A new study has revealed that proteins in our blood can predict the risk of early death.
Imagine if a simple blood test could offer a glimpse into your future health. Not just whether you have heart disease or cancer today, but whether your overall risk of dying in the next five or ten years is higher or lower than expected.
It’s the kind of idea that has hovered on the margins of medicine for decades, popping up in headlines whenever a new biomarker is discovered. In practice, however, long-term health prediction has remained frustratingly inaccurate.
Doctors continue to rely heavily on age, weight, smoking history, and a small set of routine blood tests, most of which provide only broad population-level estimates.
At the same time, modern medicine is rapidly moving towards early detection and prevention. Health systems around the world are struggling with rising rates of chronic disease and aging populations.
Clinicians increasingly need tools that can identify risks before symptoms appear, allowing for earlier intervention. The question is whether the clues of future health may already be circulating in our blood.
This is what a new study explores, in November at PLOS One.
By measuring thousands of proteins in the blood of tens of thousands of people and tracking who survived or died over time, it was discovered that certain protein patterns appear to be associated with a greater risk of dying from any cause other than accidental.
The analysis used data from more than 38,000 adults aged 39 to 70 who took part in the UK Biobank study. Participants provided blood samples and ongoing, comprehensive health and lifestyle data.
After taking into account risk factors already known to negatively affect life expectancy, such as age, body mass index (BMI) and smoking, Hundreds of proteins have been identified that are associated with the overall probability of dying from any cause, as well as the probability of dying from specific diseases, including cancer and cardiovascular disease.
As the corresponding author of the study details, Nophar Geifman In a paper in , the research team filtered through these long lists to isolate a small number of proteins known as panels of proteins. These panels contained ten proteins associated with the risk of all-cause mortality at ten years, and six proteins associated with the risk at five years.
These panels have improved predictive power over traditional models that rely on age, BMI and lifestyle factors. In statistical terms, models based only on demographic and lifestyle data performed poorly, with accuracy close to chance. Models that incorporated protein panels performed better, although gains continued to be limited.
This suggested that some proteins in the blood may carry hidden signals about long-term health that go beyond current illnesses.
Traditional risk factors such as age, weight, smoking, alcohol consumption and physical activity offer important but often inaccurate clues about declining health. To the Blood proteins, by contrast, provide real-time snapshots of what is happening inside the body.
Some may reflect slow, chronic changes, such as low-grade inflammation, tissue degradation, or subtle organ stress. Others may indicate more immediate risks linked to the heart, blood vessels or the immune system. The new study shows that the risk of dying can also be partially captured in circulating protein levels.
Far from being a perfect test
Still, this is far from a perfect test, researchers warn. Predictive power is better than chance, but remains modest. These protein signatures cannot be treated as definitive indicators of when someone will die. They may, however, with additional validation, act more as a warning that can lead to early action.
Furthermore, the study also focused only on associations. Proteins may not be causing the increased risk. They can simply be markers of underlying biological processes that have not yet produced symptoms.
The authors also note that combining all causes of death into a single outcome makes interpretation difficult. This is because the paths that lead to death vary widely. Heart disease, cancer, infections and organ failure involve very different biological mechanisms.
Even with these caveats, the results point to a future in which routine blood tests may go beyond diagnosing current disease. A simple snapshot can alert doctors that a patient faces a high risk of declining health, even when nothing obvious appears to be wrong. This could trigger earlier actions, such as closer monitoring, lifestyle advice or preventive treatments.
This type of risk stratification is becoming increasingly important as populations age and rates of chronic disease increase, placing increasing pressure on healthcare systems. Such a test could help doctors direct care more effectively.
