Scientists can read the 'rust' on a person's DNA to predict when they'll die
The results are considered to be very statistically reliable, says researcher Steve Horvath
As we age, there are a few ways you can think about how old you are. There's how old you feel, or how old the calendar says you are. And now by using "epigenetic clocks," scientists can also tell how old you are according to your DNA.
Steve Horvath, a professor in the Department of Human Genetics and Biostatistics at the University of California Los Angeles, is at the forefront of this research. He's developed an epigenetic clock that he ominously calls GrimAge.
"It really allows a researcher to predict how long a person will live," said Horvath. "And more generally, it also relates to how long somebody's so-called health span."
This tool is not available for the general public. Horvath developed it to help with aging research.
"I'm really an aging researcher in both senses of the word and I'm trying to find out why we age and what can be done against it," said Horvath. "And when you want to study aging in a rigorous manner, you need to be able to measure it very precisely."
Epigenetic clock using DNA methylation
These epigenetic clocks measure the level of epigenetic "rust" that accumulates on DNA throughout a lifetime. This rust, otherwise known as DNA methylation, is a chemical tag that goes on top of the DNA to silence certain genes. The tags are typically added as a biological reaction to different kinds of environmental stresses.
All scientists need is a drop of blood from which they can extract the DNA and measure these epigenetic modifications of the DNA molecule.
"Based on thousands of these methylation measurements, we then apply a mathematical formula, which yields a number, literally in units of years. It would say the GrimAge of your blood is, for example, 45 years old."
Horvath says the tool takes methylation measurements from specific spots on the DNA to reconstruct a person's smoking history, various stress factors related to immune function, obesity, as well as a number of other health factors.
Predicting time to death
Once they have the methylation measurements, they use statistical methods to work out an estimate of mortality risk.
"One can certainly predict time-to-death, but that would be a big error rate associated with that," said Horvath. "For example, I could come up with a prediction that would say you have another 15 years ahead of you, but then I would have to emphasize that there is an arrow of plus-minus maybe four or five years."
Since it's based on DNA methylation, which is influenced by the environment, if a person quits smoking or loses excess weight, they could bring down their GrimAge mortality risk.
Accuracy of GrimAge
To figure out how accurate his GrimAge tool is, Horvath tested it on more than 7,000 people.
"We really leveraged epidemiological cohort data where researchers had gone through the trouble of collecting blood samples back in the 1990s from individuals and then stored them in freezers," said Horvath.
It's really more likely that planet Earth will be hit by an asteroid tomorrow than that this predictor doesn't work.- Dr. Steve Horvath, UCLA
For all of these individuals, they also had follow-up information, including whether they were still alive, or if not, their cause of death.
"We are, of course, statisticians and we can calculate statistical significance levels associated with these prediction tools," said Horvath. "And it's so significant, that it's really more likely that planet Earth will be hit by an asteroid tomorrow than that this predictor doesn't work."