The impact of genetic factors and cognitive reserve on brain network organizations of healthy elderly adults

Resting-state functional magnetic resonance imaging (fMRI) has the potential to identify abnormalities in brain networks. Neuroimaging techniques have exhibited that structural and functional connectivity are impaired in the course of Alzheimer's disease. However, this condition happens on the background of aging, and it is less clear what may be the important functional changes that occur during aging that are associated with aging-related risk of conditions such as Alzheimer's disease. Therefore, it might be possible that genes play an important role in functional brain connectivity. Studies of older healthy individuals may provide a better understanding of how brain connectivity may be affected. In particular, healthy older individuals carrying specific genotypes related to dementia risk or activity-dependent plasticity may be variably associated with aging-related alterations or disruption to connectivity. This work explored the common genetic polymorphisms of APOE and BDNF Val66Met, and the lifecourse factor, cognitive reserve, in relation to age-related changes in functional and structural connectivity brain networks. Investigations in blood oxygenation level dependent (BOLD) signals found in resting-state data, maps of brain activity can be created to represent and assess structural and functional connections between different brain regions. Resting-state networks such as the default mode network (DMN), dorsal attention network (DAN), and Salience Network (SA), may provide possible explanations for differences between healthy aging and neurodegenerative diseases. Seventy-eight healthy older adults (average age 63.3) from the Tasmanian Healthy Brain Project, a longitudinal study investigating whether education later in life increases cognitive reserve and/or provides resilience to cognitive decline, were invited to participate in this project. FMRIB Software Library (FSL) was used to pre-process and clean the datasets. Functional datasets were analysed using whole-brain and seed-based analyses via FSL applying Multivariate Exploratory Linear Optimised Decomposition of Independent Components (MELODIC) and dual regression. Structural images were analysed using Statistical Parametric Mapping (SPM12) and Matlab to examine the covariance patterns of grey and white matter volume. Additionally, Freesurfer software was used to investigate region-specific alterations in grey matter volume and cortical thickness. The first aim was to investigate functional edge strength using graph theoretical measures within the DMN, DAN, and SA. Findings showed no significant differences in functional edge strength related to polymorphisms in APOE and BDNF within all three networks. Cognitive reserve was not significantly associated with edge strength. The second aim was to examine functional organization related to variations of the APOE and BDNF Val66Met polymorphisms in the DMN, DAN, and SA. Genetic interactions and the effects of cognitive reserve were also assessed. APOE ˜í¬µ3 homozygotes showed stronger functional connectivity than APOE ˜í¬µ4 carriers. No significant differences were found in functional connectivity in DMN, DAN, and SA between BDNF Met carriers and BDNF Val homozygotes. We observed a significant interaction, measuring stronger functional connectivity in Val/˜í¬µ3 homozygotes than in Met/˜í¬µ4 between the DMN and grey matter primary auditory cortex, as well as stronger connectivity in Met/˜í¬µ3 carriers than in Met/˜í¬µ4 carriers and Val/˜í¬µ4 carriers respectively, between the dorsal attention networks and default mode regions, as well as between the dorsal ventral stream and visual cortex. This analysis also revealed positive functional correlations with cognitive reserve within the DAN, and negative functional correlations with cognitive reserve within the DMN. Previous colleagues examining the THBP cohort found that increasing cognitive activities, such as later life education at university, enhanced cognitive reserve (Lenehan et al., 2016). Furthermore, this improvement within the THBP was then examined in relation to different cognitive domains, such as episodic memory, working memory, executive function, and language processing (Thow et al., 2018). Results showed that later life education enhanced language processing within the THBP cohort. Therefore, the third aim was to examine functional organization of the language network (Broca's and Wernicke's areas) using seed-based analyses within variations of the APOE and BDNF Val66Met polymorphisms and genetic interactions. The key findings were that Met/˜í¬µ3 showed stronger functional connectivity than Met/˜í¬µ4 and Val/˜í¬µ3, respectively, between the left Broca's area and right parahippocampal gyrus and left parahippocampal gyrus. Further, it was found that Val/˜í¬µ3 had stronger functional connectivity between left Broca's area and clusters of the poscentral gyrus and precuneus. Another interesting finding was that age correlated positively with functional connectivity within left Broca's area, while cognitive reserve correlated negatively with functional connectivity between left Broca's area and the left hippocampal regions. One of the most important findings to emerge from this aim was a positive correlation between age and functional connectivity within the language network, suggesting that healthy older adults are recruiting frontal lobe regions to support their language processing through neural compensation and maintaining the language function. The objective of the fourth aim was to examine differences between APOE, BDNF, and cognitive reserve relative to covariance patterns in grey and white matter. Additionally, specific brain structures, such as grey matter volume of the hippocampus and amygdala, and cortical thickness of the entorhinal cortex and parahippocampal cortex, were investigated in relation to APOE and BDNF genotypes. Further, correlations between cognitive function and whole brain structure metrics were examined. Covariance within whole brain grey matter structures were not significantly different with respect to polymorphisms. Grey matter of the hippocampal and amygdala volume, as well as cortical thickness of the parahippcampus and entorhinal cortex, showed no significant differences relative to both, APOE and BDNF genotypes. Age was significantly associated with hippocampal and amygdala volume, indicating a decrease of volume with increasing age. Episodic memory and language function did not show significant associations with grey matter volume and cortical thickness. Working memory and executive function were associated with volume and thickness of some brain regions. Overall, the results of this thesis demonstrated that the BDNF Val66Met polymorphism did not markedly affect the brain organization of healthy older adults. APOE status, however, may have an effect on functional connectivity and may decrease brain connectivity when the individual is carrying an ˜í¬µ4 allele. The findings of this project suggest that genetic interactions and the aging process influence functional connectivity in the absences of clinical signs of disease. Structurally, the brain is losing volume with increasing age, which predicted cognitive function status in healthy older participants.