Researchers used artificial intelligence to uncover connections between brown fat, cardiovascular health, and other medical conditions—while identifying a commonly prescribed blood-pressure medication as a potential starting point for new treatments.
WHAT’S HAPPENING
Artificial intelligence is helping medical researchers uncover previously overlooked connections between human biology, disease, and existing medications.
In a study published in Nature on October 14, 2026, researchers at Memorial Sloan Kettering Cancer Center analyzed medical information from 101,243 patients, using AI to investigate the potential health benefits of brown adipose tissue, commonly known as brown fat.
Unlike white fat, which primarily stores energy, brown fat burns energy to generate heat and help regulate body temperature.
Researchers have long suspected that brown fat plays a role in metabolic health. However, understanding its effects across large human populations has been difficult.
Using machine learning, scientists measured brown-fat activity in approximately 8,700 patients and compared those measurements with medical diagnoses, laboratory results, and other health indicators.
The findings revealed associations between greater brown-fat activity and lower prevalence of several conditions, including type 2 diabetes, coronary artery disease, heart failure, and certain psychiatric disorders.
Researchers also identified an unexpected possibility: losartan, a widely prescribed blood-pressure medication, may stimulate brown-fat activity.
Laboratory experiments and animal studies provided preliminary support for that discovery, although its clinical benefits in humans remain unproven.
WHY IT MATTERS
This research demonstrates how artificial intelligence can help scientists discover medically significant patterns within information that hospitals already collect.
Traditional medical research frequently begins with a hypothesis that investigators test using a relatively limited population.
AI-assisted analysis offers another approach.
By examining thousands of medical images and records, researchers can identify relationships that might otherwise remain difficult to detect.
In this study, machine learning allowed scientists to measure brown-fat activity in different regions of the body and investigate how those measurements related to specific diseases.
One particularly interesting finding involved brown fat near the armpits, known as the axillary depot.
Higher activity in this region showed strong associations with lower prevalence of several cardiometabolic conditions.
The study also connected brown-fat activity with healthier blood-glucose measurements, lower triglycerides, and certain indicators of improved cardiovascular function.
These findings do not establish that brown fat prevents disease, but they provide researchers with specific biological relationships to investigate.
WHO BENEFITS
Medical researchers: AI provides new ways to investigate complex relationships between metabolism, disease, and biological activity.
Drug developers: Identifying potential new uses for existing medications could help researchers prioritize compounds for further testing.
Healthcare institutions: Large collections of medical imaging and patient records may contain valuable information that AI can help uncover.
Patients with metabolic and cardiovascular conditions: Future research could lead to additional treatment strategies for diabetes, hypertension, and cardiovascular disease.
One important opportunity involves drug repurposing.
Because losartan is already prescribed to treat certain cardiovascular conditions, researchers have substantial existing information about its established medical uses and safety profile.
However, using it specifically to activate brown fat would require additional human clinical research.
WHO LOSES
Patients relying on unproven treatment claims: Misinterpreting early findings could encourage inappropriate medication use or unrealistic expectations.
Researchers without access to large medical datasets: Institutions lacking substantial clinical information and computing resources may struggle to conduct comparable investigations.
Healthcare systems adopting AI without adequate validation: Statistical associations can be misleading when patient characteristics, existing illnesses, and other factors are not carefully considered.
The study’s population also came from a major cancer center, meaning its findings may not apply equally to the general population.
AI can uncover patterns, but those patterns still require scientific interpretation and independent testing.
WHAT HAPPENS NEXT
Researchers now have several avenues for further investigation.
The first is determining whether increasing brown-fat activity directly improves human health or whether brown fat primarily serves as a marker associated with better metabolic function.
The second involves evaluating medications that might safely stimulate brown fat.
Losartan is a promising research candidate, but experiments demonstrating increased brown-fat activity do not establish that the medication can prevent obesity, diabetes, or cardiovascular disease through this mechanism in humans.
Additional clinical trials would be necessary.
Scientists may also apply similar AI methods to other medical imaging datasets, potentially revealing previously unrecognized biological relationships and treatment opportunities.
For healthcare organizations, the practical lesson is that existing patient data may offer substantial research value when analyzed responsibly using advanced technology.
The larger signal: Artificial intelligence is expanding medical discovery beyond designing new drugs or diagnosing diseases. By examining enormous collections of existing clinical information, AI can help researchers identify biological connections and potential new uses for familiar medications. The opportunity is significant, but the ultimate measure of success will be whether those discoveries translate into treatments that demonstrably improve patients’ lives.
