Different patterns of white matter degeneration using multiple diffusion indices and volumetric data in mild cognitive impairment and Alzheimer patients

Abstract

Alzheimeŕs disease (AD) represents the most prevalent neurodegenerative disorder that causes cognitive decline in old age. In its early stages, AD is associated with microstructural abnormalities in white matter (WM). In the current study, multiple indices of diffusion tensor imaging (DTI) and brain volumetric measurements were employed to comprehensively investigate the landscape of AD pathology. The sample comprised 58 individuals including cognitively normal subjects (controls), amnestic mild cognitive impairment (MCI) and AD patients. Relative to controls, both MCI and AD subjects showed widespread changes of anisotropic fraction (FA) in the corpus callosum, cingulate and uncinate fasciculus. Mean diffusivity and radial changes were also observed in AD patients in comparison with controls. After controlling for the gray matter atrophy the number of regions of significantly lower FA in AD patients relative to controls was decreased; nonetheless, unique areas of microstructural damage remained, e.g., the corpus callosum and uncinate fasciculus. Despite sample size limitations, the current results suggest that a combination of secondary and primary degeneration occurrs in MCI and AD, although the secondary degeneration appears to have a more critical role during the stages of disease involving dementia.

Figure 1. Distribution of diffusion tensor MRI indices for the global WM ROI in controls, MCI and AD patients.

Multivariate Analysis reported significant differences between groups for FA (F = 4.649, df = 2, P<0.05), axial diffusivity (F = 5.610, df = 2, P<0.05), MD (F = 6.821, df = 2, P<0.005) and radial diffusivity (F = 7.131, df = 2, P<0.005). *P<0.05, **P<0.01, corrected for multiple comparisons using Bonferroni correction.

Figure 2. TBSS maps showing voxelwise comparisons between patients and controls.

The mean FA skeleton (green voxels) projected on the FMRIB template brain. Low FA in AD patients in is shown in dark red (A); low FA in MCI is shown in yellow-red (B); high MD in AD is depicted in blue (C) and high radial diffusivity in orange.

Figure 3. Overlapping WM regions between controls and AD for specific DTI indices.

Areas of overlap between low FA (red) and high MD (blue) suggest gross tissue loss and are depicted in the upper panels (A) and (B); FA decreases (red) overlapping with radial diffusivity increases (orange) are shown in the lower panels (C) and (D) and suggest areas of myelin damage.

Figure 4. FA results adjusting for GM atrophy differences between groups.

Upper panels (A) and (B): TBSS maps show low FA in the AD group in comparison to controls. The figures results from two GLM models designed for group comparisons. The model 1 included age and gender as covariates and is represented by the sum of blue and red voxels. The model 2 added as covariate the GM volume to the model 1 and is identified as red voxels. Red areas representing the remaining voxels after controlling for the GM atrophy can be seen mostly in the left hemisphere. They include WM tracts such as the left anterior corona radiate and the left and right corpus callosum (A) and the right cingulate gyrus (B). For a detailed description see the results section. Lower panels (C) and (D): TBSS maps show low FA in the MCI group relatively to controls. Note that statistical significant areas resemble those from the AD-control contrast seen before in the model 2 (upper panel).