Eating After Dinner Tracked With More Fatty Liver and Cirrhosis in Nearly 200,000 Americans

I’ve never once thought about my liver while standing at the counter at 10:30 at night with a spoon in the peanut butter jar. I thought about calories, and sometimes about guilt, but never about the time. Then I read an analysis of nearly 200,000 nurses and health professionals in which the people who regularly ate after dinner or after bedtime had a 20 to 30 percent higher rate of liver disease than the people who didn’t. Now I can’t stop thinking about the time.

The work, led by hepatologist Robert Wilechansky at Oregon Health & Science University with circadian physiologist Andrew McHill and colleagues at Massachusetts General Hospital, appeared this month in Clinical Gastroenterology and Hepatology. It pooled three of the longest-running health cohorts in the country, the Nurses’ Health Study, the Nurses’ Health Study II and the Health Professionals Follow-up Study, and found the higher rates for both metabolic dysfunction-associated steatotic liver disease (MASLD, long called non-alcoholic fatty liver disease) and cirrhosis, the scarring that can come at the far end of it. HealthDay’s account of the paper put the fatty-liver figure higher, at 36 percent, with cirrhosis at 28 percent over about 16 years of follow-up, and reported that the association held after adjusting for age, diabetes, physical activity, body mass index and an estimate of body fat.

That last adjustment caught me, because I went in expecting the boring answer: late eaters are heavier, heavier people get fatty liver, done. BMI is a blunt instrument, and putting it in a model doesn’t make weight disappear as an explanation. Still, the association didn’t collapse once weight was accounted for, so the boring answer doesn’t explain away the whole finding.

Nine o’clock isn’t the same hour for every body

“We know the liver is very much in tune with the body’s circadian rhythms,” Wilechansky said. McHill points out that animal research singles out the liver as one of the organs most vulnerable when eating and sleep fall out of sync, and he argues that eating “when the internal biological clock is promoting sleep” reduces the body’s ability to metabolize food efficiently.

He has been working toward this for years. In a 2017 study in the American Journal of Clinical Nutrition, his team followed 110 adults aged 18 to 22 and used the evening rise of melatonin to mark each person’s biological night. Clock time didn’t separate the leaner participants from those carrying more body fat, because both groups ate at about the same hours on the wall. Measured against their own internal clocks, though, the higher-body-fat group ate most of their calories about 1.1 hours closer to melatonin onset.

That result stopped me. If the hour that counts is set by your melatonin and not your microwave, 9 p.m. can be early evening for one person and deep night for the person next door. Which left me with the question I couldn’t put down: why would the body handle the same snack differently depending on an hour it sets for itself? I should be careful about how far that takes us, though. The liver cohort asked people a broad question about eating after dinner or bedtime, not when their melatonin rose, so circadian mismatch is a plausible explanation for the liver finding, not a proven one.

Mouse experiments give that question its sharpest answer so far. In a 2012 Salk Institute experiment, Satchidananda Panda’s lab fed mice the same fatty food and the same calories, letting one group graze around the clock and restricting the other to an 8-hour window. The round-the-clock eaters developed liver damage, high cholesterol and high blood glucose, while the time-restricted mice weighed 28 percent less. Panda was careful even then: “we will have to wait for human studies to prove this.”

In people, the most controlled test of the mechanism is a 2022 crossover study in Cell Metabolism by Nina Vujović, Frank Scheer and colleagues, in which adults with overweight or obesity ate the same calories on an early schedule and on a late one. On the late schedule they were hungrier while awake, their 24-hour leptin (the hormone that tells you you’ve had enough) dropped, they burned less energy during the day, and their fat tissue shifted its gene activity toward storing fat.

The trial where the clock added nothing

My enthusiasm ran into a wall at a hospital in Guangzhou. In the TREATY-FLD trial, published in JAMA Network Open in 2023, researchers randomized 88 adults with obesity and fatty liver disease to the same daily calorie targets, and half of them also had to eat only between 8 a.m. and 4 p.m. After 12 months, liver fat had fallen by 6.9 percentage points in the early-window group and 7.9 in the calorie-only group, a 1.0-point gap that wasn’t statistically significant.

That stung, and I’d rather sit with it than explain it away. The trial asked whether an early eating window adds anything once calories are already being cut, and it found no measurable extra benefit. It didn’t show that timing does exactly nothing, and it didn’t test the way most of us actually eat, without a dietitian handing us a calorie ceiling. Whether timing matters by itself for people eating freely is a separate question.

The cohort can’t settle that question either. The authors describe their own data as observational and based on broadly defined, self-reported habits, a survey answer that can’t tell a 9 p.m. apple from a 1 a.m. plate of leftover pizza. The people in it are nurses and health professionals, more health-literate and better connected to care than most Americans. For cirrhosis, the first thing any liver doctor asks about is alcohol, and a nightcap and a late snack tend to travel together, so how much of that signal rides on drinking is something the full paper has to answer. Sleep is tangled up in this too, because anyone eating after bedtime is awake when they meant to be asleep. Wilechansky himself points to night shift workers, who have elevated rates of obesity and heart disease, and the late snack may partly be a marker of a disrupted night. He also stresses that only a small fraction of people with fatty liver ever progress to cirrhosis, so a relative bump in a rare outcome is no reason to panic.

No list price on an early dinner

As much as 30 percent of the U.S. population is estimated to have fatty liver disease. The first drug approved for its advanced form, Madrigal’s Rezdiffra, won accelerated approval in 2024 on the strength of its effects on disease markers, aimed initially at patients with moderate to advanced fibrosis, and launched at $47,400 a year.

An earlier dinner hasn’t been shown to treat anything, least of all the scarring Rezdiffra targets, and I’m not pretending the two are interchangeable. But follow the incentives and you can guess why the timing question has been left to academic hepatologists and a liver-society foundation (this study was funded by the AASLD Foundation) instead of a company with a sales force. That’s my inference, not anyone’s admission, but it isn’t much of a leap: nobody books revenue when a patient puts the fork down at 8 p.m.

OHSU is now recruiting for a trial that will track meals directly instead of relying on memory, and will test whether eating earlier improves clinical measures of fatty liver disease. Wilechansky says the team wants to shift the timing of eating “without changing the composition of what people eat.” TREATY-FLD wasn’t built to run that experiment, which tests timing on its own, without a calorie cut riding along.

Until the trial reports, Wilechansky’s advice is modest: “confining your calorie intake to several hours before bedtime might be beneficial.” The “might” is carrying a lot of weight, and the randomized evidence so far says not to expect timing to beat simply eating less. But all it costs to try is a closed kitchen. Mine closes when the dinner dishes are done, and if I’m hungry at 10:30, I’m drinking a glass of water and going to bed.

Sources

1. Clinical Gastroenterology and Hepatology: Wilechansky et al., late-night eating, MASLD and cirrhosis (2026)
2. JAMA Network Open: Wei et al., the TREATY-FLD randomized clinical trial (2023)
3. ScienceDaily: summary of the OHSU-led study, with funding and limitations (2026)
4. News-Medical: OHSU release on late-night snacking and liver health (2026)
5. HealthDay via MedicineNet: 36% MASLD and 28% cirrhosis risk, 16-year follow-up (2026)
6. American Journal of Clinical Nutrition: McHill et al., later circadian timing of food intake and body fat (2017)
7. Harvard Medical School Division of Sleep Medicine: Vujović et al., late isocaloric eating in Cell Metabolism (2022)
8. Salk Institute: Panda lab, time-restricted feeding in mice (2012)
9. The Leader (AAP): Wilechansky and McHill on circadian rhythms and the liver (2026)
10. Sci.News: study details and researcher quotes (2026)
11. Medical Economics: clinical context and shift-work risk (2026)
12. BioPharma Dive: Rezdiffra priced at $47,400 a year after accelerated approval (2024)