Stimulus transformation into motor action: Dynamic graph analysis reveals a posterior‐to‐anterior shift in brain network communication of older subjects

• Published in: Human brain mapping Vol. 42; no. 5; pp. 1547 - 1563
• Main Authors: Rosjat, Nils, Wang, Bin A., Liu, Liqing, Fink, Gereon R., Daun, Silvia
• Format: Journal Article
• Language: English
• Published: Hoboken, USA John Wiley & Sons, Inc 01.04.2021
• Subjects: Adult; Age; Age Factors; Aged; Aging; Aging - physiology; Analysis; Brain; Brain architecture; Brain mapping; Cerebral Cortex - physiology; Cerebral hemispheres; Cognitive ability; Cognitive tasks; Connectome; Cortical Synchronization - physiology; Cues; Data processing; EEG; Electroencephalography; External stimuli; Female; Fingers - physiology; Functional magnetic resonance imaging; hemispheric asymmetry reduction in older adults (HAROLD); Hemispheric laterality; Humans; Information processing; Locking; Low frequencies; Male; Middle Aged; Motor Activity - physiology; motor execution; Motor task performance; Nerve Net - physiology; Network analysis; phase locking of EEG signals; posterior–anterior shift in aging (PASA); Sensorimotor system; Synchronism; Synchronization; Visual Perception - physiology; Young Adult; information processing; low frequencies; hemispheric asymmetry reduction in older adults (HAROLD); phase locking of EEG signals; posterior-anterior shift in aging (PASA); motor execution
• ISSN: 1065-9471; 1097-0193; 1097-0193
• DOI: 10.1002/hbm.25313

Cognitive performance slows down with increasing age. This includes cognitive processes that are essential for the performance of a motor act, such as the slowing down in response to an external stimulus. The objective of this study was to identify aging‐associated functional changes in the brain networks that are involved in the transformation of external stimuli into motor action. To investigate this topic, we employed dynamic graphs based on phase‐locking of Electroencephalography signals recorded from healthy younger and older subjects while performing a simple visually‐cued finger‐tapping task. The network analysis yielded specific age‐related network structures varying in time in the low frequencies (2–7 Hz), which are closely connected to stimulus processing, movement initiation and execution in both age groups. The networks in older subjects, however, contained several additional, particularly interhemispheric, connections and showed an overall increased coupling density. Cluster analyses revealed reduced variability of the subnetworks in older subjects, particularly during movement preparation. In younger subjects, occipital, parietal, sensorimotor and central regions were—temporally arranged in this order—heavily involved in hub nodes. Whereas in older subjects, a hub in frontal regions preceded the noticeably delayed occurrence of sensorimotor hubs, indicating different neural information processing in older subjects. All observed changes in brain network organization, which are based on neural synchronization in the low frequencies, provide a possible neural mechanism underlying previous fMRI data, which report an overactivation, especially in the prefrontal and pre‐motor areas, associated with a loss of hemispheric lateralization in older subjects. In this article, we performed an analysis of dynamic networks based on synchronization in the delta‐theta frequency range in the context of stimulus processing and motor execution in order to understand the neural mechanisms underlying HAROLD and PASA. We were able to demonstrate a similar structure as has been shown in self‐initiated movements indicating an independence of the underlying neural mechanisms on movement initiation: The networks in older subjects displayed an overall increased connectivity, especially in motor related electrodes by establishing significantly more interhemispheric connections which is consistent with the HAROLD model. In addition, we could show that the flexibility in the networks of older subjects decreased, indicating a kind of over‐connectivity and resulting difficulties in recruiting task specific regions. Finally, an analysis of the HUB nodes, that is, the neural information flow revealed an extended stimulus processing in occipital regions and a detour via frontal regions in older subjects, which is compatible with the PASA phenomenon.