Distinct influence of specific versus global connectivity on the different Alzheimer's disease biomarkers

Recent findings suggest that the topography and propagation of lesions in Alzheimer's disease are related to functional connectivity, either showing that regions of high global connectivity are more vulnerable or that lesions propagate neuron-to-neuron from a starting area called the epicentre, thus involving specific connectivity. However, the relative influence of specific and global connectivity and their differential impact on the three main neuroimaging biomarkers of the disease (atrophy, hypometabolism and amyloid-beta deposition) have never been investigated to date. Forty-two healthy elderly subjects and 35 amyloid-beta positive amnestic mild cognitive impairment and Alzheimer's disease patients underwent resting-state functional MRI, anatomical T-1-weighted MRI, F-18-fluorodeoxyglucose-PET and florbetapir-PET scans. All patients also underwent follow-up T-1-weighted MRI, F-18-fluorodeoxyglucose-PET and florbetapir-PET scans 18 months later to assess the lesion propagation. The epicentre was defined per modality as the most altered region at baseline in patients compared to controls. Maps of global and specific functional connectivity were computed from the resting-state functional MRI data of the healthy elderly subjects. Global connectivity corresponds to the connectivity strength of each grey matter area with the rest of the brain (i.e. all other grey matter areas) while specific connectivity refers to the connectivity of a single specific brain region (the epicentre) with the rest of the brain (i.e. all other brain regions). Maps of baseline alterations and propagation were computed for grey matter atrophy, hypometabolism and amyloid-beta deposition in patients. Regression analyses were performed across the 239 brain regions to assess the links between global or specific functional connectivity in healthy elderly subjects and Alzheimer's disease-related baseline disruptions or alteration propagation. Atrophy at baseline was predicted by specific connectivity and inversely correlated with global connectivity, while hypometabolism and amyloid-beta deposition were positively influenced by both global and specific connectivity. Regarding longitudinal changes, atrophy spread in regions with high specific connectivity while hypometabolism propagated in areas showing high global connectivity. This is the first study to show that global connectivity has an opposite relationship with atrophy versus hypometabolism and amyloid-beta deposition, suggesting that the high level of functional connectivity found in hubs exerts a differential influence on these Alzheimer's disease lesions. These results sustain the hypotheses of higher vulnerability of hubs to hypometabolism and amyloid-beta deposition versus transneuronal propagation of atrophy from the epicentre to connected regions, in Alzheimer's disease. Global and specific connectivity exert a differential influence on, and provide complementary information to predict, the topography of Alzheimer's disease lesions and their propagation.