Diffusion tensor imaging detects white matter abnormalities and associated cognitive deficits in chronic adolescent TBI

Abstract

Primary objective: This study examined long-term alterations in white matter microstructure following TBI in adolescence using diffusion tensor imaging (DTI). It was hypothesized that white matter integrity would be compromised in adolescents with TBI and would correlate with measures of executive functioning and cognitive abilities.

Research design: This study employed whole-brain, voxel-wise, statistical comparison of DTI indices in youth of 12-17 years old (mean = 15.06) with TBI vs an age- and gender-matched cohort (mean age = 15.37).

Methods and procedures: This study scanned 17 adolescents with complicated-mild-to-severe TBI, 1-3 years after injury, and 13 healthy adolescents. Tract-Based Spatial Statistics (TBSS) was employed for DTI analysis.

Main outcomes and results: Overall diffusivity elevations were found in the TBI group with increases in axial diffusivity in the right hemisphere. White matter integrity was associated with word reading, planning and processing times in the TBI group, but not healthy controls.

Conclusions: The detected abnormalities in axial diffusivity may reflect neuronal regeneration and cerebral reorganization after injury. These findings provide tentative evidence of persistent white matter alteration following TBI in adolescence. Associations of DTI indices with cognitive performance following TBI provide tentative support for links between white matter integrity and performance post-TBI.

Figure 1

Results of TBSS analysis comparing axial diffusivity (AD) between adolescents with TBI (TBI) and an age-gender matched control group (CTL)., Voxels with significant differences (TBI > CTL) are shown in red/yellow (p < 0.05 corrected) and blue (p <0.1 corrected); the significant regions were ‘‘thickened’’ to enhance visualization. The results are overlaid onto the MNI152 T1 template and the white matter tract skeleton is shown in green.

Figure 2

Comparison of the DTI variables (i) AD, (ii) MD, (iii) RD, and (iv) FA between TBI and CTL groups using subject-wise mean values across all voxels in the significant cluster for AD (TBI >CTL) (see Figure 1) Group-wise mean values are denoted by horizontal lines. Significant differences between groups after correction for multiple comparisons are denoted by *(p < 0.05), **(p < 0.01). Note that AD (TBI > CTL) (i) is bound to be significant given that the ROI was chosen from voxels of significantly increased AD in the TBI group as determined by TBSS.

Figure 3

Results of the TBSS regression analysis of FA with the WRAT-4 word reading subtest (p < 0.05, corrected) (i), RD with the DKEFS Tower Total Achievement Score (p < 0.05, corrected) (ii), and AD with the WISC-IV PSI score (p < 0.1, corrected) (iii) in the TBI group (positive correlation: red, negative correlation: blue). The significantly correlated regions were ‘‘thickened’’ to enhance visualization. The results are overlaid onto the MNI152 T1 template and the white matter tract skeleton is shown in green. The p-values are thresholded at p < 0.05 corrected over voxels that were reported as significant for AD (TBI > CTL) (see Figure 1).