Abstract

     Stuttering is a neurodevelopmental disorder affecting 5–۸ % of preschool-age children, continuing into adulthood in 1 % of the population. The neural mechanisms underlying persistence and recovery from stuttering remain unclear and little information exists on neurodevelopmental anomalies in children who stutter (CWS) during preschool age, when stuttering symptoms typically first emerge. Here we present findings from the largest longitudinal study of childhood stuttering to date, comparing children with persistent stuttering (pCWS) and those who later recovered from stuttering (rCWS) with age-matched fluent peers, to examine the developmental trajectories of both gray matter volume (GMV) and white matter volume (WMV) using voxel-based morphometry. A total of 470 MRI scans were analyzed from 95 CWS (72 pCWS and 23 rCWS) and 95 fluent peers between 3 and 12 years of age. We examined overall group and group by age interactions in GMV and WMV in preschool age (3–۵ years old) and school age (6–۱۲ years old) CWS and controls, controlling for sex, IQ, intracranial volume, and socioeconomic status. The results provide broad support for a possible basal ganglia-thalamocortical (BGTC) network deficit starting in the earliest phases of the disorder and point to normalization or compensation of earlier occurring structural changes associated with stuttering recovery.

Introduction

     Developmental stuttering is a complex neurodevelopmental disorder (Smith and Weber, 2017) that disrupts the fluent flow of speech production. It is characterized by frequently occurring involuntary repetitions and prolongations in speech sounds, in addition to prolonged articulatory posture, and/or avoidance and struggle behaviors (Van Riper, 1971). Stuttering affects 5–۸ % of preschool-age children (Yairi and Ambrose, 2013) and remains as a chronic speech disorder in 1 % of the general population. Typical onset of stuttering is reported to occur at 30–۴۸ months (Bloodstein and Ratner, 2008, Reilly et al., 2013, Yairi and Ambrose, 2005) with approximately 80 % of children naturally recovering 24–۳۶ months after the onset of stuttering (Yairi and Ambrose, 2013, Yairi and Ambrose, 2005, Yairi and Ambrose, 1999). Around this same developmental period, the neural systems supporting executive function, language, and speech-motor control undergo rapid and vigorous development (Smith and Weber, 2017, Almli et al., 2007, Chang et al., 2019, Friederici, 2006, Gilmore et al., 2018).

     In the past 20 years, an increasing number of neuroimaging studies have been conducted to understand the possible neural bases of stuttering. Convergent findings from systematic reviews and meta-analyses have highlighted several neuroanatomical differences in children and adults who stutter (AWS) compared to fluent speakers (Smith and Weber, 2017, Chang et al., 2019, Belyk et al., 2015, Budde, 2014, Ingham et al., 2005, Neef et al., 2015). Structural and functional anomalies in the left speech motor neural system, including the inferior frontal gyrus (IFG), posterior superior temporal gyrus (STG), basal ganglia-thalamo-cortical (BGTC) loop and the cerebellum have been associated with stuttering (see Chang et al., 2019 for review). In addition to gray matter anomalies, children who stutter (CWS) were found to have less white matter volume (WMV) bilaterally in the forceps minor of the corpus callosum (Beal et al., 2013) while another study found no significant differences (Choo et al., 2012).

Conclusion

     Regional gray and white matter volume differences across the whole brain were examined in preschool- and school-age CWS (both persistent and recovered) compared to age matched peers in the largest MRI dataset collected for this clinical population to date. Longitudinally collected scans that spanned up to 4 years per participant allowed tracking of developmental trajectories that differentiated among the groups. The results provide broad support for a possible BGTC network deficit starting at the early phase (preschool-age) of the disorder, involving the putamen, nucleus accumbens, left IFG/vPMC, and corticostriatal tracts. The deficits in input regions of the BGTC network appear to affect the output regions of the network including the thalamus in later phases of stuttering (school-age). The current data also provide insights into neural bases of natural recovery from stuttering during childhood. Children who recover from stuttering showed increased WMV in motor projection fibers, left AF, corpus callosum, and cerebellar peduncle around the dentate nuclei, suggesting that normalization or successful compensation of stuttering-linked neural deficits. Similar volume increases in corpus callosum, and cerebellar peduncle were also observed in pCWS, indicating an incomplete compensation. These results provide substantial new insights into possible neural bases of stuttering onset, persistence, and recovery during childhood.