Changes in whole-brain functional networks and memory performance in aging

• Published in: Neurobiology of aging 2014-10, Vol.35 (10), p.2193-2202
• Main Authors: Sala-Llonch, Roser, Junqué, Carme, Arenaza-Urquijo, Eider M, Vidal-Piñeiro, Dídac, Valls-Pedret, Cinta, Palacios, Eva M, Domènech, Sara, Salvà, Antoni, Bargalló, Nuria, Bartrés-Faz, David
• Format: Article
• Language: English
• Subjects: Aged; Aging; Aging - pathology; Aging - physiology; Aging - psychology; Brain - pathology; Brain - physiology; Female; Frontal lobe; Graph theory; Gray Matter - pathology; Humans; Internal Medicine; Magnetic Resonance Imaging; Male; Memory; Middle Aged; Nerve Net - pathology; Nerve Net - physiology; Neurology; Neuropsychological Tests; Organ Size; Rest - physiology; Resting-state fMRI; Synaptic Transmission - physiology
• ISSN: 0197-4580; 1558-1497
• DOI: 10.1016/j.neurobiolaging.2014.04.007

Abstract We used resting-functional magnetic resonance imaging data from 98 healthy older adults to analyze how local and global measures of functional brain connectivity are affected by age, and whether they are related to differences in memory performance. Whole-brain networks were created individually by parcellating the brain into 90 cerebral regions and obtaining pairwise connectivity. First, we studied age-associations in interregional connectivity and their relationship with the length of the connections. Aging was associated with less connectivity in the long-range connections of fronto-parietal and fronto-occipital systems and with higher connectivity of the short-range connections within frontal, parietal, and occipital lobes. We also used the graph theory to measure functional integration and segregation. The pattern of the overall age-related correlations presented positive correlations of average minimum path length (r = 0.380, p  = 0.008) and of global clustering coefficients (r = 0.454, p < 0.001), leading to less integrated and more segregated global networks. Main correlations in clustering coefficients were located in the frontal and parietal lobes. Higher clustering coefficients of some areas were related to lower performance in verbal and visual memory functions. In conclusion, we found that older participants showed lower connectivity of long-range connections together with higher functional segregation of these same connections, which appeared to indicate a more local clustering of information processing. Higher local clustering in older participants was negatively related to memory performance.