📚 Part of our Blood Sugar Guide — an evidence-based guide.
The most common liver disease in the world has no symptoms, no widely used screening programme, and until recently no good name. Roughly one in four adults globally has fat accumulating in the liver. Most of them do not know, and most will never be told.
It used to be called non-alcoholic fatty liver disease — a name defined by what it is not. In 2023 an international panel renamed it MASLD: metabolic dysfunction–associated steatotic liver disease. The change was not cosmetic. The old name framed the condition as a liver problem that happened to occur in people who do not drink. The new one says what the evidence had been saying for years: this is a metabolic disease that shows up in the liver.
What is actually happening
The liver is not designed to store fat. It processes it, packages it, ships it out. Fatty liver is diagnosed when more than 5% of liver weight is fat — a threshold at which the organ has moved from processing to accumulating.
The main driver in most people is insulin resistance. When muscle and fat tissue stop responding well to insulin, the liver receives more free fatty acids than it can export, and it also increases its own production of fat from carbohydrate through a process called de novo lipogenesis. Fructose in liquid form is a particularly efficient substrate for this, which is part of why sugar-sweetened drinks track so closely with liver fat.
So fatty liver is usually not the first problem. It is a downstream signal that insulin resistance has been running for a while (what insulin resistance is).
The part that matters: most cases do not progress
This is where honest reporting diverges from alarm.
Simple fat accumulation — steatosis alone — is largely benign for the liver itself. Most people with it will never develop cirrhosis. The condition progresses in a minority: fat plus inflammation and liver-cell injury, now called MASH, which can drive fibrosis, and fibrosis is what eventually threatens the organ.
The distinction is clinically decisive. Fat alone is a metabolic warning. Fibrosis is liver disease. The stage of fibrosis, not the amount of fat, is what predicts liver-related outcomes and death.
And there is a second point that gets lost. Even among people whose liver never progresses, MASLD is associated with elevated cardiovascular risk — and cardiovascular disease, not liver failure, is the leading cause of death in this population. The liver finding is often most useful as a marker of what is happening everywhere else.
How it gets found, and how it gets missed
Most cases surface by accident: an ultrasound ordered for something unrelated, or mildly elevated liver enzymes on a routine panel.
Enzymes are an unreliable guide. ALT and AST are frequently normal in people with significant liver fat, and can even be normal in the presence of fibrosis. A normal liver panel does not exclude the condition.
The more useful question is not “is there fat?” but “is there fibrosis?”. That is assessed with non-invasive tools:
FIB-4 is a calculation from age, ALT, AST and platelet count — numbers that are usually already sitting in your existing blood work. It costs nothing, and it is good at ruling fibrosis out. A low score is genuinely reassuring.
Transient elastography (FibroScan) measures liver stiffness with ultrasound and is the usual next step when FIB-4 is indeterminate or high.
Biopsy remains the reference standard but is now reserved for uncertain cases, because it is invasive and samples a tiny fraction of the organ.
What actually reverses it
Weight loss is the intervention with the most consistent evidence, and it works in a dose-dependent way that is worth stating precisely:
Around 3 to 5% of body weight reduces liver fat. Around 7 to 10% is where inflammation improves in a substantial proportion of people. 10% and above is where fibrosis regression has been observed — less often, but it has been observed. Fibrosis was long assumed to be one-way. It is not always.
Composition matters alongside quantity. A recent trial found a ketogenic diet cut liver fat substantially even when weight loss was matched against a comparison diet, suggesting carbohydrate restriction acts on liver fat partly independently of the scale (what that trial found). The Mediterranean pattern has the broadest supporting evidence for this condition (the evidence behind it).
Two specific targets are well supported: cut sugar-sweetened beverages, and reduce ultra-processed food (what the research shows).
Exercise reduces liver fat even without weight loss — both aerobic and resistance training, in randomised trials. For anyone who struggles with the scale, this is the most encouraging finding in the field.
Coffee is one of the more consistent observational associations with lower fibrosis risk in liver disease. The evidence is not randomised, and the effect size is modest — but the direction has been stable across many cohorts (the wider coffee data).
On drugs: resmetirom became the first FDA-approved therapy for MASH with moderate-to-advanced fibrosis in 2024, and GLP-1 receptor agonists have shown meaningful reductions in liver fat and MASH resolution in trials (their broader effects). These are treatments for specific stages of disease, prescribed after staging — not something to seek because an ultrasound mentioned fat.
What does not work
Liver “detox” and cleanse products have no evidence for reducing liver fat or fibrosis, and some botanical supplements have themselves caused liver injury. Milk thistle has been studied and has not shown convincing benefit on hard outcomes.
Alcohol deserves a direct word. The renaming to MASLD reflects that metabolic dysfunction and alcohol can both contribute, and the categories overlap more than the old “non-alcoholic” label implied. If you have liver fat, alcohol is adding a second injury to the same organ.
When to see a doctor
Ask about liver fat if you have type 2 diabetes, obesity, or metabolic syndrome (the five numbers) — these groups carry high prevalence and are where screening has the clearest rationale. If liver fat has been mentioned on any scan, ask specifically for a FIB-4 score: it can usually be calculated from blood work you have already had.
Seek prompt care for yellowing of the eyes or skin, abdominal swelling, vomiting blood, black tarry stools, or new confusion. These are signs of advanced disease, not early fatty liver.
The honest bottom line
MASLD is common, usually silent, and in most people will not destroy the liver. But it is a reliable signal that metabolic dysfunction is established, and the cardiovascular risk that accompanies it is more likely to matter than the liver itself. The question worth asking is not whether there is fat — it is whether there is fibrosis, and a FIB-4 score answers that cheaply. Modest weight loss, less liquid sugar, and exercise are the interventions with real evidence, and they act on the underlying disease rather than the shadow it casts on an ultrasound.
Frequently asked questions
What is the difference between MASLD and NAFLD?
They describe the same condition. NAFLD (non-alcoholic fatty liver disease) defined the disease by what it was not. In 2023 an international panel renamed it MASLD — metabolic dysfunction–associated steatotic liver disease — to reflect that it is a metabolic condition expressed in the liver, and to stop treating alcohol and metabolic causes as mutually exclusive.
Can fatty liver be reversed?
Liver fat itself responds well: around 3 to 5% body weight loss measurably reduces it. Inflammation improves in many people at 7 to 10%. Fibrosis regression has been observed at 10% and above — less reliably, but it does happen, which contradicts the older assumption that fibrosis was one-way. Exercise reduces liver fat even when weight does not change.
Can you have fatty liver with normal liver enzymes?
Yes, and it is common. ALT and AST are frequently normal in people with significant liver fat, and can be normal even in the presence of fibrosis. A normal liver panel does not exclude the condition, which is why imaging or a FIB-4 score is needed to answer the question.
What is a FIB-4 score?
A calculation using age, ALT, AST and platelet count — all values usually present in routine blood work. It estimates the likelihood of advanced fibrosis. It costs nothing extra and is particularly good at ruling fibrosis out, so a low score is genuinely reassuring. It is the single most useful thing to ask for if a scan has mentioned liver fat.
Do liver detox supplements help fatty liver?
No. There is no evidence that detox or cleanse products reduce liver fat or fibrosis, and some botanical supplements have themselves caused liver injury. Milk thistle has been studied and has not shown convincing benefit on hard outcomes.