Alzheimer´s disease, cognitive decline and prediction

Alzheimers disease is associated with cognitive decline (e.g. memory impairment) and and exhibits deleterious consequences for patients, caregivers, and the social- and health care system. Structural and functional changes of the brain, such as grey matter atrophy or functional connectivity changes, precede cognitive deterioration. Modern imaging techniques, such as (functional) magnetic resonance imaging ((f)MRI) enable in-vivo examination of the brains structure and function. These imaging techniques substantially contributed to better understanding of the brains functionality. While a one- to-one correspondence of cognitive functions and anatomical regions has been proposed earlier, it is now understood that distinct large scale neurocognitive networks are the basis for effective brain function. Therefore, examinations on network-level are required to gain insights in the brains functionality. In contrast to snapshots with a single measurement, repeated brain scans allow the observation of structural and functional changes over time and enable the association of these outcomes with the cognitive performance. The worldwide prevalence of Alzheimers disease is estimated to be 15 million and progresses differently in every individual. However, the quest for so-called biomarker, parameters with prognostic value for cognition, still is in its infancies. Therefore, the aim of this study is to predict the cognitive decline in Alzheimers disease and determine whether structural and functional network changes and cognitive deterioration undergo a similar degradation process. (F)MRI of Alzheimer patients shall be acquired at three time-points at an interval of one year, and biomarker of structural and functional brain scans will be extracted. Parallel, cognitive status of the patients will be assessed. The prognostic value of the biomarker for cognitive deterioration will be evaluated by means of regression models. The results from this study (1) contribute to a better understanding of cerebral changes on network level in the context of clinical and cognitive changes, and (2) strengthen the value of structural and functional imaging as a predictor of outcome variables in the disease course. The longitudinal aspect of the study allows a refined examination of decline within one subject and therefore constitutes a major extension of current prevalent cross-sectional studies.

In this study we aimed to characterize brain structural grey matter networks as biomarker of cognitive and motor function in healthy elderly and patients with Alzheimers disease (AD). We conducted two studies. In the first study, we aimed to (1) identify networks that lose grey matter integrity with advancing age, (2) investigate if age-related impairment of integrity in grey matter networks associates with cognitive function and decreasing fine motor skills and (3) examine if grey matter disintegration is a mediator between age and cognition and fine motor skills. Data of 257 participants of the Austrian Stroke Prevention Family Study (ASPSF) were examined. In this cross-sectional study we identified 20 grey matter networks, of which fourteen showed changes with increasing age. Next to age and education, eight networks showed an association with cognition and fine motor skills. Grey matter networks partially mediated the effect between age and cognition and age and fine motor skills. We confirm an age-related decline in cognitive functioning and fine motor skills in non-demented community-dwelling subjects and showed that aging selectively affects the integrity of grey matter networks. The negative effect of age on cognition and fine motor skills is associated with distinct grey matter networks and is partly mediated by their disintegration. In study two, we aimed to (1) identify grey matter networks that discriminate between AD patients and healthy controls, (2) investigate their diagnostic accuracy in comparison and above established markers, and (3) determine if they are associated to cognitive abilities in a longitudinal study. We used the twenty grey matter networks identified in the study above. 104 AD patients and 104 age-matched controls and an independent sample to validate the results, were examined. Two of the twenty grey matter networks contributed significantly to the discrimination between AD and controls. Their diagnostic accuracy was 74% in the original cohort and 80% in the independent sample. Moreover, we found one network to be associated with verbal memory. No other associations with cognitive functions were found. Grey matter networks failed to predict the longitudinal course of cognitive decline over an average of 18 months and did not exceed the predictive value of established biomarkers. We conclude that grey matter networks have diagnostic potential, but the diagnostic information beyond conventional MRI markers is marginal.