DTI-derived indexes of brain WM correlate with cognitive performance in vascular MCI and small-vessel disease : a TBSS study

Indexes derived from diffusion tensor imaging (DTI) are sensitive to changes of both T2-hyperintense and normal-appearing brain white matter (WM) in elderly subjects with variable cognitive status. We investigated correlations between global cognitive performance and DTI-derived indexes along the WM tracts in the brain of patients with vascular mild cognitive impairment (MCI) and small vessel disease (SVD). Seventy-six patients with vascular MCI and SVD were assessed through Montreal Cognitive Assessment (MoCA) and Mini Mental State Examination (MMSE) test and underwent DTI examination on a 1.5 T MR scanner. We used Tract Based Spatial Statistics (TBSS) to assess voxel-wise in the entire brain the spatial distribution of the correlation between values of fractional anisotropy, mean, axial/radial diffusivity and global cognitive performance as assessed with MoCA and MMSE tests. All correlations were statistically tested with a significant p-value <0.05 using a family-wise error correction for multiple comparisons. The MoCA score significantly correlated with fractional anisotropy (positive correlation) and mean, axial and radial diffusivity (negative correlations) in WM tracts of cerebral hemispheres and corpus callosum, as well as in the intra-thalamic WM tracts and the superior cerebellar peduncle decussation in the midbrain. No significant correlations were observed for MMSE score. Global cognitive performance, as measured by the MoCA score, in patients with vascular MCI and SVD is associated with microstructural changes in WM tracts underlying intra- and inter-hemispheric cerebral, thalamo-cortical and cerebello-thalamic connections.