Multimodal imaging : functional, structural, and molecular brain correlates of cognitive aging

Abstract

Aging is associated with a decline in many (but not all) cognitive abilities. Although it remains largely unknown how changes in brain integrity relate to cognitive deficits, these changes are likely expressed across interrelated functional, structural, and molecular layers. This complexity calls for a multimodal imaging approach in age-related mind-brain research. Hence, in this thesis, different imaging modalities were combined in order to study the neural basis of cognitive aging.

Study I investigated functional connectivity patterns among three large-scale functional brain networks (i.e., default mode [DMN], frontoparietal control [FPN], and dorsal attention [DAN] networks) during rest and task in younger and older adults. The FPN was flexible in its affiliation to other networks, given that it was more functionally connected to the DMN during rest and to the DAN during task performance. Age-related differences were stable across states for the FPN, but were only present for connectivity between the DMN and DAN during the task. Taken together, these results suggest that resting-state is not sufficient to uncover the entire functional connectome of the human brain. Study II identified brain iron as a potential source of age-related differences in connectivity. Greater striatal iron content was associated with lower intrinsic functional connectivity of the caudate and putamen. Additionally, more iron was associated with less connectivity between the putamen and the rest of the brain. Functional connectivity within the putamen was also linked to motor ability, indicating that iron-related connectivity features are behaviorally meaningful. Study III explored the relationship between functional and structural connectivity, and showed that increased homotopic functional connectivity in the prefrontal cortex was associated with worse microstructural degeneration of the corpus callosum, and exacerbated working memory decline. However, given that the association between function and structure was weak, results also suggest that homotopic functional connectivity can be resilient to change in the integrity of its structural paths. Study IV found that dopamine and iron in the putamen were positively associated, but only up until middle age. Together with the fact that dopamine requires iron for its synthesis, these results indicate that, for individuals without excessive iron accumulation, more iron is associated with higher dopaminergic activity. Higher iron load in the putamen was also linked to better processing speed for those in middle age.

Collectively, the studies show that functional connectivity is influenced by mental state, white-matter changes, and molecular properties, with the latter also being interrelated among themselves. These different features are associated with performance and interact with each other, suggesting that cognitive decline is linked to a multitude of changes in brain integrity, and that age-related alterations in the human brain are complex and multifaceted.