Time-Restricted Eating and Exercise Did Not Cut Liver Fat in 187-Adult Trial

A four-arm randomized trial in Spain tested whether time-restricted eating, supervised exercise or the two together could lower liver fat in adults with obesity. After 12 weeks of MRI measurement, none of the three programs pulled away from usual care.

Time-Restricted Eating has become one of the most widely promoted ideas in metabolic health, and the newest randomized test of it landed on October 14, 2026 in Nature Communications. The TEMPUS trial, run by investigators based largely at the University of Granada in Spain, randomized 187 adults with obesity to usual care, an eight-hour eating window, a supervised exercise program, or both, then measured liver fat by MRI after 12 weeks.

The primary outcome was the intrahepatic fat fraction, the share of liver tissue made up of fat. Compared with usual care, the time-restricted eating group changed by minus 0.4 percent (95% CI minus 2.0 to 1.2; P = 0.93). The exercise group changed by minus 0.8 percent (95% CI minus 2.4 to 0.8; P = 0.54). The combined group changed by minus 1.1 percent (95% CI minus 2.7 to 0.5; P = 0.28).

None of those differences reached statistical significance, and the intervention groups did not separate from one another either. The authors write that the findings do not support greater short-term reductions in liver fat with time-restricted eating, exercise, or their combination than with usual care.

The design makes that result harder to wave away. Recruitment ran from May 2023 to April 2024, the trial completed on September 5, 2025, and the analysis followed intention to treat. Primary outcome assessors and statisticians were blinded to allocation, although the staff delivering the programs were not.

What the Time-Restricted Eating Group Actually Did

The published trial protocol describes the intervention in unusual detail. Participants assigned to time-restricted eating chose their own eight-hour window before the program began and held to it for the full 12 weeks, finishing the last meal by 21:00 with a tolerance of 30 minutes either side. The window applied on all seven days of the week.

That detail matters more than it looks. Self-selected windows are easier to live with than imposed ones, so the trial was testing time-restricted eating under relatively favorable conditions rather than forcing a schedule participants would resent and quietly abandon.

The Exercise Arm Was Not a Token Program

Participants in the exercise groups attended two supervised sessions a week lasting 60 to 90 minutes. Each session paired circuit-based resistance training for the major upper and lower body muscle groups with high-intensity interval work, preferably on a treadmill, structured as three to four sets of four-minute intervals separated by four minutes of active recovery.

Resistance work targeted a rating of perceived exertion above 7 and climbed toward 10 across the program. Session length grew from 55 minutes in the familiarization week to 85 minutes in the third phase, and participants also carried a walking goal that rose by 10 percent of daily steps each week.

This is a real training stimulus, not a token walking prescription. It is the kind of program that reliably improves fitness markers in other trials, which is part of what makes the liver fat result worth sitting with. Readers tracking that broader literature can compare it with our report on how combined cardio and lifting ranked in a metabolic syndrome network meta-analysis.

Why the Usual Care Group Matters Here

Usual care was not an empty condition. Every group in the trial, the control arm included, attended monthly 90-minute sessions built around Mediterranean dietary patterns and the World Health Organization physical activity recommendations. Control participants simply kept their habitual eating schedule instead of compressing it.

That raises the bar a null result has to clear. A comparison between time-restricted eating and an unsupported control group is a different experiment from a comparison between time-restricted eating and a group already receiving structured nutrition and activity education every month for three months.

Reading a Null Result Without Overreading It

A confidence interval that crosses zero is not proof that nothing happened. For time-restricted eating the interval ran from minus 2.0 to 1.2 percentage points of liver fat, which means the data are compatible with a modest reduction and also with a modest increase. The honest summary is uncertainty, not refutation.

What the trial does make unlikely, within 12 weeks and in this population, is a large and dependable drop in liver fat attributable to an eating window by itself. Trials that found bigger effects have often compared time-restricted eating against less active control conditions, or measured fat by less precise methods than MRI.

Where Time-Restricted Eating Still Has Support

Compressing an eating window remains a practical way for some people to cut total intake without counting anything, and simplicity is the usual argument in its favor rather than a unique metabolic mechanism. Meal timing also interacts with sleep and circadian rhythm, a thread we followed in our coverage of how irregular meal timing was linked to higher heart risk in men.

Liver fat specifically has responded to other dietary approaches in short trials, including the one described in our report on a ketogenic diet that cut liver fat by 67 percent in a small study. Those results do not transfer automatically, since they tested different diets, different sample sizes and different measurement tools.

The narrow reading of the new data is the defensible one. In adults with obesity, over 12 weeks, measured by MRI, time-restricted eating did not shrink liver fat more than education-based usual care did, and adding supervised exercise to time-restricted eating did not change that.

What This Means for a Weekly Routine

Nothing in this trial argues against exercising. It argues that one specific outcome, liver fat measured over three months, did not move further in the trained groups than in a control group receiving regular dietary counseling. Training still earns its place through fitness, strength, blood pressure and the long list of outcomes measured in other work, such as the links described in our report on cardio and strength training and lower diabetes risk.

For readers weighing an eating window, a few general points follow from the evidence rather than from marketing claims:

• Treat time-restricted eating as a possible aid to eating less overall, not as a liver treatment in its own right.
• Pick a window you can hold on ordinary days, since the trial used self-selected windows rather than imposed ones.
• Keep training on its own merits, because the exercise arm here was substantial and still did not change this particular outcome.
• Anyone with diagnosed liver disease, diabetes or a medication schedule tied to meals should raise meal-timing changes with a clinician first.

This is general information rather than individualized advice. Twelve weeks is also a short window by the standards of liver health, and a program that fails to separate from usual care over three months is not automatically the same as a program that fails over a year.