This study investigates the relationship between the microstructural properties of language-related white matter (WM) tracts and hand preference in typically developing school-aged children. Specifically, two research questions are formulated. First, is there a difference in WM integrity of the studied left- and right-hemispheric WM tracts between left- and right-handed children? Second, does handedness influence the lateralization pattern of the microstructural properties of the studied WM tracts? DTI imaging was carried out in 36 right-handed (17 males; mean age = 11;09 years) and 29 left-handed (12 males; mean age = 11;06 years) typically developing children on a 3T scanner (Philips, Best, The Netherlands), using a 32-channel head coil. To assess structural connectivity, DTI data were acquired using a spin-echo-echo-planar-imaging pulse sequence with 60 diffusion directions (b=1300s/mm²) and with an isotropic resolution of 2mm. Pre- and post-processing analyses were performed using ExploreDTI₁. Subsequently, deterministic fiber tractography of the superior, inferior and middle longitudinal fascicle (SLF, ILF & MdLF resp.), uncinate fascicle (UF) and extreme capsule fiber system (ECFS) was performed in both hemispheres, using the robust ROI definition protocols of Catani₂, Wakana₃ or Makris₄ on a whole brain fiber tractography data set. Next, the SLF was divided into an anterior, longitudinal (arcuate fascicle) and posterior segment₂. The white matter organization of each tract, measured by the mean fractional anisotropy (FA), was compared between age-, gender- and IQ-matched left-handed and right-handed children using dependent samples t-tests (N_pairs = 29). A lateralization index (LI) was calculated for each tract by using the following formula: (FA_{left hemisphere} - FA_{right hemisphere}) / (FA_{left hemisphere} + FA_{right hemisphere}). A single t-test was carried out for both groups separately to detect whether the LI differs significantly from zero, which indicates FA-asymmetry. Equivalent non-parametric tests were used when one or more of the parametric test’s assumptions were violated. The significance threshold was set at ɑ < 0.05 and the Benjamini-Hochberg procedure (FDR = 0.10) was applied to correct for multiple comparisons.

In all children, the SLF, ILF, MdLF, UF and ECFS could be reconstructed in both hemispheres. In contrast, reconstruction of the left-hemispheric longitudinal segment (AF) of the SLF, revealed no detectable left-hemispheric AF in 17% (5/29) of the left-handed children. However, in all right-handed children the left AF could be reconstructed. The left-hemispheric AF was significantly more absent in left-handed children compared to right-handed children (two-tailed Fisher’s exact test, p = 0.014). For the anterior and posterior segments, no significant differences in the presence versus absence ratios were observed between left- and right-handed children. Success rates are presented in Table 1.

The only significant group difference in white matter structure was found in the left anterior SLF-segment. Specifically, the mean FA was significantly higher in the left-handed group compared to the right-handed group (Hedge’s g* = 0.79, p = 0.006, Benjamini-Hochberg corrected).

Despite a large inter-subject heterogeneity in the LI of FA, our results of the single t-test analyses suggested a clear effect of handedness on LI (Table 2). In right-handed children, the mean FA of all studied tracts (except the ECFS) and all SLF-segments differed significantly from zero after applying the Benjamini-Hochberg correction. More precisely, the mean FA of the anterior SLF-segment was lateralized to the right hemisphere, whereas all other tracts had a significantly higher mean FA in the left hemisphere. In contrast, in left-handed children, none of the studied WM tracts showed a clear left- or right-hemispheric lateralization.

Our tractography results provide evidence for a different structural connectivity pattern of the language connectome in left-handed children. Handedness seems to have a major effect on the lateralization pattern of the language related white matter tracts. Whereas right-handed children show a clear left-lateralized structural language network, our group of left-handed children seems to have a more bilateral organized language system. Those observed differences in WM microstructure and lateralization might reflect an interaction between handedness and the neural processing of language in children. However, further structural and functional imaging studies are needed to disentangle the mysteries of the left-handed brain and its relation to language processing.