Cardio-metabolic diseases-including diabetes, hypertension, obesity, and dyslipidemia-are major risk factors for cognitive decline and depression. These conditions share underlying mechanisms such as insulin resistance, vascular dysfunction, and chronic inflammation. Importantly, these mechanisms interact bidirectionally with overactivation of stress-responsive neural circuits, neuroinflammation, and impaired neuroplasticity, perpetuating a vicious cycle of metabolic and neuropsychiatric deterioration.
At the clinical level, identifying individuals most susceptible to these adverse health trajectories is challenging. It requires integrating diverse risk factors to define robust predictors of metabolic and mental health outcomes.
Our project aims to:
1. Identify shared risk factors for cardio-metabolic diseases, depression, and cognitive decline. Using machine learning approaches (e.g., XGBoost), we will cluster the top predictors of these three conditions, leveraging the rich UK Biobank dataset, which includes demographic information, cardio-metabolic profiles, mental health diagnoses, cognitive function tests, and lifestyle and dietary data.
2. Determine circulating biomarkers that reflect biologically relevant pathways linking cardio-metabolic disease, depression, and cognitive decline. We will integrate metabolomics and proteomics data from UK Biobank into our machine learning models to identify key molecular signatures.
3. Develop a genetic risk score based on genes encoding these biomarkers to identify individuals at highest risk for developing cardio-metabolic disease, depression, or cognitive decline at an early stage. We will generate weighted genetic risk scores leveraging the comprehensive genomic data available in UK Biobank.
This integrative approach will enable a deeper understanding of shared pathophysiological mechanisms and support the development of predictive tools for early identification and targeted prevention strategies.