Brain developmental trajectories associated with childhood stuttering persistence and recovery

Abstract

Stuttering is a neurodevelopmental disorder affecting 5-8 % 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 (3-5 years old) and school age (6-12 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.

Keywords: Basal ganglia thalamocortical loop; Gray and white matter volume; Neurodevelopmental disorders; Speech; Voxel-based morphometry.

Fig. 1

Differences between preschool-age children with persistent stuttering and controls (3–5 years old) in gray matter volume (A) and white matter volume (B). In the left panel, between-group differences are overlaid on a single subject anatomical image. Orange and blue indicate increased and decreased volume at uncorrected p < 0.05, respectively. Clusters exhibiting a significant difference at corrected p < 0.05 are outlined by black lines. In right panel, violin plots illustrate the group differences around the peak of a significant cluster. Each marker represents the modulated volume values averaged across voxels around the peak location in each scan. The values were adjusted for the effects of sex, age, IQ, intracranial volume and socioeconomic status. SLF: superior longitudinal fasciculus, CR: corona radiata.

Fig. 2

Differences between school-age children with persistent stuttering and controls (3–5 years old) in growth rate of white matter volume. In the left panel, between-group differences in growth rate are overlaid on a single subject anatomical image. Orange and blue indicate increased and decreased growth rate at uncorrected p < 0.05, respectively. Clusters exhibiting a significant difference at corrected p < 0.05 are outlined by black lines. In the right panel, individual volume measures are plotted against age to illustrate the age-related changes in each group around the peak of a significant cluster. Each marker represents the modulated volume values averaged across voxels around the peak location in each scan. The values were adjusted for the effects of sex, IQ, intracranial volume and socioeconomic status. Longitudinal volume measures of a participant are connected using dotted lines. Solid lines represent the best-fit linear trend line in each group and the shaded areas represent the standard error of the trend line. SLF: superior longitudinal fasciculus, CR: corona radiata, AF; arcuate fasciculus, ILF: inferior longitudinal fasciculus.

Fig. 3

Differences between older children with persistent stuttering and controls (6–12 years old) in gray matter volume (A) and white matter volume (B). In the left panel, between-group differences are overlaid on a single subject anatomical image. Orange and blue indicate increased and decreased volume at uncorrected p < 0.05, respectively. Clusters exhibiting a significant difference at corrected p < 0.05 are outlined by black lines. In right panel, the violin plots illustrate the group differences around the peak of a significant cluster. Each marker represents the modulated volume values averaged across voxels around the peak location in each scan. The values were adjusted for the effects of sex, age, IQ, intracranial volume and socioeconomic status. SLF: superior longitudinal fasciculus, CR: corona radiata.

Fig. 4

Differences between older children with persistent stuttering and controls (6–12 years old) in growth rate of gray matter volume (A) and white matter volume (B). Between-group differences in growth rate are overlaid on a single subject anatomical image. Orange and blue indicate increased and decreased growth rate at uncorrected p < 0.05, respectively. Clusters exhibiting a significant difference at corrected p < 0.05 are outlined by black lines. In the scatter plots, individual volume measures are plotted against age to illustrate the age-related changes around the peak of a significant cluster in each group. Each marker represents the modulated volume values averaged across voxels around the peak location in each scan. The values were adjusted for the effects of sex, IQ, intracranial volume and socioeconomic status. Longitudinal volume measures of a participant are connected using dotted lines. Solid lines represent the best-fit linear trend lines in each group and the shaded areas represent the standard errors of the trend lines.

Fig. 5

Differences between children with persistent stuttering and controls when scans from all subjects from age 3–12 years old were analyzed together in a single model. (A) Between group differences in gray matter volume are overlaid on a single subject anatomical image. Orange and blue indicate increased and decreased volume at uncorrected p < 0.05, respectively. Clusters exhibiting a significant difference at corrected p < 0.05 are outlined by black lines. The violin plots illustrate the group differences around the peak of a significant cluster. The values were adjusted for the effects of sex, age, IQ, intracranial volume and socioeconomic status. (B) Between group differences in growth rate of gray matter are overlaid on a single subject anatomical image. The scatter plots illustrate the growth rate differences around the peak of a significant cluster in each group. Each marker represents the modulated volume values averaged across voxels around the peak location in each scan. The values were adjusted for the effect of sex, IQ, intracranial volume and socioeconomic status. Longitudinal volume measures of a participant are connected using dotted lines. Solid lines represent the best-fit linear trend lines in each group and the shaded areas represent the standard errors of the trend lines. vPMC: Ventral premotor cortex.

Fig. 6

Associations between gray/white matter volume and stuttering severity scores in children who stutter from age 3–5 years old (A) and 6–12 years old (B). Estimated association between stuttering severity and volume measures in each voxel is overlaid on a single subject anatomical image. Orange and blue indicate positive and negative association at uncorrected p < 0.05, respectively. Clusters exhibiting a significant association at corrected p < 0.05 are outlined by black lines. In the scatter plots, individual volume measures around the peak of a significant cluster are plotted against stuttering severity scores. Each marker represents the modulated volume values averaged across voxels around the peak location in each scan. The values were adjusted for the effect of sex, age, IQ, intracranial volume and socioeconomic status. Longitudinal volume measures of a participant are connected using dotted lines. Solid lines represent the best-fit linear trend lines in each group and the shaded areas represent the standard errors of the trend lines.

Fig. 7

Differences between children who recovered from stuttering and controls when scans from all participants from age 3–12 years old were analyzed together in a single model. Between group differences in growth rate of gray matter volume (A) and white matter volume (B) are overlaid on a single subject anatomical image. Orange and blue indicate increased and decreased volume at uncorrected p < 0.05, respectively. Clusters exhibiting a significant difference at corrected p < 0.05 are outlined by black lines. The scatter plots illustrate the growth rate differences around the peak of a significant cluster in each group. The values were adjusted for the effects of sex, IQ, intracranial volume and socioeconomic status. Each marker represents the modulated volume values averaged across voxels around the peak location in each scan. Longitudinal volume measures of a participant are connected using dotted lines. Solid lines represent the best-fit linear trend lines in each group and the shaded areas represent the standard errors of the trend lines. vPMC: Ventral premotor cortex, SLF: superior longitudinal fasciculus, CR: corona radiata.