A new study has found 17 proteins in the blood of older adults that were linked to their risk of dying from any cause over about eight years. Several of the same proteins also tracked with measures of how old a person’s body and individual organs seemed, and key results held up when checked in a second, separate group of people.1
The work appeared open access in Geromedicine, a journal from Science Exploration Press.12 It is an observational study: it shows which proteins go along with higher or lower risk, not whether they cause anything.1
What the researchers did
The team measured 4,265 proteins in blood plasma (the liquid part of blood) from 1,025 older adults in a long-running research project called the LonGenity study. Participants were followed for an average of 8.05 years.1
Using a standard statistical method for studying how risk changes over time, the researchers looked for proteins whose levels lined up with death from any cause. They then asked whether those proteins were also related to how many chronic diseases people had, and to “biological age” estimates built from blood proteins, both for the whole body and for specific organs.1
Biological age is an attempt to capture how fast the body and its organs are aging, as opposed to chronological age, which simply counts years. The study’s summary frames the central question this way: why do people of the same age often differ so much in health, disease and resilience?1
Finally, the researchers checked their findings in 1,124 participants from a different project, the Baltimore Longitudinal Study of Aging.1
The proteins that stood out
Out of more than 4,000 proteins, 17 were significantly linked to death from any cause. A handful drew attention because their link to death risk also showed up in the biological-age measures.1
- Higher levels, lower risk: people with more of two proteins, called ERBB1 (a receptor that responds to a growth signal) and ANTXR2, had lower death risk and looked biologically younger across several organ systems.1
- Higher levels, higher risk: people with more of three other proteins, called MIC-1 (also known as GDF15), HE4 and IGFBP-2, had higher death risk and looked biologically older across several organ systems.1
Some links were more specific. Higher MIC-1/GDF15 went along with an older biological age for the body as a whole, while ERBB1 went along with a younger biological age of the liver.1
The researchers also reported that the organ measures behaved in ways that made biological sense. A well-known heart marker, NT-proBNP, showed its strongest link with heart aging in both groups of people, and weaker or non-significant links with kidney aging. Brain and immune aging measures showed especially strong agreement between the two groups, notably for HE4, MIC-1, ERBB1, a protein called NOTUM, and IGFBP-2.1
What it does not show
The study’s own summary is direct about its limits. It does not establish that these proteins cause aging or death. It does not provide a clinical blood test that can predict a person’s lifespan or disease outcome. What it offers is a set of statistical associations in two long-term groups of people, and a list of proteins that the authors say deserve more research into how they work and whether they matter in the clinic.1
A few other gaps are worth keeping in mind:
- Links are not levers. Finding that people with more of a protein lived longer does not mean raising that protein would help anyone live longer. The protein could be a sign of something else going on in the body.
- Who was studied matters. Both groups were made up of participants in research projects, and the summary describes the first group as older adults. Results may not apply the same way to younger people or other populations.
- Biological age is itself an estimate. The organ ages in this study were derived from blood proteins, so the study compares one set of blood measurements with models built from blood measurements.1
The public summary does not state who funded the research.
Who is behind it
The study involved researchers from Albert Einstein College of Medicine, Stony Brook University, the National Institute on Aging, Stanford University School of Medicine, Kannur University and other institutions. The credit on the published figure names Sanish Sathyan and Nir Barzilai among the authors.1
Both Sathyan and Barzilai were also co-authors, with Stanford’s Tony Wyss-Coray and others, of a 2023 study in Nature that estimated the age of individual organs from blood proteins.3 The summary of the new paper does not list any company ties or products connected to the findings.1
How it fits the field
The idea that blood proteins can reveal how fast someone is aging has been building for several years, and this study adds to it rather than starting it.
In the 2023 Nature study, researchers used blood proteins thought to come from specific organs to estimate the age of 11 major organs in 5,676 adults across five groups. They reported that nearly 20% of people showed strongly accelerated aging in one organ and 1.7% in several organs, and that accelerated organ aging came with a 20% to 50% higher risk of death.3
In 2024, a study in Nature Medicine built a protein-based “aging clock” from 2,897 blood proteins in 45,441 people in the UK Biobank. It found 204 proteins that closely predicted age, and faster protein aging was linked to 18 major chronic diseases, to having several diseases at once, and to death risk. The clock performed similarly in groups from China and Finland.4 Coverage of that study noted that a smaller set of just 20 proteins gave similar accuracy.5
One protein from the new study has shown up before. A 2024 study in the journal Aging Cell looked at which blood proteins were linked to living to age 90 in 3,067 people in the U.S. and checked the results in 4,690 people in Iceland. The most significant proteins in both groups were GDF-15, the same protein also called MIC-1, and NT-proBNP.6
Another 2026 study in Aging Cell measured 5,416 blood proteins in 52 recently hospitalized older patients, 52 matched older people who were not hospitalized, and 20 healthy young adults. It identified 311 proteins that differed both with age and with disease burden, which the authors proposed as candidate markers of biological aging.7
Interest is also reaching the clinic side of longevity. One longevity clinic’s blog, for example, cites this line of blood-protein research while discussing therapeutic plasma exchange, a procedure that replaces part of a person’s plasma, and notes that the supporting findings remain mainly from animal studies.8 The new study does not test any treatment.1
What to watch next
The authors describe their 17 proteins as candidate markers to be investigated in future studies of healthy aging and longevity.1 Questions worth following include:
- whether the same proteins, especially those that held up in both groups, appear again in larger and more varied populations;
- whether lab studies can show what these proteins actually do in the body, which the study itself says is still needed;1
- whether protein-based measures of organ age ever become a validated clinical test, something this study explicitly does not provide.1
For now, the finding is a map of associations: a small set of blood proteins that rise or fall with the risk of dying in older adults, and that line up with estimates of how fast their bodies and organs are aging.
