Which measures of aging best predicted death? Brain scans and walking speed held up

Two people who are both 73 can be in very different shape. Researchers have built many tools to put a number on that difference, a biological age: epigenetic clocks that read chemical tags on DNA, measures of strength and walking speed, brain scans, and more recently “organ clocks” that estimate the age of the heart, liver, brain and other organs from proteins in the blood. Organ clocks drew attention because they suggest one organ can age faster than another in the same person.

But which of these tools tells you the most? Most studies test one at a time, in different groups of people. A team at the University of Edinburgh compared them side by side in the same people, using whether participants died during follow-up as the yardstick.

What they did

The Lothian Birth Cohort 1936 is a group of people born in 1936 who took part in a national survey of mental ability at school at age 11 and were recruited around age 70 in Scotland. They have been tested about every three years since. The team used measurements taken at about age 73 from up to 861 participants, of whom 444 had died by 2025, an average of 16 years later.

They compared 23 measures:

• Blood proteins: organ clocks for 11 organs, calculated from about 9,700 protein targets.
• Epigenetic: a clock called GrimAge2, and the length of telomeres, the protective caps on chromosomes.
• Body function: grip strength, time to walk 6 metres, and lung function (how much air a person can blow out, and how fast).
• Brain: MRI-based brain volume and grey matter volume, the amount of damage to small blood vessels visible as bright spots on scans, an MRI “brain age”, and a general thinking score from 13 tests.

For each measure, they estimated how much higher the risk of death was for people one standard deviation (roughly, one typical step away from the average) worse off, after accounting for age and sex, and then also for smoking, alcohol, body weight and common diseases.

What they found

• Almost everything predicted death. All measures except telomere length were linked to risk of dying.
• Organ clocks: real but middling. Each step of faster organ aging was linked to 16% to 43% higher risk, with the liver, immune system and heart clocks strongest.
• Others did better. A faster GrimAge2 clock was linked to 62% higher risk per step. Smaller brain volume, weaker lung function, smaller grey matter and lower thinking scores were each linked to 44% to 52% higher risk.
• Four measures stood on their own. When all measures were put into one model together, which shows which ones add information beyond the others, only four stayed significant: brain volume, small-vessel damage in the brain, the thinking score and walking time. Together they accounted for about a fifth of the variation in mortality risk; adding all the other measures raised that only slightly.
• Protein clues. Out of the 9,703 protein targets, 368 were linked to death after accounting for age and sex. Proteins tied to higher risk were mostly involved in the immune system and inflammation; the strongest was GDF15, a protein made by stressed cells. Those tied to lower risk were often involved in keeping DNA intact.

What to keep in mind

• Links, not causes. The authors stress the study is observational: a protein linked to death is not shown to cause it, and changing it may do nothing.
• One group. Participants were relatively healthy Scots of the same age and similar ancestry; the authors say results need checking in more diverse groups.
• One snapshot. Measures were taken once, at about 73. Changes over time might predict better.
• All causes of death together. The study was too small to look at deaths from heart disease or cancer separately, which is where organ clocks might show their value.
• Not a personal test. These are averages across hundreds of people. A score on any one measure says little about how long a given person will live.

Why it matters

New aging clocks appear all the time, and the newest are not automatically the best. In this group of older adults, simple or long-established measures, such as how fast someone walks, how well they think and what a brain scan shows, carried most of the information about survival, and the organ clocks, while linked to survival, were outdone by several of them. For researchers testing anti-aging treatments, that argues for combining different kinds of measures rather than relying on a single new clock.