The Hepatic Steatosis Index, usually shortened to HSI, estimates how likely it is that fat has built up in your liver, using two enzymes from a routine blood panel together with your body mass index, your sex and whether you have diabetes. It is a lab-based screen, not a home measurement: the enzyme values need a blood draw. No scanner and no specialist visit are required, but a laboratory is.
It was published in 2010 from a large Korean health-checkup cohort, with abdominal ultrasound as the reference. Its job is triage: deciding who probably needs liver imaging and who does not.
What you need
- ALT, alanine aminotransferase, in U/L. It sits on any standard liver panel.
- AST, aspartate aminotransferase, in U/L, from the same panel and the same draw.
- Your height and weight, so body mass index can be calculated; some versions ask for BMI directly.
- Your sex, which contributes a fixed amount to the score.
- Whether you have been diagnosed with diabetes, which contributes another fixed amount.
Nothing else is used. Liver enzymes do not require fasting, though they are usually drawn with a fasting panel.
How the measurement works
ALT and AST leak out of liver cells that are stressed or damaged. ALT is the more liver-specific of the two; AST also comes from muscle, heart and red blood cells. In fat-driven injury ALT usually rises more than AST, so their ratio runs above one, while in advanced scarring the pattern often flips. HSI uses that ratio, not either enzyme alone.
The score multiplies the ALT-to-AST ratio by a fixed coefficient, adds body mass index, then adds a set number of points for female sex and another for diabetes. Every input pushes the same way: higher ALT relative to AST raises the score, a higher BMI raises it, and each yes-or-no item adds its fixed step when it applies. Because BMI enters directly, body size dominates when enzymes are near normal. Rather than doing the arithmetic by hand, enter your five values into the Hepatic Steatosis Index (HSI) and let it apply the published coefficients.
How to read the result
The original study set two thresholds, which split the range into three zones.
| HSI value | Standard reading |
|---|---|
| Below 30 | Fatty liver effectively ruled out. This cut-off was set for high sensitivity, so few people with real steatosis fall below it. |
| 30 to 36 | The gray zone. The score cannot decide; imaging is the usual next step. |
| Above 36 | Fatty liver likely. This cut-off was set for high specificity, so a value here is rarely a false alarm. |
These are probabilities, not a diagnosis, and they come from a Korean population screened by ultrasound. The gray zone is wide, and many adults land in it. If your panel includes GGT and triglycerides and you have a waist measurement, the Fatty Liver Index (FLI) answers the same question from different inputs; agreement between the two says more than either alone.
What it does not tell you
HSI does not measure liver fat. It estimates the chance an ultrasound would find it; quantifying how much is there takes imaging, and MRI-based fat fraction is the accurate version.
More importantly, it says nothing about fibrosis, the scarring that drives liver outcomes. Someone can have a high HSI with an unscarred liver, or advanced fibrosis with a modest score. That question needs a separate tool such as the FIB-4 Fibrosis Index, and often elastography. Nor does it tell you why fat accumulated: it never asks about alcohol, so it cannot separate metabolic fatty liver from alcohol-related disease, viral hepatitis or drug injury.
Its inputs are easy to disturb. Hard exercise, muscle injury, a recently started statin, hemolysis in the sample or acute hepatitis will move the enzymes for reasons unrelated to fat, and because they enter as a ratio, a low AST inflates the score by itself. Normal ALT does not clear you either: a substantial share of people with confirmed steatosis have transaminases inside the reference range. And because BMI carries so much weight, a very muscular person can be pushed upward on build alone, while lean fatty liver, common in South and East Asian populations, is what the score most often misses. It is not established for children, adolescents or pregnancy.
Related reading
Disclaimer. This article is for information only and does not replace medical advice. Talk to a qualified clinician before changing anything about your health.