Fight Aging!

Epigenetic clocks to measure biological age that are derived from blood sample data (and therefore the epigenetic patterns of white blood cells) are here demonstrated to perform poorly in other cell types and tissue samples. This is known to be the case, and is an inevitable result given that (a) epigenetics differs between cell types, and (b) machine learning is used to derive an algorithm that matches up the data in hand to the epidemiological outcomes of interest. To produce a clock that performs well in multiple tissue types, one has to deliberately aim at that goal, incorporating epigenetic data obtained from all of those tissue types into the machine learning process. Some researchers have worked on this, as well as on the production of clocks that work in different species.

Epigenetic clocks are a common group of tools used to measure biological aging - the progressive deterioration of cells, tissues and organs. Epigenetic clocks have been trained almost exclusively using blood-based tissues but there is growing interest in estimating epigenetic age using less-invasive oral-based tissues (i.e., buccal or saliva) in both research and commercial settings. However, differentiated cell types across body tissues exhibit unique DNA methylation landscapes and age-related alterations to the DNA methylome. Applying epigenetic clocks derived from blood-based tissues to estimate epigenetic age of oral-based tissues may introduce biases.

We tested the within-person comparability of common epigenetic clocks across five tissue types: buccal epithelial, saliva, dry blood spots, buffy coat (i.e., leukocytes), and peripheral blood mononuclear cells. We tested 284 distinct tissue samples from 83 individuals aged 9-70 years. Overall, there were significant within-person differences in epigenetic clock estimates from oral-based versus blood-based tissues, with average differences of almost 30 years observed in some age clocks. In addition, most epigenetic clock estimates of blood-based tissues exhibited low correlation with estimates from oral-based tissues despite controlling for cellular proportions and other technical factors.

Our findings indicate that application of blood-derived epigenetic clocks in oral-based tissues may not yield comparable estimates of epigenetic age, highlighting the need for careful consideration of tissue type when estimating epigenetic age.

Link: https://doi.org/10.1101/2024.07.16.603774