Functional architecture of the aging brain
The intrinsic network architecture of the brain is continuously shaped by biological and behavioral factors from younger to older adulthood. Differences in functional networks can reveal how a lifetime of learning and lived experience can alter large-scale neurophysiological dynamics, offering a powerful lens into brain and cognitive aging. Quantifying these differences has been hampered by significant methodological challenges. Here, we use multi-echo fMRI and multi-echo ICA processing, individualized cortical parcellation methods, and multivariate (gradient and edge-level) functional connectivity analyses to provide a definitive account of the intrinsic functional architecture of the brain in older adulthood. Twenty minutes of resting-state multi-echo fMRI data were collected in younger (n=181) and older (n=120) adults. Dimensionality, the number of independent, non-noise BOLD components in the fMRI signal, was significantly reduced for older adults. Macroscale functional gradients were largely preserved. In contrast, edge-level functional connectivity was significantly altered. Within-network connections were weaker while connections between networks were stronger for older adults, and this connectivity pattern was associated with lower executive control functioning. Greater integration of sensory and motor regions with transmodal association cortices also emerged as a prominent feature of the aging connectome. These findings implicate network dedifferentiation, reflected here as reduced dimensionality within the BOLD signal and altered edge-level connectivity, as a global and putatively maladaptive feature of functional brain aging. However, greater coherence among specific networks may also signal adaptive functional reorganization in later life. By overcoming persistent and pervasive methodological challenges that have confounded previous research, the results provide a comprehensive account of the intrinsic functional architecture of the aging brain.