Functional and microstructural measures of brain aging subgroups in cognitively unimpaired subjects
Poster No:
1181
Submission Type:
Abstract Submission
Authors:
Ioanna Skampardoni1, Junhao Wen2, Ilya Nasrallah1, Zhijian Yang1, Dhivya Srinivasan1, Guray Erus1, Elizabeth Mamourian1, Randa Melhem1, Haochang Shou1, Konstantina Nikita3, Christos Davatzikos1
Institutions:
1University of Pennsylvania, Philadelphia, PA, 2University of Southern California, LA, CA, 3National Technical University of Athens, Athens, Attiki
First Author:
Co-Author(s):
Introduction:
Brain aging is accompanied by several neuropathologies, often co-occurring, heterogeneously affecting brain structure and function. Unraveling the heterogeneity of complex neuroanatomical and functional changes at early asymptomatic stages may aid in revealing vulnerability or presence of neurodegeneration with potential biological and clinical implications. Here, we examine the functional connectivity (FC) and white matter (WM) integrity of three brain aging subgroups identified via a deep learning method applied to T1- and T2-weighted magnetic resonance imaging (MRI) data of a harmonized multi-cohort sample of 27,402 cognitively unimpaired individuals from the iSTAGING consortium (Habes et al. 2021).
Methods:
The three subgroups were separately modeled in four decade-spanning age brackets along the 45-85 years range using the Smile-GAN method (Yang et al. 2021) built on regional volumetrics and white matter hyperintensities (WMH). We investigated internetwork connectivity based on 21 FC networks extracted using group-independent component analysis (ICA) on resting-state functional MRI (rsfMRI) data of the UK Biobank study (Miller et al. 2016). Additionally, fractional anisotropy (FA) maps derived from diffusion tensor imaging (DTI) data from the UK Biobank (Miller et al. 2016) were used to measure WM microstructural integrity. The mean FA values were extracted within 48 WM tracts using the Johns Hopkins University tract atlas. We used linear regression to associate the Smile-GAN subgroups with the 210 internetwork FC and 48 FA features, adjusting for age, sex, and subgroup labels.
Results:
The three subgroups of brain volumetric measures displayed consistent patterns relative to the reference group A0 across the four age intervals: typical brain agers (A1) with mild atrophy and WMH load, and two accelerated aging subgroups, one with elevated WMH burden and vascular risk factors (VRF) enrichment but moderate atrophy (A2); and a second with diffuse severe atrophy, probably driven by lifestyle factors, and modest WMH load (A3) (Figure 1).
Given the subgroup consistency across the four age intervals, functional connectivity and fractional anisotropy were examined in the entire 45-85 years age range. Internetwork connectivity analysis (N=19,143; 47% males) revealed that A3 had the most significant differences relative to the reference A0 group. We observed increased connectivity for several pairs of networks, such as the default mode - motor, somatosensory - occipital visual, and dorsal attention - occipital visual networks, and decreased connectivity for other pairs, such as the default mode - frontotemporal, subcortical - frontotemporal, and occipital visual - fronto-insular-parietal networks (Figure 2A). Our results align with the literature showing both increased (Betzel et al. 2014; Grady et al. 2016) and decreased (Onoda, Ishihara, and Yamaguchi 2012; Huang et al. 2015) internetwork connectivity, uncovering a complex functional reorganization of the brain with aging.
Regarding the fractional anisotropy analysis (N=3,443; 48% males), consistent with the known associations of WMH and VRF (Power et al. 2017; Hannawi et al. 2018; Wassenaar et al. 2019) with the WM integrity, the A2 subgroup showed significant microstructural WM integrity disruption relative to A0 for 41 tracts, with the most prominent disruption observed in posterior thalamic radiation, corona radiata, superior fronto-occipital and longitudinal fasciculus, and anterior limb of the internal capsule (Figure 2B).
Given the subgroup consistency across the four age intervals, functional connectivity and fractional anisotropy were examined in the entire 45-85 years age range. Internetwork connectivity analysis (N=19,143; 47% males) revealed that A3 had the most significant differences relative to the reference A0 group. We observed increased connectivity for several pairs of networks, such as the default mode - motor, somatosensory - occipital visual, and dorsal attention - occipital visual networks, and decreased connectivity for other pairs, such as the default mode - frontotemporal, subcortical - frontotemporal, and occipital visual - fronto-insular-parietal networks (Figure 2A). Our results align with the literature showing both increased (Betzel et al. 2014; Grady et al. 2016) and decreased (Onoda, Ishihara, and Yamaguchi 2012; Huang et al. 2015) internetwork connectivity, uncovering a complex functional reorganization of the brain with aging.
Regarding the fractional anisotropy analysis (N=3,443; 48% males), consistent with the known associations of WMH and VRF (Power et al. 2017; Hannawi et al. 2018; Wassenaar et al. 2019) with the WM integrity, the A2 subgroup showed significant microstructural WM integrity disruption relative to A0 for 41 tracts, with the most prominent disruption observed in posterior thalamic radiation, corona radiata, superior fronto-occipital and longitudinal fasciculus, and anterior limb of the internal capsule (Figure 2B).
Conclusions:
The neuroanatomical heterogeneity of brain aging was modeled with subgroups designated by regional atrophy and WMH load in a multi-cohort cognitively unimpaired population. The subgroup characterized by elevated WMH burden and presence of VRF was associated with severely disrupted white matter integrity, while the subgroup with widespread atrophy underwent multiple changes in rsfMRI internetwork connectivity.
Lifespan Development:
Aging 1
Modeling and Analysis Methods:
Diffusion MRI Modeling and Analysis 2
fMRI Connectivity and Network Modeling
Keywords:
Aging
FUNCTIONAL MRI
Machine Learning
WHITE MATTER IMAGING - DTI, HARDI, DSI, ETC
Other - Heterogeneity
1|2Indicates the priority used for review
Provide references using author date format
C. Grady, S. Sarraf, C. Saverino, and K. Campbell, 2016. “Age Differences in the Functional Interactions among the Default, Frontoparietal Control, and Dorsal Attention Networks.” Neurobiology of Aging 41 (May): 159–72.
M. Habes et al., 2021. “The Brain Chart of Aging : Machine-Learning Analytics Reveals Links between Brain Aging , White Matter Disease , Amyloid Burden , and Cognition in the ISTAGING Consortium of 10,216 Harmonized MR Scans.” Alzheimer’s and Dementia: The Journal of the Alzheimer’s Association 17 (1): 89–102.
Y. Hannawi et al., 2018. “Hypertension Is Associated with White Matter Disruption in Apparently Healthy Middle-Aged Individuals.” American Journal of Neuroradiology 39 (12): 2243-2248.
C.-C. Huang et al., 2015. “Age-Related Changes in Resting-State Networks of a Large Sample Size of Healthy Elderly.” CNS Neuroscience & Therapeutics 21 (10): 817–25.
K. L. Miller et al., 2016. “Multimodal Population Brain Imaging in the UK Biobank Prospective Epidemiological Study.” Nature Neuroscience 19 (11): 1523–36.
K. Onoda, M. Ishihara, and S. Yamaguchi, 2012. “Decreased Functional Connectivity by Aging Is Associated with Cognitive Decline.” Journal of Cognitive Neuroscience 24 (11): 2186–98.
M. C. Power et al., 2017. “Midlife and Late-Life Vascular Risk Factors and White Matter Microstructural Integrity: The Atherosclerosis Risk in Communities Neurocognitive Study.” Journal of the American Heart Association 6 (5).
T. M. Wassenaar, K. Yaffe, Y. D. van der Werf, and C. E. Sexton, 2019. “Associations between Modifiable Risk Factors and White Matter of the Aging Brain: Insights from Diffusion Tensor Imaging Studies.” Neurobiology of Aging 80: 56–70.
Z. Yang et al., 2021. “A Deep Learning Framework Identifies Dimensional Representations of Alzheimer’s Disease from Brain Structure.” Nature Communications 12 (1): 7065.