Common and unique gray matter correlates of episodic memory dysfunction in frontotemporal dementia and Alzheimer's disease

Abstract

Conflicting evidence exists regarding the integrity of episodic memory in the behavioral variant of frontotemporal dementia (bvFTD). Recent converging evidence suggests that episodic memory in progressive cases of bvFTD is compromised to the same extent as in Alzheimer's disease (AD). The underlying neural substrates of these episodic memory deficits, however, likely differ contingent on dementia type. In this study we sought to elucidate the neural substrates of episodic memory performance, across recall and recognition tasks, in both patient groups using voxel-based morphometry (VBM) analyses. We predicted that episodic memory dysfunction would be apparent in both patient groups but would relate to divergent patterns of neural atrophy specific to each dementia type. We assessed episodic memory, across verbal and visual domains, in 19 bvFTD, 18 AD patients, and 19 age- and education-matched controls. Behaviorally, patient groups were indistinguishable for immediate and delayed recall, across verbal and visual domains. Whole-brain VBM analyses revealed regions commonly implicated in episodic retrieval across groups, namely the right temporal pole, right frontal lobe, left paracingulate gyrus, and right anterior hippocampus. Divergent neural networks specific to each group were also identified. Whereas a widespread network including posterior regions such as the posterior cingulate cortex, parietal and occipital cortices was exclusively implicated in AD, the frontal and anterior temporal lobes underpinned the episodic memory deficits in bvFTD. Our results point to distinct neural changes underlying episodic memory decline specific to each dementia syndrome.

Keywords: Alzheimer's disease; episodic memory; frontotemporal dementia; medial temporal lobes; posterior cingulate cortex; voxel-based morphometry.

Figure 1

Bar chart showing overall memory performance (percentage correct) across verbal and visual recall and recognition for healthy Controls, behavioral variant FTD and AD participants. Error bars represent standard error of the mean.

Figure 2

VBM analyses showing brain areas of decreased gray matter intensity in (A) AD patients in comparison with Controls, and (B) behavioral variant FTD patients in comparison with Controls. Colored voxels show regions that were significant in the analysis with P < 0.001 uncorrected for all contrasts, with a cluster threshold of 100 contiguous voxels. All clusters reported t > 3.9. Clusters are overlaid on the MNI standard brain. [Color figure can be viewed in the online issue, which is available at http://wileyonlinelibrary.com.]

Figure 3

VBM analyses showing brain regions in which gray matter intensity correlates significantly with episodic recall performance in (A) AD in comparison with Controls, and (B) behavioral variant FTD compared with Controls. Colored voxels show regions that were significant in the analysis with P < 0.001 uncorrected with a cluster threshold of 100 contiguous voxels. All clusters reported t > 3.9. Clusters are overlaid on the MNI standard brain. [Color figure can be viewed in the online issue, which is available at http://wileyonlinelibrary.com.]

Figure 4

VBM analyses showing brain regions in which gray matter intensity correlates with recall performance in (A) both bvFTD and AD, (B) exclusively in AD, and (C) exclusively in bvFTD. Colored voxels show regions that were significant in the analysis with P < 0.001 uncorrected with a cluster threshold of 100 contiguous voxels for exploratory whole‐brain analyses (B and C) and 50 contiguous voxels for the overlap analysis (A). All clusters reported t > 3.9. Clusters are overlaid on the MNI standard brain. [Color figure can be viewed in the online issue, which is available at http://wileyonlinelibrary.com.]