Pathological Structural Changes in the Default Mode Network as Predictors of Cognitive Decline in Relation to Amyloid and Tau Deposits

According to the proposed ATN model of pathology, Alzheimer’s Disease (AD) can be classified based on three main pathological processes: Amyloid, (A), Tau (T) and Neurodegeneration (N) status. Notably, the Default Mode Network (DMN) appears particularly vulnerable to those processes, making the study of its early alterations of interest in trying to define prodromal biomarkers of pathology. With this aim, we analyzed the data of 280 participants included in the Alzheimer Disease Neuroimaging Initiative (ADNI). Structural measures of sulcal depth, gyrification and cortical thickness were extracted for all DMN nodes and tested for their predictive power regarding encoding, retrieval and recall performance at both baseline and at 2-year follow-up. Mini Mental State Examination (MMSE) scores, amyloid status and regional tau burden were considered as covariates in the model, to test for their interaction with the structural measures. Our results appear in line with the ATN model, as distinct patterns are observed as a function of tau burden and amyloid status. Furthermore, while we observe that preserved structural integrity and complexity are favourably linked to behaviour and also act as protective factors against further deterioration, we also report DMN regions exhibiting inverted patterns. To the best of our knowledge, this is also the first study to prove a tight link between baseline structural alterations within the DMN, future memory decline and tau accumulation, looking at a variety of structural and memory measures. By discussing the link between structural integrity and memory performance, we hope to shed some light on the heterogeneity of findings in the literature.