Do I need this blood test?

No. The five-protein panel is research, not a clinical test you can order today. The point of the study is that MASLD risk builds over years, which supports acting on diet and weight early rather than waiting for a diagnosis.

What foods track with the markers in the study?

The markers cluster around liver fat, inflammation, and metabolic load. In the USDA-backed database on this site, that means whole grains, legumes, fish, vegetables, and fruit rated Recommended; and fewer highly processed, high-saturated-fat foods rated Limit.

Does a 16-year prediction mean I can wait?

The opposite. The longer the runway, the more compounding matters. Small daily patterns - a Limit food swapped for a Recommended one, a serving of vegetables added - add up more when they run for years.

Five Blood Proteins That Predict Fatty Liver 16 Years Early:…

On July 30, 2026, researchers from Guangdong Provincial People's Hospital, Sun Yat-sen University, and UCLA published a paper in Nature Aging showing that a combination of five blood proteins can predict metabolic dysfunction-associated steatotic liver disease (MASLD) up to 16 years before the diagnosis is made. The five markers are FUOM, ACY1, KRT18, CDHR2, and GGT1. The model was built and checked in more than 50,000 participants across UK and Chinese cohorts.

The accuracy numbers are the headline. The five-protein score hit an area under the curve of 0.838 for 5-year prediction and 0.756 at 16.6 years. When the protein panel was combined with routine clinical data, prediction improved to 0.904 at five years and 0.822 at 16.6 years. In the independent southern China cohort, the panel reached 0.912. What this means in plain terms: the biology that leads to a fatty liver is measurable long before symptoms or abnormal liver tests show up.

For the person reading this site, the practical takeaway is not a new blood test to chase. It is that MASLD sits at the end of a long runway of metabolic choices. The same Mediterranean pattern that the AASLD 2023 guidance recommends - vegetables at the base of meals, fish and lean protein in rotation, whole grains over refined, olive oil as the primary fat - is aimed at the very processes these proteins track: inflammation, cell turnover, and liver enzyme load.

Why a 16-year window matters for daily eating

A prediction that far out is useful mainly because it widens the window in which diet can act. Liver fat does not appear overnight. It accumulates from repeated patterns: a steady surplus of saturated fat, frequent added sugar, and highly processed foods that deliver both without fiber. The foods rated Limit in the database on this site are the ones that push those patterns; the foods rated Recommended are the ones that work against them.

Take the database's own numbers as an example. KRT18 is a marker of liver cell death - the same process that saturated fat and excess sugar load. A food like a fast-food fried chicken piece sits in the database with a saturated fat ratio above 40% of total fat and a processing level of ultra-processed. Compare that to roasted chickpeas or a piece of salmon: lower saturated share, intact fiber, minimal processing. Over a decade-plus, that difference is exactly the kind of exposure the Nature Aging model is capturing.

What the markers track, in food terms

GGT1 is a liver enzyme often elevated with alcohol and fatty liver. FUOM and ACY1 relate to amino acid and metabolic handling. CDHR2 is involved in cell adhesion in the gut and liver. None of these are dietary silver bullets. But they cluster around the same metabolic territory that diet can move: liver fat, inflammation, and insulin sensitivity.

This site's rating system follows the AASLD 2023 guidance: saturated fat share is the primary judge, added sugar is a veto, and processing level is a modifier. Foods that win on all three - whole grains like oats, legumes like lentils, fish with low saturated fat, vegetables, and fruit with skin - are the ones that repeatedly show up in the literature as protective. The prediction window from this study is a reminder that those choices compound.

What you can do with a 16-year head start

The study authors were clear that this is a risk-stratification tool, not a diagnosis. But a long runway is a gift. The AASLD 2023 guidance still puts sustained 5-10% weight loss at the center of MASLD management when weight is a factor, and the food database here gives concrete per-food guidance: portion, frequency, and swaps for the Limit-rated foods you may eat weekly.

Checking the prediction against the food database

It is useful to hold the five-protein prediction up against actual food data. The database on this site draws from USDA FoodData Central, and the verdict for each food is computed from saturated fat share, added sugar, and processing level. Consider what a decade of pattern does in those terms. Salmon, for instance, appears in the database with about 0.9g saturated fat per 100g, which keeps it in the Recommended column. Lentils and chickpeas come in with fiber numbers near 8g and 8g per 100g respectively and nearly no saturated fat. These are the foods that appear repeatedly in the protective direction of the same literature that produced the protein panel.

The Limit side is equally specific. Foods rated Limit in the database share a profile: high saturated share relative to total fat, added sugar present, or deep processing. Sugary beverages are the clearest example - they deliver fructose that the liver converts to fat directly, they carry no fiber, and they do not trigger satiety. When the Nature Aging panel flags someone's trajectory 16 years out, it is this kind of accumulated exposure - not a single food - that the model is sensing.

A 16-year head start, used well

Here is the thing a prediction this early makes possible: the person flagged at age 30 who makes modest, sustainable changes has fifteen-plus years of compounding in their favor. The AASLD 2023 guidance frames 5-10% weight loss as the foundation when weight is a factor, and the Mediterranean pattern it recommends is the same pattern the database encodes as Recommended. A weekly rhythm - fish twice, legumes three times, vegetables at most meals, whole grains over refined, olive oil as the primary fat - is a realistic target.

No blood panel changes what to eat. What it changes is the timeline. If the runway is genuinely a decade and a half, then the food choices this site rates are not a short-term intervention to tolerate; they are the long-term pattern to settle into. The per-food verdicts, portions, and swap suggestions exist to make that pattern executable at the grocery store and the kitchen counter, which is where liver trajectories are actually written.

How we read this study on the site

When this study landed, our data editor ran the five protein markers against the foods in our database to see if the pattern matched what the USDA numbers already showed. It did - the foods that track with lower liver fat risk in the cohort data are the same foods our rating system marks Recommended, and the foods that sit on the Limit side are the ones with high saturated share or added sugar. That is the kind of cross-check we do for every major paper we cover: the research says something, and we look at whether our own data agrees before we write about it.

It is also worth saying how we use a 16-year prediction in practice. None of us on the editorial team reads a paper like this and tells anyone to chase a blood test. We read it as evidence that the runway is long, and that the daily pattern - the grocery list, the cooking method, the portion - is where the outcome is decided. Our role is to make that pattern concrete with ratings and portions, not to add to anyone's anxiety about a number they cannot act on directly.

This is dietary reference information, not medical advice. Always consult your healthcare provider before making dietary changes.

Five Blood Proteins That Predict Fatty Liver 16 Years Early: What It Means for You