We set out to study whether single-subject gray matter (GM) networks show disturbances that are specific for Alzheimer's disease (AD; n = 90) or behavioral variant frontotemporal dementia (bvFTD; n = 59), and whether such disturbances would be related to cognitive deficits measured with mini-mental state examination and a neuropsychological battery, using subjective cognitive decline subjects as reference. AD and bvFTD patients had a lower degree, connectivity density, clustering, path length, betweenness centrality, and small world values compared with subjective cognitive decline. AD patients had a lower connectivity density than bvFTD patients (F = 5.79, p = 0.02; mean ± standard deviation bvFTD 16.10 ± 1.19%; mean ± standard deviation AD 15.64 ± 1.02%). Lasso logistic regression showed that connectivity differences between bvFTD and AD were specific to 23 anatomical areas, in terms of local GM volume, degree, and clustering. Lower clustering values and lower degree values were specifically associated with worse mini-mental state examination scores and lower performance on the neuropsychological tests. GM showed disease-specific alterations, when comparing bvFTD with AD patients, and these alterations were associated with cognitive deficits.
Keywords: Alzheimer's disease; Behavioral variant frontotemporal dementia; Cognition; Single-subject gray matter networks; Structural networks.
Copyright © 2016 Elsevier Inc. All rights reserved.