Dynamic Functional Network Connectivity Pattern of the Amygdalohippocampal Complex in Individuals With Subjective Cognitive Decline

Subjective cognitive decline (SCD) is a potential early marker of cognitive decline and dementia. The amygdalohippocampal structure and function are closely related to cognitive decline, but few studies have investigated large-scale amygdalohippocampal brain functional network connectivity in individuals with SCD. Here, we aim to explore how the dynamic functional network connectivity (dFNC) between the amygdalohippocampal complex and other brain networks contributes to the understanding of early cognitive decline. Independent component analysis (ICA) and dFNC analysis were applied to functional magnetic resonance imaging (fMRI) data from 66 individuals with SCD to extract the amygdalohippocampal complex and identify distinct connectivity states. Cognitive performance was assessed through a composite Z score derived from a battery of neuropsychological tests. Correlation analyses were performed to examine the associations between the dFNC patterns and cognitive performance. Three distinct dFNC states were identified, each characterized by varying levels of within- and inter-network connectivity, with occurrences of 65%, 18%, and 17% respectively. Cognitive function, measured using a composite Z score, was positively correlated with amygdalohippocampal-sensorimotor network (SM) and amygdalohippocampal-visual network (VI) dFNC in State 2. Significant correlations were observed between the amygdalohippocampal complex and the left precentral gyrus (r = 0.517, FDR-corrected p = 0.005), postcentral gyrus (r = 0.487, FDR-corrected p = 0.034), and multiple visual network regions, including the lingual gyrus and lateral occipital cortex (all Ps < 0.05, FDR-corrected). These associations remained significant after adjusting for sex and age. These findings extend the current understanding of amygdalohippocampal dysfunction in cognitive decline and demonstrate that cognitive function is associated with distinct large-scale amygdalohippocampal network dynamics.