Developmental stuttering is a disorder of speech fluency, primarily characterized by prolongations, blocks and repetitions of sounds and/or syllables. The etiology of stuttering is not fully understood to date. One of the earliest theories on stuttering relates the disorder to atypical cerebral dominance (Moore, 1984, Travis, 1978, Travis and Johnson, 1934 and Webster, 1997). Although initial attempts to provide evidence for this theory were mostly unsuccessful (see Kushner, 2012), modern functional brain imaging studies have established that adults-who-stutter (AWS) indeed exhibit different functional lateralization when compared to fluent speakers (Braun et al., 1997, De Nil et al., 2001, De Nil et al., 2000, Kell et al., 2009, Neumann et al., 2005 and Pool et al., 1991). These studies, as a whole, demonstrate that regions in the right hemisphere, particularly in the frontal cortex, are over-activated in AWS (see Brown, Ingham, Ingham, Laird, & Fox, 2005).
There is an ongoing debate on the functional significance of the right frontal over-activation observed in developmental stuttering. Some authors suggest that the greater recruitment of the right hemisphere is beneficial (Braun et al., 1997, Kell et al., 2009, Neef et al., 2011 and Preibisch et al., 2003), whereas others suggest it is not (Brown et al., 2005, Chang et al., 2010, Fox et al., 2000, Kronfeld-Duenias et al., 2014 and Moore, 1984). There are also suggestions that the right hemisphere recruitment is maladaptive (Andrews et al., 1972 and Webster, 1997), that it is an outcome of negative emotions (Forster, 1995 and Webster, 1993), or causally related to overt stuttering behavior (Boberg et al., 1983, Fox et al., 1996 and Wood et al., 1980). In fact, some combination of the above explanations could be true, given that over-activations were detected in several distinct right frontal regions. As the debate is still open after more than two decades of functional imaging studies on stuttering, alternative methodological approaches may be necessary.
The right frontal over-activation observed in AWS could be better understood in the context of the underlying structural properties of their brains. Several studies conducted over the last decade have detected structural anomalies in stuttering individuals, frequently in the form of reduced fractional anisotropy (FA) in white matter regions (see Cai et al., 2014 and Cykowski et al., 2010). The majority of FA reductions are in the left hemisphere, most notably in the left Rolandic Operculum (RO) (Chang et al., 2008, Connally et al., 2014, Kell et al., 2009, Sommer et al., 2002 and Watkins et al., 2008). This is commonly attributed to white matter tracts involved in speech motor control (Civier et al., 2013 and Cykowski et al., 2010). Previous studies suggest that these left hemisphere anomalies are most likely related to the origin of the disorder (Chang et al., 2008 and Kell et al., 2009), and that the right frontal cortex is recruited to cope with the deficiency (Chang et al., 2011, Chang et al., 2008, Chang et al., 2010, Kell et al., 2009, Neef et al., 2011, Preibisch et al., 2003, Sowman et al., 2014, Tourville and Guenther, 2011 and Tourville et al., 2008). Following such an interhemispheric reorganization, the right hemisphere may carry tasks usually carried out by the left hemisphere (e.g., Karbe et al., 1998).
We propose that interhemispheric reorganization in developmental stuttering may involve changes in the main highway connecting the hemispheres, namely, the corpus callosum. Indeed, several callosal anomalies were reported in stuttering individuals, with most studies pointing to the forceps minor (Beal et al., 2013, Choo et al., 2011, Cykowski et al., 2010 and Kell et al., 2009). This interhemispheric pathway connects the lateral and medial frontal cortices and crosses the midline via the genu of the corpus callosum (Abe et al., 2004). As the corpus callosum regulates the division of labor between the hemispheres (Geschwind & Galaburda, 1985), callosal differences observed in AWS might reflect subcortical plasticity that shifts control of speech production from the dysfunctional left hemisphere to the intact right hemisphere. But plasticity is not always beneficial: in acquired disorders, such as aphasia, interhemispheric reorganization is often deleterious (Hamilton, Chrysikou, & Coslett, 2011). Similarly, we hypothesize that reorganization-related callosal differences may intensify stuttering, possibly due to recruitment of brain regions not well adapted for speech production.
The goal of this study was to examine the relation between micro-structural properties of callosal connections and the level of speech fluency in adults who stutter. For the purpose of this paper, speech fluency is defined as the ability to speak without stuttering (note that fluency here does not concern articulatory rate, language proficiency, normal interruptions in speech flow, etc.). We first established reliable group difference in FA in the corpus callosum of AWS versus matched controls. We then conducted a focused correlation analysis within the AWS group, and examined the relation between speech fluency and FA in the implicated corpus callosum region. We were interested not only in the identification of a significant correlation, but more importantly, in the direction of the relation. We reasoned that if the least fluent individuals showed the most extreme anomaly in the callosal tracts, this would be considered evidence against beneficial plasticity. Finally, we also examined the contribution of age and its interaction with stuttering, in explaining white matter variability in the callosum and in other white matter regions.
