Protecting the Aging Brain: Self-Organizing Networks and Multi-Scale Dynamics Under Energy Constraints
Approved Research ID: 37462
Approval date: December 7th 2018
In this study, we investigate the impact of diet, diabetic status, and genetics on the acceleration of brain aging and risk for age-based cognitive impairment. Dietary increase in sugar over the past 100 years has led to a national epidemic of insulin resistance (Type 2 diabetes), which has been identified as an early risk factor for late-life dementia. Our preliminary analyses of the first available batch of data to emerge from the Human Connectome Project Lifespan Study suggest that, starting as early as ~45 yrs., brain networks start to become less stable, which may reflect the first early signs of the brain's response to decreased access to energy, a process that we hypothesize to occur at an earlier age in Type 2 diabetics, as insulin resistance restricts cells' (including neurons') ability to utilize blood sugar. Over the next two years, we will test this hypothesis in a much larger sample: (1) to determine if network stability generally decreases with age; (2) if so, to determine the specific age at which the brain starts to show these aging effects and (3) to determine whether these effects are seen at an earlier age in people who eat a sugar-rich diet and/or suffer from Type 2 diabetes. fMRI data will be used to derive network stability. Structural, FLAIR, DTI data will be used to coregister and interpret fMRI results, with computational models that integrate functional networks (fMRI), structural connectivity (T1, DTI), and degree of signal transmission (hyperintensities derived from FLAIR). Proxy measures for insulin resistance will be derived from self-reported and medical diagnosis of diabetes, body fat percentage (obtained via DXA), and pancreatic volume, as well as testing association with ApoE SNPs (involved in lipid metabolism, and a biomarker for AZD risk). Covariates will be used to statistically probe the complex relationship between insulin resistance and linked comorbidities, such as vascular dementia (assessed via neuroimaging and carotid ultrasound and ECG-derived heart rate variability).
In this study, we investigate the impact of diet, diabetic status, and genetic load on the acceleration of brain aging and risk for age-based cognitive impairment. Our fMRI analyses test the assumption that brain networks are driven not only by function, but also energy constraints. If so, the fuel access-limited brain should bias towards networks with lower-energy demands. This metabolic bias phenotype can be quantified as a Markov-process asymmetry in transitioning from lower-to-higher energy networks as compared to the reverse, and can be generalized using nonequilibrium statistical physics approaches to identify network nodes most vulnerable to failure.
We wish to extend our study to involve longitudinal assessment of the impact of aging and diabetic status on the human brain. Our previous results hinted at the possibility of diabetes-related changes emerging in the brain prior to peripheral changes that otherwise facilitate diagnosis with current methods. We plan to evaluate this hypothesis by using our previously established brain metabolism related biomarkers and quantifying them in states preceding diabetes diagnosis. Our results will serve as the basis for future mechanistic studies aiming to develop effective treatments. Furthermore, annotated longitudinal brain imaging data will allow us to develop deep-learning based predictive algorithms that provide early diagnosis of diabetes.