Changes in Centrality Frequency of the Default Mode Network in Individuals With Subjective Cognitive Decline

• Main Authors: Yunyan Xie, Tiantian Liu, Jing Ai, Duanduan Chen, Yiran Zhuo, Guanglei Zhao, Shuai He, Jinglong Wu, Ying Han, Tianyi Yan
• Format: Article
• Language: English
• Published: Frontiers Media S.A. 2019-06-01
• Series: Frontiers in Aging Neuroscience
• Subjects: subjective cognitive decline; centrality frequency; resting-state functional magnetic resonance imaging; hub probability; default mode network
• Online Access: https://www.frontiersin.org/article/10.3389/fnagi.2019.00118/full

Despite subjective cognitive decline (SCD), a preclinical stage of Alzheimer's disease (AD), being widely studied in recent years, studies on centrality frequency in individuals with SCD are lacking. This study aimed to investigate the differences in centrality frequency between individuals with SCD and normal controls (NCs). Forty individuals with SCD and 53 well-matched NCs underwent a resting-state functional magnetic resonance imaging scan. We assessed individual dynamic functional connectivity using sliding window correlations. In each time window, brain regions with a high degree centrality were defined as hubs. Across the entire time window, the proportion of time that the hub appeared was characterized as centrality frequency. The centrality frequency correlated with cognitive performance differently in individuals with SCD and NCs. Our results revealed that in individuals with SCD, compared with NCs, correlations between centrality frequency of the anterior cortical regions and cognitive performance decreased (79.2% for NCs and 43.5% for individuals with SCD). In contrast, correlations between centrality frequency of the posterior cortical regions and cognitive performance increased in SCD individuals compared with NCs (20.8% for NCs and 56.5% for individuals with SCD). Moreover, the changes mainly focused on the anterior (93.3% for NCs and 45.5% for individuals with SCD) and posterior (6.7% for NCs and 54.5% for individuals with SCD) regions associated with the default mode network (DMN). In addition, we used absolute thresholds (correlation efficient r = 0.2, 0.25) and proportional thresholds (sparsity = 0.2, 0.25) to verify the results. Dynamic results are relative stable at absolute thresholds while static results are relative stable at proportional thresholds. Converging findings provide a new framework for the detection of the changes occurring in individuals with SCD via centrality frequency of the DMN.