Medicine has been breaking broad diagnoses into narrower subtypes, as it did with breast cancer, so that treatment can follow the biology rather than the label. Fatty liver disease, one of the most widespread chronic conditions on the planet, may be next in line.
Researchers at Mayo Clinic, working with Virginia Tech, report that metabolic dysfunction-associated steatotic liver disease (MASLD), the condition formerly called nonalcoholic fatty liver disease, falls into five biologically distinct subtypes. MASLD affects nearly 30% of adults worldwide. The paper appeared in Nature Communications on 2 September 2026 and has been widely shared by science news sites this month.
The team analysed more than 4,600 patients diagnosed with MASLD, combining genetic sequencing with electronic health records: liver enzymes, body mass index, blood lipids and other conditions such as diabetes, depression and sleep apnea. The records came from Mayo Clinic's Research Data Atlas, which includes the Tapestry Study and its exome data on more than 100,000 participants. Computer models then sorted patients into groups that shared genetic and clinical features.
Some of the five groups are closely tied to obesity and diabetes. Others are driven mainly by inherited genetic differences. The subtypes carried different risks of heart disease, liver failure, liver cancer and the need for a transplant, and the genetic forms were linked to a higher chance of advanced liver damage even in people without the usual metabolic risk factors.
The analysis also found, for the first time, associations between certain subtypes and depression, sleep apnea and migraine. These are statistical links; they do not show that fatty liver disease causes those conditions.
"Once you separate these subtypes, you can start to match treatments to the biology that's actually driving the disease," said Shulan Tian, Ph.D., a Mayo Clinic bioinformatician and co-senior author with Eric Klee, Ph.D.
The subtypes are a research finding, not yet a clinical test. The team plans to check them in larger and more diverse patient groups and to study whether each subtype responds differently to existing treatments, including GLP-1 receptor agonists.
