A meal sent 94 of 312 healthy volunteers past a local warning threshold for possible liver scarring, researchers in Novi Sad report in a study published online on 24 October. Before eating, just eight of the same volunteers crossed it. The authors say the gap could prompt unnecessary follow-up tests if clinics scan patients too soon after food.
The study examined transient elastography, an ultrasound-based test that estimates liver stiffness. Participants were scanned after fasting and again 45 minutes after a standard meal; the post-meal readings rose by an average of 2.1 kilopascals. The authors stress that a high reading is not a diagnosis of liver disease, but can lead a doctor to order further checks.
The contest that caught lunch
The clue emerged at a diagnostic competition in Novi Sad in March 2025, where 12 teams from Serbia, Bosnia and Montenegro assessed anonymised liver scans. Teams comparing their results noticed that afternoon cases were more likely to trigger concern. Nadežda N. Holubová, the study’s lead author at the fictional Danube Clinical Centre, said the timing made her look beyond the scan itself.
“We went back to the timestamps and saw that meals were the one pattern nobody had thought to score,” Holubová said. Her group then recruited 312 adults with normal liver blood tests and no known liver condition. Each had the same type of scan before eating and 45 minutes after a standard meal; the researchers say the paired design let them compare each person with themselves.
The post-meal change does not mean that lunch creates liver damage, the researchers say. Food temporarily changes blood flow through the liver, which can affect stiffness readings. In the study, 94 volunteers crossed the centre’s 8.0-kilopascal referral threshold after eating, compared with eight before the meal.
A normal lunch should not turn a healthy liver into a referral.
Nadežda N. Holubová, study lead author
Clinics rewrite the waiting-room rule
The result has practical consequences for clinics that use the test to decide who needs specialist assessment. The study team recommends waiting at least three hours after food before scanning, in line with the centre’s revised appointment guidance. It says the test remains useful, but clinicians should record when a patient last ate before interpreting a borderline result.
István Simon, a radiologist who took part in the competition, said the change could spare patients weeks of worry and avoid repeat appointments. “The machine did not fail; we were asking it a question at the wrong time,” he said. He added that clinics should not cancel or delay care for patients with urgent symptoms based on the study alone.
The researchers caution that their volunteers were not a representative sample of the wider population, and the findings do not show how often meal-related readings lead to unnecessary treatment. They plan to test whether different meals and longer waiting periods produce the same effect. For now, Holubová said, the simplest safeguard is to ask when the patient last ate.
