Network-specific effects of age and in-scanner subject motion: a resting-state fMRI study of 238 healthy adults

Abstract

Cognitive aging is accompanied by a range of structural and functional differences in the brain, even in the absence of neurodegenerative disease. Functional magnetic resonance imaging (fMRI) studies have reported increased bilateral activation during task performance in elderly participants compared to their younger counterparts, particularly in frontal regions. Alterations have also been observed in the functional architecture of the resting brain, suggesting that aging is associated with changes in the organization of the networks of the brain. However, previous studies have largely focused on the default mode network, and little is known about the effects of age on other resting state-networks (RSNs). The aim of the present study was to investigate age-differences in resting state functional connectivity (RSFC) using fMRI data obtained during rest from 238 healthy participants aged 21-80 years. Using independent component analysis (ICA) and dual-regression, the results revealed age-related increases in RSFC across a range of RSNs, including task-positive networks in frontal and parietal regions. In contrast, age-related reductions in the default mode network and occipital visual networks were observed. Furthermore, whereas the effects of age on the various RSNs were found independent of age-related decreases in gray matter volume, sex and subject motion, we report strong positive and widespread effects of estimated subject motion on the RSFC across RSNs. The results provide support for the notion of network-specific effects in aging, manifested as increased tonic activation of task-positive networks, supporting higher-order cognitive functions and cognitive control, along with reduced task-negative default mode network and sensory visual networks during rest. The present results also corroborate recent evidence of strong influence of subject motion on estimated functional connectivity measures and strongly suggest that studies using RSFC measures as imaging phenotypes should adjust for individual differences in in-scanner subject motion.