Brain functional connectivity in regions that exhibit age-related cortical thinning

• Ano de defesa: 2018
• Autor(a) principal: Vieira, Bruno Hebling
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Aging; Atrofia; Atrophy; Brain; Cérebro; Conectividade funcional; Envelhecimento; Functional connectivity; MRI
• Link de acesso: http://www.teses.usp.br/teses/disponiveis/59/59135/tde-17042018-130342/

The brain ages, and with it come alterations in its micro- and macro-structure which reflect in its morphology and functioning. Changes in the brain structure and functional coupling between regions can be assessed with neuroimaging, and, more specifically, magnetic resonance imaging (MRI). Using MRI data from two stages (Pilot and Enhanced) of the Nathan Kline Institute Rockland Sample (NKI-RS), totalling 613, free of neurodegenerative diseases, and right-handed, participants aged 18 to 85 years old, we measured gray-matter parameters such as cortical volume, cortical thickness, and cortical surface area, and also volume of subcortical structures. We also measured cortico-cortical functional connectivity, defined either as the Pearson correlation coefficient and partial correlation coefficient, bivariate instantaneous Granger causality and Granger causality, and generalized partial directed coherence (GPDC). GPDC was evaluated in five frequencies between the four pairs of regions displaying the strongest evidence for linear thinning, measured by their associated t-statistic, and its alterations alongside aging were assessed using a multivariate approach based on Dirichlet Regression. We also studied spatial associations between patterns of morphometric and connectivity alterations. We reproduced generalized age-related atrophy reported in the literature in cortical volume (90% of the studied structures), surface area (68%) and thickness (90%), and volumetric atrophy of several subocortical structures. We observe a positive association in the joint distribution of the expected cortical thickness at 18 years old and the yearly percentage reduction in cortical thickness. We showed, projecting these two quantitities into their principal axes and analyzing the spatial distribution of the scores, that the first principal component correlates with neocortical granularity while the second principal component represents cortical type admixture. On functional connectivity, we gathered evidence for overall increased Pearson correlation coefficient (6% of the connections in the Pilot NKI-RS and 2% in the Enhanced NKI-RS), with proportionally smaller number of decreases (0.1% in the Pilot NKI-RS and 0.3% in the Enhanced NKI-RS). The Pearson partial correlation coefficient between 12 out of 65 homotopic region pairs shows a pattern of decline with age, suggesting inter-hemispheric disconnection. However, predictive causality, as measured by both Granger causalities, do not share the same degree of changes observed in the correlational metrics. We observe increased GPDC from several regions to themselves in many frequencies (25% out of a total of 40 self-connections), indicating a degree of disconnection to the other regions. Given seed regions, we uncovered spatially distributed significant patterns of association between the standardized effect of age on the connectivity to its targets and on their targets thicknesses. Regions with smaller evidence for age-related thinning, such as several occipital areas, tend to have fewer alterations in functional connectivity than regions with greater evidence for age-related thinning, like many frontal regions. We hypothesize that regions showing a negative association (5% of the seed regions) are part of compensatory systems, being increasingly correlated with regions displaying most atrophy. Regions showing a positive association (5%) do not have compensatory mechanisms available, and therefore are losing connectivity to atrophyc regions. Overall, we found evidence for brainwide alterations in connectivity and cortical and subcortical morphometry throughout the human adult lifespan. We also found a specifc pattern of associations between the atrophic trends and age-related alterations in connectivity in the brain