After weight loss, does the fat in the liver also decrease? An analysis announced on October 1 used MRI scans to answer this question. The results are worth attention, but first we need to be clear: who was scanned, and what the numbers actually represent.
Fat in the liver is different from fat around the waist
Fat is not only stored under the skin; it can also build up in the liver. Whether someone looks heavy or not does not directly tell us how much liver fat they have; and how much weight they lose cannot directly replace a liver examination.
This study used MRI to estimate the proportion of liver fat. What it provides is information about liver fat content, not the results of taking liver tissue samples and looking at inflammation or scarring.
55 people had MRI scans, and 26 people initially had more than 5%
This post-hoc analysis came from some participants in the STEP UP trial, a total of 55 adults with obesity but without diabetes. The study combined data from two doses of semaglutide and followed participants through week 72.[1]
Among them, 26 people started with liver fat above 5%. By week 72, 23 people had values that dropped below 5%. Therefore, the news phrase "nearly ninety percent returned to the normal range" refers to 23 divided by 26, not ninety percent out of 55 people, and not that every patient with fatty liver disease would respond this way.
From 8.8% to 3.1%, how much was actually reduced?
In the combined treatment group, the average liver fat proportion fell from 8.8% to 3.1%. Subtracting the two numbers gives 5.7 percentage points; then dividing 5.7 by 8.8 gives a relative reduction of about 64.8%.[1]
Both calculations can describe the same change, but they mean different things. It is like a temperature dropping from 30 degrees to 20 degrees: first you say it fell by 10 degrees; when talking about proportion, you then explain how large that 10 is relative to the starting point. Here, too, "reduced by about 65%" cannot be understood as losing 65% of total body fat.
Why can't we directly say fatty liver disease has been cured?
A reduction in fat is a valuable change, but liver inflammation, fibrosis, and long-term complications each still require their own examinations or follow-up. This analysis was also an additional question asked after the trial was completed, involved fewer people, and combined doses.
The announcement also described a larger combined population analysis, involving 1919 people and a liver risk score. The score data and the MRI results from 55 people come from different analyses, and cannot be pieced together to say that "all 1919 people had MRI scans."
What should researchers look at next?
More convincing follow-up studies should define liver changes as an observation goal in advance, specify the control group, dose, and starting disease severity, and track inflammation or fibrosis. Only then is it easier to judge how stable the fat changes are and which people benefit more.
For readers, the most useful reminder from this news is: when evaluating weight-loss research, besides the scale, you can also look at what happens to fat inside the body.
If liver fat is below 5%, does that mean it will never rise again?
A single examination falling into a lower range only shows the state at that measurement. Whether it stays there afterward also depends on later weight, metabolism, and treatment. To answer whether it rebounds, continuous follow-up is needed.
When you receive a similar report, it is best to keep the examination date, the method used, and your weight at that time together. Comparing the same method over time is usually easier to understand than putting single results from different methods side by side.
View materials and analysis information
For experiments related to GLP-1, you can view custom peptide services and check the required sequence, modifications, and analytical requirements. The current catalog does not have a separate semaglutide page, and this article does not provide links to products that do not exist.
Custom peptide services
Describe the required sequence, terminal form, quantity, and testing requirements, and confirm the delivery scope according to the project.
Original materials
Sources & editorial method
AI-assisted writing
1 linked records are listed in the references below. Read the editorial and AI-assistance policy.
How to interpret this article
Original public sources, study populations and limitations are linked. Editorial illustrations are not experimental photographs.
Research-use boundary: Catalog materials discussed on this website are for laboratory research, development and manufacturing use only, not for human or veterinary use. This content is not medical advice and does not provide administration instructions.
