Nobody can tell you the day you’ll die. Anyone who says they can is selling something.

Risk is a different question, and it’s one I spent years on. Life insurance, pensions and annuities all run on the odds that a person of a given age dies in a given year. The industry still sets those odds mostly from birthdays, a blood panel and a questionnaire. How I lost patience with that is a story I tell in another post.

Your DNA can sharpen those odds. Here is what the research shows, and how to read it without getting fooled.

The first evidence

The early clocks were trained to guess age, not death. Even so, researchers found a signal. One study pooled four groups of older adults, about 4,600 people, and followed them for years.

People whose DNA read five years older than their actual age had about a 21 percent higher risk of dying during follow-up. After adjusting for smoking, education, heart disease, diabetes and more, the gap shrank to about 16 percent. It didn’t go away.

That last point matters. Smoking and disease explain part of the signal. Not all of it. The DNA was picking up something the usual checklist missed.

A clock trained on death

If you want a clock that predicts lifespan, train it on lifespan. That’s what the team behind a clock called GrimAge did.

They used blood samples from a long-running heart study in Framingham, Massachusetts, where participants had been followed for decades. The clock reads DNA patterns that stand in for seven blood proteins linked to disease. It adds a DNA-based estimate of how much a person has smoked. Then it combines them into one number tuned to time of death.

The smoking piece tells you something. The DNA estimate of smoking predicted lifespan better than what people reported about their own smoking. People forget. People shade the truth on forms. Your DNA keeps its own record.

In testing on about 7,000 people from five other studies, each extra year of GrimAge over a person’s actual age carried about a 10 percent higher risk of death. It also predicted heart disease and cancer. For comparison, the original clock carried about 2 percent per year, and PhenoAge, from the last post, about 5 percent.

What the numbers mean in plain English

Research papers report this as a “hazard ratio.” A hazard ratio of 1.10 means 10 percent higher risk, each year, than someone who is otherwise similar.

Ten percent sounds dramatic. It needs context. If your chance of dying this year is 1 in 100, a 10 percent higher risk makes it 1.1 in 100.

That’s relative risk versus absolute risk. Headlines love relative risk because the numbers are bigger. Your life runs on absolute risk.

Where it adds up is size and time. The extra years multiply. A person whose GrimAge reads five years older than their birthday carries about 60 percent higher risk, every year. Across millions of people, that’s the difference between a pension plan that pays out as promised and one that doesn’t.

How to read “accuracy” claims

You’ll see tests advertised as “95 percent accurate” or able to “predict when you’ll die.” Here’s my checklist before I believe any of it.

Accurate at what? Guessing your birthday is easy and not very useful. Predicting disease or death is hard and very useful. Ask which one.

Compared with what? A test that predicts death in 80-year-olds isn’t impressive, because age already does most of the work. The question is how much the DNA adds beyond age, smoking and the usual medical history.

In whom? A clock that works on the people it was built from may not hold up in anyone else. Look for results in people the model has never seen.

On what sample? Clocks built on blood don’t always give the same answer on saliva or cheek cells. More on that in the next post.

Published where? Peer-reviewed results, with the methods shown, beat a press release every time.

Risk is not destiny

GrimAge is a grim name. I find the research behind it encouraging. A clock that reads older than your birthday is a warning light, not a verdict. The same studies that tie fast clocks to early death also tie them to things people can change: smoking, diet and exercise.

For insurers, pensions and planners, a better read on the odds means fairer prices and fewer surprises. For you, it means a number you can do something about.

One practical question is left in this part of the series. If the information is in your DNA, do you need a needle to get it?

This is not medical advice.

Sources

Jon Sabes is a named inventor on U.S. Patent No. 11,817,214 and managing member of LongevityFP Technologies, LLC, which holds an exclusive license to FOXO Technologies’ epigenetics technology.