Association of omics markers of tissue aging with health outcomes and mortality

Using transcriptomic and epigenomics data across multiple tissues and matched blood from the Genotype-Tissue Expression (GTEx) database and the UK Biobank we have constructed a computational framework that calculates the rate of aging of tissues in an individual using blood gene expression, DNA methylation or blood clinical markers. The next step is to validate and optimize the resulting models to predict organ-specific aging in disease states specific to the organ of interest, and mortality using clinical and omics data from the Health and Retirement Study.

Objectives
Validate and optimize blood-based diagnostic tools using clinical and omics data from the Health and Retirement Study.
• Determine if the signatures of tissue aging in the blood are associated with disease incidence.
• Estimate the association between the blood-based organ aging and mortality.
• Build omics predictors of organ age estimations from blood gene expression, DNA methylation or blood clinical markers (i.e. prediction of the prediction).

Methods
Using the generated model to predict tissue aging using blood we will assess the association between tissue age acceleration and clinical data in the Health and Retirement Study cohort. We will study the association between age acceleration and two types of clinical information:

Chronic diseases: We expect to observe a positive tissue age acceleration in the tissues associated with the diseases diagnosed in the donors. For example, individuals diagnosed with COPD are expected to have higher age acceleration in the lungs compared to healthy individuals. For all these analyses we will perform logistic regression analysis to examine the binary outcome of disease presence or absence in relation to tissue age acceleration.

Mortality: We hypothesize that our models will reveal a significant association between accelerated aging in specific tissues and overall mortality risk. Cox proportional hazards models will be utilized to estimate the hazard ratios for mortality, with tissue age acceleration as the primary variable of interest. We will adjust these models for potential confounders such as baseline age and sex.

We've developed a computer program using data from a large database to figure out how quickly different parts of our body age, based on information from our blood. We want to see if this can help us understand how diseases develop and how long someone might live. To do this, we'll use information from the Health and Retirement Study, a study on aging and health, to check if the aging patterns we find in blood are connected to diseases people have and how long they live.