Could one blood sample say something about your risk of dying in the next five or ten years? Researchers at the University of Surrey tried to find out by measuring more than 2,900 proteins in blood samples from 38,150 UK Biobank participants. Their study, published in PLOS ONE in November 2025 and widely discussed since, found small panels of proteins that added a modest amount of predictive power to age, sex and lifestyle factors. It is a research result, not a test you can order at your lab. This article explains what was measured, how to read the numbers, and why the routine markers already on your lab report matter more for you today.
What the researchers measured
The team used data from 38,150 people in the UK Biobank, a large long-term health study in the United Kingdom. Their blood plasma had been analyzed for more than 2,900 proteins at once, an approach called proteomics. The researchers then looked at who died within 5 and 10 years from non-accident causes, while adjusting for lifestyle and health factors.
The goal was not to find a cause of death. It was to see whether the pattern of proteins in blood carries a signal about future risk that age, sex and lifestyle alone do not capture.
The main findings in plain language
Hundreds of proteins were linked to a higher risk of death, and a handful of them were combined into short panels. The table below summarizes the numbers reported by the authors.
| What was compared | Result reported |
|---|---|
| Proteins linked to higher risk of death within 5 years | 392 |
| Proteins linked to higher risk of death within 10 years | 377 |
| Proteins shared across cardiovascular, cancer and other causes | 19 |
| Size of the final protein panels | 6 proteins (5 years), 10 proteins (10 years) |
| Prediction score (AUC) with age, sex and lifestyle only | 0.49 to 0.57 |
| Prediction score (AUC) after adding the protein panels | 0.62 to 0.68 |
The panels included markers such as adrenomedullin, SERPINA1 and PLAUR. These are proteins involved in processes like blood vessel regulation and inflammation, but the study does not show that they cause anything.
How to read a score of 0.65
An AUC of 0.5 is no better than a coin toss, and 1.0 would be a perfect prediction. A move from about 0.5 to about 0.65 is a real improvement over chance, yet still far from a tool that could tell one person how long they will live. The authors themselves describe the gain as modest.
Latest scientific advances: where this study fits
This work belongs to a fast-growing field. A 2024 study in Nature Medicine built a proteomic age clock from 204 plasma proteins in 45,441 UK Biobank participants. It predicted chronological age closely and was associated with 18 major chronic diseases and with all-cause mortality. The clock held up in smaller groups from China (3,977 people) and Finland (1,990 people).
Another 2024 study in Nature Aging analyzed 1,468 proteins in about 47,600 participants. It found 963 proteins associated with 21 diseases and reported that protein-based scores carried predictive value up to a decade before diagnosis.
What this changes for you: nothing in your next check-up. What it suggests is that, in the years ahead, protein patterns may help doctors flag people who deserve closer follow-up. That would still be a tool for clinicians, used alongside standard tests and a consultation.
What this study cannot tell you
Three limits are worth keeping in mind. The design is observational, so it shows associations and cannot establish causality. The panels are small and come from one large cohort. And the authors state that external validation and cost-effectiveness analyses are still needed before any clinical use.
There is also a risk of over-reading. A protein level linked to higher risk across thousands of people does not mean that a given individual will die earlier. Risk categories are not certainties.
What you can do today with your own lab results
The markers that already exist in routine care are well understood and easy to obtain. Kidney function, blood sugar, cholesterol and inflammation are the usual starting points, and trends over time often say more than one isolated value.
To learn the basics, start by reading a simple guide to blood test results.
To understand why a change between two tests can matter, keep in mind the trend of your blood test results over time.
To see how the same idea applies to aging, explore the difference between biological and chronological age.
MedlinePlus, the US National Library of Medicine’s health site, also offers a plain-language explanation of how to understand your lab results.
When to talk to your doctor
- A result falls outside its reference range and you are unsure what it means.
- You have persistent symptoms such as unexplained fatigue, weight loss or shortness of breath.
- You have a family history of heart, kidney or metabolic disease and have not had a recent check-up.
Glossary
- Proteomics: the large-scale study of the proteins present in a sample such as blood.
- Plasma: the liquid part of blood, where circulating proteins are measured.
- Biobank: a large collection of samples and health data from volunteers, followed over many years.
- AUC: a score from 0.5 (chance) to 1.0 (perfect) that summarizes how well a model separates people who will and will not have an outcome.
- Biomarker: a measurable substance, such as a protein, that reflects a state of health or disease.
Frequently asked questions
Can a blood test predict how long I will live?
No. This study shows that protein patterns carry a modest signal about groups of people, not a reliable forecast for one person.
Can I get this protein panel at my lab?
No. The panels were built from research-grade measurements and still need validation in other populations before any clinical use.
What are adrenomedullin, SERPINA1 and PLAUR?
They are proteins involved in blood vessel regulation, inflammation control and tissue remodeling. Their presence in the panels reflects an association, not proven cause and effect.
Which routine blood tests are worth following?
Standard markers such as kidney function, blood sugar, cholesterol and inflammation are established tools. You can explore the kidney function panel or the lipid panel explained to see what each result means.
Sources
- Koziar N, Whetton AD, Geifman N. A plasma-based protein signature association with all-cause mortality. PLOS ONE, 2025
- PubMed record of the same study (PMID 41270070)
- Argentieri MA et al. Proteomic aging clock predicts mortality and risk of common age-related diseases in diverse populations. Nature Medicine, 2024
- Blood protein assessment of leading incident diseases and mortality in the UK Biobank. Nature Aging, 2024
- MedlinePlus, How to Understand Your Lab Results
More on AI DiagMe
- CRP blood test: read your inflammation results
- Muscle loss, eGFR and the cystatin C test
- C3 complement blood test: aging and inflammation
Understand your lab results with AI DiagMe
