Different glucose metabolic brain networks between Subjective Cognitive Decline and Health Control based on graph theory

Abstract

Recently, more evidences manifest that the subjective cognitive decline (SCD) of unimpaired individual may represent first symptom of Alzheimer's disease (AD). This study investigated the differences of intrinsic glucose metabolic functional connectivity between SCD and healthy subject (HC) groups from the perspective of brain network topology. In this study we attained 18F-Fluorodeoxyglucose positron emission tomography (18F-FDG PET) scans from Xuanwu Hospital, Beijing, China, including 85 SCD subjects (male = 16, mean age = 66, MMSE = 28.4) and 74 HC subjects (male = 37, mean age = 65,MMSE=29.0). Graph theory method has been used in this study. Network parameters, including global efficiency, local efficiency, characteristic path length, clustering coefficient, betweenness centrality, sigma and modularity were calculated and compared between two groups. As a result, both SCD and HC groups showed the small-world property. Meanwhile, SCD showed loss of small-world properties, for example, sigma in SCD was significantly lower than HC (p<0.05). In addition, the clustering coefficient and local efficiency of SCD were both higher than HC significantly (p<0.05). In contrast, the characteristic path length and global efficiency of SCD were lower than HC, which led to the regularization of brain network in SCD group. Furthermore, we found global modularity of SCD was lower than HC and the number of modules also decreased. Our findings suggested that there exist differences in glucose metabolic brain network between two groups, demonstrating that the graph theory analysis method could be useful and helpful to predict risks in the preclinical stage of AD.