Nature study of 101,243 scans links brown fat to lower diabetes and heart disease odds

What the study found

Rockefeller University researchers analysed 18F-FDG PET-CT scans from 101,243 adults taken between June 2009 and July 2024, using an AI system to find brown fat, according to the Nature paper [9].

Detectable brown fat turned up in 8,709 patients [9]. Those patients had 58% lower odds of type 2 diabetes and 47% lower odds of chronic ischaemic heart disease, and were less likely to have anxiety or depression [9][7]. More active brown fat meant a bigger apparent benefit [1]. The dataset is roughly double the lab's 2021 analysis of about 50,000 patients [10]. The design is retrospective, so the paper shows association, not cause [1][7].

Brown fat burns energy as heat instead of storing it, which is why it has been chased as a metabolic target for years [2][11]. Until the mid-2000s, adults were thought to lose the tissue after infancy; three New England Journal of Medicine papers in 2009 showed it persists and responds to cold [2]. The team also reported better glycemic parameters, lipid profiles and echocardiographic indices in the brown fat group, along with lower inflammatory markers [10]. Higher activity in one depot, the armpit, was linked to lower odds of high blood pressure [10].

The losartan signal

Patients taking angiotensin receptor blockers were more likely to have detectable brown fat on their scans [10][11].

Losartan, an ARB with a well-known safety profile, then raised metabolic activity in cultured brown fat cells and in mice, which burned more energy and consumed more oxygen [9][11]. Treated mice also produced more carbon dioxide, with no significant difference in body weight, fat mass or physical activity [11]. The effect ran through molecular pathways distinct from the ones that lower blood pressure [1][10]. "The fact that its different effects are mediated by different molecular pathways is advantageous, since it means we could potentially isolate them in future therapies," said Nicolás Gómez-Banoy, an endocrinologist at Rockefeller University [1]. Isolating them would in principle let someone stimulate brown fat without touching blood pressure [1].

The study did not show the drug raises metabolism or causes weight loss in humans, and losartan remains a prescription medicine for approved uses [11].

How the data was built

The work came from Rockefeller University's Laboratory of Molecular Metabolism, led by Paul Cohen, with Memorial Sloan Kettering Cancer Center researchers Heiko Schöder and Daniel Lafontaine [10]. A machine learning tool built at the cancer centre quantified activity across six depots: supraclavicular, cervical, axillary, mediastinal, paraspinal and abdominal [9][10]. Brown fat positive patients were matched against controls of similar age, sex, BMI and outdoor ambient temperature using propensity scores [9]. The new analysis screened more than 500 diagnostic codes, against a handful in the 2021 study [10]. "The scope of this new analysis is far more expansive," Cohen said [10].

The caveats

The population is narrow, and that limits how far the results travel. Every participant had been diagnosed with cancer, and more than three-quarters were women [1]. Mice are also far more reliant on brown fat than people. In small mammals the tissue can reach 10 per cent of body weight, while human deposits shrink with age and are often dormant [1][2]. The mental health result is the least explained of the set. A co-author suggested that people with brown fat have a healthier metabolic profile and therefore fewer psychiatric problems, but the mechanism is still under investigation [7][17].

"More work is needed to prove that brown fat is indeed causing all these benefits," said Irfan Lodhi [7][8]. He is a molecular biologist at Washington University in St. Louis.

What happens next

Finding practical ways to raise brown fat activity is also unsolved: cold exposure is not always practical and existing activating drugs carry side effects [7]. The study's pharmacological profiling identified losartan as a candidate for further investigation, but its effect on human brown fat has not been tested in clinical trials [9]. Cohen framed the dataset as a screening platform. "The dream is that we can use these kinds of data as a platform for not only learning new biology, but maybe repurposing already-approved drugs to activate brown fat in humans," he said [7][8].