Heterogeneous Network-based Brain Ageing in Alzheimer's Disease Continuum
Poster No:
254
Submission Type:
Late-Breaking Abstract Submission
Authors:
Yanxi Huo1, Weijie Huang1,2, Zhenzhao Liu1, Tianyu Bai1, Yichen Wang1, Kexin Wang1, Ni Shu1
Institutions:
1State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University, Beijing, China, 2College of Artificial Intelligence, Nanjing University of Aeronautics and Astronautics, Nanjing, China
First Author:
Yanxi Huo
State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University
Beijing, China
State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University
Beijing, China
Co-Author(s):
Weijie Huang
State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University|College of Artificial Intelligence, Nanjing University of Aeronautics and Astronautics
Beijing, China|Nanjing, China
State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University|College of Artificial Intelligence, Nanjing University of Aeronautics and Astronautics
Beijing, China|Nanjing, China
Zhenzhao Liu
State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University
Beijing, China
State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University
Beijing, China
Tianyu Bai
State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University
Beijing, China
State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University
Beijing, China
Yichen Wang
State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University
Beijing, China
State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University
Beijing, China
Kexin Wang
State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University
Beijing, China
State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University
Beijing, China
Ni Shu
State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University
Beijing, China
State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University
Beijing, China
Introduction:
Alzheimer's disease (AD) exhibits marked heterogeneity in clinical phenotypes, genetic profiles, and neuropathology (Robinson et al., 2023). Mild cognitive impairment (MCI) is a well-established precursor to AD conversion (Chen et al., 2022). Revealing the heterogeneity across the AD continuum its underlying biological basis is crucial for advancing precision medicine tailored to this disease (Zhao et al., 2024). In this study, we employed brain age prediction model to investigate network-specific patterns and longitudinal trajectories across distinct stages of the AD continuum. By leveraging individual-level prediction differences, we aim to uncover distinct progression patterns and their biological correlates.
Methods:
Datasets: Two independent cohorts were analyzed: UK Biobank (n=28,341 NCs) and ADNI (n=1,412; 525 NCs, 611 MCI, 276 AD patients) (Fig. 1).
Brain Age Modeling: All T1-weighted MRI scans were processed using FreeSurfer (Fischl, 2012) and parcellated into 7 networks using the Yeo-7 atlas (Thomas Yeo et al., 2011). An enhanced Simple Fully Convolutional Network model (Peng et al., 2021), pretrained on UK Biobank data, estimated network-specific brain age, then transferred to unseen UK Biobank and ADNI datasets with linear correction (Beheshti et al., 2019).
Nested Case–Control Approach: ADNI participants were grouped by diagnostic transitions: (1) NC (n=189, stable cognition), (2) Stable MCI (sMCI, n=157), (3) Progressive MCI (pMCI, n=300, MCI-to-AD converters), and (4) AD (n=276, baseline AD).
Statistical Analysis: Baseline network-specific PAD estimates were derived using the emmeans package in R (Lenth, 2017). Linear mixed models assessed longitudinal PAD changes, adjusting for group, time, and covariates (age, sex, education, APOE status), with a time × group interaction term to capture group-specific trajectories. Individual PAD change rates were estimated as the sum of fixed-effect time slopes, group-specific time interactions, and subject-specific random slopes. Generalized linear models examined associations between PADs, genetics, pathology, mental health, and cognition. Path analysis tested mediation of genetic and pathological effects on cognition by network PADs. RandomForestClassifier evaluated PADs' predictive power for MCI-to-AD progression.
Brain Age Modeling: All T1-weighted MRI scans were processed using FreeSurfer (Fischl, 2012) and parcellated into 7 networks using the Yeo-7 atlas (Thomas Yeo et al., 2011). An enhanced Simple Fully Convolutional Network model (Peng et al., 2021), pretrained on UK Biobank data, estimated network-specific brain age, then transferred to unseen UK Biobank and ADNI datasets with linear correction (Beheshti et al., 2019).
Nested Case–Control Approach: ADNI participants were grouped by diagnostic transitions: (1) NC (n=189, stable cognition), (2) Stable MCI (sMCI, n=157), (3) Progressive MCI (pMCI, n=300, MCI-to-AD converters), and (4) AD (n=276, baseline AD).
Statistical Analysis: Baseline network-specific PAD estimates were derived using the emmeans package in R (Lenth, 2017). Linear mixed models assessed longitudinal PAD changes, adjusting for group, time, and covariates (age, sex, education, APOE status), with a time × group interaction term to capture group-specific trajectories. Individual PAD change rates were estimated as the sum of fixed-effect time slopes, group-specific time interactions, and subject-specific random slopes. Generalized linear models examined associations between PADs, genetics, pathology, mental health, and cognition. Path analysis tested mediation of genetic and pathological effects on cognition by network PADs. RandomForestClassifier evaluated PADs' predictive power for MCI-to-AD progression.
Results:
Network-specific brain age models achieved strong performance across all unseen independent datasets, with Pearson correlations >0.85 and mean absolute error (MAE) <3.60 years. Fig. 2A highlights distinct network PAD patterns across AD spectrum subgroups: AD exhibits the highest deviations, followed by pMCI and sMCI, with NC near 0, reflecting progressive brain aging along the disease continuum. Longitudinally, pMCI shows accelerated PAD increases prior to AD diagnosis, especially in SM, DAN, VAN, and FPN (Fig. 2B). Meanwhile, the DMN consistently shows the highest PAD, indicating it is the earliest and most affected network (Fig. 2C). Fig. 2D reveals significant positive correlations between baseline PADs in DMN and limbic networks and pathological biomarkers, alongside negative associations with cognition; PAD change rates in networks with pre-AD accelerated aging (DAN, VAN, FPN) correlate strongly with genetic risk, pathology, mental health, and cognition (P<.05). Genetic and pathological biomarkers (APOE-ε4, PHS, pTau181) mediate cognitive decline via network-specific PAD changes in SM, DAN, VAN, and FPN (P<.05, Fig. 2E). Figure 2F shows that integrating genetics, pathology, and network PADs (AUC=0.95) significantly boosts diagnostic accuracy for MCI-to-AD progression over genetics and pathology alone (AUC=0.89).
Conclusions:
Network-specific brain age models uncover heterogeneous spatial aging patterns across the AD continuum from both cross-sectional and longitudinal perspectives, linking biological drivers (genetics, pathology) to clinical outcomes. These findings highlight PADs as a promising, personalized biomarker for precision AD prognosis and therapeutic targeting.
Disorders of the Nervous System:
Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s) 1
Lifespan Development:
Aging 2
Modeling and Analysis Methods:
Multivariate Approaches
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
Anatomy and Functional Systems
Novel Imaging Acquisition Methods:
Anatomical MRI
Keywords:
Aging
Degenerative Disease
Statistical Methods
STRUCTURAL MRI
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using APA citation style.
Chen, J., Wang, S., Chen, R., & Liu, Y. (2022). Editorial: Neuroimaging Biomarkers and Cognition in Alzheimer’s Disease Spectrum. Frontiers in Aging Neuroscience, 14, 848719. https://doi.org/10.3389/fnagi.2022.848719
Fischl, B. (2012). FreeSurfer. NeuroImage, 62(2), 774–781. https://doi.org/10.1016/j.neuroimage.2012.01.021
Lenth, R. V. (2017). emmeans: Estimated Marginal Means, aka Least-Squares Means (p. 1.10.7) [Dataset]. https://doi.org/10.32614/CRAN.package.emmeans
Peng, H., Gong, W., Beckmann, C. F., Vedaldi, A., & Smith, S. M. (2021). Accurate brain age prediction with lightweight deep neural networks. Medical Image Analysis, 68, 101871. https://doi.org/10.1016/j.media.2020.101871
Robinson, J. L., Xie, S. X., Baer, D. R., Suh, E., Van Deerlin, V. M., Loh, N. J., Irwin, D. J., McMillan, C. T., Wolk, D. A., Chen-Plotkin, A., Weintraub, D., Schuck, T., Lee, V. M. Y., Trojanowski, J. Q., & Lee, E. B. (2023). Pathological combinations in neurodegenerative disease are heterogeneous and disease-associated. Brain: A Journal of Neurology, 146(6), 2557–2569. https://doi.org/10.1093/brain/awad059
Thomas Yeo, B. T., Krienen, F. M., Sepulcre, J., Sabuncu, M. R., Lashkari, D., Hollinshead, M., Roffman, J. L., Smoller, J. W., Zöllei, L., Polimeni, J. R., Fischl, B., Liu, H., & Buckner, R. L. (2011). The organization of the human cerebral cortex estimated by intrinsic functional connectivity. Journal of Neurophysiology, 106(3), 1125–1165. https://doi.org/10.1152/jn.00338.2011
Zhao, K., Chen, P., Wang, D., Zhou, R., Ma, G., & Liu, Y. (2024). A Multiform Heterogeneity Framework for Alzheimer’s Disease Based on Multimodal Neuroimaging. Biological Psychiatry, 0(0). https://doi.org/10.1016/j.biopsych.2024.12.009