Neuropsychiatric correlates of white matter hyperintensities in Alzheimer's disease

Abstract

Objective: To investigate the association of behavioral and psychological symptoms of dementia (BPSD) in Alzheimer's disease (AD) and magnetic resonance imaging (MRI) measures of brain atrophy and white matter hyperintensities (WMH).

Methods: Thirty-seven patients with probable AD received the Neuropsychiatric Inventory (NPI), the Mini Mental Status Exam (MMSE), and an MRI scan as part of their initial evaluation at the Outpatient Memory Diagnostic Clinic at McLean Hospital. MRI-based volumetric measurements of whole brain atrophy, hippocampal volumes, and WMH were obtained. Analysis of covariance models, using age as a covariate and the presence of specific BPSD as independent variables, were used to test for differences in whole brain volumes, hippocampal volumes and WMH volumes.

Results: Increased WMH were associated with symptoms of anxiety, aberrant motor behavior, and night time disturbance, while symptoms of disinhibition were linked to lower WMH volume. No associations were found for whole brain or hippocampal volumes and BPSD.

Conclusions: These findings suggest that white matter changes are associated with the presence of BPSD in AD.

Figure 1. Automated measurement and normalization of whole brain volumes

1A) T1-weighted images were loaded into the SIENAX algorithm. 1B) This program identifies brain and skull surfaces using a deformable model and produces images containing only brain tissue. 1C) The skull surface is registered to a standard atlas to create a normalization scaling factor. 1D) Voxels representing brain tissue are segmented from those representing cerebrospinal fluid and used to calculate raw whole brain volumes. This raw volume is multiplied by the scaling factor to produce normalized whole brain volumes.

Figure 2. Coronal and sagittal views of the hippocampus

Figure 3. Quantification of FLAIR hyperintensities

A.) Original fluid attenuated inversion recovery (FLAIR) image. B) Initial threshold parameters set to include all voxels containing hyperintensities. C) Final hyperintensity region of interest with non-brain voxels excluded.

Figure 4. Dendrogram derived from cluster analysis showing the relatedness between the 11 behavioral symptoms of the NPI based on their presence or absence

Figure 5. White matter hyperintensity volumes and the presence of behavioral and psychological symptoms of dementia (BPSD) in probable Alzheimer’s disease

Least squares means and standard error estimates after covarying for age and other BPSD. † p < 0.10, * p < 0.05, ** p <0.005