New AI blood test predicts heart disease 15 years early

The system, known as CardiOmicScore, uses information from a single blood test to estimate a person's future risk of six major cardiovascular diseases (CVDs): coronary artery disease, stroke, heart failure, atrial fibrillation, peripheral artery disease and venous thromboembolism. In people at elevated risk, the model was able to detect warning signals as far as 15 years before clinical onset.

The findings were published in Nature Communications.

A Blood Test That Captures Current Health

Cardiovascular diseases remain the leading cause of death worldwide, accounting for approximately 19.8 million fatalities in 2022 alone.

Doctors commonly assess cardiovascular risk by examining factors such as age, blood pressure, smoking history and other standard clinical measurements. These indicators are useful, but they may not reveal the earliest biological changes taking place inside the body before a disease becomes apparent.

As a result, some people may not be identified as high risk until the best opportunity for prevention has already begun to narrow.

Genetic risk tests offer another way to estimate a person's likelihood of developing disease. Polygenic risk scores, for example, combine the effects of many genetic variants into a single measure of inherited risk. However, a person's genetic makeup is largely fixed at birth.

That means genetic scores cannot fully reflect more immediate changes caused by diet, exercise, aging, illness, environmental exposures or other influences on health.

CardiOmicScore was designed to provide a more current picture of what is happening inside the body.

AI Combines Thousands of Biological Signals

To build the tool, the HKUMed team used deep learning to combine several layers of biological information. This approach is known as multiomics because it brings together data from different areas of biology, including genomics, metabolomics and proteomics.

Genomics examines genetic information. Proteomics focuses on proteins, which carry out many essential functions in the body. Metabolomics studies small molecules called metabolites, which are produced as the body processes food, generates energy and responds to disease.

The researchers analyzed large scale population data from the UK Biobank. Their model examined 2,920 circulating proteins and 168 metabolites measured in blood samples.

Together, these molecules can provide a detailed snapshot of a person's current biological state. They may reflect subtle changes in immune activity, metabolism and vascular health before noticeable symptoms develop.

Professor Zhang Qingpeng, Associate Professor in the Department of Pharmacology and Pharmacy at HKUMed, explained, "Genes determine where we start -- they define our baseline health risk. However, proteins and metabolites reflect our current physical health. Our AI tool is designed to decode these complex molecular signals, enabling doctors and patients to identify risks much earlier, which can potentially change the trajectory of disease through timely lifestyle modifications and early prevention."

Predicting Six Cardiovascular Diseases

The results showed that CardiOmicScore could turn complex molecular measurements into personalized estimates of cardiovascular risk.

The system performed substantially better than conventional polygenic risk scores. Its accuracy improved further when researchers added clinical information such as age and gender.

The model was designed to assess the risk of coronary artery disease, stroke, heart failure, atrial fibrillation, peripheral artery disease and venous thromboembolism.

Atrial fibrillation is an irregular heartbeat that can increase the risk of stroke and other complications. Peripheral artery disease occurs when narrowed blood vessels reduce circulation to the limbs. Venous thromboembolism refers to dangerous blood clots that form in a vein and may travel to the lungs.

Among high risk individuals, CardiOmicScore could flag elevated cardiovascular risk up to 15 years before symptoms emerged.

Moving From Treatment to Earlier Prevention

The research reflects a broader shift in precision medicine.

Traditional genetic approaches provide a relatively fixed estimate of inherited risk. Multiomics tools may offer a more dynamic assessment by tracking biological signals that change over time.

In the future, a small blood sample could potentially be used to create a detailed risk profile covering several cardiovascular diseases at once. That information could give patients and doctors more time to respond with lifestyle changes, closer monitoring or other preventive measures.

Professor Zhang added, "We aim to leverage technology to identify and prevent diseases before they develop. By shifting health management from reactive treatment to proactive prediction and intervention, we aim to create a lasting impact for both public health and individual patient care."

About the Research Team

The study was led by Professor Zhang Qingpeng, Associate Professor in the Department of Pharmacology and Pharmacy, HKUMed, and the HKU Musketeers Foundation Institute of Data Science (IDS).

The first author is Luo Yan from the HKU IDS.