Hepatic steatosis: what is proven in treatment and what is merely sold

Hepatic steatosis is found by accident more often than it is looked for: someone has an abdominal ultrasound for an unrelated reason, and the report comes back with a line about increased liver echogenicity. What usually follows is one of two things. Either "nothing to worry about, lose some weight" — and that is the end of it. Or a list of "liver support" products, none of which is approved for this indication.

Here is what is known about this condition today: what it is now called, which feature is linked to prognosis, how stage is measured, and what has actually been tested — with a mandatory note on exactly who it was tested in.

The name changed, and it is not cosmetic

In 2023 the professional societies agreed on new nomenclature (Rinella et al., Journal of Hepatology and Hepatology, 2023 — a multi-society Delphi consensus of AASLD, EASL and ALEH). NAFLD — non-alcoholic fatty liver disease — was replaced by MASLD, metabolic dysfunction-associated steatotic liver disease; NASH became MASH. SLD became the umbrella term for fat accumulation in the liver from any cause.

The point of the change is the logic of the diagnosis. The old name rested on a negation: "non-alcoholic". The new one requires a positive feature: hepatic steatosis plus at least one of five cardiometabolic criteria — waist circumference above a threshold, impaired glycaemia, elevated blood pressure, elevated triglycerides, low HDL. The diagnosis stopped being a diagnosis of exclusion and became a statement about metabolic state.

The practical consequence: an older label in a report does not mean a different disease. But under the new name the search is not confined to the liver — it covers the whole cardiometabolic set, including the lipid panel.

What is linked to prognosis is fibrosis stage, not the amount of fat

The most common misconception here is that severity tracks the amount of fat. The data say otherwise.

In a retrospective cohort of 619 adults with liver biopsy and a median follow-up of 12.6 years, the histological feature associated with the outcome — death or liver transplantation — was fibrosis stage, not the degree of steatosis and not the presence of steatohepatitis as such (Angulo et al., Gastroenterology, 2015). The population matters: these were adults who had already undergone biopsy, which is not the average person with an incidental ultrasound finding.

The word "associated" is deliberate. This is an observational cohort: it shows that fibrosis stage travels with the outcome, not that it causes it.

Hence the current management logic. The question is not "is there fat" but "is there fibrosis, and at what stage". Steatosis without fibrosis and steatosis with advanced fibrosis are two different clinical stories carrying the same word in the report.

How it is measured — and why a normal ALT closes nothing

Normal liver enzymes exclude neither steatosis nor advanced fibrosis. In a cohort of adults with biopsy, among those with both ALT and AST below 40, steatohepatitis with intermediate-stage fibrosis was found in 19% and cirrhosis in 7% (Gawrieh et al., American Journal of Gastroenterology, 2019). In children this is stated outright in guidance: a normal ALT excludes neither steatosis nor severe fibrosis (ESPGHAN, 2012).

What is used instead:

• Calculated indices. FIB-4 is derived from age, ALT, AST and platelet count — four values already in routine bloodwork. A first-step tool: it rules out a high probability of advanced fibrosis but does not assign a stage.

• Elastography. Transient elastography measures liver stiffness; the CAP parameter on the same device estimates fat content.

• MRI. MRI-PDFF quantifies the fat fraction — not fibrosis — and serves as an endpoint in early-phase trials.

• Biopsy remains the reference standard for staging, but is used selectively.

🔴 A separate note on age, because this is where the error is most common. Below the age of 35 FIB-4 performs poorly: in a validation study the area under the curve in that age group was 0.60 and the sensitivity of the lower threshold was 0%, meaning the index produced false negatives (McPherson et al., 2017). For that reason the AASLD 2023 guidance calls for a secondary assessment under 35 — elastography or an ELF test — rather than discarding the index. Above 65 the problem is the opposite: false positives increase and the threshold is raised. In children and adolescents FIB-4 is not validated at all.

What has actually been tested — and in whom

Weight reduction — the best-studied line

The best-known work is a prospective cohort of 293 adults with biopsy-confirmed steatohepatitis, with paired biopsies in 261 of them before and after 52 weeks of lifestyle intervention (Vilar-Gómez et al., Gastroenterology, 2015). The findings tracked the size of the weight reduction: with a reduction of 5% or more, resolution of steatohepatitis was recorded in 58%; with 10% or more, in 90%, and regression of fibrosis in that same group in 45%.

Two limits, without which those numbers read wrong. First, the study was not randomised. This is a relationship observed within a cohort, not a demonstrated effect: people who achieved a larger reduction may have differed in other ways too. Second, the proportion reaching the largest reduction was small — so 90% describes the outcome inside that small group, not what usually happens.

The stepped thresholds that guidelines are built around come from exactly this evidence and are addressed to adults: the EASL–EASD–EASO 2024 guidelines name 5% for liver fat, 7–10% for inflammation and 10% or more for fibrosis; AASLD 2023 uses a different split — 3–5% for steatosis and more than 10% for steatohepatitis and fibrosis. No such histological targets have been set by the professional societies for children and adolescents.

Physical activity

Regular aerobic exercise is associated with reduced liver fat content, including without meaningful weight change: a meta-analysis of ten randomised trials in 316 adults found a reduction in intrahepatic lipid with a standardised mean difference of −0.76 (Babu et al., Nutrients, 2021). In a separate randomised trial with weight held stable, fat content fell while markers of inflammation and fibrosis did not change (Houghton et al., 2017).

So the endpoint here is fat on imaging, in adults. An effect on fibrosis stage or on histology has not been shown for exercise in the available data.

Bariatric surgery

In a prospective cohort of adults with morbid obesity, 64 patients underwent repeat biopsy five years after surgery: resolution of steatohepatitis without worsening of fibrosis was recorded in 84%, and improvement of fibrosis by at least one stage in 70% (Lassailly et al., Gastroenterology, 2020). This too was not a randomised trial, and the population is narrow — people for whom surgery was indicated by weight.

Drugs

Until 2024 no drug was approved for this indication and everything prescribed was off-label. In two years that changed — for adults.

Resmetirom was approved by the FDA on 14 March 2024 for adults with MASH and fibrosis stages F2–F3, and this was an accelerated approval: granted on a surrogate histological endpoint from the MAESTRO-NASH trial (Harrison et al., NEJM, 2024), with confirmation of an effect on clinical outcomes still ongoing and expected at month 54. In the European Union it received a conditional marketing authorisation on 18 August 2025. Accelerated and conditional approvals are not the same as a demonstrated effect on outcomes: they are permission to use while verification continues.

Semaglutide. In the phase 3 ESSENCE trial (Newsome et al., NEJM, 2025) in adults with MASH and F2–F3 fibrosis, resolution of steatohepatitis without worsening of fibrosis at week 72 was 62.9% versus 34.3% on placebo, and improvement in fibrosis without worsening of steatohepatitis was 36.8% versus 22.4%. The FDA approved semaglutide for MASH on 15 August 2025, also under accelerated approval and for adults only; the corresponding European authorisation followed in March 2026.

Vitamin E and pioglitazone were tested in the randomised PIVENS trial (Sanyal et al., NEJM, 2010) in 247 adults without diabetes over 96 weeks. Vitamin E at 800 IU daily improved the histological picture of steatohepatitis — 43% versus 19% on placebo — while fibrosis did not change. Pioglitazone did not meet the primary endpoint: a difference was present but fell short of the pre-set significance threshold, and that group gained weight. This is not "a liver vitamin off the shelf": the dose was high, the population narrow, and that dose carries its own discussion of risk. The decision belongs to a physician.

In children a separate randomised trial, TONIC (Lavine et al., JAMA, 2011) in 173 children aged 8–17, produced a different result: the primary endpoint — sustained reduction in ALT — was not met, by either vitamin E or metformin. On a secondary endpoint, resolution of steatohepatitis occurred more often with vitamin E than with placebo, 58% versus 28%. A trial that misses its primary endpoint supports few claims; but even this one shows that the adult conclusion does not transfer to adolescents automatically.

What is sold but does not survive scrutiny

The "liver support" market is large, and it runs on the fact that a person with an ultrasound finding wants something to take.

• Essential phospholipids — the most recognisable hepatoprotector class on the shelf. They hold no regulatory approval for MASLD/MASH and do not appear in professional society guidelines as a treatment.

• Silymarin (milk thistle). Here it is important not to swap populations. The 2007 Cochrane review usually cited found no convincing effect on outcomes — but it covered alcohol-related liver disease and hepatitis B and C, not MASLD. For steatohepatitis specifically there is a separate randomised trial in 99 patients in which the primary endpoint was not met. AASLD 2023 explicitly recommends against silymarin in steatohepatitis.

• Ursodeoxycholic acid. A randomised placebo-controlled trial in 166 patients over two years showed no difference from placebo in steatosis, necroinflammation or fibrosis (Lindor et al., Hepatology, 2004). AASLD 2023 does not recommend it as treatment.

• "Liver detox", lipotropic drips, cleanses. There is nothing here to test: the endpoint is usually declared to be how you feel.

The common signature of this group is a promised result with no named endpoint. If it is not stated what was measured and in whom, there is nothing to check.

Why an adolescent is a separate conversation

The prevalence of hepatic steatosis in children with obesity is markedly higher than in the general paediatric population: an autopsy study of children and adolescents aged 2–19 found 9.6% overall and 38% among children with obesity (Schwimmer et al., Pediatrics, 2006); a 2015 meta-analysis reported 7.6% in the general population versus 34.2% in obesity clinics.

At the same time no drug is approved by regulators for MASLD/MASH in children and adolescents: both resmetirom and semaglutide are authorised for this indication in adults only. Semaglutide is indeed approved from age 12 — but for obesity, not for the liver, and the two registrations should not be conflated.

Paediatric screening is also built differently from adult screening: NASPGHAN recommends starting at age 9–11 in children with obesity and using ALT as the test rather than routine ultrasound, repeated every two to three years. Lifestyle intervention remains the basis of management — not because it outperforms a drug, but because a drug tested in this population does not exist.

The horizon differs too. In an adult with a long-standing condition the conversation is about not letting what already exists progress; in an adolescent, decades lie ahead over which fibrosis may or may not have time to form.

What to do with all of this

1. Establish the stage, not just the fact. An ultrasound finding answers "is there fat" and says nothing about fibrosis. Next comes a calculated index and, at a physician's discretion, elastography.

2. Do not treat normal enzymes as case closed. An ALT within range does not exclude an advanced stage.

3. Check that the scale matches the age. Adult thresholds do not apply to children and adolescents, and under 35 they require a second assessment.

4. Look wider than the liver. Under the current logic the diagnosis is metabolic: blood pressure, glucose and lipids are part of the same picture.

5. Ask any proposed product for its endpoint. What was measured, in whom, by what design. The question clears most of the shelf.

6. Measure the same way each time. A trend is readable when the device and the laboratory stay the same; otherwise you are comparing methods, not people.

This material is informational and does not replace medical advice. Decisions about investigation and treatment are made with a physician.