Study: A Routine Blood Pressure Reading May Also Predict Cancer Risk

Heart Risk Calculators Show Surprising Skill at Spotting Cancer

Cancer and heart disease share more common ground than most people realize. Both rank among the world’s biggest killers, and both are shaped by the same habits: smoking, inactivity, a poor diet. Now a new study suggests something surprising: the calculators doctors already use to size up someone’s heart disease risk, once mathematically adjusted, can also flag who is more likely to develop cancer. The comparison holds for several cancer types, though not all.

Researchers recalibrated four widely used heart risk calculators using cancer outcomes, then tested how well the adjusted versions predicted a person’s chances of developing cancer over the next 10 years. Across several specific cancer types, the adjusted heart-risk tools performed with accuracy approaching a dedicated cancer model built specifically for that purpose. For lung cancer, the dedicated tool outperformed the recalibrated heart models overall, though among people with no smoking history, both approaches performed about the same.

This European Journal of Cancer study runs on a simple idea: rather than building an entirely new system to flag cancer risk, doctors could get double duty out of infrastructure already sitting in clinics and health records. That’s the appeal. The catch is that the study only measured how accurately these scores rank risk on paper. Nobody has yet tested whether handing a patient or doctor one of these numbers actually changes what happens next.

Four Heart Calculators Were Tested Against More Than 5 Million Patient Records

Researchers evaluated four heart calculators: QRISK3, the Pooled Cohort Equations, SCORE2, and SCORE2-OP, all used across the UK, Europe, and the United States to estimate heart attack or stroke risk. To test cancer prediction, the team drew on the UK Biobank, covering more than 500,000 participants recruited between 2006 and 2010 with an average age of 56.6, then checked results against the Clinical Practice Research Datalink, containing records for nearly 4.9 million more people.

Over up to 10 years of follow-up, cancer developed in 53,821 UK Biobank participants and 752,661 CPRD participants; cardiovascular disease developed in 23,660 and 244,799, respectively. More participants developed cancer than cardiovascular disease in both datasets, consistent with research showing cancer has overtaken cardiovascular disease as a leading cause of death in many high-income countries over the past few decades. The team fine-tuned the heart models using 20 percent of the UK Biobank data, and the adjusted versions showed strong calibration between predicted and observed cancer risk.

Repurposed Heart Tools Nearly Matched a Dedicated Cancer Model’s Accuracy

Prediction accuracy is often measured with a “c-statistic,” where 0.5 means guessing and 1.0 means perfect prediction. The heart tools scored between 0.71 and 0.74 for their original purpose. For all cancers combined, the repurposed tools scored 0.63, versus 0.65 for the dedicated cancer tool, QCancer. That overall performance was modest: the models ranked risk somewhat better than chance but were far from reliably determining which individual patients would develop cancer.

For certain cancers, including gastro-oesophageal, liver and bile duct, voice box, and renal tract cancers, the heart tools reached accuracy between 0.70 and 0.81. Testing against the separate CPRD dataset produced similar results, suggesting the findings held up reasonably well across two large UK datasets, though nobody has tested how the approach performs outside the UK.

Blood Pressure and Smoking Top the List of Cancer Predictors

Analyzing QRISK3, the most widely used calculator in the UK, researchers found that after age, smoking status and systolic blood pressure were the most influential predictors across nearly all cancer types. Smoking mattered most for voice box and lung cancers but less for prostate, uterine, ovarian, and skin cancers. Cholesterol, socioeconomic deprivation, and family history of cardiovascular disease also showed broad importance.

Researchers cautioned that these associations do not mean high blood pressure causes cancer; clinical trials have not shown that lowering blood pressure reduces cancer risk. Blood pressure may partly reflect other health and lifestyle factors shared by heart disease and cancer, such as inactivity and alcohol use, although this study could not determine the reason.

Existing Health Records Could Ease Adoption, But Questions Remain

QRISK3 is already built into UK electronic health records and used in the NHS Health Check, a national program for adults 40 to 74. In theory, that infrastructure could piggyback cancer-risk alerts onto data doctors are already collecting, no extra tests or appointments required. Researchers also made the adjusted models freely available through a public web application, making it easier for other scientists to examine and test the approach further.

Because the analysis covered adults aged 40 to 84, it says nothing about risk in younger or older people. Some variables, including cholesterol, BMI, and smoking history, were partly missing in the CPRD dataset, though statistical gap-filling did not substantially change results.

Across more than 5 million participants and nearly 800,000 cancer cases, this work suggests the same math doctors already trust for heart risk can flag cancer risk too. What isn’t clear yet is whether acting on that number actually catches cancer earlier or saves lives.

---

Disclaimer: This article is based on peer-reviewed research and is intended for general informational purposes. It is not medical advice. Anyone with questions about their personal cancer or heart disease risk should speak with a qualified healthcare provider.

---