The virtual aging brain: a model-driven explanation for cognitive decline in older subjects

Abstract

Healthy aging is accompanied by heterogeneous decline of cognitive abilities among individuals, especially during senescence. The mechanisms of this variability are not understood, but have been associated with the reorganization of white matter fiber tracts and the functional co-activations of brain regions. Here, we built a causal inference framework to provide mechanistic insight into the link between structural connectivity and brain function, informed by brain imaging data and network modeling. By applying various degrees of interhemispheric degradation of structural connectivity, we were not only able to reproduce the age-related decline in interhemispheric functional communication and the associated dynamical flexibility, but we obtained an increase of global modulation of structural connectivity over the brain function during senescence. Notably, the increase in modulation between structural connectivity and brian function was higher in magnitude and steeper in its increase in older adults with poor cognitive performance. We independently validated the causal hypothesis of our framework via a Bayesian approach based on deep-learning. The current results might be the first mechanistic demonstration of dedifferentiation and scaffolding during aging leading to cognitive decline demonstrated in a large cohort.

Personalized virtual brains of Alzheimer’s Disease link dynamical biomarkers of fMRI with increased local excitability

“…The temporal evolution of the brain FC is represented by sliding window technique and it is quantified via switching index (for further detail Material and Methods section). The index related with brain fluidity (Figure 1E) (Hansen et al, 2015; Melozzi et al, 2019; Courtiol et al, 2020) quantifies the switching behaviour of resting state networks (Rabuffo et al, 2021; Lavanga et al, 2022). Empirical resting-state fMRI analysis showed that patients with cognitive decline had a significant reduction in both homotopic dysconnectivity (Figure 2B), and brain fluidity as measured by FCD variance of the limbic network (Figure 2C) (Pillai’s Trace = 0.86, F (4, 218) = 40.97, p < 0.001).…”

“…Specifically, reduced inter-hemispheric homotopic FC (Figure 1D) is also associated with AD (Wang et al, 2015, Qiu et al, 2016). Functional connectivity dynamics (FCD) (see Figure 1.E) refers to the flexibility of the brain to flow between different states to facilitate cognition and has been shown to be altered in healthy lifespan (Battaglia et al, 2020, Petkoski et al, 2022), normal mild-life and older adults (Lavanga et al, 2022), and AD (Córdova-Palomera et al, 2017, Jie et al, 2018). These changes in FCD may provide insight into the neural mechanisms underlying cognitive decline in AD.…”

Interrelating differences in structural and functional connectivity in the older adult's brain

“…This is thought to reduce the capacity to also behaviorally switch between concepts and to slow the rate of functional adaptations to external influences (Escrichs et al, 2021 ; Lee et al, 2019 ; Xia et al, 2019 ). Computational models showed that reduced metastability is a response to SC decline (Deco & Kringelbach, 2016 ; Lavanga et al, 2022 ; Naik et al, 2017 ). Hence, the functionally highly interconnected system found in the second component could point towards a low capacity to switch between functional states potentially resulting from the incipient SC decline and would explain the strongest association with cognitive performance decline.…”

“…Higher overall communication in the brain, that is, connectivity, requires higher energy consumption (Tomasi et al, 2013 ). At the same time, a highly interconnected functional system reduces the ability to efficiently switch between brain states (Chan et al, 2014 , 2017 ; Colcombe et al, 2005 ; Goh, 2011 ; Nashiro et al, 2017 ; Park et al, 2004 ), which is associated with lower cognitive performance (Lavanga et al, 2022 ). Accordingly, the functional connectivity system of this third component could reflect a rather low‐energy state of the resting brain, which at the same time may involve a high ability to efficiently adapt to external stimuli.…”

“…Computational models showed that reduced metastability is a response to SC decline (Deco & Kringelbach, 2016;Lavanga et al, 2022;Naik et al, 2017). Hence, the functionally highly interconnected system found in the second component could point towards a low capacity to switch between functional states potentially resulting from the incipient SC decline and would explain the strongest association with cognitive performance decline.…”