The NAFLD Liver Fat Score estimates whether your liver holds more fat than it should, from five inputs — metabolic syndrome yes or no, type 2 diabetes yes or no, fasting insulin, AST, and the AST-to-ALT ratio — so it needs a fasting blood draw including an insulin assay most routine panels leave out. It is a laboratory measurement, not a home one, and no scanner is involved: the score exists as a cheap stand-in for imaging.
A Finnish group published it in 2009, validated against magnetic resonance spectroscopy, the reference way to quantify liver fat without a biopsy. It answers one question: is there probably more fat here than the usual threshold of about 5.5 percent of liver weight?
What you need
- A fasting blood sample, drawn after roughly 8 to 12 hours without food.
- Fasting serum insulin — the awkward input. It is not on a standard panel, and assays differ between laboratories.
- AST (aspartate aminotransferase) and ALT (alanine aminotransferase), both on a routine liver panel from the same draw.
- Whether you have diagnosed type 2 diabetes.
- Whether you meet the criteria for metabolic syndrome.
That last item is a small assessment in itself: metabolic syndrome is defined by waist circumference, blood pressure, fasting glucose, triglycerides and HDL cholesterol together, so if it has never been settled, work through Metabolic syndrome — the hidden pandemic of the 21st century first.
How the measurement works
The score never looks at the liver; it reads the company a fatty liver keeps. Fat accumulates when insulin stops restraining fat delivery and synthesis, so insulin resistance carries most of the signal, and the transaminases add what the liver is leaking.
Each input pushes in a known direction. Metabolic syndrome raises the score; type 2 diabetes raises it further and weighs more than the syndrome does; higher fasting insulin raises it; higher AST raises it. A higher AST-to-ALT ratio lowers it, which surprises people until you see what the ratio does. In simple fatty liver ALT usually sits above AST, so a ratio below one is typical; a ratio above one points toward alcohol-related injury or advanced scarring, problems this score was not built to find.
Two terms pull against each other, insulin must be in the right units, and the coefficients are not memorable. Put your five values into the NAFLD Liver Fat Score (NAFLD-LFS) and let it do the arithmetic.
How to read the result
The output is one continuous number, which can be negative. The published cut-off is minus 0.640: above it, liver fat beyond the spectroscopy threshold is likely; below it, unlikely. There are no further graded bands, so do not read severity into the decimal — a high score is not a measured fat percentage.
The cut-off was set to catch most people who do have a fatty liver, at the cost of flagging some who do not, so the answers are not symmetrical. A score below it is reasonably reassuring; a score above it is a reason to look properly, usually with ultrasound, rather than a diagnosis.
What it does not tell you
It reports fat, not damage. It cannot separate simple steatosis, largely reversible, from steatohepatitis, where inflammation is destroying cells, and it says nothing about fibrosis — the finding that decides prognosis. That needs another tool: the FIB-4 Fibrosis Index asks about scarring instead, and a high fat score with a low fibrosis score means something quite different from both being high.
Nor does it test the word “non-alcoholic”, which is assumed about you before the score is run. Alcohol makes livers fatty by its own route, so if intake was never asked about honestly, the label is wrong even when the number is right.
The inputs bring their own weaknesses. Insulin assays are not standardized, so the same blood can give different values in two laboratories and the score moves with them. In anyone taking insulin, or a drug that makes the pancreas secrete more, the level reflects the prescription, not the physiology. Transaminases drift with hard exercise, muscle injury, illness and common medications, so one draw in an unusual week can mislead.
Finally, note where it came from: derived in Finnish adults, with performance that varies across populations, and not established for children or in pregnancy. It was built to classify at one moment, not to track treatment, so a small change after months of a new diet is not a measure of fat lost.
Related reading
- Steatohepatosis
- Insulin resistance
- Metabolic syndrome
- How to Improve Metabolic Health: A Practical Guide
Disclaimer. This article is for information only and does not replace medical advice. Talk to a qualified clinician before changing anything about your health.