Characterizing brain-cardiovascular aging using multi-organ imaging and machine learning
Poster No:
875
Submission Type:
Abstract Submission
Authors:
Yalda Amirmoezzi1, Vanessa Cropley2, Sina Mansour L1,3, Caio Seguin1,4, Andrew Zalesky1,5, Ye Tian6
Institutions:
1Systems Lab, Department of Psychiatry, The University of Melbourne, Melbourne, Australia, 2Centre for Youth Mental Health, The University of Melbourne, Melbourne, Australia, 3National University of Singapore, Singapore, 4Department of Psychological and Brain Sciences, Indiana University, Bloomington, IN, 5Department of Biomedical Engineering, The University of Melbourne, Melbourne, Australia, 6Department of Psychiatry, The University of Melbourne, Melbourne, Australia
First Author:
Yalda Amirmoezzi
Systems Lab, Department of Psychiatry, The University of Melbourne
Melbourne, Australia
Systems Lab, Department of Psychiatry, The University of Melbourne
Melbourne, Australia
Co-Author(s):
Sina Mansour L
Systems Lab, Department of Psychiatry, The University of Melbourne|National University of Singapore
Melbourne, Australia|Singapore
Systems Lab, Department of Psychiatry, The University of Melbourne|National University of Singapore
Melbourne, Australia|Singapore
Caio Seguin
Systems Lab, Department of Psychiatry, The University of Melbourne|Department of Psychological and Brain Sciences, Indiana University
Melbourne, Australia|Bloomington, IN
Systems Lab, Department of Psychiatry, The University of Melbourne|Department of Psychological and Brain Sciences, Indiana University
Melbourne, Australia|Bloomington, IN
Andrew Zalesky
Systems Lab, Department of Psychiatry, The University of Melbourne|Department of Biomedical Engineering, The University of Melbourne
Melbourne, Australia|Melbourne, Australia
Systems Lab, Department of Psychiatry, The University of Melbourne|Department of Biomedical Engineering, The University of Melbourne
Melbourne, Australia|Melbourne, Australia
Introduction:
As individuals age, both the brain and cardiovascular systems undergo significant changes, making them susceptible to various neurodegenerative and cardiovascular conditions and increased risk of comorbidity and mortality. Despite the increasingly recognised physiological interplay between the brain and the cardiovascular system, brain networks and circuits that may drive this brain-heart aging axis remain largely unknown. Moreover, the concurrent brain-heart aging may be driven by shared biological and behavioural risk factors. Identifying these risk factors may facilitate the development of synergistic intervention strategies to slow aging and mitigate disease risk.
Methods:
We included both cross-sectional and longitudinal data from the UK Biobank. For the cross-sectional analysis we used 2,904 healthy individuals and for the longitudinal analysis, we included data from 1,236 participants (age range:46–81 years).
We used both morphological and connectivity metrics derived from brain MRI to estimate brain age. Cardiovascular features comprised cardiac magnetic resonance imaging, carotid ultrasound, and physiological data, which were grouped into two categories based on their structural and functional relevance and were used to estimate the biological age of the heart. Separate predictive models were developed for the cardiovascular system (structure, function and all features), and the brain (the whole-brain, the seven brain networks, and the subcortex) to predict each individual's chronological age, followed by calculation of the age gap index. The age gap index thus indicates whether an individual's brain or heart appears older or younger than their chronological age.
In the longitudinal analysis, cardiovascular phenotypes were categorized by artery sections and heart chambers, including the aorta, carotid arteries, ventricle and atrium. Then a structural equation model was applied separately for each cardiovascular compartment to assess whether risk factors for coronary heart disease including total cholesterol, HDL cholesterol, systolic blood pressure, blood pressure medication, age, BMI, diabetes and smoking status, and the total Framingham risk score associates with the rate of brain aging, and whether this relationship is medicated by changes in the heart and arteries.
We used both morphological and connectivity metrics derived from brain MRI to estimate brain age. Cardiovascular features comprised cardiac magnetic resonance imaging, carotid ultrasound, and physiological data, which were grouped into two categories based on their structural and functional relevance and were used to estimate the biological age of the heart. Separate predictive models were developed for the cardiovascular system (structure, function and all features), and the brain (the whole-brain, the seven brain networks, and the subcortex) to predict each individual's chronological age, followed by calculation of the age gap index. The age gap index thus indicates whether an individual's brain or heart appears older or younger than their chronological age.
In the longitudinal analysis, cardiovascular phenotypes were categorized by artery sections and heart chambers, including the aorta, carotid arteries, ventricle and atrium. Then a structural equation model was applied separately for each cardiovascular compartment to assess whether risk factors for coronary heart disease including total cholesterol, HDL cholesterol, systolic blood pressure, blood pressure medication, age, BMI, diabetes and smoking status, and the total Framingham risk score associates with the rate of brain aging, and whether this relationship is medicated by changes in the heart and arteries.
Results:
We found a significant positive association between the brain age gap and the cardiovascular age gap estimated based on combination of all features (r = 0.07, P = 〖7×10〗^(-5)). Next, our network specific analysis revealed significant correlations between the cardiovascular age gap calculated using functional phenotypes and age gaps for somatomotor (r = 0.08, P = 〖 2×10〗^(-5)), salience (r = 0.07, P = 0.0002), and default mode (r = 0.06, P = 0.001) networks and subcortex (r = 0.07, P = 0.0001) (Figure 1). In contrast, the structural cardiovascular age gap was significantly correlated only with the subcortical age gap. For morphology, we found a widespread association with the functional cardiovascular age gap extending across different networks.
Using longitudinal data, we found that the aorta and carotid arteries significantly mediated the relationship between the FRS and the rate of brain aging (Figure 2). Among different risk factors, total cholesterol was associated with the rate of brain aging, with this effect was mediated by the aorta and carotid arteries. Additionally, the aorta and carotid arteries mediated the relationship between the FRS and cognitive performance.
Using longitudinal data, we found that the aorta and carotid arteries significantly mediated the relationship between the FRS and the rate of brain aging (Figure 2). Among different risk factors, total cholesterol was associated with the rate of brain aging, with this effect was mediated by the aorta and carotid arteries. Additionally, the aorta and carotid arteries mediated the relationship between the FRS and cognitive performance.
Conclusions:
The aging of multiple cortical networks and subcortex is selectively associated with the aging of the cardiovascular system, with somatomotor, salience and default mode networks being most predominantly implicated. Moreover, individuals at higher risk of developing cardiovascular disease experience an accelerated rate of brain aging and reduced cognitive performance in later life, with this relationship mediated by morphological changes in aorta and carotid arteries.
Lifespan Development:
Aging 1
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural) 2
Other Methods
Keywords:
Aging
Computational Neuroscience
MRI
Other - Brain-heart
1|2Indicates the priority used for review