Impact of Genetics on the Aging of Functional Connectivity with BrainAGE using the UK Biobank
Poster No:
1141
Submission Type:
Abstract Submission
Authors:
Karen Ardila1,2,3,4, Aashka Mohite1,2,3,4, Emily Munro1,2,3,4,5, Abdoljalil Addeh1,2,3,4, Charlotte Curtis2,6, Quan Long4,7, M. Ethan MacDonald1,2,3,4
Institutions:
1Department of Biomedical Engineering, University of Calgary, Calgary, Alberta, Canada, 2Department of Electrical Engineering & Software Engineering, University of Calgary, Calgary, Alberta, Canada, 3Department of Radiology, University of Calgary, Calgary, Alberta, Canada, 4Hotchkiss Brain Institute, University of Calgary, Calgary, Alberta, Canada, 5Department of Chemical Engineering, University of Calgary, Calgary, Alberta, Canada, 6Department of Mathematics & Computing, Mount Royal University, Calgary, Alberta, Canada, 7Department of Biochemistry & Molecular Biology, University of Calgary, Calgary, Alberta, Canada
First Author:
Karen Ardila
Department of Biomedical Engineering, University of Calgary|Department of Electrical Engineering & Software Engineering, University of Calgary|Department of Radiology, University of Calgary|Hotchkiss Brain Institute, University of Calgary
Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada
Department of Biomedical Engineering, University of Calgary|Department of Electrical Engineering & Software Engineering, University of Calgary|Department of Radiology, University of Calgary|Hotchkiss Brain Institute, University of Calgary
Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada
Co-Author(s):
Aashka Mohite
Department of Biomedical Engineering, University of Calgary|Department of Electrical Engineering & Software Engineering, University of Calgary|Department of Radiology, University of Calgary|Hotchkiss Brain Institute, University of Calgary
Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada
Department of Biomedical Engineering, University of Calgary|Department of Electrical Engineering & Software Engineering, University of Calgary|Department of Radiology, University of Calgary|Hotchkiss Brain Institute, University of Calgary
Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada
Emily Munro
Department of Biomedical Engineering, University of Calgary|Department of Electrical Engineering & Software Engineering, University of Calgary|Department of Radiology, University of Calgary|Hotchkiss Brain Institute, University of Calgary|Department of Chemical Engineering, University of Calgary
Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada
Department of Biomedical Engineering, University of Calgary|Department of Electrical Engineering & Software Engineering, University of Calgary|Department of Radiology, University of Calgary|Hotchkiss Brain Institute, University of Calgary|Department of Chemical Engineering, University of Calgary
Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada
Abdoljalil Addeh
Department of Biomedical Engineering, University of Calgary|Department of Electrical Engineering & Software Engineering, University of Calgary|Department of Radiology, University of Calgary|Hotchkiss Brain Institute, University of Calgary
Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada
Department of Biomedical Engineering, University of Calgary|Department of Electrical Engineering & Software Engineering, University of Calgary|Department of Radiology, University of Calgary|Hotchkiss Brain Institute, University of Calgary
Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada
Charlotte Curtis
Department of Electrical Engineering & Software Engineering, University of Calgary|Department of Mathematics & Computing, Mount Royal University
Calgary, Alberta, Canada|Calgary, Alberta, Canada
Department of Electrical Engineering & Software Engineering, University of Calgary|Department of Mathematics & Computing, Mount Royal University
Calgary, Alberta, Canada|Calgary, Alberta, Canada
Quan Long
Hotchkiss Brain Institute, University of Calgary|Department of Biochemistry & Molecular Biology, University of Calgary
Calgary, Alberta, Canada|Calgary, Alberta, Canada
Hotchkiss Brain Institute, University of Calgary|Department of Biochemistry & Molecular Biology, University of Calgary
Calgary, Alberta, Canada|Calgary, Alberta, Canada
M. Ethan MacDonald
Department of Biomedical Engineering, University of Calgary|Department of Electrical Engineering & Software Engineering, University of Calgary|Department of Radiology, University of Calgary|Hotchkiss Brain Institute, University of Calgary
Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada
Department of Biomedical Engineering, University of Calgary|Department of Electrical Engineering & Software Engineering, University of Calgary|Department of Radiology, University of Calgary|Hotchkiss Brain Institute, University of Calgary
Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada|Calgary, Alberta, Canada
Introduction:
Brain aging leads to altered connectivity in rfMRI studies and is influenced by genetic factors, yet the full impact of genetics remains unclear (Pang et al., 2019). Recent advances in neuroimaging and machine learning have resulted in models that predict brain aging by calculating the Brain Age Gap Estimate (BrainAGE), a useful biomarker for neurological diseases (Sanford et al., 2022), from the difference between the model's predicted age and the chronological age (Foo et al., 2020; Wilms et al., 2020; Gonneaud et al., 2021; Ardila et al., 2023). BrainAGE models suggest that brain systems vulnerable to aging overlap with those susceptible to neurodegeneration but are limited in cohort sizes and approach (Zhao et al., 2019; Brouwer et al., 2022). This research aims to investigate the genetic underpinnings of functional connectivity changes during brain aging using machine learning, neuroimaging, and genetic data from the UK Biobank.
Methods:
This cross-sectional study utilized genetic and rfMRI partial correlation matrices (upper diagonal) data from 37,449 subjects (aged 44-82, mean 64 years, 52.2% female) from the UK Biobank. BrainAGE was computed by inputting measurements from each correlation matrix (Fig. 1) to a machine learning based age prediction model with bias correction (Beheshti et al., 2019).
A Genome-wide association study (GWAS) was conducted using the BrainAGE as a phenotype input. The GWAS was implemented with PLINKv1.9 (Chang et al., 2015), with rigorous data quality control including filtering based on missingness, sex discrepancy, minor allele frequency, Hardy-Weinberg equilibrium, and familial relatedness (Marees et al., 2018). Associations with a p-value less than 1e-04 were considered significant genetic variants to accelerated aging of functional connectivity.
A Genome-wide association study (GWAS) was conducted using the BrainAGE as a phenotype input. The GWAS was implemented with PLINKv1.9 (Chang et al., 2015), with rigorous data quality control including filtering based on missingness, sex discrepancy, minor allele frequency, Hardy-Weinberg equilibrium, and familial relatedness (Marees et al., 2018). Associations with a p-value less than 1e-04 were considered significant genetic variants to accelerated aging of functional connectivity.
Results:
The BrainAGE model performance presented an R-squared score of 0.86 and a mean absolute error of 2.18. The GWAS identified 78 candidate single-nucleotide polymorphisms (SNPs), 10 independent significant SNPs, and 7 lead SNPs. The phenogram of the GWAS results related to accelerated brain aging shows the corresponding effect of each independent significant SNP across the human genome (Fig. 2).
The genes found were TOMM40, previously related to Alzheimer's disease, body mass index, cardiovascular disease risk factors, age-related macular degeneration, longevity, and cerebrospinal fluid t-tau levels (GWAS Catalog, 2023a); APOE, with Alzheimer's disease biomarker, cholesterol levels, hippocampal volume, cerebrospinal fluid t-tau levels, parental longevity, and C-reactive protein levels (GWAS Catalog, 2023b); APOC1, with Alzheimer's disease biomarker, cholesterol levels, longevity, cerebrospinal fluid amyloid beta 42 levels, and C-reactive protein levels (GWAS Catalog, 2023c); and finally, SRXN1 with body mass index, height, autism spectrum disorder, attention deficit-hyperactivity disorder, bipolar disorder, major depressive disorder, and schizophrenia (GWAS Catalog, 2023d).
The genes found were TOMM40, previously related to Alzheimer's disease, body mass index, cardiovascular disease risk factors, age-related macular degeneration, longevity, and cerebrospinal fluid t-tau levels (GWAS Catalog, 2023a); APOE, with Alzheimer's disease biomarker, cholesterol levels, hippocampal volume, cerebrospinal fluid t-tau levels, parental longevity, and C-reactive protein levels (GWAS Catalog, 2023b); APOC1, with Alzheimer's disease biomarker, cholesterol levels, longevity, cerebrospinal fluid amyloid beta 42 levels, and C-reactive protein levels (GWAS Catalog, 2023c); and finally, SRXN1 with body mass index, height, autism spectrum disorder, attention deficit-hyperactivity disorder, bipolar disorder, major depressive disorder, and schizophrenia (GWAS Catalog, 2023d).
Conclusions:
This study investigated the genetic predisposition of specific SNPs and genes in relation to rfMRI brain correlation matrix changes and accelerated brain aging. The main contributions are the associations with genes predominantly related with longevity, Alzheimer's disease, cerebrospinal fluid t-tau levels and amyloid beta, C-reactive protein levels, and mental traits.
Suggestions for future research to enhance the BrainAGE prediction model include the potential the inclusion of additional or different imaging modalities for more precise or varied detection of SNPs.
The study identified several gene associations with brain aging, contributing to the knowledge of the genetic underpinnings of brain aging related to connectivity, and paving the way for potential precision medicine targets.
Suggestions for future research to enhance the BrainAGE prediction model include the potential the inclusion of additional or different imaging modalities for more precise or varied detection of SNPs.
The study identified several gene associations with brain aging, contributing to the knowledge of the genetic underpinnings of brain aging related to connectivity, and paving the way for potential precision medicine targets.
Genetics:
Genetic Association Studies
Lifespan Development:
Aging 1
Modeling and Analysis Methods:
fMRI Connectivity and Network Modeling 2
Keywords:
Aging
Machine Learning
Other - Neurogenetics; Resting-state fMRI; Genome-wide association study
1|2Indicates the priority used for review
Provide references using author date format
Beheshti, I. et al. (2019) ‘Bias-adjustment in neuroimaging-based brain age frameworks: A robust scheme’, NeuroImage: Clinical, 24, p. 102063.
Brouwer, R.M. et al. (2022) ‘Genetic variants associated with longitudinal changes in brain structure across the lifespan’, Nature Neuroscience, 25(4), pp. 421–432.
Chang, C.C. et al. (2015) ‘Second-generation PLINK: Rising to the challenge of larger and richer datasets’, GigaScience, 4(1), p. 7.
Foo, H. et al. (2020) ‘Genetic influence on ageing-related changes in resting-state brain functional networks in healthy adults: A systematic review’, Neuroscience & Biobehavioral Reviews, 113, pp. 98–110.
Gonneaud, J. et al. (2021) ‘Accelerated functional brain aging in pre-clinical familial Alzheimer’s disease’, Nature Communications 2021 12:1, 12(1), pp. 1–17.
GWAS Catalog (2023a) Gene TOMM40.
GWAS Catalog (2023b) Gene APOE.
GWAS Catalog (2023c) Gene APOC1.
GWAS Catalog (2023d) Gene SRXN1.
Marees, A.T. et al. (2018) ‘A tutorial on conducting genome-wide association studies: Quality control and statistical analysis’.
Pang, S.Y.Y. et al. (2019) ‘The interplay of aging, genetics and environmental factors in the pathogenesis of Parkinson’s disease’, Translational Neurodegeneration 2019 8:1, 8(1), pp. 1–11.
Sanford, N. et al. (2022) ‘Sex differences in predictors and regional patterns of brain age gap estimates’, Human Brain Mapping, 43(15), pp. 4689–4698.
Wilms, M. et al. (2020) ‘Bidirectional Modeling and Analysis of Brain Aging with Normalizing Flows’, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 12449 LNCS, pp. 23–33.
Zhao, B. et al. (2019) ‘Genome-wide association analysis of 19,629 individuals identifies variants influencing regional brain volumes and refines their genetic co-architecture with cognitive and mental health traits’, Nature Genetics, 51(11), pp. 1637–1644.