According to a research study published in the European Journal of Cancer, cardiovascular health calculators can be repurposed to predict a person’s odds of developing cancer. A few small calibration tweaks allowed statisticians to use widely used heart health prediction tools QRISK-3, SCORE-2, SCORE-2OP and pooled cohort equations to highlight people in danger of developing various types of cancer.
In fact, the cardiovascular health calculators performed just as well as existing cancer risk calculators in predicting bowel, prostate and breast cancer. Surprisingly, the tests were also better at spotting people on their way to developing kidney, laryngeal (throat) and gastro-oesophageal (stomach and gullet) cancer than at flagging folk at risk of heart problems.
Testing the Tests
How did they figure out that a cardiovascular health calculator predicts cancer risk?
Researchers at University College London tested the QRISK-3, SCORE-2, SCORE-2OP and pooled cohort equations against population data gathered from the UK biobank to see how well cardiovascular prediction tools performed versus cancer diagnoses. They discovered that key information about an individual’s current health and lifestyle can predict their likelihood of developing one or more of six cancers within ten years.
The investigators validated their prediction tool using a second database, the Clinical Practice Research Datalink, to make sure that the patterns held true in different conditions. Overall the researchers used data from over five million participants to train and test the prediction tools. Using training data from the UK Biobank, the team searched for people who scored as high risk for coronary heart disease and stroke on the QRISK-3, SCORE-2, SCORE-2OP and pooled cohort equations and were diagnosed with cancer. They found an association between the results of the cardiovascular health calculators and six cancers: bowel, prostate, breast, kidney, throat and stomach/gullet cancer.
The team also looked at whether the cardiovascular health calculators predicted risk for cancers of the liver and biliary tree, ovarian, uterine, melanoma, brain and other intracranial cancers, oral, bladder, blood cancer and lung. They did not spot a useful correlation between cardiovascular risk scores and the odds of developing any of these cancers, using these tools.
The during the testing phase, the models consistently identified people as at risk of bowel, prostate, breast, kidney, throat and stomach/gullet cancer within the data sets who would later go on to be diagnosed with these cancers.
Shared Risk Factors for Cancer and Cardiovascular Disease
Much of the cross over between cardiovascular health and cancer risk was driven by smoking and high blood pressure, with smoking a known factor in both conditions. While high blood pressure is probably not a cause of cancer in itself, it is triggered by a set of lifestyle and health factors, including lack of exercise and excessive alcohol. Other key predictors included total cholesterol to high-density lipoprotein cholesterol ratio, the Townsend Deprivation Index and family history of cardiovascular disease.
While these tests don’t predict whether a person will definitely develop bowel, prostate, breast, kidney, laryngeal or gastro-oesophageal cancer, they are a useful indicator that a person might need to change things up a bit.
Cardiovascular Health Calculators and Cancer
Lead researcher on the project, Dr Sam Quill, said in a press release for UCL, ‘In the UK, we currently offer statins to people at higher risk of heart disease using the QRISK3 tool. Our study suggests the risk of certain cancers could be estimated at the same time. However, the value of these tools in cancer prevention needs to be weighed carefully against wider efforts to reduce cancer risk for everyone. More research is needed to explore which approach and which interventions work best.’
Source
Quill S, Hingorani AD, Chaturvedi N, Schmidt AF. Repurposing cardiovascular disease prediction models for cancer. European Journal of Cancer. 2026;243. doi:10.1016/j.ejca.2026.116848
