People whose scans showed active brown fat had 58% lower odds of type 2 diabetes and 47% lower odds of chronic ischaemic heart disease than closely matched people without it. That comes from a new Nature paper that used artificial intelligence to read more than 100,000 clinical scans. It is a striking association, but it is not proof that brown fat protects anyone, and here is what the study does and does not show. This is general information, not medical advice.
What the new brown fat study did
The paper, titled “Artificial intelligence-enabled clinical phenotyping of human brown fat,” was published in Nature in 2026 (DOI 10.1038/s41586-026-11162-7). The first author is Gomez-Banoy N. News-Medical summarized the findings on Oct. 15, 2026.
The team looked back at clinical 18F-FDG PET-CT scans from 101,243 adults. The scans were taken at a cancer center between June 2009 and July 2024. In the part of the analysis that measured metabolic activity, 8,709 patients were classed as brown adipose tissue (BAT) positive.
Those BAT-positive patients were compared with BAT-negative controls. The groups were matched for age, sex, body mass index (BMI) and outdoor temperature, using a statistical method called propensity matching.
Brown fat vs white fat: a plain-language primer
Most body fat is white fat. It stores energy and cushions organs. Brown fat is a different tissue. It burns energy to make heat, which is why it is often linked to staying warm in the cold.
Brown fat cells are packed with mitochondria, the tiny structures that turn fuel into energy. A protein called UCP1 helps these cells release that energy as heat instead of storing it. This is general background, and it is the reason researchers keep studying brown fat as a possible lever on metabolism.
Babies have a good deal of brown fat. For years, scientists debated how much adults keep. Scans like the ones in this study have helped show that many adults still have some, though amounts vary a lot from person to person.
How AI measured brown fat on PET-CT scans
A PET-CT scan combines two images. The PET part tracks where the body takes up a radioactive sugar tracer (FDG), which shows metabolic activity. The CT part shows the anatomy. Brown fat that is active takes up the tracer, so it can light up on the scan.
These scans were ordered for cancer care, not to study brown fat. Reading them by hand at this scale would be impractical. The researchers used AI to semi-automatically measure BAT activity across six areas of the body:
- Supraclavicular (above the collarbone)
- Cervical (neck)
- Axillary (armpit)
- Mediastinal (chest, between the lungs)
- Paraspinal (along the spine)
- Abdominal
Among these, axillary BAT activity was the most strongly tied to favorable cardiometabolic outcomes and to less visceral fat, which is the fat packed around internal organs.
What the brown fat findings show about heart and metabolic health
Compared with matched BAT-negative people, those with brown fat had 58% lower odds of type 2 diabetes and 47% lower odds of chronic ischaemic heart disease, according to the News-Medical summary. They also had lower odds of primary hypertension (high blood pressure), depressive episodes and anxiety disorders. The summary gave no percentages for those three.
The blood and imaging results pointed the same way in several areas:
- Blood sugar and fats: lower glucose, HbA1c and triglycerides, and lower non-HDL cholesterol.
- Cholesterol that went the other direction: higher HDL, total and LDL cholesterol.
- Heart structure: echocardiography showed lower measures of pathological left ventricular remodeling.
- Inflammation and liver: lower inflammatory markers and lower liver enzymes.
- Bone: lower vertebral bone-density scores. The summary did not make clear whether that difference is significant.
The higher LDL and total cholesterol is a reminder that the picture is mixed. Not every marker moved in the “better” direction.
What “lower odds” means
“Odds” is a way of comparing how likely something is in one group versus another. A 58% lower odds of type 2 diabetes means that, after matching, the odds in the brown fat group were 58% lower than in the comparison group.
It does not mean that 58 of every 100 people with brown fat avoided diabetes. It also does not mean an individual’s risk dropped by that amount. Odds ratios describe groups in a particular dataset, here a group of people who all happened to be scanned at one cancer center.
The losartan clue: what mice and cells showed
The team noticed something about medication. Angiotensin receptor blockers, a class of blood pressure drugs, were positively associated with BAT status. The researchers then tested one of them, losartan, in cultured fat cells and in mice.
In those lab experiments, losartan increased UCP1 expression and oxygen consumption. In mice, it increased energy expenditure at room temperature and in the cold, without measurable weight loss. It also activated PPARgamma signaling independent of the classical AT1 receptor.
This is early, preclinical work. The summary states plainly that losartan’s effect on human brown fat has not been tested in trials, and that “its effectiveness for this purpose in humans remains unknown.”
Please do not read this as a reason to ask for losartan or any other drug. Blood pressure medicines are prescribed for specific reasons and need a clinician’s oversight.
Why this brown fat study cannot prove cause and effect
The limits matter as much as the headline numbers. According to the summary, the study:
- Was retrospective, meaning it looked back at existing records rather than following people forward.
- Drew on a cancer-center population, who may differ from the general public.
- Used scans taken without prior cold stimulation, so brown fat that was dormant on the day could have been missed.
- Relied on glucose uptake, which is an indirect measure of brown fat activity.
- Cannot show that brown fat prevents disease. It shows an association.
There is also a simple alternative explanation. People who are metabolically healthier may be more likely to have active brown fat, rather than brown fat making them healthier. Matching on age, sex, BMI and temperature helps, but it cannot remove every hidden difference between groups.
Questions the study leaves open
Several basic questions remain unanswered in the summary we reviewed. It does not say how much active tissue a person needs before any benefit appears, or whether the link holds in people who are not being scanned for cancer.
It also does not tell us whether brown fat activity changes over time within one person, or whether a scan taken in warmer or cooler conditions would have sorted some patients differently. Matching on outdoor temperature was meant to address part of that, but scans without cold stimulation remain a limit.
Finally, the mental health findings deserve extra caution. Lower odds of depressive episodes and anxiety disorders were reported, but with no percentages in the summary, it is hard to judge how large those differences are. Follow-up work in other hospitals and in people followed forward in time would help show which of these links hold up.
What to take from the brown fat research
No study described here tested cold exposure, supplements or any lifestyle change as a way to build or activate brown fat. So there is no evidence from this paper to support doing any of them for your health.
What the research does offer is a better map. Large, AI-assisted reads of routine scans may help scientists work out which kinds of brown fat matter most, and where in the body. The axillary finding is one lead worth following up in prospective studies.
For now, the habits with the strongest evidence for heart and metabolic health are unchanged. Our earlier coverage looks at how many steps per day may help your heart and at how 11 extra minutes of sleep was linked to lower heart risk. If you are weighing diet changes, see our report on a high-fiber diet vs metformin trial.
Practical takeaway: treat this as an interesting early signal, not a to-do list. If you have concerns about diabetes, heart disease, blood pressure or mood, talk with your doctor about screening and about the proven steps for your situation. For the study itself, you can find the journal at nature.com.
