Many people take medicine for high blood pressure for years without knowing there may be an underlying cause that can be treated.
New research suggests artificial intelligence could help change that by spotting patients who should be tested for a hidden hormone disorder called primary aldosteronism.
The study was presented at ENDO 2026, the Endocrine Society’s annual meeting. Researchers from Mayo Clinic created an AI program that searched through decades of electronic health records to identify patients who might have the condition long before it was officially diagnosed.
Primary aldosteronism develops when the adrenal glands produce too much aldosterone, a hormone that controls salt and potassium levels.
Too much aldosterone raises blood pressure and increases the risk of stroke, coronary artery disease, atrial fibrillation, heart failure and kidney disease. Experts estimate that up to 20% of people with high blood pressure may have this condition, but many are never screened.
To build the AI model, researchers analyzed anonymous medical records collected between 1986 and 2025. The program learned from common clinical information, including blood pressure readings, potassium blood tests, medications, age, sex and diagnosis codes. It was then tested on more than 225,000 adults with hypertension.
The AI successfully identified most patients who were later found to have primary aldosteronism. It predicted risk as early as one year before diagnosis. At the chosen setting, the model detected more than 90% of true cases while selecting about two out of every three patients for additional testing.
One advantage of this approach is that it uses information already stored in electronic medical records. Hospitals would not need expensive new equipment or complicated tests simply to decide who should be screened.
Even so, the results should be interpreted carefully. The model is designed as a screening tool rather than a diagnostic test. People identified by the AI would still need further laboratory testing to confirm the diagnosis. In addition, the findings were presented at a scientific conference and should be confirmed in future studies involving other healthcare systems.
Overall, this research highlights how AI could help doctors find a treatable cause of high blood pressure that is often overlooked. Earlier diagnosis could reduce complications, improve patient care and lower healthcare costs if the technology performs well in future real-world studies.
If you care about blood pressure, please read studies about unhealthy habits that could increase high blood pressure risk, and people with severe high blood pressure should reduce coffee intake.
For more information about blood pressure, please see recent studies that early time-restricted eating could help improve blood pressure, and results showing plant-based foods could benefit people with high blood pressure.
Source: Mayo Clinic.
