MANY MINDEDA blood test that reads six heart risks well before symptoms
Researchers at the University of Hong Kong's medical faculty have built an AI tool, called CardiOmicScore, that uses a single blood test to estimate a person's future cardiovascular risk. It covers six major conditions — coronary artery disease, stroke, heart failure, atrial fibrillation, peripheral artery disease, and venous thromboembolism — and, in people at elevated risk, picked up warning signals up to 15 years before disease appeared. The work was published in Nature Communications.
The method uses deep learning across several layers of biology, combining genomics, proteomics, and metabolomics. It drew on 2,920 circulating proteins and 168 metabolites measured in blood samples from the UK Biobank, and performed substantially better than conventional polygenic risk scores, improving further when age and sex were added. Crucially, unlike genetic scores fixed at birth, it is designed to reflect changes from diet, exercise, aging, illness, and environmental exposure.
Cardiovascular disease is the world's leading cause of death, tied to roughly 19.8 million deaths in 2022, and most of that burden builds silently over decades. A cheap blood test that flags risk years ahead — and that responds to how someone actually lives — could move care from expensive late-stage treatment toward earlier, lower-cost prevention, which is where the biggest health gains usually sit.
The caveats matter. The team frames this as a tool that may help predict disease and possibly shift its course, not an established clinical test, and the results rest on UK Biobank population data that skews toward one region and demographic. Any benefit also depends on what follows the score — the monitoring, lifestyle changes, or preventive treatment it is meant to prompt. Watch for validation in other populations.
Source: ScienceDaily