Diurnal dynamics of multilayer brain networks predict cognitive trajectories in aging

Resting-state functional connectivity (rsFC) is a highly dynamic process that varies across different times of the day within each individual. Although this variability was long considered to be noise, recent evidence suggests it may allow for an optimal adaptation to changes in the environment. However, the way rsFC is shaped on a circadian scale and its association with cognition are still unclear. We analyzed data from 90 late middle-aged participants from the Cognitive Fitness in Aging study (61 women; 50-69 years). Participants completed five electroencephalographic (EEG) recordings of spontaneous resting-state activity spread over 20 h of prolonged wakefulness. Using a temporal multilayer network approach, we characterized the diurnal variations of the dynamic recruitment and integration of resting-state brain networks. We focused on the theta and gamma frequency bands within the default mode network (DMN), central executive network (CEN), and salience network (SN). Additionally, we investigated the relationship between the recruitment and integration of these networks with baseline cognitive performance and at a 7-year longitudinal follow-up, as well as with positron emission tomography (PET) early neuropathological markers of Alzheimer's disease such as β-amyloid and tau/neuroinflammation. Diurnal changes in theta and gamma dynamics were associated with distinct cognitive aspects. Specifically, higher baseline memory performance was associated with higher theta dynamic integration of the SN and the CEN, as well as higher theta dynamic recruitment of the DMN. Moreover, lower longitudinal memory decline at 7 years was associated with higher theta dynamic integration of the SN, CEN, and DMN. In contrast, higher gamma diurnal dynamic integration of the SN and the CEN was associated with lower executive and attentional performance, as well as higher early β-amyloid accumulation, at baseline. These findings suggest that maintaining a balance between network flexibility and stability throughout the diurnal phase of the circadian cycle may play a crucial role in cognitive aging, with stable theta-band connectivity supporting memory, whereas excessive gamma-band stability in the SN and CEN may contribute to executive decline and early amyloid accumulation. These insights highlight the importance of considering time-of-day in brain rsFC studies, calling for a temporal multilayer approach to capture these dynamic patterns more effectively.