Investigation of metabolic brain network changes due to normal aging with ¹⁸F-FDG PET/MR

Introduction: Normal aging has multiple effects on brain including morphologic, functional and metabolic changes. Recently, principal components analysis (PCA) has been used to reveal the specific metabolic brain patterns of Parkinson’s disease (PD) and Alzheimer’s disease (AD) [1]. However, how normal aging affects brain metabolism and whether it also has specific patterns is yet to be answered. The objective of this study was to investigate the brain metabolic pattern for normal aging using 18F-fluorodeoxyglucose (FDG) combined positron emission tomography and magnetic resonance imaging (PET/MRI).

Methods: Resting-state brain FDG PET imaging was performed in a cohort of 200 gender-matched healthy subjects across a wide age range from 20 to 70 years old. All subjects had given their informed consent and were screened clinically to exclude organic brain disease. SVD-based PCA was used to identify the aging-related metabolic brain patterns for the subjects [2]. The expression scores of principal components (PCs) were evaluated on each subject and its correlation with age was also calculated. In addition, all the subjects were divided into 5 age-groups (20-29, 30-39, 40-49, 50-59 and ≥60), the average expression scores were assessed for each age-group.

Results: The first two PCs accounted for 61.2% of the variance, with PC1 and PC2 accounting for 43.7% and 17.4%, respectively, and were selected for further analysis. Other components had low variance and were not physiologically meaningful. PC1 was characterized by a relative decrease in the calcarine, cingulum, postcentral cortex, angular cortex, putamen, temporal cortex, insula, precuneus, and fusiform, which was topographically similar to the default mode network (DMN) as revealed by fMRI [3]. The PC1 expression scores did not show correlation with age (r = 0.135, P = 0.0912). The PC2 had the highest positive loadings in the primary somatosensory cortex, motor cortex, temporal gyrus, retrosubicular area, post cingulum, and some white matter areas. PC2’s expression scores had a significantly negative correlation with age (r = -0.3043, P < 0.001). For the group analysis of PC1, the expression scores (-3.50, -5.62, -1.29, -4.50 and 15.52) were relatively stable and only increased in the group of ≥60. The sharp increment may result from local ischemia which is very common for people over 60 years old. PC2’s group-scores (7.75, 6.10, -0.79, -4.97 and -6.60) were continuously decreasing with age.

Conclusions: In conclusion, PCA based on PET/MR data is an efficient way to obtain aging-related brain metabolic patterns. According to our results, normal aging has two main effects on brain metabolism. The first effect is related to the dysfunction of DMN and follows microvascular pathology in the elder population. The second one is continuously reduced with aging which can be regarded as the aging-related metabolic brain pattern. [1] Spetsieris, Phoebe G., et al. "Metabolic resting-state brain networks in health and disease." Proceedings of the National Academy of Sciences 112.8 (2015): 2563-2568. [2] Yuan, Li-Xia, et al. "Intra-and inter-scanner reliability of scaled subprofile model of principal component analysis on ALFF in resting-state fMRI under eyes open and closed conditions." Frontiers in neuroscience 12 (2018): 311. [3] Raichle, Marcus E. "The restless brain." Brain connectivity 1.1 (2011): 3-12. Figure 1. Metabolic resting-state networks (a and b) identified in healthy subjects across a wide age range from 20 to 70 years old.

• Download figure
• Open in new tab
• Download powerpoint