ALZHEIMER’S DISEASE PATIENTS WITH HIGHER COGNITIVE RESERVE PRESENT MORE EFFICIENT NETWORK TOPOLOGY

Abstract

AD is the most common form of dementia worldwide, and yet there is still no means of curing or even preventing the disease. It may be possible, however, to delay the onset of AD by reducing some modifiable factors. In this context, innumerable epidemiological studies report that greater cognitive reserve (CR) may postpone the clinical expression of AD. In the present study we sought to investigate: a) the relationship between education and resting-state fMRI graph metrics in healthy controls, amnestic mild cognitive impairment subjects (aMCI) and mild Alzheimer's disease (AD) patients; b) the relationship between cerebrospinal fluid (CSF) biomarkers’ levels and graph metrics in aMCI and mild AD patients; c) the relationship between CSF biomarkers’ levels and education in aMCI and mild AD patients. 14 mild AD patients, 31 aMCI, and 28 controls were scanned on a 3.0T MR scanner, and CSF samples were obtained from aMCI and AD patients. Education was used as a measure of cognitive reserve. We used the UF2C toolbox to construct functional connectivity matrices based on 70 functional regions. Graph metrics were computed with the GraphVar toolbox, network-based statistics corrected for multiplicity. The relationship between CSF and education was assessed with multiple regression models. Education positively correlated to measures of centrality (betweenness centrality – Bc and eigenvector centrality - v), segregation (clustering coefficient - C), and integration (efficiency - E) in several regions in controls, aMCI and mild AD patients. CSF abnormal levels of proteins (beta amyloid, total-tau and phospho-tau) significantly related to disruption in network topology in aMCI and mild AD patients. No relationship was found between CSF biomarkers and education. Although cognitive reserve may not be protective against the development of AD pathophysiology, it does seem that individuals with higher cognitive reserve present a more efficient network topology. Whereas healthy individuals with more years of education presented increased network efficiency, AD patients more educated might be better able to recruit compensatory mechanisms to maintain cognitive function.