Contents
The aging process, fast and slow
We appear to age at a steady pace when we consider the days on the calendar and the earth’s orbits around its own axis and around the sun. In the end, every day we live is a day we grow older. As a result, our chronological age is determined by the passing of the days and the ticking of the clock.
There is also a biological age, which doesn’t always align with the calendars and clocks. Though we cannot age backward (yet? ), not everyone ages at the same rate biologically. People can be younger or older than their chronological age. Our biological age cannot be measured by clocks and calendars, so we need to find other aging clocks. Recently, much research has been conducted on internal biological signals that correlate with various aging processes in our bodies.
Protein levels in the blood can, for example, be used to measure biological age or inflammation. Using epigenetic aging clocks is a recent method for capturing biological age. Different aspects of aging are thought to be reflected in the dynamics of epigenetic tags that latch onto DNA and change gene activity. (Epigenetic clocks have been used, for example, to study embryonic development or to find out why some animals seem to age slowly.)
Synchronizing clocks
In light of this diversity of aging clocks, it might be a good idea to combine them. This is already happening with different epigenetic clocks, but what if we looked at several physiological variables, such as blood markers and cardiovascular function? (Epigenetic changes can affect all of these things.)
According to a new study:
There are few biological age measurements that integrate traits across hierarchical levels (e.g., proteome, metabolome, and phenome), and across the many organ systems involved in aging.
You can be fairly certain that when you read something like that in a paper’s introduction, the authors will do something about it. Scientists used datasets from SardiNIA (ave age of 14 – 101, with a mean of 43.7 years, and 57% were females), and InCHIANTI (measures of blood markers, cardiovascular function, and personality tests).
In order to identify which characteristics strongly correlated with biological aging, the researchers applied a variety of machine learning models to a large portion of that data. In the next step, the models were asked to predict the ages of the individuals in the datasets. The most accurate model would be based on a large number of individuals. In large datasets, this assumption seems reasonable. Later, the most appropriate model can be applied to individuals to determine how far they deviate from the average biological age.)
Random forest classifiers performed best. It depends on making several decision trees for each prediction, and then looking at the ‘forest’ of all those trees, and if a lot of them look the same, you’ve probably got a good model.
Eventually, the researchers identified 52 traits, from blood cell counts to cholesterol to blood pressure and pulse wave velocity. Among these, cardiovascular risk markers were the most influential. The researchers were able to calculate a physiological aging rate (PAR) based on these traits.
These are a few interesting correlations:
There was a connection between PAR and mortality (no surprise there).
Several epigenetic aging clocks correlated fairly well with the PAR (this indicates that epigenetic changes trickle down, or perhaps up?).
Heritability of PAR was moderate. There may be reasons for this, such as a shared environment for parents and children, shared habits, etc.
But the scientists did have some genetic data to examine, and they discovered a number of SNPs (single-letter DNA changes) associated with PAR. (The majority are found in the CFI and GAR1 genes, with several in or near the LINC00202 region, which codes for a long non-coding RNA that influences gene activity). Despite genetic factors, however, the researchers emphasize that: