- Design
- Multinational retrospective cohort, biopsy-confirmed
- Population
- 18 326 adults with MASLD from 41 countries
- Primary outcome
- All-cause mortality and liver-related clinical events
- Effect
- Lean vs overweight/obese: aHR 0.90 (0.69–1.16) for death; advanced fibrosis aHR 2.10 for death
A multinational retrospective study in JAMA analysed 18 326 adults with biopsy-confirmed MASLD from 41 countries. About 6.7% were lean by BMI, rising to 11.3% in Asia.
Lean patients had less type 2 diabetes (38.2% vs 48.3%) and less advanced fibrosis (29.3% vs 37.2%) than overweight or obese patients. But after adjustment, lean status was not associated with lower mortality (aHR 0.90, 95% CI 0.69–1.16) or fewer liver events (aHR 0.94, 0.76–1.17). Advanced fibrosis was what tracked prognosis: aHR 2.10 for death and 3.41 for clinical liver events. FIB-4 was slightly less accurate in lean patients (AUC 0.76 vs 0.79), while transient elastography was more accurate (0.87 vs 0.83).
A normal BMI is easily read as reassurance in a patient with fatty liver on ultrasound, particularly in South Asian clinics where lean MASLD is common. This cohort argues against that. The existing pathway — FIB-4 first, then liver stiffness measurement where indicated — still applies; the point is to apply it to lean patients with the same rigour. Patients here were biopsied, so they may be sicker than an average clinic population.
- Calculate FIB-4 in every patient with MASLD and diabetes, whatever their weight.
- Where FIB-4 is indeterminate in a lean patient, consider transient elastography rather than relying on the score alone.
- Record waist circumference as well as BMI; the two classified lean status differently.
- Counsel lean patients that their liver risk depends on fibrosis stage, not on how much they weigh.
Why it matters
It challenges the assumption that lean fatty liver is a benign variant.
Don't overread it
This was observational and biopsy-selected; it shows similar outcomes, not that lean patients need different treatment.
The statistics, in plain English
An adjusted hazard ratio of 0.90 with a confidence interval from 0.69 to 1.16 crosses 1.0, so the data are consistent with lean patients having somewhat lower, equal or higher risk — no clear difference. By contrast, advanced fibrosis doubled the hazard of death, a consistent and large association.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for diabetes & endocrinology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free