Age-Related Differences of Frequency-Dependent Functional Connectivity in Brain Networks and Their Link to Motor Performance

Abstract

Background:

Aging affects the brain at the anatomical and functional levels, resulting in a decline in motor and cognitive performance. Functional magnetic resonance imaging (fMRI) studies documented lower connectivity within brain networks and higher connectivity between them, for older as compared with young adults. However, it is still unclear whether the reduced segregation between networks, as observed with fMRI, has neurophysiological underpinnings.

Methods:

We collected high-density electroencephalography (hdEEG) data in 24 young and 24 older adults at rest. Bimanual coordination performance was also measured in the same participants, using a computerized test. Using the hdEEG data, we reconstructed oscillatory power and functional connectivity for six large-scale brain networks, in delta, theta, alpha, beta and gamma frequency bands. We evaluated age-related differences in network power and connectivity between young and older participants, and their possible relationships with bimanual coordination performance.

Results:

We observed that the level of network segregation generally decreased with age, in line with fMRI findings. However, there was a relatively strong dependence on the frequency band and the brain network being considered. EEG connectivity in the sensorimotor network predicted motor performance differences across older individuals, particularly when neural oscillations in the beta frequency band were considered.

Discussion:

Our study provides electrophysiological evidence in support of the “de-differentiation hypothesis” for the aging brain, and for the existence of a clear link between the strength of EEG connectivity at rest and motor performance.

Impact statement

Using advanced analyses of high-density electroencephalographic data, we could show how band-limited network connectivity differs between young and older adults, and how these differences can predict motor performance. The analyses revealed a complex pattern of connectivity–behavior associations, which not only depended on the network considered, but also on the specific neural oscillations. This may provide novel targets for age-related interventions inducing neural entrainment, based for instance on transcranial alternating current stimulation.