Observational case-control study.
Children will be assessed at Erasmus MC Sophia Children*s Hospital. First,
standard hearing and articulation tests will be taken to exclude children with
hearing and articulation problems (other than stuttering). Then, the main,
experimental speech tasks will be assessed in a sound-proof booth and children
will be visually and auditory recorded. The probe task exists of words and
non-words. In a child friendly presentation, the target (non-)words will be
presented auditory and children will be requested to repeat every item 8 times,
enabling measuring the variability in the acoustic signal between these
repetitions of the same item. The first part of the task consists of non-words
of an existing perceptual speech task: the Oral Motor Assessment Scale (1). The
second part consists of target (non-)words in two complexity levels (simple vs.
complex) on three variables, a-c: (a) articulation rate (normal rate vs.
up-tempo rate) (b) mouth opening at word initiation (large/open vowel words vs.
small/closed vowel words) and (c) word length (2 syllable vs. 4-5 syllable
words). The assessment, including sufficient pause time for the child, can be
completed within one hour.

The nature and duration of the hearing test and speech tasks are comparable to
tasks children are used to do in speech therapy treatment. The assessments are
not invasive, not cognitively difficult and not social-emotional burdensome.
Therefore, the risks of this study are negligible and the burden will be
minimal.
An individual*s results of the speech tasks may contribute to his/her diagnosis
of stuttering and may add to a tailormade treatment advice.
The research is regarded as group-related since it was aimed to ascertain the
speech motror skills of children who stutter, children who recovered from
stuttering, compared to children who do not stutter.