Spatial Distribution and Hierarchical Clustering of β-Amyloid and Glucose Metabolism in Alzheimer’s Disease

• Published in: Frontiers in aging neuroscience Vol. 14; p. 788567
• Main Authors: Zhou, Da-An, Xu, Kai, Zhao, Xiaobin, Chen, Qian, Sang, Feng, Fan, Di, Su, Li, Zhang, Zhanjun, Ai, Lin, Chen, Yaojing
• Format: Journal Article
• Language: English
• Published: Lausanne Frontiers Research Foundation 06.06.2022; Frontiers Media S.A
• Subjects: Aging; Alzheimer's disease; Biomarkers; Brain; Cluster analysis; Clustering; Dementia; Glucose; Glucose metabolism; hierarchical organization; Metabolism; Neurodegeneration; Neurodegenerative diseases; Neuroscience; Positron emission tomography; Spatial distribution; Tomography; β-Amyloid; Beijing China; China
• ISSN: 1663-4365; 1663-4365
• DOI: 10.3389/fnagi.2022.788567

Increased amyloid burden and decreased glucose metabolism are important characteristics of Alzheimer’s disease (AD), but their spatial distribution and hierarchical clustering organization are still poorly understood. In this study, we explored the distribution and clustering organization of amyloid and glucose metabolism based on 18 F-florbetapir and 18 F-fluorodeoxyglucose PET data from 68 AD patients and 20 cognitively normal individuals. We found that: (i) cortical regions with highest florbetapir binding were the regions with high glucose metabolism; (ii) the percentage changes of amyloid deposition were greatest in the frontal and temporal areas, and the hypometabolism was greatest in the parietal and temporal areas; (iii) brain areas can be divided into three hierarchical clusters by amyloid and into five clusters by metabolism using a hierarchical clustering approach, indicating that adjacent regions are more likely to be grouped into one sub-network; and (iv) there was a significant positive correlation in any pair of amyloid-amyloid and metabolism-metabolism sub-networks, and a significant negative correlation in amyloid-metabolism sub-networks. This may suggest that the influence forms and brain regions of AD on different pathological markers may not be synchronous, but they are closely related.