Functional connectivity of key resting state networks and objectively measured physical activity in older adults with...

Greater brain network integrity may associate with physically active lifestyles. Three resting state networks may provide unique insights into known physical activity-mediated brain health benefits: the default mode network (involved with self-monitoring), the salience network (involved in orienting oneself to salient external and internal stimuli), and the central executive network (responsible for higher level cognitive task). The current study explored relationships between system-wide neural network integrity measured by functional magnetic resonance imaging and objectively-measured physical activity. We hypothesize connectivity patterns as measured by fMRI networks will relate to actigraphy markers such that 1) there will be higher connectivity within the central executive network in more physically active individuals, and 2) there will be higher connectivity within the default mode network and salience network in those with higher levels of physical activity.

Eighteen non-demented older adults with orthopedic pain (age 67.11 ± 5.61, 50% female, education 15.94 ± 2.51 years) completed brain magnetic resonance imaging, and wore an actigraphy device to objectively measure types of physical and sedentary engagement.

Results showed a negative relationship between central executive network connectivity and sedentary time (β = -0.108, p = .039), and a positive relationship with both moderate-to-vigorous physical activity (β = 0.629, p = .029) and total activity time (β = 0.645, p = .039). Results also showed positive relationships for the default mode network (β = 0.588, p = .033) and the salience network (β = 0.608, p = .037) with mean cadence (i.e. steps per minute).

Our work adds to the existing literature on specific types of activity measurement (i.e. sedentary time, cadence and moderate-to-vigorous physical activity) which will be useful for interventions aimed at improving the integrity of underlying neural networks.